Participant Presentations And Session Transcript For June 22
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
1 | C O N T E N T S |
2 | |
3 | MORNING SESSION (SELLING): |
4 | Introduction |
5 | Presentations: |
6 | Kenneth Elzinga |
7 | Janusz Ordover |
8 | Patrick Bolton |
9 | A. Douglas Melamed |
10 | Moderated Discussion |
11 | Lunch Recess |
12 | |
13 | AFTERNOON SESSION (BUYING): |
14 | Introduction |
15 | Presentations: |
16 | Tim Brennan |
17 | John Kirkwood |
18 | Janet L. McDavid |
19 | Steven C. Salop |
20 | Frederick R. Warren-Boulton |
21 | Moderated Discussion |
22 | |
23 | |
24 | |
25 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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, |
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. |
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, |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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. |
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 |
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 |
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 |
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 |
1 | de-emphasize the cost test, we want to add as a safe |
2 | harbor the -- allow for an efficiency defense. |
3 | So, how do we prove those elements? Well, some |
4 | of them are straightforward, and I will not go into -ï½ |
5 | so, facilitating market structure is any evidence of |
6 | market power. The scheme of predation and supporting |
7 | evidence, I want to give you an example of how you go |
8 | about doing this. So, I will in particular take out of |
9 | our article the example we have on financial market |
10 | predation, and so under this element, what is important |
11 | is to establish that the conditions to implement a given |
12 | strategy are present and to provide direct or |
13 | circumstantial evidence showing that this strategy is |
14 | being implemented. |
15 | Recoupment, again, this is relatively |
16 | straightforward. You would want to show evidence of |
17 | exclusion and disciplining of rivals, and we stressed |
18 | the idea that second, that you should emphasize probable |
19 | recoupment instead of actual recoupment, because what |
20 | matters is whether at the time when this strategy was |
21 | being chosen, whether at that time, at the time of the |
22 | information the incumbent had at that time, whether it |
23 | made sense to implement such a strategy, and we know, as |
24 | in our own investments in finance, we know that at the |
25 | time when we make a decision of investment, we make an |
1 | analysis using this kind of cash flows that suggests |
2 | that we have a positive net present value investment, |
3 | but that is no guarantee that when we actually undertake |
4 | the investment, it will end up being profitable. So, we |
5 | would emphasize probable recoupment, and in particular, |
6 | put a lot of weight on market structure that makes |
7 | recoupment likely in the future. |
8 | Let me also emphasize here the "or related" in |
9 | brackets, and this is a point that Professor Ordover |
10 | emphasized, that recoupment shouldn't just be seen in |
11 | the narrow market where predation takes place. It could |
12 | be obtained in a related -- I forget the term you |
13 | used -ï½ |
14 | DR. ORDOVER: Noncoincident. |
15 | DR. BOLTON: -- noncoincident market. |
16 | On price below cost, I do not have much to add |
17 | to what I have said already except that in the paper we |
18 | emphasize that a better measure than average variable |
19 | cost would be average avoidable cost, and a better |
20 | measure for long-run average cost would be long-run |
21 | average incremental cost. I do not want to go further |
22 | into this, because making fine distinctions about these |
23 | definitions could end up being a mind-numbing exercise, |
24 | and it just highlights the difficulty with applying the |
25 | cost test. |
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, |
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 |
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 |
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. |
1 | (Applause.) |
2 | MR. POTTER: Our final speaker today is also the |
3 | only lawyer on the panel, although Doug is very used to |
4 | dealing with economists, so I am sure it will not be a |
5 | problem for him to follow them. |
6 | Doug is a partner at Wilmer Hale, and he is the |
7 | co-chair of the firm's Antitrust and Competition |
8 | Department. He has significant experience in a number |
9 | of government investigations, both government and |
10 | private litigation, substantial antitrust counseling, |
11 | and some of that counseling in investigatory work, in |
12 | litigation work, has involved predatory pricing. |
13 | From 1996 to 2001, Mr. Melamed served as the |
14 | Principal Deputy Assistant Attorney General in the |
15 | Antitrust Division and then as the Acting Assistant |
16 | Attorney General in the Antitrust Division. He's a |
17 | prolific writer, a frequent speaker, always has |
18 | interesting viewpoints that are well thought out. His |
19 | most recent -ï½ |
20 | MR. MELAMED: Don't raise the bar, please. |
21 | MR. POTTER: -- his most recent article, which |
22 | appears in the summer 2006 Antitrust Law Journal |
23 | provides a thought-provoking commentary on whether there |
24 | are unifying principles under Section 2. |
25 | Mr. Melamed, welcome. |
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 |
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 |
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 |
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 |
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, |
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 |
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 |
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 |
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. |
1 | So, proposition four, in applying the Brooke |
2 | Group price-cost test, the price of the product at issue |
3 | is the appropriate price to compare to cost. That in my |
4 | view is only partially correct. Obviously you look at |
5 | the price of or the revenues generated by the additional |
6 | sales attributable to the predatory conduct. You don't |
7 | look at the price of, of course, the inframarginal |
8 | units, the units that would have been sold anyhow, |
9 | because those units didn't exclude the rival or at least |
10 | they didn't exclude them by reason of the |
11 | anticompetitive conduct. |
12 | But that's not all there is to it. Suppose |
13 | we're in a two-sided market. Suppose you're cutting |
14 | price on circulation of the newspaper in order to |
15 | generate more readers and therefore more advertising |
16 | revenues. Surely you want to take into account the |
17 | incremental advertising revenues. |
18 | Suppose you have complimentary revenues. You |
19 | know, the Government didn't accuse Microsoft of |
20 | predatory pricing because the browser was free when |
21 | bundled with the operating system. Because it was a |
22 | plausible story that it increased revenues, we didn't -ï½ |
23 | increased revenues for the operating system. What about |
24 | revenues lost from inframarginal sales; that is to say, |
25 | the sales that the defendant would have made anyhow even |
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, |
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 |
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 |
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. |
1 | I think the position of McGee and the Chicago |
2 | School generally is that predatory pricing can be a |
3 | rational business strategy, it's just it's a very |
4 | unusual one, defined where it's successful, where it |
5 | works. |
6 | DR. ORDOVER: Well, I certainly agree with the |
7 | statement, with a couple of -- I don't know how many |
8 | caveats, but first -- five caveats -- one, two, three -ï½ |
9 | the first caveat is we have got to define what predation |
10 | means. Second, we have to figure out what the price is. |
11 | Third, we have got to figure out how to engage |
12 | rationality. Other than that, I think it's all fine. |
13 | Other than that, how was the performance, right? So, |
14 | this is exactly the way I see it. |
15 | I mean, this is surely a statement that has a |
16 | meaning as long as we can agree on the meaning of the |
17 | terms or words that go into the statement. None of |
18 | these things are relatively or clearly defined. We |
19 | already have different standards for predation. In the |
20 | airline case that Ken and I are in, pricing to whom is |
21 | an issue that -- average price on the aircraft? Is it |
22 | the price to the business passengers, the leisure |
23 | passengers? A huge amount of disagreement. Is it the |
24 | price of the incremental unit? Is the price averaged |
25 | out over the volume that is being sold? |
1 | Anyway, what's rational? I guess |
2 | profit-maximizing, over what horizon, what discount |
3 | rates we are going to use? All of these things enter |
4 | into what we have been struggling with, which is to say |
5 | that we have something -- we understand a basic core set |
6 | of issues, but I think these remaining areas of |
7 | disagreement are really needed to breathe light -ï½ |
8 | MR. POTTER: And with the later slides, we will |
9 | get into those specific areas. |
10 | DR. ORDOVER: I haven't looked at them. Ex ante |
11 | assessment, huh? |
12 | MR. MELAMED: Just a comment provoked by what |
13 | Janusz said, there is always -- at least in my |
14 | experience in cases I have dealt with, I am not involved |
15 | in Spirit -- difficult questions about what are the |
16 | products you're talking about, what prices are you |
17 | talking about, is it the leisure passengers or whoever |
18 | it may be. There is, if there is discipline in the |
19 | overall case, however, some discipline on the parties on |
20 | that issue, if the plaintiff wants to argue that the |
21 | price is predatory because he found one passenger in |
22 | seat 14B where the price was below cost, he is probably |
23 | not going to be able to prove that he was driven out of |
24 | the market on account of the predatory price, and so if |
25 | the courts are rigorous in connecting the allegedly |
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, |
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 |
1 | we mean pricing below cost is rarely tried and even more |
2 | rarely successful, it's rationally then used in |
3 | Matsushita, but it doesn't support Brooke Group, because |
4 | that would be like saying killing with an ice pick is |
5 | very rare, so let's define murder as consisting solely |
6 | of killing with an ice pick. |
7 | The question, if you want to justify or explore |
8 | the wisdom of defining predatory pricing as pricing |
9 | below cost, the question is, what about the conduct that |
10 | isn't deemed to be predatory pricing by that definition, |
11 | some kind of profit sacrifice at above cost levels, is |
12 | that rarely tried and rarely successful? And I'm just |
13 | not sure that there has been rigor in thinking about |
14 | what this statement means. |
15 | MR. POTTER: Building off of this slide, does |
16 | anybody have a view on whether predatory pricing is more |
17 | or less likely in certain industries because of the |
18 | characteristics of those industries? |
19 | DR. ELZINGA: Yes, I certainly do. I have a |
20 | belief that predatory pricing is more likely to occur |
21 | where the target firm will exit quickly and be unlike -ï½ |
22 | either the target firm or other capacity will be |
23 | unlikely to enter again, and just picking up on |
24 | something that Doug said, where you are trying to look |
25 | for some more simple benchmark, he suggested just |
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 |
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 |
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 |
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 |
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 |
1 | down. Everybody would be left buying a full revenue |
2 | cigarette if they were a smoker. |
3 | The discount segment continues to grow. It's |
4 | about 40 percent of the industry now. So, if all of |
5 | these cases had been decided differently using game |
6 | theoretic approach or a concern that Patrick expresses, |
7 | I think consumers would be worse off. I really do. |
8 | MR. POTTER: Janusz? |
9 | DR. ORDOVER: One comment. I think that there |
10 | is an issue that we may want to talk about a little bit |
11 | more, and that is to say, the rigor and the reviews of |
12 | the galaxy of predation models that are based on really |
13 | state-of-the-art game theory, and the question, what |
14 | follows from those in terms of public policy? To me, |
15 | that is the biggest problem that I have been totally |
16 | incapable of resolving in my own head, but in the end, |
17 | coming down on the proposition that while we cannot be |
18 | as perhaps lackadaisical about anticompetitive |
19 | exclusionary behavior as the Court in this famous quote |
20 | was, we still need to take some kind of tools that the |
21 | courts can use to say, yes, yes, I agree, things can |
22 | happen, and Milgrom and Roberts and Kreps-Wilson, they |
23 | all have shown all those things, and many others follow |
24 | and, you know, your lovely paper with Scharfstein on -ï½ |
25 | what, signal jamming or -- it was, signal jamming paper, |
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 |
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 |
1 | today, you have a slide later on, is it hard to counsel |
2 | your client? No, I say not to worry about it, because |
3 | the -- but actually -- actually -ï½ |
4 | DR. ORDOVER: Can we go home now? |
5 | MR. MELAMED: But actually, I say more than |
6 | that. First I say you may lose the characterization |
7 | issue, you may not be able to prove predatory pricing, |
8 | but then I say, "Wait a minute, there are certain |
9 | settings in which you could get hurt. Is your target |
10 | likely not only to withdraw from this market but, for |
11 | example, to go out of business and become bankrupt and |
12 | his only asset may be a lawsuit? How litigious is he? |
13 | Is this a part of some broader commercial strategy?" |
14 | So, there are situations I think even with the |
15 | law today totally in favor of the weight of false |
16 | positives where it probably does deter some |
17 | procompetitive pricing. Whether on balance at this rule |
18 | or at some other rule that harm is greater than the harm |
19 | of false negatives I'll leave to the economists. |
20 | MR. POTTER: All right. |
21 | Next one, establishing a reasonable prospect of |
22 | recoupment should be essential in any analysis of |
23 | predatory pricing. Is there anyone who disagrees with |
24 | this statement? |
25 | MR. MELAMED: Only to the extent I already said |
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 |
1 | visible signal for the rest of the players that may be |
2 | sitting out on the outskirts and waiting what to do." |
3 | Or it could be that the firm which you are |
4 | trying to induce to exit or to restrain its expansion is |
5 | what I called in the first slide a scarce competitor, |
6 | and, in fact, there is something very special, very |
7 | particular about that rival which cannot be replicated, |
8 | and in that case, yes, you get to the point in which the |
9 | assessment of this later recoupment or the implications |
10 | of this strategy is critical, and if you cannot show -ï½ |
11 | you, the plaintiff -- that if you exit the marketplace |
12 | or if you get cut back in the marketplace, economic |
13 | welfare is going to be hurt in some way, then I think |
14 | you have gone very far in challenging the conduct at |
15 | issue. |
16 | MR. POTTER: Next slide -- oh, I'm sorry. |
17 | DR. ELZINGA: I was just going to say, I think I |
18 | am saying just the same thing that Janusz said but |
19 | perhaps in just a couple words. I do not think you need |
20 | to do a recoupment analysis for many predation |
21 | allegations, because entry conditions or prices and |
22 | costs will tell you you needn't take that extra step. |
23 | DR. BOLTON: Can I just add -ï½ |
24 | MR. POTTER: Sure, go ahead. |
25 | DR. BOLTON: -- one comment? So, I agree with |
1 | Janusz that in principle, recoupment is important, it is |
2 | the right question to ask, but in terms of how do you |
3 | administer a recoupment test, I think the weight has to |
4 | be on what you call the reasonable prospect, and I think |
5 | a narrow reading of a recoupment test, as you criticized |
6 | earlier, I would criticize as well. |
7 | MR. POTTER: Okay, fair enough. |
8 | Next slide. Prices above some measure of cost, |
9 | and you can all pick your own measure of cost that you |
10 | think is the best cost, whatever it is, should not be |
11 | considered predatory. Is there anyone who disagrees |
12 | with this? |
13 | Patrick, do you want to say anything? |
14 | DR. BOLTON: So, from the -- well, we know that |
15 | a policy of -- after Brooke Group is that a price -- at |
16 | least a price above average total cost should not be |
17 | considered as predatory. I am happy to live with that, |
18 | although I am not sure that it is always a wise policy. |
19 | MR. POTTER: When it is -ï½ |
20 | DR. BOLTON: I would disagree, though, with the |
21 | statement that prices above average variable cost should |
22 | not be considered as predatory. |
23 | MR. POTTER: You just mentioned that you might |
24 | disagree in certain instances that even prices above |
25 | average total cost should not be predatory. |
1 | DR. BOLTON: Could be predatory. In principle, |
2 | in theory, there are situations where prices above -ï½ |
3 | even price above total cost can be predatory. |
4 | MR. POTTER: Can you give an example of those? |
5 | DR. BOLTON: Well, an example of a large |
6 | incumbent with increasing returns, scaled technology, |
7 | facing a small entrant that has not been able to reach |
8 | minimum cost capacity, you could exclude that entrant by |
9 | pricing lower than monopoly price but still above your |
10 | average total cost and exclude the entrant. |
11 | DR. ORDOVER: Maybe I could just ask you a |
12 | question. Would you comment on the cost principles -- I |
13 | have been puzzled by them myself -- that follow from |
14 | these various game theory like models of, say, |
15 | Kreps-Wilson? They do not seem to give clear cost |
16 | benchmarks. Is that true? Is that your reading as |
17 | well? |
18 | DR. BOLTON: Yes, that is correct. They do not |
19 | give a clear reading on cost benchmarks, and I think |
20 | there is a whole group of economists who have been |
21 | working on predatory pricing who think that costs are a |
22 | very poor way of discriminating between anticompetitive |
23 | effects and procompetitive effects, that there are as |
24 | likely to be false positives as there are to be false |
25 | negatives. There are many situations where pricing is |
1 | below even average variable cost, and it is efficient. |
2 | It is not predatory. So, a lot of economists feel it is |
3 | just a poor test. |
4 | MR. POTTER: Doug? |
5 | MR. MELAMED: Let me ask Patrick this question. |
6 | I understand the theory, even if I cannot understand the |
7 | game theory, of why an above cost, even above total |
8 | cost, but below profit -- monopoly profit-maximizing |
9 | test could be predatory in the sense that it could |
10 | exclude a rival and in the long run we are all going to |
11 | be worse off for it. |
12 | What I don't understand and I am interested in |
13 | your reaction to is how one turns that into a legal rule |
14 | that companies can comply with. I mean, how do you -ï½ |
15 | you know, if -- sure, if you imagine -- if you posit a |
16 | stable market on day one and then the entrant comes and |
17 | maybe you have a good historic benchmark and you can |
18 | say, "Gee, he's changed his pricing," but even then you |
19 | have to ask the question, "Well, what would the monopoly |
20 | profit maximizing price be with the new entrant?" Is |
21 | each company supposed to hire a game theorist and work |
22 | out the game and figure out -ï½ |
23 | DR. ORDOVER: Yes. |
24 | MR. MELAMED: -- what the price is? In other |
25 | words, how do we implement that test? |
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 |
1 | economic rationale rather than intent. |
2 | DR. ELZINGA: And in like fashion, Doug, I |
3 | suspect you have encountered clients who just aren't |
4 | aware that when they write things, they have to be |
5 | written with an eye towards antitrust enforcement, and |
6 | so you do find documents coming -- to have militaristic |
7 | or powerful metaphors that have nothing to do with |
8 | consumer welfare and may, in fact, represent exaggerated |
9 | views of the company's prowess and stature in the |
10 | marketplace. |
11 | MR. MELAMED: I find almost invariably they are |
12 | written by lunatic middle managers, but -ï½ |
13 | MR. POTTER: That wasn't your position when you |
14 | were Deputy Assistant Attorney General. |
15 | MR. MELAMED: Well, you know, you learn. |
16 | DR. ORDOVER: Or those documents are usually by |
17 | investment bankers who are pedaling a particular deal or |
18 | something like that, alleging that as a result of action |
19 | X or Y, the firm would be able to leverage its market |
20 | power from one market to another. If one were to take |
21 | these arguments -- take these documents seriously, that |
22 | would be the end of most of the Chicago Business School, |
23 | presumably, investment banking, but also, the ability of |
24 | business people to compete in the marketplace, because |
25 | this is what these guys are selling. So, you have to |
1 | read their signal, which is the investment advice or |
2 | business advice which they proffer, as being an attempt |
3 | to market the product at above competitive price to the, |
4 | you know, willing or unwilling buyers, and I think that |
5 | that is why I am very worried about reading all kinds -ï½ |
6 | I mean, I have seen documents probably as good or as bad |
7 | as Doug or anyone, and I try to discount their value |
8 | because they are frequently misleading. |
9 | Now, this is not to say that people who run the |
10 | companies do not have an insight into the marketplaces |
11 | in which they are competing, but I think there are |
12 | limits of the kind of inferences you can really make |
13 | from those types of documents, especially when they are |
14 | also written by third parties with a very special agenda |
15 | of their own in my view. |
16 | MR. POTTER: Next slide, as long as we're on |
17 | costs, let's throw this out, a variety of cost tests. |
18 | The proposition for the panel is, average avoidable |
19 | cost, which for definition, cost per incremental unit |
20 | that does not have to be paid if the incremental units |
21 | are not produced, is the best cost measure to use if |
22 | forced to use the Brooke Group analysis. Is there any |
23 | disagreement with that? |
24 | Doug? |
25 | MR. MELAMED: Yes, I actually disagree with it |
1 | as phrased. I think it would be more useful to use |
2 | avoidable cost compared -- and then add up the revenues, |
3 | because when you say average avoidable cost compared to |
4 | price, you are limiting yourself to the revenues from |
5 | the product in question, and you can't take account by |
6 | that formula, at least, of two-sided markets and |
7 | everything else. |
8 | DR. ORDOVER: And networks. I think the issue, |
9 | just to pitch it to the folks if you want to raise it or |
10 | discuss it further, I think in the American Airlines |
11 | case, there was a lot of debate as to what the right |
12 | benchmark of cost was in my view, and at least the |
13 | Government had I think proposed four, if I am not |
14 | mistaken -ï½ |
15 | MR. POTTER: Correct. |
16 | DR. ORDOVER: -- that may be three too many, |
17 | but, you know, we have offered at least one or two |
18 | ourselves. |
19 | The same thing in Spirit, I think we have come |
20 | up with two different measures of cost, but they would |
21 | apply to different types of outputs, all passengers |
22 | rather than leisure versus business. So, there is a lot |
23 | of wiggle room as to what it is that this cost measure |
24 | is going to be applied to, and as Doug pointed out, it |
25 | is also key to figure out what is the measure of |
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 |
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 |
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 |
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 |
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? |
1 | Because one view of opportunity costs is he had an |
2 | opportunity to get more profit. How do you distinguish |
3 | the two, if at all? |
4 | Patrick? |
5 | DR. BOLTON: Yes, so, this is where Doug drew a |
6 | very clear line that you should not count lost revenues |
7 | and inframarginal sales as an opportunity cost. I would |
8 | say as a matter of theory, you should count that as an |
9 | opportunity cost. The real question is just one of |
10 | administerability. |
11 | The other point I would make in this respect is |
12 | that -ï½ |
13 | DR. ELZINGA: I am sorry, could you just explain |
14 | for the benefit of at least me why, why you would count |
15 | that, the inframarginal? |
16 | DR. BOLTON: Well, because the question we are |
17 | trying to answer is, what is driving the price |
18 | reduction? Is this a move by the incumbent that will |
19 | raise profits irrespective of its anticompetitive |
20 | effects or not? To be able to answer that question, we |
21 | need to understand the nature of the profit sacrifice, |
22 | the size of the profit sacrifice and what justifies it. |
23 | Is there an efficiency rationale or is there an |
24 | anticompetitive rationale? So, we cannot avoid it, and |
25 | that is what I was going to say with respect to |
1 | recoupment. |
2 | Recoupment is the right question to ask. When |
3 | you try and make a recoupment analysis, you are |
4 | comparing profit sacrifice and return on investment. |
5 | So, it is inescapable, you have to look at lost revenue |
6 | and inframarginal sales when you do your recoupment |
7 | analysis. |
8 | Now, if you do it for your recoupment analysis, |
9 | I don't see why we should not take that into account for |
10 | the cost test, but I will leave that for -- I think I |
11 | see that as a practical problem and not a conceptual |
12 | problem. |
13 | MR. POTTER: Doug, do you want any extra time or |
14 | not? |
15 | MR. MELAMED: Well, the woman who is |
16 | transcribing this said I spoke so fast that maybe my |
17 | words were not caught before, but I will assume she got |
18 | them. |
19 | MR. POTTER: Okay. What about, how do we |
20 | distinguish in situations the appropriate costs when -ï½ |
21 | essentially price discrimination or nonlinear pricing? |
22 | Does anybody have any views on that? |
23 | Doug? No, that is not on the slide. |
24 | MR. MELAMED: What are the avoidable costs? |
25 | MR. POTTER: Fine, let's go to the next slide. |
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 |
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 |
1 | the reputation of incumbent firms for being tough, |
2 | aggressive, low-cost competitors. |
3 | My concern -- and again, so much of the |
4 | difference, perhaps, between Patrick and myself is one |
5 | of administerability -- is once you start bringing in |
6 | reputation effects as a potential hammer for antitrust |
7 | plaintiffs, what is the consequence of that for all the |
8 | good things that reputations do of incumbent firms to |
9 | keep people, even for their own good, out of markets in |
10 | which they have no business competing because they will |
11 | not be efficient utilizers of society's scarce resources |
12 | in those settings? |
13 | DR. ORDOVER: I think that the reputation |
14 | effects are almost a cornerstone of the new game |
15 | theoretic model of anticompetitive behavior of the sort |
16 | that Patrick summarized in his talk and his paper, and |
17 | there is no question that firms act in ways that try to |
18 | convey signals to the outside world and to the inside |
19 | world, and I think I would agree with Ken, not going so |
20 | far as we should not find entrants to be all such great |
21 | participants -- I like entrants. I think they should -ï½ |
22 | you know, let them slug it out. |
23 | Let's not create a presumption that some of them |
24 | may be inefficient ones or some of them are efficient |
25 | ones. Who the hell knows? It is the crucible of the |
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 |
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. |
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 |
1 | lot of money if it stays, because it cannot exit, and |
2 | therefore, it is not willing to try. If exit and entry |
3 | are easy, then I don't believe there is much power to |
4 | that informational signal. |
5 | MR. MELAMED: I agree completely with that. |
6 | MR. POTTER: Next slide. I know at least one of |
7 | our panelists said in his presentation that he would |
8 | disagree with this, so, Patrick, I believe that is you, |
9 | but we will see what the others think. Meeting |
10 | competition should not be a defense to predatory |
11 | pricing. Is there anybody who agrees with that? |
12 | MR. MELAMED: Agrees that it should not be? |
13 | MR. POTTER: Should not be. Doug, one |
14 | agreement. Patrick, you disagree. Okay, Doug, why -ï½ |
15 | DR. ORDOVER: Wait, should not be -ï½ |
16 | MR. POTTER: Should not be a defense. So, you |
17 | have essentially a high-cost producer. A lower-cost |
18 | producer comes in, more efficient, at lower price. The |
19 | high-cost producer cuts his price, lowers cost to meet |
20 | competition. Should that be protected or not in a |
21 | predatory pricing case? |
22 | DR. BOLTON: So, on the meeting competition |
23 | defense, if meeting the competition is a best response, |
24 | then this should be a defense. So, in principle, this |
25 | is an admissible defense. Administerability, again, |
1 | concerns are important here. For example, what do we |
2 | mean by meeting the competition? Is matching the price |
3 | of the entrant meeting the competition? Is that how we |
4 | define it? I would argue that's dangerous, because the |
5 | products may not be the same. If the incumbent's |
6 | product is higher quality than the entrant's, then |
7 | matching the price of the entrant is not meeting |
8 | competition. It's -ï½ |
9 | MR. POTTER: So, a jury is going to decide what |
10 | the quality-adjusted price is? |
11 | DR. BOLTON: (Nodding.) |
12 | MR. POTTER: Doug? |
13 | MR. MELAMED: I think Patrick and I might not |
14 | actually disagree but just use different words. He said |
15 | if this is the best response. If it's the best |
16 | response, then it would seem to me that the revenues |
17 | generated by the response are in excess of the avoidable |
18 | costs, in which case it passes the price-cost test, but |
19 | if that's not the case, if it fails that test, it's an |
20 | inefficient response. The fact that he's meeting |
21 | competition I don't think should make it a safe harbor. |
22 | MR. POTTER: On a more general basis, it's not |
23 | one of the slides, but what role should efficiencies |
24 | play as a defense to predatory pricing? I know, |
25 | Patrick, you think they should play a central role. Any |
1 | of the other panelists have a view on that? |
2 | (No response.) |
3 | MR. POTTER: Seeing none, we will go on to the |
4 | next slide. |
5 | It is extremely difficult to craft an effective |
6 | injunctive remedy in predatory pricing cases, and I'm |
7 | thinking of the -- I think there was at least one TRO in |
8 | a case where the judge said for purposes of the TRO, the |
9 | company couldn't price above the price that it had set |
10 | on August 1st, you know -ï½ |
11 | DR. ORDOVER: Of any year? |
12 | MR. POTTER: Well, it was a particular year, you |
13 | know, subject to changes in raw material costs. You |
14 | know, is that the remedy that -ï½ |
15 | DR. ORDOVER: That was the Baumol Test, right, |
16 | you can cut the price, but you can't raise it for five |
17 | years? |
18 | MR. POTTER: What is the injunctive remedy? I |
19 | understand what the damage remedy is if it's a private |
20 | case. What is the effective injunctive remedy? Is |
21 | there any? |
22 | Doug, have you given that some thought? |
23 | MR. MELAMED: Yes, that question I have. I |
24 | think it is very difficult, very dicey. There may be a |
25 | circumstance in which it makes sense for a court to |
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 |
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. |
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 |
1 | each of the panelists. |
2 | The final slide is, if there was one thing you |
3 | could change with the current legal approach to |
4 | predatory pricing, what would it be? And since we |
5 | started with Ken, I think this time we will end with |
6 | Ken, in reverse order. Doug, why don't you take this |
7 | one first. |
8 | MR. MELAMED: I think I would just try to |
9 | demystify it a little bit and think of it simply as part |
10 | of a complicated set of strategies that companies use |
11 | that under some set of circumstances can be |
12 | anticompetitive. |
13 | MR. POTTER: Patrick? |
14 | DR. BOLTON: Yes, I would agree with that, and I |
15 | would also vote for de-emphasis of the cost test and |
16 | putting intent back as a possible way of proving |
17 | predatory pricing, and here, I think it would be helpful |
18 | to maybe articulate the guidelines on how one would -ï½ |
19 | what's a legitimate way of proving intent and perhaps, |
20 | you know, move in that direction. |
21 | MR. POTTER: Janusz? |
22 | DR. ORDOVER: I don't think I have a favorite. |
23 | I will just say that I will agree with Doug, that we |
24 | need to get clarity on what are the public policy or |
25 | economic principles that either underlie the tests that |
1 | are being proposed, where does the -- you know, the cost |
2 | tests, where do they come from? There has to be -ï½ |
3 | going back, I think at this point we have enough |
4 | learning to try to go back to first principles and try |
5 | to understand what it is that we are trying to |
6 | accomplish, taking full account of the administerability |
7 | of whatever provisions are going to ultimately be |
8 | developed, but I think it would be foolish for us -- for |
9 | me, anyway -- to vote for the least favorite aspect of |
10 | what is out there at this point. |
11 | MR. POTTER: Ken? |
12 | DR. ELZINGA: Well, I certainly can't argue |
13 | against Janusz's call for clarity, but I think we are in |
14 | a pretty congenial equilibrium right now. |
15 | MR. POTTER: Good. I just want to point out |
16 | that over on the side, we have some of this afternoon's |
17 | panelists that were kind enough to be here this morning. |
18 | They are John Kirkwood, Tim Brennan and Rick |
19 | Warren-Boulton, and just right before we leave, I will |
20 | give each of you 30 seconds to say anything you wanted |
21 | to say about this morning's panel, or if you just want |
22 | to wait until this afternoon, feel free. Anyone want to |
23 | take a go? |
24 | DR. WARREN-BOULTON: I think I will wait for |
25 | this afternoon, but -ï½ |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
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, |
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 |
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 |
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 |
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 |
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. |
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 |
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. |
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 |
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 |
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 |
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 |
1 | actually, even if you care about total welfare, for |
2 | several reasons, the consumer welfare standard actually |
3 | could lead to higher total welfare, and I will go into |
4 | that if that comes up later on. |
5 | Finally, I think the consumer welfare standard |
6 | supports innovation better than does the total welfare |
7 | standard. Again, I will go into that if there is time |
8 | later on. |
9 | So, thank you very much. |
10 | (Applause.) |
11 | MS. SCHULTHEISS: Okay, our next speaker is |
12 | Dr. Rick Warren-Boulton, a principal of MiCRA, |
13 | Microeconomic Consulting and Research Associates, an |
14 | economics consulting firm and research firm specializing |
15 | in antitrust litigation and regulatory matters. Before |
16 | joining MiCRA -ï½ |
17 | DR. WARREN-BOULTON: We like to call that |
18 | "MiCRA." |
19 | MS. SCHULTHEISS: I like to call it "MiCRA." |
20 | Rick Warren-Boulton served in a number of positions |
21 | involving economics and antitrust law. He was an |
22 | Associate Professor of Economics at Washington |
23 | University in St. Louis. He was the Chief Economist for |
24 | the Antitrust Division at the Department of Justice. He |
25 | was also a resident scholar at the American Enterprise |
1 | Institute. |
2 | Rick has authored numerous publications -- you |
3 | laugh -ï½ |
4 | MR. KIRKWOOD: Are you still welcome there? |
5 | DR. WARREN-BOULTON: Yes, for lunch. |
6 | MS. SCHULTHEISS: Old friends, huh, Rick? |
7 | He has authored numerous publications, primarily |
8 | in the application of industrial organization economics |
9 | to antitrust and regulation. He's served as an expert |
10 | witness in numerous cases, including the Department of |
11 | Justice, U.S. versus AT&T, and for the FTC in our FTC |
12 | versus Staples and Office Depot merger. He also was an |
13 | expert with the Department of Justice in the United |
14 | States versus Microsoft. |
15 | In addition, Dr. Warren-Boulton was recently a |
16 | consultant in support of the Ross-Simmons side in its |
17 | effort to convince the Supreme Court to reject |
18 | Weyerhaeuser's petition for certiorari. Rick is yet |
19 | another one of our panelists who has some connection to |
20 | the Weyerhaeuser case. |
21 | Rick, with that... |
22 | DR. WARREN-BOULTON: Thank you. You know you |
23 | are getting old when you start talking about when you |
24 | were an expert witness in the AT&T case. |
25 | MS. SCHULTHEISS: And now they have all |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
1 | curves usually are horizontal. So, the reason why we |
2 | get that is I think because there's a real difference. |
3 | There is diminishing returns of consumption and |
4 | relatively constant returns on production. |
5 | The second thing is I think monopsony is much |
6 | easier to identify, monopsony power, for several |
7 | reasons. One is that supply inelasticity is really only |
8 | observed in a few situations, and when I stop and I |
9 | think about what they are, they are most importantly, in |
10 | my experience, when a product is a by-product of some |
11 | other product or when it is an exhaustible natural |
12 | resource. Both of those are very unusual, and, you |
13 | know, both of them actually apply in this particular -ï½ |
14 | in the Ross-Simmons case. Alder turns out to be a |
15 | by-product or what they call a come-along product in the |
16 | industry, and it is an exhaustible product. So, you can |
17 | use the screen essentially of a by-product as a -- to |
18 | limit the scope of any decision. |
19 | By the way, there is a second reason, which I |
20 | always find kind of fun, which is why observing |
21 | monopsony is so much rarer than observing monopoly. |
22 | Monopsony is inefficient, and so nearly all monopsony |
23 | situations or potential monopsony situations are solved |
24 | by vertical integration, okay? If you are an aluminum |
25 | plant and you are right next to an electricity dam and |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
1 | respect to the predatory purchasing conundrum that can |
2 | reasonably be applied by businesspeople and their |
3 | lawyers. |
4 | Remember, the purchasing department of even a |
5 | Fortune 100 company does not include these guys. They |
6 | would not know a downward sloping demand curve if it hit |
7 | them in the face. And so the notion that that is a |
8 | standard that can be applied by a businessperson in a |
9 | purchasing department is just unrealistic. |
10 | It is actually quite funny, I think, that Rick |
11 | referred to Justice Stewart's definition of pornography, |
12 | because that is also the analogy that I had here. I |
13 | think the -ï½ |
14 | DR. WARREN-BOULTON: Like minds. |
15 | MS. McDAVID: -- Ninth Circuit standard in |
16 | Weyerhaeuser provides much less guidance than Justice |
17 | Stewart's standard with respect to pornography. There |
18 | is no basis here for advising a client, and this is a |
19 | serious mistake, because ultimately, businesspeople have |
20 | to understand the rules. Justice Breyer, when he was |
21 | still on the First Circuit, said that antitrust rules |
22 | must be clear enough for lawyers to explain them to |
23 | clients and be administratively workable, and that in |
24 | formulating antitrust liability standards, the courts |
25 | must consider what advice the lawyer is going to give. |
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 |
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 |
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 |
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. |
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 |
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 |
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' |
1 | costs. |
2 | Now, nonprice predation sometimes involves |
3 | prices, like here, prices are involved, and when you |
4 | call it predation, that tends to plug into Brooke Group, |
5 | and as I said, I think there are two vastly different |
6 | paradigms. |
7 | MS. SCHULTHEISS: So, in terms of what we are |
8 | speaking about here today, Rick focused on the narrow |
9 | with the predatory overbidding or overbuying that is in |
10 | the Weyerhaeuser case, but then there is a separate area |
11 | of conduct that involves some type of exclusionary |
12 | conduct, what you call raising rivals' costs, what Tim |
13 | might call exclusion, and looking at the complementary |
14 | market. |
15 | DR. SALOP: I think what all three of us were |
16 | taking issue with in your question was the word |
17 | "predatory." Had you just said "conduct that gives |
18 | market power in the input market" and "conduct that |
19 | gives market power in the output market," we all would |
20 | have -ï½ |
21 | MS. SCHULTHEISS: Do you all agree with that |
22 | one? |
23 | DR. BRENNAN: More or less. |
24 | MS. SCHULTHEISS: What about you, Rick? |
25 | DR. WARREN-BOULTON: First of all, yes, I think |
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 |
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. |
1 | MS. SCHULTHEISS: Steve, I take it you agree. |
2 | DR. SALOP: Yes, I agree with what they said, |
3 | but they didn't answer your question, which is, well, |
4 | what the agency should not be doing, and frankly -ï½ |
5 | MS. SCHULTHEISS: I do not know if it is |
6 | necessarily what we should or should not, but our |
7 | primary focus. Should we be worried about one more than |
8 | the other? |
9 | DR. SALOP: The way you phrased it is something |
10 | that I think is all too apparent in this administration, |
11 | which is you are spending an awful lot of time deciding |
12 | what you should not do, and you are not doing a heck of |
13 | a lot, and so I think that in the exclusionary conduct |
14 | area, the agencies ought to get involved and start |
15 | looking for exclusionary conduct cases rather than |
16 | protecting monopolists. |
17 | MS. SCHULTHEISS: Okay. |
18 | DR. WARREN-BOULTON: I think I am probably |
19 | agreeing with everybody. First of all, I agree that -ï½ |
20 | I mean, the whole point is that, you know, predatory |
21 | overbidding to get monopsony power, as I said, is |
22 | incredibly rare. So, the answer to the first question |
23 | is yes, and oddly enough, I would agree with Steve that |
24 | I think exclusion is a real problem, and so if you are |
25 | going to spend more money on something, I would have you |
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 |
1 | not think we should assume that it will continue to be |
2 | an unusual circumstance. |
3 | MS. SCHULTHEISS: So, you -ï½ |
4 | MS. McDAVID: And virtually every major firm in |
5 | this country sues in the Ninth Circuit. So, unless the |
6 | law gets clarified, I think we will see these cases |
7 | because they will get syndicated. |
8 | MS. SCHULTHEISS: Steve, did you want to say |
9 | something in response to that? |
10 | DR. SALOP: Yes, I just wanted to -- I just |
11 | jotted down some overbuying cases before, and so I know |
12 | we see sort of the classic, you know, Ross-Simmons cases |
13 | being rare, but one, as I said before, Ross-Simmons |
14 | alleged raising rivals' costs as well as -- as well |
15 | as -ï½ |
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 -ï½ |
22 | I mean, the beef case was a conspiracy in overbuying, |
23 | the Monfort case was about overbuying, went to the |
24 | Supreme Court, Socony-Vacuum was about overbuying, Alcoa |
25 | had a piece that was involved in overbuying, and in |
1 | terms of what is going on right now, stuff the agency |
2 | has in its shop, all the stuff about shelf space is all |
3 | about overbuying, too, you know, could be thought of as |
4 | an overbuying issue. There could be overbuying of |
5 | patents. Alcoa, you know, allegedly tied up some |
6 | patents, so it made it harder for people in -- not sort |
7 | of classic predatory bidding, but, you know, you could |
8 | certainly have overbuying of patents, too. |
9 | So, I do not think it is a -- I think this idea |
10 | that it is just exhaustible resources that are |
11 | by-products, I know that covers the old molybdenum case, |
12 | one of my favorite cases when I was at the FTC, but I |
13 | think it is rather broader than that. |
14 | DR. WARREN-BOULTON: Can I just make a -- I |
15 | mean, I have not gone back and looked at the cases, but |
16 | I would question how many cases are there -ï½ |
17 | MS. SCHULTHEISS: Can you speak into the |
18 | microphone? |
19 | DR. WARREN-BOULTON: I'm sorry, how many cases |
20 | are there in overbidding to gain monopsony power. Steve |
21 | cites a tobacco case. I do not think that was to gain |
22 | monopsony power for tobacco growers. I thought that was |
23 | to exclude other, you know, cigarette producers. The |
24 | beef case, I do not think they ever managed to -- it was |
25 | ever shown that they ever managed to raise the price of |
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 |
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. |
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 |
1 | monopsony. They would be driving down the price they |
2 | paid for the natural gas. |
3 | The triangles, the welfare losses, the dead |
4 | weight stuff, all that stuff would be exactly the same |
5 | in both cases, at least qualitatively it would be the |
6 | same, but one's monopoly and one's monopsony, and so |
7 | it's purely in some sense a -- I don't know, I won't say |
8 | it's arbitrary or artificial, but it is the way those |
9 | markets just happen to work. So, one needs to be |
10 | careful about making a distinction in that way, that one |
11 | can turn monopoly into monopsony in some sectors without |
12 | a great deal of difficulty. |
13 | DR. WARREN-BOULTON: Yes, I -ï½ |
14 | MS. SCHULTHEISS: Rick is going to have to get |
15 | the last word I see. |
16 | DR. WARREN-BOULTON: -- I very much agree. You |
17 | can be regarded as buying a service or selling, and it |
18 | can be based on form. I think one implication that I |
19 | think is kind of interesting is that in terms of a |
20 | welfare loss, you know, the idea that somehow welfare |
21 | losses are only important if they happen downstream |
22 | rather than upstream strikes me as kind of bizarre. |
23 | MS. SCHULTHEISS: Let's get to that in a second, |
24 | because I want to get to that issue. |
25 | DR. WARREN-BOULTON: But that is one of the |
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. |
1 | DR. WARREN-BOULTON: The question is, what does |
2 | the audience grasp? |
3 | DR. SALOP: You think saying they bought more |
4 | than necessary is a good proxy for recoupment? Is that |
5 | what you just said? |
6 | DR. WARREN-BOULTON: No, paying more than |
7 | necessary, it seems to me to be to a -- I don't know, an |
8 | ordinary person's idea of was there a profit sacrifice |
9 | here, all right, which is at least one element. |
10 | MS. SCHULTHEISS: But in terms of precedent, how |
11 | can that possibly -- you are just agreeing with the |
12 | statement I just said, then. You would not have that |
13 | kind of an objection to the standard that was set out in |
14 | the jury instruction. |
15 | DR. WARREN-BOULTON: You know, I am saying |
16 | anybody here, I think anybody here could probably today |
17 | come up with a better set of jury instructions than |
18 | that. How bad it is, I do not know. |
19 | MS. SCHULTHEISS: I do want -ï½ |
20 | DR. SALOP: Do you think a reasonable jury could |
21 | reach the right conclusion if they tried to apply that |
22 | jury instruction? |
23 | DR. WARREN-BOULTON: You know, actually, I have |
24 | been on a jury, and I have to tell you that my personal |
25 | experience is that what a jury decides has little, if |
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 |
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. |
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 |
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 |
1 | the Supreme Court has made it very clear that harm to |
2 | competitors is not enough, and I think the same thing |
3 | should be true with harm to suppliers. |
4 | The tricky part of this -- and I am sorry, this |
5 | is kind of a long answer -- the tricky part is that it |
6 | is quite clear to make an agreement -ï½ |
7 | MS. SCHULTHEISS: Could you speak more into the |
8 | mic? I'm sorry. |
9 | DR. SALOP: I'm sorry. |
10 | It's quite clear that naked agreements among |
11 | competitors, buyer-side competitors, to fix prices that |
12 | they make to inputs is illegal and should be per se |
13 | illegal, and -ï½ |
14 | MS. SCHULTHEISS: Well, that is horizontal, |
15 | though. |
16 | DR. SALOP: But that is where antitrust starts. |
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 |
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 |
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 |
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 |
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 |
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 |
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 |
1 | distinction. I do not think you have to say that |
2 | somehow by bringing in producer surplus, we are somehow |
3 | worrying about competitors. |
4 | DR. SALOP: Well, you said something very |
5 | different. You said you would be willing to find an |
6 | antitrust violation where there is no consumer harm. |
7 | That is different from saying that you are adopting the |
8 | total welfare standard as the overarching standard to |
9 | govern antitrust cases, because if you were adopting the |
10 | total welfare standard to govern antitrust cases, then |
11 | the simple business tort of burning down your neighbor's |
12 | fields would lower total welfare, and it could support |
13 | an antitrust violation. |
14 | Certainly if you and your neighbor burned down |
15 | everybody else's fields, so you would not get into all |
16 | this complexity -ï½ |
17 | MS. SCHULTHEISS: Let's deal with a single firm, |
18 | which is what we are really focusing on. |
19 | DR. SALOP: Actually, we are not. We are |
20 | focusing on what the welfare standard should be. |
21 | MS. SCHULTHEISS: Right, but in connection with |
22 | Section 2. |
23 | DR. SALOP: Well, why would you have a different |
24 | welfare standard for Section 1 and Section 2? Why would |
25 | you want to make -- why would you want to gerrymander |
1 | antitrust and make it incoherent? You know, Tim pointed |
2 | out this notion of the pipeline, you know, if you have |
3 | different rules governing how -- if the rules are very |
4 | different according to whether the pipeline does a |
5 | tolling agreement where they sell services or whether |
6 | they buy the gas and resell it at the other end, you are |
7 | going to have incoherent antitrust. If you are going to |
8 | have dramatically different rules for tying, exclusive |
9 | dealing, vertical mergers, where one's per se illegal, |
10 | one's rule of reason, and one's virtually per se legal, |
11 | well, since lawyers can characterize conduct pretty |
12 | easily as any one of those three boxes, you are going to |
13 | end up with very incoherent antitrust, which we did |
14 | until the eighties. So, I think saying this is Section |
15 | 2, not Section 1, that is a recipe for disaster. |
16 | MS. SCHULTHEISS: Janet, did you want to respond |
17 | to that? |
18 | MS. McDAVID: No. |
19 | MS. SCHULTHEISS: Okay. Rick? |
20 | DR. WARREN-BOULTON: Well, the broad question |
21 | is, gee, do we have to have the same rules in every |
22 | situation? My answer is no. If you can distinguish |
23 | between situations, then you can have the same rules -ï½ |
24 | you can have different rules. There is no -ï½ |
25 | MS. SCHULTHEISS: But would the test be the |
1 | same? |
2 | DR. WARREN-BOULTON: Sure. I mean, I think the |
3 | test for naked price-fixing is different than the |
4 | test -- you know, for the safe harbor for naked |
5 | price-fixing is very small. I think the safe harbor |
6 | for, you know, predatory pricing should be quite large. |
7 | MS. SCHULTHEISS: No, I mean in terms of the |
8 | consumer welfare versus total welfare, that test should |
9 | be the same across all of the violations or different? |
10 | Steve is arguing I think that it should be the same |
11 | welfare test regardless of the violation. |
12 | DR. WARREN-BOULTON: Oh, I see, yes. You can |
13 | argue what is the best welfare, but somehow suppliers |
14 | are upstream or downstream, welfare does not change |
15 | depending on -ï½ |
16 | MS. SCHULTHEISS: So, you agree with Steve, |
17 | then, that whatever welfare test -- whatever welfare |
18 | test is chosen should apply across all of the various |
19 | violations you're looking at? |
20 | DR. WARREN-BOULTON: As long as he chooses my |
21 | welfare test, then -ï½ |
22 | MS. SCHULTHEISS: Janet, would you agree with |
23 | that? Yes, no? |
24 | MS. McDAVID: I think we should probably move |
25 | on. |
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? |
1 | DR. WARREN-BOULTON: Total welfare -- I mean -ï½ |
2 | MS. SCHULTHEISS: Let's do this in order. Jack |
3 | looked like he wanted to respond. Let me let Jack |
4 | respond first. |
5 | MR. KIRKWOOD: Yes, that was a point I didn't |
6 | comment on before, but I do agree with Steve there. If |
7 | you do have a case in which there is supplier harm but |
8 | consumer benefit, then I would go with the consumer |
9 | welfare standard. I think that does make antitrust more |
10 | coherent. |
11 | DR. BRENNAN: Total welfare went up in that |
12 | standard, so I would stick with it. |
13 | DR. SALOP: Yes. Now, how do you know that |
14 | total welfare went up? Did you do that calculation? |
15 | DR. BRENNAN: It is a -- I guess it is a |
16 | presumption for me, kind of 101, yes. |
17 | MS. SCHULTHEISS: I mean, I am going to give |
18 | Rick and Janet one more chance, and then we will get off |
19 | this. |
20 | DR. WARREN-BOULTON: How about we get a group of |
21 | consumers that get together a group of monopsonized |
22 | consumers, is this somehow a good thing? I don't think |
23 | so. |
24 | DR. SALOP: That is the cartel case decided by |
25 | Judge Breyer at the time, and he said it was okay. |
1 | DR. WARREN-BOULTON: Well -ï½ |
2 | MS. McDAVID: I think in this -- in the |
3 | circumstance that Steve has posited, we have got an |
4 | integrated joint venture that will be evaluated under |
5 | the rule of reason, there is an efficiency and a |
6 | business rationale, and under the rule of reason, you |
7 | probably do not conclude it is illegal. |
8 | MS. SCHULTHEISS: Well, I want to get to |
9 | something else that is completely off of this topic for |
10 | a few minutes, because I want to make sure we have time |
11 | for it, and that is related to relief and remedies. |
12 | How do you deal with relief and remedies and in |
13 | particular in the overbidding situation? Should a court |
14 | enjoin the defendant's pricing? How do you deal with |
15 | that issue in the context of this type of conduct? |
16 | MS. McDAVID: Well, in the private party |
17 | litigation, it is going to be damages. |
18 | MS. SCHULTHEISS: I mean in a government case. |
19 | MS. McDAVID: In a government case, probably -ï½ |
20 | well, then you need to have a standard, and a standard |
21 | of fairness does not allow a remedy. |
22 | MS. SCHULTHEISS: But even if you have the |
23 | Brooke Group standard, what is the Government's remedy |
24 | in that situation on the predatory bidding, you know, |
25 | raise your prices or lower your prices? I mean -ï½ |
1 | MS. McDAVID: It is pretty unlikely to me that |
2 | this is going to be a government case, although I think |
3 | there is a role for the Government to play in Section 2 |
4 | enforcement, an important role. These cases are going |
5 | to come up in disputes among rivals. |
6 | MS. SCHULTHEISS: So, you are thinking the |
7 | overbidding context, it is not the type of case that the |
8 | Government should be getting into? |
9 | MS. McDAVID: It is the type of case that I |
10 | doubt the Government will get into. It is the type -ï½ |
11 | the Government typically intervenes in cases where there |
12 | is kind of a broader principle to be generated. This is |
13 | typically an intrafirm dispute or interfirm dispute, and |
14 | they rarely get into those circumstances. |
15 | MS. SCHULTHEISS: Well, when we are looking at |
16 | what test to apply, though, you know, for example, you |
17 | know, we did file an amicus brief in this case, |
18 | obviously concerned with the test and the jury |
19 | instructions from the Ninth Circuit, if it had been a |
20 | government case, and I mean this sincerely, what kind of |
21 | relief could the Government get in any kind of an |
22 | overbuying monopsony type case like this? Does anybody |
23 | have any -ï½ |
24 | DR. WARREN-BOULTON: I am about to agree with |
25 | Janet. I think there are situations in which we rely |
1 | primarily on deterrents -- there are situations in which |
2 | deterrents, ex post penalties, are simply much, much |
3 | less expensive than ex ante penalties. I mean, you |
4 | think of how would you have hypothetically solved |
5 | Weyerhaeuser. Well, I suppose to me the answer is you |
6 | deter basically through either private litigation or a |
7 | fine. How could you prevent this situation? Would you |
8 | have to divest? You would have to sort of break up |
9 | Weyerhaeuser. |
10 | It seems to me that facing an alternative, |
11 | looking at the costs of preventing bad behavior through |
12 | structural means, that is to say, you know, forcing |
13 | Weyerhaeuser to sell off sawmills versus simply having |
14 | an ex post, you know, you will face damages if you, you |
15 | know, behave badly. A behavioral remedy is probably |
16 | better than a structural remedy, and if it is |
17 | behavioral, it seems to me I agree with Janet, it is |
18 | really something -- it is for private litigation. |
19 | Steve? |
20 | MS. SCHULTHEISS: Would you agree with that, |
21 | DR. SALOP: No, I thought -- I think there is an |
22 | answer. First of all, what the Government usually does |
23 | is it gets an injunction. You tell them not to violate |
24 | the standard anymore, and once you know what your |
25 | standard is -ï½ |
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 |
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 |
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 |
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 -ï½ |
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 |
1 | the person who should be suing for damages is the timber |
2 | owner. I mean, they are the guys who supposedly have |
3 | been harmed by this, who are being monopsonized. |
4 | MS. McDAVID: They weren't sure, they were |
5 | getting a little extra. |
6 | DR. WARREN-BOULTON: The question is, why do we |
7 | have the competitor suing rather than the timberer, and |
8 | the answer is because the competitor gets hurt first, |
9 | right? Now, we could, of course, simply have said |
10 | forget it, let's wait, right? But having the competitor |
11 | sued is a sort of an ex ante, prophylactic way to |
12 | prevent presumably the monopsony harm to the person you |
13 | are really worried about, you know, who probably never |
14 | turned up in this litigation, which was the people who |
15 | were actually sawing the timber. They are the guys who |
16 | were supposedly monopsonized. |
17 | MS. SCHULTHEISS: I am going to ask Ken and |
18 | Patrick, since they have spent the afternoon with us and |
19 | have clearly given a lot of thought to these issues, and |
20 | we talked about it extensively this morning on the |
21 | sell-side, whether you have any response to the remedies |
22 | on this side of it, of the issue. |
23 | DR. ELZINGA: I just have a couple reactions. |
24 | First of all, I must confess -ï½ |
25 | MS. SCHULTHEISS: Is there a mic near you? |
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 |
1 | about damages, it occurred to me, some of the issues |
2 | were brought up similar to the case, the |
3 | Sotheby's-Christie's -ï½ |
4 | MS. SCHULTHEISS: The auction case? |
5 | DR. BOLTON: -- cartel, yes, where there was an |
6 | issue who was harmed and how do you -- was it the -- was |
7 | it the buyers of art or was it the sellers and who |
8 | should be the recipient of the damages, and in that |
9 | case, I forget how it was decided in the end, but there |
10 | was reason to believe that the wrong party was |
11 | collecting the damages. Maybe someone else will -ï½ |
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 -ï½ |
20 | Social Security taxes, who pays it, is it labor or is it |
21 | the firm? If you ask anybody on the street, they say, |
22 | oh, I pay half of it and my employer pays half of it. |
23 | If you ask an economist, and he will say 95 percent is |
24 | paid for by the workers, but the incidence of a tax |
25 | bears very little resemblance to the accounting and |
1 | collection of that tax, and I think in the same way, it |
2 | is extraordinarily difficult to find out who actually |
3 | pays for antitrust violations by simply looking at the |
4 | accounting incidence. |
5 | DR. SALOP: Yes, I -ï½ |
6 | MS. SCHULTHEISS: We have two minutes. |
7 | DR. SALOP: Okay, so I will just make a law |
8 | professor remark. There is Illinois Brick and there is |
9 | Hanover Shoe, and that says the direct purchasers get to |
10 | sue the direct purchasers or sellers of the art. They |
11 | get the money. There is no pass-on defense by the |
12 | cartel, and the -- you know, the Supreme Court had a |
13 | reason for that, and so the -- I would think the proper |
14 | people to get the money under the current law are the |
15 | sellers of the art, not the buyers. |
16 | MS. SCHULTHEISS: Okay. Well, I think with |
17 | that -ï½ |
18 | DR. ELZINGA: I worked for the judge on that |
19 | case. Most of it went to the lawyers. |
20 | MS. SCHULTHEISS: Well, on that note -ï½ |
21 | MS. McDAVID: The entrepreneurial plaintiff's |
22 | bar. |
23 | DR. WARREN-BOULTON: Well, at least some of it |
24 | went to you. |
25 | MS. SCHULTHEISS: On that note, we will wrap |
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 |
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 |
Updated December 29, 2023