# FEDERAL TRADE COMMISSION > Briefs, arguments, decisions, and more. URL: https://www.frixlaw.com/law-library/documents/agency%3Aftc%3A05957757fb1fed5b ## Record - **Collection:** Agency decision - **Document type:** Agency decision ## Text 1 1 FEDERAL TRADE COMMISSION 2 I N D E X 3 4 5 OPENING REMARKS PAGE 6 CHRISTINE VARNEY 4 7 8 9 PANELS PAGE 10 1 15 11 2 57 12 3 99 13 4 139 14 5 183 15 16 17 18 19 20 21 22 23 24 25 For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 2 1 FEDERAL TRADE COMMISSION 2 3 4 FEDERAL TRADE COMMISSION ) 5 WORKSHOP ON: ) 6 HORIZONTAL MERGER GUIDELINES ) Matter No. 7 REVIEW PROJECT ) P092900 8 ) 9 ---------------------------------) 10 11 TUESDAY, JANUARY 26, 2010 12 13 Conference Center 14 Federal Trade Commission 15 601 New Jersey Avenue, N.W. 16 Washington, D.C. 20580 17 18 19 The above-entitled hearing was held, pursuant to notice, at 9:00 a.m. 20 21 22 23 24 25 For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 3 1 P R O C E E D I N G S 2 - 3 - MR. SHELANSKI: - - Good morning. Thank you very 4 much for coming to this fifth of our Horizontal 5 Guidelines Workshops. 6 started, one a chore and one a pleasure. 7 to read you a security briefing, which we're required to 8 read, and the pleasure is to introduce our opening 9 speaker. 10 I have two tasks before we get The chore is So here's the security briefing. Anyone that 11 goes outside the building without an FTC badge will be 12 required to go through the magnetometer and x-ray 13 machine prior to re-entering into the conference center. 14 In the event of a fire or evacuation of the building, 15 please leave the building in an orderly fashion. 16 worries me with this crowd. 17 building, you need to orient yourself to New Jersey 18 Avenue. 19 That Once outside of the Across from the FTC is the Georgetown Law 20 Center. 21 rallying point. 22 not relevant to us. 23 person accounting for everyone in the conference center. 24 That person will make themselves known to you. 25 Look to the right front sidewalk. That is our Everyone will rally by floors. That's You need to check in with the In the event that it is safer to remain inside, For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 4 1 you will be told where to go inside the building, and if 2 you spot suspicious activity, please alert security, and 3 that does not refer to my colleague Phil Weiser's 4 remarks. 5 Towards the end of each panel or during each 6 panel, somebody will distribute these question cards. 7 We ask you, please, to fill them out, and they will be 8 brought up front. 9 most effective way to get questions. 10 the audience, however, we may opt to just take your 11 questions. 12 people here. 13 That has tended, we think, to be the Given the size of It might be simpler, given the number of Finally, there will be a lunch break from 12:30 14 to 2:00. 15 here, but there are a couple of cafes. 16 a left out the front door, there are a couple of places 17 to eat on that block, and then there's Union Station 18 just a couple blocks away. 19 closest bets. 20 The offerings are not enormous right around If you just make Those are probably your Finally, I would like to introduce Assistant 21 Attorney General Christine Varney, of course, she needs 22 no introduction, who will give us some opening remarks 23 today, and just so you all know, all remarks today will 24 be on a transcript that will be posted online within a 25 week or so. Thank you. For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 5 1 MS. VARNEY: Thanks, Howard. Good morning, 2 everybody. Don't take any notes because my remarks are 3 going to be posted on the DOJ website at greater length. 4 I wanted this morning is talk about what we've learned 5 from the workshops and the comments we've received so 6 far and give you some preliminary views about what we've 7 heard during the process and where I, and only me, 8 believe consensus may be emerging. 9 As we complete the five workshops, we welcome 10 additional comments, including comments on the topics 11 that I'm going to outline this morning. 12 A consistent theme running through the panels is 13 that there are indeed gaps between the guidelines and 14 actual Agency practice, gaps in the sense of both 15 omissions of important factors that help predict the 16 competitive effects of mergers and statements in the 17 guidelines that may now be inaccurate. 18 something that we are all aware of. 19 These gaps are The guidelines need to inform practitioners and 20 the business community of the Agencies' standards for 21 evaluating mergers. 22 what we actually do run counter to our goal of being 23 transparent. 24 accurate predictions about our likely enforcement 25 intentions adjust new behavior accordingly. Gaps between what we say we do and Transparency helps businesses make Lack of For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 6 1 transparency creates uncertainty, and uncertainty 2 results in unpleasant surprises. 3 We want to avoid that. Similarly, the agencies rely heavily on the 4 merger guidelines in our competition advocacy efforts 5 both here and abroad. 6 pervasive, the guidelines must reflect our best thinking 7 about the competitive effects of mergers and appropriate 8 merger enforcement policy. 9 To be effective as well as Courts also rely on the guidelines, in the words 10 of the Fifth Circuit, "to provide persuasive authority 11 when deciding if a particular acquisition violates 12 antitrust laws." 13 inaccurately reflect enforcement or omit crucial 14 considerations, we do a disservice to the law as well as 15 the business community. 16 When the guidelines either At the same time, I do not want to overstate the 17 magnitude of these gaps. 18 whether a merger is likely to create or enhance market 19 power, resulting in anti-competitive effects, remains 20 the heart of merger analysis. 21 The focus in the guidelines on The guidelines articulation of possible 22 unilateral and coordinated effects entry and 23 efficiencies accurately reflects the key concerns of 24 merger analysis. 25 that I did not envision radical review of the I said at the outset of this project For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 7 1 guidelines. 2 workshops has changed my assessment. 3 guidelines, however, does appear to be worthwhile in a 4 number of areas. 5 Nothing so far in the comments or the Updating the Turning to some specifics, there are a few areas 6 where consensus appears to be emerging. To begin, many 7 of our panelists have noted that the Agency's do not 8 mechanically apply the five-step process set forth in 9 the guidelines where markets and market shares are first 10 assessed, followed by a sequential consideration of 11 potentially adverse competitive effects, entry, 12 efficiencies, and then failing firm defenses. 13 our panelists advocated following that sequence as the 14 best way to either assess every merger's likely 15 competitive effects or to reach an enforcement decision. None of 16 To be sure, the guidelines themselves offer a 17 note of caution regarding the potentially misleading 18 results that can follow from mechanical application of 19 the guidelines. 20 flexibility is both the norm of actual Agency practice 21 and appropriate given the diversity of considerations 22 that are presented in the range of transactions viewed 23 by the agencies. 24 25 Panelists have noted that far more Thus, as a matter of actual practice and sound theory, some adjustment of the description of the For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 8 1 analytical process used by the Agency seems appropriate. 2 Implicit in deemphasizing the sequential nature of the 3 guidelines inquiry is a recognition that defining 4 markets and measuring market shares may not always be 5 the most effective starting point. 6 Remember the purpose of defining a market and 7 assessing shares is to assess the potential harm. 8 it is clear that either certain vulnerable customers are 9 likely to be harmed by a merger or that certain 10 customers have, in fact, been harmed by a consummated 11 merger, the need to define a market to assess likely 12 competitive effects is obviously diminished. 13 When For instance, the consumer harm that followed 14 from the consummated Evanston Hospital transaction 15 lessened the importance of the Commission's market 16 definition and market share analysis. 17 have largely confirmed the view that market definition 18 should not be an end-all exercise. 19 something to be incorporated in a more integrated, fact 20 driven analysis directed at competitive effects. 21 Of course this is not news. Our panelists Rather, it is The commentary on 22 the Horizontal Merger Guidelines explains that the 23 agencies apply the guidelines flexibly, and those 24 practicing before the agencies have been aware for some 25 time that market concentration is more important in some For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 9 1 cases than others. 2 new, disruptive entrant may well impact competition far 3 more than market shares might suggest. 4 For instance, a merger involving a Similarly, the Division's merger review process 5 initiative has recognized the appropriateness of a 6 tailored second request schedule designed to enable the 7 division to take a quick look at potentially dispositive 8 issues such as failing firm and entry at the outset of 9 an investigation; thus precluding the need for full, 10 sequential review outlined in the guidelines. 11 acknowledging this flexibility in the guidelines 12 themselves seems to me to be prudent. 13 Expressly The next area I would like to discuss is one 14 where the guidelines appear to be inaccurately 15 describing the Agencies' enforcement policy. 16 come as no surprise to you, the merger challenges data 17 that we collect confirms that it is rare for the 18 agencies to challenge mergers that will lead to HHI 19 concentration levels below 1,800. 20 indicate that such mergers potentially raise significant 21 competitive concerns. 22 It will Yet the guidelines Similarly, the guidelines suggest that a 100 23 point increase in HHI concentration level raises 24 anticompetitive concerns. 25 agencies have only infrequently, indeed I say rarely, In actual practice, the For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 10 1 challenged a merger unless they increased concentration 2 several times that much. 3 More broadly, our panelists have generally 4 confirmed that the guidelines overstate the importance 5 of HHIs in merger analysis. 6 you that HHIs have not been the focus of any party 7 presentation or any staff recommendation since I've been 8 the Assistant Attorney General. 9 the current state of merger analysis where HHI levels 10 are given far less prominent place as a predictive tool 11 for assessing competitive effects than suggested by the 12 guidelines. 13 Again, it will not surprise That reality reflects In that vein, I note that while many panelists 14 have acknowledged their usefulness as a tool for 15 assessing likely competitive effects, none has 16 maintained that HHIs should be the key driver in an 17 enforcement decision. 18 It is clear that the HHI threshold set forth in 19 the guidelines no longer capture Agency practice or 20 economic learning about the kinds of mergers that are 21 most likely to lead to consumer harm. 22 thresholds to express accurately how the agencies use 23 them seems not just appropriate but also necessary to 24 overcome what is at this point an affirmative 25 misstatement. Revising the For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 11 A third area where the guidelines may be 1 2 usefully updated is unilateral effects. 3 where economic thinking and Agency practice have 4 progressed significantly since 1992, when the concept of 5 adverse unilateral effects was first explicitly 6 introduced in the guidelines. 7 This is an area That introduction was a major step forward, but 8 the treatment of unilateral effects was sparse, and 9 several of our panelists and commentators have noted 10 that significant advances in thinking have taken place 11 since 1992. 12 Unlike the HHI threshold where gaps are more in 13 the nature of misstatements, in unilateral effects, the 14 gaps are more in the nature of omissions. 15 important considerations that the agencies routinely 16 employ when assessing unilateral effects that are not 17 mentioned or even alluded to in the guidelines. There are 18 Our panelists identified a number of 19 considerations routinely used to assess unilateral 20 effects. 21 reports, customer switching patterns, the views of 22 competitors, customers and industry observers, for 23 instance, are all tools we use to analyze mergers of 24 firms selling differentiate products. 25 Diversion ratios, price cost margins, win/loss Yet the guidelines say little about how these For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 12 1 types of evidence are used to assess unilateral effects. 2 In fact, when assessing pricing effects in markets with 3 differentiated products, both agencies employ a variety 4 of techniques to evaluate whether or not the merger is 5 likely to lead to higher prices. 6 There is a growing body of evidence that 7 measures of upward pricing pressure, which focus on 8 diversion ratios and price cost margins, more accurately 9 evaluate the likelihood of unilateral pricing effects 10 where HHIs may be more productive in a coordinated 11 effects analysis. 12 Unilateral effects can arise along many 13 dimensions of competition, including pricing of 14 differentiated products, negotiations between buyers and 15 sellers, output and capacity for more homogeneous 16 products, product variety and innovation. 17 have accumulated a great deal of experience analyzing 18 such effects, and that expertise is not reflected in the 19 current guidelines. 20 The agencies Updated guidelines can enhance transparency by 21 explaining how the agencies currently evaluate 22 unilateral effects. 23 areas where clarification of the guidelines appears to 24 be worthwhile. 25 I'll briefly mention five other First, the discussion in the guidelines of For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 13 1 targeted customers and price discrimination could be 2 clarified. 3 discrimination. 4 discrimination is quite abbreviated. Many of our cases involve price Yet the guidelines treatment of price Second, the guidelines could more accurately 5 6 convey actual Agency practice by indicating that market 7 shares are normally assessed using recent or projected 8 sales in the relevant market while explaining the 9 conditions under which other measures, such as capacity, 10 may be used. Third, different parts of the guidelines employ 11 12 closely related concepts of supply side responses by 13 non-merging firms: 14 in the relevant market, uncommitted entry, repositioning 15 and committed entry. 16 more unified approach to these concepts. Expansion by firms already selling A number of panelists suggested a Fourth, the guidelines could clarify that 17 18 coordinated effects can arise through accommodating 19 behavior among a small number of rivals without the 20 necessity of reaching the terms of coordination. Lastly, several panelists have pointed out that 21 22 the guidelines are virtually silent regarding 23 innovation, despite wide spread recognition innovation 24 generates enormous value for consumers over the long 25 run. For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 14 A revision could move the guidelines into the 1 2 21st Century by explaining how the agencies account for 3 market dynamics, the pro-competitive role of disruptive 4 entrants and a merger's effect on innovation. It's now time to turn to our expert panelists, 5 6 but I would first like to reiterate several notes of 7 thanks. 8 have submitted or will submit comments, thank you for 9 volunteering your time and expertise. To all our panelists and to those who either I would also like to offer my warm thanks to the 10 11 Federal Trade Commission staff and Antitrust Division 12 staff who have worked very, very hard to organize these 13 workshops, and I'm very appreciative for the very 14 wonderful and cooperative relationship the FTC and the 15 DOJ have enjoyed on this project, and I thank the FTC 16 for their hospitality today. So I'm going to turn it over to my deputy, Phil 17 18 19 Weiser. Thanks, everyone. (Applause.) 20 21 22 23 24 25 For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 15 1 PANEL 1: 2 PRESUMPTION. 3 MODERATOR: 4 General 5 PANELISTS: 6 MARK COOPER, Director of Research, Consumer Federation 7 of America 8 ALBERT A. FOER, President, American Antitrust Institute 9 ANDREW I. GAVIL, Professor of Law, Howard University 10 School of Law 11 CHARLES F. RULE, Partner, Cadwalader, Wickersham & Taft, 12 LLP MARKET CONCENTRATION AND THE STRUCTURAL PHIL WEISER, Deputy Assistant Attorney 13 MR. WEISER: 14 So thank you, Christine. That was 15 a terrific way to kick off our last session, and thank 16 you so much for your engagement and support in this 17 effort. 18 panel to come up, hopefully sitting where your name card 19 is, I will start the process of introducing you all. If I could ask the panelists for the first 20 The idea of this workshop in general was to get 21 a variety of perspectives, and there are many people who 22 said, Well, are there things you want us to say, and the 23 answer was, No, we want you to have a thoughtful, 24 engaging discussion, and with the folks we have here, 25 I'm very confident we will. For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 16 Sitting to my left, Mark Cooper, who is one of 1 2 the premier consumer advocates, probably by all accounts 3 would be in the consumer advocate hall of fame. 4 the research director at the Consumer Federation of 5 America. 6 there, and I was at the Antitrust Division, and he is an 7 economist trained at Yale where he has a Ph.D. and is, 8 among other things, a fellow at the University of 9 Colorado's Silicon Flatirons Center. 10 He's I first met Mark in the '90s when he was Next to him, Bert Foer, who is what I guess you 11 would have to call a policy entrepreneur. 12 the American Antitrust Institute and has turned that 13 entity into a force. 14 not familiar with AAI, it brings together a bunch of 15 people on an advisory board as well as some resident 16 fellows to advocate on antitrust policy, and we're so 17 glad to have him here. 18 background in both private practice and also at the FTC 19 and in the industry. 20 He founded It is a unique enterprise. Those Bert comes from a rich Next to him Andy Gavil, who is one of the 21 leading lights in the academy, a law professor here at 22 Howard University. 23 antitrust case book with Bill Kovacic -- He is also a coauthor of an 24 MR. GAVIL: And John Baker. 25 MR. WEISER: -- and John Baker. John is now For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 17 1 chief economist with the FCC formerly of the FTC, and 2 Bill Kovacic, of course, is a Commissioner at the FTC. 3 Finally, Rick Rule, who is one of the former 4 Assistant Attorney Generals, who has been so kind to be 5 supportive and engaging on this project. 6 of a long-standing place at the bar. 7 Cadwalader where he is head of their antitrust group. 8 He represents a number of major clients and was head of 9 the Antitrust Division in the '80s. 10 Rick has sort He is now at The folks we have gathered here are going to 11 talk about market concentration. The way we're going to 12 do that is have a series of questions, have a give and 13 take, and I want to start with what Christine teed up 14 for us, which is we have these HHI figures, which many 15 acknowledge are, as Christine put it, a misstatement of 16 actual Agency practice, and also some would suggest out 17 of sync with economic learning. 18 question is: 19 makes it appropriate, and possibly if you would want to 20 supplement it, some have said maybe HHIs are not the 21 right framework. I guess the broad How do we think about this issue? What 22 We could think instead of significant 23 competitors or market shares of the merging firms, what 24 have you. 25 think, were you there when the '82 guidelines were done Rick, you've been around this for awhile. For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 I 18 1 or the '84 guidelines? MR. RULE: 2 I came shortly after the '82 3 guidelines were done and was there and in theory, sort 4 of the deputy, who was managing the process for 5 rewriting the guidelines in '84. MR. WEISER: 6 What I heard is back then they were 7 using four firm concentration ratios, and they tried to 8 reverse engineer from that an HHI figure, and they came 9 up with 1,800, which has been -- was 1,800 taking it 10 back in the '80s and they kept the '92? MR. RULE: 11 Well, what happened was in the '68 12 guidelines, the original guidelines, they used four firm 13 concentration ratios. 14 Werden, who is in the audience I think, is the source of 15 my information on this. 16 conversion in '82 was the introduction of what was 17 viewed at that time as a relatively revolutionary tool, 18 the HHI. 19 I wasn't there in '82, but Mr. As I understand it, the big As I understand it, Bill Baxter decided that it 20 made sense in making that change not to change the 21 thresholds, to largely have something that was similar. 22 Also the numbers, a thousand, 1,800, are nice round 23 numbers and sort of equate to certain things that make 24 sense, and so that's how that came about. 25 I should say it's also important to keep the For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 19 1 historical perspective in mind. 2 performance paradigm, so to speak, was alive and well, 3 reflected in the courts, and in fact even, one might 4 say, more extremely so, and that's where the four firm 5 concentration ratios came from. 6 In '68, the structure I can remember as a summer associate, in fact, 7 in the late '70s writing a paper that sort of summarized 8 the things other than structure that were relevant in 9 antitrust analysis and doing that for Ed Zimmerman, who 10 had also been an AAG, and he found that quite amazing, 11 that there were things that were relevant other than 12 structure. 13 In '82, I think it is fair to say that there was 14 the new learning that had occurred. 15 recognition that structure might be less important, but 16 there was still a strong sense within the Division, and 17 I think in Bill Baxter's case, that structure still was 18 the significant factor, and a lot of the analysis really 19 took the form of market definition issues, and of course 20 the focus in '82, in addition to the HHI, was the SSNIP 21 test, which of course was the principal lasting 22 innovation of the '82 guidelines. 23 There was a In '84, there was a sense that -- and if you go 24 back to the '84 guidelines, you will see that the HHI 25 thresholds, even there, there's a line that sort of For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 20 1 indicates they are at the beginning of the analysis. 2 There were some of us in the Division who thought we 3 should be even more explicit, that they really were safe 4 harbors as opposed to determinative guidelines, but 5 there was the sense that politically, that would not be 6 a fruitful exercise. 7 So instead of changing the numbers to reflect 8 what was emerging as the reality in '84, there was that 9 line. 10 analyses that have been done of mergers really are from 11 1990 forward, and if you go back into the 1980s and 12 around that time, there actually were mergers that were 13 getting scrutinized, and at times challenged, as I 14 recall it where the HHI post merger was below 1,800. Now, I will say, if you go back, most of the 15 What I think was happening over that period of 16 time, and you can see it in some of the speeches, you 17 can actually see it in the international guidelines that 18 came out at that time, there was a recognition in the 19 Division that the notion that structure or market shares 20 were really a beginning and that what you had to do, and 21 the way I like to say it, is you have to tell a story. 22 Inevitably you have to focus on structure 23 because what a merger does is changes the structure of 24 the market, and what you have to decide is, as the 25 agencies do, whether or not that change in structure For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 21 1 makes it likely, in light of all of the relevant 2 circumstances, that prices may go up, and I think 3 through the '80s there was a recognition that it was 4 more complex than just structure. 5 than could just be handled in a SSNIP test. It was more complex Entry was the factor that everybody focused on 6 7 in the '80s, but there were other factors that were 8 developing. 9 that the unilateral effects analysis was specifically 10 stated. 11 being looked at by the Division in the '80s. 12 I think notwithstanding it wasn't until '92 It was something that, in various forms, was So by the late '80s, by early '92, I think it 13 was very clear to anybody inside the Division that those 14 numbers were really only safe harbors and that they were 15 just the starting point, and at that point, structure, 16 market share was really only one part of the holistic 17 effort to tell the story, to show what the linkage was 18 and decide whether or not that linkage was a concern in 19 light of all of the relevant factors. 20 So I think that the history is important. 21 me, I think that if you look at those numbers that were 22 generated by the FTC and the DOJ in the early part of 23 the last decade now, you will see that I think probably 24 2,000 is the cut off or 2,500, as it now turns out, and 25 the few outliers where there's still cases that had been For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 To 22 1 brought below 1,800, I think there are kind political 2 explanations for them, and if anything, those political 3 explanations led the agencies to bring cases in those 4 areas because the guidelines were there. So as far as I'm concerned, they are, as 5 6 Christine said, safe harbors. 7 because I think today they're inaccurate, but I think 8 that the sort of trend away from just focusing on 9 structure is again a 30 year or longer occurrence. 10 ought to continue and the guidelines ought to be very 11 clear that structure really in and of itself can only be 12 a starting point and really can only be part of a much 13 larger effort of looking at a variety of factors. MR. WEISER: 14 15 They ought to be raised It Andy, how do you conceive of the HHIs and their proper role in the guidelines? MR. GAVIL: 16 I think what I would add to what 17 Rick said is that this 30 year evolution is larger than 18 just merger law. 19 years. 20 Section 1, Section 2, our thinking and our reliance on 21 structure has changed generally in many ways in 22 antitrust law. Antitrust law has changed in those 30 In you look back at cases in other carries, 23 So I think it's clear that we don't rely on 24 structure to the same degree that we once did, but 25 having said that, I worry about two things in throwing For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 23 1 the baby with the bath water out. 2 structure is of some utility? 3 utility at certain very high levels? 4 Do we still think Do we think it's of some One of the big challenges in antitrust generally 5 is that there is a trade-off between reducing error 6 costs, false positives, false negatives, and increasing 7 the direct costs of deciding and litigating cases, so 8 does the structural assumption have some utility? 9 think it still does. I 10 The other thing I worry about is every sentence 11 the agencies now add to the guidelines will be cited and 12 held against them in court when they litigate as a 13 constraint on their discretion, so to the degree you 14 move away from the structural presumptions and you 15 started adding, Well, we ought to look at this factor, 16 look at that factor, when you get into court and 17 litigate, people will say, Well, you didn't look at that 18 factor in this case, and I think there's a long history 19 under the guidelines since 1982 of courts holding the 20 agencies to their own guidelines and it not always 21 working out well for the agencies. 22 I completely agree, however, that moving away 23 from the 1,800 makes sense. The assumption there was 24 sort of a six to five was the threshold where we would 25 start getting concerned, it looks more like where we are For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 24 1 today is five to four or four to three, but I get very 2 concerned about the agencies saying that, creating the 3 impression of safe harbor and actually constraining 4 their own discretion, if a case based on other factors 5 that happens to be six to five or happens to be five to 6 four or gives them some concern, how you draft the 7 guidelines could wind up making it more difficult to 8 litigate and win that case if you had to. 9 So it makes sense to me to change it. 10 the case for changing the increased thresholds, the 11 1,500, that stuff is clearly -- that's too small to be 12 of some use, but I would just caution a little bit about 13 balancing the value of increased guidance against 14 constraints you can place on the Agency by adding 15 additional factors that you want to look at, which will 16 become de facto requirements when you litigate. MR. WEISER: 17 I think So you don't think the old lawyerly 18 construct of including, but not limited to, or 19 illustrative, but not necessarily required, is going to 20 do the job because there's a tension between providing 21 guidance and giving people transparency into what we do, 22 and the other side is you worry about pinning yourself 23 down? 24 25 MR. GAVIL: The guidelines currently say and even the announcement of this process said, this is just For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 25 1 how we make decision, not how we litigate, and then it 2 proceeded to list all of the cases in which the courts 3 had used the guidelines as a framework with shifting 4 burdens of production and proof, and I think that's just 5 a reality that you have to be aware of. 6 You can put the conditionality in there, but 7 thinking back to Baker Hughes and the language of the 8 guidelines on entry at the time, it didn't stop the 9 court from saying, Well, the language you've used is not 10 persuading us. 11 MR. WEISER: Bert, you at a conference last year 12 in Colorado said something to the effect of there is I 13 guess an indisputable gap between practice and the 14 guidelines, let's say it's 1,800 and 2,500 as Rick 15 suggested, and many people have said, as Rick noted, you 16 can just raise it, and I think you said at the time, 17 Well, you can start bringing more cases that are in the 18 1,800, 2,000 range. 19 How do you approach that issue? Is that still kind of your view? 20 MR. FOER: My view is that we're moving in the 21 wrong direction. 22 direction would be to conform practice to the 23 guidelines. 24 metric is that we should look for, and it seems to me 25 that the proper metric is Congressional intent. We have a gap, but that the proper You've asked us to talk about what the It's For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 26 1 not economic theory. And, the Congressional intent, as 2 defined unfortunately perhaps long ago, by the Supreme 3 Court is that this is an incipiency statute. The whole purpose is to avoid high levels of 4 5 concentration, and if you step back the way, for 6 instance, the Antitrust Modernization Commission did not 7 and you ask, Where are we, where have we come, how 8 concentrated has industry become? 9 to fail issue that everybody is worried about today? What is this too big I think you have to say that we have not 10 11 succeeded in fulfilling the Congressional mandate. Now, 12 why else do we hear our marching orders if not from 13 Congress and the Supreme Court? 14 is whether the Incipiency Doctrine can be utilized more 15 than it is in the guidelines. 16 in one sentence and virtually ignore it. So the question I ask The guidelines mention it It does not seem to me that this brings us back 17 18 to Von's, which nobody wants, but I thought that the 19 original guidelines, looking at basically a six to five, 20 and basically saying, okay, five companies competing 21 should be kind of a model, not inevitable, not 22 irrebuttable, but when you get to five, you should be 23 worried. 24 25 I think that the reality today is much more when you get to three, you're worried. Well, by then it's For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 27 1 awfully late to be worried. 2 good at is creating competition. 3 in theory at least, preserving competition, and we know 4 that by the time you get to three or four or five, 5 collusion is much easier to accomplish. 6 accomplished in a more fragmented industry, but common 7 sense tells us it's easier to accomplish. 8 One thing antitrust is not It's much better at, It can be Also, we don't really have a way of getting at 9 parallel behavior, so since we can't get at parallel 10 behavior very well, we should try to maintain a 11 structure in which it's less likely to occur. 12 Therefore, I would shift the burden, when we get 13 to high levels of concentration, and I would say instead 14 of starting with the proposition that underlies our 15 current policy: 16 good. 17 to the overall welfare. 18 Mainly, that mergers by and large are They're efficient. They're likely to be useful That's the golden proposition, and it works 19 pretty well up until high levels of concentration, but 20 then it no longer works, and when we look at the results 21 of mergers, most of them don't work out very well. 22 There's some that work out very well, and we've got to 23 not preclude those, but generally speaking, they're not 24 terribly successful, because the externalities of a 25 merger are not calculated into the analysis. There are For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 28 1 real externalities, and I think we're coming out the 2 wrong way. 3 level of concentration, there should be a strong 4 presumption against it, and the burden of demonstrating 5 that it's in the public interest should be on the 6 parties that want to go forward. 7 In other words, when we get to a very high MR. WEISER: So there are two ideas here. I'll 8 take them both. One is the virtues and vices of a safe 9 harbor and what should that be, and the second is the 10 virtue and vices of a I think it's called a structural 11 presumption, which is where Bert is going, and, Mark, I 12 want to ask you to address the second one. 13 At what point, and Bert suggested six to five or 14 five to four, others I think would suggest four to three 15 or even three to two, should a structural presumption 16 give some weight? 17 as a predictor of actual competitive effects has become 18 more questioned, although the guidelines still today 19 have a commitment to a structural presumption. 20 21 22 As Rick noted, the focus on structure Is that something that should be retained, and how should the agencies look at it? MR. COOPER: Well, I think the critical point is 23 if you're going to set a threshold, it's important to 24 know what the threshold means. 25 what it's going to mean, where you set it is a shot in Unless you really know For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 29 1 the dark. I like to analogize and a couple people have 2 heard this before, for the last 25 years, the merger 3 guidelines have been sort of like the pirate's code in 4 the Pirates of the Caribbean. 5 the movie, so now I get to tell the story. Not too many people saw It's a comedic device throughout the movie that 6 7 actually really gives you some insight into life in 8 pirate society, and essentially what happens is at each 9 key point -- it's an older crowd, you don't have young 10 kids here. 11 someone is about to do the morally incorrect thing, 12 another character says, But wait a minute, what about 13 the pirate's code, which of course tells you to do the 14 opposite thing, don't abandon your friend, right, when 15 you're about to jump ship? At each key moment in the movie, when The pirate's code will always tell you to do the 16 17 opposite thing, and everyone violates it, except of 18 course for the heroine, who is not a member of pirate 19 society, and they play this routine throughout the 20 movie. 21 committed, someone says, What about the pirate's code, 22 and they go off and do the wrong thing anyway. 23 Every time some dastardly act is about to be At the end of the movie, the chief villain is 24 challenged, and they say, Well, what about the pirate's 25 code, and he says, The pirate's code, they's only For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 30 1 guidelines, and he does the wrong thing. The fascinating thing is in the beginning of the 2 3 second movie, they introduce this very early, and 4 someone is about to challenge him, and he just waves his 5 hand and says, Don't give me that parlay stuff, he's not 6 going to hear the pirate's code, and the pirate's code 7 disappears from the last two movies. 8 The key here is that if the thresholds are going 9 to be meaningful, they will be useful, but lax law 10 enforcement is bad in antitrust, just like every place 11 else in law enforcement, so if you're going to give me 12 thresholds, they have to be meaningful, and I would say 13 the following: 14 many, I'm a ten guy, but that's okay, times have 15 changed. 16 I can live with four is few and six is I can live with four is few and six is many if, 17 when you get above 2,500, you pretty well know that 18 you're going to end up in court. 19 be meaningful, and you need to know that between six and 20 four, there's going to be a parlay going on. 21 the central theme in the pirate's code is whenever 22 you're about to get off, you would say, wait, parlay, 23 and in theory the pirates were supposed to negotiate. 24 If you come in with a six to five or a five to 25 four, you should know that there's going to be a really The threshold has to That was For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 31 1 tough conversation about the harm to competition, so for 2 me, I think the reality has moved there, and I agree 3 with Bert. 4 there's a gap between practice and the guidelines, we 5 assume that the guidelines are wrong and the practice is 6 right. 7 right, and that practice is wrong, but four to six is a 8 number that I think we can begin to live with. When we have this conversation, we say Some of us actually think the guidelines were Let me say, the other question, non structural 9 10 issues. 11 structures still have utility one step further. 12 the current state of the economic discipline, the 13 question is: 14 some utility? 15 I would take Andy's statement about that Given Does neoclassical economics still have Let's be clear. The fundamental assumptions 16 that we've used to analyze the performance of markets 17 has been shaken, sometimes I like to say buried, if not 18 dead, beneath the financial rubble of Wall Street, and 19 so we need to ask ourselves the question that 20 transaction costs economics and behavioral economics 21 teaches us things about economic performance that are 22 directly contradictory to neoclassical assumptions and 23 predictions, and the question we should ask ourselves: 24 Does the teaching of these two disciplines make it more 25 or less likely that market power will be abused? For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 32 1 I believe structure still has some importance, 2 and I frankly believe that behavioral economics teaches 3 us that once you have market power, given the role of 4 inertia and social influence and things in human 5 society, market power is liable to be more durable than 6 you thought, not less, because the assumptions you make 7 about human behavior are incorrect. 8 reality. They don't reflect Clearly a good clear statement of four is few, 9 10 six is many, with a precise understanding that this 11 stuff is going to meaningfully dictate future Agency 12 behavior would, in fact, be a better place to live than 13 where we've been for the last 25 years. MR. WEISER: 14 So we're going to come back to the 15 following formulation as you put it. You said in some 16 context, you need to be able to put on the spot, so we 17 can tell the story. 18 safe harbor, and in other contexts, some argue there 19 should be a structural presumption and maybe likely to 20 challenge. In some contexts you need maybe a Mark has put on the table if you have a six to 21 22 five or five to four merger, you have to have a good 23 story. 24 expect to be challenged, and the Agency should get a 25 presumption. If you have a four to three merger, you should What's your take on that proposal? For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 Is it 33 1 2 something you can live with? MR. RULE: Well, let me start by saying, as I 3 get older and mellower, I find it hard to resist 4 propositions from Mark and Bert and others. 5 probably has something to do with the fact that I tend 6 to represent more plaintiffs these days. It also 7 MR. WEISER: There is a transcript of this. 8 MR. RULE: 9 plaintiff clients. 10 the structural numbers are relevant frankly to the world 11 at large, and I think all of us know, we run into 12 clients who have heard this thing called HHIs, and so it 13 becomes a big issue for them. That's okay. I will give it to my No, what I would say is, look, to me It is relevant to them. 14 It's relevant to people who are planning, who 15 don't want to necessarily go out and hire an antitrust 16 lawyer when they're putting together two Kansas wheat 17 farms to basically say, Look, if you're under this 18 level, there is not a problem. 19 people maybe don't like this, that it ought to be a safe 20 harbor, and that may tell you that you want to set these 21 numbers a little lower, so maybe not 2,500. 22 is the right number, but again, you also have a thousand 23 in the guidelines. 24 25 That's why I say, and Maybe 2,000 I think that's what gets communicated. The problem I have with what Mark says, as I say, even For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 34 1 though I find some attraction to it is I think the 2 experience of the last 25 to 30 years has taught people 3 who do this, both inside the government and outside the 4 government, that there really are potentially a lot of 5 relevant considerations so that in some cases, I will 6 grant you that six might be too few. 7 I mean, I'm old enough I guess that I don't find 8 1,800 to be completely appalling, although I think if 9 you go from six to five, maybe that's a better area or 10 even five to four. 11 and under some circumstances that might be a problem. 12 I think though what the agencies have found is I mean, I can understand the theory, 13 there are a lot of other pieces of evidence. 14 a lot of other facts that can inform one as to whether 15 or not a merger that goes from six to five, five to 16 four, four to three, three to two, in fact is a problem. 17 If we can know the answer better than relying on 18 something like a market definition, which is not 19 perfect, and then some heuristics that don't necessarily 20 have support empirically, that's what we pay the 21 government to do, to try to get the answer right. 22 There are Unless they can tell a story that the merger may 23 substantially lessen competition in some line of 24 commerce, in some section of the country, then frankly 25 the law doesn't allow them to challenge it, but to me, I For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 35 1 2 guess we do know more today than we did 25 years ago. I think analysis has to take that into account, 3 and one of the things we do know is that structure 4 doesn't hold the same significance for determining the 5 outcome of an effect of a merger as we thought it did 30 6 years ago, and I think given that reality, the 7 guidelines ought to be changed to reflect it. 8 much as on some days Mark's proposal might make sense, I 9 think it ignores the fact that we may actually be able 10 to get it right in a particular case more often than 11 using what is a rather crude rule that Mark plays out. 12 MR. WEISER: Again, as So let me take that line of 13 discussion, and I have several different ones that I 14 want to follow but we'll follow this one: 15 industries have what you might call different minimum 16 efficient scale, meaning it's hard to sustain, let's 17 say, five competitors in certain types of industries. Different 18 The DOJ filed comments recently in the Broadband 19 Plan, noting in that broadband markets you're not going 20 to see textbook competition. 21 broadband providers for that matter. 22 You may well not see six What do you say if there are claims in an 23 industry where it's moving from, let's say, four to 24 three, and they're saying it's four and three merging, 25 and we need to be stronger. On a pure structural case For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 36 1 you might not want to allow that, but as Rick says, 2 there might be other reasons to believe that merger is 3 benign and may be pro-competitive. 4 How does that square with a concern that Mark 5 articulated about departing from what you're pre 6 committing to as a particular code? 7 start with that one, Mark? 8 MR. COOPER: Do you want to Well, yes, because I have the 9 experience of working lots of industries, and so in the 10 first year of this administration, I've had 11 conversations with the antitrust authorities over 12 airlines, railroads, newspapers, wireless companies, 13 broadband service providers, all of which are industries 14 where four would be heaven. 15 small number of competitors, and the antitrust 16 authorities lose their primary weapon, which is lots of 17 competitors, to ensuring an efficient economy. 18 We have this problem of a So I have a series of principles, five quick 19 principles, and I will file them. Basically, when I'm 20 looking at a situation, first of all, you really do have 21 to test the limits of minimum efficient scale. 22 is going to come in and say: 23 support more than two or three or four. 24 challenge that, but if it's true, you really have to 25 make sure that you get the maximum number of competitors Everyone Hey, this market won't You need to For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 37 1 you can, and that's going to be a fight about whether or 2 not the weaker of the two merged parties is viable. 3 In the case when you conclude that there is 4 going to be less than four, then you have to be really 5 worried about market power because economics teaches us 6 that the ability of the small number of players to 7 extract rents and otherwise avoid the inconveniences of 8 competition is great when there's that small number of 9 competitors. So we need to really worry about things like 10 11 artificial barriers to entry, refusals to deal, efforts 12 to monopolize neighboring markets. 13 a tremendous need to analyze small number of competitors 14 from the Agencies' point of view, both eventually 15 prophylactically setting out a policy by which you might 16 bring other cases under other sections of the Act, but 17 also as a framework for analyzing what we understand 18 about the conditions we have to put on these mergers. So, for me, there's So if I'm confronted with a four to three and I 19 20 conclude that it's a necessary outcome in terms of 21 minimum efficient scale, then I have to really worry 22 about the ways that the resulting market power would be 23 abused. 24 You will notice I stopped at four to three. 25 concepts of a dynamic duopoly or a benign monopoly For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 The 38 1 simply don't exist in my vocabulary. I just don't think 2 this Agency or the antitrust authorities in this country 3 can, in fact, be comfortable with the theories that led 4 us get down to those extremely low numbers. 5 MR. WEISER: Bert? 6 MR. FOER: 7 I don't have five points. 8 encapsulating point, and that is the principle that as 9 the level of concentration increases, the size and the 10 certainty of the offsetting benefits have to increase. 11 In other words, the higher the level of concentration, 12 the more skeptical, the more intensive the 13 investigation, the greater certainty that these 14 efficiencies are going to be there, and that they will 15 be passed on, in substantial part, to consumers. I don't think I disagree with Mark. I have kind of one It's a sliding scale. 16 It's the Heinz Baby Food 17 test where you had apparently very high level of 18 efficiency demonstrated, but it wasn't high enough 19 because the level of concentration was going to be so 20 high. 21 very difficult to become more scientific about it 22 because in part, we've created a pseudoscience. 23 I think that's the right approach, and it gets Sorry all my economist friends, but I think that 24 we've made it into more of a science than it really is 25 or it can be, and that one of the prices we pay for that For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 39 1 is a lack of intelligibility to the public in our merger 2 policies. 3 three than it is about HHIs. 4 internally inside the Beltway, but we've also got an 5 audience outside the Beltway that we've largely ignored, 6 and I don't think they are very supportive of what we 7 do. 8 9 It's so much easier to talk about a four to MR. WEISER: Maybe we can do it Andy, you can jump in here. If not, I have another question for you. 10 MR. GAVIL: No. 11 MR. WEISER: The other question goes like this: 12 We've talked now for the last half hour or more about 13 concentration broadly speaking, not differentiated 14 between coordinated effects and unilateral effects, and 15 part of what happened I think is that the '82 guidelines 16 and '84 guidelines largely were thinking about and 17 governing the concern about coordinated effects, and 18 since '92, most of the Agency's cases have been on the 19 unilateral effects side, still also invoking the HHI 20 structural presumption. 21 So let me start with coordinated effects. On 22 coordinated effects, the question would be as follows, 23 and this gets to something Rick said earlier: 24 have a structural case, say a four to three merger or a 25 three to two merger, is that enough based upon what we If you For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 40 1 know that we should be worried about, coordination which 2 I believe Bert said or Mark said is something that if 3 it's tacit coordination is not a problem under Sherman 4 Section 1? 5 that's not something Sherman 1 does anything about. 6 You can have conscious parallelism, and So is that enough of a reason to worry about it 7 or might we want to say, as the guidelines do, that 8 there are certain pre conditions we need to look at to 9 understand whether or not collusion or coordination or 10 even, as Christine said, accommodating behavior is 11 likely, and the Agencies need to have some evidence of 12 that in addition to the structural conditions? 13 you think about that coordinated effects question? 14 MR. GAVIL: Two things. How do First, this is sort of 15 a broader comment, and I'm glad we made that transition 16 because in the revisions that are being talked about for 17 the guidelines, we're trying to separate out what is a 18 dilemma of the guidelines. 19 multiple strands of intellectual history in merger and 20 economic thinking. 21 strong because of the '68 guidelines, because of the 22 influence of the structure conduct paradigm. The guidelines reflect The structural paradigm was very 23 We then introduced oligopoly theory, game 24 theory, and we've sort of layered different strands on, 25 and I think the tension that's now being addressed For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 41 1 between the structural concepts and unilateral effects 2 is a good example of it. 3 away from, in all of our discussion to this moment this 4 morning, is that all that structural stuff may not be as 5 relevant in unilateral effects cases. What you really want to get 6 That's a reflection of there being these 7 different strands, but the structural paradigm was very 8 well established in the case law. 9 established in the literature, so it got written into 10 the '68 guidelines, carried over in '82 as not to appear 11 to be a too radical departure. It was very well 12 So we have these competing strands of 13 intellectual history, and I think part of the challenge 14 in the rewrite is to explain that and separate that out 15 and explain which models work under what circumstances. Now, to get more directly to the question. 16 The 17 guidelines in essence already answered your question, 18 made that decision, that structure alone was not enough, 19 that there's a separate inquiry about anticompetitive 20 effects. 21 guidelines, but again, as I said earlier, the agencies 22 found they were being held to that when they went and 23 litigated. It's all in the same section of the 24 When you went out and said, Well, here's 25 structure, well, the statistical case is pretty much no For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 42 1 longer going to be enough, except at maybe a very high 2 level. 3 are the conditions for coordination in this market? 4 will this merger alter those conditions and facilitate 5 better coordination? You have to tell the coordination theory. That's become part of the analysis. 6 What How It is 7 clearly added to the burden of the agencies in 8 challenging a coordinated effects case. 9 should be? 10 shouldn't have to do that? Is that as it Are there some levels beyond which we Again, I think if you're going to write that 11 12 into the guidelines, two to one, do you need to show it? 13 Bert mentioned Heinz. 14 delineating what theory of anticompetitive effects is 15 there. 16 is just fine with us, and the sliding scale approaches, 17 and we're not going to really demand that, but you look 18 at cases like Arch Coal, and the Court wants to know 19 where your evidence is of coordinated effects. Heinz is not very careful about It's just saying at some point the presumption 20 The last point I would make is, yes, it is very 21 important that we use the merger laws to stop structures 22 from forming which could lead to coordination that we 23 could not reach under Section 1. 24 traditional purpose of Section 7, because we recognize 25 that oligopolistic coordination, which can't be reached That has always been a For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 43 1 under Section 1, is still bad. 2 higher price, and the only tool we really have in 3 antitrust, because we've essentially walked away from 4 the idea that Section 1 can reach interdependent 5 pricing, is stopping the structure from forming that 6 will make it easier. 7 It still results in So I think on that sense, Section 7 really does 8 provide a very important -- it goes back to the 9 incipiency idea to some degree. 10 barrier that keeps us from getting to structures and 11 problems that we can't reach under other parts of the 12 antitrust laws. It is an important 13 So I think that role is still important. 14 does, I think, require us to tell a coordination story. 15 How will this merger incrementally increase the ability 16 of firms post merger to coordinate is an important 17 question to answer, especially if we're talking about 18 six to five, five to four. 19 three to two and two to one, maybe the story doesn't 20 matter as much. 21 It Like I said when you get to We're just too scared to go there. MR. WEISER: Rick, to kind of capture Andy's 22 point, if you have let's say a four to three or three to 23 two merger where you have conditions that you would seem 24 to facilitate coordination, let's say very difficult to 25 enter, homogenous product and maybe some story you can For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 44 1 tell about how coordination happens, should that be 2 enough for a Court to, under incipiency theory and under 3 the structural presumption, be able to stop a merger? 4 Others have argued in comments that the whole 5 idea of a structural presumption and this concern is not 6 one the Agency should focus on. 7 that question? 8 MR. RULE: How do you come down on Again maybe it's a reflection of my 9 age, but I tend to agree with Andy on this one. To me, 10 one of the issues, and I think a number of commentators 11 have raised this, I think that currently the guidelines 12 are a little confusing in the use of the term 13 coordinated effects versus unilateral effects, and I 14 think you should probably get away from that. 15 I think I heard Christine say that there should 16 be a more detailed description of what an adverse price 17 effect means, and I agree with that. 18 that, and I think the evidence is consistent with the 19 fact that in some industries, for example, the 20 characteristics that you laid out, a reduction in the 21 number of competitors can raise a threat of a price 22 increase. But I also believe 23 Now, I would say that even in that circumstance, 24 one ought to be willing to look at efficiencies and that 25 sort of thing, and so I would say in appropriate For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 45 1 circumstances, that could be a basis for concluding that 2 a merger violates the antitrust laws. On the other hand, and this is I think part of 3 4 experience, but it's also part of the change in the 5 economy: 6 modern industries just doesn't work very well, and so 7 the notion that there's going to be some sort of 8 coordinated interaction in some industries, for example, 9 the information industry, is just, to me, not very 10 credible. Trying to apply that paradigm to a lot of 11 I don't think there's a lot of empirical basis 12 for that, so I think you've got to reach the conclusion 13 first that this is an industry that is likely to witness 14 tacit collusion, in the old term, before you reach that 15 conclusion. 16 The other point I would make, and I think this 17 is just an interesting observation, while I agree with 18 everything Andy said and what I just said, it's also 19 kind of interesting that the law actually, under Section 20 1, has moved in the direction of capturing more of what 21 might be called tacit collusion, leaving Twombly aside 22 and the difficulty of pursuing those cases, if you look 23 at Posner's opinion in high fructose corn syrup, there 24 are ways I think today that I would have been much more 25 skeptical about 25 years ago of actually creating an For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 46 1 inference of a conspiracy using some of the analysis of 2 Posner. 3 So that there's an argument today that maybe we 4 can reach some of that conduct under Section 1, that 25 5 years ago when all this was developed, there was kind of 6 the sense that you just couldn't find an agreement under 7 those circumstances, even though there was tacit 8 collusion going on, and so approaching it and trying to 9 stop it structurally was more important. 10 least that argument. 11 MR. COOPER: There's at I really agree with that, except I 12 don't want to call it tacit collusion. 13 it noncooperative games because I think we're talking 14 about the same thing, and I think that analysis of 15 noncooperative games is the bridge between coordination 16 and unilateral action, and he did win a Nobel Prize for 17 it, and we have spent 25 years, -- and it's almost 18 exactly 25 years that the theory has received an immense 19 amount of attention. 20 I want to call While I'm not a lawyer and haven't reviewed the 21 cases very closely, I don't think the influence of 22 noncooperative game theory has been fully felt, nor has 23 the influence, as I said, of behavioral economics, and I 24 think that that needs to get reflected, so I'm agreeing 25 with that. I just don't want to call it tacit collusion For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 47 1 because that has that old style ring to it of, there's a 2 collusion here; no, these are just people who, as the 3 lion in the movie says at the second bar room scene, 4 Adam Smith was wrong. 5 6 7 It's a wonderful line because -- MR. WEISER: Which movie are you talking about MR. COOPER: This is in A Beautiful Mind. now? I'm 8 sorry, I'm a veracious consumer of popular culture, so 9 in A Beautiful Mind, Nash is struggling with his theory, 10 and in the second bar room scene, there are nine guys 11 and nine gals, and one very pretty gal and other very 12 intelligent women, and he looks at it. 13 going to happen here, right? 14 all compete for the one good looking woman, eight of 15 them will be disappointed. He says, what's He realizes that if they 16 He then goes back and writes his theory of how 17 the nine guys will learn very quickly to allocate who 18 ought to chase whom, and the ability of a small number 19 of people to capture the monopoly rents available 20 without colluding is a really important observation to 21 which the economics discipline has devoted a great deal 22 of attention for exactly a quarter of a century since 23 the guidelines were adopted. 24 MR. WEISER: 25 I just want to point out for those who missed it, we have Mark Cooper and Rick Rule in For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 48 1 agreement, so the idea of an emerging consensus, we can 2 stop right here. 3 We can stop right here and see if we have 4 questions from the audience. I think, as Howard said, 5 we have a small enough group of folks that rather than 6 asking you to submit written ones, if there are people 7 who have questions that they want to ask, we have time 8 for a question or two, and as sort of a professor on 9 leave, I'm not afraid to call on people either. 10 Any questions folks want to ask? 11 too, but if there were any questions? 12 the back, do you have a question? 13 MR. ABBOTT: I have more Is that Alden in Yes, thank you very much. The 14 question would be directed at Rick. 15 Posner's opinion in high fructose corn syrup, but given 16 recent case law, some might argue that it's becoming 17 very, very hard to win a Section 1 case. 18 is viewed by many as a minority view. 19 He pointed out Posner's view I would say there are lots of other commentators 20 who have challenged that, so if that is the case, how 21 likely are you going to be able to pursue a Section 1 22 case, and does this get back to the notion that Section 23 7 is an incipiency statute, and because of the very 24 difficulties in approving a quote, unquote agreement, 25 despite Posner's views, Section 1 may be a less than For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 49 1 ideal vehicle? MR. RULE: 2 I mean, I don't disagree with that, 3 Alden, and as I said, I agree that under the right 4 circumstances where you believe that whether you want to 5 call it tacit coordination or some sort of game theory 6 tells you that there's going to be a likelihood that 7 prices will increase, I think that's a basis for 8 stopping a merger. 9 The only point I would make on Section 1, the 10 fact that Posner's decision is out there, I think it 11 lends credibility to an argument that frankly ten years 12 ago would not have gotten you very far. 13 principal issue on Section 1, for what it's worth, in 14 terms of being difficult to win is Twombly actually. 15 Twombly is the one that creates the biggest obstacles, 16 but that's a different panel. I think the I do think that that is one theory that could 17 18 motivate a merger challenge. 19 that structure is not by itself determinative of whether 20 or not an industry is going to exhibit that sort of 21 conduct. 22 experience of the last 30 years, and I think that's 23 really what needs to be captured by the revision of the 24 guidelines. 25 Again I think the point is You have to look at other factors. MR. WEISER: That's the Do we have another question? For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 50 MR. CARY: 1 George Cary. I guess I am finding 2 this discussion really fascinating, especially some of 3 Rick Rule's comments. 4 have is: 5 explication of the role of non price coordination in 6 merger analysis? 7 I guess the question that I would Is there room in the guidelines for a greater The guidelines seem to focus on pricing. They 8 don't seem to elaborate very much about how non price 9 competition might be the subject of coordination. 10 tends to be relegated into the unilateral effects part 11 of the guidelines, and I wonder whether there isn't room 12 for some discussion about competitors channeling their 13 competitive efforts into elements of competition where 14 consumers could be harmed, where they still compete, for 15 example, on marketing rather than on price or on some 16 forms of innovation rather than other forms of 17 innovation, or is that too big a project? 18 That I guess the sub theme here is whether the 19 unilateral effects analysis has ignored the role of 20 coordination among firms producing differentiated 21 products and whether that ought to be spelled out 22 somewhere? 23 MR. WEISER: So, George, that's a great 24 question. I was going to add, let me put my related 25 point on the question, and then I'll let the panelists For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 51 1 answer. More broadly: 2 Should the structural presumption 3 not only be tied to and motivated by a story about 4 price, but other elements of competition, be it quality, 5 product variety what have you? MR. COOPER: 6 Mark? Well, as the consumer advocate who 7 is always accused about only caring about price, let me 8 say we care about a lot more, and we always have. 9 are other two areas that are really important, one, is 10 terms of service. 11 termination fees in cell phones, wireless, for instance 12 is a really onerous condition on consumers, and there 13 will be people who will disagree with that, but we look 14 out at bundling in the cable industry as a term of 15 service. There We have got complaints about -- 16 We look out at the competition of big fat 17 bundles in the triple play, and these are key questions 18 about everybody's offering me the same package, and it 19 only serves a quarter of the market, so, yes, I think 20 the terms of service is a second area that's really 21 important in addition to the price, and then the big 22 enchilada is innovation and long-term competition. 23 We have tried very hard not to go for the 24 short-term, near term buck, so frequently people will 25 come forward and argue that, hey, the prices will be For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 52 1 lower next week, and we say, yeah, but what about next 2 year or ten years from now, so the second area is 3 innovation and long-term competitive structure. 4 The guidelines have a footnote here and there, 5 they need to be much more prominent because they are at 6 least as important, and, in the case of the second one, 7 innovation and long-term competition, probably more 8 important than price and terms of service. 9 guidelines should be oriented around that, they should 10 be forward looking. 11 I think the It's interesting, Bert talks about the 12 Congressional mandate, and I ask myself: What would the 13 Congressional mandate look like if this Congress were 14 working on it? 15 I think the most important thing they would talk about 16 is long-term innovation and production, and that 17 wouldn't be a bad thing. 18 about price and a lot more about building an economy for 19 the 21st Century, and I think that would be a useful 20 thing for the guidelines to say. Obviously they can't agree on much, but 21 MR. WEISER: 22 MR. FOER: They would talk a lot less Other comments? Bert? If I can challenge Joe Farrell for a 23 speech he once gave: Price is usually a pretty good 24 surrogate for the things that we want from competition, 25 from the market. We want fair price. We want For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 53 1 innovation. We want choice for the consumer, but in 2 some industries and in some circumstances, price is not 3 a very good surrogate. 4 For instance, right now there's an investigation 5 of a voting machine merger where the bottom line is not 6 so much what the price for a voting machine is going to 7 be as different aspects of the effects that can come, 8 including some possibly very important innovation 9 effects. 10 In information industries, we may not care as 11 much about price as we do about choice, so there's got 12 to be a loosening up that permits these other objectives 13 to become part of the analysis. 14 Exactly how you do that, George, I'm not sure, 15 but I am sure that your question is the right question. 16 How do we make certain that what we're getting out of 17 our policy are the outputs that we really want, and 18 price alone is insufficient. MR. RULE: 19 I mean, here's another one where I 20 will agree, this time with Bert. I've always kind of 21 viewed price as an easier, sort of more quantitative 22 variable to do things like understand how you define 23 markets. 24 competitive parameters can be understood or reduced in 25 some ways to price. And actually I think a number of different For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 54 However, I agree that -- and it's contrary. 1 It 2 would frankly be inconsistent with my experience 3 recently to say that the Agency should ignore other 4 effects, other than price, because my sense is that if 5 there is a reason to be concerned about sort of non 6 price elements of competition, the agencies will look at 7 it. 8 Conversely, again my experience has been that 9 where one has an explanation, that even though there 10 might appear to be some minimal price effect, if there 11 is a countervailing non price benefit like a quality 12 improvement or a technological innovation improvement, 13 the agencies will consider that. 14 I think the only thing I would counsel the 15 Department and the FTC as they go through the process of 16 doing guidelines, I think it is incredibly difficult to 17 generalize. 18 again if you explain the process, then that will help 19 counsel. 20 think there's anybody else in this room who was involved 21 in it -- who came up with the sort of structural 22 presumption for R&D joint ventures. 23 and no particular empirical reason for doing it, but we 24 were trying to come up with something to put in 25 legislative history in the old NCRA, and that's where we I think that's one of those areas where, But, I can tell you as the person -- I don't We just made it up, For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 55 1 came up with this notion that so long as it was possible 2 to create I think we said three other joint ventures, 3 equal capability, there shouldn't be a problem. 4 It sounded good. There was a logic to it, but 5 I'm not sure that -- and again it had the benefit of 6 sort of weeding out the things that probably aren't 7 going to be very interesting, but I do think that when 8 you start getting into non price areas, it's much harder 9 to make generalizations, and I personally think it would 10 be unwise to try. 11 MR. WEISER: Andy, you get the last word. 12 MR. GAVIL: I think it's hard to make 13 generalizations, but there are industries where it's 14 obvious that innovation, quality and service, those 15 three things, can be very important and are vulnerable 16 to being lost. 17 I realize we're out of time. Healthcare I think 18 is an industry where you can see lots of examples where 19 you have pressure from payers to reduce payments. 20 allow mergers. 21 is service and quality and innovation as well. 22 We One thing that could get lost in the mix I think there have been some examples of that, 23 so you do, I think, have to go beyond price. Whether or 24 not -- and I think George's narrow question is whether 25 concentration has really been linked to losses of non For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 56 1 price competition. 2 to support that, but I think it is important for the 3 agencies to not just function on price, and I think it's 4 easy to identify industries where these other components 5 of competition are especially important and are 6 vulnerable to being lost. 7 MR. WEISER: 8 9 I don't know that there are studies I want to thank our panelists for a great discussion. (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 57 1 PANEL 2: PRICE DISCRIMINATION/POWER BUYERS. 2 MODERATOR: 3 Economics 4 PANELISTS: 5 SUSAN CREIGHTON, Partner, Wilson, Sonsini, Goodrich & 6 Rosati 7 MARC SCHILDKRAUT, Partner, Howrey, LLP 8 JOE SIMS, Partner, Jones Day 9 JOHN THORNE, Senior Vice President and Deputy General 10 Counsel, Verizon Communications, Inc. HOWARD SHELANSKI, Deputy Director, Bureau of 11 MR. SHELANSKI: 12 Okay. Well, I would like to 13 welcome you to our panel on price discrimination and 14 powerful buyers, and we have a wonderful panel, as we do 15 really throughout the day. 16 exception of one member of our panel, has both serious 17 private antitrust as well as government enforcement 18 experience, and the one, John Thorne, who does not, has 19 vast experience being pursued by public enforcement 20 agencies, so this is really a very fit panel for this 21 topic. 22 Everybody here with the, I would just like to briefly introduce the panel 23 and then open up with a couple of questions. 24 immediately to my left is Susan Creighton, who is a 25 partner at Wilson Sonsini, and a former Bureau of Seated For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 58 1 Competition Director here at the Federal Trade 2 Commission. 3 To the left of her is Marc Schildkraut, a 4 partner at Howrey, who has had a very distinguished 5 career in both private practice and public enforcement 6 and is another FTC alum. 7 Joe Sims is well known to everyone as a leading 8 antitrust partner at Jones Day, who spent a long part 9 his career at the Department of Justice, Antitrust 10 Division as a Deputy Assistant Attorney General, and 11 then John Thorne, who is senior vice president and 12 deputy general counsel of Verizon, who has been a 13 contributor on many panels through his writing, and also 14 as a litigant in many regulatory and antitrust matters. 15 I would like to start with a very broad question 16 for our panel, which is how the Agency should judge a 17 merger's effects on price discrimination? 18 and criteria are relevant to judging a merger's effect 19 on price discrimination? 20 with you? 21 MS. CREIGHTON: And, Susan, why don't we start Sure. It seemed to me that 22 really was sort of two questions. 23 and the other is: 24 relevant? 25 What evidence One was the criteria, What kinds of evidence should be In terms of the criteria, I should say by way of For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 59 1 preface that it's interesting how often, particularly in 2 say high technology markets, price discrimination is 3 almost always the first thing staff is looking for, so 4 it comes up all the time, and because it comes up all 5 the time, I think that the criteria that really should 6 be used is whether you are able to identify the infra 7 marginal customers, and can you engage in price 8 discrimination? 9 price discrimination, and third is: Is it profitable? 10 So those are the criteria. I think that's the What's the mechanism for engaging in 11 easy part. The question is just how heavy should the 12 burden of evidence be to go from sort of just presuming 13 that, gee, you should be able to discriminate between 14 the buyers and actually having to prove it? 15 It seems to me that the evidence should have to 16 be relatively compelling that you actually would be able 17 both to identify the customers, and that they would have 18 no means of avoiding having sort of some recourse, 19 whether arbitrage or something else. 20 kind of econometric evidence that it would be 21 profitable. 22 And also, some So just to use a high tech example, I would say 23 that the ability to price discriminate is -- I don't 24 know if John Baker is here, but I think he had used the 25 great example in an article a long time ago, something For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 60 1 like a unicorn or a white tiger. 2 I guess I would say it's neither. It's probably 3 just a regular tiger, that it can be found in some 4 places in nature relatively frequently, but the things 5 that you need to be looking for are, for example, how 6 does the seller relate to the buyer? 7 reseller channel, or is it direct contact with the 8 particular customers? 9 be: 10 customer, or is it extensive hands on, deep knowledge of 11 the individual customer? Is it through a Then, further, an example would Is it just sort of an infrequent dealing with the 12 In technology, for example, with heavy supply of 13 services, you have people on the premises all the time, 14 then all of a sudden it starts to become plausible that 15 maybe you actually do have some ability to know the 16 ability of the customer to have some kind of ability to 17 avoid price discrimination or not. So that would be the kind of evidence that I 18 19 would be looking for. I think that kind of thing can 20 bedevil agencies trying to figure out why one customer 21 likes one thing and not the other, so I think, for 22 example, the SunGuard case was probably a great example 23 of that. 24 It wasn't possible to draw a circle around 25 saying, well, it's the big customers that can self For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 61 1 supply, or it's sort of this type of customer versus 2 that type of customer, but because of the nature of the 3 customer relationships, I think if price discrimination 4 had been better understood and better supported in the 5 merger guidelines, that might have been an easier case 6 to say just because we have 60 declarations and they 7 have 80 declarations, that doesn't mean you just throw 8 up your hands. 9 It may mean, in fact, that there are 60 10 customers that, in fact, the Agency knows don't have 11 alternatives that self supply. 12 So bottom line, I guess what I would say is I 13 think there needs to be not only plausible but 14 demonstrable evidence that would tell a story about how 15 it is that you actually would be able to engage in that 16 kind of price discrimination, so it isn't just a story. 17 Of course it would be great if there's evidence 18 that supports that where you can show through 19 econometric evidence or otherwise that in fact that kind 20 of price discrimination already has been going on. 21 can to it, but I thought, for example, that was the 22 thrust of the econometric evidence in Oracle. 23 was to show that there had already been that kind of 24 price discrimination evidence. 25 MR. SHELANSKI: Thanks. We I think Marc? For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 62 MR. SCHILDKRAUT: 1 I think you need to go back 2 one step and ask yourself something more about price 3 discrimination, because price discrimination is 4 basically pervasive. 5 time someone uses a coupon, that's price discrimination. 6 Any time someone turns around and goes to a movie 7 theater and has a child or a senior citizen with them, 8 they'll probably all be at different prices. Just to give you examples, any Almost every airline discriminates left and 9 10 right no matter whether they have market power or not, 11 so this is a quite pervasive thing, price 12 discrimination, and the problem in the guidelines with 13 using price discrimination is you can make millions of 14 markets. 15 arbitrary, particularly to a Court more used to general 16 criteria, to all of a sudden have a case where you say, 17 we're going to identify this group of customers that can 18 be targeted. 19 It all seems very arbitrary. It can seem very That becomes very, very difficult because you 20 can slice and dice 500 different ways, and being able to 21 do that suggests to me that you actually need to be more 22 rigorous when you have a theory of price discrimination 23 to define a market than when you have a general theory 24 of a market definition, and that further means to me 25 that Susan's last remark was very important, which is For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 63 1 ongoing evidence of price discrimination is important. 2 It's important to show that that ongoing 3 evidence is not the kind of evidence that relates to 4 simply the every day kind of price discriminations which 5 I was talking about, which is pervasive, but there are 6 industries where we're talking about something else, and 7 those are usually multiple players. 8 There's systematic price discrimination ongoing, 9 particularly if you're dealing with a fungible 10 commodity, and then you need to ask yourself the 11 question: 12 possible that an industry like that can really price 13 discriminate? 14 What the heck is going on here? How is it Typically, when I was back at the FTC and I 15 would ask questions like that in depositions, I would 16 say, Well, why are you doing this, I mean, wouldn't you 17 be better off shaving the price to the people who are 18 disfavored, and you can make more money? 19 answer I got was, Well, if I did that, everybody would 20 do that, and how would I ever be better off? 21 when I knew that I had something I had to think about 22 much harder because that was indicating to me that the 23 propensity to price discriminate was actually 24 meaningful. 25 Usually the That's If you have something that's that meaningful and For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 64 1 you can then turn around and prove something like that 2 to a court, then you have a theory that's workable and 3 something that you can do something with and something 4 that doesn't seem arbitrary. 5 thing is to avoid this potential for arbitrariness, and 6 that requires not only having something special, it 7 requires being able to target. 8 possible. 9 I think the most important It means no arbitrage is All those different things need to come into 10 play, and one more thing that needs to come into play: 11 I think you have to think about your underlying theory 12 when you're doing this. 13 you're dealing in a case that is a coordinated 14 interaction case, it is very possible that, unlike most 15 cases, a small fringe firm that couldn't really increase 16 its output is going to be able to undermine that 17 collusion very, very easily because all it needs to do 18 is to shift to the disfavored customers. 19 What I mean by that, is that if It doesn't have to produce another unit, so I 20 think all these things need to come into play. I 21 probably answered all of your questions at once in doing 22 this, but I think all of those things need to come into 23 play when you're thinking about defining markets in 24 terms of price discrimination. 25 MR. SHELANSKI: We have a couple follow ups, but For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 65 1 I want to hold off on them until we hear from Joe on 2 this. MR. SIMS: 3 Well, first I think I feel compelled 4 to correct the historical record as laid out by Mr. 5 Cooper. 6 movie called Ghost Busters, and there was a great scene 7 in Ghost Busters where Sigourney Weaver, possessed by 8 the demon, is pursuing Bill Murray throughout her 9 apartment, and he's trying to resist her, and finally 10 she tackles him on the bed in the bedroom, and he says, 11 Wait a minute, wait a minute. 12 absolute hard and fast result, no fraternizing with the 13 customers, and then he looks directly at the camera and 14 says, Well, actually it's more of a guideline. 15 think that's the origin of the Pirates of the Caribbean 16 remark. 17 For those of you who are my age, you remember a He says, We have an So I A theme that will run through my comments today 18 is practical versus theoretical. 19 criteria and evidence? I would start with: 20 parties doing it today? If it's not happening pre 21 merger, then there needs to be a really compelling story 22 about why it's going to happen post merger. 23 You asked: What's the Are the Even if it is happening pre merger, there should 24 be some explanation of why the merger is going to make 25 it worse or why the merger is going to make the effect For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 66 1 2 of that price discrimination worse. This seems credible only in some pretty limited 3 circumstances. If it doesn't happen pre merger, how is 4 the merger going to change that fact? 5 is a potential for this, will the merger cause some 6 dynamic changes by customers to minimize that risk? Indeed if there 7 Buyer statements, declarations, having obtained 8 declarations on both sides of this question from buyers 9 in multiple matters, I'm not very enthusiastic about 10 their probative value. 11 many customers don't really know what their options are 12 until they're incentivised to think about them. It is I think pretty common that 13 Inertia plays a very strong role in business 14 behavior, and until they've actually been forced to 15 examine the possibilities, a lot of people will 16 automatically revert to the: There isn't really a 17 realistic option available. There's a lot of laziness 18 in preferring the status quo, so you ought to have some 19 evidence that these concerns are real as opposed to just 20 the statement of the concerns. 21 the same test to statements contained in documents or 22 otherwise from the merging parties. Indeed, I would apply 23 Anybody who is an experienced practitioner in 24 this field knows how often they get deeply involved in 25 looking at a merger and come to the conclusion that one For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 67 1 or both of the parties don't understand their business 2 as well as you would expect them to. 3 The facts, as evaluated by somebody who knows 4 how to look at it from an antitrust perspective, as 5 opposed to a business perspective frequently lead you to 6 different conclusions than the parties reached on their 7 own without the advice and input of that training. 8 So sometimes companies don't know they have 9 market power. 10 the FTC would accept that as a defense, and the other 11 way ought to work too. 12 that they have market power doesn't mean they have 13 actually. 14 for hard evidence as opposed to conclusionary assertions 15 by either side. 16 I don't think the Antitrust Division or The fact that they say or think So my general point here is you need to look MR. SHELANSKI: John, as one of these 17 representatives of one of these high tech companies that 18 Susan alluded to, have you now or have you ever engaged 19 in price discrimination? 20 MR. THORNE: I withdraw the question. That's a great question. If you've 21 seen any of the recent Verizon television commercials 22 like during the NFL playoffs, you see the guy come out, 23 and there's a big white sign, and it says $99 for this 24 package of all the voice calls you want to make in a 25 month, and he flips around the 9 to become a 6. On For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 68 1 national television in front of all the NFL viewers, our 2 pricing went from $99 to $69. 3 When you have large economies of using national 4 advertising, it's hard to target Albuquerque for a price 5 increase that's different than that. 6 the transaction costs often overwhelm any desire on a 7 tiny market basis to price discriminate, so in echo of 8 what Joe said, the practical constraints make it more 9 difficult to price discriminate than some of the 10 theoreticians would anticipate. 11 MR. SHELANSKI: In the real world, Let me ask a follow-up to that. 12 Would a hypothetical telephone company that had merged 13 seriatim with a number of other hypothetical telephone 14 companies have found its ability to engage in this kind 15 of price discrimination be affected by those 16 transactions, and if so, in which direction? 17 MR. THORNE: Telecom is a hard industry to talk 18 about as an example because some price discrimination 19 increases output. 20 don't think if you can fly an airplane today that didn't 21 have differently priced seats and still fill up all the 22 seats, so there's some industries where price 23 discrimination may be output increasing, and in telecomm 24 in some aspects may be that. 25 It allows you to build a system. I think the general trend, if you look at For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 I 69 1 telecom mergers, for example, in wireless or wire line 2 has been to unify, not to fragment the pricing 3 structure. 4 MR. SHELANSKI: Right. That actually leads very 5 nicely to a follow up question I would like to ask the 6 whole panel. 7 different kinds of effects that price discrimination can 8 have for consumers. 9 It gets exactly to this question of the Certainly the fact that some people are willing 10 to pay an enormous amount of money for business class 11 may enable the airline to offer some very cheap fares in 12 the back of the plane and fill seats that otherwise 13 would not have been filled, so I guess the question I 14 would like to follow-up with, we'll start with Susan 15 again and just work down the line is: 16 merger investigation does produce the kind of compelling 17 evidence that you and Marc and Joe have talked about of 18 price discrimination or of an increased ability to 19 engage in price discrimination, how should the agencies 20 balance harms to vulnerable groups of consumers against 21 possible benefits to other consumers? 22 MS. CREIGHTON: If a I think that's a great question 23 because I think that, maybe just to step back on the 24 question of whether you call it sort of a localized 25 effect within a larger market or a price discrimination For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 70 1 market or a sub market, you really are asking or 2 potentially getting to the point where you're saying: 3 For this group of six customers, there's the ability to 4 price discriminate, but it could very well be that 5 that's in the context of a merger where the other 94 6 customers really want to be able to have the integration 7 that you're now going to be able to supply, these six 8 don't need it. 9 So I guess it seems to me that as much as the 10 agencies have resisted historically the notion of very 11 narrowly defining efficiencies and sort of offsetting 12 pro-competitive effects, that you can't say benefits in 13 this market can't be offset by benefits or sort of 14 detriments in that market, it has to be merger specific, 15 sort of all this very narrow defining down of what kind 16 of benefits will credit to the merger. 17 I think the concomitant of saying, Yes, we will 18 look at -- and it may in fact be sufficient for us to 19 challenge a merger if there are these localized 20 competitive effects, that it's incumbent on the agencies 21 simultaneously to step back and broaden their view with 22 respect to the offsetting competitive benefits. 23 MR. SCHILDKRAUT: Yeah, I would agree with that, 24 and I might want to go a step further than that. 25 have, in the efficiencies section of the guidelines now, For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 We 71 1 something that says we're not going to trade-off 2 different markets, but if you have price discrimination 3 markets, you can have millions of markets, and any 4 individual consumer could be an individual market on 5 that basis. 6 If a million consumers are going to benefit and 7 one is going to be harmed, in theory under the 8 guidelines, you have an anticompetitive effect, and the 9 guidelines are telling us we must prevent that from 10 happening. 11 agencies actually practice. 12 I think that that is not the way the The agencies do make trade-offs under those 13 circumstances. They don't announce them that way, but 14 if the guidelines are going to be honest about this, we 15 ought to look at these trade-offs and think about 16 whether you want to bring a case where the 17 pro-competitive effects to most consumers outweigh the 18 anticompetitive effects to some. 19 MR. SHELANSKI: Would you envision doing this 20 within the effects analysis, or would you envision the 21 pro-competitive aspects as something that would come in 22 under the efficiency analysis? 23 MR. SCHILDKRAUT: Well, it could come under the 24 efficiencies analysis, but it doesn't have to be a 25 traditional efficiency. It is standard analysis that it For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 72 1 is ambiguous whether price discrimination is going to 2 lead to adverse welfare effects, so it could very well 3 be, I suppose, that in a merger, you're going to get the 4 better ability to price discriminate, but the better 5 ability to price discriminate could lead, on average, to 6 lower prices. 7 That's not what one would normally think of as a 8 standard efficiency analysis, so my answer to that is it 9 depends on what kind of effect you would have as to 10 where you would balance it, but in either case, I don't 11 think you should let the guidelines where tradeoffs are 12 verboten unduly effect the analysis where we think there 13 is going to be positive welfare effects from the 14 acquisition. 15 MS. CREIGHTON: If I could just maybe interject 16 something Marc says triggers, and this is a bit of a 17 detour, but one of the issues that you have is the 18 question of who your audience is for the guidelines, and 19 I guess I would encourage you to be thinking about 20 District Courts as your audience, over and above 21 everybody else, and one of the benefits that the 22 guidelines have had is a tremendous amount of buy in 23 from the court system. 24 To preserve that, I think if you're going to 25 have buy in on the notion of sort of more localized For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 73 1 competitive effects, you are going to have to sell this 2 as reasonable. 3 to sort of the common sense kinds of reactions you're 4 going to get from judges along the lines of what Marc 5 was describing. You're going to have to sort of respond If you're trying to say: 6 Under our guidelines 7 we don't have to look at the fact that overall our 8 prices are going to go down, you're just not going to 9 get the kind of buy in that the '92 guidelines had. 10 So just at a very practical level, I think it's 11 important to be taking cognizance of that, that this has 12 to be a realistic and accord at some level with the 13 intuitions of general stretches. 14 Sorry. 15 MR. SHELANSKI: 16 MR. SIMS: Joe? Let me just first follow-up on that 17 point. 18 seems to me are two: 19 world who are trying to figure out how to look at 20 mergers based on the way the government will look at 21 mergers, number 1, counselors, internal and external 22 counselors, and number 2, the courts. 23 I mean, your basic audiences for guidelines it All those folks out there in the I agree with Susan, the courts are a lot more 24 important than the counselors. 25 figure it out over time. The counselors can The courts will hold you to For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 74 1 what you say and what you layout as the standards. 2 only will they but they should. 3 boundaries are there? 4 boundaries, so there is a tension. 5 Not Otherwise, what The statute provides very little There's a tension between trying to write 6 guidelines that are descriptive, in fact, of what the 7 agencies do and descriptive enough so that people can 8 actually figure out what the agencies do in some detail 9 on the one hand and writing guidelines that will 10 actually be followed and useful to courts when you're 11 dealing with that. 12 Now, most mergers don't go to courts, and so a 13 reasonable person could say: 14 over counselors? 15 It's not the guidelines, and it's not the agencies. 16 It's the courts, and those are, to me, the most 17 important audience. 18 Well, why emphasize courts The courts set the rules in the end. The other sort of side point to this, and I'll 19 come back to your basic question, is: Economics is 20 critically important in intelligent analysis of mergers, 21 and for that matter almost anything else in antitrust, 22 but courts deal in English. 23 and so you can't really in my view assert economics, an 24 economic analysis only or primarily as the basis for 25 challenging a transaction. They don't deal in math, For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 75 1 Economic analysis has to be supportive of an 2 intelligible and pervasive competitive effects story. 3 You have to tell a story. 4 sitting on the bench. 5 are like me, who have to have economics interpreted to 6 them by intelligent economists, and what happens in 7 trials is we have our intelligent economists, and you 8 have your intelligent economists, and the court sits up 9 there and says, I don't have a clue which one of these 10 is right and they wash, and you end up with a decision 11 based on something else, so I apologize for the 12 divergence. 13 There are damn few Posners More of the people on the bench On the question of how do you balance, I think 14 this is actually the single most important question 15 that's connected to price discrimination, and I think I 16 agree with both Susan and Marc, if I understood them. 17 You really can't, as a practical matter, expect 18 to be successful in challenging transactions which have 19 apparent anticompetitive effects only on very small 20 audiences and positive or neutral effects on much larger 21 audiences. 22 A court is going to look at this not as an 23 exercise in trying to find the group of consumers who 24 might be injured, but they're likely to look at it as an 25 exercise in figuring out whether this transaction is in For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 76 1 the public interest from an antitrust perspective? 2 net good or a net bad? 3 94 does not come out to net bad. Is a To use Susan's example, 96 plus Now, the commentary recognizes that the agencies 4 5 will deal with this thing. That Guy Bakery case that's 6 noted in the commentaries talks about maybe there was 7 some anticompetitive effects on some institutional 8 customers, but there were procompetitive effects on 9 everybody else, and the institutional customers are only 10 20 percent of the customers, and the efficiencies were 11 uniform across the board, and so we didn't challenge the 12 transaction, even though there arguably was a basis for 13 challenging it. That is the kind of analysis that I think the 14 15 agencies have to do, and more importantly, it's the kind 16 of analysis that they have to be prepared to defeat, if 17 they don't do that analysis and try to go to court to 18 protect that 20 percent or in many cases, that 1 or 2 19 percent of the potential audience. MR. SHELANSKI: 20 21 MR. THORNE: I thought you were about to switch to a new topic. MR. SHELANSKI: 24 25 John, do you want to comment? 22 23 Thanks. topic. I'm about to switch to a new We may look back and follow-up on some of this, For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 77 1 but I would like to switch to our distinct but related 2 topic of large buyers, and I would like to start with 3 you on this, and I have a couple of different questions. 4 Why don't we start with this one, which is 5 basically: How powerful buyers , and I actually want to 6 use the term powerful buyers instead of large buyers. 7 MR. THORNE: 8 term, but go ahead. 9 I can tell you that's the wrong MR. SHELANSKI: That can be part of your answer. 10 I would like to hear your thoughts on that, but how 11 should powerful buyers factor into the analysis of 12 competitive effects, and specifically how should 13 agencies determine whether powerful buyers will protect 14 all buyers or just themselves? 15 dictate the market price or just their own price? 16 MR. THORNE: To what extent do they That's a good question for me 17 because most of my experience, most of Verizon's 18 experience with the agencies is as a buyer. 19 frequently called by Agency staff about other people's 20 mergers, and we do a little bit of merging ourselves, 21 but most of our interactions is in the context of: 22 You've been named as one of the 20 largest buyers of so 23 and so's product, they're merging, can you put somebody 24 on the phone that can explain the jargon of the industry 25 because nobody knows what these products are or even We're very For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 78 1 know who these various participants are. 2 Then usually the interview goes a bit deeper. 3 They ask for the views of a customer, and I think the 4 views sometimes matter. 5 and some papers by Joe Farrell on how often customer 6 viewpoint is important, but the important question comes 7 after that, and that's: 8 through? 9 What are you going to do to protect yourself? 10 do anything? There's a paper by Ken Heyer Well, what if the merger goes What if we, the Agency, allow it to happen? Can you Who would you turn to? 11 There we usually -- I don't think Verizon is 12 unique in this, but usually we have something to say. 13 Occasionally I get a guy on the phone that's being 14 interviewed, and I don't know, I'm expecting you to 15 block it, but often as a buyer, there's a strategy for 16 dealing with a merger or with anything that might 17 threaten the price increase or a change in terms of 18 dealing. 19 I'm tempted to tell a story. Maybe I'll tell it 20 quickly and then get away with it. Just on the 21 visibility into this thing over the whole period, when 22 the Bell System was broken up, you had the Baby Bell 23 Companies freed from buying Western Electric Gear. 24 had always for -- not a whole hundred years, but for 25 most of the century been buying their house product, and For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 They 79 1 all of a sudden, like being subject to a merger to 2 monopoly, you're freed to try to do something else. 3 You've been buying Western Electric but now 4 you're allowed to do something else. 5 immediately went to Canada and bought Nortel, the 6 supplier there to the American market, so there were 7 two. 8 The Bell Companies Bell Atlantic, the precursor to Verizon, wasn't 9 happy with just two suppliers, and went to Germany. 10 Siemens, which was on a totally different standard than 11 the North American telecom standards, agreed with some 12 nudging and promises, to bring its gear to the North 13 American market. 14 gear, get it tested, guaranteed enough purchasing to 15 make it worth it as well. 16 Bell Atlantic helped to qualify the Now, we had three suppliers. Lucent, the 17 renaming of Western Electric, felt threatened and 18 retreated to a strategy of: 19 locked in supply, let's milk it for all its worth, let's 20 make it hard to or expensive to get increased capacity 21 and new features on the locked in devices. 22 Atlantic again, with the help of Bellcore, the standards 23 groups of the Bell Companies, set standards to break the 24 points at which we were locked so we could add capacity 25 in other people's gear that connected through a standard Well, I guess we have some Bell For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 80 1 interface, add features to get another box that was 2 connected through another standard interface. 3 Lucent, again I'm just going to tell you what 4 the allegations were because that case -- although the 5 trial started but we didn't finish, Lucent tried to 6 thwart the standards. 7 East Texas. 8 I brought an antitrust case in We settled. Moral of the story was that the strategy of 9 opening up what was at one point a total hundred percent 10 self supply to multiple competitors and evading even the 11 lock in on the residue, the strategy succeeded and Bell 12 Atlantic, Verizon was not the only beneficiary, but tiny 13 Seelex, the competitive local telephone companies under 14 the '96 Act, entered the market getting all sorts of 15 cheap product as a spillover from probably a group of 16 sophisticated buyers doing the work of opening up that 17 market, attracting supply, supporting and qualifying new 18 entrants; then with standard setting and redefining the 19 product, making it possible to have mix and match 20 capability for the things you had to add to the locked 21 in piece of it. 22 That's the long story I meant to make short, but 23 let me just outline how I think buyers are important. 24 First, as a matter of fact, buyers often can 25 self protect, and if you think in symmetry terms, this For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 81 1 is a problem with being a math major, the guidelines 2 spend so much time on what other suppliers do. 3 enter to discipline a merger? 4 The other half of the transaction, what can they do to 5 self protect? Can they Well, what can buyers do? 6 We submitted some comments that cite some of the 7 relevant articles on the subject, but there's one thing 8 we missed and I want to point out. 9 the year after the '92 guidelines, in the '93 Antitrust 10 Law Journal Winter Edition, has a wonderful little 11 article canvasing the ways that buyers can defeat 12 oligopoly pricing. 13 Mary Lou Steptoe, So the first point is the buyer's side of the 14 market is important. 15 my own experience, the agencies seem to be looking at 16 whether buyers can self protect. 17 the interview phone call and say, yeah, we think we can 18 take care of this, that helps an Agency decide not to 19 challenge. 20 The second point, and this is just If the buyers answer The third thing, recently courts have begun 21 taking seriously the buyer's self protection. For 22 example, Verizon supplied a witness for the DOJ case 23 against Oracle PeopleSoft. 24 these witnesses seem like they can take care of 25 themselves and disregarded their concern about the The good Judge wrote that For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 82 1 merger. So I think the merger guidelines, as presently 2 3 written, insufficiently reflect what the courts are 4 doing, what the Agency staff is doing and the importance 5 of this. 6 sort of come back to price discrimination. 7 Now, that's not the end of the story, and we The fact that some buyers can protect themselves 8 doesn't mean that all buyers can protect themselves or 9 the spillover is perfect. 10 protect themselves. The same skepticism that applies on 11 the supplier side: Will entry be timely, likely, 12 sufficient, you can apply a similar skepticism to a 13 story about buyer self protection, but it's still an 14 important element of how the markets or some markets 15 tend to work, and I think it deserves some attention in 16 the guidelines, the way the supplier side entry stories 17 get attention. 18 MR. SHELANSKI: Not all buyers can always Okay. I've got some follow ups, 19 but I think before we go to those, I would like to hear 20 from the rest of the our panelists. 21 some thoughts on this? 22 MR. SIMS: Yes. Joe, do you have I guess the only thing I would 23 add to what John said, all of which I agree with, is 24 that the key question to me is not so much can some 25 buyers protect themselves, but can most buyers protect For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 83 1 themselves? There are lots of different ways to protect 2 yourself. 3 consumer products, and when we're talking about the 4 reaction of distributors like supermarkets or the 5 Wal-Marts of the world, and they sell those distributors 6 30 or 40 different products, misbehaving with respect to 7 one product creates serious dangers with respect to 8 other products. I've done a lot of work for people who make I know a lot of economists find that nonsense, 9 10 but as a practical matter, it is real and business 11 people believe it, so that constrains their behavior 12 because these people are important buyers. 13 the term large buyers because large really isn't the 14 issue. 15 the competitive environment in which they operate? 16 I don't like It's how much bargaining power do they have in Same thing can happen on geographical 17 differences. You're selling to people in multiple 18 geographies. You have the ability to exercise market 19 power in one but not in the others. 20 in one, does that cause them to change their behavior in 21 another, or the fact that the buyer has some strategic 22 importance to the seller? 23 some ways. 24 25 If you exercise it It's a validating buyer in So there's lots of different ways that this can happen, but the critical question, as John says, is: For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 84 1 How far does that reach? We get back to price 2 discrimination in the end. 3 differently with respect to some meaningful group of 4 consumers that don't have the same leverage, bargaining 5 leverage that this one group has? 6 MR. SCHILDKRAUT: Can you, in fact, behave Often this issue is self 7 correcting. 8 is very difficult, and in fact you have a group of firms 9 that can attempt to coordinate, to target some 10 unfortunate, small buyers while they can't target the 11 large buyers. 12 those circumstances is some of our colluding firms are 13 going to end up much better off than others, because 14 unless they can perfectly allocate the customers, 15 everybody is sharing equally, the seller who ends up 16 selling more to the big buyers and less to the small 17 guys who are at higher prices is going to turn around 18 and say everybody else is doing better than I am, and 19 that seller is going to start cheating. 20 What I mean by that is, let's say arbitrage Typically what's going to happen under And because he's just not doing as well, so 21 coordination becomes very hard when you're dealing with 22 big buyers versus small buyers, and so you have to watch 23 out and make sure you're actually dealing with a stable 24 situation, even if it looks like small firms can't 25 protect themselves as well, because they may not be able For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 85 1 to, but the market may end up protecting them. 2 Go ahead. 3 MS. CREIGHTON: 4 MR. SCHILDKRAUT: No. I thought you were done. On the other hand, one other 5 point I want to stay in the opposite direction, I'm not 6 saying it's impossible that you could have a stable 7 situation like that and you have to, under those 8 circumstances if you're the Agency, watch out for what 9 large buyers are saying to you because if the large 10 buyers think they can do better off than the small 11 buyers can, they may not want to say that because there 12 may be benefits to them because the price will stay high 13 downstream, and they're getting the benefit and they may 14 think the merger is good simply because it's 15 anticompetitive and they're going to be able to share in 16 the anticompetitive effects. 17 Go ahead. 18 MS. CREIGHTON: I was going to mention on a more 19 pedestrian level, I think it may be that power buyers 20 are the issue most often raised by the parties about 21 which the guidelines are completely silent, and I would 22 certainly say that after entry, my guess is power 23 buyers, you would have to poll the staff, is the defense 24 invoked most often as a defense. 25 So although I had previously said I think the For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 86 1 principal audience for the guidelines is the courts, I 2 said that because I think there are other ways of 3 educating practitioners, as Joe was suggesting. 4 commentary was an effort at that. 5 speeches. 6 educate practitioners, but it is my impression that 7 powerful buyers get invoked way too often by the parties 8 and rejected out of hand way too often by the staff, 9 whereas entry I think staff really agrees is an issue. 10 I think their skepticism about power buyers is 11 reflected in the commentary, and that that was an undue 12 skepticism. 13 infrequent solution if you otherwise have a competitive 14 problem, but it happens with some frequency, and so sort 15 of things like I think you were just mentioning, Howard. 16 I would be curious why John thought power buyers was the 17 wrong term, but certainly I think many practitioners 18 tend to equate power buyers with large buyers. The You can give There's a lot of other things you can do to The power buyers are going to be relatively 19 You always see it when it's like state and local 20 government, oh, well they're large, but they may be the 21 perfect example of a customer who can't in fact on a 22 particular kind of purchase defend themselves. 23 it would behoove the agencies to follow the example in 24 this respect of the European Commission, and 25 specifically address power buyers. I think For The Record, Inc. (301) 870-8025 - www.ftrinc.net - (800) 921-5555 87 1 Going back to my final argument about the courts 2 being your ultimate audience here. 3 this often before the agencies, eventually it's going to 4 start showing up in litigation as well, and you will 5 want to have staked out your ground for when it is that 6 the power buyers are sufficient and hence, why in a 7 particular case that condition hasn't been met. 8 MR. SHELANSKI: If it's coming up Before I get with John on what 9 the right term is, let me just ask a question that 10 follows very quickly from that. 11 John and Sue and Marc is a suggestion that there are 12 certainly circumstances where buyers are, we call them 13 powerful buyers, large buyers, who obviously can't 14 protect themselves. 15 that too large of a presumption. 16 all of merger analysis, and then you brought in these 17 buyers with the suggestion that the existence of a 18 powerful buyers could be a defense. 19 What I was hearing from We obviously don't want to make Otherwise you swallow Is your thought that powerful buyers should be 20 elevated in the guidelines to the level of efficiency as 21 sort of a defense, or more along the lines that I 22 thou [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%3A05957757fb1fed5b. Public record. Not legal advice.