# UNITED STATES FEDERAL TRADE COMMISSION > Briefs, arguments, decisions, and more. URL: https://www.frixlaw.com/law-library/documents/agency%3Aftc%3A8a9f400fc8c77b99 ## Record - **Collection:** Agency decision - **Document type:** Agency decision ## Text 1 1 UNITED STATES FEDERAL TRADE COMMISSION 2 3 4 5 UNILATERAL EFFECTS ANALYSIS 6 AND LITIGATION WORKSHOP 7 8 9 10 11 U.S. Federal Trade Commission 12 601 New Jersey Avenue, N.W. 13 Conference Center 14 Washington, D.C. 20001 15 16 Tuesday, February 12, 2008 17 9:00 a.m. to 5:00 p.m. 18 19 20 21 22 23 24 25 Reported by: Susanne Bergling, RMR-CLR For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 2 1 PARTICIPANTS 2 3 Introductory Remarks: 4 Deborah Platt Majoras 5 Panel 1: 6 Andrew I. Gavil Janet L. McDavid Robert Willig David P. Wales, Moderator 7 8 Panel 2: 9 10 11 Jonathan B. Baker Kathryn M. Fenton Richard G. Parker Daniel M. Wall Jeffrey Schmidt, Moderator 12 Panel 3: 13 14 15 Hon. Douglas Ginsburg Hon. Diane Wood Michael J. Bloom Richard Liebeskind William E. Kovacic, Moderator 16 Panel 4: 17 18 19 William Baer Susan Creighton Richard Rapp Constance Robinson J. Thomas Rosch, Moderator 20 Panel 5: 21 22 23 Orley Ashenfelter Dennis Carlton Carl Shapiro Joseph Simons Michael R. Baye, Moderator 24 25 For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 3 1 CONTENTS 2 3 Introductory Remarks............................. 5 4 Deborah Platt Majoras............................ 7 6 Panel 1.......................................... 19 7 Andrew I. Gavil.................................. 22 8 Janet L. McDavid................................. 39 9 Robert Willig.................................... 29 10 Discussion....................................... 46 12 Panel 2.......................................... 66 13 Jonathan B. Baker................................ 69 14 Kathryn M. Fenton................................ 76 15 Richard G. Parker................................ 84 16 Daniel M. Wall................................... 91 17 Discussion....................................... 107 5 11 18 19 Panel 3.......................................... 129 20 Michael J. Bloom.................................130/157 21 Richard Liebeskind............................... 143 22 Discussion....................................... 167 23 24 Panel 4.......................................... 178 25 Discussion....................................... 181 For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 4 1 Panel 5.......................................... 227 2 Orley Ashenfelter................................ 244 3 Dennis Carlton................................... 231 4 Carl Shapiro..................................... 237 5 Joseph Simons.................................... 254 6 Discussion....................................... 258 7 8 Conclusion....................................... 284 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 5 1 P R O C E E D I N G S 2 - 3 - - - - INTRODUCTION 4 5 MR. SCHMIDT: Good morning. I think we are 6 going to try to start the program. 7 Workshop on Unilateral Effects. 8 Director of the Bureau of Competition, and we are very 9 glad to have you here today. 10 about this program. 11 workshop is the brainchild of Chairman Majoras, and it 12 represents the best of the FTC in trying to better 13 understand some of the important competition policy 14 issues that we face. 15 Welcome to the FTC's I am Jeff Schmidt, the We are really excited As some of you may know, this I have the chore of doing a couple housekeeping 16 tasks here, so if you will indulge me as I go through 17 this to make sure that I have covered the requirements. 18 I think the -- let's see, the first thing is I have been 19 asked to remind you that the agenda today is a full one, 20 so that if you can try to be back in your seats by the 21 time lunch is over with and breaks are over with, we can 22 hopefully stay on schedule. 23 And I have also been asked to ask you to use the 24 side doors instead of the center doors, for reasons that 25 are not particularly clear to me. For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 6 1 Secondly, if you will turn off the ringers on 2 your cell phones, BlackBerries, pagers, and the like, 3 and I will do likewise when I get down from here. 4 And third, the restrooms are out the glass 5 doors, past the security desk, and then behind the 6 elevator bank to the left. 7 restrooms are located there. 8 Both the men's and women's And then fourth, if you do leave the building 9 during the day, unfortunately, for those of you who are 10 not FTC employees, you will need to go through security 11 again. 12 couple extra minutes to do that. 13 So, if you can be sure to give yourselves a And then finally, as a federal government 14 agency, we do practice certain safety measures. 15 Probably the most important thing for you to know is -- 16 obviously you know the one exit that you came in through 17 -- if you need to leave the building in the event of an 18 emergency. 19 There will be FTC people who will also be obviously here 20 and are on site in the event that we have any problems, 21 but, of course, we are not anticipating that. There is also an exit immediately behind us. 22 So, with that, I'd like to welcome the Chairman 23 of the Federal Trade Commission, Deborah Platt Majoras, 24 to open our workshop. 25 (Applause.) For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 7 1 2 OPENING REMARKS 3 4 CHAIRMAN MAJORAS: Well, thank you very much, 5 everyone. 6 the morning in Washington, especially on election day. 7 It is always good to see a robust crowd in I welcome you to this workshop at the FTC. As 8 many of you know, the FTC has found that when we are 9 working through particular policy issues, we often find 10 it very valuable to bring in experts from the outside 11 who can then, in a public forum, communicate their views 12 and help us think through the issue. 13 discussions can take whatever form or length is required 14 for the issue. 15 Our public Just last week, for example, we held a one-day 16 round table with DOJ to explore our Joint Technical 17 Assistance Program in the international arena. 18 about a year ago this week, we had a two-day forum on 19 the broadband access issue, which has been dubbed Net 20 Neutrality. 21 past 18 months, we and DOJ have hosted 29 sessions of 22 experts discussing the appropriate application of 23 Section 2 of the Sherman Act to business conduct. Just And then, as many of you know, over the 24 So, today, you have been good enough to join us 25 as we gather to discuss unilateral effects analysis in For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 8 1 merger review and in the litigation context, and I am 2 pleased to say that we have gathered really a highly 3 knowledgeable and thoughtful group of panelists, and I 4 am very grateful to all of you for agreeing to lend your 5 views. 6 Back in February of 2004, the FTC and DOJ held a 7 merger enforcement workshop, which focused on whether 8 the analytical framework set forth in the 1992 9 Guidelines, which, of course, had its roots in the 1982 10 Guidelines, was adequately serving the dual purposes of 11 leading to the correct decisions in horizontal merger 12 review and providing reasonably clear guidance to 13 businesses and their counselors. 14 The workshop participants generally agreed that, 15 in fact, the Guidelines framework was serving those 16 purposes. 17 a reworking of the Guidelines, but rather, the agencies' 18 commentary on the Horizontal Merger Guidelines, through 19 which we explained, by reference to specific cases, 20 including cases where we had closed the investigation, 21 how we have applied the Guidelines to actual mergers. So, borne out of that workshop, then, was not 22 If you reviewed the section on unilateral 23 effects, it shows a large number of enforcement actions, 24 most of which resulted in consent decrees. 25 little doubt, I think, among antitrust practitioners There can be For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 9 1 that unilateral effects is recognized as a central 2 antitrust concern, and that the Government has a record 3 of success in obtaining relief in these cases. 4 Of course, the record is not perfect. 5 litigated matters, both the FTC and DOJ have suffered 6 some losses in differentiated products cases under a 7 unilateral effects theory. 8 in the Whole Foods case, the district court did not 9 grant the preliminary injunction that the FTC sought, 10 and before that, DOJ lost the SunGard and Oracle 11 challenges. 12 cases in which a unilateral effects theory of harm has 13 been alleged, as in Staples, Swedish Match, and Libbey, 14 the courts' decisions have really not expressly 15 discussed the application of unilateral effects theory. 16 In Most recently, for the FTC, Even when the Government has prevailed in Now, there may, of course, be no meaningful 17 pattern in these losses. 18 likely will lose some cases over time, as only the 19 toughest cases result in litigation; and try as we do, 20 we cannot determine with absolute precision on which 21 side of the line a close case will fall according to a 22 court. 23 if we believe that we have the evidence to support our 24 position that a merger is likely to be anticompetitive. 25 If we are doing our jobs, we Still, we cannot shy away from the tough cases Clearly, though, if you look at the cases and For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 10 1 particularly the losses, they do show, I think, what we 2 experience, which is that there are challenges in 3 proving a relevant market in which we allege that the 4 likely harm will arise out of the loss of competition 5 between two competitors that have served as next-best 6 substitutes to one another for a significant number of 7 customers. 8 Recall that, for example, in the Oracle case, 9 the Justice Department sought to bar Oracle's 10 acquisition of PeopleSoft. 11 incumbent manufacturers in a market defined as 12 enterprise resource planning system software that 13 handles human resources management and financial 14 management systems for customers that made minimum 15 purchases of $500,000. 16 of course, argued for a much broader market that 17 included not just those programs, but also other forms 18 of ERP programs, as well as non-ERP software solutions, 19 and would not have limited the market by size of 20 customer sales. 21 proposed market expanded the number of market 22 participants. 23 These were two of the three By comparison, the defendants, So, not surprisingly, defendants' I am obviously simplifying in the interest of 24 time here, but there, the court found that DOJ failed to 25 prove its alleged product market, at least in part For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 11 1 because it was not consistent with business delineations 2 recognized within the industry. 3 presented testimony from numerous customers that they 4 might prefer defendants' products over some of the 5 alternatives, but, said the court, none testified about 6 how they would respond in actual purchases to a 7 post-merger SSNIP. 8 the Government to rely principally on qualitative 9 materials like market research reports and declarations 10 from customers and industry consultants. 11 The Government had Lack of hard, quantitative data led The defendants countered with examples of users 12 that had implemented alternatives to the defendants' 13 products. 14 Government had failed to define the alleged, narrow, 15 relevant market, which meant that the shares that you 16 then calculate to show concentration levels weren't 17 correct and that ultimately, the Government's estimates 18 of competitive effects, based on that market definition, 19 also had to be disregarded. 20 Ultimately, the court found that the Then you go to the SunGard case. The district 21 court there rejected DOJ's market definition in refusing 22 to bar SunGard from acquiring the assets of Comdisco. 23 These companies, as well as IBM, were in the business of 24 providing shared hot-site services which are backup 25 computer centers that you use in the event of a For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 12 1 disaster. 2 shared hot-site services for customers with mainframe 3 and midrange computer processing centers. 4 contended that there were a lot of alternatives to these 5 that customers could and did turn to to safeguard 6 themselves in the event of disasters. 7 The Government alleged a market that was Defendants Both sides offered customer testimony to support 8 their contentions, but there the court rejected the 9 customer testimony, finding that both sides were 10 engaging in cherry-picking sampling and that neither 11 side's witnesses were representative of all existing and 12 future customers. 13 relevant market that was neither the narrow market that 14 DOJ had alleged or the broader market that the 15 defendants had alleged. 16 market somewhere in between. 17 Ultimately, the court found a In fact, the court found a And finally, if you look at the Commission's 18 challenge to Whole Foods' acquisition of Wild Oats, the 19 court there rejected the contention that the relevant 20 market was the premium natural and organic supermarket. 21 There, the Government presented not only economic 22 evidence but evidence that was taken from the parties 23 themselves that, in fact, showed that the two were 24 uniquely close competitors. 25 Whole Foods and Wild Oats competed at a certain level There was no doubt that For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 13 1 with other supermarkets, and we never denied that, but 2 staff presented evidence that the companies believed 3 that the other was a uniquely close competitor, and 4 thus, made decisions on that basis; and as the Whole 5 Foods CEO told his board in justifying the transaction, 6 that the acquisition would eliminate Wild Oats as a 7 platform for conventional supermarkets to get into the 8 organic market segment, and the entry through that 9 avenue would be only a threat to his market position. 10 And in addition, after paying a premium for stores, 11 Whole Foods made clear it had the intention to close 12 dozens of stores and to scrap plans to build new stores. 13 Of course, the district court did not see the 14 evidence there as we did and concluded that we were 15 wrong about what constituted the relevant market, and 16 that case is now on appeal. 17 Don't get me wrong. The courts play an 18 absolutely critical role in U.S. merger enforcement. 19 Indeed, almost uniquely so if you look at our courts' 20 role in comparison with many courts around the world. 21 And after every litigated case, it is very important 22 that we carefully evaluate the courts' decisions, our 23 own analysis, and our evidentiary presentations. 24 You know, the fact that litigated cases happen 25 so infrequently -- indeed, the three cases litigated by For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 14 1 the FTC over the past year were virtually unprecedented 2 over the past couple of decades. 3 haven't litigated so many cases in a year. 4 makes it all the more important that we learn from each 5 and every court decision. The agencies just So, that 6 In addition, because most merger decisions are 7 not litigated, we have a great responsibility to ensure 8 that we are basing those decisions, most of which result 9 in consent decrees, on solid analysis which would be 10 supportable in the courts if litigation were necessary. 11 And if we lose, it is essential that we take a critical 12 look at our legal analysis and presentation to 13 determine, to the extent we can, how and why we were 14 unable to convince the court of our position. 15 In this regard, I am very proud of the 16 debriefing efforts that are being undertaken and have 17 been for the last six months within our agency among the 18 economists and the lawyers to think these things 19 through, and today's workshop is another step in our 20 process. 21 but given the human limitations on objectivity, we may 22 be so close to a case or an approach or a set of 23 strategies that our own introspective evaluation is just 24 simply not enough. 25 We can identify ways to improve internally, The workshop combines a lot of our thinking, For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 15 1 covering many of the areas that we and others outside 2 have identified as worthy of discussion. 3 has market definition, which has been such an important 4 tool in analysis, become an end in unilateral effects 5 cases rather than a means to determine if the merged 6 entity will have the ability to exercise power? 7 is it because, as Professors Farrell and Shapiro argue 8 and probably will talk about today in a preliminary 9 draft paper, the Guidelines have shoehorned unilateral 10 effects analysis into the traditional market definition 11 concentration framework that has its roots in 12 coordinated effects analysis? 13 For example, If so, We will define markets in unilateral effects 14 cases in problematical ways in litigation, because given 15 the nature of the analysis of closeness of substitution, 16 they appear to judges to have been gerrymandered and not 17 always consistent with our views as consumers; and, of 18 course, we are all consumers, including judges. 19 ready to touch the third rail and discuss whether market 20 definition is necessary in a case in which we can 21 present direct evidence of competitive effects? 22 regard, are we just getting tripped up over our own 23 terminology and our step-by-step analysis, and should we 24 do a better job of explaining, as I tried in the 25 Evanston opinion, that in differentiated product Are we In that For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 16 1 unilateral effects cases, market definition and 2 competitive effects are simply two sides of the same 3 coin no matter how we label? 4 might argue, stick to traditional market definition and 5 concentration calculations because, while sometimes 6 imperfect, they provide important disciplines on legal 7 analysis? 8 the fact that a huge percentage of mergers we review 9 have to be analyzed within only 30 days or less, 10 necessitating that we have to have some tools to be able 11 to find the right answer quickly? 12 evidence and how we present it? 13 reject customer declarations, customer testimony, 14 parties' unvarnished statements about competition and 15 mergers in favor of litigation declarations and economic 16 evidence at different times, all of which, some of us 17 believe, at least at some points, to be very important 18 evidence in these cases. 19 Or should we, as some Should our thoughts on this be influenced by What about our We have had judges Are we moving toward a system where fancy 20 econometrics will win the day, much like we hear about 21 jurors who have seen so much CSI and Law & Order on TV 22 that they insist on fancy DNA or fingerprint evidence in 23 order to find guilt in a case? 24 noneconomic and economic evidence are most probative in 25 these cases, and how does our answer vary by factual What types of For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 17 1 conditions, where we have dynamic versus static markets; 2 if we have industrial products cases versus retail 3 cases, direct to consumer? 4 How do we handle new economic learning when we 5 go in to court? 6 not a static discipline, and we want to learn as the 7 economics develop. 8 litigation standpoint? 9 experts? 10 especially if the market definition -- and you heard 11 some of the ones that I mentioned in some of these 12 cases -- are just simply not intuitive to us as 13 consumers? 14 This is very important, because ours is So, how do we handle that from a How important are industry And how can we best tell the story to a judge, Now, later today, I am very excited that we are 15 going to have a mock closing argument over a 16 hypothetical ice cream merger, and as you will see from 17 the facts there, the Government in that hypothetical 18 case alleged that superpremium ice cream is a separate 19 market from other types, with the defense taking the 20 position that ice cream is ice cream. 21 the economics and facts are not necessarily completely 22 in alignment with what our intuition might be. 23 panel will provide us with really an exceptional 24 opportunity to hear how two experienced judges go about 25 weighing the often complex and contradictory testimony As we will see, So, this For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 18 1 in economics, which is typically presented in an 2 antitrust merger case. 3 So, with that, I would like to thank you all for 4 being here to discuss with us this important topic, and, 5 again, many thanks to our panelists who have agreed to 6 be here with us. 7 introduce to you, to begin the first panel, David Wales, 8 who's the Deputy Director of the Bureau of Competition. 9 I will stop now, and I would like to (Applause.) 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 19 1 PANEL 1: 2 FOUNDATIONS OF UNILATERAL EFFECTS THEORIES: 3 CORE FEATURES, ECONOMIC BASES, 4 AND POTENTIAL GROUNDS FOR ATTACK 5 6 MR. WALES: Great. Thanks a lot, Debbie. 7 We are, to reiterate, very excited today about 8 our various panels, and I personally am very excited 9 about this panel. 10 participants and hopefully we will have some great 11 dialogue. 12 I think we have some great The way we would like to kick it off is just to 13 talk about some of the foundations of unilateral 14 effects, some of its core features, economic bases, and 15 potential grounds for attack, and other general topics 16 to set up some of the additional discussions that we 17 will have. 18 The format is going to work this way: Each of 19 the three -- now three -- panelists will have brief 20 presentations to talk about some of the issues they 21 think are important, that they want to convey, and then 22 what we would like to do is open it up to discussion, 23 hopefully get an active discussion as to some of these 24 issues and drill down a bit further on some of the key 25 points. For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 20 1 So, with that I am going to go ahead and give a 2 brief introduction of the panelists, and then I am going 3 to ask them to go ahead and start their presentations. 4 First off, we have, all the way down at the end, 5 Andrew Gavil. 6 University School of Law. He has been a member of the 7 Howard faculty since 1989. Prior to joining the 8 faculty, he practiced antitrust law and commercial 9 litigation with law firms in Chicago and Denver. He is 10 the lead author of Antitrust Law in Perspective: Cases, 11 Concepts and Problems in Competition Policy, and is 12 currently at work with the co-author, Professor Harry 13 First, on Microsoft and the Globalization of Competition 14 Policy: 15 received the Warren Rosmarin Award for Excellence in 16 Teaching and Service at the Law School and serves as a 17 faculty advisor to the Howard Law Journal. 18 Professor Gavil teaches law at Howard A Study in Antitrust Institutions. Next up we have Robert Willig. In 2004, he Professor Willig 19 teaches economics at Princeton University. He's a 20 former supervisor of economics research at Bell 21 Laboratories. 22 of Policies Affecting Prices and Products, and 23 Contestable Markets and the Theory of Industry 24 Structure, and co-editor of The Handbook of Industrial 25 Organization and Can Privatization Deliver? He is the co-author of Welfare Analysis For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 21 1 Infrastructure for Latin America, and numerous articles. 2 A fellow of the Econometric Society, he has served on 3 the editorial boards of the American Economic Review and 4 the Journal of Industrial Economics. 5 Antitrust Division in the U.S. Department of Justice as 6 Deputy Assistant Attorney General for Economics. 7 Finally we have Jan McDavid. He served in the She is a partner 8 at Hogan & Hartson here in D.C. She focuses primarily 9 on antitrust and trade regulation litigation and 10 counseling. 11 Antitrust Section of the American Bar Association, 12 including Chair. 13 Council of the U.S. Chamber of Commerce, and has served 14 on antitrust task forces with the U.S. Department of 15 Defense. 16 and articles involving antitrust, including the 17 Antitrust Evidence Handbook, Mergers & Acquisitions, and 18 Antitrust & Trade Associations Practice Guide. 19 Ms. McDavid's recognition includes The Best of the Best 20 Competition and Antitrust Section; Legal Times of 21 Washington Top Antitrust Lawyers; The International 22 Who's Who of Business Lawyers; and Guide to the World's 23 Leading Competition Lawyers. She has served in multiple positions of the She also is a member of the Antitrust She is the author or co-author of many books 24 We are thrilled to have each of you here today. 25 With that I think what we would like to do is For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 22 1 start off, Professor Gavil, with your presentation. 2 take it away. 3 PROFESSOR GAVIL: 4 MR. WALES: 5 PROFESSOR GAVIL: So, The slides? Yes. Good morning, everyone. I am 6 delighted to be here, and I thank Chairman Majoras and 7 Andrew for inviting me to join you. 8 To start off our first panel, I was asked to see 9 if in about five or seven minutes I could sum up the 10 history of unilateral effects. 11 that. 12 So, I will try and do I thought that in just a few slides I would talk 13 a little bit about the roots of unilateral effects 14 doctrine, both legal and economic, and how it fits into 15 the larger picture of merger analysis. 16 thinking about various phases we have gone through in 17 terms of merger enforcement analysis. That got me 18 I start with a hypothesis, and it was really 19 late last night when I typed this, so maybe it should 20 have a question mark at the end. 21 my hypothesis, so I will pose it more so as a 22 question -- a possible hypothesis. I am not sure this is 23 In a sense, unilateral effects is both the 24 oldest and the newest theory of anticompetitive harm for 25 mergers. The underlying legal and economic theories are For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 23 1 neither novel, nor new. 2 theory has certainly been refined; it has been 3 elaborated. 4 will talk a little bit about that, which have clearly 5 been aided by technology and there is increased access 6 to data, which also, aided by technology, has been very 7 significant. 8 mind, and as Chairman Majoras already put it for us, is 9 why has the contemporary theory of unilateral effects 10 proven to be such a difficult sell in the courts? 11 What is newer? Well, the There are new empirical techniques, and we But the question, of course, on everyone's The basic larger idea of merger to monopoly, of 12 course, is original to the Sherman Act. 13 quotation from Hans B. Thorelli, Federal Antitrust 14 Policy: 15 Here is a That "Sherman" -- talking here about John 16 Sherman -- "wanted the bill to cover the great 17 industrial trusts proper as well as mergers and other 18 tight combinations when of a monopolistic nature there 19 can be no doubt." 20 So the idea that we should prohibit mergers to 21 monopoly is a very old idea in antitrust. It was 22 supposed to be covered by the Sherman Act. In many of 23 the early merger cases that came out of the great merger 24 wave, Northern Securities, U.S. Steel, although of 25 varying success in terms of enforcement, the basic For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 24 1 theory was merger to monopoly, and the trusts themselves 2 were combines. They were viewed as mergers to monopoly. 3 The 1950 amendments ushered in the non- 4 monopolistic merger period, somewhat in response to the 5 Columbia Steel case of 1948, although there are other 6 factors as well. 7 these merger challenges from the twenties to the 8 forties. 9 a different set of concerns. 10 altered the focus from a focus on merger to monopoly to 11 what we might call nonmonopolistic mergers. The Government was losing a number of Congress decided to step in. They clearly had They broadened out and 12 We might also call these the wilderness years, 13 as the anchor, even in early thinking about merger to 14 monopoly, was a little bit more clear than what happened 15 in this period. 16 "trend towards concentration," a concept which is 17 typified by cases like Brown Shoe, Von's, and Pabst, and 18 which we now teach against in casebooks, toward the 19 structural approach, and the general concerns it raised 20 about market shares that were obviously elevating. 21 was the idea of making predictions from market structure 22 that took form in the Philadelphia National Bank 23 presumption, and, of course, was reflected in the first 24 Merger Guidelines in 1968. 25 There was an evolution from emphasis on Here From 1968 to 1992, there was an effort to better For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 25 1 define what the true anticompetitive theory was, and in 2 a sense this period led to a commingling and a fusion of 3 two competing traditions. 4 tradition going back to the 19th Century in economics, 5 and the other was the structural presumption, which had 6 developed in some of the writings on industrial 7 organization economics in the 1950s. One was the oligolopy 8 If you go back, as I did, looking at Stigler and 9 Posner and Bork and contrast them with Kaysen and Turner 10 (1959), you really see these two very different sets of 11 ideas competing for influence in terms of merger policy. 12 Their first offspring was the coordinated effects theory 13 in the 1982 Guidelines and the way the Guidelines are 14 structured. 15 Shapiro explore in their paper, I will mention that a 16 little later on. 17 that combine pieces of different theories I think is one 18 of the issues that is going to emerge today as 19 important. 20 are reflected in different pieces of the Guidelines, and 21 like a puzzle where the lines between the pieces are 22 still very defined, they do not always quite fit 23 together very well, and sometimes they can even work at 24 cross-purposes. 25 This is a point that Joe Farrell and Carl The attempt to structure Guidelines We have different intellectual thoughts that From the mid-1980s to the present, there was For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 26 1 something of a reintegration of the pre-1950 and 2 post-1950 models. 3 mergers are reintegrated in the Guidelines. Coordinated 4 and unilateral effects are both introduced. Both, of 5 course, have roots in oligopoly theory, but both are 6 still tethered to the structural concepts in the 7 Guidelines. 8 Monopolistic and nonmonopolistic For more sources on this history of unilateral 9 effects and its roots, I just cited a few of the 10 articles here on the slides, all of the authors being in 11 the room, Baker, Willig, and Denis, all go through some 12 of these issues of the intellectual roots of modern 13 unilateral theory. 14 Well, where do we go from here and what is the 15 discussion about today? I think one issue that I wanted 16 to put out is, how do we relate developments in 17 unilateral effects to the larger context of modern 18 antitrust? 19 I wanted to put the idea out there. 20 unilateral effects parallels, in a sense, the tension 21 that now exists in Section 1 between actual effects and 22 the quick-look doctrine on the one hand and 23 circumstantial effects under the Sherman Act. 24 Coordinated cases tend still to be structural in some 25 sense, economic, and more sophisticated in others. And this I am not quite sure I believe, but Coordinated versus For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 But 27 1 to the degree they are relying on creating a 2 circumstantial, predictive case for coordinated effects, 3 they are more like the circumstantial approach to merger 4 analysis. 5 I tried to give a new name -- I don't know if it 6 will work or stick -- but unilateral effects is more 7 akin to "predicting actual effects" based on empirical 8 evidence, and in that sense, it really can be located in 9 the circle with cases like NCAA and Indiana Federation 10 and California Dental and Polygram, cases that try to, 11 as the Chairman was talking about earlier, try to look 12 at actual effects and market definition, market power, 13 as flip sides of an issue. 14 As the court said in NCAA and again in Indiana 15 Federation, traditional market power analysis involved 16 defining a relevant market, calculating market shares, 17 and predicting market power and consequence 18 anticompetitive effects from large and durable shares. 19 The Court has held, however, that doing so was just a 20 surrogate for actual anticompetitive effects. 21 have the actual anticompetitive effects, you shouldn't 22 need to do those things. When you 23 The tension about that has arisen with respect 24 to such actual effects cases is similar to the tension 25 that exists now around unilateral effects. Concerns For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 28 1 about the reliability of actual effects evidence have 2 also caused some push-back in non-merger areas. 3 productive step we could take would be to get merger 4 analysis, instead of in its own pigeonhole, relocated in 5 the larger picture of what is happening in antitrust. 6 So, one The irony of precision -- last slide here -- why 7 are unilateral effects cases a tough sell in court? 8 economists, there is the appeal of empiricism. 9 very appealing. 10 quotation out of one of Jon Baker's articles: 11 For They are They -- based on data -- I pulled this "[i]f the facts support a unilateral theory, it 12 is clear as a matter of economic logic why the 13 particular merger would likely lead to higher prices." 14 This reminded me a little bit of the language in 15 Polygram where the FTC talked about anticompetitive 16 effects being "intuitively obvious" based on economic 17 analysis. 18 decision-makers? 19 But what is the challenge for Why the resistance? Well, in a sense, the models can be more complex 20 than the traditional PNB presumption. This is somewhat 21 ironic since the models were designed to yield a greater 22 degree of precision, a greater degree of understanding, 23 yet the models themselves are more complex. 24 presumption was by comparison easy, like per se rules, 25 like other burden-shifting devices. The PNB It did not require For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 29 1 a lot of understanding to say: 2 that's a lot!" 3 "40 plus 20 is 60. Ooh, Empirical evidence also may be confusing when 4 combined with traditional structural evidence. 5 appear highly dependent on assumptions, and, therefore, 6 subject to manipulation if the assumptions change. 7 can be a little bit more rigorous in theory than 8 practice. 9 And I think there is a larger issue, one that David 10 Meyer talked about in a speech last fall. 11 whether we like it or not, at something of a historical 12 moment in antitrust, where courts are proving very 13 skeptical about antitrust cases, and unilateral effects 14 has run into that skepticism as it tries to develop and 15 evolve in the courts. 16 It can It Sometimes the data do not match the theory. We are, Those are my opening comments, and I will turn 17 it back over to the panel. 18 MR. WALES: 19 Next we have Professor Willig with some brief 20 Great. Thanks, Professor Gavil. remarks. 21 PROFESSOR WILLIG: Brief? 22 I face an interesting challenge. I was asked to 23 cover the Merger Guidelines, a short overview to be 24 sure, unilateral effects therein, the history of 25 antitrust, and the economics of unilateral effects, and For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 30 1 I wasn't given five to seven; I was given three to five. 2 MR. WALES: 3 PROFESSOR WILLIG: 4 I lied. can talk really fast. 5 With another cup of coffee, I So, who's got the coffee for me? Elements of the Guidelines in an historical 6 context: 7 a relevant market? 8 algorithms, those of us who love that kind of thing, but 9 the idea of a relevant market is so simple that I think 10 we should remember its basic concept all day long 11 throughout the discussions. 12 collection of the principal sources of competitive 13 discipline on the products of the merging firms, 14 especially the overlapping products of the merging 15 firms. 16 First and foremost, relevant market. What is I know we talk about all the A relevant market is a If you collect all the sources of competitive 17 discipline and you put them all under a single source of 18 control, then you should be seeing some elevation of 19 monopoly power, and hence, the hypothetical monopoly 20 test as the way to make sure that you have got all of 21 the principal sources of competitive discipline 22 identified and collected in the relevant market. 23 idea of it is simple. 24 just the way to make sure that you have actually got 25 market power there collected in these various sources of The The hypothetical monopoly test is For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 31 1 competitive discipline. 2 This was the idea of the 1982 Guidelines, along 3 with a way of counting concentration within a relevant 4 market. 5 away from the technocratics, the Herfindahls and the 6 like -- remember when that was a bizarre thing? 7 remember that. 8 feel like it was yesterday and I was already old when 9 these things happened. 10 dream of mine. 11 great, exactly the way he looked -- God knows when. 12 improvement, but no change. 13 (Laughter.) 14 PROFESSOR WILLIG: The concentration question, again, taking it I I mean, I hate to be an historian and That is sort of a dangerous Never mind how Jon looks. He looks No So, why do we count 15 concentration and change in concentration? Well, a 16 relevant market is a place where a hypothetical 17 monopolist could or would exercise monopoly power. 18 change in concentration and the level asks, well, what 19 does the merger do to bring us to the status of that 20 hypothetical monopolist? 21 actually bring us to that hypothetical monopoly? 22 goes hand in glove with the idea of the relevant market. 23 The Herfindahl is a very clever way to measure 24 concentration. 25 collect share data and see how concentrated they are. The How close will the merger It It is nothing but an arithmetic way to For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 32 1 Now, everybody keeps saying today -- and I have 2 heard this before as well -- that the 1982 Guidelines 3 are all about collusion, coordinated effects, as we 4 would call it today. 5 there. 6 was being published as coming right out of a Cournot 7 model. 8 Waterson, and, in fact, Ordover and I were asked to 9 write a review of those '82 Guidelines. 10 consulting on the Division on them when they were being 11 written with Larry White, and in '83, Ordover and I 12 wrote, "Why do they keep using the word collusion in the 13 Guidelines? 14 models like Cournot with what we would call today 15 unilateral effects." 16 a lack of language, than a distortion of the ideas. 17 obviously did better a decade later by looking it in the 18 face, but to say that the '82 Guidelines were really 19 about collusion I think is a grave intellectual error if 20 we are doing history, and that was my assignment. Hey, I was there; Larry White was It turns out that the Herfindahl Index, by 1982, You all remember this, economists Cowling and I had done some They are actually talking about oligopoly I think it was more a mislabeling, We 21 Now we move on to the current Guidelines -- 22 hopefully still current -- and we have coordinated 23 effects, which we are not talking about today, and we 24 have unilateral effects, and I'd like to highlight three 25 different cases of unilateral effects that are squarely For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 33 1 in the Guidelines, and here, too, I am worried that we 2 are losing track about which one it is that we are 3 speaking of. 4 First of all, unilateral effects apply in the 5 Guidelines to the case of "homogeneous products," 6 commodities in the common parlance. 7 this a market in which firms are distinguished by their 8 capacities rather than by the characteristics of their 9 products, because they are all basically the same; 10 hence, homogeneous products. 11 totally good sense in a market of homogeneous products. 12 The economics of it are very simple. 13 The Guidelines call Unilateral effects make The idea is that if a firm gets bigger in a 14 space of homogeneous products, then it has got a bigger 15 base of capacity on which to enjoy a price rise, and so 16 a big merger tends to enhance the incentives of the 17 newly merged firm to cut back on output so as to push 18 the price up, because now, it has got more capacity on 19 which to enjoy the positive profit effects of that price 20 rise. 21 Not elaborate, not fancy, not about merger 22 simulation models, although we have lots of analytics to 23 handle that if we want to, but it is not what we are 24 usually talking about on a day like today, but it is 25 still unilateral effects. So, I think we need to For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 34 1 sharpen our language away from just unilateral effects 2 to unilateral effects in markets where the products are 3 importantly differentiated to separate out the case of 4 the commodities. 5 Second of all, within the category of 6 differentiated products, there is a main case really in 7 the Guidelines where the differentiated products are -- 8 I am calling it today generally differentiated. 9 Baker and I and Paul Denis debated this stuff for much 10 of two years together. 11 products are ones that compete with others in the 12 relevant market, but kind of generally, without any 13 specific product-to-product relationships. 14 Think about cold remedies. Jon Generally differentiated I mean, does anybody 15 really know what the subcategories are of cold remedies? 16 Everybody's got their favorites, and yet each cold 17 remedy basically competes with all the other ones. 18 Maybe a pharmacologist would know the difference, but we 19 consumers sure don't. 20 are all kind of mushed together in one big pot, no 21 specific competitive relationships. Or midsize cars, you know, they 22 Well, in a market like that, it makes sense to 23 think that the share of a product is indicative of its 24 competitive significance as an alternative to whatever 25 your favorite product is; that shares really connote For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 35 1 competitive significance, because the competitive 2 relationships are general throughout the marketplace. 3 That is the lead case of differentiated products under 4 the Guidelines, and there, relevant market makes just as 5 good sense as it does for a homogeneous product industry 6 that collects all these products that interact 7 importantly; concentration makes sense as a measure of 8 significance, and off we go. 9 There is a lot of economics lying behind this. 10 The Logit model of demand handles this. We all grew up 11 on the CES Utility model of monopolistic competition, 12 and in markets like that, this is exactly the kind of 13 interaction among the products. 14 differentiated products stuff. This is really classic 15 What we are all getting confused about is the 16 third case where the competition among differentiated 17 products is not general; instead, it is local, and where 18 differentiation is local, market share is not indicative 19 of competitive significance as a matter of substitution 20 for any other product. 21 products, no. 22 Some products yes; other Think about Toyota Camrys. They are very 23 successful cars, and yet they are in no way interesting 24 substitutes for the BMW drivers in the crowd. 25 maybe an Audi with a low market share is a much closer Instead, For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 36 1 source of substitution for the BMW than would be the 2 best-selling Camry. 3 are discernible; they are different; people recognize 4 them as such; and they drive the importance of different 5 substitution relationships. 6 of unilateral effects. 7 talking about the third one, and I think it would really 8 help to clarify that in our discussions. 9 So, here, product characteristics So, three different kinds Today, we are really only When we have localized effects, we are going to 10 have small, narrow relevant markets. You know, 11 Bimmer-oriented relevant markets instead of all cars or 12 all midsize cars, and what we are hearing is all judges 13 who I guess do not drive Bimmers find it a little bit 14 harder to understand. 15 A proposal I would make today -- and I am not 16 going to wait for the question, I just want to slip it 17 in -- the proposal is that we accept the idea that 18 markets can be narrow where competition is localized -- 19 bite that bullet -- and accept the idea that sometimes 20 the best evidence for what constitutes the true, narrow 21 relevant market is not our normal kind of intuition 22 about, "Oh, a car is a car; a grocery store is a grocery 23 store; a stationery story is a stationery store," but we 24 allow ourselves, where appropriate and where the 25 evidence is there, to deduce market definition from For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 37 1 evidence about competitive effects; that after we look 2 at the Staples/Office Depot evidence, that where there 3 are only two superstores instead of three, prices are 4 higher, that teaches us that the office superstores are 5 not in the same relevant market as your corner 6 drugstore, which I would have thought intuitively, but 7 the evidence proves that is not true. 8 proves that, indeed, the relevant market is office 9 superstores. 10 sources of evidence, but the statistics that show that 11 are our best evidence for market definition. 12 The evidence I wouldn't have known that through other Why shouldn't we allow markets to be defined 13 using best evidence? 14 kinds of data, that would be our best evidence. 15 not that markets are irrelevant. 16 should be willing to test them and to prove them, 17 sometimes using the same kind of information that we use 18 for competitive effects, where we have such solid 19 evidence. 20 And in cases where we have those It is It is just that we It is not wrong in Whole Foods for the judge to 21 be debating what the relevant market is -- all 22 supermarkets or just organically oriented ones. 23 very much the right question, and I think the judge was 24 on the right beam in trying to figure out what the best 25 source of persuasive evidence was. That is I don't know what For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 38 1 the right answer is either. 2 would have been all over the lot just like the judge 3 was. 4 hard question. 5 aware of those data, but, I mean, maybe you are right. 6 But I think the judge was grappling with the right 7 question, and why not allow competitive effects and 8 natural experiments to be part of the evidence that does 9 drive a determination of the relevant market, along with 10 competitive effects? 11 with that. 12 So, if I were the judge, I I don't know if it was a wrong process. Maybe the FTC knows better. It is a I am not I think there is nothing wrong I think there is a danger in eliminating the 13 idea of a relevant market, because not forcing ourselves 14 to actually enumerate, out loud, all the sources of 15 important competitive discipline creates the danger that 16 in our weaker moments, when we are not absolutely on our 17 game -- and I know mostly we are in this room, but 18 sometimes we are off our game -- when you are on the 19 other side of me, for example -- that under those 20 circumstances, you should be impelled by the process to 21 enumerate all of what you think are the important 22 sources of competitive discipline, and the process of 23 relevant market is the force that makes us do that. 24 Just saying, "Oh, it is obvious that these two products 25 are the closest substitutes, end of story," is a For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 39 1 dangerous way to lead our process as far as the law is 2 concerned. 3 Thank you. 4 MR. WALES: 5 We are now going to turn to Jan McDavid with her Thank you, Professor. 6 opening statement. 7 Jan? 8 MS. McDAVID: 9 In recent years, as we have been talking about, Thanks, David. 10 the agencies have increasingly relied on unilateral 11 effects theories. 12 economists in the room, can tell us whether the 13 techniques underlying these theories are appropriate and 14 debate which theory is appropriate in a particular case. 15 I am not an economist; I don't play one on television. 16 I hire people like Bobby for that. Other panelists, and especially the 17 Instead, I'd like to discuss these issues from 18 the perspective of an antitrust practitioner who has to 19 explain them to business people who are making decisions 20 about potential transactions and who interact with the 21 staff of the agency about particular transactions. 22 Now, it has always seemed logical to me to 23 consider whether a merger that eliminates direct 24 competition between the merging parties substantially 25 reduces overall competition within the meaning of For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 40 1 Section 7. 2 very common sense notion that a merger is likely to have 3 more a harmful competitive effect if the merging parties 4 are particularly close competitors. 5 Unilateral effects analysis is based on the The most obvious example, of course, is a merger 6 to monopoly in which there is no competition remaining 7 following a transaction. 8 transactions in which some rivals remain could produce 9 those competitive effects. 10 won't. 11 before the courts, is how do you distinguish between all 12 of these different formulations? 13 But it also seems logical that In other circumstances, they The question before us, before the agencies and I have always found that the easiest way to 14 explain these concepts to business people is the next 15 best substitutes formulation, and so that is basically 16 what I have done. 17 Now, as a Colorado skier, I often use the Vail 18 case as the paradigm that I walk my clients through in 19 trying to have them understand competitive effects. 20 About ten years ago, Vail resorts, which operates both 21 Vail and Beaver Creek, proposed to acquire the Ralston 22 resort ski properties in Colorado. 23 thought Ralston only made dog food will be surprised to 24 know that they actually operated Breckinridge, Arapahoe 25 Basin, and Keystone, and did not do so especially well. Those of you who For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 41 1 The Division concluded that there were two kinds 2 of skiers: 3 get on an airplane and fly somewhere to ski, and if 4 prices go up for us, we could go somewhere else. 5 could get on an airplane to Salt Lake rather than to 6 Denver if I wanted to go skiing. 7 what they called the front-range skiers, the folks who 8 get in their cars somewhere in the Denver metropolitan 9 area and drive about two-and-a-half hours to a ski area, 10 and they concluded that that was the market in which 11 they needed to analyze the effects of the proposed 12 Vail-Ralston transaction. 13 There were destination skiers, like me, who I And then there were The competitive impact statement made it clear 14 that the Division was applying a unilateral effects 15 theory to the case. 16 deterred from increasing its prices at Vail and Beaver 17 Creek by the fact that skiers could go to Keystone 18 instead, if prices were to be increased at Vail and 19 Beaver Creek, or Breckinridge or Arapahoe Basin. 20 Vail also owned Keystone, Breckinridge, and A-Basin, 21 they would also pick up the revenues on the sales of 22 those tickets, and therefore, a price increase might 23 become profitable. Before the merger, Vail was But if 24 Based on an econometric analysis, using largely 25 survey data -- and that is a point I really do want to For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 42 1 come back to -- and data on margins, the Division 2 concluded that a price increase of a dollar per ticket 3 was likely in the event that Vail owned Vail, Beaver 4 Creek, and Keystone, because Keystone was the next best 5 substitute. 6 would fix this problem. They also concluded that divesting A-Basin 7 Now, the antitrust agencies' ability to engage 8 in the type of analysis that they used in the Vail case 9 or in the other cases we have been talking about has 10 been made possible by the kinds of rich data sources 11 that are available, as well as computers. 12 involving branded food products, for example, IRI and 13 Nielsen data permit very elaborate econometric models in 14 which we can actually use transaction data to test these 15 propositions. 16 in branded food products are not available most of the 17 time, and even in branded food product transactions, 18 they actually focus on competition at the wrong level, 19 because they are focusing on the prices set by 20 retailers, not the prices set by the manufacturers of 21 the food products who are actually engaged in the 22 merger. 23 In cases But the retail scanner data that we have So, what substitutes for these kind of data are 24 available and how does the quality of the data affect 25 the quality of the analysis in which we are engaging? For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 43 1 It seems logical to me that differences in the quality 2 of the data are very likely to lead to differences in 3 the quality of the economic analysis that is being done 4 and that use of data that is not reliable may lead to 5 skewed and unreliable results. 6 An awful lot of the debate is also about the 7 kinds of assumptions that are being used, and if you 8 vary the assumptions, you vary the outcome. 9 possible, under the Guidelines and under the Commentary, 10 to find unilateral effects at even low market shares. 11 Many of us believed there was a 35 percent safe harbor 12 in the Guidelines, but the Commentary says there isn't. 13 Where is the right line? 14 effects predicts some kind of a price increase absent 15 some significant efficiencies. We all know how reliable 16 the efficiency estimates are. All of this can skew the 17 outcome in ways that may render the results at least 18 suspicious and make people skeptical. 19 It is very Every model of unilateral Now, I bring to this process the skepticism that 20 I also bring to the HHI analysis. The HHIs lead to a 21 mathematical result which looks precise on its face, but 22 we all know that it varies entirely based on market 23 definition and market shares, neither of which are very 24 reliable, and then you just square it and add it up. 25 So, it all depends on where you start as to where you For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 44 1 end up. 2 For that reason, I rarely try to define markets 3 in the transactions I am working on. 4 almost immediately, on competitive effects analysis, 5 because that is where I have always thought the game was 6 going to be played. 7 are a very useful first screen for thinking about the 8 transactions into which we should start conducting that 9 kind of elaborate analysis, but they create an 10 artificial sense of precision where no real precision is 11 possible, and I am concerned that some of the same 12 things happen with respect to the kinds of unilateral 13 effects analyses that we have been undertaking. 14 I always zero in, I have always thought that the HHIs Let's go back to the Vail case as an example. 15 People who ski in Colorado who probably agree that 16 Keystone was the most likely next best substitute for 17 Vail and Beaver Creek, with Breckinridge being a close 18 second. 19 survey data would allow you to conclude that prices 20 would go up one dollar or we would be especially 21 skeptical that divesting Arapahoe Basin was going to fix 22 that problem. 23 I think we would have been very skeptical that I have never skied at Arapahoe Basin. It is way 24 too hard for me. There are people there who sleep with 25 their dogs in their Volkswagen buses in the parking lot. For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 45 1 It is not a substitute for Vail. 2 an economic model that suggests that it is. 3 So why should we trust So, I think where all of this takes me is that 4 we have to bring some common sense to these kinds of 5 analyses, and that is where I am concerned that the 6 agencies are running into resistance. 7 they've been doing appears to be gerrymandered or 8 jury-rigged and doesn't pass the common sense test. 9 When your judge is someone who's been sentencing drug 10 offenders in the morning and is handling unilateral 11 effects analysis in the afternoon, you have to be 12 conscious of the limitations of your audience. 13 don't do the math either. 14 Some of what They Judge Wood, who handled the cereals transaction, 15 brought Fred Kahn in to advise her as effectively her 16 law clerk when she tried that case, even though she was 17 a very experienced antitrust lawyer and very good at the 18 economics. 19 sorts of problems that we have to be conscious of. And that is, I think, an illustration of the 20 So, I would like to use the unilateral effects 21 analysis as part of a holistic analysis of all of the 22 evidence. 23 results with the more traditional models, considering 24 the company's strategic planning documents; who do they 25 think are their most significant rivals; what do the I have always thought we get to pretty good For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 46 1 customers say; what is the evidence about entry; is one 2 of the companies failing; is one of the company's 3 ability to compete on an ongoing basis impaired in the 4 future. 5 the evidence consistent? 6 place? Does this tell us an overall story? Is all of Does it take you to the same 7 If that is the case, I think you can be 8 reasonably confident about the kind of decision you are 9 reaching. 10 skeptical, and the agencies will encounter a skeptical 11 audience in a federal judge. 12 lessons are things we have to keep in mind as we do 13 these sorts of analyses. 14 MR. WALES: 15 We will kick things off a little bit. If it does not, then the agencies should be I think those kinds of Thanks, Jan. I thought 16 I would ask some questions and hopefully get the 17 dialogue going. 18 It seems that there is not a lot of dispute that 19 unilateral effects is a valid theory and one that we 20 think should be applied in the appropriate cases, 21 especially in differentiated product merger cases, but 22 the reality is it has been a tough sell to judges, and I 23 guess the question is, what do we take from that? 24 are the reasons why we think that judges are having a 25 hard time? Is it the fact that perhaps unilateral For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 What 47 1 effects is not a sound theory? Is it more practical in 2 the sense that there are assumptions, intuitive 3 problems? 4 think the problems are? 5 PROFESSOR GAVIL: Are the Guidelines to blame? What do you Well, the Guidelines are a 6 product of a long history and tradition, and again, I 7 would say that you need to look at it in the larger 8 context of antitrust. 9 relevant markets and market definition and market shares 10 and assumptions that you draw from that, connections 11 between that and the possibilities of anticompetitive 12 effects, for a long time. 13 going to be an easy process, and the evidence is going 14 to have to be especially compelling. 15 We have been thinking about So, shaking that loose is not I think if something does differentiate Staples, 16 it is that the evidence was especially compelling. 17 is difficult from the outside to evaluate how compelling 18 the evidence is in cases still pending, like Whole 19 Foods, where we just don't know all of the evidence that 20 was introduced. 21 It And I think a second part of it is Bobby's 22 comment that maybe we shouldn't be trying to persuade 23 anyone to totally let go of that structural tradition. 24 I combine that with Jan's comment -- this has been true 25 in nonmerger cases -- when the two kinds of evidence are For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 48 1 pointing in the same direction, you are going to have 2 the strongest case. 3 Now, that means a lot of work maybe, but when 4 the direct and circumstantial evidence in non- merger 5 cases is pointing towards market power, those cases are 6 pretty hard to rebut. 7 combination of thoughts here that lead to that 8 conclusion. 9 MS. McDAVID: So, maybe there is this sort of I think one of the things about 10 Staples we should remember is that although we had very 11 complicated economic analysis by Professor Ashenfelter, 12 there was also some really simple stuff. 13 higher where there was one firm and prices were higher 14 where there were two firms than they were when there 15 were three. 16 people who don't do the math. 17 Prices were That was a pretty simple paradigm for even PROFESSOR WILLIG: It seems to me that the basic 18 thought behind differentiated products or local 19 competitive effects, the basic thought is totally 20 intuitive. 21 breakfast table test at home, which is to say that, 22 look, it turns out that when my favorite car is being 23 priced by the marketing people, the first thing they 24 look to is this closely competing car, and maybe we 25 actually have evidence from the companies of that or I mean, it passes my dinner table, even my For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 49 1 maybe we can readily deduce that, but as an expert, that 2 would be my opening line if I am talking to my family or 3 to a common lay judge, is to say, look, what is keeping 4 prices where they are today is largely and importantly 5 competition with this other product, and guess what, 6 after the merger, that product will be in the same 7 executive suite, the margin will be just going into the 8 same pocket as the margin on the BMW, my favorite car, 9 and so that source of price competition will be gone. 10 Now, Your Honor, believe me, I have looked at 11 other possible sources of competition, and there are 12 other ones, but they are just nowhere near as important 13 to the pricing of the BMW as that Audi car, and now Audi 14 and BMW are threatening to merge. 15 a broader relevant market, I have tabulated all the 16 other possible sources of competition, and they do have 17 some effect, but not nearly as important as the effect 18 that would be lost because of this merger. 19 about that? 20 Jan? 21 MS. McDAVID: So, I have looked at What is hard No, I think that is pretty simple, 22 Bobby. By the way, I have always thought that the 23 Division's case in Oracle made a great deal of sense. 24 The problem was that the market, as defined, was not 25 really a product the company sold. It, therefore, For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 50 1 looked jury-rigged, and I think that is just part of the 2 problem. 3 It failed the common sense test. MR. WALES: What about one of the -- I guess in 4 the merger commentaries it talks about the fact that you 5 can have both quantitative and qualitative evidence that 6 may be probative of the closeness of substitution of the 7 various products and, of course, the potential 8 competitive effect. 9 Is it the case now that you must have 10 quantitative evidence, despite the fact that the 11 commentaries talk about how you can have either 12 quantitative or qualitative information, like business 13 documents? 14 judge was more focused on the quantitative as opposed to 15 the qualitative evidence, where there was some pretty 16 good qualitative evidence in the business documents. 17 Obviously in Whole Foods, it seemed like the MS. McDAVID: We have to do both. The reality 18 is when we are proposing a transaction, we have to do 19 both. 20 the matters that I handle before the agencies, I 21 encourage my economists to share all of their data, all 22 of their analyses, almost sit in a room with the agency 23 economist and be as cooperative as possible. 24 get to the right kinds of outcomes. 25 what we did in the cruise lines case, and many people There is no alternative, and, you know, in all of We will That is absolutely For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 51 1 hold that out as the model. 2 agency's going to do it. 3 destruction circumstance. 4 PROFESSOR WILLIG: We have to do it. The It is a mutually assured I mean, to me, the 5 quantification, aside from our satisfaction in using 6 professional standards as economists, but the 7 substantive question that has to be addressed -- and 8 this brings us back to relevant market, I think -- is 9 suppose that we can all agree, intuitively, that B is 10 the closest substitute for A, and A would be the sellers 11 are threatening to merge, but that really is not the end 12 of the story, nor is it even the end of the story to say 13 how closely substitutable A and B are, because in many, 14 many local or bigger markets, there is a C, D, and E 15 lurking behind A and B. 16 Those of you who know Princeton, if you get off 17 Route 1 to make a right turn to come to the campus down 18 Washington Road, there is a little traffic circle, and 19 on that traffic circle there is two gas stations, and 20 they are head-to-head competitors. 21 literally head to head on the traffic circle. 22 always use this in class. 23 stations merge? 24 they are close substitutes, so wouldn't you bust the 25 merger right away? I mean, they are So, I What if those two gas What do you say, class? You can see For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 52 1 So, anybody who says yes never makes it to the 2 midterm as far as I am concerned. 3 Half a mile down Route 1, there are five other gas 4 stations. 5 merge, we would lose that head-to-head competition, but 6 it would not be a substantial or it might not be a 7 substantial change in the state of competition, because 8 there is all these other gas stations just a half a mile 9 down the road. 10 But you know what? Now, it is true if those two gas stations This is what scares me about getting rid of 11 relevant market when it comes to localized competition 12 among differentiated products. 13 say, right away, "No, no, we have got to stop that 14 merger," without asking what else is there right behind 15 that pair of closest substitutes? 16 question that the relevant market forces us to answer, 17 to pick it up, saying, "Well, yeah, there are other 18 sources of competition, but you know what, they are not 19 nearly as important." 20 Half of my class will And that is the But we need some quantification to get us to the 21 ability to conclude whether or not those other gas 22 stations are closely enough competitive to these two 23 that are head-on to see whether their merger will 24 significantly tend to raise price, or whether, instead, 25 C, D, and E will provide ample competitive discipline to For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 53 1 stop there from being a significant price increase 2 because of the merger. 3 quantification is necessary. 4 That means some kind of When I tell you half a mile, you know the 5 answer, but when we are talking about cold remedies or 6 supermarkets of different kinds, we have no ready such 7 quantification, and now we are into a real debate that 8 is frustrating a lot of people. 9 MS. McDAVID: I do not think it matters what you 10 call it -- or whether you focus on relevant market or 11 market shares, what you have to determine are what are 12 the -- as Bobby put it -- the sources of competitive 13 discipline post-transaction on the merging parties? 14 you are going to have to identify them and talk about 15 how significant they are. 16 MR. WALES: And It seems that judges have had a hard 17 time, though, in terms of applying the Guidelines and 18 understanding the difference between identifying that 19 localized competition that we think matters in terms of 20 the unique constraint on the merging, differentiated 21 products, and defining a broader market that might 22 contain more distant competitive constraints. 23 need to rethink how the Guidelines work in 24 differentiated product cases? 25 MS. McDAVID: Do we Well, the Commentary made an For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 54 1 effort to do that, but we are regularly reminded that 2 the Guidelines are not law. 3 reminded us of that in his Whole Foods opinion. 4 Guidelines are sources of explanation and an 5 extraordinarily useful framework for us to use before 6 the agencies, but fundamentally, they are not going to 7 bind a court. 8 think is what you really need. 9 10 11 I think Judge Friedman So, the Some explanation, in whatever format, I MR. WALES: Would anyone support amending the Guidelines? PROFESSOR GAVIL: The Guidelines have become 12 kind of a two-edged sword I think for the agencies. 13 Yes, formally, they are not law. 14 all state -- not only the Merger Guidelines, but all of 15 the enforcement agency guidelines -- that they are not 16 intended to establish a litigation format; they do not 17 specify burdens of proof. 18 agencies use them in courts, the degree to which parties 19 use them and hold the agencies to them, means that they 20 have become very influential documents in court. 21 are looked to as demarking lines for burden-shifting 22 when you look at the steps of the Guidelines. 23 Guidelines, on their face, would seem to suggest that 24 you always start by defining a relevant market and 25 calculating market shares. Yes, formally, they But the degree to which the They And the For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 55 1 So, when you say change the Guidelines, ask 2 should we change the Guidelines, well, to what end and 3 with what consequences? 4 difficult challenge for the agencies to articulate 5 enforcement standards to two communities. 6 articulating to the business community their intentions 7 with respect to enforcement efforts, but then when they 8 go to court, in part, given the Supreme Court's absence 9 from mergers for so long, when they go to court, they 10 are kind of trying to use the cases that are available, 11 that are the best cases. 12 Guidelines as if it were law, as if it were their own 13 law. I think it has become a They are Yet they have to live with the 14 So, it is a challenging question, what to do 15 with the Guidelines, and can you fix the problem in 16 court by changing the Guidelines, by further developing 17 the theories? 18 said, when those first '82 Guidelines came out with HHIs 19 and SSNIP, you know, there was giggling in the room at 20 the ABA meeting -- "what could this be and what court 21 would ever do this?" 22 changed. 23 Maybe. Coming back to something Bobby And with time, that has clearly So, maybe part of the answer is that changing 24 the Guidelines could change things, but it may not 25 change things in the next case or it may take some time For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 56 1 until those ideas filter through and gain the confidence 2 of lawyers and judges as well. 3 PROFESSOR WILLIG: Well, let me ask, just to 4 pose my own question, if you were to think with me that 5 the best way to go is to deliver the message that the 6 way to determine relevant market is through best 7 evidence, which sometimes may be consumer survey -- God 8 help us -- sometimes through your own stomach as a 9 consumer, but sometimes through real consideration of 10 marketing data or natural experiments, like in office 11 products case, get the message out that we do need to 12 determine relevant markets, but we can sometimes do it 13 backwards. 14 we would do for competitive effects but use that as the 15 source of best evidence for relevant market. 16 Sometimes we can do the same analysis that What is the best way to get that message out? 17 Is it a revision of the Guidelines? 18 it next time there is a document that talks about best 19 practices, that that becomes a prominent example? 20 in court explicitly that way? 21 courts better than I, what is the best way to deliver a 22 message of that kind? 23 MR. WALES: Is it a speech? Is Do it Those of you who know One additional point, is the 35 24 percent threshold in the Guidelines. We have seen some 25 courts reject that, actually in Oracle, there were some For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 57 1 pretty negative comments on it; other courts were 2 willing to accept it as another proxy in the attempt to 3 measure the closeness of substitution between the 4 merging products. 5 The Merger Commentaries talk about it as merely 6 a screen and not a safe harbor. 7 place in antitrust cases? 8 that something we should consider changing? 9 MS. McDAVID: Does it still have a Should we be using it? Is Well, if you go back to my common 10 sense notion, when the agencies challenge a transaction 11 where the market shares are below 35 percent, it 12 suggests that there are a number of rivals that really 13 matter out there. 14 lot of skepticism about a challenge under those 15 circumstances. 16 compelling case about why the other 65 percent is not 17 sufficient to constrain the exercise of market power in 18 that circumstance. 19 I think that you are going to find a You are going to have to have a pretty PROFESSOR WILLIG: I think it is a form of 20 prosecutorial discipline, because it does force the 21 agency to articulate a narrow enough relevant market to 22 get past the 35 percent threshold and to confess that, 23 indeed, we are talking about localized competition, that 24 is the theory of the case. 25 is articulated, that is what is driving the bringing of No matter how explicitly it For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 58 1 the case -- maybe appropriately, there is no doubt about 2 it -- but then the relevant market has to be articulated 3 as a narrow one, and then the 35 percent threshold will 4 be met easily. 5 The question is, will the court find that narrow 6 market to be credible? 7 be credible. 8 court is weighing in from a lay point of view. 9 And if not, maybe it shouldn't It really is a matter of judgment, and the MS. McDAVID: Think back to the Grinnell case 10 where the Court talked about the market definition as a 11 red-haired, green-eyed man with the limp. 12 that the kind of thing you want to argue to a judge who 13 is going to be viewing this through his or her prism, 14 which may or may not include an economics background? 15 16 PROFESSOR WILLIG: I mean, is Or maybe the judge will like to sleep in a van with the dogs and go skiing. 17 MS. McDAVID: Exactly. 18 PROFESSOR GAVIL: One thought just to add here 19 is I think safe harbors are important. And I think that 20 not all market definition is going to be rocket science. 21 And the challenge is, if you have got a market 22 definition that does require more data, that is one that 23 is a little bit more complex, stating safe harbors can 24 suggest a false level of certainty -- using a safe 25 harbor that is based on a numerical threshold suggests a For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 59 1 degree of precision that may not be there with more 2 ambiguous markets. 3 terms of the less rocket science market definition, so 4 transactions can be identified that just are not going 5 to be on the table. 6 do not know, but the concept of having some easily 7 discernible area of safe behavior is an important one in 8 enforcement. 9 of antitrust enforcement. 10 But it does give some guidance in Whether that is the right number, I We talk about it again in all other areas MS. McDAVID: The cruise lines case is an 11 interesting example of market definition, because the 12 Commission's statement defined a market limited to 13 cruise lines, but then it became really clear that in a 14 competitive effects analysis, the exercise of market 15 power would be constrained by other vacation choices. 16 Therefore, we focused on competitive effects, which is 17 where I think the game really needs to be played. 18 MR. WALES: Okay, put your agency hats on. You 19 are back at the agencies. 20 the agency be looking for in terms of good unilateral 21 effects cases? 22 circumstances you think necessary, perhaps even 23 including some of the most recent cases -- were they 24 ones we should have brought? 25 agency be focusing on? What types of matters should What are the specific factual Which ones should the Obviously merger to monopoly is For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 60 1 the easiest, but I do not think anyone's going to say 2 that is all we should be looking at. 3 MS. McDAVID: I would go back to circumstances 4 in which the evidence aligns, where the economic 5 evidence is consistent with the parties' internal 6 strategic planning documents. 7 strategic planning documents as a first screen. 8 particularly focus on one another, that may be an 9 indication of next best substitutes, and, therefore, a 10 transaction should be subject to additional analysis. 11 But I'd use a combination of all of the evidence and be 12 sure it points in the same direction. 13 PROFESSOR WILLIG: You can almost use their If they Yeah, Jan, we have both seen 14 an awful lot of collections of business documents where 15 a company is very fond of naming one competitor over and 16 over again strategically and where the sum total of the 17 competitive forces from all the others, on analysis, 18 turns out to be every bit as important. 19 MS. McDAVID: I said first screen. 20 PROFESSOR WILLIG: 21 MS. McDAVID: 22 PROFESSOR WILLIG: 23 MS. McDAVID: Yeah. First screen. But caution to that. Of course. It has got to be the 24 whole collection of all evidence, not just the strategic 25 planning documents, but including the views of the For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 61 1 customers, evidence of recent entry, the competitive 2 problems the particular firms face, the whole array of 3 evidence. 4 PROFESSOR GAVIL: I think we have come to a 5 point where there is something of a paradox that makes 6 the question hard to answer. 7 need to bring the best case; the Government needs a win. 8 It is easy to say that. 9 to say that, well, all the evidence ought to be pointing 10 in the same direction. 11 It is easy to say they And it is relatively easy, too, Here is the reason I think it is somewhat 12 paradoxical. The blatant merger to monopoly, like the 13 blatant cartel, is not going to happen, presumably, very 14 often. 15 going to be harder cases. 16 to be represented by people like Jan, who are making the 17 best possible arguments with the best possible 18 economists about why a particular transaction should be 19 permitted. 20 the general skepticism of the courts about antitrust 21 now, means there are not going to be any easy cases. 22 is going to be hard to choose the best case. The cases that are going to be presented are The merging firms are going So, I think, in a sense, that, combined with It 23 It's not to say that people do not still propose 24 extreme things and that that may come along and you may 25 get lucky and have a fish in the barrel to shoot, but I For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 62 1 think that we are more likely to be facing complex fact 2 patterns, complex economics, and close calls, and it may 3 have more to do, in terms of winning, with the luck of 4 the draw in which judge you get and how that judge 5 reacts to the package of evidence than all that much 6 that the agency can do or the parties can do. 7 going to be tough cases. 8 of areas of antitrust. 9 PROFESSOR WILLIG: Those are That is where we are in a lot And, of course, don't forget 10 that how tough the cases are is, in a way, a testament 11 to the remaining credibility of the agencies, because 12 the cases that would be easy do not get to court. 13 the ones that are left to go to court are the really 14 hard ones, inevitably, and that is still true, despite 15 the somewhat checkered record of the agencies in courts 16 lately, and that is a testament to the lasting view of 17 this marketplace of the skills and the abilities of the 18 agencies. 19 So, So, look on the bright side. MR. WALES: I think there has been a lot of talk 20 lately about the general skepticism about antitrust. 21 That skepticism is something that we feel more generally 22 in terms of talking to judges and others. 23 How do we deal with that? How do we reduce that 24 skepticism and somehow renew the interest in strong 25 antitrust enforcement? For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 63 1 MS. McDAVID: It is a forensic exercise. It's 2 got to be. 3 educated about the concepts that underlie some of this. 4 The Antitrust Bar tries to do a good bit of that, and we 5 do supply copies of Antitrust Law Developments. 6 And I think the bench is becoming better PROFESSOR GAVIL: The only thing I would add 7 here is, again, I think context is important. 8 to get narrowly focused on our little corner of the 9 world in antitrust. 10 antitrust cases. 11 expensive process. 12 our antitrust case -- I am working on a symposium at 13 Howard on the history of Conley and Twombly -- and 14 Conley, in 1957, 50 years ago, was a civil rights case. 15 The five lawyers working on the case were all 16 African-American. 17 the nut of getting at intent to discriminate by a union 18 that was complicit in employer discrimination, and in 19 that context, at that moment in time, the court said, 20 "lower the pleading barrier, these cases have to go 21 forward." 22 civil litigation for 50 years. 23 We tend Judges are not skeptical just about Litigation has become a costly and Twombly, which we think of as They were basically trying to crack That became the standard that we used in all And then if you had to imagine what would be the 24 antithesis of that case, Twombly was potentially the 25 antithesis of that case -- a nationwide class action For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 64 1 involving potentially hundreds of millions of consumers 2 against all of the leading telecommunications companies, 3 and the court recoiled from Conley in that case. 4 Now, partly, that is a challenge of using the 5 same procedural standards in every kind of case that we 6 do, but what does that mean? 7 litigation system today with over a quarter of a million 8 cases filed each year in the federal courts. 9 of cases; a lot of them are complex; habeus can be just 10 as complex for a judge as antitrust; and there is 11 generally resistance to litigation. 12 looking outside antitrust is helpful in locating 13 ourselves in the larger world of federal court 14 litigation. 15 PROFESSOR WILLIG: It means that we have a It's a lot So, again, I think Do you think the public who 16 forms these troubling views, including the judges, 17 distinguishes adequately enough between cases brought by 18 the United States, by the FTC, and cases brought by the 19 adventuresome private bar? 20 I mean, maybe some of the bad rap that antitrust 21 has is because of the activist plaintiff's bar. It 22 could be. 23 variable in their superficial and end validity than are 24 the cases brought by the agencies. 25 PROFESSOR GAVIL: I think on average those cases are far more Bobby, I think it is a good For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 65 1 point. 2 extent the agencies have fed the fires of hostility to 3 private actions, the courts' hostility to antitrust is 4 coming back and constraining the agencies as well. 5 One of the ironies, though, is that to the But yes, clearly, if you look at the Supreme 6 Court decisions of the last two terms, there is a lot of 7 anti-private action rhetoric going on, and some of it 8 was coming from the government agencies that were 9 encouraging that view, and it came back to bite them in 10 a case like Credit Suisse, for example. 11 MS. McDAVID: 12 truth in that. 13 Trinko. 14 15 Certainly it was driving Twombly and MR. WALES: today. Okay, I'd like to thank our panel We had an excellent discussion. 16 (Applause.) 17 MR. WALES: 18 break. 19 very much. 20 I think there is a good bit of The plan is to take a 15-minute So, let's be back at 10:35, if we could. (A brief recess was taken.) 21 22 23 24 25 For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 Thanks 66 1 PANEL 2: 2 THE ROLE OF MARKET DEFINITION IN 3 UNILATERAL EFFECTS ANALYSIS AND 4 IN THE LITIGATION OF UNILATERAL EFFECTS CASES 5 6 MR. SCHMIDT: The next panel is going to focus 7 on the role of market definition in unilateral effects 8 analysis. 9 panel that it is difficult to separate these panel 10 discussions so that they do not overlap at all, but our 11 focus is going to be on the requirement or the lack of 12 requirement to prove a relevant product market and the 13 various implications of that. 14 I think you have already seen from the first We have a terrific panel to focus on that issue 15 with us today, and let me just take a minute to go 16 through the introductions, and then we will start right 17 in. 18 To my far left, Jon Baker. Jon is a Professor 19 of Law at American University's Washington College of 20 Law, where he teaches courses primarily in the areas of 21 antitrust and economic regulation. 22 senior consultant with CRA International. 23 experience includes being the Director of the Bureau of 24 Economics -- we won't hold that against him -- at the 25 Federal Trade Commission, Senior Economist -- sorry, Professor Baker is a His previous For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 67 1 Mike, wherever Mike is -- Senior Economist at the 2 President's Council of Economic Advisors, Special 3 Assistant to the Deputy Assistant Attorney General in 4 the Antitrust Division, and Assistant Professor at 5 Dartmouth's School of Business Administration. 6 sure you know, Jon is co-author of an antitrust case 7 book and past editorial chair of the Antitrust Law 8 Journal and a past member of the Council of the ABA 9 Antitrust Section, and in 2004, he received American 10 University's Faculty Award for Outstanding Scholarship, 11 Research, and Other Professional Accomplishments, and in 12 1998, he received the FTC's Award for Distinguished 13 Service. 14 To my immediate left is Kathy Fenton. As I am Kathy is 15 a partner at Jones Day. She's practiced antitrust law 16 for more than 25 years. She is currently the Chair of 17 the Antitrust Section of the ABA and has served in 18 numerous positions, including editorial chair, of the 19 Antitrust Law Journal. 20 professional service committee and served as chair of 21 the ethics subcommittee. 22 on issues of professional responsibility, conflicts of 23 interest, and legal ethics, including serving as an 24 instructor on legal ethics for the D.C. Bar's new 25 admittees course. She is a member of Jones Day's She has written and lectured Her recognitions include Who's Who in For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 68 1 American Law, The Best Lawyers in America, 2007. 2 previously served as an Attorney Advisor to the Chairman 3 of the FTC and was a law clerk here in the District of 4 Columbia, the District Court. 5 She To my far right is Dan Wall, partner at Latham & 6 Watkins. 7 Competition Practice Group. 8 has been active in the Antitrust Section of the ABA, 9 also. 10 of the Antitrust magazine; was chair of both the 11 Computer Industry Committee and Sports and Entertainment 12 Industry Committee; organized and chaired The Stanford 13 Conference on Antitrust in the Technology Economy. 14 has also authored numerous articles on application of 15 economic theory to antitrust issues and on high 16 technology antitrust. 17 lawyer in the Antitrust Division of the U.S. Department 18 of Justice, and his recognitions include Chambers USA, 19 America's Leading Business Lawyers, The Best Lawyers in 20 America, Legal Media Group's Expert Guide to Competition 21 and Antitrust Lawyers, and Global Competition Review's 22 GCR 100. 23 Dan is Chair of Latham's Global Antitrust and Throughout his career, Dan Dan was a founder and served four years as editor He He began his career as a trial Then to my immediate right is Rich Parker, a 24 partner at O'Melveny & Myers. Rich is Co-Chair of that 25 firm's Antitrust/Competition Practice. He returned to For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 69 1 O'Melveny in February 2001 after serving three years at 2 the FTC, as first Senior Deputy Director and then 3 Director of the Bureau of Competition. 4 recognized as a Leading Lawyer in Antitrust by the Legal 5 Times; named by the Global Competition Review as one of 6 the best antitrust defense lawyers in the United States; 7 and recognized as a leading antitrust practitioner by 8 Global Competition Review, Chambers Global, Chambers 9 USA, and Super Lawyers Magazine, and probably others. 10 He received the Distinguished Service Award also from 11 the FTC. 12 Rich has been So, with that, I think we are going to try to 13 follow the same format that the first panel used, which 14 is to ask each of the panelists to give a short 15 presentation, and then we will go right into questions 16 and hopefully have a lively discussion. 17 going to start with Jon. 18 PROFESSOR BAKER: I think we are Good morning, everyone. I am 19 delighted to have been asked to be here, and I see some 20 old friends. 21 for future reference, Bobby and Andy, I prefer to be 22 discussed for my ideas, not for how I look, okay? 23 It is also very nice to be discussed, but My assignment is to talk about -- is to be a law 24 professor and to talk about the -- I can't help it, I 25 will be an economist, too -- talk about the pros and For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 70 1 cons of using market definition in unilateral effects 2 cases to set up the panel. 3 into three categories, so I am going to talk about legal 4 arguments, economic arguments, and litigation tactic 5 pros and cons. 6 The arguments neatly divide So, on the legal side, we have to start with the 7 words of the statute, of Clayton Act Section 7, which 8 objects to acquisitions that substantially lessen 9 competition, and now I will quote, "in any line of 10 commerce or in any activity affecting commerce in any 11 section of the country," and that language, that 12 statutory language, arguably, makes proof of a market an 13 element of the offense. 14 On the other hand, if the Government can prove 15 harm to competition directly, there has to be some 16 market within which competition takes place, and, why 17 isn't that inference good enough to satisfy the statute? 18 I once wrote an article where I called that kind of 19 approach a res ipsa loquitur market definition. 20 words of the statute is one legal issue. 21 So Another legal issue is the Oracle decision. 22 Judge Walker held that the Government must prove that 23 the merger must -- in a unilateral effects case, that 24 the merger must -- would create a monopoly or near 25 monopoly. Monopoly is almost always demonstrated by For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 71 1 high market share, so the Government essentially has to 2 define a market to satisfy this element of what Judge 3 Walker sees as part of the offense. 4 that Judge Walker's holding in that decision is based on 5 a clear error in economic reasoning. 6 believe that other courts will follow it. 7 commonly used horizontal differentiation model that 8 Judge Walker seems to have in mind, unilateral effects 9 can arise in mergers that involve firms that are not the 10 largest in the market and that do not create a dominant 11 firm, just as a matter of economics. 12 legal pros and cons. 13 The con here is So, I don't Even in the So, that is the Now, economic pros and cons of defining a 14 market. I think here I am going to start with the cons 15 and not the pros. 16 among sellers of differentiated products does not turn 17 on market shares. 18 as arising because the merger lets the firm recapture 19 profits that previously it would have lost were it to 20 have raised price, and so it now has, after the merger, 21 an incentive to raise price. The economics of unilateral effects You can think of unilateral effects That is one intuition. 22 Another way of thinking about unilateral effects 23 is that they arise because the merger allows the firm to 24 remove the competitive response of an important rival, 25 and that makes the initial firm's residual demand less For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 72 1 elastic. 2 incentive to raise its price. 3 about unilateral effects among sellers of differentiated 4 products, the market shares do not directly matter to 5 the economic analysis. 6 Again, you can see how that would give it an Either way you think Now the other side of the story. The market 7 shares would be a good indicator of pressure to raise 8 price if the diversion ratios or the demand elasticities 9 are related to them. 10 second choices are distributed similarly to customer 11 first choices, which is what Bobby was getting at this 12 morning when he talked about generally differentiated 13 products. That could occur if the customer's 14 Also, high market shares likely indicate that 15 the diversion ratios are so high or that they are high 16 enough that they will generate some sort of unilateral 17 effects, unless the merging firms' products appeal to 18 very different groups of customers. 19 a 50 percent market share merges with a firm with a 20 20 percent market share, the two would have to be in very 21 different niches in order to not have a unilateral 22 effects problem. 23 burden. 24 25 So, if a firm with The high shares almost shift the Also on the pro side of using market definition in the economics category, if the way you collect the For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 73 1 evidence relies on econometric evidence of diversion 2 ratios or demand elasticities, then, some sort of an at 3 least informal market definition is required to specify 4 the list of potential rivals that you have to include in 5 order to avoid bias in your analysis. 6 So, if you leave out an important rival when you 7 conduct the estimation, then the elasticity estimates 8 are most likely biased in the direction of overstating 9 the unilateral effects. 10 Bobby was also getting at this morning when he talked 11 about collecting the important sources of competitive 12 discipline. 13 in this context, as biasing the estimate of unilateral 14 effects because you left out the others down the road, 15 in Bobby's theory. 16 This is something that I think The gas station example could be understood The third area where I want to talk about pros 17 and cons of defining markets and proving unilateral 18 effects cases has to do with litigation tactics. 19 the pros and cons depend on whether the Government would 20 define a narrow market or a broad market or not one at 21 all. 22 market. 23 through superstores rather than all office supplies, or 24 superpremium ice cream rather than ice cream, the kind 25 of things that we talk about in our professional world. Here, Let's suppose the Government defines a narrow Here we have in mind, office supplies sold For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 74 1 The pro of defining a narrow market is that it 2 allows the Government to portray the case as a merger to 3 monopoly or near monopoly, and it also creates a causal 4 inference of unilateral effects when the market shares 5 are high, as with the 50 percent firm merging with a 20 6 percent firm, as we said before. 7 On the other hand, a narrow market may not be 8 persuasive if it looks gerrymandered. 9 particular problem if some of Bobby's Audi drivers would 10 go to BMW and some would go to Lexus. 11 is harder for him to sell his Audi/BMW market to a 12 court, particularly if more of the Audi customers would 13 go to Lexus than to BMW. 14 That could be a It may be that it Also, this approach potentially focuses 15 attention on the wrong issue. 16 primary attention to the extent of buyer substitution to 17 the third firms, the rivals outside the market, rather 18 than to the extent of the buyer substitution between the 19 merging firms, which is the source of the unilateral 20 effects. 21 substitution between the merging firms, but you are busy 22 worrying about, in market definition, the substitution 23 to the third firms. 24 25 That is, it directs your The first thing you want to know is the Now, let's suppose the Government defines a broad market. The pro here is that the market may seem For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 75 1 more intuitive, like Jan suggested this morning, unless 2 gerrymandered in its appearance. 3 allows the competitive effects case to take primary 4 place in telling the competitive effects story in 5 litigation for the Government and focus attention on the 6 way that the merger lets the firm recapture lost profits 7 or alter the competitive response of an important rival, 8 consistent with the economic theory. 9 on the theory, the economic theory. 10 The broad market You are focusing On the other hand, if you define a broad market, 11 you may essentially admit that a large number of firms 12 are rivals to the merging firms, that merging firms' 13 shares are small and that competitive effects are not 14 uniform, because they are concentrated in a small part 15 of the market. 16 trying the case, and they make the competitive effects 17 look small. 18 Government embroiled in this question of whether there 19 is a 35 percent safe harbor for unilateral effects or 20 not in the Merger Guidelines that was alluded to in the 21 last panel. 22 All those things are bad optics for And there is also the danger of getting the The final litigation choice would be not to 23 define a market at all. Again, the benefit of that is 24 it focuses the case on the way the merger lets the firm 25 recapture the lost profits or removes the competitive For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 76 1 response of an important rival, consistent with the 2 economic theory. 3 avoid litigation problems with defining a broad market 4 when market shares are low, but the con is that may be 5 illusory, because the defendant would presumably define 6 a broad market, and so the Government may not actually 7 avoid the problems arising from defining a broad market. It would seem the Government could 8 So, there you have it, an even-handed view of 9 pros and cons of proving markets in unilateral effects 10 cases. 11 MR. SCHMIDT: Thanks, Jon. 12 Kathy? 13 MS. FENTON: 14 I was asked to share some thoughts on the legal Thank you, Jeff. 15 need to prove market definition in unilateral effects 16 cases, and as Jon Baker already indicated, the reason we 17 are having this discussion goes back to the basic 18 language of Section 7, the requirement to show effects 19 "in any line of commerce in any section of the country," 20 a mandate that some -- you may call them a strict 21 constructionist -- have identified as being the source 22 for any obligation to prove markets as part of your 23 affirmative showing of a Section 7 violation. 24 25 But I think the more interesting issue to focus on in this area is the fact that much of the current For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 77 1 debate can be directly traced to the lack of recent 2 and/or relevant Supreme Court opinions on this subject. 3 I am sure there is a great analogy to children's 4 literature that is possible here, whether it is Rip Van 5 Winkle or The Sleeping Princesses, but your last 6 substantive merger case goes back to 1975, and the last 7 time the court spoke on this issue was a year earlier, 8 in 1974, in the Marine Bancorp case, where it set forth 9 a fairly traditional three-part analysis that says: 10 "The analysis of likely competitive effects from 11 a merger requires determinations of, one, a line of 12 commerce, a product market in which to assess the 13 transaction; two, the section of the country or 14 geographic market in which to assess the transaction; 15 and three, the transaction's probable effects on 16 competition in the relevant product and geographic 17 market." 18 Now, judges, tending to be relatively 19 conventional creatures, look at that language and see, 20 not surprisingly, a mandate to define a relevant market. 21 The silence on the subject for the ensuing years from 22 the Supreme Court has simply added to the proliferation 23 of approaches we see at the district court. 24 those approaches have been responding to other 25 developments occurring at the Supreme Court level Some of For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 78 1 outside of Section 7, outside of the merger context, in 2 areas involving either collusion or monopoly claims, 3 because you have a whole series of cases, some of which 4 were briefly touched on by the opening panel, NCAA, Cal. 5 Dental, Polygram, and perhaps, most dramatically, 6 Indiana Federation of Dentists, that seem to eliminate 7 the need for formal market definition if there is actual 8 proof of anticompetitive effects. 9 And I think the quote from Indiana Dentists 10 probably captures this line of development outside the 11 merger area most dramatically, because there the Supreme 12 Court said: 13 "Since the purpose of the inquiries into market 14 definition and market power is to determine whether an 15 arrangement has the potential for genuine adverse 16 effects on competition, 'proof of actual detrimental 17 effects, such as a reduction of output,' can obviate the 18 need for inquiry into market power, which is but a 19 'surrogate for detrimental effects.'" 20 Needless to say, that precedent from the Supreme 21 Court has surfaced in numerous briefs, often by the 22 private plaintiffs or government agencies prosecuting a 23 unilateral effects merger, seeking to argue that the 24 formalities of market definition are not essential as an 25 element of proof, and the argument in that regard, I For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 79 1 think, is perhaps very nicely captured in a recent 2 article by Katz & Shelanski in the Antitrust Law 3 Journal, called "Mergers and Innovation," that takes a 4 slight detour through unilateral effects analysis and 5 says: 6 "If the formalities of market definition can be 7 skipped in favor of direct analysis of harm in 8 monopolization and collusion cases, there is no reason 9 why the same should not hold true for merger analysis 10 where the issue, likely competitive harm, is similar." 11 They go on to recognize that merger analysis has 12 some limitations. 13 prospective and predictive than other kinds of antitrust 14 cases where the conduct at issue frequently has been 15 ongoing for some time," but this simply means that 16 direct effects may be easier to show in nonmerger cases 17 and not that direct evidence of market power should not 18 have the same priority in merger cases where such 19 evidence is available. 20 They say it is "more often I would suggest that economists probably have 21 more flexibility than district court judges in offering 22 that alternative as a way of resolving these cases, but 23 the debate continues, and as you look at the recent 24 district court opinions involving unilateral effects, 25 you know, Oracle, Whole Foods, Arch Coal, you really For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 80 1 could go down the litany, you see judges struggling with 2 this question of what is their obligation to formally 3 make findings of fact and conclusions of law on the 4 relevant market question, and they tend to engage in 5 activities that could be characterized as a market 6 definition exercise without necessarily acknowledging 7 their obligation to do so. 8 can identify for resolving this question is the 9 possibility of further Supreme Court statements on this 10 question. 11 And I think the only hope I Now, in the world post Hart-Scott-Rodino 12 notification, that is going to be a difficult 13 proposition, just because most mergers that are 14 challenged by a government enforcement agency do not 15 hold together long enough to ever reach the point of 16 Supreme Court review, but I think there is one possible 17 candidate on the horizon that I offer for your 18 consideration. 19 definition not with respect to a product market but a 20 geographic market, and the case, of course, is the 21 Commission decision in Evanston, which is still 22 awaiting, as far as I know -- and I will bow to more 23 superior information sources -- a determination by the 24 parties to file an appeal with one of the U.S. circuit 25 courts. It poses the question of role of market For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 81 1 But in that case, the Commission opinion dealing 2 with a post-closing challenge to a hospital merger 3 concluded: 4 "It is not necessary to define the relevant 5 geographic market, because it is possible to show, 6 through direct evidence, that the merger enabled the 7 merged parties to exercise market power unilaterally." 8 Thus, the Commission concluded, because the merger 9 enabled the parties to raise prices by a substantial 10 amount, at least equal to a SSNIP, through a unilateral 11 exercise of market power, the geographic area alleged by 12 the FTC to constitute a relevant market constituted a 13 well-defined antitrust geographic market under Section 14 7. 15 Now, if that issue were preserved through the 16 appellate process, we certainly have the prospect of a 17 court of appeals chiming in on the need for relevant 18 market definition and, as I said, a possibility for 19 Supreme Court review since a concluded merger, a 20 divestiture challenge essentially, is sufficiently high 21 stakes that the parties might be incented to take that 22 step. 23 But in the absence of that, I think we are going 24 to continue to see a struggle at the district court 25 level as they look back to precedents, and it is not For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 82 1 just Marine Bancorp. It is Philadelphia National Bank, 2 it is DuPont, that all contain the language about 3 defining relevant markets, as well as what I would 4 suggest are some practical limitations imposed by the 5 Merger Guidelines themselves and the Merger Guidelines 6 structure, because there, the five-part organization 7 embodied in the Guidelines has, in a sense, provided a 8 road map for a lot of subsequent district court 9 analysis. 10 You start with market definition and 11 concentration; you consider potential adverse effects; 12 you do an entry analysis; you consider efficiencies; you 13 deal with failing or exiting assets. 14 sounds like a mandate for relevant market definition, 15 and as a result, to borrow Andy's phrase from the 16 initial panel, it is probably a very hard sell for the 17 courts to try and avoid or escape that exercise, and in 18 particular, this combines with a number of other 19 practical aspects, including judicial skepticism of 20 economic analysis. That, again, 21 And I was reminded in preparing for this 22 exercise of a fascinating quote from Ken Auletta's book, 23 World War 3.0, which, of course, is on the Microsoft 24 case, but he had, you might recall, conducted fairly 25 extensive interviews as part of the process for that For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 83 1 book. One of the people he interviewed was Judge Hogan 2 of the district court here in the District of Columbia, 3 who some might view as one of the godfathers or patron 4 saints of unilateral effects analysis since he is the 5 author of the opinion not just in Staples, but also 6 Swedish Match a few years earlier. 7 They somehow got off the topic of Microsoft in 8 the discussion for Auletta's book and started talking 9 about the Staples/Office Depot case, and Auletta reports 10 in his book: 11 "When Judge Hogan presided over the Government's 12 antitrust action to block the proposed merger of Staples 13 and Office Depot, Hogan reported, 'We had a lot of 14 economic evidence, we had a lot of documentary evidence, 15 although in that case, the economic evidence that the 16 Government had was not at all convincing to me. 17 the internal company documents were more convincing. 18 That is why I stopped the merger.'" I think 19 And that reality, I think, is something that you 20 are going to see reflected in perhaps less overt fashion 21 in many of the judicial decisions dealing with that 22 question. 23 MR. SCHMIDT: 24 Rich? 25 MR. PARKER: Thanks, Kathy. I am supposed to give the For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 84 1 government perspective on trying these cases, and as you 2 know, I am now playing on the other team, so it would 3 really be unfair if you quote this stuff back to me when 4 I am sitting next to a client. 5 trying to convince you to go away. 6 down as a rule. 7 When I'm down here So, let's get that What I want to talk about is how to put a case 8 like this together. We have people who understand the 9 law and economics better than I do. 10 hear that from me. 11 and trying cases is an art, and everybody has a 12 different style, but here is the way I think about it. You do not need to So, here is my own personal view, 13 I was privileged, my first job out of law 14 school, to clerk for Judge William Matthew Byrne, 15 Junior, in Los Angeles, who passed away a year ago, who 16 was one of the best trial lawyers in Southern California 17 before he went on the bench. 18 And was a great trial judge and was a great teacher. 19 And I remember, when I was down there, we had this 20 really boring patent case. 21 dry than listen to this testimony in this chemical 22 patent case, but that was my job and my co-clerk's. 23 He won a lot of big cases. I would rather watch paint And the trial ended, and we went back to 24 chambers, and the judge said, "Well, "Justice West of 25 the Pecos" says that the plaintiffs ought to win here." For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 85 1 I said just looked at him. He said, "By that I mean, 2 common sense, logic, my gut sense of what is fair and 3 reasonable," and then he went through and told a story 4 about what happened here, which is exactly the way 5 counsel probably should have tried the case, and said, 6 "Now, that is what my opinion ought to say, and you tell 7 me if we can get to a plaintiff victory under the case 8 law, and if we cannot, then we better have a meeting 9 and, figure something else out." 10 Pecos" has always been in the back of my mind. 11 stopped being a mentor to me, and that is the way I view 12 these cases. "Justice West of the He never 13 In my opinion, the Government ought to try these 14 cases with effects, and I do not think what I am saying 15 is anything inconsistent with what was said in the first 16 panel. 17 have an advantage in being the Government, and the 18 advantage is inherent judicial conservatism. 19 market that is working. 20 coming in with their fancy economists saying, "Well, we 21 are going to change this structure radically, but don't 22 worry, our efficiencies are going to do this, that, and 23 the other thing." 24 skepticism with a judge or with most judges about 25 radical changes in a functioning market, and you are You start with effects. Remember this. You You have a And now you have these guys And so I think you have an inherent For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 86 1 trying to stop that from happening. 2 advantage. That is an 3 So, you play on that, and you build it by 4 showing what is going to happen that is bad here. 5 are people going to get hurt? 6 and others said this morning, there is an inherent 7 dinner table logic to unilateral effects. 8 not care about Bimmers and Audis, but Whoppers and Big 9 Macs or something like that they do. 10 competition from other burgers and maybe from Taco Bell, 11 but those two are unique competitors, and they look at 12 each other when they price their products, and if one 13 buys the other, that constraint is gone. 14 logic that makes a lot of "Justice West of the Pecos" 15 sense. 16 How And as Dr. Willig said Judges may Sure there is That is a In my opinion, the most important support for 17 that case is the company's business documents. 18 they look at? 19 board? Do they look at this fringe or do they look at 20 tacos? Do they look at whatever? 21 each other? 22 build on that. 23 What do What do they look at when they go to the Or do they look at That is the number one point. And you And the second thing you build on are customers. 24 Customers. The Government cannot try, effectively, a 25 case without strong customer support, and by "customer For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 87 1 support," I don't mean just, "I hate the merger." I 2 mean, "I have dealt with these people day-in and 3 day-out, for year after year, and I play them off each 4 other, and this, that, and the other thing, and I have 5 detailed knowledge, and in my opinion, I have benefited 6 from that competition, and let's not let it go away." 7 You cannot put on the stand a lot of people who 8 simply don't like the merger because they don't like the 9 merger but do not have any real experience in dealing 10 with the entity being purchased. 11 Arch Coal, where at least -- and this is Monday morning 12 quarterbacking -- but at least some of the witnesses in 13 that case had that problem. 14 Now, relevant market. I am going back to You have to prove a 15 relevant market. Every case says that. You can't 16 pretend like they do not say that, including your 17 favorite cases, starting with Chicago Bridge, your 18 latest victory, Swedish Match, every one of them, Baker 19 Hughes, Staples, Drug [Text truncated at 120,000 characters. The full text is on the page linked above.] --- Source: Frix Law Library, https://www.frixlaw.com/law-library/documents/agency%3Aftc%3A8a9f400fc8c77b99. Public record. Not legal advice.