Monopsony in Law and Economics

This page intentionally left blank

Monopsony in Law and Economics

Most readers are familiar with the concept of a monopoly. A monopolist is the only seller of a good or service for which there are not good substitutes. Economists and policy makers are concerned about monopolies because they lead to higher prices and lower output. The topic of this book is monopsony, the economic condition in which there is one buyer of a good or service. It is a common misunderstanding that if monopolists raise prices, then monopsonists must lower them. It is true that a monopsonist may force sellers to sell to them at lower prices, but this does not mean consumers are better off as a result. This book explains why monopsonists can be harmful and the way law has developed to respond to these harms. Roger D. Blair is Walter J. Matherly Professor of Economics at the University of Florida, where he has taught since 1970. He received his Ph.D. from Michigan State University. Professor Blair is the author or coauthor of numerous books, including Antitrust Economics (with David Kaserman), Law and Economics of Vertical Integration and Control (with David Kaserman), Monopsony: Antitrust Law and Economics (with Jeffrey Harrison), Intellectual Property: Economic and Legal Dimensions of Rights and Remedies (Cambridge University Press, with Thomas Cotter), The Economics of Franchising (Cambridge University Press, with Francine Lafontaine), and Volume II of Antitrust Law (with Herbert Hovenkamp, Christine Durrance, and the late Philip Areeda). He is also the editor or coeditor of many volumes, including Proving Antitrust Damages. Professor Blair has written more than 170 articles or chapters in professional economics journals, law reviews, and books. Jeffrey L. Harrison is the Stephen C. O’ Connell Chair and Professor of Law at the University of Florida College of Law. He received his M.B.A. and Ph.D. from the University of Florida and his J.D. from the University of North Carolina. He has held teaching positions at the University of North Carolina at Greensboro, the University of North Carolina at Chapel Hill, the University of Texas, the University of Houston, and the Sorbonne, Paris. Among the books he has published are Understanding Antitrust and Its Economic Implications (with E. T. Sullivan); Law and Economics in a Nutshell; Law and Economics: Positive, Normative, and Behavioral Perspectives; Law and Economics (with Jules Theeuwes); and Regulation and Deregulation (with Thomas Morgan and Paul Verkuil).

Monopsony in Law and Economics Roger D. Blair and Jeffrey L. Harrison University of Florida

CAMBRIDGE UNIVERSITY PRESS

Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo, Delhi, Dubai, Tokyo Cambridge University Press The Edinburgh Building, Cambridge CB2 8RU, UK Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9780521762304 © Roger D. Blair and Jeffrey L. Harrison 2010 This publication is in copyright. Subject to statutory exception and to the provision of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published in print format 2010 ISBN-13

978-0-511-90236-9

eBook (NetLibrary)

ISBN-13

978-0-521-76230-4

Hardback

ISBN-13

978-0-521-74608-3

Paperback

Cambridge University Press has no responsibility for the persistence or accuracy of urls for external or third-party internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

To the memory of my parents – Duncan and Eleanor – who gave me so much. –Roger D. Blair

For Sarah, McCabe, Casey, and Connor –Jeffrey L. Harrison

Contents

Preface

page xiii

1

Introduction 1.1 Introduction 1.2 Some Recent Examples 1.2.1 Bid Rigging at Antique Auctions 1.2.2 Information Sharing at Treasury Auctions 1.2.3 American Express and the Boston Fee Party 1.2.4 Collusion on Campus: Financial Aid 1.2.5 Collusion in Baseball’s Free Agent Market 1.2.6 The Market for College Athletes and Coaches 1.2.7 Monopsony and Agriculture 1.2.8 Upheaval in Health Care Markets 1.3 Plan of the Book

1 1 2 2 3 5 6 8 10 11 12 14

2

The Antitrust Laws and Monopsonistic Forms of Conduct 2.1 Introduction 2.2 The Sherman Act 2.2.1 Agreements in Restraint of Trade 2.2.2 Monopolization and Attempts to Monopolize 2.3 The Clayton Act 2.3.1 Price Discrimination 2.3.2 Tying and Exclusive Dealing 2.3.3 Mergers 2.4 A Taxonomy of Monopsony Cases 2.4.1 Collusive Monopsony 2.4.2 Single-Firm Conduct 2.4.3 Mergers 2.5 Concluding Remarks

16 16 17 17 20 22 23 24 27 29 30 36 40 40

vii

viii

Contents

3

Economic Theory of Monopsony 3.1 Introduction 3.2 A Simple Model of Monopsony 3.2.1 The Welfare Effects of Monopsony 3.2.2 The Effect of Monopsony on Output Price 3.3 Collusive Monopsony 3.3.1 The Economics 3.3.2 Conditions Conducive to Collusive Monopsony 3.3.3 Organizing and Implementing a Buyer Cartel 3.3.4 Problems for Collusive Buyers 3.4 Measuring Buying Power 3.4.1 The Measurement of Buying Power 3.4.2 The Buying Power Index: Pure Monopsony 3.4.3 The Buying Power Index: The Dominant Buyer 3.4.4 The BPI and the Department of Justice Threshold 3.5 Market Definition and the BPI 3.6 The Buying Power Index in Practice 3.7 Concluding Remarks

41 41 41 43 45 48 48 49 50 51 53 54 54 55 60 61 64 67

4

The Antitrust Response to Monopsony and Collusive Monopsony 4.1 Introduction 4.2 Judicial Assessment of Market Power 4.3 Monopsony Pricing 4.4 Abuses of Monopsony Power 4.4.1 Price Effects 4.4.1.1 Classical Model 4.4.1.2 Inelastic Supply and Perishable Commodities   4.4.1.2.1  The Role of Perishability   4.4.1.2.2 Welfare Consequences of Collusion in the Context of Inelastic Supply   4.4.1.2.3  All-or-None Supply Cases 4.4.2 Nonprice Abuses 4.4.2.1 Horizontal Market Division 4.4.2.2 Bid Rigging 4.4.2.3 Refusals to Deal 4.4.2.4 Monopsony Leverage and Tying 4.5 Monopsony and Merger Policy 4.5.1 Horizontal Mergers 4.5.2 Vertical Mergers 4.6 Monopsony and Price Discrimination

68 68 69 70 78 78 78 79 81 82 83 85 85 86 88 91 93 93 96 99

Contents

5

6

7

ix

4.6.1 Monopsony and Unexploited Scale Economies 4.6.2 Different Supply Elasticities 4.7 Concluding Remarks

101 103 104

Cooperative Buying Efforts 5.1 Introduction 5.2 The Theory of Joint Purchasing 5.2.1 Efficiency-Enhancing Cooperative Buying Ventures 5.2.2 Enhanced Efficiency and Increased Buying Power 5.3 Antitrust Responses to Cooperative Buying Efforts 5.3.1 The Ancillary Restraints Doctrine 5.3.2 Application of the Ancillary Restraints Doctrine to Efficiency-Producing Buying Agreements 5.3.2.1 Cooperative Buying 5.3.2.2 Product Standardization 5.3.2.3 Northwest Wholesale Stationers and U.S. Department of Justice Guidelines 5.3.2.4 A Structural Dilemma 5.4 Concluding Remarks

106 106 107 108 109 113 113

Bilateral Monopoly 6.1 Introduction 6.2 Countervailing Market Power in Antitrust 6.3 The Bilateral Monopoly Muddle 6.4 The Correct Solution 6.4.1 A Simple Model of Bilateral Monopoly 6.5 A Formula Price Contract Solution 6.5.1 Difficulties with Vertical Integration 6.5.2 Vertical Integration by Contract 6.5.3 Formula Price Contracts 6.5.4 Performance Characteristics 6.6 Limitations of Countervailing Power Considerations 6.6.1 Transaction Costs 6.6.2 The Risk of Seller Collusion 6.6.3 Oligopoly and Oligopsony 6.7 Concluding Remarks Appendix: Price Determination in Bilateral Monopoly

123 123 124 126 127 128 131 131 132 133 134 136 136 138 140 141 142

Monopsony and Antitrust Enforcement 7.1 Introduction 7.2 Private Enforcement

146 146 146

115 115 117 119 121 122

Contents

x

7.3

7.4 7.5 7.6

7.2.1 Antitrust Injury 7.2.2 Antitrust Standing Applying the Standards 7.3.1 Collusive Monopsony 7.3.1.1 Price Fixing 7.3.1.2 Market Division 7.3.1.3 Boycotts 7.3.2 Monopsonization 7.3.2.1 Nonprice Efforts 7.3.2.2 Predatory Pricing Mergers Price Discrimination Concluding Remarks and Measuring Antitrust Damages

150 152 156 157 157 162 163 163 163 164 166 167 168

8

Monopsony in Action: Agricultural Markets 8.1 Introduction 8.2 The Monopsony Problem 8.3 Monopsony Power and Contract Power 8.4 Tying and Reciprocal Dealing 8.5 Alternative Approaches to Agricultural Buying Power 8.6 Concluding Remarks

172 172 173 174 179 182 187

9

Monopsony in Action: The NCAA 9.1 Introduction 9.2 Introduction to the NCAA 9.3 Collusive Monopsony 9.4 The NCAA and Collusive Monopsony 9.4.1 Organizing Limits on Prices and Quantities 9.4.2 Revenue Sharing 9.4.3 Sanctions for Cheating 9.5 Antitrust Challenges to the NCAA Monopsony 9.5.1 Walk-on Athletes: In re NCAA I-A Walk-on Football Players Litigation 9.5.2 Undercompensation: White v. NCAA 9.5.3 Numbers of Coaches: Hennessey v. NCAA 9.5.4 Coaches’ Compensation: Law v. NCAA 9.6 Concluding Remarks

188 188 188 189 190 191 194 194 196

10 Monopsony in Action: Physician Collective Bargaining: Monopoly or Bilateral Monopoly 10.1  Introduction 10.2  Economic Rationale of the State Legislation

197 198 199 201 204 205 205 207

Contents 10.3  The Effect of Cooperative Bargaining 10.4  Reaction of the Federal Trade Commission 10.5  Economic Rationale of FTC Concerns 10.6  State Action Doctrine 10.6.1  Background 10.6.2  Clear Articulation 10.6.3  Active State Supervision 10.6.4  Information Available 10.7  Diagnosing Monopsony Power 10.7.1  Statutory Guidance 10.7.2  Monopsony and Monopsony Power 10.7.2.1 Analysis of the Relevant Product (or Service) Market 10.7.2.2  Analysis of the Relevant Geographic Market 10.7.2.3  Measuring Monopsony Power 10.7.2.4  Importance of Entry Barriers 10.7.2.5  Implications of the BPI 10.7.3  Calculating Market Shares 10.8  Postnegotiation Review 10.9  Concluding Remarks

xi 208 208 211 211 212 212 213 215 215 216 217 217 218 218 220 220 221 222 222

11  Final Comments 11.1  Introduction 11.2  The Evolving Nature of Antitrust Law 11.3  Issues of Antitrust Injury and Standing 11.4  Cooperative Buying and Bilateral Monopoly 11.5  The Pervasiveness of All-or-None Supply 11.6  Single-Firm Behavior

224 224 224 225 228 230 231

Bibliography

233

Index

245

Preface

In 1993, we published a relatively short book, Monopsony: Antitrust Law and Economics (Princeton University Press). At that time, although mono­psony had been a mainstay in the study of economics, the courts addressed it only rarely. No doubt, this neglect was due at least in part to two things. First, plaintiffs (and their lawyers) focused their attention on sellers, and, therefore, monopsony did not come up. Second, the ill effects of mono­psony are somewhat counterintuitive. There was a mistaken belief among some that if monopsony power is used to lower prices, this must ultimately be bene­ ficial to consumers. In most instances, this is faulty economic ­reasoning. This state of affairs was particularly difficult to understand since mono­ psony conditions and behavior abounded even then. Moreover, mono­psony results in the same types of distributive and allocative issues raised by the exercise of monopoly power. In the past eighteen years, much of this has changed. Increasingly, courts have decided cases that deal with the behavior of buyers. These cases arise in markets ranging from agriculture to health care to college and professional athletes. In addition, the number of contributions of other legal scholars to the monopsony literature has exploded since our initial effort. This focus on buyer behavior gives rise to complex issues. For example, when is a ­buying cooperative an illegal collusive monopsony? Does bilateral monopoly re­present a “solution” to the problems monopsony presents? In addition, when the antitrust theory involves monopsony, how are the concepts of antitrust standing and injury to be applied? We believed that these changes warranted our revisiting monopsony. This book represents far more than a new edition of our 1993 offering. It includes a number of additional chapters, applying monopoly theory and law to specific industries:  agriculture, health care, and sports. In some xiii

xiv

Preface

measure, it also represents a reexamination of some of our own ideas as presented in our earlier book. As the final chapter of this book indicates, there is still work to be done. This is, in part, a direct result of the nature of antitrust law more generally. As it evolves, it is important that the legal treatment of monopsony also evolve. A work of this nature is never the product of the efforts of the authors alone. Thus, the authors would like to thank Kristine Coffin, Christine Durrance, Jessica Haynes, Jill Herndon, and John Lopatka for past collaboration. Sarah Goldberger, Casey Harrison, Jessica Haynes, and Corinne Turcotte all helped in various ways in manuscript preparation. Scott Parris, our ­editor at Cambridge, was supportive and encouraging throughout the ­process. Naturally, all errors and omissions are our own. We sincerely invite comments from interested readers. Roger D. Blair Jeffrey L. Harrison 2010

ONE

Introduction

1.1 Introduction Technically, monopsony exists when there is but one buyer of a well­specified good or service. Thought to be rare, until recently monopsony received scant attention in most antitrust casebooks and texts.1 This changed somewhat in the 1990s and, as a result of a recent decision addressing monopsony conduct by the U.S. Supreme Court, monopsony has been thrust into the forefront of consideration by antitrust academics and lawyers.2 This is as it should be because monopsony is far more prevalent than many have recognized. Consider the following examples: The owners of professional football teams agree on which players each team will have the exclusive right to negotiate with; the National Collegiate Athletic Association (NCAA) regulates both the number of athletic scholarships and the amount of compensation that the athletes can receive;3 financial aid officers of elite colleges and universities meet to avoid a bidding war for the most desirable students;4 tuna canneries in California allegedly fix purchase prices at A notable early exception is provided by Richard Posner and Frank Easterbrook, who analyze monopsonistic price fixing. They did not, however, examine monopsony in other contexts. Richard Posner & Frank Easterbrook, Antitrust:  Cases, Economic Notes and Other Materials 146 (West Publishing Co. 1981). 2 Weyerhaeuser Co. v. Ross-Simmons Hardwood Lumber, 127 S.Ct. 1478 (2006). 3 One of many cases addressing the monopsony power of the NCAA involved the refusal to pay walk-on players. In re NCAA I-A Walk-On Football Players Litigation, 2006 WL 1207915 (W.D.Wash. 2006). Robert Barro awarded the first annual prize for best monopoly in America to the NCAA for two major accomplishments. First, its interference with the market mechanism transfers wealth from poor ghetto residents to rich colleges. At the same time, it manages to maintain the moral high ground by convincing the majority that such controls are good and payment is evil. Gary Becker, The NCAA: A Cartel in Sheepskin Clothing, Business Week, Sept. 14, 1987, at 24. 4 See United States v. Brown Univ., 5 F.3d 658 (3d Cir. 1993). 1

1

2

Introduction

artificially low levels;5 and antique dealers rig the bids in public auctions and then divide the spoils later.6 In all of these instances, the parties were exercising monopsony power in one way or another. What follows are more detailed examples of the forms the monopsonistic conduct has taken and a look forward to the plan of this book.

1.2 Some Recent Examples 1.2.1  Bid Rigging at Antique Auctions Many antique auctions7 have been plagued by the formation of auction pools, which are groups of buyers who agree among themselves not to bid against one another. Although auction pools or buyer rings are considered to be unsavory, they have been seen as customary and even inevitable at antique auctions.8 In spite of the obviously anticompetitive purpose and effect of auction pools, many of the participants are quite proud of their pooling activity. Cox quoted one pool member as saying “the day I was allowed to go into the pool was a banner day.” Another reported that he considered his first invitation to pool as a “mark of distinction.”9 We discuss antique auctions as an example, but pools have been discovered in auctions for timber rights,10 real estate, used commercial equipment, and such mundane items as burlap, cattle, scrap metal, sewing machines, and tobacco.

Eagle v. Star-Kist Foods, Inc., 812 F.2d 538 (9th Cir. 1987). The plaintiffs in this case were fishermen who were paid by vessel owners on the basis of a price per ton of fish caught. Since the plaintiffs were employees of the vessel owners, their injuries were derived from those suffered by the vessel owners. On that basis, they were denied standing. 6 United States v. Pook, 1988 WL 36379 (E.D. Pa. 1988); United States v. Howe, Criminal No. 87–00262 (E.D. Pa. July 21, 1987). 7 For a wonderful account of auctions, see Ralph Cassady, Jr., Auctions and Auctioneering (Berkeley, CA:  Univ. of Cal. Press 1967). In addition, see Paul Milgrom, Auctions and Bidding: A Primer, 3 J. Econ. Persp. 3 (Summer 1989). More technical treatments are provided by Eric Rasmusen, Games and Information 245 (New York: Basil Blackwell 1989); and Louis Philips, The Economics of Imperfect Information 89 (New York: Cambridge Univ. Press 1988). Other useful contributions include R. Preston & John McMillan, Bidding Rings, 82 Am. Econ. Rev. 578 (1992); and Ken Hendricks, Robert Porter & Gunfu Tan, Bidding Rings and the Winner’s Curse, 39 RAND J. Econ. 1018 (Winter 2008). 8 Meg Cox, At Many Auctions, Illegal Bidding Thrives as a Longtime Practice Among Dealers, Wall Street J., Feb. 19, 1988, at C1. Some fascinating insights are also provided by Daniel A. Graham & Robert C. Marshall, Collusive Bidder Behavior at Single Object Second-Price and English Auctions, 95 J. Pol. Econ. 1217 (1987). 9 Graham and Marshall, supra note 8. 10 For an account, see Lee Baldwin, R. Marshall, & J. F. Richards, Collusion at Forest Service Timber Sales, 105 J. Pol. Econ. 657 (1997). 5

Some Recent Examples

3

Antique auction pools seem to enjoy continuing popularity.11 No doubt, this is due in part to the substantial collusive profits that may result and the fact that the practice is very difficult to police if the participants are clever. Usually, the opportunity for collusive profit arises when there is asymmetric information: An astute buyer recognizes that an item is undervalued or knows that he will be more expert than the average buyer at, say, an estate auction. There is an incentive for this buyer to identify others who may also be knowledgeable and form an auction pool. The members of the pool may select a designated bidder who will face no competition from pool members. Some pools are so brazen that the members sit together while only one of the participants bids. In other cases, the members will try to disguise the existence of the pool by not sitting together and by pretending to bid against one another. After the items have been purchased at the rigged auction, they must be resold in a private auction, or “knockout.” The difference between the true value as determined at the knockout and the rigged auction price represents profit for the pool members. This surplus is then divided among the members. Although systematic evidence on collusive profit does not exist, anecdotal evidence is suggestive. Cox, for example, reported an instance where a pool purchased a desk for $1,325, whereas the knockout price was $5,000. Similarly, Cassady provided an example of a Chippendale commode that sold for £750 at a public auction, but subsequently sold for £4,350 at a knockout. When pooling is suspected, the auctioneer can set a reserve price below which the item will not be sold. If the auctioneer has a good idea about the competitive price, this may be quite effective in thwarting the collusion. Alternatively, the auctioneer can run up the price by accepting phantom bids from nonexistent bidders. This tactic is possible only when the bidding ring does not contain all of the potential buyers.

1.2.2  Information Sharing at Treasury Auctions The scandal surrounding the activities of Salomon Brothers at Treasury auctions provides another example of collusive monopsony.12 The U.S. Treasury In addition to the cases alluded to by Cox, supra note 8, there have been several recent cases: United States v. Kay & Gross, Inc., Crim. No. 91-CR411 (D.C. SNY May 9, 1991); United States v. Thomas Schwenke, Inc., Crim. No. 91-CR487 (D.C. SNY June 7, 1991); United States v. Howe, Crim. No. 87–00262 (E.D. Pa. July 21, 1987). 12 Apparently, Salomon Brothers was not the only one engaged in improper behavior; see Michael Siconolfi, Michael R. Sesit, & Constance Mitchell, Hidden Bonds, Wall Street J., Aug. 19, 1991, at A1. 11

4

Introduction

sells Treasury bills, notes, and bonds at auction in order to generate the funds necessary to finance the budget deficit and to pay the government’s bills. It resorts to the auction in an effort to obtain the needed funds at the least cost possible. The 200-year-old Treasury market is the largest securities market in the world. This market is extremely important as the interest rates that are set on sales of Treasury bills, notes, and bonds provide a benchmark for borrowing rates for businesses and consumers. Thus, what happens in this market has far-reaching effects in the ­economy.13 Nonetheless, bidding activity had been restricted to the 40 primary dealers who had been licensed by the Federal Reserve Bank of New York. These dealers were the only ones permitted to bid on their own behalf as principals or on behalf of their customers. Competition among these 40 dealers for a specific issue sets a price for the debt instrument and thereby determines the actual yield or interest rate. When the Treasury announces the details of an auction, dealers begin active trading in the “when-issued” market. Information from this market provides data on how strong the demand is for the securities being offered. This, of course, is crucial because the primary dealers submit bids in the following form: “I will buy $10 million worth of bonds at 98.” At the Treasury auction, all such bids are tabulated and the highest offers are accepted. As a result, one successful bidder may pay a price of, say, 99 while another pays 98. When the dealers resell this issue to investors, the prices will be the same. But the cost to the dealer will depend on whether he was a low successful bidder or a high successful bidder. Consequently, it is extremely important to not be a high successful bidder. A dealer can lose millions of dollars if his bid is even slightly above the average bid. One way for a dealer to ensure that a bid is not too high is to collude with others on the actual bids that will be submitted. Sharing information from the when-issued market is permissible, but sharing confidential bidding information is illegal.14 When the auction deadline nears, the major dealers are in constant contact sharing information on expected demand from major clients, the amount they expect to bid, and the level of the market. Apparently, Salomon’s collusion was facilitated by the existence of a number of former Salomon dealers who had gone to work for other firms. In theory, the exchange of See Tom Herman, A Primer on the Treasury Market: How the Government Sells Its Debt, Wall Street J., Aug. 19, 1991, at A5. 14 The prohibition on anticompetitive information exchanges can be traced to American Column & Lumber Co. v. United States, 257 U.S. 377 (1921). Recent decisions that clearly extend this judicial hostility are United States v. Container Corp. of America, 393 U.S. 333 (1969), and United States v. United States Gypsum Co., 4313 U.S. 422 (1978). 13

Some Recent Examples

5

information will tend to keep the price of the security down and the interest rate up.15 Milton Friedman suggested a way to eliminate some of the risks to the bidders in the current system and thereby reduce the incentive to collude.16 He recommended the use of a Dutch auction.17 Each bidder offers to take a certain quantity of a particular issue at a specified price. The Treasury then reviews the offers and sells the entire issue at the highest price that clears the market. Under this modification, every successful bidder will pay the same market-clearing price, and, therefore, no one will be at a cost disadvantage in the resale market. This has the advantage of eliminating completely the risk of being out of line on the price paid and thereby reduces the incentive to collude on the bids.18

1.2.3  American Express and the Boston Fee Party American Express offers its cardholders somewhat better terms than the most popular bank cards, Visa and MasterCard.19 As a result, consumers who qualify for an American Express card prefer to use it. But the merchants are not similarly infatuated with accepting the American Express card because the fees that they must pay are higher than those charged by Visa and MasterCard. While the bank cards impose fees on the merchants in the 1–2 percent range, American Express charges 3–5 percent depending on the merchant’s location and volume of business. In the Boston area, most of the restaurants were being charged about 3.5 percent by American Express. For a somewhat different view, see Narasimhan Jegadeesh, Treasury Auction Bids and the Salomon Squeeze, XLVIII J. Fin. 1403 (1993). 16 Milton Friedman, How to Sell Government Securities, Wall Street J. Aug. 28, 1991, at A8. 17 In a classic Dutch auction, the auctioneer starts by offering to sell at a high price. The offer then is reduced gradually until a buyer is found. Dutch auctions have been used extensively in selling fresh fish, cut flowers, and fresh fruit. There was a time when art treasurers were sold in Dutch auctions, but the familiar English or ascending price auction is more popular today. For an extremely interesting empirical analysis of price dispersion over time at a Dutch auction of cut flowers, see Christopher D. Hall, A Dutch Auction Information Exchange, 32 J. L. & Econ. 195 (1989). 18 Friedman pointed out that this change would mean that the raison d’etre of the authorized dealers would vanish as final purchasers could deal directly with the Treasury. In addition, there would be no need for a when-issued market to exist since the information gathered there would not be needed. 19 These events were chronicled in the Wall Street Journal. See Peter Pae, Boycott Threat Spurs American Express to Rethink Fees, Wall Street J., Mar. 28, 1991, at B1; Peter Pae, Today’s Special: Cut in American Express Fees, Wall Street J., Apr. 22, 1991, at B1; and Johnnie L. Roberts, FTC Probes American Express Restaurant Fee Revolt, Wall Street J., Apr. 26, 1991, at B1. 15

Introduction

6

During difficult economic times, beleaguered restaurants resented the way that these fees ate into their profits. In March 1991, about 100 Boston restaurants decided that they had had enough of American Express’s high fees for using its credit card. Unlike some well-known restaurants that unilaterally dropped the American Express card, the Boston group relied on strength in numbers to negotiate a better deal from American Express. Faced with an ultimatum that the entire Boston group would drop the American Express card unless some concessions were made on fees, American Express agreed to rethink its fee structure. The American Express concessions varied depending on a restaurant’s volume and whether it filed its charge records electronically. For those restaurants that did not file their charge records electronically, there was no relief at all; the fees remained at their original level. For those restaurants that did file their charge records electronically, however, there were reductions in fees that depended on the annual volume. Those restaurants that processed over $10 million annually with American Express saw their fees drop from 3.25 percent to 2.75 percent – a saving of at least $50,000 per year. Restaurants with volumes between $1 million and $10 million saw their fees fall to 3 percent. For the least-intensive users, those with volumes under $1 million, there was no change. Needless to say, the small restaurants that claimed to have started the revolt against American Express were sorely disappointed that they did not stand to benefit from their efforts.

1.2.4  Collusion on Campus: Financial Aid Financial aid for college students is a big business. Some 5 million students received $26 billion in financial aid from federal, state, university, and private sources in 1988.20 In 1989, the Department of Justice became interested in how financial aid decisions were being made by some of the elite colleges and universities in the United States.21 In particular, it was interested in whether there was collusion among some of these prestigious schools in awarding financial aid. When a student applied for financial aid, he or she filled out a standard form that provided information on the applicant’s family income and assets. The purpose of the form was to assess how much the applicant’s family could be expected to contribute toward the cost of educating the student. Connie Leslie & Sue Hutchinson, An Ivy League Cartel, Newsweek, Aug. 21, 1989, at 65. Division Seeks Documents From Colleges in Probe of Financial Aid and Tuition, 57 Antitrust Trade Reg. Rep. 278 (1989).

20 21

Some Recent Examples

7

Presumably, each university’s financial aid officer could use this information to reach a unilateral decision on how much to offer a particular student. But many of our nation’s elite colleges and universities did not reach unilateral decisions. Instead, they colluded. The so-called Overlap Group was comprised of 23 prestigious, mostly East Coast, colleges and universities, such as Harvard, Yale, Princeton, Bryn Mawr, Williams, and Amherst. This was a financial aid cartel that pooled information on financial aid applicants. Each member sent its records on financial aid applicants to Harvard, which employed Student Aid Services, Inc. to sort the applicants and identify which ones had applied to two or more of the cartel members. Twice a year, the financial aid officers met to decide how much financial aid each student would be offered. If there was disagreement on how much should be offered to a particular applicant, the matter was discussed and resolved – offers were raised or lowered as the case may have been and uniformity was established. The end result was that the student was not the beneficiary of competition among the schools for his or her enrollment. Financial aid is a form of payment that a student receives in exchange for agreeing to attend a certain college or university. What these cartel members were doing was agreeing not to compete for the best of the applicants on the basis of price. In fact, they readily acknowledged that the purpose of the Overlap Group’s efforts was to extend uniform financial aid offers to prevent a bidding war for students. They claimed that they wanted each student to make his or her decision on the basis of academic considerations rather than on the basis of price.22 But the net result was that the Overlap Group pooled its buying power and thereby reduced the cost of attracting the most promising students. There are some who agreed with this effort. For example, a New York Times editorial concluded that “… the system’s larger goals are surely worthy: to prevent a needless and costly scramble for the most talented ….”23 A second editorial in the New York Times concluded that “… colleges ought to have some freedom in determining how best to spend their resources.”24 This is a peculiar view to take because of the obvious implications in other In National Society of Professional Engineers v. United States, 435 U.S. 679 (1978), the Supreme Court’s considerations dealt with a similar argument. The engineers felt that prospective clients should select an engineer for a specific job before even knowing the price. They argued that price competition was socially undesirable. The Court rejected this contention: “a defense based on the assumption that competition itself is unreasonable” will not be accepted. 23 Campus Conspirators?, New York Times, Aug. 18, 1989, at A30. 24 Bidding for the Best and the Poorest, New York Times, Feb. 19, 1990, at A16. 22

8

Introduction

areas. For example, one might infer from this that the New York Times editors would approve if all newspaper publishers colluded on salaries of editorial writers so an individual would select his or her employer on the basis of nonwage terms. Somehow it seems doubtful that these editors would so readily agree that newspaper publishers “ought to have some freedom in determining how best to spend their resources.”

1.2.5  Collusion in Baseball’s Free Agent Market In major league baseball, a free agent is a player with at least six years in the major leagues, whose contract has expired. A free agent can enter the market, negotiate with any and all teams, and sign with the highest bidder. Competition on both sides of the transaction was protected by Article 18, Clause H of the collective bargaining agreement between the owners and the players’ union,25 which requires that each player and each team act independently. The result of this competition was a significant surge in salaries for major league baseball players.26 During the collective bargaining negotiations in 1985, baseball Commissioner Peter Ueberroth ordered the clubs to exchange information on their financial situations. This revealed that many teams were not doing too well financially. At the same time, some statistics indicated that players with multiyear contracts experienced significant declines in their performance.27 Apparently, the owners agreed among themselves to not pursue any free agents. In contrast to the free agency period between the 1984 and 1985 seasons, which saw Bruce Sutter receive a six-year contract with the Atlanta Braves for $10.1 million and Rick Sutcliffe negotiate a $9.6 million five-year deal with the Chicago Cubs, only one of the twenty-nine free agents available after the 1985 season received a bona fide offer from another club. A similar plight met the seventy-nine players who filed for free agency in 1986. The players’ union filed a grievance that was heard by a labor arbitrator.28 In This clause, “Individual Nature of Rights,” provides that “[t]he utilization or nonutilization of rights under this article 18 is an individual matter to be determined solely by each player and each club for his or its own benefits. Players should not act in concert with other players and clubs should not act in concert with other clubs.” In the current collective bargaining agreement, a nearly identical provision is now found in Article 20, Clause E. 26 From 1976 to 1980, salaries rose 279 percent, and from 1980 to 1985, they rose another 287 percent. 27 Kenneth Lehn, Property Rights Risk Sharing and Player Disability in Major League Baseball, 25 J. L. Econ. 343 (1982), is an early study of how contract changes influence performance in major league baseball. 28 The players could not file an antitrust suit since baseball had received an exemption in Federal Baseball Club of Baltimore, Inc. v. National League of Professional Baseball Clubs, 25

Some Recent Examples

9

September 1987, Thomas T. Roberts ruled that the owners were guilty of collusion (Collusion I).29 While Roberts was trying to figure out what the relief should be for the players that were affected adversely, a second collusion case was being heard by another arbitrator, George Nicolau. In August 1988, the owners were found guilty of collusion during the 1987 free agency period. Nicolau found the total lack of interest in any of the 79 free agents to be inexplicable absent collusion. Nicolau ruled that “[b]y common consent, exclusive negotiating rights were, in effect, ceded to former clubs. There was no vestige of a free market, as that term is commonly understood.” The owners had contended that there had been no collusion. The absence of interest in free agents, the owners asserted, was due to a simultaneous, unilateral return to fiscal sanity. Nicolau flatly rejected this contention: The owners had exhibited a uniform pattern of behavior that was “simply unexplainable by the rubric of financial responsibility or by any other factors….”30 In spite of the adverse ruling in Collusion I, the owners developed and used a salary-offer data bank so they could compare offers being made to free agents. This data bank was used to keep down the offers that were made to free agents during the 1988 free agency period. This resulted in the Collusion III arbitration, which also went against the owners. Nicolau, who was the arbitrator for Collusion III, found that the data bank was used by the owners to “quietly cooperate” so that “prices won’t get out of line and no club will be hurt too much.”31 The damages awarded in collusion I were $10.5 million for 1986 and $102.5 million for 1987 and 1988 in Collusion II and III. Before the arbitrator could determine the losses for the 1989 and 1990 seasons as well as the interest that was due, the owners and the union settled all three cases for a final total of $280 million. This amounted to $10.77 million per club. Each player who felt that the collusion injured him could file for compensation from the $280 million fund. When the dust settled, some 843 players filed 3,173 claims that totaled $1.3 billion. 259 U.S. 200 (1922). The reserve clause, which bound a player to one team, was unsuccessfully challenged by Curt Flood; see Flood v. Kuhn, 407 U.S. 258 (1972). Free agency began in 1975 when a labor arbitrator granted free agency to Dave McNally and Andy Messersmith. 29 Hal Lancaster, Baseball Owners Conspire to Shut Down Market for Free Agents, Arbitrator Rules, Wall Street J., Sept. 22, 1987, at 10. 30 Id. 31 Claire Smith, Arbitrator Finds 3d Case of Baseball Collusion, New York Times, Jul. 19, 1990, at B9.

10

Introduction

The actual allocation of the settlement to the players was left to the Major League Baseball Players Association, which relied on arbitration. The money was not fully distributed until 2005. One player, Steve Garvey, was not happy that the arbitrator did not award a portion of the $280 million pool to him and appealed all the way to the Supreme Court.32 The collusion issue arose again in 2002–3.33 In this instance, the alleged practice was not refusing to negotiate with players but making the same or very similar offers to free agents.34 A settlement was reached in 2006 when the owners agreed to a $12 million settlement.

1.2.6  The Market for College Athletes and Coaches Schools that are members of the NCAA must, by necessity, enter into agreements. For example, it would be difficult to have a football game without agreeing on a standard set of rules, who the officials will be, and a variety of other matters. Without substantial uniformity, games could not take place. In addition, even if the games could literally take place, they would be of little interest to anyone if the outcomes were predictable and lopsided. It is when one thinks about the need for parity that the issue of agreements about players and coaches comes into play. Since colleges and universities essentially “hire” both coaches and players, the monopsony issue arises. In recent years, there have been a number of challenges to the NCAA rules with respect to coaches and players. For example, in In re Walk-On Football Players Litigation,35 players challenged the agreement among schools to limit the number of scholarship players. In effect, schools had used collusive monopsony power to limit the “purchases” of player services.36 In a similar action, football and basketball players sued the NCAA for limiting the amounts paid for scholarships. The players complained that covering tuition, books, and room and board did not constitute a full scholarship since athletes also incurred expenses for travel, phone calls, and laundry.37 In more familiar antitrust terms, the schools purchasing player Major League Baseball Players Ass’n v. Garvey, 532 U.S. 504 (2001). See Marc Edelman, Has Collusion Returned to Baseball? Analyzing Whether a Concerted Increase in Free Agent Player Supply Would Violate Baseball’s Collusion Clause, 24 Loy. Ent. L. Rev. 159 (2004). 34 MLB, Players Union Settle Potential Collusion Claims, http://sports.espn.go.com/mlb/ news/story?id=2652091 (last visited Sept. 8, 2008). 35 2006 WL 1207915, supra note 3. 36 Ultimately, the case was unsuccessful as the court was unwilling to certify the class of walk-on players. 37 NCAA Settles Antitrust Lawsuit; Will Create $218 Million Fund for Athletes, University Business, Jan. 31, 2008, http://www.universitybusiness.com/newssummary.aspx?news=yes 32 33

Some Recent Examples

11

services were allegedly fixing the maximum prices. Eventually this suit settled for $218 million and an agreement that schools would expand the benefits available to athletes.38 The use of monopsony power in college athletics was also extended to coaches’ salaries. Law v. National Collegiate Athletic Association39 was a reaction to a 1991 vote by the NCAA to limit the salaries of entry-level coaches. Other than in football, one coach in each sport was designated a “restricted earnings” coach, and his or her salary was limited to $12,000 for the academic year or $16,000 for the calendar year.40 The agreement thus had an impact on the price paid for coaching talent. Although the case eventually settled for $54.5 million, this did not resolve all issues. As with the baseball free agent settlement, there remained the complicated matter of distributing the settlement fund. The complexity of the allocation problem can be understood by noting that coaches paid the least were harmed less by the salary cap than those earning more. In fact, for some the salary cap may have resulted in no damage because even without the cap they would not have been paid more than the collusively determined wage. The eventual allocation of the settlement fund reflected these considerations.

1.2.7  Monopsony and Agriculture For years, there have been complaints by small farmers and ranchers that they are the victims of monopsony power. In effect, small sellers facing a single buyer or a limited number of buyers have little bargaining power. For example, in one study of the catfish market, it was found that 225 farmers in Western Alabama were selling to a single buyer and were receiving 13.6% less than the national average.41 Similarly, three firms do most of the meat packing in the United States, meaning they have significant buying power vis-à-vis ranchers. 42

40 38 39



41



42

&postid=15221 (last visited Oct. 14, 2008). See In re NCAA I-A Walk-On Football Players Litigation, supra note 3. White v. NCAA, Case 2:06-cv-00999-VBf-MAN (U.S.D.C. 2008). 134 F.3d 1010 (10th Cir. 1998). This case is examined in Roger Blair & Richard Romano, Collusive Monopsony In Theory and Practice: The NCAA. 42 Antitrust Bull. 681 (Fall 1997). Henry Kinnucan & Gregory Sullivan, Monopsonistic Food Processing and Farm Prices: The Case of the West Alabama Catfish Industry, S. J. Agric. Econ. 15 (December 1986). The monopsony issue in agricultural markets was brought to the forefront in a series of Congressional Hearings. Hearing on Monopsony Issues in Agriculture Before the S. Comm. on the Judiciary, 108th Cong. (2003).

12

Introduction

One of the more common issues in agricultural markets is the so-called merger to monopsony problem. Typically, one envisions mergers as increasing the power of sellers. In fact, when buyers merge they too can increase their leverage with respect to sellers. Mergers in the markets for dairy products, soybeans, and hogs have all raised this issue.43 The argument is that greater concentrations of buying power result because sellers become, in effect, captives – they have no alternative outlet for their commodities. The captive supplier question can be illustrated with a common practice in poultry markets in which large processors depend on many small farmers to “grow out” chickens or turkeys. In these instances the processor supplies the chicks or poults and pays the farmer to raise the chickens or turkeys to an appropriate size for slaughter. One centrally located processing plant may buy from many suppliers.44 To some extent the monopsony problem would seem to be lessened because no farmer is required to enter the growing out business. In fact, the potential sellers may be relatively unsophisticated, have low opportunity costs, and be urged to invest in facilities that can only be used for the growing out process. Having made the investment, the balance of bargaining power shifts to buyers.45

1.2.8  Upheaval in Heath Care Markets If one had to pick the industry most affected by monopsony power in the last 30 years, it would undoubtedly be the health care industry. At one point, physicians and other providers, including hospitals and pharmacies, were able to set fees or prices without a great deal of concern about competition. It was not that they possessed monopoly power as conventionally understood but that asymmetric information and search costs made it difficult for patients to move easily from one provider to another in response to price differences. This changed considerably with the development of third-party payors that act as surrogates for thousands of patients. These third-party payors See Testimony of Peter Carstensen, Single-Buyer Markets in Agriculture, Senate Judiciary Committee, October 30, 2003. 44 See C. Robert Taylor, The Many Faces of Power in the Food System, Presented at the DOJ/ FTC Workshop on Merger Enforcement, February 17, 2004. For a case responding to the same potential “squeeze,” see In re Southwestern Milk Antitrust Litigation, 555 F.Supp.2d 934 (E.D. Tenn. 2008). 45 See Been v. O.K. Indus., Inc., 495 F.3d 1217 (10th Cir. 2007). For instances of alleged bid rigging or collusive monopsony in an agricultural markets, see Knevelbaard Dairies v. Kraft Foods, Inc., 232 F.3d 979 (9th Cir. 2000); Pease v. Jasper Wyman & Son, 202 WL 1974081 (Me. Super. 2002). See generally, C. Robert Taylor, Buyer Power Litigation in Agriculture: Pickett v. Tyson Fresh Meats, Inc., 53 Antitrust Bull. 455 (2008). 43

Some Recent Examples

13

wield substantial monopsony power since they may represent a large percentage of a provider’s potential customers.46 Thus, they can refuse to buy services from a specific provider unless the terms offered are favorable to the third party.47 As one would expect, these consolidations on the buyer’s side have led to challenges from providers. One of the earliest challenges came in Kartell v. Blue Shield of Massachusetts, Inc.,48 in which physicians challenged a requirement by the third-party payor that patients not be billed for any amount in excess of that paid by the third-party payor. In effect, the complaint was that the amount reimbursed by Blue Cross Blue Shield was “too low.” The court rejected the doctors’ theory noting that the antitrust laws did not respond directly to high or low prices in the absence of additional unlawful behavior. The pinch health care providers have found themselves in has set off a number of reactions on the selling side of the market. First, there is the possibility of unionization. Obviously this is not an option for hospitals and pharmacies. In principle, it could work for physicians, but for the most part, physicians do not qualify as “employees” and, thus, may not ­unionize.49 Second, physicians and hospitals have also attempted to make use of an arrangement called the “messenger model,” which permits them to employ the same agent to bargain with third-party payors.50 The messenger model was designed by the U.S. Justice Department and the Federal Trade Commission to lower the costs of contracting between providers and third-party payors. It is also possible that it facilitates the sharing of information that will assist in efforts to stabilize prices. Third, there seems to be little doubt that the increase in monopsony power has led physicians to reconsider their business models. The reason for this has to do with potential antitrust liability. If several physicians with separate practices meet to discuss fees and make demands on third-party payors, they are in violation of the antitrust laws because they have entered into an agreement that restrains trade. On the other hand, if they form a See Roger D. Blair & Jill Boylston Herndon, Physician Cooperative Bargaining Ventures: An Economic Analysis, 71 Antitrust L. J. 989 (2004). 47 The impact of third-party payors may extend beyond those dealing with price. For example, the specific treatment may be determined by whether it is “covered” as opposed to what is optimal. Whether it is covered will depend on the contract between the provider and the actual buyer of the coverage – the third-party payor. 48 749 F.2d 933 (C.A. Mass. 1984). 49 An agreement among physicians outside the protection of the labor laws would itself be a horizontal agreement in violation of Section 1 of the Sherman Act. For a brief discussion see R. Blair & J. Herndon, supra note 46. 50 See Jeffrey L. Harrison, The Messenger Model: Don’t Ask, Don’t Tell? 71 Antitrust L. J. 1017 (2004). 46

14

Introduction

single business entity in the form of a clinic, they may have the same discussions and make the same demands without fear of liability simply because there is no agreement among separate legal entities. Finally, several times in the past few years, there have been legislative attempts to “level the playing field.” For example, in 2000, the Quality HeathCare Coalition Act passed in the U.S. House of Representatives but failed to proceed any further. Under that Act, health care professionals would have been exempt from the antitrust laws. The bill has been repeatedly introduced since that time. In fact, in 2007 the bill was introduced by Libertarian and presidential candidate Ron Paul, who described it as “a first step towards restoring a true free market in health care by restoring the rights of freedom of contract and association to health care professionals.”51

1.3  Plan of the Book We offer a systematic treatment of the law and economics of monopsony in the chapters that follow. It may be useful to sketch our plan so the reader can understand our organization. Chapter 2:  Antitrust Laws and Monopsonistic Forms of Conduct. In this chapter, we provide a general antitrust foundation for the analysis of monopsony. Chapter 3:  Economic Theory of Monopsony. We develop the fundamental theory of monopsony and its impact on social welfare. In addition, we examine collusive monopsony and present a measure of monopsony power. Chapter 4:  The Antitrust Response to Monopsony and Collusive Monopsony. The antitrust tools presented in Chapter 2 are applied to monopsony in this chapter. We review the case law as it has evolved in response to monopsonization, collusive monopsony, abuses of monopsony power, mergers, and price discrimination. Chapter 5: Cooperative Buying Efforts. There is an apparent inconsistency in the treatment of cooperative buying arrangements. As the preceding chapter finds, collusive monopsony is justly condemned while cooperative buying ventures have been applauded. We reconcile this apparent discrepancy and offer a principled basis for distinguishing between socially desirable and undesirable cooperative buying. Ron Paul, Introducing the Quality Health Care Coalition Act, Aug. 8, 2007, http://www. reasontofreedom.com/introducing_the_quality_health_care_coalition_act_by_us_rep_ ron_paul.html (last visited Oct. 21, 2008).

51

Plan of the Book

15

Chapter 6:  Bilateral Monopoly. We turn our attention to the case of monopoly supply and monopsony demand. After developing the economic theory of bilateral monopoly, we examine the appropriate antitrust rules. Chapter 7:  Monopsony and Antitrust Enforcement. The antitrust laws provide a legal structure within which business is conducted. Violations of the antitrust laws must be detected and prosecuted by either the public antitrust authorities or private citizens. This chapter reviews the public and private enforcement of antitrust as it pertains to monopsony. Chapter 8: Monopsony in Action: Agricultural Markets. In this chapter begins a sequence of three chapters that focus on specific industries and monopsony issues in those industries. Chapter 8 is devoted to agriculture. In that context a great deal of concerns has been expressed about the relationship between producers and processors. The chapter examines whether an antitrust response is warranted. Chapter 9:  Monopsony in Action:  The NCAA. It will surprise no one who follows sports that athletes at both the amateur and professional levels are often bargaining with employers with market power. In this chapter the NCAA as a monopsonist is explored. Chapter 10:  Monopsony in Action:  Physician Collective Bargaining: Monopoly or Bilateral Monopoly. Health care is the focus of this chapter. In many of the markets within heath care, there has been a great deal of consolidation on the buying side of the market. Physicians who were once able to set their own fees find that they are, in effect, selling to managed care organizations that have great power in determining fees. This is a system in which there is a trend toward bargaining by physicians and bilateral monopoly. Chapter 11: Final Comments. This chapter offers some final remarks and suggests several areas for future research.

T WO

The Antitrust Laws and Monopsonistic Forms of Conduct

2.1 Introduction This chapter provides an overview of the antitrust laws and describes the wide variety of buyer conduct that has been and can be scrutinized under the antitrust laws. The economic theory of monopsony is examined in Chapter 3, and the question of whether the application of the antitrust laws to buyers makes economic sense is addressed in Chapters 4 and 5. The antitrust laws include both substantive and procedural components. Our focus in this chapter is on the substantive element. Our examination of the procedural provisions of the antitrust laws is covered in Chapter 7. When considering the substantive provisions, the relevant legislation is Sections 1 and 2 of the Sherman Act, Sections 2, 3, and 7 of the Clayton Act and, to some extent, Section 5 of the Federal Trade Commission Act.1 As it turns out, it is in many ways an overstatement to refer to these sections as “substantive.” Except for Section 2 of the Clayton Act, which prohibits price discrimination, the statutes are quite loosely worded, and their real content has been left to judicial interpretation.2

In this chapter, Section 5 of the Federal Trade Commission Act is not separately analyzed. Section 5 prohibits “[u]nfair methods of competition.” 15 U.S.C. §45 (2000). The Federal Trade Commission and the Supreme Court have interpreted Section 5 as extending to those violations that would be prohibited by Sections 1 and 2 of the Sherman Act. 2 Several commentators have noted the resulting common law nature of the antitrust laws. See, e.g., Roger D. Blair & Carolyn D. Schafer, Antitrust Law and Evolutionary Models of Legal Change, 40 U. Fla. L. Rev. 379 (1988). For a critical view, see Thomas C. Arthur, Farewell to the Sea of Doubt: Jettisoning the Constitutional Sherman Act, 74 Cal. L. Rev. 263 (1986). 1

16

The Sherman Act

17

2.2  The Sherman Act 2.2.1  Agreements in Restraint of Trade Section 1 of the Sherman Act, which prohibits “[e]very contract, combination … or conspiracy in restraint of trade,”3 has been applied to both sellers and buyers. This wording has now undergone over 120 years of varying interpretation. A literal interpretation would result in the prohibition of many day-to-day agreements that are desirable. For example, a simple agreement to form a partnership would seem to be a violation since the partners would not compete with one another. In fact, the earliest substantive application of the Sherman Act, United States v. Trans-Missouri Freight Association4 effectively held that every restraint had been declared illegal. In the years that followed, this rigid view gave way to attempts to reconcile Section 1 with the practicalities of business practices. What emerged was the ancillary restraints doctrine, which was first suggested in the 1898 case United States v. Joint Traffic Association5 and more fully developed in United States v. Addyston Pipe and Steel.6 Under the ancillary restraints doctrine, an agreement that would otherwise violate Section 1 might be regarded as “reasonable” and legal if it were ancillary to an otherwise lawful agreement and was not more restrictive than necessary.7 In 1911, the Supreme Court decided Standard Oil Co. v. United States8 and rendered the ancillary restraints doctrine far less critical as a means of reconciling the broad language of Section 1 of the Sherman Act with business realities. Standard Oil removed the requirement that a restraint be ancillary in order to have any chance of passing antitrust muster. Instead, the issue with respect to both ancillary and direct restraints was one of “reasonableness.” 15 U.S.C. § 1 (2000). 166 U.S. 290 (1897). The early opinions of Justice Peckham have been examined by John R. Carter, From Peckham to White: Economic Welfare and the Rule of Reason, 25 Antitrust Bull. 95, (1980). 5 171 U.S. 505 (1898). 6 85 F. 271 (6th Cir. 1898) aff ’d 175 U.S. 211(1899). 7 Often, in early interpretations of the Sherman Act, Sections 1 and 2 of the Act were not distinguished. For more on the early years of the Sherman Act, see Hans Thorelli, The Federal Antitrust Policy (Baltimore:  Johns Hopkins Press 1954); Donald Dewey, The Common Law Background of Antitrust Policy, 41 Va. L. Rev. 759 (1955). 8 221 U.S. 1 (1911). See E. Thomas Sullivan & Jeffrey L. Harrison, Understanding Antitrust and Its Economic Implications 113 (New York: LexisNexis, 2009). 3 4

18

The Antitrust Laws and Monopsonistic Forms

Although the Court announced the rule of reason in 1911, a series of “bright line” tests of legality emerged in the years that followed.9 In effect, the Court created a list of agreements that conclusively would be presumed to violate the rule of reason. They were, in effect, unlawful per se. At its peak in the late 1960s, the list of practices that were illegal per se included horizontal10 and vertical price fixing11 (resale price maintenance), horizontal division of territories,12 vertical restraints on distribution,13 group boycotts,14 and tying arrangements.15 Thus, some arrangements were unlawful without further inquiry into their actual economic impact, while others were examined under the “rule of reason” and were unlawful only if the actual effects were shown to be harmful. Of course, the difficult issue of whether or not to classify the challenged activity as falling within one of the per se unlawful categories could not be escaped. The methodology involved was often not that different from the rule of reason standard itself. In the early 1970s, the expansion of the per se category came to a close. This change occurred as the composition of the Supreme Court shifted and economic analysis became the Court’s methodology of choice.16 This new philosophy was evidenced primarily by two elements. First, the Court emphasized that, while the rule of reason standard involved a process of weighing, the factors to be weighed were the procompetitive and   9 The per se status of resale price maintenance was actually established in the same year as Standard Oil. Dr. Miles Med. Co. v. John D. Park & Sons, 220 U.S. 373 (1911). Dr. Miles was overturned by Leegin Creative Leather Products, Inc. v. PSKS, Inc., 551 U.S. 877 (2007). 10 United States v. Socony-Vacuum Oil Co., 310 U.S. 150 (1940). Horizontal price fixing occurs when competitors agree to avoid price competition by fixing a uniform price or by some other means. 11 Dr. Miles, supra note 9. Vertical price fixing or resale price maintenance occurs when a seller supplies a product to a buyer who will resell it and the resale price is agreed on. 12 United States v. Topco Assoc., 405 U.S. 596 (1972). A horizontal division of territory occurs when competing sellers divide a market geographically with the understanding that each one will be the only seller in a specified area. 13 United States v. Arnold, Schwinn & Co., 388 U.S. 365 (1967). Vertical restraints on distribution involve limitations by sellers who sell products that will be resold or the areas or customers to whom purchasers may resell. 14 Klor’s Inc. v. Broadway–Hale Stores, Inc., 359 U.S. 207 (1989). Group boycotts fall into two categories. In some cases, competitors unite and refuse to deal with a supplier or customer who is dealing with a competitor of the group. The term “group boycott” has also been applied to agreements among competitors fixing a term of exchange other than price. The Supreme Court made it clear that boycotts were not actually illegal per se in Northwest Wholesale Stationery. 15 Northern Pacific Railway Co. v. United States, 356 U.S. 1 (1958). In a tying arrangement, a seller conditions the sale of a desired product (the tying product) on the buyer’s purchase of a less desired produce (the tied product). 16 By 1977, only four members of the 1969 Court remained.

The Sherman Act

19

anticompetitive effects of a particular activity. Thus, appeals to more general notions of “product quality” were only relevant if they were designed to intensify competition.17 Second, the Court overruled a prior case18 in which it had applied a per se standard and narrowed other per se cases. Clearly, the pivotal case announcing this new era was Continental T.V. v. G.T.E. Sylvania,19 in which the Court rejected the view that vertical restraints on distribution were to be treated as unlawful per se. In returning to a rule of reason standard, the Court announced its preference for a purely economic approach to antitrust enforcement and exhibited a newfound willingness to consider the possible procompetitive effects of practices that were thought to be per se unlawfu1.20 In actuality, the methodology the Court has adopted and has applied in the post-Sylvania era, even when considering what were previously per se offenses, is remarkably similar to the ancillary restraints approach. For example, in National Collegiate Athletic Association v. Board of Regents,21 the Supreme Court examined a variety of practices of the NCAA that had the effect of limiting the number of times teams could appear on television and the compensation they would receive. The Court conceded that the practices would ordinarily be condemned under the per se standard as price fixing. Instead of such condemnation, however, it permitted a rule of reason examination noting that the industry was one “in which horizontal restraints on competition are essential if the product is to be available at all.”22 It is interesting to note that the Court still found that the activity violated Section 1 of the Sherman Act.23 As a result of cases like NCAA, unless the activity clearly involves some type of price fixing or territorial division, defendants are more likely to ­escape per se treatment. Even activities that have a direct anticompetitive effect may not be summarily condemned until the Court has had an ­opportunity to consider any procompetitive effects.24 If the defendant is able Nat’l Soc’y of Prof. Eng’r v. United States, 435 U.S. 679 (1978). Since the 1970s, the Court had followed through and removed other practices from the per se category. 19 433 U.S. 36 (1977). See also United States Steel Corp. v. Fortner Enterprises, Inc., 429 U.S. 610 (1977). 20 For an interesting analysis of Sylvania and its possible implications, see Robert Bork, Vertical Restraints: Schwinn Overruled, 171 Sup. Ct. Rev. 1977 (1978); and Robert Pitofsky, The Sylvania Case: Antitrust Analysis of Non-Price Vertical Restrictions, 78 Colum. L. Rev. 1 (1978). 21 468 U.S. 85 (1984). 22 Id. at 101. 23 Id. at 120. 24 See, e.g., Federal Trade Commission v. Indiana Federation of Dentists, 476 U.S. 447 (1986). 17 18

20

The Antitrust Laws and Monopsonistic Forms

to make a colorable argument that such an effect does exist, the standard applied will be the “rule of reason.” This is not to say, however, that rule of reason means that the plaintiff will always have to demonstrate an actual overall anticompetitive effect. In fact, as the per se standard has evolved, so has the rule of reason. Indeed, the standards are best viewed as ends of a continuum. Today, activities that are nominally subjected to a rule of reason may be condemned with far less than a “full blown” weighing of effects,25 and some that are nominally per se are only labeled such after a great deal of analysis.26

2.2.2  Monpolization and Attempts to Monopolize The language of Section 2 of the Sherman Act is hardly any more helpful than that of Section 1. The lack of clarity is accentuated by the fact that the Supreme Court has considered Section 2 relatively rarely. Section 2 prohibits efforts to “monopolize, or attempt to monopolize.”27 It has been applied to conduct by both sellers and buyers. The case adding the most influential substance to Section 2 is United States v. Aluminum Co. of America (Alcoa).28 Alcoa established that “monopolization” involved both a structural component and a conduct component. That is, in order to monopolize illegally, the firm must possess monopoly power or a high degree of market power and have engaged in some form of conduct that was designed to maintain that position. Critical in Alcoa and its forerunners was the recognition that simply being a monopolist did not violate Section 2.29 Thus, monopoly power that was “thrust upon” the firm or that was the result of “skill, foresight and industry” was not prohibited by Section 2.30 In the years after Alcoa, and especially in the post-Sylvania era, the structure and conduct components of Section 2 have become more focused. For See NCAA v. Board of Regents of the Univ. of Okla., 468 U.S. 85; F.T.C. v. Ind. Fed’n of Dentists, 476 U.S. 447. See generally Lawrence Sullivan & John Wiley, Recent Antitrust Developments:  Defining the Scope of Exemptions, Expanding Coverage and Refining the Rule of Reason, 27 UCLA L. Rev. 265 (1979). 26 See, e.g., Arizona v. Maricopa County Medical Society, 57 U.S. 332 (1982). For an analysis of this decision, see Jeffrey L. Harrison, Price Fixing, the Professions, and Ancillary Restraints: Coping with Maricopa County, U. Ill. L. Rev. 925 (1982). 27 15 U.S.C. §2 (2000). 28 148 F.2d 416 (2d Cir. 1945). Alcoa was decided by the U.S. Court of Appeal for the Second Circuit when four members of the U.S. Supreme Court disqualified themselves. 29 In addition to Alcoa, see United States v. United States Steel Corp., 251 U.S. 417 (1920); Standard Oil v. United States, 221 U.S. 1 (1911). 30 148 F.2d at 427–30. 25

The Sherman Act

21

example, earlier cases, like Alcoa, seem to have placed undue reliance on market share as an indicator of market power. Today, it is generally recognized that market share is but a starting point and that the implications of market share must be qualified by estimates of demand and supply side substitutability.31 The conduct element of monopolization has become more focused but remains less precise than the market power issue. Two trends, however, do seem to have emerged. First, courts are defining the areas of questionable conduct increasingly narrowly.32 Second, the business behavior that most clearly satisfies the conduct element of the monopolization claim is that which is predatory in nature. Predatory activity is unprofitable in the short run and has as its primary business justification the exclusion of competitors, which permits monopoly profits in the long run. This latter point is illustrated in the 1985 case, Aspen Skiing Co. v. Aspen Highlands Skiing Corp.,33 in which a ski resort operating on three mountains with a number of lifts was sued for ceasing to offer a joint lift ticket with a smaller neighboring competitor. The Supreme Court considered the business justifications offered by the defendant and concluded that none were compelling. Instead, the Court noted that the defendant, in offering a ticket that was less comprehensive, may have “elected to forgo … short run benefits because it was more interested in reducing competition in the Aspen ­market over the long run by harming its smaller competitor.”34 The “attempt to monopolize” component of Section 2 has understandably tracked the monopolization element. The general approach was outlined in 1905, by Justice Holmes in Swift & Co. v. United States: Where acts are not sufficient in themselves to produce a result which the law seeks to prevent – for instance, monopoly – but require further acts in addition to the See, e.g., Telex Corp. v. IBM Corp., 510 F.2d 894 (9th Cir. 1975), cert. dismissed, 423 U.S. 802 (1975). Also see William Landes & Richard Posner, Market Power in Antitrust Cases, 94 Harv. L. Rev. 937 (1981). For more on monopolization standards, see Einer Elhauge, Defining Better Monopolization Standards, 56 Stan. L. Rev. 253 (2003), and Keith N. Hylton, The Law and Economics of Monopolization Standards, in Antitrust Law & Economics, Keith N. Hylton ed. (Boston: Edward Elgar Publishing, 2010). 32 See Berkey Photo Inc. v. Eastman Kodak Co., 603 F.2d 263 (2d Cir. 1979), cert. denied, 444 U.S. 1093 (1980). 33 472 U.S. 585 (1985). This opinion has come under some heavy criticism. See, e.g., Dennis W. Carlton, A General Analysis of Exclusionary Conduct and Refusals to Deal – Why Aspen and Kodak Are Misguided, 68 Antitrust L. J. 659 (2001). 34 472 U.S. at 608. After Aspen, the Supreme Court, in Verizon Communications, Inc. v. Law Offices of Curtis V. Trinko, 540 U.S. 398 (2004), seems to narrow those circumstances in which a dominant firm would be required to cooperate with a competitor. The basic methodology with a focus on predation has, however, not changed. 31

22

The Antitrust Laws and Monopsonistic Forms

mere forces of nature to bring that result to pass, an intent to bring it to pass is necessary in order to produce a dangerous probability that will happen. But when that intent and the consequent dangerous probability exist, this statute, like many others, and like the common law in some cases, directs itself against that dangerous probability as well as against the completed result.35

Obviously, direct determination of “intent” is likely to be impossible. Thus, as a general matter, the conduct of the would-be monopolist must create the inference of “intent.” In addition, market share is typically used as the gauge for whether the firm is likely to become a monopoly. Again, it is not the intent or the effort to become a monopolist that the law prohibits. Instead, it is, as aptly stated by Professors Areeda and Turner, “the intention to prevail by unfair means.”36 Defining “unfair means” has not been smooth sailing for the courts, but again, the most likely areas of questionable conduct involve the sacrifice of short-run profits, not in an effort to promote a product, but primarily to exclude a competitor.

2.3  The Clayton Act The relevant substantive sections of the Clayton Act are Section 2, which prohibits price discrimination; Section 3, which prohibits conditional sales such as exclusive dealing and tying arrangements; and Section 7, which prohibits some mergers. The Clayton Act provisions are somewhat less openended than Sections 1 and 2 of the Sherman Act in that the specific types of prohibited practices are described. Furthermore, the practices are prohibited only when the effect “may be substantially to lessen competition.”37 The Clayton Act can also be distinguished from the Sherman Act in that Sections 2 and 3 exclude buyer activity from the scope of coverage. Although this is an important point, it is equally important to note that most buyer activities that would fall outside literal interpretations of a Clayton Act section, but that would otherwise be anticompetitive, would in most instances fall under Sherman Act scrutiny.

196 U.S. 375, 296 (1905). Phillip Areeda & Donald Turner, Antitrust Law, vol. 3, 315 (Boston: Little, Brown 1978). See also Edward Cooper, Attempts and Monopolization: A Mildly Expansionary Answer to the Prophylactic Riddle of Section Two, 72 Mich. L. Rev. 357 (1974). 37 Section 2 of the Clayton Act, however, describes three types of discriminatory practices that seem to be per se unlawful. See E. Sullivan & Harrison, supra note 8, at 384–7. These are not discussed in detail here. 35 36

The Clayton Act

23

2.3.1  Price Discrimination Section 2 of the Clayton Act as amended by the Robinson-Patman Act is said to prohibit “price discrimination.”38 Under a true economic definition of price discrimination, Section 2 would prohibit any price that was not based on the difference in the marginal cost of serving two buyers. As it turns out, the implementation of Section 2 has not relied on the economist’s definition of price discrimination. Instead, it is applied primarily to prohibit price differences.39 The section’s coverage is further limited in that it applies to sales only. Thus, discrimination in lease terms would not fall within Section 2 of the Clayton Act. The Act contains two primary defenses for sellers: meeting competition and cost justification. These defenses will not be discussed at length here and, for the most part, have not been interpreted in a manner that makes them readily available.40 More problematic from the point of view of monopsony is that Section 2 is aimed primarily at prices charged by sellers.41 As it turns out, this sellerside emphasis may be misguided. The reason is that the Clayton Act, as originally enacted, was designed to avoid the possibility that powerful sellers would discriminate in price in order to harm competing sellers.42 That is, they would drop prices in competitive markets while keeping them higher in markets in which they had market power. This is called primary line price discrimination or injury. Between 1914, when the Clayton Act was enacted, and 1936, when it was amended by the Robinson-Patman Act, concerns about the harms of price discrimination changed. The new focus was on the possibility that powerful buyers could use their monopsony power as a means to exact price concessions from their suppliers that less powerful buyers would not be able to obtain.43 This is called secondary line price discrimination or injury. The 15 U.S.C. §13 (2000). See F.T.C. v. Anheuser-Busch, Inc., 363 U.S. 536, 549 (1960). 40 For an extensive analysis, see Frederick Rowe, Price Discrimination Under the RobinsonPatman Act (Boston: Little, Brown 1962). See also Richard Posner, The Robinson-Patman Act:  Federal Regulation of Price Differences (Washington, D.C.:  American Enterprise Institute 1976); Hugh Hansen, Robinson-Patman Law: A Review and Analysis, 5 Fordham L. Rev. 1113 (1983). 41 One of the so-called per se prohibitions of Section 2 of the Clayton Act may apply directly to buyer behavior. Section 2(c) makes it unlawful “to pay grant, or to receive or accept, anything of value as a commission, brokerage, or other compensation, or any allowance or discount in lieu thereof, except for services rendered.” 42 See F.T.C. v. Anheuser-Busch, Inc. 363 U.S. 536, 543 (1960). 43 See generally, Rowe, supra note 40, at 3–23. 38 39

24

The Antitrust Laws and Monopsonistic Forms

1936 amendments made it clear that the harms to be avoided were both primary and secondary line. Despite this later emphasis on the misuse of monopsony power, Section 2 continues to be interpreted in a manner that requires focusing on the seller’s reaction to the use of monopsony power as opposed to focusing on buyers who employ it. At least nominally, this possibly mistaken emphasis on sellers, when the actual concern of the 1936 amendments was on buying power, is offset by Section 2(f), which makes it “unlawful for any person … knowingly to induce or receive a discrimination in price which is prohibited by this section.”44 Although it might seem to apply directly to buyers employing monopsony power, the prohibition was interpreted by the Supreme Court, in Great Atlantic & Pacific Tea Co. v. FTC,45 to depend on the perspective of the seller. Consequently, if the seller has a meeting competition or cost justification defense, which very well might be provided by the buyer’s claims, the seller would not be viewed as violating the section. Since the seller has not violated Section 2 of the Clayton Act as amended by the RobinsonPatman Act, the buyer has not “induced” an illegal price and, therefore, is not in violation itself. In effect, a section of the Clayton Act that seems to have been aimed directly at the use of monopsony power hardly seems much of a threat to those possessing and exercising that power. This, as it turns out, may not be all that worrisome as the impact of a broad application of the Robinson-Patman Act is thought by many to be anticompetitive itself.46

2.3.2  Tying and Exclusive Dealing Section 3 of the Clayton Act prohibits conditional sales such as tying and exclusive dealing arrangements involving “goods, wares, merchandise, machinery, supplies or other commodities.”47 It does not apply to sales of services. The textbook tying arrangement is one in which a seller sells a more desired product (the tying product) only if the purchaser also buys a less desired product (the tied product).48 By definition, tying must involve 15 U.S.C. §13. 440 U.S. 69 (1979). 46 See Roger D. Blair & David L. Kaserman, Antitrust Economics, 2d ed. (New York: Oxford University Press 2009); Robert Bork, The Antitrust Paradox 382 (New York:  Basic Books 1978); Herbert Hovenkamp, Economics and Federal Antitrust Law 272 (St. Paul, Minn.: West Publishing Co. 1985). 47 15 U.S.C. §14. 48 For a compact survey of the law and economics of tying, see Blair & Kaserman, supra note 46; Sullivan & Harrison, supra note 8 at 227–9. 44 45

The Clayton Act

25

two distinct products. At one point, the generally accepted view was that the competitive harms of tying were twofold. First, competing sellers of the tied product would be foreclosed from selling to those buyers who had acquiesced to the tying firm’s demands. Second, entry costs would be increased because a firm wishing to compete in the tied good market would be required to enter both the tying product and the tied product markets. Against this background, tying was analyzed under a rather peculiar per se standard. In fact, the 1958 case of Northern Pacific Railway Co. v. United States49 can be read as prohibiting tying whenever a substantial amount of commerce is affected. Even today, under the standards announced in the 1984 case, Jefferson Parish Hospital v. Hyde,50 tying is at least nominally per se unlawful when the seller possesses substantial market power in the tying product market, the buyer is coerced into purchasing the tied product, and a “not insubstantial” volume of commerce is affected in the tied product market. One of the trends in the post-Sylvania era of antitrust has been to reevaluate the legal status of tying arrangements. Increasingly, economists and legal scholars have come to doubt the likelihood that tying arrangements, except in special circumstances, are harmful to consumers.51 The primary rationale for this new approach is the recognition that the firm with market power has a finite amount of that power and a single profitmaximizing price it is able to charge. To “force” consumers to buy a product that they would prefer not to buy at all or to buy from some other seller is really tantamount to a price increase. Presumably, however, the firm is already making full use of its market power and would not engage in a tying practice that amounted to raising prices above profit-maximizing levels. If the tying cannot be explained as an effort to raise monopoly profit, so the argument goes, it must be the result of some other motivation. It is important to note that this is not an argument that the tying firm will not increase profit. Indeed, profits might increase if the tying arrangement resulted in 356 U.S. 1 (1958). 466 U.S. 2 (1984). 51 See Tyler A. Baker, The Supreme Court and the Per Se Tying Rule:  Cutting the Gordian Knot, 66 Va. L. Rev. 1235 (1980); Bork, supra note 46, at 365; E. Thomas Sullivan & Herbert Hovenkamp, Antitrust Law, Policy and Procedure, 5th ed. 510 (Charlottesville, Va.; LexisNexis. 2007); Ward Bowman, Tying Arrangements and the Leverage Problem, 67 Yale L. J. 19 (1957). For a contrary view, see W. David Slawson, A New Concept of Competition: Reanalyzing Tie-In Doctrine After Hyde, 30 Antitrust Bull. 257 (1985); and Victor Kramer, The Supreme Court and Tying Arrangements: Antitrust as History, 69 Minn. L. Rev. 1013 (1985). 49 50

26

The Antitrust Laws and Monopsonistic Forms

efficiencies in marketing the two items. Instead, the argument is that the tying firm may not increase the returns to its market power.52 This new approach is reflected in the courts, which have made it increasingly difficult for plaintiffs to prevail under a tying theory. They have done this primarily by finding that the two “products” were actually properly viewed as a single product53 and by emphasizing the requirement that the tying firm have true market power in the tying product market.54 Exclusive dealing involves a theme similar to that of tying except that only one product is involved.55 This results in a requirements contract in that the buyer is required to purchase all of its needs for a particular item from the same supplier. Originally, the economic objections to exclusive dealing were basically the same as those expressed about tying. The concerns were not as pronounced, however, as they were for tying primarily due to the recognition that exclusive dealing arrangements were often beneficial to both buyers and sellers.56 Thus, although the standard originally applied to exclusive dealing seemed to condemn it whenever a “substantial share of a market was foreclosed,57 this view softened well before Sylvania.58 In the post-Sylvania period, recognition of the fallacies in the reasoning with respect to the economic harms of exclusive dealing and a better understanding of the possible procompetitive effects have meant that the practice is now effectively afforded rule of reason treatment.59 For an interesting dissenting view, see Louis Kaplow, Extension of Monopoly Power Through Leverage, 85 Colum. L. Rev. 515 (1985). 53 See, e.g., Jack Walters & Sons Corp. v. Morton Building, Inc., 737 F.2d 698 (7th Cir. 1984), cert. denied, 469 U.S. 1018 (1984); Principe v. McDonalds Corp., 631 F.2d 303 (4th Cir. 1980), cert. denied 451 U.S. 970 (1981). 54 See Jefferson Parish Hosp. v. Hyde, 466 U.S. 2 (1984); United States Steel Corp. v. Fortner Enter., Inc., 429 U.S. 610 (1977). 55 For a brief survey of the law and economics of exclusive dealing, see Blair & Kaserman, supra note 46, Chapter 20. 56 See William B. Lockhart & Howard R. Sacks, The Relevance of Economic Factors in Determining Whether Exclusive Arrangements Violate Section 3 of the Clayton Act, 65 Harv. L. Rev. 913 (1952); Howard P. Marvel, Exclusive Dealing, 25 J. L. & Econ. 1 (1982); and Richard M. Steuer, Exclusive Dealing After Jefferson Parish, 54 Antitrust L. J. 1229 (1986). More recent analyses are provided by Eric B. Rasmussen, J. Mark Ramseyer, & John S. Wiley, Jr., Naked Exclusion, 8 Am. Econ. Rev., 1137 (1991); B. Douglas Bernheim & Michael D. Winston, Exclusive Dealing, 106 J. Pol. Econ. 64 (1998); and Jonathan M. Jacobson, Exclusive Dealing, Foreclosure, and Consumer Harm, 70 Antitrust L. J. 311 (2002). 57 See Standard Oil v. United States, 377 U.S. 293 (1949). 58 See Tampa Electric Co. v. Nashville Coal Co., 365 U.S. 320 (1961). 59 See, e.g., Roland Mach. v. Dresser Indus., 749 F.2d 380 (7th Cir. 1984); Joyce Beverages v. Royal Crown Cola, 555 F. Supp. 271 (S.D.N.Y. 1983). 52

The Clayton Act

27

Similar to tying and exclusive dealing is reciprocal dealing, which involves firm A’s conditioning its purchases from firm B on firm B’s agreement to make purchases from firm A.60 Although there are suggestions that such practices may be per se unlawful,61 recent judicial attitudes have probably tracked those found with respect to tying and exclusive dealing arrangements. A tying arrangement by a buyer would involve buying an item from a seller only if that seller would also sell another item, which it would prefer not to sell at all or to sell to another purchaser. Exclusive dealing would amount to an outputs contract under which the seller would be required to sell all of its output to the buyer if it were to sell any at all. Certainly, the arguments and counterarguments that apply to tying and exclusive dealing by sellers would apply with equal force to buyers. Section 3 of the Clayton Act, however, does not apply to buyers. Section 3 is restricted expressly to those who “lease or make a sale.” This does not mean that buyer-initiated arrangements would necessarily escape antitrust scrutiny. As in the case of tying and exclusive dealing arrangements by sellers involving services, tying and exclusive dealing by buyers would be examined under Sections 1 and 2 of the Sherman Act. There is nothing to suggest that Sherman Act standards would differ from those applied under Section 3 of the Clayton Act.62

2.3.3  Mergers Although the application of Sections 2 and 3 of the Clayton Act to buyers is problematic, there seems to be little doubt that the merger prohibitions of Section 7 apply to buyers and sellers alike. Typically, mergers are characterized as horizontal, vertical, or conglomerate. The case for limiting mergers of competing sellers is obvious in that it increases concentration and the market power of the merged firm. Similarly, a merger of competing buyers combines their monopsony power. Like all of antitrust, merger law has moved from a period in the 1960s during which virtually every horizontal merger was scrutinized very closely63 to the current more relaxed posture. The increased likelihood that K. Walters & J. Stephen, Reciprocity Reexamined:  The Consolidated Foods Case, 29 J. L. & Econ. 423 (1986), offers an interesting economic analysis. In addition, see Blair & Kaserman, supra note 46, Chapter 19. 61 See Betaseed Inc. v. U and I, Inc., 618 F.2d 1203 (9th Cir. 1982). 62 H. Hovenkamp, supra note 46, at 218. 63 See e.g., United States v. Von’s Grocery, 384 U.S. 270 (1966). An exasperated Justice Stewart remarked that “[t]he sole consistency that I can find is that in litigation under §7, the Government always wins.” 60

28

The Antitrust Laws and Monopsonistic Forms

horizontal mergers will now pass antitrust muster can be traced to four factors. First, the courts have abandoned the view that the interests to be protected in prohibiting mergers are other than economic. Thus, justifications like that found in the 1962 case, Brown Shoe Co. v. United States,64 that “Congress appreciated that occasional higher costs and prices might result from the maintenance of fragmented industries and markets. It resolved these competing considerations in favor of decentralization”65 now seem outmoded. Second, courts have become more sophisticated at the process of market definition, with the result being greater recognition of factors other than market share as determinants of market power.66 Third, still evolving interpretations of the standing and antitrust injury requirements that must be met by private plaintiffs mean that competitors of the merging firms are unlikely to bring successful private actions.67 Finally, public enforcement agencies – specifically, the U.S. Justice Department and the Federal Trade Commission (FTC) – have narrowed the scope of the mergers challenged.68 Clearly, under the guidelines now applied by these agencies, many of the public enforcement efforts of earlier years would not have taken place. Although Section 7 did not initially extend to vertical mergers,69 it is now clear that it does and that its prohibitions apply whether the acquiring firm is the supplier or buyer. Indeed, two of the more important vertical merger cases, Ford Motor Co. v. United States70 and Fruehauf Corp. v. Federal Trade Commission,71 involved acquisitions of suppliers by buyers. Initially the supposed economic problem posed by vertical mergers was the same as that cited for tying arrangements – vertical foreclosure and increased entry barriers resulting from the necessity of entering the market at two levels. The tendencies in antitrust law generally, however, have been reflected in areas of vertical mergers policy. The doubts expressed by economists and legal scholars about the supposed harms of tying arrangements are expressed with equal force to the alleged harms of vertical mergers.72 Indeed, all of the 66 67 64 65



68



69

72 70 71

370 U.S. 294 (1962). Id. at 344. See, e.g., United States v. Gen. Dynamics Corp., 415 U.S. 486 (1974). See Cargill, Inc. v. Monfort of Colo., Inc., 479 U.S. 104 (1986). See also Chapters 6 and 11 in this volume. Guidelines were initially published by the Antitrust Division of the Justice Department in 1968. These guidelines were revised in 1982 and 1984. In 1992, the Justice Department and the Federal Trade Commission issued joint guidelines. In 1950, the Celler-Kefauver Amendment to Section 7 expanded its coverage to vertical mergers. 405 U.S. 562 (1972). 603 F.2d 345 (2d Cir. 1979). For comprehensive analysis, see Alan A. Fisher & Richard Sciacca, An Economic Analysis of Vertical Merger Enforcement Policy, 6 Res. L. & Econ. 1 (1984).

A Taxonomy of Monopsony Cases

29

factors accounting for the decreased scrutiny of horizontal mergers apply at least as persuasively to vertical mergers. Currently, the Justice Department Merger Guidelines express concern about vertical mergers primarily when there is a resulting increase in the difficulty of entry that increases the likelihood of monopolization or collusion.73 This view is reflected in Freuhauf, in which the manufacturer of truck trailers acquired the manufacturer of a variety of components.74 The claim was that the merger would reduce competition in the truck trailer market as well as in markets for the components manufactured by the acquired firm. Although the court focused on the typical foreclosure problems, the risk of foreclosure was regarded as relevant to the extent that horizontal competition would decrease. In Fruehauf, the horizontal effects of interest to the court were on the selling side of the market. This is not to say that the fears expressed about the horizontal impact on the selling side of the market would not apply to the buying side as well. For example, a series of vertical mergers could enable the integrated firms to act as a collusive monopsony.

2.4 A Taxonomy of Monopsony Cases Monopsony power has been examined by the courts under the federal antitrust laws in a variety of contexts ranging from price fixing to group boycotts and mergers. For the most part the analysis mirrors that found in the more typical seller-side cases. As in seller-side cases, this does not mean that it is always easy to correctly identify the practice as fitting into one well-defined category. For example, horizontal price fixing may begin to become blurred with boycotts, and efforts to monopolize may actually involve the use of monopsony power. The analysis of monopsony is also complicated by the fact that courts occasionally examine the use of monopsony power while speaking the language of seller-side economics.75 In addition, conduct that would fall under the Clayton Act if engaged in by sellers must be examined under the Sherman Act when buyers are involved. For purposes of evaluating monopsony cases from the standpoint of economic efficiency, it makes sense to adopt a “purpose”-oriented classification system. After all, all monopsony cases can be reduced to either a unilateral or collusive use of buying power in order to promote the interests of the buyer. This means obtaining a lower price or some other concession in the terms of the exchange with sellers. This approach to classification 1984 United States Department of Justice Merger Guidelines, §4.22. 603 F.2d 345 (1979). 75 See, e.g., United States v. Griffith, 334 U.S. 100 (1948). 73 74

30

The Antitrust Laws and Monopsonistic Forms

is followed in Chapter 4. Here, however, it is more instructive to survey the richness of monopsony by attempting to track the relevant legislation. Section 2.4.1 examines the types of violations by buyers that are similar to those that would be examined under Section 1 of the Sherman Act if sellers had undertaken them. Section 2.4.2 is devoted to single-firm conduct. The question of mergers and monopsony power is examined in Section 2.4.3.76

2.4.1  Collusive Monopsony Whether buyers are as frequently involved in horizontal price-fixing schemes as sellers is an empirical question that cannot be answered with certainty. What is clear is that buyers have given in to the temptation to fix prices and have, for the most part, been treated in the same manner as sellers.77 An early example is Live Poultry Dealers’ Protective Association v. United States,78 in which 178 wholesale poultry buyers in New York united and appointed a committee of seven individuals to determine the market price to be paid for poultry each day. According to the defendants, the “stabilizing” of prices would “eliminate opportunities for bad trade practices.”79 Judge Hand did not hesitate to affirm the lower court’s grant of an injunction against the poultry buyers. Accordingly, he wrote, “it is somewhat surprising at this day to hear it suggested that a frank agreement to fix prices … may be defended in the notion that the results are economically desirable. … [I]f one thing were definitely settled, it was that the Sherman Act forbade all agreements preventing competition in price among a group of buyers, otherwise competitive, if they are numerous enough to affect the market.”80 Although Judge Hand stops slightly short of declaring the agreement per se unlawful, this does not represent a deviation from the judicial attitude at that time toward price fixing by sellers. In fact, as late as 1927, the Supreme This effort to categorize cases is necessarily imprecise. In many cases, the same conduct is analyzed under both sections of the Sherman Act. 77 Two commentators point out that cases in which buyers have been found to have engaged in price fixing by and large involve instances in which there are competing sellers. See Jonathan Jacobson & Gary Dorman, Joint Purchasing, Monopsony and Antitrust, 36 Antitrust Bull. 1 (1991). Whether this means that price fixing by buyers is only per se unlawful when competing sellers are involved, as the authors intimate, is not supported by this correlation and does not seem to be supported by the reasoning of courts in buyer price-fixing cases. Our view that collusive monopsony alone would violate the antitrust law seems to be shared by Judge Richard Posner; see Vogel v. American Society of Appraisers, 744 F.2d 598, 601 (7th Cir. 1984). 78 4 F.2d 840 (2d Cir. 1924). 79 Id. at 841. 80 Id. at 842–3. 76

A Taxonomy of Monopsony Cases

31

Court, in United States v. Trenton Potteries Co.,81 noted that price fixing was unlawful “if effective.”82 Probably the best known buyer-side price fixing case is Mandeville Island Farms v. American Crystal Sugar Co.,83 which addressed the efforts of sugar refiners to fix or control the price paid to sugar beet growers in California. The lower court had dismissed the complaint on the basis of the assertion that the purchase of sugar beets in California did not involve interstate commerce. The Supreme Court reversed and in so doing stated its policy toward price fixing by buyers: “It is clear that the agreement is the sort of combination condemned by the Act, even though the price-fixing was by purchasers, and the persons injured under the treble damage claim are sellers not customers or consumers.”84 The Court relied expressly on the landmark price-fixing case, United States v. Socony Vacuum, which announced as per se unlawful all efforts aimed at “raising, depressing, fixing, pegging or stabilizing” prices.85 It has been suggested that the Court’s approach to price fixing by buyers in Mandeville Island Farms is somewhat less rigorous than its approach to price fixing by sellers.86 Certainly, in its analysis, the Court does indicate that the sellers of sugar beets in California had little recourse than to sell at artificially depressed prices.87 This language can be viewed as indicating that the Court required a show of effectiveness in the buyer-side case, a view that would differ from that afforded sellers. On the other hand, the Court seems more concerned with expressing how the price fixing could have interstate ramifications. That likelihood along with its express reliance on Socony Vacuum indicates that buyers were regarded no differently than sellers.88 83 84 85 86

273 U.S. 392 (1927). Id. at 397. 334 U.S. 835 (1947). Id. at 235 (notes deleted). 310 U.S. 150, 223 (1940). See Jack Rovner, Monopsony Power in Health Care Markets: Must the Big Buyer Beware Hard Bargaining? 18 Loy. U. Chi. L. J. 857 (1987). 87 Supra note 83 at 240. 88 On remand, American Crystal was found to have engaged in a price-fixing combination. American Crystal Sugar Co. v. Mandeville Island Farms, 195 F.2d 622 (9th Cir.) cert. denied, 343 U.S. 957 (1952). A more recent case in which price fixing by a collusive monopsony is held out as a possibility is St. Bernard General Hospital v. Hospital Service Association of New Orleans, 712 F.2d 978 (5th Cir. 1982). For a modern case involving an allegation of price fixing in the pharmaceutical industry, see North Jackson Pharmacy, Inc. v. Express Scripts, 345 F.Supp.2d 1279 (N.D. Ala. 2004); North Jackson Pharmacy, Inc. v. CaremarkRX, Inc., 385 F.Supp.2d 740 (N.D. Ill. 2005). 81 82

32

The Antitrust Laws and Monopsonistic Forms

Of course, not all price-fixing efforts by sellers or buyers take the simplistic form of an actual horizontal agreement on price. Indeed, it appears that buyers have been as imaginative as sellers in designing ways to depress prices without actually agreeing on the prices to be paid. Following the broad mandate of Socony Vacuum, however, courts have found these price stabilization efforts illegal. Probably the best known of these efforts was considered in the 1965 case, National Macaroni Manufacturers Association v. Federal Trade Commission.89 To understand Macaroni – the case – it is important to understand a little about macaroni  – the product. High-quality macaroni is made from 100 percent semolina, which is defined as “middlings of durum wheat in granulated form with a tolerance of 3% flour.” The use of durum makes the macaroni easier to work with and results in “superior cooking tolerances and color.”90 In the early 1950s and early 1960s, the trade association made up of 84 of 125 domestic manufacturers of macaroni evidently became concerned that crop failures would cause the price of durum wheat to increase. In addition, exporters were entering the market and purchasing from domestic growers, further increasing the upward pressure on price. The solution, in the view of the domestic producers, was twofold. First, the Secretary of Agriculture was asked to limit the export of durum wheat.91 Second, the Association in both 1953 and in 1961 agreed to a uniform recipe for macaroni that was 50 percent durum and 50 percent farina. The Federal Trade Commission concluded that both the purpose and effect of the agreement was to depress the price of durum wheat. Accordingly, the FTC issued a cease and desist order, which was approved by the Court of Appeals. In its opinion, the court reviews the structural characteristics of the market and indicates that the macaroni industry is the only market for wheat that the Association members were able to dominate through their agreement. Despite this language, it would be incorrect to view the court as applying anything other than a per se standard to the Association’s efforts. This discussion was for the purpose of summarizing the findings of the Commission. The court closes with the observation that “[t]he combination found in the instant case is illegal per se.”92 Macaroni is not the only instance in which limiting purchases has been linked to lowering prices. Probably the most well-known recent case involved 91 92 89 90

345 F.2d 421 (7th Cir. 1965). Id. at 434. This measure was evidently taken only during the second period of possible shortages. Supra note 90, at 437.

A Taxonomy of Monopsony Cases

33

a suit by “walk-on” or non-scholarship football players challenging the NCAA’s limit on the number of players who could receive scholarships.93 Another variety of price fixing takes the form of bid rigging. When done by would-be competing buyers, the buyers determine in advance who will have the right to purchase. For example, in the 1946 case of American Tobacco Co. v. United States,94 the three major producers of cigarettes were charged with conspiring to eliminate competition in the purchase of leaf tobacco. Tobacco was sold at auctions, primarily in the Carolinas, Virginia, Georgia, and Florida. The producers would not participate in an auction unless all three producers were present. Thus, they were able to control the locations of markets and ultimately the access of competing producers to tobacco because sellers found it unattractive to market tobacco if the “Big Three” were not present. The purchasers evidently also agreed prior to the auction on the maximum high bid. In addition, they formulated grades of tobacco and did not compete for those grades.95 In order to avoid cheating, the purchaser would remain in the market even after having purchased the amount desired. This would ensure that the other purchaser would not buy at a lower price. The scheme was conceived not to achieve the lowest input prices but to ensure that the major producers had uniform input prices.96 More recently a 1983 case, Reid Brothers Logging Company v. Ketchikan Pulp Co.,97 dealt with the activities of two logging companies operating in the Tongass National Forest in southeast Alaska. The two companies were allotted, under long-term contracts with the U.S. Forest Service (USFS), the logging rights to specific areas. In addition, they were expected to bid for the rights to other timber, also offered by the USFS or by independent loggers, in order to use the full capacity of the pulp plants and sawmills they had established. In a combination of arrangements that were found to be per se unlawful, the defendants refused to bid against each other for timber even during a shortage.98 The bid rigging in this instance was effectuated by a horizontal division of the Tongass National Forest with each party agreeing to restrict its purchasing activities to its designated portion of the forest. In a sixteenyear period from 1959 through 1975, the buyers had bid against each other Ultimately the case was unsuccessful as the court was unwilling to certify the class of walk-on players. In re NCAA I-A Walk-On Football Players Litigation, 2006 WL 1207915 (W.D.Wash. 2006). It is discussed more fully in Chapter 9. 94 326 U.S. 781 (1946). 95 Id. at 801–2. 96 Id. at 802. 97 699 F.2d 1292 (9th Cir. 1983), cert. denied, 464 U.S. 916 (1983). 98 This refusal is analogous to a seller’s refusal to bid for new business in the presence of excess productive capacity. 93

34

The Antitrust Laws and Monopsonistic Forms

in only 3 of the 143 USFS sales. The scheme was further supplemented by an agreement on the method of determining the prices to be paid to loggers.99 United States v. Champion Int’l Corp.100 is another logging case, similar to Reid Bros. It is particularly interesting because the conspiracy allegedly arose quite “innocently” from a “tit-for-tat” strategy when one of the parties was “surprised” to find that he was bidding alone. As an experiment, he decided to reciprocate by declining to bid against his chief rival at another sale. The rest, as they say, is history. Another variation of what is effectively bid rigging occurs in the theater “split” cases. An example is United States v. Capitol Service Inc.,101 in which the defendants operated 90 percent of the first-run motion picture theaters in Milwaukee. Their understanding amounted to a film by film division of the market under which specific films were allocated to certain theaters. An interesting feature of the case is the position taken by the defendants that the arrangement was a “good” split and, therefore, not subject to the per se rule. The split was not a complete ban on competitive bidding but resulted in a “right of first negotiation,” which meant that, in theory, other theaters could bid on individual films. In finding the so-called “good” split to be per se unlawful, the Court of Appeals for the Seventh Circuit ruled that even though there was the theoretical possibility that competing buyers could bid on the films once they were allocated, the time pressure under which such efforts would have to take place really precluded distributors from turning to other exhibitors. In effect, the so-called right of first negotiation amounted to the creation of a transaction cost that would impede competitive bidding.102 Id. at 1296. See also Cackling Acres, Inc. v. Olson Farms, Inc., 541 F.2d 242 (10th Cir. 1976), cert. denied, 429 U.S. 1122 (1977). For bid rigging in an agricultural context, see Knevelbaard Dairies v. Kraft Foods, Inc., 232 F.3d 979 (9th Cir. 2000). 100 557 F.2d 1270 (9th Cir. 1977), cert. denied, 434 U.S. 938 (1977). 101 756 F.2d 502 (7th Cir. 1985), cert. denied, 474 U.S. 945 (1985), aff ’d, 568 F. Supp. 134 (E.D. Wis. 1983); see also Harkins Amusement Enter. v. Gen. Cinema Corp., 850 F.2d 477 (9th Cir. 1988), cert. denied, 488 U.S. 1019 (1989); Gen. Cinema v. Buena Vista Distrib. Co., 532 F.Supp. 1244 (C.D. Cal. 1982). For a “split” case expressly declining to apply the per se standard, see Balmoral Cinema, Inc. v. Allied Artists Pictures Corp., 885 F.2d 313 (6th Cir. 1989). 102 Id. at 505–6. The court relied on National Society of Professional Engineers v. United States, 435 U.S. 679 (1978), in which the Supreme Court reviewed an agreement among engineers not to engage in competitive bidding. Although the Capital Service court clearly regarded its buyer-side version of that case as per se unlawful, it is not clear that the Supreme Court applied the per se standard in Professional Engineers. See generally, Harrison, supra note 26. 99

A Taxonomy of Monopsony Cases

35

A recurring theme in the context of collusive monopsony cases is an effort by employers to lower the compensation paid to employees. Such was the case in Law v. National Collegiate Athletic Association103 in which colleges and universities agreed to a maximum compensation for some basketball coaches. The practice of the NCAA in limiting the compensation to scholarship players has also been challenged as a collusive monopsony.104 In a different employment context, the issue was addressed in Todd. V. Exxon Corporation,105 in which professional-level employees claimed that petrochemical companies had exchanged data in an effort to collusively set maximum salaries. In Hall v. United Air Lines Inc.,106 the complaint was that airlines had conspired to lower commissions paid to travel agents. Aside from horizontal agreements to literally fix prices or to alter the bidding process in order to depress prices, buyers have also joined in order to use their monopsony power as a means of eliminating competitors. For the most part, the ultimate goal in these cases has been to affect the selling side of the market. On the other hand, there are often independent advantages of decreased competition among buyers,107 and the leverage employed comes from the buying side. Three cases that have been important in the evolution of antitrust law reflect the indifference of the courts to whether pressure is applied by buyers or by sellers. In the 1914 case, Eastern States Retail Lumber Dealers’ Association v. United States,108 retailers of lumber circulated among themselves the names of wholesalers who were attempting to compete at the retail level. The list was part of a blackballing scheme used by retailers who felt they had “been unduly interfered with by the whole­ salers in selling to customers.”109 Even though the plan was designed to protect the interests of the retailers as sellers, the “engine” for the plan was buying-side power. Without monopsony power, the threat by the dealers could have been ignored by the wholesalers. In United States v. Crescent Amusement Co.,110 the Supreme Court reviewed a similar arrangement involving nine companies that operated 185 F.R.D. 324 (D.C. Kansas, 1999). NCAA Settles Antitrust Lawsuit; Will Create $218 Million Fund for Athletes, University Business (Jan. 31, 2008), http://www.universitybusiness.com/newssummary.aspx?news= yes&postid=15221 (last visited Oct. 14, 2008). See In re Walk-On Players Litigation, 2006 WL 1207915. 105 275 F.3d 191 (2d Cir. 2001). See also U.S. v. Brown University, 5 F.3d 658 (C.A. Pa. 1993). 106 296 F.Supp. 652 (E.D. N.C. 2003). In this case, the plaintiffs were unable to establish that an agreement had been made. 107 For a formal economic analysis, see Chapter 3 in this volume. 108 234 U.S. 60 (1914). 109 Id. at 606. 110 323 U.S. 173 (1945). 103 104

36

The Antitrust Laws and Monopsonistic Forms

motion picture theaters in seventy small towns located in the southeastern United States. In the words of the Court, “the combination used its buying power for the purpose of restricting the ability of its competitors to license films.”111 In effect, the members of the conspiracy insisted that they be given the exclusive rights to exhibit films in towns in which they faced competition or they would not exhibit a distributor’s films in those towns where they had monopoly power. The Court noted that competing exhibitors would frequently go out of business or sell out to the conspiring exhibitor “‘because his ‘mule was scared.’”112 The final variation of the use of monopsony power under a conspiracy theory as a means of decreasing seller competition is found in Klor’s v. Broadway-Hale Stores, Inc.113 Broadway-Hale operated a chain of department stores, one of which was next to Klor’s, with which it competed in the sale of radios, televisions, and appliances. Evidently, Broadway-Hale did not conspire with other retailers, but was able to use its “‘monopolistic’ buying power”114 to convince a number of manufacturers and distributors of national brands to either stop selling to Klor’s or to sell only at a discriminatory price. Broadway-Hale was granted summary judgment, and the Supreme Court reversed in an opinion that, until it was “reinterpreted” recently by the Court,115 was seen as standing for the per se illegality of group boycotts.116

2.4.2  Single-Firm Conduct The examination of collusive monopsony is not limited to cases in which horizontal agreements are involved. There are also instances in which single firms are examined under the antitrust laws as a result of their agreements with firms vertically aligned or as a result of unilateral action. The best known of these instances is United States v. Griffith,117 which the Court analyzed in Section 2 terms even though it involved an agreement among four corporations to use a common agent and approach to dealing with film distributors. The facts in Griffith were similar to those in Crescent Amusement discussed Id. at 181. Id. Whatever that means. 113 359 U.S. 207 (1959). 114 Id. at 209. 115 See Bus. Elec. v. Sharp Elec., 485 U.S. 717 (1988). 116 Klor’s, 358 U.S. at 212. For two failed efforts to use a collusive monopsony boycott theory, see Les Shockley Racing v. National Hot Rod Association, 884 F.2d 504 (9th Cir. 1989); Pan-Islamic Trade Corp. v. Exxon Corp., 632 F.2d 539 (5th Cir. 1980). 117 334 U.S. 100 (1948). 111 112

A Taxonomy of Monopsony Cases

37

earlier. According to the Court, the cases were factually different in that, in the latter, the defendants “through the pooling of their buying power increased their leverage over their competitive situations by insisting that they be given monopoly rights in towns where they had competition.”118 The suggestion was that the defendants in Griffith were more subtle in achieving the same goals in that a single circuit-wide contract was negotiated. The Supreme Court found the distinction largely irrelevant and reassured that “circuit buying power” that had the effect of creating a monopoly violated Sections 1 and 2 of the Sherman Act.119 Cases like Griffith in which the objective is exclusive access to some input that would then result in an advantage for the firm as a seller are similar in theme to the exclusive dealing and tying arrangements that are often scrutinized under the Clayton Act when sellers are involved. A good example is Federal Trade Commission v. Motion Picture Advertising Service Co.,120 in which a producer and distributor of advertising motion pictures purchased exhibition rights from theaters. At issue was the requirement that those theaters selling rights to Motion Picture Advertising sell only to that firm. As noted previously, Section 3 of the Clayton Act extends to requirements contracts – the seller’s equivalent of such an exclusive dealing arrangement. It does not, however, extend to outputs contracts. Thus, the FTC relied on Section 5(a) of the Federal Trade Commission Act121 and the Sherman Act in challenging the agreements. The Supreme Court approved the Commission’s limitations on the length of the contracts.122 Another version of the use of monopsony leverage occurs in the reciprocal dealing cases. Again, these cases do not fit neatly into the Clayton Act but do have a tying-type theme.123 For example, in Federal Trade Commission v. Consolidated Foods,124 a 1965 case, the acquisition of a producer of dehydrated Id. at 105. For three more recent, albeit failing, efforts to employ a Griffith-type theory, see White Mule Co. v. ATC Leasing Co. LLC, 540 F.2d 869 (N.D. Ohio 2008); Christensen v. Colt Industries, 766 F. Supp. 670 (C.D. Ill. 1991); Camellia City Telecasters v. Tribune Broadcasting Co., 762 F. Supp. 290 (D. Colo. 1991). 120 344 U.S. 392 (1953). 121 Supra note 1. 122 A case involving the use of leverage in a context similar to both Griffith and Motion Picture Advertising Services is Reading International Inc. v. Oaktree Capital Management, LLC, 317 F.Supp.2d 3–1 (S.D.N.Y. 2003) in which an independent movie theater claimed that a chain had pressured distributors not to permit the independent to show first-run films. 123 See Lawrence Sullivan, Antitrust 490 (St. Paul, Minn.: West Publishing Co. 1976). Note, A Reevaluation of Reciprocal Dealing Under the Federal Antitrust Laws: Spartan Grain v. Ayers, 11 Loy. U. Chi. L. J. 577 (1980). 124 380 U.S. 592 (1965). 118 119

38

The Antitrust Laws and Monopsonistic Forms

vegetables by a major food processor was challenged. The rationale was that Consolidated, the acquiring firm, could condition its future purchases from other suppliers on their agreement to purchase dehydrated vegetables from the newly acquired firm. In effect, Consolidated would be able to use its monopsony power as a means of increasing sales. The Supreme Court approved the Commission’s divestiture order, reasoning that “reciprocity made possible by such an acquisition is one of the congeries of anticompetitive practices at which the antitrust laws are aimed.”125 A more direct attack under Sections 1 and 2 of the Sherman Act is found in Betaseed Inc. v. U and I Incorporated,126 in which the seller of sugar beet seeds challenged the practice of a competing seller of buying sugar beets for processing only from growers who had bought its seeds. The defendant was the only processor of sugar beets in the relevant market. The Court of Appeals for the Ninth Circuit found the practice analogous to tying and indicated that the same standard should apply.127 An issue that has emerged as courts have become more acquainted with monopsony is whether the antitrust laws prohibit a firm with monopsony power from using its power to lower prices. This is an odd question in some respects because it is virtually a given that price setting by a seller to take full advantage of any monopoly power it might have is not a violation of the antitrust laws.128 Courts considering monopsony pricing and seeing the analogy to monopoly price setting have indicated that the use of buying power to obtain a lower price does not violate the antitrust laws.129 There is, as noted previously, the possibility that a seller may lower its prices to predatory levels in order to eliminate competitors. In the aftermath of their departure from the market, so the theory goes, the predator raises prices in order to recoup the losses incurred during the period of predation. The monopsony version of this would involve raising the prices paid for inputs in order to raise the costs of its rivals or deny them access to necessary inputs. The ultimate goal is to exclude the competitors from Id. at 594–5. See Walters, supra note 60, for a different economic analysis of the facts in this case. 126 681 F.2d 1203 (9th Cir. 1982). 127 Id. at 1221. See also Spartan Grain & Mill Co. v. Ayers, 581 F.2d 419 (5th Cir. 1978) cert. denied 444 U.S. 831 (1979); United States v. Gen. Dynamics Corp., 258 F. Supp. 36 (1966). 128 For more on this, see Section 4.1. 129 See e.g., White Mule co. v. ATC Leasing LLC, 540 F. Supp.2d 869 (N.D. Ohio 2008); Kamine/Besicorp Allegany L.P. v. Rochester Gas & Electric Corp., 908 F. Supp.2d 1194 (W.D.N.Y. 1995); Ball Mem’l Hosp., Inc. v. Mutual Hosp. Ins., Inc., 784 F.2d 1325 (7th Cir. 1986); Kartell v. Blue Shield of Mass., Inc., 749 F.2d 922 (1st Cir. 1984) cert. denied, 471 U.S. 1029 (1985). 125

A Taxonomy of Monopsony Cases

39

the input market by investing in overpaying for the input. The firm then may recoup its investment by lowering the price for those inputs. Or, if the competing buyer also sells in the same output market, the firm may recoup its losses there. This theory and a discussion of the likelihood that such a strategy would be successful are reserved for Chapter 4. It is important to note that the practice does not exist simply in theory. For the purposes at hand, it is useful to note that this theory has been employed as a basis for an antitrust claim on a number of occasions, sometimes with apparent success. One successful use of such a theory came in American Tobacco v. United States, a 1946 case130 involving the leading producers of cigarettes and their effort to deny access to tobacco by competitors offering cheaper “ten cent brands.” In an effort to stave off the competition, the three leading producers bid aggressively for the less expensive tobacco used in those brands even though they did not incorporate that tobacco into their own brands. Accordingly, it was claimed that “such purchases of cheaper tobacco evidenced a combination and a purpose … to raise the price of such tobacco to such a point that cigarettes made there could not be sold at a sufficiently low price to compete with the petitioners’ more highly advertised brands.”131 Another successful use of this theory came in Reid Brothers Logging,132 previously discussed in the context of horizontal price fixing. In addition to the practices already discussed, the logging companies also sought to exclude competing buyers of logs. Thus, when bidding against another firm, one of the defendants would “run the bidding up … to the point it will really hurt.”133 This practice was among those condemned by the court. In 2007, in Weyerhaeuser Co. v. Ross-Simmons Hardwood Lumber Co.,134 the Supreme Court had an opportunity to address directly the standard to be employed in these instances of “predatory bidding.” It relied on the similarity of monopsony to monopoly and applied the standard developed in earlier predatory pricing cases.135 368 U.S. 781 (1946). Id. at 804–5. 132 699 F.2d 1292 (9th Cir. 1983), cert. denied, 464 U.S. 916 (1983). 133 Id. at 1297. For an unsuccessful use of the predatory buying theory, see In re Beef Industry Antitrust Litigation, 907 F.2d 510 (5th Cir. 1990). 134 127 S.Ct. 1069 (2007). 135 The general rule that nonpredatory monopsony price determination alone does not violate the Sherman Act was challenged more recently in U.S. v. Delta Dental of Rhode Island. There, Delta Dental, as insurer, reimbursed dentists for care provided to insureds. Under the contracts with the dentists, if they accepted a lower reimbursement rate, that rate would automatically be applied to Delta as well. The practice was challenged as ultimately leading to less competition in the market for insurance and higher insurance rates. 943 F.Supp. 172 (D. Rhode Island 1996). 130 131

40

The Antitrust Laws and Monopsonistic Forms

2.4.3  Mergers As discussed in the preceding overview, concerns about mergers, whether horizontal or vertical, are about the effect on horizontal competition. This competition can take place on either the buying or selling side of the market. Although not possible to determine as an empirical matter, it does appear that, recently, there has been increased attention to potential decreases in competition among buyers.136 An excellent example is United States v. Rice Growers Association,137 in which the U.S. Department of Justice successfully challenged the acquisition of the facilities of one rice-milling firm by another rice-milling firm. The product market was “the purchase or acquisition for milling of paddy rice grown in California.”138 Together the two firms accounted for approximately 60 percent of the rice purchased in the market.139 Similarly, in United States v. Pennzoil,140 the Department of Justice successfully challenged the acquisition of Kendall Refining Company by Pennzoil based on the position that “actual and potential competition between these corporations in the purchase of Penn Grade [crude oil] would be eliminated.”141

2.5  Concluding Remarks This general overview of the antitrust laws and the ways in which they can be used to attack monopsony power illustrate that monopsony may be more important than previously understood. In the next chapter, we set out the basic monopsony model and demonstrate the adverse consequences for social welfare. This will provide the intellectual foundation for an antitrust policy designed to curb the competitive harms of monopsony. For example, one recent merger between health insurers was challenged not simply because there would be less competition in the sale of insurance but because there would be less competition for the purchase of provider sevices. See U.S. v. Aetna, Inc., 1999 WL 1419046 (N.D. Tex 1999). See also Alliedsignal v. B. F. Goodrich Co., 183 F.3d 568 (7th Cir. 1999). 137 1986 WL 12561 (E.D. Cal. 1986). 138 Id. at 1. 139 The court also considered the buyer-side Herfindahl-Hirshman Index before and after the merger. 140 252 F. Supp. 962 (W.D. Pa. 1965). 141 Id. at 967. For an analysis of merger to monopsony in the context of a highly regulated industry, see Environmental Action, Inc. v. F.E.R.C., 939 F.2d 1057 (D.C. Cir. 1991). 136

THREE

Economic Theory of Monopsony

3.1 Introduction Pure monopsony is the demand-side analog of monopoly. Just as a monopolist is a single seller, the monopsonist is a single buyer. In the same sense that a monopolist has market power in selling its output, the monopsonist has buying power in purchasing some of its input requirements. The economic objections to monopoly and monopsony are similar: The exercise of market power reduces social welfare. These are the topics that we explore in this chapter.1 Initially, we examine the profit-maximizing behavior of a monopsonist and contrast that to a competitive buyer. The welfare effects of monopsony are developed and extended to the case of collusive monopsony. Here, we review the problems associated with organizing and implementing a buyer cartel. Finally, we turn our attention to the measurement of monopsony power.

3.2 A Simple Model of Monopsony In order to examine the welfare consequences of monopsony, we begin with a simple, stylized model of a textile mill in an isolated town in North Carolina. This mill produces textiles (T) by employing labor (L) and machinery (M) according to the following production function:2 T = T(L, M)

(3.1)

A somewhat more technical treatment of the theory can be found in Roger D. Blair & Christine Piette Durrance, The Economics of Monopsony, in Issues in Competition Law and Policy, W. Dale Collins, ed. (Chicago: American Bar Association 2008). 2 We have assumed a two-input production function for expositional convenience. Our results readily generalize to more complicated settings where the firm employs multiple inputs. 1

41

42

Economic Theory of Monopsony

Due to the town’s isolation, the mill enjoys some monopsony power in hiring labor in the local labor market. As a result, the wage (w) that the textile mill pays is a function of the number of people that are hired: w = w(L)

(3.2)

We assume that this supply curve for labor is positively sloped (i.e., dw(L)/ dL > 0).3 Since this textile mill purchases machinery in a national market, it can buy as many machines as it wishes at the market-determined price. In other words, it is a competitor in the machinery market. Since the firm has already selected the plant size, the number of machines and their configuration are fixed in the short run.4 As a result, the manager of the textile mill will adjust the number of employees to maximize profits (π). Assuming that this textile mill sells its output in a competitive output market, the firm’s profit function can be written as π = P∙T – w(L)L – pM

(3.3)

where P is the price of the textiles, and M is the fixed number of machines, and p is the machine price. In particular, the plant manager will adjust the quantity of labor until the marginal impact on profit is zero (i.e., until the first partial derivative of (3) vanishes): ∂π ∂T =P. ∂L ∂L

(

(L)+ Ldw(L)/dL ) = 0

(3.4) 

Since ∂T/∂L is the change in output flowing from a one unit increase in the quantity of labor employed, it is the marginal product of labor (MPL). The product of the output price (P) and the marginal product of labor is the value of the marginal product of labor, which we may denote by VMPL. When the textile mill employs an additional worker, it must pay a higher wage because the supply of labor is positively sloped. But it cannot pay the higher wage to just that added worker; it must pay everyone the higher wage.5 Thus, when the manager adds a worker, there are two effects on the total wage bill: (1) the new employee must be paid at the prevailing wage, which is w(L), and (2) the compensation of all the other employees (L) will It is well known that even a single buyer will have no monopsony power if the supply curve is flat. This will become apparent when we examine the buying power index in Section 3.5. 4 Focusing our attention on the short run makes the exposition less cumbersome, but our results will generalize to the long run. 5 If the mill tried to discriminate by paying a higher wage to only the added worker, others would quit and be re-hired at the higher wage. In this way, arbitrage would destroy these efforts at wage discrimination. 3

A Simple Model of Monopsony

43

Price MFC

a

b S w1 w2

c

f

d D

e 0

Q2

Q1

Quantity

Figure 3.1.  Competitive and Monopsony Conditions Compared.

rise by the increase in the wage, (dw(L)/dL). The sum of these two effects is the marginal factor cost of labor (MFCL): MFCL = w(L) + L dw(L)/dL

(3.5)

Thus, profit maximization requires expanding employment until the marginal value contributed by that expansion (VMPL) is just equal to the marginal cost of the expansion (MFCL).

3.2.1  The Welfare Effects of Monopsony The consequences of this behavior on social welfare can be seen with the aid of Figure 3.1. If the textile mill ignored the effect that its employment decisions actually have on the wage rate (i.e., if it had behaved as a competitor in the labor market), then it would have hired the quantity where demand and supply were equal, which is shown as Q1 in the figure. The wage that corresponds to this quantity is shown as w1. This wage–quantity combination maximizes social welfare, which is the sum of the consumer surplus and producer surplus.6 Consumer surplus represents the difference between 6

See e.g., Hal Varian, Intermediate Microeconomics 258 (New York: W. W. Norton 1987).

44

Economic Theory of Monopsony

what consumers are just willing to pay for a good and what they have to pay in the market. Producer surplus analogously represents the difference between the price that producers are just willing to accept and what they receive in the market. In Figure 3.1, consumer surplus is the area under the demand curve and above w1, which is the area afw1. The producer surplus is the area above the supply curve and below w1, which is efw1. Absent monopsony power, the forces of supply and demand will lead to a maximization of social welfare.7 We cannot expect the monopsonist to behave as though it were a pure competitor since this would require myopia or behavior that is inconsistent with profit maximization. Because the monopsonist’s purchase decisions influence the wage rate, the monopsonist will take this effect into account when making its decisions on the quantity of labor to employ. As we have seen, it will hire where the marginal factor cost (MFC) equals demand (D), which is the value of the marginal product.8 Consequently, it will purchase quantity Q2. Instead of paying a wage of w1 for labor, it will pay w2, which is the lowest price consistent with the purchase of Q2. The social welfare effects of monopsony are analogous to those of monopoly – too few resources will be employed. At the point where supply and demand are equal, the value of the labor services as measured by the value of the marginal product (or the demand price) is just equal to the cost to society of providing that quantity as measured by the supply price. At this point, the employment level is optimal in a social sense because all of the gains from trade have been realized.9 But the monopsonist will not hire this number of units because it is not privately optimal to do so. Private profit maximization will lead the monopsonist to employ the smaller quantity of Q2. As a result, too few resources will be employed. In Figure 3.1, one can see that the value of the marginal product at Q2 exceeds the social cost, which is measured by the height of the supply curve.10 This means that This is the intellectual foundation for the economist’s preference for competition over other market structures. Once the total pie is maximized, distributional considerations can be resolved (in principle, at least) through transfer mechanisms such as taxes and subsidies. 8 Technically, a monopsonist does not have a demand function because the monopsonist may buy the same quantity at two different prices depending on the supply conditions. But the value of the marginal product curve still contains the same information  – the addition to total revenue that an increment of the input will generate. As a result, it serves much the same purpose as a demand function. See Milton Friedman, Price Theory 189 (Chicago: Aldine Publishing 1976). 9 This, of course, assumes that there are no externalities associated with that employment. In that event, what is privately optimal is also socially optimal. 10 This difference between the marginal value of an input to an employer and the price paid for the input led Joan Robinson to complain about “monopsonistic exploitation.” See Joan Robinson, The Economics of Imperfect Competition 293 (London: Macmillan & Co. 1933). 7

A Simple Model of Monopsony

45

there are unrealized gains from further trade (i.e., that some social value is forgone by the restricted hiring decision). In Figure 3.1, the producer surplus declines by the area w1fdw2 while the consumer surplus rises by the difference between the rectangle w1cdw2 and triangle bcf. As a result, there is a deadweight social welfare loss equal to the area of triangle bdf. The rectangle w1cdw2 measures the distributional change as it represents a wealth transfer from suppliers to buyers. Typically, we do not worry about distributional effects, as everyone’s preferences count. As Gordon Tullock and Richard Posner have warned, however, these distributional gains provide an incentive for socially wasteful expenditures of resources in pursuit of those gains.11 For our purposes here, we focus on the traditional concern with the allocative inefficiency of monopsony.

3.2.2  The Effect of Monopsony on Output Price Since the monopsonist extracts a lower price from its suppliers, it is tempting to infer that this will lower its costs and thereby benefit consumers through lower prices of the final goods or services that the monopsonist produces. This is precisely the mistake that the Sixth Circuit Court of Appeals made in its Balmoral decision.12 This case involved a “split” agreement among movie theaters in which the exhibitors agreed not to engage in competitive bidding for films offered by the distributors. Standard bright line antitrust analysis might have resulted in this practice being treated as per se unlawful as an agreement to fix price, to engage in a group boycott, or to allocate customers. Instead, the court noted that the practice of the colluding buyers “may simply lower prices paid by exhibitors to distributors.”13 The action, according to the court, “may lower prices to movie goers at the box office and may serve rather than undermine consumer welfare.”14 Although the See Gordon Tullock, The Welfare Costs of Tariffs, Monopolies, and Theft, 5 W. Econ. J. 224–232 (1967); Richard A. Posner, The Social Cost of Monopoly and Regulation, 83 J. Pol. Econ. 807 (1975). 12 Balmoral Cinema v. Allied Artists Pictures, 885 F.2d 313 (6th Cir. 1989). 13 Id. at 316. In Weyerhaeuser Co. Ross-Simmons Hardwood Lumber Co., Inc., 549 U.S. 312 (2007), the Supreme Court considered the use of monopsony power. It avoided this mistake by noting there is likely to be little or no impact on consumer prices. Id. at 324. See discussion in Chapter 2. 14 Id. at 317. This is the same sort of mistake made by earlier courts confronting movie theatre splits. See, e.g., Admiral Theatre Corp. v. Douglas Theatre Co., 427 F. supp. 1268 (D. Neb. 1977), aff ’d on other grounds, 585 F. 2d 877 (8th Cir. 1978). Subsequent courts have gotten what we consider to be the correct answer. See, e.g., General Cinema Corp v. Buena Vista Distrib. Co., 532 F. Supp. 1244 (C.D. Cal. 1982), aff ’d on other grounds, 681 F. 2d 594 (9th Cir. 1982). Also, see the discussion of splits in Chapter 2. 11

46

Economic Theory of Monopsony

monopsonist does pay a lower price for some inputs, it does not pass on these lower costs simply because its relevant costs for decision-making purposes are marginal costs, and these are not lower. At first blush, this seems counterintuitive, but comparing a monopsonist with a competitive buyer illustrates the point. Recall that our textile mill’s optimal employment decision requires that P ∙ MPL = MFCL, where P denotes the price of the monopsonist’s output, MPL represents the marginal product of labor, which is the addition to total output resulting from employing an additional unit of labor, and MFCL is the marginal factor cost of labor, which is the increase in total expenditures on labor resulting from employing an additional unit of labor. In effect, the firm will hire an additional worker as long as the market value of the incremental output that worker generates exceeds the incremental cost of hiring that worker. We also know from standard economic theory that competitive firms produce at the point where output price equals the marginal cost of production. We can see what this means in the current context by rearranging the foregoing condition. Dividing both sides of the equation by MPL, we have P = MFCL/MPL. This means that the firm’s marginal cost must be equal to the right-hand side of this second equality.15 We can compare this result with that of a textile mill that must compete in both the output market and in the labor market. In this case, the firm faces a constant wage rate no matter how much labor it employs. In effect, the MFC is constant and equal to the wage. Thus, its optimal hiring decision requires P ∙ MPL = w where w is the wage rate. Put differently, the firm will hire additional workers as long as the market value of the units of output produced by adding workers, which is price per unit of output times the marginal product of labor, exceeds the wage rate. Again, rearranging this equality by dividing both sides by MPL yields P = w/MPL, which means that the marginal cost for this firm is w/MPL. Now, we have seen from Figure 3.1 that the marginal factor cost exceeds the input price for a monopsonist (i.e., MFCL is larger than w). Consequently, the marginal cost for the monopsonist, MFCL/MPL, is actually higher than the marginal cost for a firm with no monopsony power. In other words, the marginal cost curve of a monopsonist lies above the marginal cost curve of a competitor in the labor market. Since the marginal cost is what drives The logic is fairly obvious: the numerator is the incremental input expenditure necessary to expand employment, and the denominator is the incremental output generated by the expanded employment. The ratio is the incremental cost of output, which is the definition of marginal cost. For a rigorous derivation see C. E. Ferguson, The Neoclassical Theory of Production and Distribution 176 (New York: Cambridge University Press 1971).

15

A Simple Model of Monopsony

47

the firm’s output decision, the monopsonist will actually reduce its output below the level that a seller without monopsony power would select. This output reduction by one textile mill in an isolated area of North Carolina will have no impact on the competitive output market. That mill will sell its output at the market-determined price. Thus, the decrease in the input price is not passed on to consumers.16 Suppose, however, that our textile mill produces a unique product and, as a consequence, enjoys some market power17 in the sale of its output. This means that the monopsonist faces a negatively sloped demand curve in its output market (i.e., the output price is a declining function of the quantity sold). The firm’s profit function is π = P(T) ∙ T – w(L)L – pM

(3.6)

Again, the textile mill must expand its employment of labor until the marginal impact on profit is zero: ∂π = (P(T)+ TdP(T)/dT)∂T/∂L ∂L

w + Ldw(L)/dL) = 0

(3.7)

Rearranging algebraically yields P(T)+ TdP(T)/dT =

w + L(dw(L)/dL) ∂T/∂L

(3.8)

or in more familiar terms MR = MFC/MPL

(3.9)

where MR is marginal revenue, MFC is marginal factor cost, and MPL is the marginal product of labor. As before, the right-hand side represents marginal cost. Since marginal revenue declines as output expands, an increase in marginal cost will result in a decrease in the firm’s profit-maximizing output. Since the demand curve has a negative slope, a decrease in ­quantity Since the actual price paid for the input falls, the firm’s average cost falls. This, however, will not be passed on to consumers in the form of lower output prices. Instead, it will be retained as profit by the firm. If the output market is competitive as we have assumed in the text, the monopsonist will sell its reduced output at the market-determined price. If the monopsonist has any power in the output market, its reduction in output will cause the output price to rise. Thus, the effect of monopsony on average cost similarly has no salutary effect on output price. 17 For an examination of market power in antitrust, consult William Landes & Richard Posner, Market Power in Antitrust Cases. 94 Harv. L. Rev. 937 (March 1981), and Thomas G. Krattenmaker, Robert H. Lande, & Steven C. Salop, Monopoly Power and Market Power in Antitrust Law, 76 Geo. L. J. 241 (1987). 16

48

Economic Theory of Monopsony

leads to an increase in price to consumers. Thus, if a firm with some ­market power in the sale of its product acquires monopsony power in the purchase of its inputs, the prices paid for those inputs fall, but the marginal cost of production rises, the monopsonist’s output falls, and the price to its ­customers actually increases. In sum, monopsony power is to the demand side of a market what monopoly power is to the selling side. Monopoly power is marked by the ability of sellers to raise price above competitive levels, which requires the ability to limit output. Monopsony power involves the power to lower input prices below competitive levels. This requires the ability to restrict demand for the input. In either case, the quantity exchanged is less than the quantity exchanged under competitive conditions and is said to be allocatively inefficient. Ironically, the reduced input prices do not lead to reduced output prices. In fact, when the monopsonist has market power in its output market, the reduced input prices clearly translate into higher output prices. Even when the monopsonist has no market power in the output market, its reduction in output will have some impact on the output price. This effect may be miniscule, but it is there nonetheless.

3.3  Collusive Monopsony 3.3.1  The Economics In some markets, buyers may collude among themselves in an effort to influence the price that they pay for a product or an input.18 Just as collusion among sellers can lead to economic results that are similar to monopoly (i.e., elevated price and reduced output), so can collusion among buyers lead to monopsonistic outcomes. Acting in concert, the buyers would ­orchestrate their purchase decisions to achieve precisely the same results as the ­monopsonist described in the preceding section.19 The ­buying cartel would collude on terms so as to depress the price to w2 in Figure 3.1 by restricting the aggregate purchases to Q2. Consequently, collusive ­monopsony has the same deleterious effects on social welfare as does pure monopsony: Too Oligopsony is a market structure with a few large buyers that cannot fail to recognize their mutual interdependence. The standard oligopoly models of Cournot, Bertrand, Stackelberg, Chamberlin, and others can be adapted to buyers. A similar range of outcomes from the purely competitive to the purely monopsonistic can be found in this array of models. 19 The buyer’s motivation for colluding is, of course, to obtain greater profits. In particular, the buyers convert some of the suppliers’ producer surplus into consumer surplus. This can be seen in Figure 3.1. 18

Collusive Monopsony

49

few resources will be employed. It is clear that the suppliers of the collusively monopsonized product are hurt by a collusive reduction in the purchase of their outputs. Moreover, since the suppliers provide less of their product, the quantity of the final good must be reduced, or its quality must be reduced. Consequently, such restraints are inconsistent with consumer welfare and should be condemned. Two interesting examples are provided by the NCAA and major league baseball. The NCAA is comprised of colleges and universities participating in intercollegiate athletics.20 Member schools present their views on a variety of issues, and collective decisions are reached at periodic meetings of the schools. This has resulted in limitations on the number of scholarships offered (i.e., the quantity) and the compensation that can be offered to the student-athletes (i.e., the price). Thus, collusion among these ostensible rivals under the auspices of the NCAA has reduced the quantity of and compensation for intercollegiate student-athletes on scholarships. In addition, the members limit the number of coaches that can be hired, which reduces the quality of the teams. Similarly, the major league baseball owners can agree legally on the number of players that each team may carry on its active roster. At one point, the roster size was reduced from twenty-five to twenty-four, thereby depressing the number of players employed in major league baseball. Interestingly, the roster size was increased temporarily to twenty-seven following the 1990 lockout, but was returned to twenty-five for the 1991 season. In any event, restrictions on roster sizes limit employment and tend to depress salaries.

3.3.2  Conditions Conducive to Collusive Monopsony There are a number of structural conditions that facilitate collusion among buyers.21 These include the following: 1. Few buyers. The smaller the number of buyers, the lower the costs of decision making within the group. Moreover, the ability to police the agreement is enhanced when the group is small. For a more detailed analysis, see Roger D. Blair & Richard E. Romano, Collusive Monopsony in Theory and Practice: The NCAA, 42 Antitrust Bull. 681 (1997). 21 For an examination of these conditions as they pertain to sellers, see Frederick M. Scherer & David Ross, Industrial Market Structure and Economic Performance, 3d ed. 235 (Boston: Houghton Mifflin Co. 1990); George Hay & Daniel Kelley, An Empirical Survey of Price Fixing Conspiracies, 17 J. L. & Econ. 13 (1974); and George Hay, Oligopoly, Shared Monopoly, and Antitrust Law, 67 Cornell L. Rev. 439 (1982). 20

Economic Theory of Monopsony

50

2. Product homogeneity. When the product is homogeneous, the agreement on price is simplified. For a homogeneous product, there is a single price to fix. In contrast, product heterogeneity requires a complex price schedule that maintains equilibrium price differentials. This can lead to differences of opinion and serve to undermine stability of the buying cartel. 3. Sealed bids. When buyers must submit sealed bids on an item or a collection, everyone knows who bid how much when the bids are opened publicly. As a result, collusion is facilitated simply because no one can cheat on the agreement without being discovered. Consequently, any agreement that is reached is more apt to be stable. As we will see later, bid rigging and buyer rings at auctions have not been uncommon. 4. Inelasticity of supply. When supply is relatively inelastic, the collusive buyers need not restrict their purchases much in order to achieve a significant price reduction. Consequently, the profits that flow from collusive monopsony will be larger, the less elastic the supply at the competitive price. The greater the rewards to successful collusion, the more likely such collusion is to occur.

3.3.3  Organizing and Implementing a Buyer Cartel There are three major tasks that must be performed if a buyer cartel is to be successful:22 1. Agreement 2. Implementation 3. Monitoring Failure to handle any element will undermine the success of the collusive venture. 1. Agreement. Any cartel must reach some mutually satisfactory agreement on a host of issues. First, of course, the members have to decide on the price that they will offer. This necessarily requires a corresponding agreement on how much they will purchase collectively since price cannot be depressed without an appropriate restriction on purchases. But this is not all. To be fully successful in extracting the monopsony profits, the buyers must agree on the quality of the product, the service that the seller is to provide, the credit terms, the delivery terms For a brief discussion of these tasks for colluding sellers, see Kenneth G. Elzinga, New Developments on the Cartel Front, 29 Antitrust Bull. 3 (1984).

22

Collusive Monopsony

51

and so on. All dimensions of the transaction must be agreed on or some of the monopsony profit will be lost. In practice, some collusive arrangements may not resolve all of these issues and thereby only approximate the results of a successful cartel. In this event, the adverse consequences for social welfare are still present. 2. Implementation. Once an agreement has been reached on the terms that the buyers will offer, this agreement must be implemented. Since a restriction on total purchases is necessary, the group must decide on how much each participant will curtail its purchases. If this is not resolved, the buyers will dissipate some of the monopsony profits through nonprice competition or indirect price competition for the limited quantity of the product. 3. Monitoring. As a general proposition, there is no honor among thieves. One must anticipate that cheating will occur since the same greed that led to the original price-fixing conspiracy provides a powerful economic incentive to cheat. As a consequence, the conspirators must try to deter cheating on their collusive arrangement. In theory and practice, such efforts have proved elusive.23 The idea is to make cheating unprofitable, but this is hard to do without imposing significant costs (i.e., lower profits) on the loyal cartel members. If the threatened punishment is severe enough to deter cheating, then it will not actually have to be imposed, and no costs will be borne. For this to work, however, the threat must be credible.

3.3.4  Problems for Collusive Buyers There are two major problems that will plague collusive buyers: (1) entry and (2) cheating on the agreement. These difficulties serve to undermine the success of collusive monopsony. These problems are illustrated with the aid of Figure 3.2, which shows the value of the marginal product curve (VMP) for the good in question for one of the colluding buyers. The competitive price is w1 and the noncollusive employment decision is Q1. Collusive buying leads to a restriction in purchases to Q2 and a drop in the input price to w2. There have been numerous academic attempts to devise ways of deterring cheating. For a few examples, see George J. Stigler, A Theory of Oligopoly, 72 J. Pol. Econ. 44 (1964); Daniel Orr & Paul MacAvoy, Price Strategies to Promote Cartel Stability, 32 Economica 186 (1965); D. K. Osborne, Cartel Problems, 66 Am. Econ. Rev. 835 (1976); Edward Green & Robert Porter, Noncooperative Collusion Under Imperfect Price Information, 52 Econometrica 87 (1984); Lester Telser, A Theory of Self-Enforcing Agreements, 53 J. Bus. 27 (1980); and Ian Ayres, How Cartels Punish: A Structural Theory of Self-Enforcing Collusion, 87 Colum. L. Rev. 295 (1987).

23

Economic Theory of Monopsony

52 Price

VMP2 A

w1

w2

0

C

B

D

VMP Q2

Q1

Q3

Quantity

Figure 3.2.  Value of Marginal Product.

For this firm, the cost of restricting its purchases is given by the triangular area A, which represents the forgone profit that the buyer could have earned. The benefits of collusion flow from the reduced expenditures on the input, which equal (w1 – w2)Q2 and are represented by area B. The net gain is B – A, which must be positive or collusion would not occur. Entry. The supra-competitive profits shown in Figure 3.2 provide an incentive for outsiders to enter. As entry occurs, the price of the input will be bid up above w2, which will undermine the profitability of collusion. It is not obvious how the cartel could prevent outsiders from entering in response to these profit opportunities. This is not to say that the threat of entry is high. Presumably, the colluding sugar mills in Mandeville Island Farms were enjoying monopsony profits, which would attract entry. But entering as a buyer of sugar beets requires building a sugar mill and entering the sugar business, which would not be done on a whim. Cheating. In addition to potential entrants, there are enemies within the cartel: cheaters. In Figure 3.2, for example, the value of the input’s marginal product (VMP2) is well above the input price (w2). If all of the other firms hold the price down to w2, this firm has an incentive to expand its purchases from Q2 to Q3. By cheating on its agreement, this firm can increase

Measuring Buying Power

53

its profits from B – A to B + C + D for a net gain of A + C + D. Since each member of the cartel will be confronted with the same temptation, it is not unlikely that the agreement will be undermined through clandestine defections.

3.4  Measuring Buying Power There is little doubt that a showing of market power has assumed increasing importance in antitrust cases over time. In the context of assessing monopsony power, the issue we are interested in is buying power, which is the power to reduce price below the competitive levels by restricting purchases. The consideration of market power on the buying side is not novel in antitrust analysis. In its Merger Guidelines, for example, the Department of Justice points out that “[t]he unifying theme of the guidelines is that mergers should not be permitted to create or enhance ‘market power’ or to facilitate its exercise.”24 After expressing the usual concern for the market power of sellers, the guidelines recognize that “[t]he exercise of market power by buyers has wealth transfer and resource misallocation effects analogous to those associated with the exercise of market power by sellers.”25 The guidelines do not provide any guidance on how to measure buying power, but the Antitrust Division of the Department of Justice looks to market share as the prime indicator of buying power.26 In this section, a buying power index is developed that captures the competitive significance of monopsony power. It extended it to the case of a dominant buyer and employed it to evaluate the Antitrust Division’s position that a firm with less than a 35 percent market share is unlikely to possess significant market power. The exercise illustrates that focusing solely on market share as a measure of monopsony power can result in incorrect conclusions. This is not to say, however, that the buying power index can always be readily calculated. Nonetheless, it can be quite useful as a framework for assessing monopsony power. U.S. Department of Justice Merger Guidelines, 49 Fed. Reg. 26,823 (1984). Id. 26 For example, in a business review letter from Charles F. Rule, Assistant Attorney General, dated November 18, 1987, the Department’s focus is clearly on market share: “In general, if the total projected volume of shipments by the members of a group purchasing transportation service is less than 35 percent of available transportation capacity, it is unlikely that it will be profitable for the group’s members to restrict usage of transportation services in an effort to drive down rates.” 24 25

54

Economic Theory of Monopsony

3.4.1  The Measurement of Buying Power In a classic article, Abba Lerner examined the consequences of ­monopoly.27 He recognized that the monopolist will select a price that yields the maximum profit. From the consumer’s perspective, this monopoly profit constitutes a tax on the consumer that the monopolist can collect simply because of the monopolist’s restrictive powers.28 Lerner also pointed out that the popular hostility toward monopoly stems from the consumer’s resentment toward the monopolist’s tax. He went on to suggest that the degree of monopoly in an industry should measure “the divergence of the system from the social optimum that is reached in perfect competition.”29 Here, an analogous concept is used for measuring the degree of monopsony power. This measure is presented first for the case of pure monopsony and subsequently for a dominant buyer. In either case, the buying power index applies to collusive variants of the alternative market structures.

3.4.2  The Buying Power Index: Pure Monopsony As already illustrated, when a buyer has no market power, it expands its purchases until the value of the marginal product equals the price of the good or service in question (i.e., until VMP = w). In contrast, a monopsonist will stop short of that and purchase the quantity at which the value of the marginal product equals the marginal factor cost: VMP = MFC. Returning to Figure 3.1, it can be seen that the monopsonist restricts its employment below the competitive outcome: Q2 as opposed to Q1. The measure of buying power should capture this deviation. One way to do this is to follow Lerner and define a buying power index as the percentage deviation from the competitive result: BPI =

VMP − w w

(3.10)

Abba Lerner, The Concept of Monopoly and the Measurement of Monopoly Power, 1 Rev. Econ. Stud. 157 (1934). 28 Id. Lerner also discussed the social welfare losses associated with monopoly. It is apparent from this article, however, that the concept of consumer surplus was still quite controversial, and Lerner did not pursue this. Paul Samuelson, Foundations of Economic Analysis 195 (Cambridge, MA:  Harvard University Press 1947), dismissed consumer surplus as well. Sir John Hicks, A Revision of Demand Theory (Oxford:  Oxford University Press 1956), however, rehabilitated the concept, and it has become widely used as a welfare measure. 29 Lerner, supra note 27 at 157. 27

Measuring Buying Power

55

Since profit maximization will lead the monopsonist to equate the value of the marginal product to the marginal factor cost, we can substitute and have BPI =

MFC − w w

(3.11)

Note, however, that MFC = w + L dw(L)/dL and, therefore, BPI =

w + Ldw(L)/dL −w w

BPI =

Ldw(L)/dL w

(3.12) (3.13)

which is the reciprocal of the elasticity of supply. Thus, the buying power index can be written as BPI = 1�

(3.14)

where ε is the elasticity of supply, which measures the responsiveness of the quantity supplied to changes in price. As ε rises, the firm’s monopsony power declines. Less technically, to the extent that sellers can alter the quantity offered for sale in response to price changes, monopsony (buying) power declines.

3.4.3  The Buying Power Index: The Dominant Buyer30 There is no doubt that there are few, if any, pure monopsonists in the real world. This, however, does not mean that a large buyer cannot impose monopsonistic welfare losses. Suppose, for example, that a series of mergers created one large buyer that recognized that its purchase decisions would influence the price of some input. The firm still would not be a pure The dominant firm model was developed to examine the pricing behavior of a large seller with a fringe of small competitors. See Karl Forchheimer, Theoretisches zum unvollstandigen Monopole, Schmollers Jahrbuch 1 (1908), and George J. Stigler, Notes on the Theory of Duopoly, 48 J. Pol. Econ. 521 (1940). The analysis of the dominant seller has found its way into many standard microeconomics textbooks; see e.g., John Gould & Edward Lazear, Microeconomic Theory, 6th ed. 403 (Homewood, IL:  Richard D. Irwin 1989). William Fellner, Competition Among the Few 136 (New York: Alfred A. Knopf 1949) presented an oligopolistic variant of this model in which several large sellers act in concert. What follows is our adaptation of this model to the buying side.

30

Economic Theory of Monopsony

56 Price

MFC

S Sr w* Ddf

Df O

Q*f

Q*df

Q*

Quantity

Figure 3.3.  Dominant Firm and Competitive Fringe.

monopsonist, however, since there remains a fringe of competitive buyers that will accept the price that the dominant firm pays as the market determined price. The fringe firms will buy this input up to the point where their demand equals this price. The dominant buyer’s problem is to adjust its purchases to maximize profit subject to the competitive behavior of the other buyers. This is shown in Figure 3.3 where Df represents the demand for the input in question by the competitive fringe, Ddf represents the demand for the dominant firm, and S is the supply curve. The dominant firm recognizes that at any price it selects, the competitive fringe will purchase as any competitive firm would, namely, where Df equals the price. The dominant buyer incorporates this behavior into its profit function by subtracting Df from S to obtain the residual supply, which is denoted by Sr in Figure 3.3. Then the curve marginal to Sr, which is labeled mfc, represents the marginal cost to the dominant buyer of expanding its purchases (i.e., its marginal factor cost). The balance of the analysis is familiar: The dominant buyer purchases Q*df where mfc equals Ddf, which determines price equal to w* from the residual supply. At a price of w*, sellers will provide Q*, which is equal to the sum of Q*f and Q*df. As we can see, the marginal factor cost (mfc) exceeds the price of the input (w).

Measuring Buying Power

57

The buying power index for a dominant firm facing a fringe of competitive buyers can also be derived.31 Figure 3.3 reveals that the quantity supplied to the dominant buyer (QDB) equals the market supply (QM) minus the quantity demanded by the competitive fringe: QDB = QM – QF

(3.15)

It is easy to derive the elasticity of supply for the dominant buyer (εDB). To this end, differentiate both sides with respect to w: dQDB dQM dQF = − dw dw dw

(3.16)

Multiplying both sides by w/QDB provides dQDB . w dQM . w dQF . w = − dw QDB dw QDB dw QDB

(3.17)

Further algebraic multiplication yields dQDB . w dQM . w . QM dQF . w . QF = − dw QDB dw QM QDB dw QF QDB

(3.18)

Substituting the usual definitions for demand and supply elasticities, we have

� DB = � QM + � QF QDB

QDB

(3.19)

where ε is the market supply elasticity, η is the elasticity of demand for the fringe firms, and QDB, QF, and QM represent the quantities of the dominant buyer, the fringe, and the market, respectively. If we define the dominant buyer’s market share as s = QDB/QM, then we can express the elasticity of supply for the dominant buyer as �DB = � + �(1−s) s s

(3.20)

It follows by substitution that the buying power index in the dominant buyer case is

This derivation provides the buying power analog to that on the selling side. For the latter, see Thomas R. Saving, Concentration Rations and the Degree of Monopoly, 11 Int’l Econ. Rev. 139 (1970), or William M. Landes & Richard A. Posner, Market Power in Antitrust Cases, 94 Harv. L. Rev. 937 (1981).

31

58

Economic Theory of Monopsony

BPI =

s 1 1 = = �DB �/s +� (1−s)/s � + � (1−s)

(3.21)

Thus, the BPI for the dominant buyer depends on the overall elasticity of supply (ε) and the elasticity of demand (η) for the fringe as well as the dominant buyer’s market share (s). Comparative Statics For purposes of antitrust analysis, it is useful to determine how each variable influences the measure of buying power. First, we shall derive the result analytically and then show some illustrative calculations. Market Share (s). There is no doubt that the market share of the large buyer is an important determinant of buying power. In fact, we may observe that, holding all else constant, buying power increases as the dominant buyer’s share of total purchases increases. This is determined analytically by evaluating the sign of the partial derivative of BPI with respect to s. In this case, it is easy to show that ∂BPI/∂s > 0. In addition, we can get a feel for the importance of market share by using the BPI. Suppose, for example, that the elasticity of supply (ε) and the elasticity of fringe demand (η) are each equal to 1. Then, as market share increases, the percentage deviation from the competitive result increases. This is shown in Table 3.1. These results are consistent with expectations and do not need further elaboration. In addition, the results displayed in Table 3.1 are entirely consistent with the Department of Justice’s instincts regarding the importance of market share. Note, however, that the BPI is positive for all values of s; not just for those values in excess of 35 percent. Elasticity of Supply (ε). The elasticity of supply measures the responsiveness of the quantity supplied to the changes in the price. Ceteris paribus, as the elasticity of supply increases, the BPI falls. Analytically, this is determined by examining the sign of the partial derivative of BPI with respect to ε. In this instance, one can readily see that ∂BPI/∂ε < 0. The effect on the BPI of changing ε can be seen in Table 3.2. We assume that the elasticity of fringe demand (η) equals 1 and that the dominant buyer’s share (s) equals 60 percent. Then, as the elasticity of supply increases, the percentage deviation from the competitive result falls. The sensitivity of the BPI to changes in the elasticity of supply is apparent. As the quantity supplied becomes more responsive to changes in price, the large buyer’s power falls. This is because the suppliers can redirect their efforts to other products where prices may be higher. In the limit, the

Measuring Buying Power

59

Table 3.1.  Alternative values of BPI for ε = 1 and η = 1 s

10

20

30

40

50

60

70

80

90

100

BPI

5

11

18

25

33

43

54

67

82

100

Table 3.2.  Alternative values of BPI for s = 60 percent and η = 1 ε

0

0.5

1.0

2.0

3.0

∞

BPI

150

67.0

43.0

25.0

18.0

0

Table 3.3.  Alternative values of BPI for s = 60 percent and ε = 1 η

0

0.5

1.0

2.0

3.0

∞

BPI

60

50.0

43.0

33.0

27.0

0.0

e­ lasticity of supply goes to infinity (i.e., the supply curve is flat and, therefore, perfectly elastic), and the BPI goes to zero.32 Elasticity of Fringe Demand (η). Finally, we may examine the influence of demand elasticity of the fringe buyers. As this elasticity increases, the buying power of the dominant firm falls. This can be seen by examining the sign of the derivative of BPI with respect to η. It is easy to show that ∂BPI/∂η < 0. Some illustrative calculations will make this result more tangible. Table 3.3 presents alternative calculations for BPI under the assumptions that s equals 60 percent and ε equals 1. As these calculations show, the BPI declines with increases in the elasticity of fringe demand. This follows because any reduction in price implemented by the dominant buyer’s curtailed purchases is tempered by the enhanced purchases of the fringe. The more responsive they are to price decreases, the more difficult it is for the dominant buyer to make such a decrease stick. In the limit, the elasticity of fringe demand goes to infinity, and the dominant buyer’s power goes to zero. Jonathan Jacobson & Gary Dorman, Joint Purchasing, Monopsony, and Antitrust, 36 Antitrust Bull. 1 (1991), made this point when they expressed their skepticism regarding the empirical importance of monopsony problems.

32

Economic Theory of Monopsony

60

Table 3.4.  Alternative values of BPI for s = 35 percent η

0.5

1.0

2.0

3.0

0.0

108

54

27

18

0.5

42

30

19

14

1.0

26

21

15

12

2.0

15

13

11

9

3.0

11

10

8

7

ε

3.4.4  The BPI and the Department of Justice Threshold As mentioned earlier, the Antitrust Division uses a 35 percent market share as a critical value for s.33 But an examination of the BPI formula suggests that relying solely on market share may be unwise. Table 3.4 shows the percentage deviations from competitive prices associated with a market share of 35 percent and several different values for ε and η. The significance of a 35 percent market share varies considerably. When ε = 3 and η = 3, a firm or a cooperative buying venture with a 35 percent market share will cause only a 7 percent deviation. In contrast, when ε = 0.5 and η = 0.5, a market share of 35 percent leads to a 42 percent deviation. The rest of the entries have similar interpretations. The results of these calculations are consistent with the theoretical predictions on ε and η, namely, that the greater these elasticities are, the smaller the BPI is. As a final illustration of the danger of relying too heavily on market shares without considering the other determinants of buying power, we examine two situations. First, consider a cooperative buying venture involving optometrists who collectively purchase 25 percent of the optometric supplies in question. Next, consider a group of hospitals that purchase 75 percent of the medical supplies in question. Based on the Antitrust Division’s rule of thumb, the optometrist group would pass muster while the hospital group would be challenged. Suppose, however, that ε = 0.5 and η = 0.2 for the optometrists while ε = 2.0 and η = 1.0 for the hospital group. The BPI for the optometrists then would be 38.5 percent, but only 33.3 percent for Areeda and Turner propose a 25 percent threshold as part of their merger rules. As with the Department of Justice threshold of 35 percent, reliance on market share can be risky, and errors may result. See Phillip Areeda & Donald Turner, 4 Antitrust Law 200–6 (Boston, MA: Little Brown 1980).

33

Market Definition and the BPI

61

the hospitals.34 Thus, the hospital group with triple the market share of the optometrists would actually have less market power.

3.5  Market Definition and the BPI Defining the relevant market is one of the most important aspects of antitrust cases dealing with monopolization and mergers. Ceteris paribus, the greater a firm’s market share, the greater will be its market power, and how the market is defined may greatly influence the market shares of the firm or firms involved. A broad market definition will tend to result in relatively low market shares, whereas a narrow definition will result in comparatively high shares. Consequently, the outcome of many antitrust cases can hinge crucially on the market definition adopted by the court. As one might expect, plaintiffs typically argue for a narrow market definition, whereas defendants typically argue for a broad definition. The court’s problem, then, is to decide which (if either) of the proposed market definitions is correct. Defining the relevant market seems simple enough:  One merely identifies a group of sellers and a corresponding group of buyers whose supply and demand decisions determine the equilibrium price.35 This exercise requires the identification of both geographic and product boundaries that separate buyers and sellers whose actions influence price from buyers and sellers whose actions do not influence price. In practice, however, this is far easier said than done. This is not the place to attempt a resolution of this difficult problem. Instead, the focus is on the change in perspective that is required when buying power is the issue. In addition, we explain how the BPI compensates for the inherent inexactness of market definitions. Market Definition:  the Seller’s Perspective. Again, market definition requires delineating the relevant product and geographic markets. When examining the market power of a seller, one adopts the buyer’s view and asks what goods are reasonably substitutable for the defendant’s product.36 0.25 or BPI = 0.3846 or 38.5 percent. For the 0.50 + 0.20 (0.75) 0.75 = 0.33 or 33.3 percent. hospital group, BPI = 2.0 + 1.0 (0.25)

34

For the optometrists, BPI =

See, e.g., George J. Stigler & Robert A. Sherwin, The Extent of the Market, 28 J. L. & Econ. 555 (1985), who define the market as “… that set of suppliers and demanders whose trading establishes the price of a good.” 36 The leading Supreme Court case is United States v. E. I. du Pont de Nemours & Co., 351 U.S. 377 (1956). 35

62

Economic Theory of Monopsony

In the product market case, one asks what products satisfy the same need. As the number of substitutes increases, the defendant’s market share and its market power decrease. Having thus defined the product market, the question then becomes one of identifying the locations of those buyers who can reasonably compete with the defendant.37 For example, one might examine the ability of distant sellers to ship goods into the defendant’s sales areas. Probably, the most important factor affecting the ability of distant sellers to compete locally is the cost of transporting the product. If transshipment readily occurs, sellers in other locations are fairly regarded as being within the defendant’s geographic market.38 When the defendant is a buyer, the focus necessarily shifts. First, one must adopt the seller’s point of view. The “product” market is composed of competing uses for the seller’s output. In effect, the defendant buyer’s market share and power are diluted as the number of possible uses increases. This result makes sense because the availability of reasonably substitutable uses decreases the likelihood that any one buyer can force prices down and maintain them below a competitive market level. Sellers will simply turn to alternative buyers. The different perspective in terms of product market definition is seen more clearly in All Care Nursing Service, Inc. v. Bethesda Memorial Hospital,39 where suppliers of temporary nursing services complained that hospitals were colluding to drive down the price of temporary nursing services. The defendants alleged that the price restraints were ancillary to the purpose of increasing the quality of care. If the rule of reason were applied, the initial step would be to determine the proper market definition. Ordinarily, one would think in terms of services that are reasonably interchangeable for those of nurses. In the monopsony context, however, the focus would be quite different. The question would be whether there are uses for nursing services beyond those required by hospitals. Thus, one might consider whether the agencies view nursing homes and private home care as reasonably interchangeable outlets for their product. Geographic market definition also requires a search for outlets that could be regarded as reasonable substitutes. The relevant question is whether sellers can ship to users outside the local geographic area when prices are depressed. If so, those distant buyers should be regarded as part of the relevant geographic market. Again, the distant buyers provide a way for sellers See generally Kenneth G. Elzinga & Thomas Hogarty, The Problem of Geographic Market Delineation in Antimerger Suits, 18 Antitrust Bull. 45 (1973). 38 Id. See also Saving or Landes & Posner, supra note 31. 39 135 F. 3d 740 (11th Cir. 1981). 37

Market Definition and the BPI

63

to avoid taking the lower price offered by the local buyer. Unlike the typical monopoly market definition case, where the “import” transportation costs of distant suppliers are likely to be the most important factor in determining whether their sales should be included in the market being defined, the focus in defining the market in the case of monopsony must be on the “export” transportation costs of the local sellers. Obviously, transportation costs will vary by destination (and, of course, by product). Those distant locations where prices are high enough to compensate buyers for the transportation costs should be included in the market. A possible method of bypassing the step of examining transportation costs would be simply to observe whether local sellers are making sales in distant markets. Distant areas in which sales are currently made are usually part of the geographic market. This does not mean that sales could not be made in locations where sales are currently not made if the local price were depressed. Thus, a definition restricted to locations in which sales actually are made may understate the size of the geographic market. The proper focus was applied in United States v. Rice Growers Association of California,40 where a merger of rice millers was challenged. Opposition to the merger was passed on the possibility of increased buying power. In effect, the merged millers would be able to lower the price paid to growers of rice. The market affected was identified by the Department of Justice as the “purchase or other acquisition for milling of paddy rice grown in California.”41 The court correctly analyzed the alternatives available to the sellers in response to depressed prices. Specifically, in defining the geographic market, the court noted that shipment to out-of-state purchasers was not feasible.42 Self-Correction and the BPI. If one relies solely on market share as an indication of market power, the market delineation must be precise. Particular care is necessary in examining the elasticity of demand. If buyers respond quickly to the efforts of another buyer to depress prices by increasing purchases when prices fall, they must be included as part of the defendant’s market; otherwise, the market share calculation will not reflect market power. A comforting aspect of the BPI is that it does not require absolute precision in delineating the market in order to be accurate.43 For example, suppose the defendant buyer is regarded as competing with a few local buyers with 42 43 40 41

1986 WL 12562 (E.D. Cal. 1986). Id. at 4. Id. at 7–8. This analysis is based on an analogous discussion found in Landes & Posner, supra note 31, at 962–3.

64

Economic Theory of Monopsony

the result being a 60 percent market share. In addition, suppose there are a number of distant buyers who encourage transshipment when prices are depressed. To the extent that these buyers should have been, but were not, included in the defendant’s market share calculation, the market share of 60 percent will overstate the defendant’s market power. On the other hand, since the distant purchasers have been left out of the market, their impact will be reflected in the calculation of elasticity of demand. A higher elasticity of demand means lower buying power, and this would be an offsetting factor in the calculation of the BPI.

3.6  The Buying Power Index in Practice Usually, measuring the BPI is no trivial matter. In fact, obtaining a range of estimates for the BPI is a far more realistic prospect than calculating a single number. But the index suggests a way of thinking about buying power that is easy to understand and useful in application. Moreover, the methodology is already available and in use in the context of more traditional examinations of market power. Of course, the issues in the context of monopoly power are to what extent market share may overstate or understate the ability of sellers to raise prices and what other factors are relevant determinants of that ability. In the case of traditional monopoly power, the factors that “qualify” the inference about market power flowing from market share are (1) the industry elasticity of demand and (2) the supply elasticity.44 In short, the ability to raise and maintain prices above competitive levels decreases – for any given market share – to the extent that buyers can shift to products in different industries or to the extent that more output can be made available by rivals in the defendant’s market. In the case of the BPI, the ability of buyers to depress prices – for any given market share – decreases to the extent sellers are free to vary their output and new customers become available.45 Recently, more enlightened examinations of monopoly power have employed the methodology of qualifying the inference of market power based on market share alone. A brief look at a few cases indicates that that process is hardly difficult and easily could be adapted to examinations of See Landes & Posner, supra note 31. The ability of sellers to vary output in the monopsony context is analogous to the ability of consumers to shift their demand in the monopoly context. Similarly, the capacity of sellers to find new customers in the monopsony context is analogous to the ability of buyers to turn to new sources of supply in the monopoly context.

44 45

The Buying Power Index in Practice

65

buying power. For example, in United States v. Waste Management,46 the Court of Appeals for the Fifth Circuit reviewed a lower court opinion holding that a merger of garbage collection companies resulting in a combined market share of nearly 50 percent was prima facie illegal. The court reasoned that a 50 percent market share would not be sufficiently high to prohibit the merger if it were misleading as to the actual competitive impact. The court then qualified the market share by noting that entry into the business was relatively easy and that the new firm would not be free to raise prices. In reversing the trial court opinion, the Fifth Circuit essentially relied on the influence of supply elasticity. In another merger case, Hospital Corporation of America v. FTC,47 the Seventh Circuit Court of Appeals reviewed a finding of the Federal Trade Commission that a merger of hospitals in the Chattanooga, Tennessee, area violated Section 7 of the Clayton Act. The merger increased the acquiring firm’s market share from 14 to 26 percent and increased the market share of the four largest firms from 79 to 91 percent.48 Judge Richard Posner, certainly one of the most ardent supporters of the view that reliance on market share alone may lead to incorrect results, wrote the opinion for the court affirming the Commission. He specifically cited the low probability that patients in Chattanooga would readily seek hospital services elsewhere (demand elasticity) and the difficulty of new hospital entry (supply elasticity) as supporting the inference that firms in the area would be able to raise and maintain prices above competitive levels. Broadway Delivery Corporation v. United Parcel Service49 is another interesting example primarily because it suggests that low market shares may actually understate market power. There the trial court had instructed the jury that a market share of less than 50 percent was insufficient to establish market power.50 The Court of Appeals for the Second Circuit held that this instruction was erroneous, indicating that the jury might reasonably decide otherwise in light of information about “demand and supply elasticities.”51 48 49 50 51 46 47

743 F.2d 976 (2nd Cir. 1984). 807 F.2d 1381 (7th Cir. 1986), cert. denied, 481 U.S. 1038 (1987). 807 F.2d at 1384. 651 F.2d 122 (2nd Cir. 1981), cert. denied, 454 U.S. 908 (1981). 651 F.2d at 127. Id. at 128. The Second Circuit cautioned that a trial court might have to assist the jury in assessing the significance of these factors. Interestingly, the Second Circuit did not hold that the erroneous instruction required reversal as the plaintiff ’s evidence was insufficient to allow a finding that the defendant possessed market power.

66

Economic Theory of Monopsony

Turning to the BPI, one would want to assess each of the components: market share, the elasticity of supply, and the elasticity of fringe demand. Once one settles on the market definitions – both product and geographic – the calculation of market share is still not trivial because one must determine the total industry sales. Ordinarily, this information is not readily available and must be gathered through the discovery process.52 This may be a rather cumbersome chore, but it is surely manageable. The second element in the BPI is supply elasticity. Ideally, to measure the supply elasticity, one would want an econometric estimate of the supply function.53 From that mathematical expression, one could calculate the elasticity of supply at any point along the supply curve. There are, however, formidable problems in estimating supply functions. As a result, the precise measure of supply elasticity is apt to be elusive. Nonetheless, one can obtain some qualitative sense of how responsive the quantity supplied is to changes in price simply by asking suppliers how they would react to a 5–10 percent permanent change in price.54 Important in this regard is the ability of the suppliers to switch production facilities from one product to another. The easier and cheaper it is to convert capacity to other uses, the more elastic is supply. In monopoly cases, an important consideration is the ease of entry,55 but in monopsony cases we are concerned about exit. The easier exit is, the more elastic the supply response will be. Important in this regard is the speed of exit – how long does it take to convert a plant to new uses. In order to move into another industry, other considerations are important: whether production technology is available or subject to patent In some industries that are subject to special regulation or taxation – for example, alcoholic beverages, gasoline, cigarettes, and the like – industry totals and market shares are relatively easy to obtain. In most industries, however, the antitrust authorities will have to obtain sales data from each firm in the industry and do its own calculations. 53 Econometric methods are examined in detail by Jan Kmenta, Elements of Econometrics (New York:  Macmillan 1971). A useful introduction is provided by Franklin Fisher, Multiple Regression in Legal Proceedings, 80 Colum. L. Rev. 702 (1980). Also see Phillip Areeda, Herbert Hovenkamp, Roger D. Blair, & Christine Piette Durrance, Antitrust Law ¶394 (New York: Aspen Publishers 2007). An interesting application is contained in Jeffrey L. Harrison, The Lost Profits Measure of Damages in Price Enhancement Cases, 64 Minn. L. Rev. 751 (1980). The econometric estimation proceeds from economic principles, which identify the important determinants of supply, to the specification of a functional form and to the application of statistical techniques to economic data. The result is a mathematical expression relating the quantity supplied to the various determinants of supply. 54 The hypothetical 5–10 percent price change is how the Department of Justice and the Federal Trade Commission have interpreted the “small but significant” language in the Merger Guidelines. 55 A useful discussion of ease of entry is provided by Steven Salop, Measuring the Ease of Entry, 31 Antitrust Bull. 551 (1986). 52

Concluding Remarks

67

control, whether the converted facility will be sunk,56 whether brand preferences must be established,57 and whether the firm must go into the new industry at a relatively large scale.58 Putting all of this information together, one can form an opinion regarding the elasticity of supply in a qualitative sense. Finally, the third element in the BPI is the elasticity of fringe demand. Again, ideal measures of this demand elasticity are unlikely to be available. Nonetheless, one can form some judgment about this elasticity by asking the fringe buyers how they would respond to a 5–10 percent change in price. As price falls, for example, one wants to know whether the fringe buyers will increase their purchases significantly. If so, their elasticity of demand is large and, correspondingly, the BPI will be low. If, however, they cannot easily substitute the good in question for other goods, a 5–10 percent decrease in price may not lead to significant increases in purchases. Their elasticity of demand will be low, which will, in turn, indicate that the BPI is high. Thus, one can obtain a qualitative feel for the elasticity of fringe demand. Forming an opinion about the existence of substantial market power is relatively easy if all elements of the BPI point in the same direction. When market share is large, supply is relatively inelastic, and fringe demand is also relatively inelastic, then one can be quite confident that the BPI is large. Similarly, when market share is small and the elasticities are fairly large, the BPI is bound to be small. In other instances, however, the signals will conflict. For example, a large share may be accompanied by a qualitative assessment that the supply elasticity is relatively large as is the elasticity of fringe demand. In that case, a categorical opinion is more problematic, but the analyst’s experience may permit an informed judgment.

3.7  Concluding Remarks There is no doubt that pure monopsony is quite rare in the U.S. economy. But where monopsony power exists because of a dominant buyer or collusion among buyers, it causes social welfare losses analogous to those occasioned by monopoly. In particular, too few resources will be employed. The economic analysis in this chapter provides the foundation for the antitrust analysis that follows. Id. at 562. Id. at 560. 58 Id. at 560–1. 56 57

FOUR

The Antitrust Response to Monopsony and Collusive Monopsony

4.1 Introduction This chapter is about the antitrust response to monopsony and collusive monopsony. In some instances, it is possible to rely on actual examples of litigation involving monopsony power. In other scenarios involving monopsony and collusive monopsony conduct, however, judicial experience may be limited or nonexistent. In those areas, the judicial response can only be inferred from the judicial treatment of analogous behavior by firms on the selling side of the market. In both cases  – actual or inferred judicial response – an evaluation is presented based on economic principles. By way of preview, the actual judicial responses to monopsony generally make economic sense. In the opinions that are available, this may seem surprising since many of these opinions followed bright-line rules developed before economic principles were expressly applied to antitrust analysis. In this chapter, a modern economic foundation for these cases is offered as is a framework for analyzing monopsony practices that are yet to be addressed. In the process, we demonstrate that abuses by monopsonists and collusive monopsonists have not always taken the form suggested by the traditional models examined in Chapter 3. Regardless of the section of the antitrust laws applied, an important step in the analysis is an assessment of market power or, in the case of monopsony, buying power. The first section examines the few instances in which the question has been addressed by the courts. The sections following that one are divided into the basic areas of possible antitrust violations by sellers. First, there is a limited discussion of “monopsonization,” the buyer-side analog of monopolization, that would presumably fall within the ambit of Section 2 of the Sherman Act. It focuses on the pricing practices of monopsony. Second, the more general abuses of monopsony power are discussed 68

Judicial Assessment of Market Power

69

in connection with collusive monopsony because much of the prohibited conduct would also be prohibited if engaged in by a group of firms.

4.2  Judicial Assessment of Market Power In Chapter 3, the theory of monopsony power was discussed in detail. In practice, the precise issue has only rarely presented itself. This is not a consequence of limited numbers of monopsony cases. More likely, the assessment of market power is rare because monopsony cases have generally arisen in the context of per se offenses and, thus, a market analysis is unnecessary. Even in reported cases in which market power is addressed, the analysis is not as detailed as that suggested in Chapter 3. The first step in assessing market power is the proper definition of the market. In Campfield v. State Farm Mutual Automobile Insurance Co.,1 the plaintiff supplied auto glass repair services. Typically cracks longer than six inches in length require that the glass be replaced. The technology that the plaintiff favored involved using a chemical patching technique as opposed to glass replacement for cracks longer than six inches. State Farm, which processed 1.7 million insurance claims per year related to glass damage, would not reimburse policy holders for the process employed by the plaintiff. The actual handling of glass repair claims was outsourced to another firm. The plaintiff made antitrust claims under Sections 1 and 2 of the Sherman Act alleging that he was, in essence, the victim of a boycott. At the outset, the court noted that it was necessary to define the relevant market. In this instance, the plaintiff claimed that the relevant market was “State Farm insured repairable windshield repair,” and that State Farm had used its power to force prices down in that market.2 The court noted that the claim was actually that State Farm possessed monopsony power and found that the plaintiff ’s proposed market definition was underinclusive. The key was to define the market in terms of all potential buyers of plaintiff ’s repair services, not simply those insured by State Farm. Market definition was also critical in Todd. v. Exxon Corp.3 Exxon and several other oil companies were alleged to have artificially lowered compensation levels for certain types of employees. Plaintiffs defined the relevant market as “the services of experienced, salaried, non-union, managerial, professional and technical employees in the oil and petroleum industry in 2008 WL 2736656 (10th Cir. 2008). Id. 3 275 F.3d 191 (2d Cir. 2001). 1 2

70

The Antitrust Response to Monopsony

the continental United States and various submarkets thereof.”4 Defendants possessed 80–90 percent of the market so defined. The lower court dismissed the complaint relying on a selling side analysis. It reasoned that managerial, professional, and technical employees were not interchangeable and could not be included in the same “product” market.5 The appellate court recognized the case as involving monopsony, requiring a different approach to market definition. It correctly noted that the interchangeability of those attempting to sell their services was not ­relevant. Instead, the question was whether the defendant buyers were reasonably interchangeable with other buyers. Thus, the question was whether the types of employees listed in the market definition could be regarded as selling services in the same market. The plaintiffs argued that they possessed distinctive oil-industry-specific qualifications that made the relevant market oil companies and not other firms in other industries that also hired managerial, professional, and technical workers. The appellate court described the market definition as not “implausible” and observed that the plaintiffs would have to show that a drop in the relative salaries of managerial, professional, and technical workers in the oil industry would not result in their departure from that industry. At least as far as the proper approach to buyer-side market definition, the appellate court’s analysis was the correct one.

4.3  Monopsony Pricing The application of monopsony power leads a profit-maximizing firm to restrict the quantity employed and thereby to reduce social welfare.6 This raises the question of whether an economically sensible antitrust remedy exists for monopsony pricing. The simple answer is “probably not.” Consider the case of a “natural” monopsony, where productive efficiency requires that there be a single buyer of an input.7 If this buyer makes its purchase decisions Id. at 199. This seems strange since the idea of “cluster” markets has been employed in other contexts. See, e.g., United States v. Philadelphia National Bank & Trust Co., 399 U.S, 350 (1970). Also, see Ian Ayres, Rationalizing Antitrust Cluster Markets, 95 Yale L. J. 109 (1985). 6 See Figure 3.1 and the accompanying text. 7 This is analogous to the case of natural monopoly, where productive efficiency requires that a single firm produce all of the output. Familiar examples include local telephone service, cable televisioin, and electric power generation and distribution. See Sanford Berg & John Tschirhart, Natural Monopoly Regulation (New York: Cambridge University Press 1988), for an extended examination of the theory of natural monopoly and society’s response. 4 5

Monopsony Pricing

71

according to the model developed in Chapter 3, the only way that the antitrust laws could be used to eliminate the welfare losses would be to require a restructuring of the demand side of the market. For example, a single buyer could be dismantled into small units to eliminate its buying power. This, however, is not economically efficient and, therefore, restructuring would impose welfare losses of uncertain magnitude.8 Accordingly, based on conventional economic reasoning, one cannot advocate such a solution. This is analogous to the situation of a natural monopoly that emerged due to economies of scale.9 No structural relief is desirable because productive inefficiency would result. Just as Section 2 of the Sherman Act does not forbid the structural condition of monopoly, it has not been applied to forbid the structural condition of monopsony. It is textbook antitrust law that Section 2 of the Sherman Act responds only to abuses of power.10 The question then arises: In the case of monopsony, should it be an abuse of market power to use that power simply to influence the price paid? Both practical and theoretical reasons suggest that monopsony pricing alone should not fall within the prohibitions of Section 2. From a practical standpoint, such a response would require a court to actually engage in price regulation.11 As a theoretical matter, it is the lure of supra-competitive profits, which result from the exploitation of market power, that leads to a dilution of monopsony power as new firms enter the market. As a matter of legal interpretation, it has been clear for some time that merely charging a monopoly price is not an abuse of monopoly ­power.12 Now, the same reasoning has been applied to the use of monopsony power as a means of lowering price. For example, in Kartell v. Blue Shield of Massachusetts,13 Blue Shield was characterized as a buyer of physician   8 A single employer may have monopsony power in a local input market, but sell its output in a competitive national market. Imposing any inefficiency on such a firm will raise its costs and render it uncompetitive, which could lead to bankruptcy.   9 “Natural monopoly generally refers to a property of productive technology, often in conjunction with market demand, such that a single firm is able to serve the market at less cost than two or more firms.” D. Spulber, Regulation and Markets 3 (The MIT Press 1989). See also Berg & Tschirhart, supra note 7. 10 See E. Thomas Sullivan & Jeffrey L. Harrison, Understanding Antitrust and Its Economic Implications, 5th ed. 277 (New York:  LexisNexis 2009); Berkey Photo, Inc. v. Eastman Kodak Co., 603 F.2d 263, 297 (2d Cir. 1979), cert. denied, 444 U.S. 1093 (1980). 11 See U.S. v. Trenton Potteries Co., 273 U.S. 392, 397–8 (1927). 12 See sources in note 10, supra. 13 749 F.2d 922 (1st Cir. 1984), cert denied, 471 U.S. 1029 (1985). See also, Ball Mem’l Hosp., Inc. v. Mut. Hosp. Ins., Inc., 784 F.2d 1327 (7th Cir. 1986). Some courts, however, have suggested that monopsony pricing could violate the antitrust laws. See Med. Arts Pharmacy of Stanford, Inc. v. Blue Cross & Blue Shield of Conn., Inc., 675 F.2d 502 (2nd Cir. 1982)

72

The Antitrust Response to Monopsony

services. A group of physicians challenged the use of Blue Shield’s market power as a buyer on the grounds that Blue Shield could set rates below levels that would be set by a “freely competitive market.” Judge (now Supreme Court Justice) Breyer, writing for the First Circuit Court of Appeals, noted the courts’ general reluctance to monitor price in the case of monopoly and reasoned that the use of market power by a buyer to obtain lower prices was similarly not a violation of the antitrust laws.14 Although there is evidently no antitrust response to a single firm with monopsony power using that power to force prices down, there may be a response to efforts to raise prices. This may seem to be an unlikely strategy for a buyer to adopt, but it is the buyer’s counterpart to a predatory pricing strategy by a seller.15 Focusing first on sellers, the label “predatory pricing” is below costs and, therefore, irrational unless the result is to exclude competitors. When competitors exit the industry, the remaining firm will raise its prices and recoup the losses incurred during the period of predation. Whether a firm would actually engage in predatory pricing is a matter of considerable doubt. For such a policy to be effective, the firm would sell not only its original output at below-cost prices but would have to be prepared to meet the total demand in the market. Furthermore, the quantity demanded may exceed current levels as buyers substitute away from other products in favor of the below-cost offer. The expansion in demand may mean that the predatory firm will find its plant and capacity stretched to levels where its average cost of production increases, further expanding the losses incurred.16 (per curium); Custom Auto Body, Inc. v. Aetna Casualty & Surety Co., 1983 WL 1873 (D.R.I. 1983). 14 See also Kamine/Besicorp Allegany P.A. v. Rochester Gas and Elec. Corp., 908 F.Supp. 1194 (W.D.N.Y. 1995). 15 There is an expansive literature on predatory pricing by sellers. See, e.g., Phillip Areeda & Donald Turner, Predatory Pricing and Related Practices Under Section 2 of the Sherman Act, 88 Harv. L. Rev. 687 (1975); F. Easterbrook, Predatory Strategies and Counterstrategies, 48 U. Chi. L. Rev. 263 (1981); Paul Joskow & Alan Klevorick, A Framework for Analyzing Predatory Pricing Policy, 89 Yale L. J. 213 (1979); Oliver Williamson, Predatory Pricing: A Strategic and Welfare Analysis, 87 Yale L. J. 204 (1977). For explanation of just what is meant by “predatory pricing,” see Joseph Brodley & George Hay, Predatory Pricing: Competing Theories and the Evolution of Legal Standards, 66 Cornell L. Rev. 738 (1981), which summarizes and synthesizes the flood of commentary engendered by Areeda and Turner. See also Wesley Liebeler, Whither Predatory Pricing? From Areeda and Turner to Matsushita, 61 Notre Dame L. Rev. 1052 (1986). 16 See Robert Bork, The Antitrust Paradox 107 (New York:  Basic Books 1978). John S. McGee, Predatory Pricing: The Standard Oil (N.J.) Case, 1 J. L. & Econ. 137 (1958), and John S. McGee, Predatory Pricing Revisited, 23 J. L. & Econ. 289 (1980), argues that predation is unprofitable and thus of little economic relevance. Kenneth Elzinga, Predatory Pricing: The Case of the Gunpowder Trust, 13 J. L. & Econ. 223 (1970), and Roland Koller,

Monopsony Pricing

73

All of these losses can be regarded as an investment that can only be profitable if competitors abandon the market and the monopolist can raise its price above competitive levels and thereby earn monopoly profit. These future profits must be discounted to present value, further decreasing the likelihood of a profitable investment.17 Finally, and perhaps most importantly, either market characteristics or actions of the monopolist must retard entry into the market while the monopolist is earning an economic profit. This unlikely scenario has led to considerable debate about whether the threat of predatory pricing is a realistic antitrust concern. The Supreme Court has observed that “there is ample evidence to suggest the practice does occur.”18 On the other hand, it has also noted that “predatory pricing schemes are rarely tried, and even more rarely successful.”19 In Brook Group Ltd. v. Brown & Williamson Tobacco Co.,20 a 1993 case, the Supreme Court announced the standard to be applied in predatory pricing claims.21 According to the Court, a claim could only be supported if the plaintiff demonstrated a period of below-cost pricings22 plus a dangerous probability that the losses incurred could be recouped.23 From the point of view of a monopsonist, predatory conduct would amount to setting prices so high for an input that it would make sense only if it thereby denied other buyers access to the input.24 By analogy to the case of monopoly, the likely standard that would signal a predatory effort by a buyer would be offering a price for an input greater than its marginal revenue product. For example, in Figure 4.1, the profit-maximizing



17

20 21 18 19



22



23 24

The Limitations of Local Price-Cutting as a Barrier to Entry, 64 J. Pol. Econ. 329 (1956). A cost-benefit approach to testing the rationality of predation along with some empirical evidence has been offered by Charles McCall, Predatory Pricing: An Economic and Legal Analysis, 1 Antitrust Bull. 32 (1987). Kenneth Elzinga, Collusive Predation: Matsuchita v. Zenith, in J. Kwoka & L. White, The Antitrust Revolution (Glenview, IL: Scott, Foresman 1989), and Kenneth Elzinga & David Mills, Testing for Predation: Is Recoupment Feasible? 34 Antitrust Bull. 869 (1989), look at predation as an investment. The present value calculations demonstrate the economic implausibility of predatory pricing. Cargill, Inc. v. Monfort of Colo., Inc., 107 S.Ct. 484, 495 (1986). Matsushita Elec. v. Zenith Radio Corp., 106 S.Ct. 1348, 1357–8 (1986). 509 U.S. 209 (1993). The Court addressed predatory pricing as both a form of attempt to monopolize and a primary line of price discrimination. The Court did not announce a precise cost standard, which leaves considerable ambiguity and invites confusion in the lower courts. Id. at 224. For an extensive analysis of predatory buying by a would-be monopsonist, see Roger B. Blair & John E. Lopatka, Predatory Buying and the Antitrust Law, Utah L. Rev. 415 (2008).

The Antitrust Response to Monopsony

74 Price

MFC S

P4 PC P2 P1

MRP O

Q1 Q2

QC

Q4

Quantity

Figure 4.1.  Monopsony Pricing and Predation.

monopsonist would ordinarily pay price P1. Increases in competition could result in a price of P2 or, in the extreme, a competitive price of PC. Above price PC, however, the monopsonist would be paying in excess of its marginal revenue product. This is critical because marginal revenue product represents the additional revenue generated by employing one more unit of the input. Paying a price greater than marginal revenue product therefore makes no sense unless it leads to a future payoff. In the context of monopsony, the payoff could come in two forms. First, when competing buyers leave the market, the monopsonist will lower price below the competitive level. In addition, to the extent that the monopsonist had competed in the output market with some of the same firms, it stands to gain in the output market from their departure. Indeed, the use of monopsony power in order to gain market power as a seller is a common phenomenon in antitrust.25 Successful monopsony predation is probably as unlikely as successful monopoly predation. First, the predatory firm would have to raise the price not only for the inputs it originally purchased, but be prepared to purchase all of the input currently available in the market at the higher price. Thus, in Figure 4.1, if the price offered were P4, the quantity purchased in the See United States v. Griffith, 334 U.S. 100 (1948); Klor’s v. Broadway-Hale Stores, Inc., 359 U.S. 207 (1959).

25

Monopsony Pricing

75

predatory effort would be Q4, not Q1. In addition, other input suppliers that could switch into the production of this input would have an economic incentive to do so. Taking the added supply off market would increase the predator’s financial burden. Second, the predator has the problem of what to do with the excessive input. One possibility is to destroy it; another is to store it for future use. Any effort to process the input into its final output would likely increase the quantity of that output available and depress its price. All of these options create further financial burdens. Finally, once the predator begins to take advantage of its monopsony power by depressing the price of the input, either market conditions or actions by the monopsonist must retard the reappearance of competing buyers. In effect, the buyer must be able to profit from the lower price for a long enough period of time that it can recoup its predatory investment. All of these hurdles mean that the prospects for the predatory buyer are probably as unlikely as they are for the predatory seller. Although the notion of predatory pricing by buyers may seem somewhat fanciful, it is a theory that has been asserted on more than one occasion. In each instance, however, the courts have not focused on pricing greater than the marginal revenue product as the standard for buyer predation. Instead, the standard has been either the buyers’ marginal cost or the production cost of the “victim.” For example, in American Tobacco Co. v. United States,26 the major producers of cigarettes were charged with conspiring to monopolize the sale of tobacco products. One practice was to outbid smaller competitors who were making inroads into the market by purchasing relatively cheap tobacco and offering “ten-cent brands.” One of the claims on which the Supreme Court seemed to base its finding that Sections 1 and 2 of the Sherman Act had been violated was that the input costs to the smaller producers were being driven up to the point that they were unable to underprice the more expensive brands.27 Similarly, in Reid Brothers Logging Co. v. Ketchikan Pulp Co.,28 the claim was against firms operating pulp plants and sawmills engaged in a number of activities designed to reduce competition between them as buyers. One element of the claim was that they paid prices in excess of those paid by competing buyers as a means of denying those buyers access to logs. The 328 U.S. 781 (1946). Because the case involved a conspiracy, there was no need to address the specific issue of predatory buying. As a result, the opinion is not particularly instructive on that point. 27 Id. at 804. 28 699 F.2d 1292 (9th Cir. 1983), cert. denied, 464 U.S. 916 (1983). 26

76

The Antitrust Response to Monopsony

district court finding that this was “predatory bidding” was challenged by the defendants, who argued that there had been no showing that “the high prices paid for standing timber would prevent the defendants from covering their marginal costs on the ultimate sale of the processed lumber.”29 The Ninth Circuit rejected defendants’ reasoning and noted a simple cost-based test of predation had been rejected by that court in William Inqlis v. ITT Continental Baking.30 A different result is found in the beef industry litigation which spanned three decades.31 In this case, the meat packers were charged with paying artificially high prices for fed cattle as a way of maintaining monopsony power in the “fed cattle procurement market.” The theory failed for a number of reasons; the primary one being the lack of proof that the defendants had offered a predatory price. In this instance, the court defined as predatory “a price higher than that which would allow the packer to make a profit.”32 It was not until 2007 that the U.S. Supreme Court examined predatory bidding in considerable detail. Unlike prior cases, the question arose after the Court’s opinion in Brooke Group and after the development of costbased standards for predation. This made for a much more focused analysis than had previously been the case. In Weyerhaeuser Co. v. Ross-Simmons Hardwood Lumber Co.,33 plaintiff claimed that Weyerhaeuser had bid up the prices of alder sawlogs as part of a plan to drive it, a smaller producer, from the industry. The plaintiff noted that the price of logs had increased while Weyerhaeuser’s profits declined. The Court was faced with the issue of whether the standard announced in Brooke Group in the context of predatory pricing by a seller should be applied to the claims of predatory pricing by a buyer. The Ninth Circuit Court of Appeals had held that the same standard did not apply. That court’s reasoning was that seller predation does, for some period of time, benefit buyers. In the case of predatory buying, the benefits enjoyed by the sellers during the predatory period were not comparable. The Supreme Court rejected the Ninth Circuit’s reasoning and applied the Brooke Group standards. The Court correctly described the parallels between predation by sellers and by buyers. In the case of buyers, the Court recognized that recoupment could occur if the defendant were able 31 32 33 29 30

Id. at 1298. 389 F.Supp. 1334 (D.C.Cal. 1975). In re Beff Industry Litigation, 907 F.2d 510 (9th Cir. 1990). Id. at 515. 549 U.S. 312 (2007).

Monopsony Pricing

77

to reduce prices in input markets and maintain them at below-competitive levels long enough to recoup losses incurred during the period of predation.34 Consequently, it held that predation by buyers would be assessed under a two-step standard similar to that announced in Brooke Group. The outcome was supported in part by the Court’s view that predatory bidding was unlikely to occur and be successful. The Court also noted that predatory bidding and the expanded purchases necessitated by offering supracompetitive prices may actually result in higher output.35 Unfortunately, the Court did not fully take into account the difference between the buying and selling side of the market and announced a standard that is underinclusive.36 When identifying predatory bidding, according to the Court, “[a] plaintiff must prove that the alleged predatory bidding led to below-cost pricing of the predator’s outputs. That is, the predator’s bidding on the buying side must have caused the cost of the relevant output to rise above the revenues generated in the sale of those outputs.”37 In effect, while the Court noted that monopsony was the mirror image of monopoly, it rejected a mirror image standard for the determination of predation. To understand the problem with this standard, it is important to note why predation is important at all. Predatory pricing and bidding are regarded as evidence that a firm is attempting to monopolize a market by unfair means. The term predation is reserved for instances in which a firm makes itself economically worse off or acts in an economically irrational fashion and deliberately suffers losses. The only reasonable explanation is that it will more than offset those losses after competitors fail. In the case of predatory bidding, that kind of conduct, as demonstrated previously, occurs when a firm pays more for an input than that input’s marginal revenue product. A firm that pays in excess of the marginal revenue product for an input, however, may be able to sell its output at above-cost prices depending on its power in other input markets. For example, the firm may buy multiple inputs and pay competitive prices for some of them. This may reduce average costs enough to offset the losses in the market in which it is engaged in predatory conduct and the ultimate output price is above cost. Similarly, The case at hand did not involve an instance in which the defendant created a risk of monopolizing the output market and, thus, could recoup in output markets. 35 Id. at 377. This assumes that the inputs are not subsequently destroyed. See Blair & Lopatka, supra note 24. 36 One author observes that the Court “created a safe harbor” for the predatory buyer. See John B. Kirkwood, Controlling Above-Cost Predation: An Alternative to Weyerhaeuser and Brooke Group, 53 Antitrust Bull. 369 (2008). 37 Weyerhaeuser Co., 549 U.S. at 325. 34

78

The Antitrust Response to Monopsony

if the firm has market power in the output market, it may be able to sell at prices above both marginal and average cost in spite of engaging in predation in one input market. What this means is that evidence that output is sold at above-cost prices is not a reliable indicator that predatory bidding is not taking place, nor does it eliminate the possibility that the harms of predatory bidding will follow.

4.4 Abuses of Monopsony Power The practices assessed in this section fall into two groups: those by a monopsonist that would violate Section 2 of the Sherman Act and those by colluding buyers that would violate Section 1 of the Sherman Act. As in cases of monopoly power, a monopsonist or a group of firms that collectively possess monopsony power may use that buying power to influence a term of the exchange other than price. Thus, we consider both price and nonprice effects.

4.4.1  Price Effects 4.4.1.1  Classical Model As noted in the previous section, unilateral efforts by a monopsony to force prices downward traditionally do not violate the antitrust laws. When collusion is involved, however, Section 1 of the Sherman Act has been brought to bear. For example, in the leading collusive monopsony case, Mandeville Island Farms v. American Crystal Sugar Co.,38 sugar refiners in northern California agreed to adopt a pricing formula that resulted in uniform price offers for sugar beets. The effect was to reduce the average price paid to sugar beet growers to a level below that which would have prevailed in the absence of the agreement. In finding the agreement unlawful, the Supreme Court cited the general per se illegality of price-fixing agreements and announced: “It is clear that the agreement is the sort of combination condemned by the Act, even though the price-fixing was by purchasers and the persons specially injured under the treble damage claim are sellers, not customers or consumers.”39 Mandeville Island Farms seems to be a clear example of a collusive monopsony practice that has the economic results predicted by the traditional theory. It is important to note, however, that if the case were reconsidered 334 U.S. 219 (1948). Id. at 235.

38 39

Abuses of Monopsony Power

79

today and the effect of the collusive effort examined, a court might erroneously conclude that since the effort lowered prices, the eventual outcome could be beneficial to consumers. This impact analysis, which is wrong on economic grounds, could result in a substantive finding of no violation, or a finding that the plaintiff-sellers had suffered no antitrust injury. This problem is addressed in Chapter 7. Although price fixing by buyers is as likely to be per se unlawful as that by sellers, in both instances courts are reluctant to classify all price-affecting conduct as “per se price fixing.” This distinction between “literal” price fixing and price fixing that would be subject to the per se standard was first announced in Broadcast Music, Inc. v. Columbia Broadcasting System.40 The distinction arose in the monopsony context in U.S. v. Brown University.41 Ivy League colleges had agreed on the financial assistance to be offered to commonly accepted students.42 In effect, if a student were accepted by more than one school, the schools would not compete by offering different levels of assistance. Literal price fixing was involved. The Court of Appeals for the Third Circuit held, however, that the rule of reason should be applied. It noted that MIT (the lone defendant after the other conspirators had entered into a consent decree) was a qualified charitable organization and was not guided by the desire to maximize profits. This, it reasoned, meant the anticompetitive effects could be different than in the conventional price-fixing context. In short, even in buying-side cases, the characterization of conduct will play a role in determining the antitrust standard to be applied.43 4.4.1.2  Inelastic Supply and Perishable Commodities Similar in effect to that predicted by the traditional model is the response of a collusive monopsony to supply that is inelastic in the short run. In these instances, if there were a shortage, the free interaction of supply and demand would result in an increase in price. One of the earliest collusive monopsony cases involved the fixed supply model. In National Macaroni Manufacturers Association v. Federal Trade Commission,44 the Seventh Circuit Court of Appeals found objectionable a buying cartel of macaroni manufacturers that attempted to control the price paid for one of its primary inputs. High-quality macaroni requires the use of 100 percent durum 42 43

441 U.S. 1 (1979). 5 F.3d 658 (3d Cir. 1993). Id. For a similar discussion see North Jackson Pharmacy, Inc. v. Caremart RX, Inc., 385 F.Supp. 2d 740 (N.D. Ill. 2005). 44 345 F.2d 421 (7th Cir. 1965). 40 41

The Antitrust Response to Monopsony

80 Price

Supply After Crop Failure

Supply

P2 P1

Demand Collusive Demand 0

Q2

Q1

Tons of Durum Wheat

Figure 4.2.  Collusive Monopsony and Shortages.

wheat, which is easier for the manufacturers to work with and yields a macaroni product that has the most desirable cooking properties. In the absence of any supply disruptions, the intersection of demand and supply, which occurs at a price of P1 and a quantity of Q1 in Figure 4.2, determines the free-market equilibrium price and quantity of durum wheat. The original supply curve is drawn as a vertical line at quantity Q1 to reflect the fact that during a growing season, the maximum quantity supplied cannot respond to changes in price. Everyone’s expectations were upset by major crop damage, which curtailed the supply of durum wheat. For the growing season in question, the maximum supply fell to Q2 as shown in Figure 4.2. In the normal course of events, competition among buyers for the reduced quantity of durum wheat would have driven the price up to P2. But this was not to be. The macaroni manufacturers agreed among themselves to alter their macaroni recipe.45 Instead of using 100 percent durum wheat, they agreed to use a blend of 50 percent durum wheat and 50 percent farina. This, of course, reduced the quality of the final product. For the producers, however, the agreement artificially depressed the demand for durum wheat to the level shown as the “collusive demand” and thereby depressed its price to the original level. For a discussion of horizontal agreements designed to standardize products, see Section 5.2.

45

Abuses of Monopsony Power

81

The Seventh Circuit found that this was a per se violation of Section 5 of the Federal Trade Commission Act.46 When farmers have bumper crops, the supply may be so great that prices fall and the resulting profits are low or nonexistent. Antitrust policy would not permit the farmers to collude in order to prop up their prices and profits in the presence of bumper crops. The Macaroni decision provides symmetry: When shortages boost prices and perhaps profits, buyers may not collude to reduce those prices. 4.4.1.2.1  The Role of Perishability In Macaroni, the assumption was that the sellers had a perfectly inelastic supply curve, which is probably not quite accurate since the farmers may have had the option of withholding some output from the market in hopes of higher prices in future years. The assumption of inelastic supply is entirely appropriate, however, for the case of perishable commodities. Perishability makes the current supply perfectly inelastic because there is no cost-saving way to restrict output once the crop has been harvested. Two examples of this case are the protracted litigation involving beef producers and the continuous controversy surrounding the free agency rights of professional athletes.47 In In re Beef Industry Antitrust Litigation, the plaintiffs alleged that grocery store chains were colluding to depress the price paid for beef.48 The plaintiffs noted that the supply of beef was fixed in the short run because a fattened steer must be sold within three weeks of the time it becomes “choice grade.” This left the plaintiffs completely unable to withhold output in response to a collusively determined price; supply was inelastic. In other words, the supply curve was vertical at quantity Q2 in Figure 4.2. Once again, the alleged collusion was designed to move the demand curve to a lower level and thereby decrease the price paid. A similar, but actually more severe, problem besets professional athletes and other workers. Labor is an extremely perishable commodity – an hour not worked today can never be recovered. A professional athlete has alternative occupations, but high-quality professional athletes are so scarce that Id. at 427. Perishability also seems to have played a role in American Tobacco Co. v. U.S., 328 U.S. 781. (1946). 47 Professional athletes are subject to a form of bid rigging. See Section B.2, infra. 48 In re Beef Industry Antitrust Litigation, 600 F.2d 1148 (5th Cir. 1979) (reversing and remanding unfilled opinion), cert. denied, 449 U.S. 905 (1980), 542 F. Supp. 1122 (N.D. Texas 1983), aff ’d, 710 F.2d 216 (5th Cir. 1983), cert. denied, 465 U.S. 1052 (1984); 713 F. Supp. 971 (N.D. Tex. 1989), aff ’d, 907 F.2d 510 (5th Cir. 1990). The issue in much of the Beef litigation was whether the producers of beef, who were indirectly affected by the alleged bid rigging, had antitrust standing. 46

82

The Antitrust Response to Monopsony

their wages as athletes generally are far greater than their next most lucrative endeavor. Accordingly, the supply of labor effort for each individual athlete will be quite inelastic. Collusion among employers can drive the wage paid down to the individual’s reservation wage. This is the purpose of the team owners’ agreement to prohibit free agency (i.e., the ability to bargain with any and all teams). As a result, each team will be a pure monopsonist in dealing with its players. 4.4.1.2.2  Welfare Consequences of Collusion in the Context of Inelastic Supply There is nothing to recommend collusive monopsony. The efforts of collusive monopsonists to obtain lower prices do not translate into lower prices for consumers, but only into higher profits for themselves. Beyond that, there are economic consequences in the context of inelastic supply, in particular, that require specific attention. Since supply is perfectly inelastic, the collusion among buyers will have no impact on quantity in the short run. Thus, in these circumstances, collusion among buyers to depress demand has only distributive significance in the short run.49 In the durum wheat example, since the quantity was fixed at Q2 after the crop failure, the effect of collusion was to redistribute wealth:  Suppliers will receive less for their wheat. Instead of receiving P2Q2 dollars for the crop, farmers will receive only P1Q2. The difference, (P2  – P1) Q2, was redistributed to the colluders. Ordinarily, the economic objection to collusion does not involve distributive issues, and in the short run, this is true here as well. But collusion creates expectations that have long-run consequences. As producer profits are reduced by the collusion, their incentives to plant durum wheat are reduced, and they may reduce supply in the future. Reductions in supply have adverse consequences for consumer welfare in the future. There is little doubt that the Macaroni decision was simply a result of Socony-Vacuum’s blanket prohibition on price fixing. Macaroni is sensible in a modern antitrust analysis because it recognizes that, even in cases where supply is fixed and reductions in quantity are not possible in the short run, consumers are hurt because the producer’s profits are reduced and their response may be to reduce supply in the future. Consequently, under certain conditions, such restraints are inconsistent with consumer welfare in the long run.

Mountain States Tel. & Tel. Co. v. FCC, 939 F.2d 1035, 1043 (C. Cir. 1991).

49

Abuses of Monopsony Power

83

4.4.1.2.3  All-or-None Supply Cases According to the classical model of monopsony, the monopsonist must reduce the quantity that it purchases in order to depress price. It would be more desirable from the monopsonist’s perspective to pay the lower price without reducing the quantity purchased. Ordinarily, this is not possible unless the monopsonist can push the sellers onto their all-or-none supply curve.50 Since this concept may be somewhat unfamiliar, a digression is in order. The standard supply curve reveals the answer to the following question: What is the quantity that suppliers will provide at a given price? By answering this question for various prices, one obtains the price–quantity combinations that form the usual supply curve. The choice confronting the seller is how much to supply at a particular price. The all-or-none supply curve, however, is a different matter. It reveals the answer to the question of what the maximum quantity is that suppliers will make available at each price when the alternative is to sell nothing at all.51 By framing the question in this fashion, the buyer seeks to extract the entire producer surplus. Accordingly, the all-or-none supply curve lies below the standard supply curve.52 This enables the monopsonist to exploit its power by pushing the sellers off their traditional supply curves and onto their all-or-none supply curves. Consider the demand that a group of firms have for a particular input – for example, the venerable widget, as shown in Figure 4.3. The usual supply curve of widgets is also shown. It is clear that the interaction of this supply and demand determines an equilibrium price and quantity in a competitive widget market of P1 and Q1. Now, suppose that the buyers get together and exert their collective buying power over the diverse group of suppliers. One strategy would be to restrict purchases below Q1 and thereby depress price below P1. Alternatively, however, the buyers could impose all-or-nothing decisions on the unorganized group of suppliers. In fact, if this threat of buying nothing is credible, they can push the suppliers off the traditional supply curve and onto the all-or-none supply curve at the quantity Q1, which is the privately optimal quantity for the colluding buyers. The price actually paid falls from P1 to P2 without any reduction in the quantity transacted. See Milton Friedman, Price Theory 16 (Chicago: Aldine Publishing Co. 1976) Richard G. Layard & A. A. Walters, Microeconomic Theory 224 (New York:  McGrawHill 1978):  “Lying below the supply curve is the seller’s all-or-nothing price, showing the minimum price per unit at which he is willing to sell each quantity.” For an interesting application of the all-or-none model, see Jill B. Herndon, Health Insurer Monopsony Power: All-Or-None Model, 21 J. Health Econ., 197 (2002). 52 See Friedman, supra note 50, at 118. 50 51

The Antitrust Response to Monopsony

84 Price

Supply All-or-None Supply

A B

P1

F

P2

E Demand

C

0

Q1

Quantity of Widgets

Figure 4.3.  All or None Supply.

The short-run consequences of the all-or-nothing scenario are purely distributional, which is not to say they are unimportant.53 The entire producer surplus is transferred to the collusive buyers. In Figure 4.3, the noncollusive consumer surplus is the triangular area ABP1 and the producer surplus is CBPI. After imposing all-or-none conditions on the suppliers, the collusive monopsonists increase their consumer surplus by the rectangular area P1BEP2. This comes at the expense of producers whose producer surplus has been reduced by the same area. Note that the area above the supply curve and below P2 – that is, area CFP2 – is equal to area EFB.54 Thus, the buyers have extracted the entire producer surplus through their collusive efforts. This type of collusion does not reduce output in the short run. Thus, consumers will be no worse off. The long-run impact on output and consumers may be different, however, as suppliers may exit the market if the price offered is below average total cost. The all-or-none model seems to fit recent cases involving challenges to the monopsonistic pricing practices of health care insurers.55 In these See Herbert Hovenkamp, Distributive Justice and the Antitrust Laws, 51 Geo. Wash. L. Rev. 1 (1982); and Robert Lande, Wealth Transfers as the Original and Primary Concern of Antitrust: The Efficiency Interpretation Challenged, 34 Hastings L. J. 65 (1982). 54 This can be proven with plane geometry. Triangles CFP and EFB are similar right triangles. Since the bases FP2 and FE are of equal length, the triangles are congruent and, therefore, the areas are the same. 55 See, e.g., Kartell v. Blue Shield of Mass., 749 F.2d 922 (1st Cir. 1984), cert. denied, 471 U.S. 1029 (1985); Med. Arts Pharmacy of Stamford, Inc. v. Blue Shield of Conn., Inc., 675 F.2d 53

Abuses of Monopsony Power

85

cases, the providers typically object to the maximum price the insurer has offered. It seems likely that the insurers prefer to have little or no impact on the quantity of medical services available. The long-run consequences are, however, difficult to predict. For example, in Kartell v. Blue Shield of Massachusetts,56 a group of physicians challenged the pricing policies of Blue Shield, which offered reimbursement on a take-it-or-leave-it basis. Plaintiffs contended that the rates were set so low that it would discourage entry into the physician services market. The argument, if accurate as an empirical matter, would support the possibility of long-run misallocative effects of monopsony pricing. In some sense, this long-run effect may seem counterintuitive because it is unlikely that the interests of Blue Shield, or any other monopsonist, are served by pricing so low that there is a long-run exit of sellers. In principle, this exit would alter the supply and thereby alter the all-or-none price and quantity. Thus, the practice can be employed to deny competitors in output markets access to critical inputs.

4.4.2  Nonprice Abuses The purest collusive monopsony cases, like Mandeville Island Farms and Macaroni, involve efforts to control demand and lower prices. Just as in monopoly cases, there are instances in which the “abuse” of power is not designed to lower prices directly but to exact some other form of advantage for the monopsonist or collusive monopsony. Typically, these buyer-side cases take the form of common arrangements among sellers. 4.4.2.1  Horizontal Market Division One of the relatively firm per se rules applied to sellers prohibits horizontal market division.57 For example, sellers may agree to not compete in certain 502 (2d Cir. 1982); Travelers Ins. Co. v. Blue Cross of Conn., Inc., 675 F.2d 505 (3rd Cir. 1973), cert. denied, 414 U.S. 1093 (1973); Pa. Dentist Ass’n v. Med. Serv. Ass’n of Pa., 577 F.Supp. 457 (1983). The all-or-none possibility was also addressed in Sony Elec., Inc. v. Soundview Tech., Inc., 157 F.Supp.2d 180 (D. Conn. 2001) and White Mule Co. v. ATC Leasing Co., LLC, 540 F.Supp. 2d 869 (N.D. Ohio 2008). The all-or-nothing model is also likely to be consistent with agricultural markets. See Chapter 8. 56 749 F.2d at 923–4. Another example of an all-or-none situation appears in All Care Nursing Services V. Bethesda Memorial Hospital, Inc., 887 F.2d 1535 (11th Cir. 1989). Areeda and Kaplow point out that challenges brought under Section 1 of the Sherman Act usually fail because an insurer is a single buyer and can unilaterally shop around for favorable terms. See Phillip Areeda & Louis Kaplow, Antitrust Analysis 251, n. 27 (Boston: Little, Brown 1988). 57 United States v. Topco Assoc., 405 U.S. 596 (1972). The continued strength of this per se rule is in doubt. See Polk Bros., Inc. v. Forest City Enter., Inc., 776 F.2d 185 (7th Cir. 1986).

86

The Antitrust Response to Monopsony

territories, for the same customers, or sell the same products. On the buying side of the market, the comparable agreement would be one in which buyers agree to divide suppliers up and not compete in their purchases from suppliers. Interestingly, the landmark case that is viewed as firmly establishing that price fixing by competing sellers is per se unlawful involved a horizontal market division by buyers. In United States v. Socony-Vacuum Oil Co.,58 the defendants were the ostensibly competing vertically integrated major petroleum refiners. These firms sold large quantities of gasoline to jobbers (wholesalers) at prices that were contractually tied to prices in the spot market. At the time, independent refiners were dumping a considerable quantity of gasoline into the spot market, which depressed the spot market price. This, in turn, had a depressing effect on the prices that the jobbers paid. The majors hit on a plan to prop up the spot market price. Each major refiner was assigned one or more independent so-called “dancing partners.” The majors would purchase gasoline from their dancing partners and thereby prevent that supply from reaching the spot market.59 The obvious ultimate goal was the stabilization of the prices received by the major refiners. The actual anticompetitive action of the defendants was a horizontal market division by buyers. A horizontal market division by buyers that exactly parallels that by sellers is found in Reid Brothers Logging Co. v. Ketchikan Pulp Co.,60 in which buyers of logs in the Tongass National Forest in southeast Alaska apparently divided the Forest into “spheres of influence” and bid only on those timber sales in their assigned territory. In the sixteen-year period 1959–5, the parties bid against each other a total of three times out of a possible 143 sales by the U.S. Forest Service. This discipline evidently held even throughout a period of timber shortage when defections might have been expected.61 4.4.2.2  Bid Rigging Bid rigging is an especially popular form of collusive monopsony, especially in circumstances of inelastic supply. The most visible version of bid rigging is found in the market for professional athletes.62 In addition, bid 60 61

310 U.S. 150 (1940). Id. at 181. 699 F.2d 1292 (9th Cir. 1983), cert. denied, 464 U.S. 916 (1983). Id. at 1296. More recently, blueberry growers in Maine alleged that buyers were depressing prices by allocating certain growers to specific buyers. In effect, each buyer only purchased from a predetermined group of growers. Pease v. Jasper Wyman & Son, 2002 WL 1974081 (Me. Super. 2002). 62 A well-known case is Flood v. Kuhn, 407 U.S. 258 (1972). 58 59

Abuses of Monopsony Power

87

rigging has occurred among antique dealers in the purchase of antiques at estate auctions.63 In addition, bid rigging has occurred at auctions of used commercial equipment.64 In all cases, the ostensible rival buyers agree in advance not to bid against each other. In essence, the sellers are allocated to designated winning buyers. Given the inelastic supply of athletes, antiques, or used commercial equipment, the winning “bidder” can offer a low price at or just above the reserve price with the assurance that the seller will be relatively unable to respond. In the case of athletes, the allocation of players to employers takes place through a “draft” process. In the more typical public auction, the bid riggers conduct a private auction among themselves. The difference between the winning bid in the private auction and the rigged bid in the public auction on each antique purchased provides a pool of collusive profits. These profits are shared among the collusive monopsonists in proportions that cannot be determined on an a priori basis.65 The result of such bid rigging is to deny to the owners of the antiques the full market value of their goods. Split agreements among supposedly competing movie theaters provide another example of bid rigging.66 Competitive bidding among exhibitors (i.e., the movie theaters) for soon-to-be released films could result in large cash guarantees, high license fees, and other unfavorable terms. One way to avoid such an outcome is for the theaters to agree among themselves to refrain from competing with one another (i.e., to “split” the films among themselves). The intended result, of course, is to obtain the same number of films, but on more favorable terms from the exhibitors’ perspective. A further example of bid rigging is United States v. Portac, Inc., which involved a timber auction.67 The U.S. Forest Service decided to auction off timber on forest service land. Astoria Plywood Corporation “won” the purchase Charles F. Rule reported the Antitrust Division of the Department of Justice had “a number of investigations focusing on auction pools in a variety of commercial areas.” See Antique Dealers Face Price Fixing Charges, 53 Antitrust Trade Reg. Rep. 117 (1987). 64 United States v. Perfection Mach. Sales, Inc., Crim. No. 88–00281 (E.D. Pa. 1988); United States v. Perfection Mach. Sales, Inc., Crim. No. 88–0391 RFP (N.D. Cal. 1988). 65 Any profit-sharing rule could be used as long as it does not lead to so much dissatisfaction among the colluders that the buying cartel breaks down. There is no unique profit-sharing rule to guarantee cartel stability. Frederick M. Scherer & D. Ross, Industrial Market Structure and Economic Performance, 3d ed. 248 (Boston: Houghton Mifflin 1990). 66 For examples, see Movie 1 & 2 v. United Artists Commc’n, Inc., 909 F.2d 1245 (9th Cir. 1990); Balmoral Cinema v. Allied Artists Pictures, 885 F.2d 313 (6th Cir. 1989); Southway Theatres v. Ga. Theatre Co., 672 F.2d 485 (5th Cir. 1982); United States v. Capitol Serv. Inc., 568 F. Supp. 134 (E.D. Wis. 1983), aff ’d., 756 F.2d 502 (7th Cir. 1985), cert. denied, 474 U.S. 945 (1985). 67 869 F.2d 1288 (9th Cir. 1989). 63

88

The Antitrust Response to Monopsony

due to an agreement with the other potential bidders to refrain from competing. The logs from the sale were subsequently allocated among the conspirators and Portac, Inc., which was a sawmill company. The purpose once again was to keep log costs down, and to deny full value to the USFS.68 More recently, big rigging was the issue in prolonged litigation involving tobacco grower and buyers.69 The growers grew tobacco as permitted under a government quota system and sold it at auction. Tobacco that was not sold was put in reserve and sold at a discount. Each year’s quota was determined in part by the prior year’s sales. Plaintiff growers complained that major cigarette companies’ leaf buyers had conspired to keep auction prices and purchases low. This increased the tobacco in reserve and obtainable at lower prices. In addition, it meant lower allocations in future years. Plaintiffs’ theory was that the cigarette companies were acting to eventually end the auction system so that purchases would be made by contracts under which the buyers could make even more effective use of their monopsony power. The theory was sufficiently sound to avoid dismissal, and the case was eventually settled. 4.4.2.3  Refusals to Deal Typically, a refusal to deal involves a seller or sellers who will not sell a much needed input to potential competitors. On the buying side of the market, it means that firms possessing monopsony power may refuse to buy from firms if those firms make sales to firms that compete with the monopsonist in output markets. An example of this is the well-known “monopolization” case of United States v. Griffith.70 In Griffith, the defendant operated a chain of movie theaters, which purchased exhibition rights from film distributors. In some towns, Griffith was the only exhibitor, but in others, it faced competition as an exhibitor. It possessed monopsony power in the towns in which it faced no competition since it was the only buyer there. The defendant made its purchases of film exhibition rights contingent on the condition that Griffith would have exclusive rights to films in the markets in which it faced competition. In essence, rather than use its monopsony The court regarded the arrangement as violative of Section 1 of the Sherman Act but dealt with a different issue. 69 Deloach v. Philip Morris Co., Inc., 2001 WL 1301221 (M.D. N.C. 2001). Other recent cases in which bid rigging by buyers is addressed are Knevelbaard Dairies v. Kraft Foods, Inc., 232 F.3d 979 (9th Cir. 2000); Granite Partners L.P. v. Bear Stearns & Co., Inc., 58 F.Supp. 2d 228 (S.D. N.Y. 1999). 70 334 U.S. 100 (1948). 68

Abuses of Monopsony Power

89

power to demand the lowest possible rental fees in each market, the concession exacted was in the form of exclusive exhibition rights for the entire circuit. Monopsony power was used to enhance the buyer’s monopoly in those markets where it faced competition as a seller. Without specifically noting the monopsony character of the power exercised, the Supreme Court held that the practice violated Section 2 of the Sherman Act.71 Reminiscent of Griffith in a modern high-tech context is Camellia City Telecasters v. Tribune Broadcasting Company.72 Camellia, a syndicator of television programming, claimed that Tribune would only agree to purchase Camellia’s programming for broadcast in New York and Chicago if it would sell the same programming to the Tribune’s Denver television station. As in Griffith, the alleged aim of Tribune was to disadvantage a competing broadcaster in Denver. The court, while evidently accepting the monopsony tying theory, granted summary judgment for the defendant due to Camellia’s inability to sufficiently substantiate its factual claims.73 Monopsony power is also frequently at issue in instances involving more than one firm. An example is Eastern States Lumber Dealers’ Association v. United States,74 a 1914 case leading to the eventual classification of some group boycotts as per se unlawful. The defendants in Eastern States were retailers of lumber who passed among themselves lists of wholesalers who were attempting to sell to retail customers. The apparent purpose was to facilitate an agreement not to purchase from wholesalers who were also competing at the retail level. In effect, the retailers were using their collective power as buyers to decrease the competition they faced as sellers. An antitrust “standard” in which the abuse of monopsony power is more expressly condemned is Klor’s v. Broadway Hale.75 The defendant, Broadway Hale, owned a chain of department stores and used what the Court described as its “monopolistic buying power”76 as a means of convincing several manufacturers of appliances to not deal with Klor’s, a small singleoutlet competing retailer. Rather than concentrate its monopsony power on getting the lowest possible price from the manufacturers, the concession exacted this time was the refusal to sell to a competitor. To some extent, See also United States v. Crescent Amusement Co., 323 U.S. 173 (1944). For a recent case in which exhibitors allegedly conspired to disadvantage a competing exhibition, see Harkins Amusement Enterprises v. General Cinema, 850 F.2d 477 (9th Cir. 1988). 72 762 F. Supp. 290 (D. Colo. 1991). 73 Id. at 294. 74 234 U.S. 600 (1914). 75 359 U.S. 207 (1959). 76 Id. at 209–10. 71

90

The Antitrust Response to Monopsony

Broadway-Hale or its suppliers, may have been concerned about liability under the Robinson-Patman Act, which deals with price discrimination. In an opinion that can be regarded as declaring group boycotts illegal per se, the practice was held to violate the Sherman Act.77 Similarly, in Federal Trade Commission v. Motion Picture Advertising Service,78 the defendant was a producer and distributor of advertising films that were shown in movie theaters. It bought time from theaters only on the condition that they would not sell time to competing producers of advertising films. In essence, the defendant required sellers to enter into an outputs contract under which all their advertising time was sold to the defendant.79 The Supreme Court agreed with the FTC’s finding that the agreements were unreasonable.80 In Griffith, Klor’s, and Motion Picture Advertising Service, monopsony power was not used to negotiate a lower price for the buyers. Instead, the impact was to reduce supply availability to rival sellers, thereby enhancing the position of the defendants as sellers.81 The actual impact in the input market seems to conform more to the all-or-none model with the most serious economic impact actually being felt in the output markets. In these instances, the courts have reached outcomes that are consistent with established economic objectives by focusing on the results of increased mono­ poly power. The use of monopsony power to enhance a firm’s monopoly power was also at issue more recently in Sunshine Cellular v. Vanguard Cellular System, Inc.,82 a case that also illustrates the complexities that can arise when both buying-side and selling-side markets must be defined. In fact, to some extent the case raises the issue of what it means to be a “buyer.” Sunshine was a cell phone service provider in a geographic area that was effectively surrounded by areas served by Vanguard. Vanguard refused, however, to In Business Electronics v. Sharp Electronics, 108 U.S. 1515, 1525 (1988), Justice Scalia reinterpreted Klor’s, suggesting that per se liability in that case depended on a horizontal agreement among the boycotting manufacturers. 78 344 U.S. 392 (1953). 79 Cases like Griffith, Tribune Broadcasting, and Motion Picture Advertising Service have a “tying” element in them in that the buyers threatened to withhold purchase of a product, service, or right unless a more desired product, service, or right was included in the transaction. Typically, the use of market power by sellers to tie the sale of a less desirable product to the purchase of a more desirable one is treated under Section 1 of the Sherman Act or Section 3 of the Clayton Act. When services are involved, only Sections 1 and 2 of the Sherman Act are applicable. In the case of tying by buyers, only the Sherman Act is available, as the Clayton Act applies expressly to sellers. 80 Id. at 394–8. The FTC limited the length of the agreements. 81 Id. 82 810 F.Supp. 486 (S.D. N.Y. 1992). 77

Abuses of Monopsony Power

91

enter into the typical “roaming” agreement. Under those standard agreements, customers who want to make calls from areas not covered by their providers can make those calls. Their own providers bill them and remit the funds to the provider in the areas from which the calls were placed. This can be viewed as buying the service from a provider in the area into which a customer has roamed. A great deal of Sunshine’s revenue came from customers of Vanguard. Essentially, Sunshine argued that Vanguard’s refusal to buy the roaming services from Sunshine greatly reduced its income and was part of a Vanguard effort to eventually monopolize the market for cell phone services.83 To survive a motion to dismiss, which it did successfully, Sunshine was required to identify both the relevant seller-side and buyerside markets. 4.4.2.4  Monopsony Leverage and Tying The use of monopsony leverage in the preceding cases suggests that they represent the buyer-side analog to exclusive dealing requirements84 or tying by sellers. To understand the limitations of applying a tying theory to buyers, it is useful to first examine tying by sellers. In the context of sellers, it can be a violation of the Sherman Act or Section 3 of the Clayton Act to sell a product – the tying product – only on the condition that the purchaser also buy another product – the tied product.85 The seller supposedly is able to use its market power in the tying product market to “force” the purchase of the tied product.86 In recent years, scholars have pointed out the dangers of summarily condemning tying arrangements.87 The argument is that the seller has a single Vanguard’s refusal limits its customers’ ability to use their cell phones when they roam into Sunshine’s area. This, of course, reduces the attractiveness of Vanguard’s service. It is hard to see a procompetitive motive for Vanguard’s policy. 84 Monopsonistic exclusive dealing would involve a demand that a supplier sell all of its output to a single buyer. Almost certainly the impact is to eliminate firms that would compete with the monopsonist on the selling side of the market. For an example see White Mule Company v. ATC Leasing Company LLC, 540 F.Supp.2d 869 (N.D. Ohio 2008). 85 The Clayton Act applies expressly to “goods, wares, merchandise, machinery, supplies or other commodities.” The Sherman Act would apply in the case of services. 86 See Roger D. Blair & Jeffrey Finci, The Individual Coercion Doctrine and Tying Arrangements: An Economic Analysis, 10 Fla. St. U. L. Rev. 531 (1983). In Jefferson Parish Hospital District No. 2 v. Hyde, 466 U.S. 2 (1984), the Court emphasized that “the essential characteristic of an invalid tying arrangement lies in the seller’s exploitation of its control over the tying product to force the buyer into the purchase of a tied product.” It went on to say that “[w]hen ‘forcing’ occurs, our cases have found the tying arrangement to be unlawful.” 87 See Ward Bowman, Tying Arrangements and the Leverage Problem, 67 Yale L. J. 19 (1957); Roger D. Blair & David L. Kaserman, Antitrust Economics, 2d ed. 406 (New York: Oxford University Press 2009); Sullivan & Harrison, supra note 10 at 227–9. 83

92

The Antitrust Response to Monopsony

amount of monopoly power and a single profit maximizing price for the tying product. Forcing customers to buy a product that they would prefer either not to buy at all or to buy from a different seller is the equivalent of raising the price of the tying product. Consequently, any increased profit from the tying can only occur because consumers actually find the “packaged” product more attractive, there are economies associated with the manufacture and distribution of the packaged product, or there is some possibility of establishing a separate base of monopoly power in the tied product market. This range of possibilities suggests to many that tying should not be regarded as a per se violation of the antitrust laws.88 As the cases in the preceding section point out, firms with monopsony power do sometimes condition the purchase of a good or service on the agreement of the seller to also sell another good or service to that purchaser. The tying purchase takes place in the market where the firm has monopsony power and the tied purchase occurs in the market in which it lacks power. Thus, in Griffith, the exhibitor would purchase film rights for some markets only if it were also sold the exhibition rights for other markets.89 Those markets in which the defendant was the only exhibitor would be the tying markets, and the markets in which there were other exhibitors would be the tied markets. In all of the instances cited, the purchasers’ ultimate aim was to deny access to the tied input by competing sellers. As in the case in which the tying seller hopes to establish market power in the tied product market, these cases obviously involve efforts to extend market power to an additional market, albeit on the selling side of the market. More problematic are instances in which the monopsonist does not sell in competition with other purchasers of the tied product. Here, like the typical seller tying case, economic theory suggests that the monopsonist can get the full benefit of its power simply by selecting the proper price for the tying purchase. Any effort to also force the sellers to sell another product or service would be comparable to lowering the price even further for the tying purchase. As in the seller example, this only makes sense if the packaged sale is more attractive to the seller, there are economies to the buyer of buying the items as a package, or there is a likelihood of establishing independent monopsony power in the market for the tied good. In Jefferson Parish Hospital Dist. No. 2 v. Hyde, 466 U.S. 2 (1984), and in Eastman Kodak v. Image Technical Severices, 504 U.S. 451 (1992) the Supreme Court reaffirmed the per se status of tying. The requirements for a per se ruling in tying cases are, however, far more extensive than for other per se violations. 89 This amounts to a different “product” in the sense that the rights were to exhibit in different geographic markets. 88

Monopsony and Merger Policy

93

4.5  Monopsony and Merger Policy A monopsony issue in the area of horizontal mergers is raised when one buyer acquires a rival buyer and thereby increases the possibility that there will be an undesirable concentration of power on the buying side of the market. The monopsony issue in the context of vertical mergers arises when a firm acquires one of its suppliers. The motivations and the welfare significance, however, are quite distinct. Consequently, horizontal and vertical mergers are treated separately in this section.

4.5.1  Horizontal Mergers A series of horizontal mergers among firms that buy the same inputs can lead to a case of pure monopsony.90 In other instances, the merger may result in a dominant firm in a market with several smaller or fringe buyers. In either case, the merger to monopsony may or may not involve monopoly in the output market. For example, a series of mergers could result in a single exhibitor controlling all of the movie theaters in a geographic area.91 One could argue that the single exhibitor has a monopsony in the area since no one else licenses movies while recognizing that television, radio, plays, concerts, shows, and the like provide enough reasonable substitutes to blunt the exhibitor’s monopoly power.92 In the following analysis, monopsony power without any corresponding monopoly power is assumed. In this case, the merged monopsonist still imposes welfare losses on society. On these grounds, the result, whether due to pure monopsony or the emergence of a dominant buyer, is similar to that of the classical monopsonist. The economic harm will be as described in Chapter 3, and, unless these social welfare losses are offset by productive efficiencies, the merger should be proscribed.93 For a brief treatment of buyer concentration, see Phillip Areeda & Donald Turner, Antitrust Law, vol. 4, 963 (Boston: Little, Brown 1980). 91 This was an issue raised de novo in a posttrial brief by the government in United States v. Syufy Enterprises, Inc., 712 F. Supp. 1386 (N.D. Cal. 1989). The government had been challenging Syufy’s theater acquisitions in the Las Vegas market on traditional (i.e., monopolistic) grounds, but subsequently claimed that the mergers were designed to achieve monopsony power. 92 Since United States v. Paramount Pictures, Inc., 67 U.S. 915 (1948), the relevant product in motion picture cases has been first-run exhibition of motion pictures. See United States v. Syufy Enter., 712 F. Supp. 1347. Charles McCoy, The Paramount Cases: Golden Anniversary in a Rapidly Changing Marketplace, Antitrust 32 (1988), calls for a rethinking given modern realtities. 93 The original efficiencies defense was offered by Oliver Williamson, Economies as an Antitrust Defense: The Welfare Tradeoffs, 58 Am. Econ. Rev. 18 (1968), who considered a 90

94

The Antitrust Response to Monopsony

As in a case of a horizontal merger of competing sellers, the critical issue is to determine the level of market power enjoyed by the newly merged firm. The most commonly employed indicator of market power is market share. Whether one relies on market share or the more sophisticated Lerner Index, as discussed in Chapter 3, the difficult problem is to define the market. All products that are “reasonably interchangeable” to buyers should be included in the market. The level of interchangeability will depend on how quickly buyers respond to price changes by altering their buying patterns. Thus, if buyers turn quickly to other sellers of the same product or sellers of different products as soon as one seller raises prices, the alternatives are part of the same market. Technically, the propensity to switch in response to price changes is measured by the elasticity of demand or the cross-elasticity of demand. Another factor limiting market power is the responsiveness of potential sellers to price increases as measured by supply elasticity. Thus, even when a firm has a large market share, if a price increase encourages firms to enter the market, the market power of the firm is diminished.94 In the case of monopsony, market definition is in many respects the reverse of that applied in the case of sellers. First, the market is not the market of competing sellers but of competing buyers. This market is comprised of buyers who are seen by sellers as being reasonably good substitutes. The greater the number of good substitutes from the point of view of sellers, the more easily sellers may substitute away from low-paying buyers in favor of higher-paying buyers. This is measured by the sellers’ supply elasticity for the merged buyers. Also dampening the power of the monopsonist is the ability of new buyers to enter the market in response to the artificially depressed price. Here, a high elasticity of demand will lessen monopsony power. Although hardly an everyday occurrence, the courts do consider cases of merger to monopsony.95 In United States v. Rice Growers Association of California,96 a merger of firms engaged in the purchase, milling, and resale of rice was challenged by the Department of Justice on a number of bases, including the possibility that it would “substantially lessen competition in the market … [for] [t]he purchase or acquisition of paddy merger that led to productive efficiencies, which reduced costs, and allocative efficiency due to price exceeding marginal cost. He demonstrated how, in priniciple, one might weigh the conflicting effects. 94 See the discussion in Section 3.5. 95 Omnicare Inc., v. Unitedhealth Group, Inc., 524 F.Supp. 2d 1031 (N.D. Ill. 2007); U.S. v. Aetna, Inc., 1999 WL 1419046 (N.D. Tex. 1999). 96 1986 WL 12562 (E.D. Cal. 1986).

Monopsony and Merger Policy

95

rice grown in California.”97 The court held that the acquisition violated Section 7 of the Clayton Act because of the decrease in competition among purchasers. In so holding, the court engaged in the appropriate market analysis. For example, it explored the possibility of “substitute” buyers and noted that California rice growers “perceive California rice mills as their best alternative for the sale of paddy rice.”98 In addition, it was not feasible for growers to shift to other crops. In effect, the rice growers’ supply was relatively inelastic. With respect to possible new demands for the rice as its price was forced down, the court found that shipment to other parts of the country or overseas was not economically feasible. United States v. Syufy Enterprises99 illustrates the problems that arise when markets are confused. The government challenged the acquisition practices of a movie theater chain in the Las Vegas market as being in violation of Section 2 of the Sherman Act and Section 7 of the Clayton Act. Evidently, at trial the government based its case on the possible anticompetitive effects of the acquisition in the market for the first-run exhibition of films; that is, the output market. In defining that market, the issues would be whether there were good substitutes in the eyes of consumers for first-run films and whether good substitutes might appear in the market if prices were elevated. In a posttrial brief, however, the government made a merger to monopsony argument based on Rice Growers and cited the impact on film distributors as the harm to be avoided.100 The court dismissed the argument as really an effort by the government to reduce the relevance of the demand elasticity of consumers.101 The court noted that the government could not begin with one theory of impact and then, when the market definition began to favor the defendants, change to a theory that had not been presented at trial. What seems clear from the opinion is that the government discovered the merger to monopsony theory too late. Moreover, if there was any harm due to Syufy’s acquisitions, it was to the distributors of films. Consequently, the proper focus for the government should have been the availability of alternative outlets to the sellers.102 Id. at 12. Id. at 5. 99 712 F. Supp. 1386 (N.D. Cal. 1989), aff ’d., 903 F.2d 659 (9th Cir. 1990). 100 Id. at 1398. 101 Id. at 1397–8. 102 For the merger to monopsony issue in the context of a regulated industry, see Environmental Action, Inc. v. F.E.R.C., 939 F.2d 1057 (D.C. Cir. 1991). 97 98

96

The Antitrust Response to Monopsony

4.5.2  Vertical Mergers Since 1971, a great deal of attention has been paid to the motives for vertical integration.103 Most of this attention has focused on the incentives for an intermediate good monopolist to integrate forward to the downstream stage. Not nearly as much attention has been paid to the theory of vertical integration by a monopsonist buyer of an input.104 As the following model illustrates, there is a strong incentive for a monopsonist to integrate upstream and acquire its input suppliers. In Figure 4.4, the monopsonist’s derived demand curve for an input is graphed as the downward sloping marginal revenue product curve (MRP). The industry average cost of producing the input is represented by AC. The industry’s marginal cost curve, which corresponds to the industry supply curve, is denoted by MC. The industry supply curve shows the average purchase price to the monopsonist for any particular quantity. In contrast, the marginal factor cost (MFC) represents the marginal cost to the monopsonist of increasing the employment of this input by one unit. The marginal factor cost lies above the supply curve because increases in the employment of this input require paying increased prices. Since the higher price must be paid for all units of the input, the marginal impact on total expenditures (MFC) exceeds the price, which is the average impact.105 In the absence of vertical integration, the monopsonist will employ Q1 units of the input as determined by the intersection of MRP and MFC at point B. At this level of employment, the monopsonist’s total revenue is given by the area ABQ10. Total cost is equal to price times quantity P1Q1, which is equal to the area P1CQ10. Consequently, the nonintegrated monopsonist’s maximum profits are given by the area ABCP1 in the figure. Interestingly, the burst of attention was spawned by a two-page note that rediscovered an earlier result. See John Vernon & Daniel Graham, Profitability of Monopolization by Vertical Integration, 79 J. Pol. Econ. 924 (1971); and M. L. Burnstein, A Theory of Full-Line Forcing, 55 Northeastern U. L. Rev. 62 (1960). 104 Vertical integration by a monopsonist has been examined most thoroughly by Martin Perry, Vertical Integration:  The Monopsony Case, 68 Am. Econ. Rev. 561 (1978). The mathematics in this paper may make the message inaccessible to many readers. Other treatments that are not as thorough but are still useful include S. Y. Wu, The Effects of Vertical Integration on Price and Output, 2 W. Econ. J. 117 (1964), John McGee & Lowell Bassett, Vertical Integration Revisited, 19 J. L. & Econ. 17 (1976); and Roger D. Blair & David L. Kaserman, Law and Economics of Vertical Integration and Control 114 (New York: Academic Press 1983). 105 Since the total cost of employing x is PxQx where Px is an increasing function of x, the marginal effect of expanding employment is the marginal factor cost dPxQx/dQx = Px+Qx (dPx/dQx), which is necessarily greater than Px since dPx/dQx is positive. 103

Monopsony and Merger Policy

97

Price and Cost MFC MC

A B

D

AC

C

P1 F

G

E

MRP 0

Q1

Q2

Quantity

Figure 4.4.  Monopsony and Vertical Merger.

With complete vertical integration, the optimal employment of the input increases from Q1 to Q2 as determined by the intersection of MRP and MC at point D.106 Here, total revenue is area ADQ20, and total cost is area FEQ20. Thus, the fully integrated firm’s maximum profits are given by the area ADEF in the figure. The gross increase in profit that results from vertical integration by the monopsonist is given by the area P1CBDEF in the figure. This, of course, is what provides the incentive for vertical integration, but whether this is accomplished through merger or internal expansion depends on the relative costs of the two avenues. The increased profits that result from vertical integration flow from two sources. Area P1CGF represents that portion of the upstream industry’s pre-integration rents that are captured by the monopsonist.107 The second source of increased profits arises from eliminating the efficiency loss due to underemploying the monopsonized input in the absence of vertical integration.108 This is the “efficiency effect,” and it is shown as area GBDE in the ­figure. Clearly, the relative sizes of these two effects depend Following vertical integration, the firm will look at the marginal cost of producing the input rather than the marginal cost of buying the input in reaching its employment decisions. 107 See Perry, supra note 104, at 568–9. 108 See Blair & Kaserman, supra note 87, at 118. 106

98

The Antitrust Response to Monopsony

on the slope of the MRP curve. As MRP becomes flatter, the efficiency effect becomes larger, and the “rent effect” becomes smaller. The sum of the profits resulting from these two effects provides the total increase in profits from vertical integration under monopsony as shown by the shaded area in the figure. This increase in profits measures the gross benefits to the monopsonist of acquiring productive capacity at the upstream stage. As long as the costs of acquisition are smaller than these gross benefits, the acquisition will occur. In this case, the welfare effects of the monopsonist’s vertically integrating backward are unambiguously positive. As the monopsonist vertically integrates, the employment of the monopsonized input is expanded. This results in greater output of the final good and, of course, a lower price. If the monopsonist is a perfect competitor in the output market, the price reduction will be very slight indeed. But if the monopsonist has some market power in the final good market, the price reduction may be quite substantial. In either event, consumers are not hurt by the vertical integration. In the case considered here, the acquiring firm was a pure monopsonist. The same effects, however, follow in the case of purchasers with a lower level of monopsony power. Generally, market foreclosure is the main objection to vertical mergers.109 In the case of pure monopsony, however, there is no actual foreclosure. Since the monopsonist buys all of the upstream industry’s output, no rival purchaser is foreclosed from this source of supply. Similarly, since the monopsonist wants to buy all of the input produced, no suppliers are foreclosed from the opportunity of selling to the monopsonist. A stronger possibility of foreclosure arises if the acquiring firm is not a pure monopsonist. Here, the argument would be that other purchasers would be unable to obtain the input. In addition, the independent suppliers would have one less outlet. As has been pointed out elsewhere, it is not clear that this should be a concern.110 If the acquiring firm obtains its output from the acquired firm, the output of other suppliers would seem to be available to the competing purchasers. In an extreme case, the acquiring firm may The most recent Supreme Court decision on vertical integration is Ford Motor Co. v. United States, 405 U.S. 562 (1972), which involved Ford’s acquisition of Electric Autolite, a spark plug manufacturer. The trial court objected to the acquisition because of “the foreclosure of Ford as a purchaser of about 10 percent of the total industry output.” Id. at 569 (quoting United States v. Ford Motor Co., 315 F. Supp. 372, 375 (1971)). An issue of foreclosure also arose in AlliedSignal, Inc. v. B. F. Goodrich, Co., 183 F.3d 586 (7th Cir. 1999), in which the plaintiff argued that a merger between firms manufacturing components of aircraft landing gear could render it unable to acquire necessary parts. 110 Bork, supra note 16, at 231–8. 109

Monopsony and Price Discrimination

99

acquire all producers of the input. In effect, it will have become a monopolist in the input market. If so, the horizontal concerns about monopolists in general would apply. Another variation on the foreclosure issue involves potential foreclosure. If another final good producer were to enter that industry, it would have no source of supply. This is often seen as a barrier to entry because a new entrant must also be vertically integrated. But entry occurs in response to excess profits. If there are excess profits, the fact that a new entrant must enter at both levels can, under conventional analysis, only delay entry – it cannot prevent it.111. This delay is not without value to the established firm because the period during which excess profits are earned is longer as a result. In the case at hand, one must weigh the certain and immediate gain in consumer welfare against this problem of potential foreclosure.

4.6  Monopsony and Price Discrimination The Robinson-Patman Act of 1936, which amended Section 2 of the Clayton Act, forbids price discrimination by sellers when it is likely to have an anticompetitive effect.112 Most cases involved either primary line or secondary line price discrimination. In primary line cases, the seller sells at a higher price in the market in which demand is relatively inelastic and for a lower price in the market in which it is relatively elastic and the seller faces a competitor. The theory is that the discriminating firm is attempting to eliminate a direct competitor. The Supreme Court has indicated that these cases will be assessed as predatory pricing cases.113 Under secondary line, the seller’s efforts are not directed at its own competitor. Instead, the competitive impact is at the level of the customer. Here the theory is that a powerful buyer receives the lower prices in an effort to eliminate competition at its own level. Secondary line discrimination, until recently, was designed to protect small firms from more powerful firms that possessed monopsony power.114 For the most part, the Robinson-Patman Act does not address the behavior of buyers; usually the seller comes under fire when Under “real-world” conditions, entry is not free, that is, it involves some transaction costs. Thus, entry can only occur if the transaction costs associated with entry do not exceed the gains to be had by entry. 112 Under Section 2, price discrimination is prohibited when it “may tend substantially to lessen competition,” 15 U.S.C. §13 (1988). 113 Brooke Group Ltd., v. Brown & Williamson Tobacco, 509 U.S. 209 (1993). 114 Recent cases suggest that this concern is of secondary importance. See Volvo Truck N.A. v. Reeder-Simco GMC, 546 U.S. 164 (2006). 111

100

The Antitrust Response to Monopsony

there is an allegation of price discrimination. But the powerful buyer is not ignored totally as Section 2(f) provides “[t]hat it shall be unlawful for any person engaged in the course of such commerce, knowingly to induce or receive a discrimination in price which is prohibited by the section.”115 In principle, this would seem to impose liability on large buyers who extract discriminatory price concessions from a seller.116 But if that seller were also selling to other buyers at a different price, there would be no liability under the Robinson-Patman Act. From the point of view of buyers, price discrimination could also describe primary and secondary line effects. A primary line case would involve a situation in which a buyer offers higher prices in a market in which it is attempting to eliminate a competing buyer. Since the Supreme Court has ruled that primary line cases on the selling side of the market will be treated under predatory pricing standards, it seems likely that primary line cases on the buying side of the market, if addressed at all, will be assessed under predatory buying standards.117 In a secondary line case, the buyer would buy from firms in competition and favor one over the other by paying a higher price for its output. Presumably, this would put the favored firm at an advantage vis-à-vis competing buyers, in the market for its own upstream purchases. As in the case of seller-side secondary line price discrimination in which the less powerful buyer is eliminated, secondary line discrimination by a buyer would tend to eliminate the less powerful seller. In a secondary line case, the market power of the discriminating party is not an issue. Instead, the relevant power is that held by the favored seller. As in the conventional seller-side cases of secondary line price discrimination, it is not clear that this is necessarily anticompetitive. As noted earlier, the Robinson-Patman Act, with its emphasis on the sale of the same product to different buyers, is unlikely to be extended to the price-discriminating buyer. That is not to say that price discrimination by a buyer would not raise concerns under other sections of the antitrust laws. Consequently, it is useful to examine in greater detail the economics of monopsonistic price discrimination. The most likely sources of monopsonistic price discrimination would seem to arise in connection with

15 U.S.C. §13(f) (1988). Great Atl. & Pac. Tea Co. v. FTC, 440 U.S. 69 (1979); Automatic Canteen Co. of Am. v. FTC, 356 U.S. 61 (1953); Am. Motors Specialties Co. v. FTC, 278 F.2d 225 (2d Cir.) cert. denied, 364 U.S. 884 (1960). 117 See discussion of Weyerhaeuser, supra Section 4.3. 115 116

Monopsony and Price Discrimination

101

Cost

Average Cost AC2 AC1

0

Q2

Q1

Quantity

Figure 4.5.  Monopsony and Scale Economies.

unexploited economies of scale or as a means of exploiting differences in supply elasticities. These concepts are explored in turn later.

4.6.1  Monopsony and Unexploited Scale Economies In industries with unexploited scale118 economies, there is excess capacity, which simply means that some firm or firms could exit and the costs of production would not rise. When suppliers generally have excess capacity, large buyers may recognize that they have some buying power that could be used to extract favorable terms. In Figure 4.5, the seller initially is able to produce and sell Q1 units of output while incurring per unit costs of production and distribution equal to AC1. The existence of excess capacity is signaled by the fact that the producer’s average cost curve (AC) is decreasing at output Q1. Suppose a large buyer accounts for Q1 – Q2 units of the total output sold. If that buyer threatens to withdraw its business and go elsewhere for its supplies of this product, there will be two effects on the seller, and neither is good. First, Economies of scale are said to exist when the firm is producing on the negatively sloped portion of its average cost curve. Expansions in the quantity produced will result in reductions in per unit costs. For an extensive examination of this concept and its implication for market structure, see Frederick M. Scherer, Industrial Market Structure and Economic Performance 81 (Chicago: Rand McNally 1980).

118

102

The Antitrust Response to Monopsony

it will lose whatever profit it had been earning on that buyer’s purchases. This, of course, is fairly obvious. Second, and somewhat less obviously, the seller will lose some of the profit that it had been earning on sales to its other customers. In the event that the large buyer in question actually goes elsewhere, the seller’s quantity will fall to Q2 and the average cost of production and distribution will rise to AC2. As a consequence, the total cost of serving the customers that remain rises by the striped area, equal to (AC2 – AC1) Q2. A sophisticated large buyer may engage in strategic behavior in an effort to extract some added benefits from the seller with excess capacity. For example, it could threaten to withdraw its business unless the seller agrees to a price equal to AC1. This would mean no profits on the sales to the large buyer, but the seller would preserve all of the profit on its sales to its other customers. In the limit, the large buyer could require that the seller actually agree to a price below cost such that losses would be incurred on sales to the large buyer equal to the striped area. At that point, the seller would be indifferent between losing that account and retaining it since total profits are the same in either event. This business behavior appears to be abusive, but what makes it possible is the presence of excess capacity. To the extent that some of the sellers will suffer losses and have to leave the industry, it is economically beneficial. Society has too many resources invested in the production of the input and, therefore, it is socially desirable that some capacity exit the industry. Thus, the large buyer’s behavior may appear unsavory, but it leads to desirable social results. The relevant section of the Robinson-Patman Act for assessing buyer liability is 2(f).119 There can be no violation of 2(f), however, unless the seller itself has violated the Act. One way for the seller not to have violated the Act is for it to have a valid defense. In the excess capacity example, the buyer with monopsony power is likely to solicit bids from multiple sellers with excess capacity. Each seller would then have a valid “meeting competition” defense for its discriminatory offer as long as it had done no more than make a good faith effort to meet the prices of competing sellers.120 See Great Atl. & Pac. Tea Co., supra note 116. Although price discrimination is illegal if it has or is likely to have an anticompetitive effect, a price concession made in good faith to meet competition is legal irrespective of the competitive consequences. Section 2(b) provides that “nothing herein contained shall prevent a seller rebutting the prima facie case thus made by showing that his lower price … was made in good faith to meet an equally low price of a competitor.” The Supreme Court has interpreted this defense as absolute. See Standard Oil Co. v. Federal Trade Commssion, 340 U.S. 231 (1951).

119 120

Monopsony and Price Discrimination

103

Consequently, it is likely most large buyers would escape liability in most instances.

4.6.2  Different Supply Elasticities The discriminating monopsonist may exploit differences in supply elasticities to pay different prices for inputs of identical quality.121 In doing so, the monopsonist maximizes its profits. This is shown in Figure 4.6 where S1 and S2 represent alternative sources of supply of a specific input. The analysis assumes that the quality of the input is uniform across both sources of supply. The monopsonist will purchase from the two supply sources such that the marginal factor costs are equal to the marginal revenue product: MRP = MFC 1 = MFC2

If this condition is not met – say MFC1 exceeds MFC2 – then the monopsonist can reduce purchases from source 1 thereby saving MFC1 and replace those inputs with purchases from source 2, thereby expending MFC2. If, as hypothesized, MFC1 exceeds MFC2, such a rearrangement will reduce total expenditure and, therefore, increase profits. Consequently, such shifts will continue until the marginal factor costs are equal across all sources of supply. In Figure 4.6, the horizontal summation of MFC1 and MFC2 is shown as ΣMFC. The monopsonist will buy a total quantity equal to Q. These purchases will be split between sources 1 and 2 such that MFC1 equals MFC2, resulting in Q1 being purchased from source 1 and Q2 being purchased from source 2. Due to the differences in supply elasticities, the price paid to source 1 suppliers is w1, whereas the price paid to source 2 suppliers is w2. Since the quality of the inputs (and, therefore, their productivity) is identical across sources of supply, we say that the price difference is discriminatory. The consequences of monopsonistic price discrimination under the conditions described in this section can be summarized briefly. First, the monopsonist’s total purchases are unaffected and, therefore, the subsequent total output produced by the monopsonist will not change. Thus, there is no impact on the monopsonist’s customers. Second, the single price paid for the input by a nondiscriminating monopsonist would lie between w1 and w2. The discriminatory monopolist similarly exploits differences in demand elasticities to charge different prices. See Roger Blair & Larry Kenny, Microeconomics with Business Applications, 268 (New York: John Wiley 1987).

121

The Antitrust Response to Monopsony

104

Price and Cost MFC1

S1

MFC2 ΣMFC S2

w1 w2

Demand

0

Q1

Q2

Q=Q1+Q2

Quantity

Figure 4.6.  Monopsony and Different Supply Elasticities.

This means that the effect of price discrimination is to expand purchases from source 1 and contract them from source 2. Finally, it should be noted that some producer surplus is converted to consumer surplus, which is why the monopsonist is motivated to discriminate. As already noted, the Robinson-Patman Act’s prohibitions do not appear to extend to the discriminatory monopsonist. Nevertheless, “favoring” one supplier over another could be viewed as an abuse of monopsony power, but no cases on this point have been discovered.

4.7  Concluding Remarks This chapter examined the actual antitrust treatment of monopsony as we have found it in various court opinions. In addition, it suggests the most likely path that the courts will follow in cases of first impression by analogy to decisions dealing with sellers. In addition, some variations on the classical model of monopsony have been developed to explain the exercise of buying power implicit in several cases in which it was not expressly addressed. For the most part, the extant decisions are consistent with the promotion of consumer welfare. Many of these cases were decided before the emphasis on economic reasoning became as prominent as it is today. Thus, it is a bit

Concluding Remarks

105

surprising that the results have been so consistent with that approach. Of course, there is the possibility that the courts reached the right conclusions for the wrong reasons. Now, however, there appears to be a greater awareness of the economic consequences of various business practices. Decisions in cases dealing with monopsony should be on firm ground and should continue to promote consumer welfare.

FIVE

Cooperative Buying Efforts

5.1 Introduction As Chapter 3 illustrates, collusive monopsony can present many of the same economic concerns as pure monopsony. Accordingly, the antitrust response to price fixing by buyers generally has been similar to the response to price fixing by sellers. In actuality, determining the appropriate antitrust response to joint decision making by competing buyers can be a complicated matter. The problem is that when buyers cooperate, one cannot predict the economic results on an a priori basis. Two common examples of buyer cooperation illustrate this point. First, buyers may agree to use the same inputs or to use them in a specified combination resulting in product standardization. For example, all the child care providers in a community may form an associate and agree that, as a condition of membership, no less than a certain portion of the caregivers will have college training. By adhering to the rule and publicizing their standard, it would be possible to lower the transaction costs of prospective customers. This, of course, will tend to increase the demand for the services of association members. On the other hand, such an agreement would also amount to an agreement to limit demand for other inputs, thereby driving their prices down. It could also increase cost to a common level and, therefore, lead to price uniformity. A second common example is a buying cooperative. When buyers join together to make purchases, there are three basic outcomes. First, “cooperative buying” may be nothing more than a euphemism for collusive monopsony that drives prices below competitive levels and has negative economic effects on social welfare similar to those caused by price-fixing sellers. Second, a cooperative buying effort may not result in an aggregation of buying power at all and, therefore, the use of this power cannot be 106

The Theory of Joint Purchasing

107

the motivation. Instead, these joint ventures are formed for the purpose of exploiting productive efficiencies that result from conducting businesses on a larger scale.1 In fact, in Northwest Wholesale Stationers v. Pacific Stationery and Printing,2 a 1985 decision, the U.S. Supreme Court recognized that cooperative buying enabled the cooperating members “to achieve economies of scale in purchasing and warehousing that would otherwise be unavailable.”3 A third possibility is to have elements of both monopsony, with the attendant losses in allocative efficiency, and joint purchasing efficiency, resulting in productivity gains, at the same time. This chapter focuses on the economics and antitrust law applicable to these two types of cooperative buying endeavors. The analysis can also be extended to other joint buyer activities. The economics of product standardization is fairly straightforward. Before proceeding, however, it is necessary to examine in far greater detail the economic theory in the case of buying cooperatives. Since the theory of collusive monopsony with no off-setting efficiencies was discussed in Chapter 3, the following analysis is confined to instances in which efficiencies do flow from the cooperative’s efforts. Following this theoretical section, the antitrust law relevant to both product standardization and cooperative buying is discussed.

5.2  The Theory of Joint Purchasing When buyers cooperate, it is possible to realize pure efficiency gains in production without the emergence of buying power. We examine this case These efficiencies may be realized in distribution and promotion as is probably the case with the Topco or IGA grocery cooperatives. Alternatively, they may well be due to reduced transaction costs. On the latter, see Oliver Williamson, The Economics of Antitrust: Transaction Cost Considerations, 122 U. Pa. L. Rev. 1429 (1974). 2 472 U.S. 84 (1985). Northwest Wholesale Stationers was a purchasing cooperative comprised of about 100 office supply retailers. It was the primary wholesaler for its members. The suit involved a dispute over the expulsion of Pacific Stationery & Printing and whether such expulsion constituted an illegal boycott. The Supreme Court held that it did not. 3 Id. at 286–7. In the 1972 case of U.S. v. Topco Associates, 405 U.S. 596 (1972), a group of small and medium-sized supermarket chains combined to procure for and distribute to its members food and nonfood items under the Topco label in order to compete with the private brands of larger chains. One element of their scheme was the designation of territories within which the various sellers of the Topco brand would not compete. This horizontal market division was found to be a per se violation of Section 1 of the Sherman Act. Id. at 608. Another possible objection to the arrangement could have been that it involved horizontal price fixing by buyers. In fact, the horizontal agreement on price was necessary in order for the members to enjoy the economies of large-scale purchases. Thus, rather than exercise market power the buyers were simply combining their orders in order to operate more efficiently. 1

Cooperative Buying Efforts

108 Price

S w2 w1

VMP2

VMP1 0

Q1

Q2

Quantity

Figure 5.1.  Cooperative Buying with Productive Efficiencies.

first. Then we turn our attention to the case where the productive efficiency is accompanied by market power and the allocative efficiency losses that result.

5.2.1  Efficiency-Enhancing Cooperative Buying Ventures When a cooperative buying venture results in greater productive efficiency, this is reflected in higher marginal products of the inputs employed. For given output prices, this will cause the value of the marginal product to increase and, in turn, will cause the value of the marginal product curve (VMP) to shift to the right.4 This is illustrated in Figure 5.1. Prior to the formation of the cooperative, the value of the marginal product was VMP1. As a result, the quantity purchased was Q1 and the price was w1. Since cooperative purchasing enhances efficiency, VMP1 shifts to VMP2. The result of For example, the Cobb-Douglas production function is Q = ALαKβ where Q denotes output, L and K represent labor and capital inputs, and A is an efficiency parameter. The marginal product of labor is MPL = αALαKβ/L, which is obviously affected by the efficiency parameter, A. If production and distribution are organized more efficiently, A will increase, and both inputs will be more valuable to the firm. Clearly, an increase in A will cause the VMP curve to shift to the right.

4

The Theory of Joint Purchasing

109

cooperation in this case is to expand employment from Q1 to Q2, which necessarily requires an increase in the price from w1 to w2. In this case, we can see that both producer surplus and consumer surplus increase – everyone gains. This, of course, should be applauded. Thus, it is easy to distinguish the case of collusive monopsony from the case of an efficiency-enhancing buying cooperative. Collusive monopsony is accompanied by a decrease in purchases and a decrease in the price paid. In contrast, the case of the efficiency-enhancing buying cooperative is characterized by an increase in the price paid for the intermediate good.5 Both buyers and sellers are better off. Due to these economic results, as a practical matter, it is extremely unlikely that there will be any antitrust complaint about efficiency-enhancing buying cooperatives.

5.2.2  Enhanced Efficiency and Increased Buying Power A more difficult case arises when the buying cooperative’s efforts result in a trade-off between the productive efficiency illustrated in Figure 5.1 and allocative inefficiency.6 Whether society is better off or worse off depends on the magnitudes of the conflicting effects, because there are losers as well as winners. Two possibilities emerge. First, if the effect on productive efficiency swamps the allocative inefficiency, the result will be an increase in purchases and an increase in price, which means that the welfare effects are necessarily positive.7 It is tempting to conclude from this that agreements with this combination of effects should pass antitrust muster, but that is not necessarily the case. It is possible to see why by examining Figure 5.2. Initially, the demand curve for this intermediate good is represented by the value of the marginal product curve labeled VMP1. Prior to the formation of the cooperative buying venture, the market equilibrium occurred at the intersection of the demand (VMP1) and the supply (S). The quantity was Q1 and the price was w1. The formation of the cooperative buying venture changed all that because it enhanced efficiency at the same time that it conferred buying power on the former purchasing rivals. An analogous output test for horizontal price fixing among sellers was proposed by Roger D. Blair & Richard E. Romano, Distinguishing Participants from Nonparticipants in a PriceFixing Conspiracy: Liability and Damages, 28 Am. Bus. L. J. 33 (1990). 6 A similar welfare trade-off was noted by Oliver Williamson, Economics as an Antitrust Defense: The Welfare Tradeoffs, 58 Am. Econ. Rev. 18 (1968), in analyzing mergers among rival producers that increased productive efficiency due to some horizontal integration at the same time that market power on the selling side increased. 7 When price rises and quantity rises, both producer surplus and consumer surplus must rise. As a result, the welfare effect must be positive. 5

Cooperative Buying Efforts

110 Price

MFC

S

w2

w3 w1 VMP2 VMP1 0

Q1 Q3

Q2

Quantity

Figure 5.2.  Buying Cooperative with Buying Power: 1.

In Figure 5.2, the positive influences of the enhanced efficiency gained through cooperative buying is shown by the shift to the right of the value of the marginal product curve from VMP1 to VMP2. The buying cooperative’s buying power is displayed by the presence of the marginal factor cost curve (MFC). In the absence of any buying power, the cooperative would buy at the point where the supply curve (S) intersects the VMP2 curve. The price would be w2 and the quantity Q2. If, however, the buying cooperative exerts market power, the price will be w3 and the quantity purchased will be Q3. The deceptive factor is that the outcome with w3 and Q3 is superior in terms of social welfare to w1 and Q1. But it is not as beneficial as the outcome when the buying cooperative does not exert market power. Thus, the question for antitrust purposes is whether the benefits from the increased productive efficiency are only available when the collusive monopsony makes use of its market power. The second possibility is similar but slightly more complicated. It is possible that the price and quantity will both fall in the presence of efficiency gains. This is illustrated in Figure 5.3. As in Figure 5.2, prior to the formation of the buying cooperative, price will be w1 and quantity Q1. Again, the positive influence of cooperative buying is represented by a shift of the value

The Theory of Joint Purchasing

111

Price

MFC

S

w1 w2

VMP2 VMP1 0

Q2

Q1

Quantity

Figure 5.3.  Buying Cooperative with Buying Power: 2.

of the marginal product curve from VMP1 to VMP2. If the buying cooperative also asserts monopsony power, the price will be w2 and the quantity will be Q2. The striped area in the figure represents the efficiency gains. The cross-hatched triangular area represents the losses associated with the use of monopsony power. If this striped area exceeds the cross-hatched triangular area, then society is better off as a result of the formation of the cooperative buying venture. If the deadweight social welfare loss (i.e., the triangular area) is larger than the striped area, then society is worse off as a result of the cooperative’s formation.8 In this instance, it is tempting to conclude that one should compare the losses due to the horizontal agreement with the gains in order to determine whether the agreement should be prohibited or permitted.9 Again, We have assumed that the formation of the buying cooperative does not result in market power in the sale of the buyers’ output. This was certainly the case, for example, in Topco where the Topco members had to compete at retail with many retail grocers. If such selling power were to develop, the welfare analysis would be far more complicated. 9 In weighing these costs and benefits, it is difficult to determine the distributional consequences. The enhanced productive efficiency means that society can produce the same output with fewer resources. The resources saved can be used to produce more of other goods and services, which benefits consumers generally. 8

112

Cooperative Buying Efforts

this is not necessarily the case. As in the first example, the need for comparison is only required if the benefits were unobtainable in the absence of the cooperative buying agreement. If we can get the efficiency gains without the allocative efficiency loss, that would be preferable on social welfare grounds. Accordingly, the agreement should not get an antitrust pass under these circumstances. If, however, the buying cooperative creates efficiencies and monopsony power, the trade-off analysis is necessary. Of course, it would be socially beneficial if the buying cooperative refrained from exercising its monopsony power, but that is unrealistic. Such behavior would not be profit maximizing and, therefore, economically irrational. This conundrum can be highlighted most effectively by examining some well-known examples of cooperative buying in practice. In an effort to compete with successful chain stores during the 1930s, independent retailers formed cooperative buying groups designed primarily to concentrate their buying power. Subsequently, these groups expanded the scope of their efforts on behalf of their members. These cooperatives now focus on economies in purchasing, warehousing, promotion, distribution, and the like, which do not require the exercise of monopsony power. But if the cooperative is large enough to have monopsony power, it is fanciful to suppose (or even hope) that it will not be exercised. As an antitrust policy matter, we cannot expect economic agents to behave in economically irrational ways. In some instances, these cooperatives are organized by wholesalers that can provide strong leadership. In such cases, there is some likelihood that the wholesaler will recognize and exercise the group’s collective monopsony power, which in turn, will offset some of the procompetitive effects of cooperative buying. In other instances, most prominently in the food and hardware industries, retailers organize cooperatives, and the wholesale operation is run democratically on behalf of the retailers. It is less likely that one of these cooperatives will exercise monopsony power. Each independent member of the cooperative will submit orders for the supplies that it needs for retail sales, and the manager of the wholesale operation will just fill that order. Thus, the collective quantity purchased by the group is the sum of the independently determined quantities of the membership. The procompetitive benefits of joint purchasing are not offset by the use of monopsonistic buying power. It is clear that an agreement is necessary to achieve the desirable economies, and a second agreement is needed to exploit the collective buying power. These are separable decisions and, therefore, no sacrifice in productive efficiency need be made in order to proscribe the use of buying power.

Antitrust Responses to Cooperative Buying Efforts

113

5.3 Antitrust Responses to Cooperative Buying Efforts In order to ascertain the appropriate antitrust response to joint purchasing arrangements and product standardization, it is necessary to review some of the pertinent standards in antitrust law and their recent evolution. Horizontal restraints, like many of those described in Chapters 3 and 4, are typically assessed under one of two antitrust standards. Under the rule of reason, the practice is a violation if it can be shown that the anticompetitive effects of agreement outweigh any possible procompetitive effects. Some arrangements, however, have been declared unlawful per se, thus relieving the plaintiff from showing the actual economic impact. In theory, the per se standard is reserved for practices “which because of their pernicious effect on competition and lack of redeeming virtue are conclusively presumed to be unreasonable and therefore illegal without elaborate inquiry as to the precise harm they have caused or the business excuse for their use.”10 As a general matter, the per se standard has been reserved for price fixing, horizontal market division, group boycotts, and tying arrangements.11 As cases like Mandeville Island Farms, National Macaroni Manufacturers, and Eastern States Lumber illustrate, the courts have tended to apply these per se rules without regard to whether buyers or sellers are involved in the alleged violation. Over the last thirty years, as economics has played an increasingly important role in antitrust analysis, the per se standard has become less rigorous, and courts have been increasingly willing to apply the rule of reason standard to practices that previously would have been regarded as per se unlawful. As a consequence, the “classification” process has become more difficult, and the status of particular agreements less predictable. Most important in the context of agreement among buyers is the reemergence of the “ancillary restraints doctrine.”

5.3.1  The Ancillary Restraints Doctrine The transition to a rule of reason standard has been facilitated by using the ancillary restraints doctrine, a methodology that was first articulated in the 1898 case of United States v. Addyston Pipe & Steel,12 and revived in 1979 by N. Pac. Ry. Co. v. United States, 356 U.S. at 5 (1958). In light of the events described next, this listing overstates the actual reach of the per se rules. 12 85 F. 172 (6th Cir. 1898), aff ’d, 175 U.S. 211 (1899). 10 11

114

Cooperative Buying Efforts

the Supreme Court in Broadcast Music Inc., v. Columbia Broadcasting System (BMI).13 The ancillary restraints doctrine was introduced as a way around the broad language of Section 1 of the Sherman Act that condemned “every contact … in restraint of trade.” Such language, if literally applied, would have made even the formation of a simple partnership illegal. In Addyston Pipe & Steel, Judge Taft interpreted Section 1 as exempting restraints that were ancillary to a lawful main purpose, essential to its success, and no more restrictive than necessary. The best recent example of the doctrine’s use is found in BMI, where the Supreme Court considered the practice of the American Society of Composers, Authors, and Publishers (ASCAP) and BMI of issuing blanket licenses for musical compositions and their inventories. The practice technically involved price fixing because it required competing and potentially competing composers to agree on a price for their combined offerings. Following the per se proscription of price fixing, the Court of Appeals held that the arrangement was per se unlawful. The Supreme Court reversed and held that the practice, although literally price fixing, should be evaluated under the rule of reason. In essence, the Court seemed to view the practice as one that was ancillary to an otherwise lawful joint marketing arrangement. Moreover, it was necessary in order for the product, the blanket license, to even exist. The same approach is found in NCAA v. Board of Regents of the University of Oklahoma (NCAA),14 where the Court reviewed the NCAA practices that forbade member universities from individually selling television rights to their football games. Again, the Court noted that the activity was the type that ordinarily would be “presumed unreasonable without inquiry into the particular market context in which it is found.”15 It opted for a limited rule of reason standard, however, noting that the case involved “an industry in which horizontal agreements on competition are essential if the product is to be available at all.”16 In both BMI and NCAA, the Court distinguished “per se price fixing” from mere literal price fixing. The classification process in both instances reflected Addyston Pipe & Steel in that the horizontal restraints were afforded rule of reason treatment due to the possibility that they were the least restrictive methods of obtaining otherwise lawful and potentially competitive results. 15 16 13 14

441 U.S. 1 (1979). 468 U.S. 85 (1984). Id. at 100. Id. at 101. Interestingly, when under the rule of reason, the practices of the NCAA were held to be unlawful.

Antitrust Responses to Cooperative Buying Efforts

115

It is informative to compare BMI and NCAA with Arizona v. Maricopa County Medical Society,17 where the Court considered a horizontal agreement among physicians on maximum prices. The defendants relied on BMI to justify applying the rule of reason standard for what was otherwise per se price fixing. They claimed that, by agreeing to maximum prices, insurance carriers would be able to limit and calculate the risks they underwrote with the ultimate result being “cost containment” and lower prices for consumers. The Court rejected this argument, applied the per se rule, and found that the horizontal agreement was not the least restrictive method of reaching the purported goal. The Court noted that the insurance carriers could contract with each physician individually, thereby achieving the same level of predictability.18

5.3.2  Application of the Ancillary Restraints Doctrine to Efficiency-Producing Buying Agreements Critical to the reclassification of an otherwise unlawful per se agreement is the making of a colorable argument that (1) the agreement is ultimately procompetitive, (2) the anticompetitive agreement makes the procompetitive results possible, and (3) the anticompetitive agreement is no more restrictive than necessary.19 Presumably, this policy with regard to reclassification of agreements from per se to rule of reason would apply to horizontal agreement among buyers in a straightforward fashion. Here, the proper classification of two important types of buyer agreements is analyzed: cooperative buying and product standardization. 5.3.2.1  Cooperative Buying As indicated earlier, a cooperative buying agreement may result in any one of three outcomes. Thus, application of the inflexible per se standard is probably inappropriate.20 An approach that seems to follow from the developments described earlier, in the treatment of other horizontal restraints, is 457 U.S. 332 (1982). Id. at 356–7. 19 For a recent application, see Polk Bros. v. Forest City Enterprises, 776 F. 2d 185 (7th Cir. 1985). See, generally, Martin Louis, Restraints Ancillary to Joint Ventures and Licensing Agreement:  Do Sealy and Topco Logically Survive Sylvania and Broadcast Music? 66 Va. L. Rev. 879 (1980); Jeffrey Harrison, Price Fixing, the Professions and Ancillary Restraints: Coping with Maricopa County, U. Ill. L. Rev. 925 (1982). 20 This has also been termed a “self executing” per se standard, See Lawrence Sullivan & John Wiley, Recent Antitrust Developments: Refining the Scope of Exemptions, Expanding Coverage and Refining the Rule of Reason, 27 UCLA L. Rev. 265 (1980). 17 18

116

Cooperative Buying Efforts

to regard price fixing by members of a buying cooperative as per se unlawful unless the members of the cooperative make the kind of arguments required under the ancillary restraints doctrine. Application of that doctrine suggests that the pricing practices of the cooperative should be treated as per se unlawful unless the buyers are able to offer colorable arguments that 1. The horizontal agreement is necessary to achieve some productive efficiency, 2. A price agreement is required to achieve the claimed efficiency, and 3. The exercise of monopsonistic buying power to force prices below competitive market levels is required to achieve the procompetitive efficiencies. If the defendant can satisfy these requirements, the standard of analysis would then shift to the rule of reason, which requires a balancing of the procompetitive and anticompetitive effects to determine the cooperative’s legality. The latter two requirements deserve emphasis. As for requirement 2, a great many efficiencies may result from horizontal agreements among competing buyers. When the horizontal agreement concerns price, however, the defendant must demonstrate that “but for” horizontal price uniformity, the efficiency cannot be obtained. With respect to requirement 3, it is important to recall Figures 5.2 and 5.3. In those instances, the horizontal price agreement had both procompetitive and anticompetitive effects. That is not to say, however, that the anticompetitive effects, a consequence of buying power, necessarily led to the procompetitive effects. In other words, the same efficiencies may have been obtained through a joint buying program that did not exercise its buying power to force prices down. Thus, it is up to the buyers to link the use of monopsony power to the procompetitive effects. The application of an ancillary restraints approach is either simple or likely to be unnecessary when the cooperative buying does not result in the trade-off between productive efficiency and allocative inefficiency. First, in instances where the price-fixing buyers have no colorable efficiency justification, the agreement is per se unlawful. Second, a cooperative buying venture that only enhances efficiency will not result in any private suits by complaining sellers, nor should there be any such suits.21 The reason is This is not to say there will be no suits by buyers who are excluded from the buying cooperative. See Nw. Wholesale Stationers v. Pac. Stationery & Printing Co., 472 U.S. 284 (1985).

21

Antitrust Responses to Cooperative Buying Efforts

117

straightforward: As illustrated in Figure 5.1, the quantity sold and the price received both rise. Consequently, sellers have no motivation to complain. For example, Topco, a buying cooperative, is owned by its members and achieves certain efficiencies in purchasing, product development, packaging, quality control, and so on. The result is that the Topco members are more effective competitors in the retail grocery trade. Consequently, sales to Topco members have expanded to the benefit of the members and of their suppliers. Everyone gains; there are no anticompetitive effects and no complaints regarding the pricing agreement. Finally, there is the situation in which there are procompetitive efficiencies of joint buying as well as the anticompetitive exercise of monopsony power. In that event, two cases can be distinguished on the basis of whether the exercise of buying power is inextricably intertwined with the achievement of procompetitive efficiencies. If the exercise of buying power cannot be divorced from the horizontal integration that yields the procompetitive efficiencies, then the cooperative venture should be accorded rule of reason treatment; otherwise, it is per se unlawful. 5.3.2.2  Product Standardization A second intriguing possibility arises when sellers agree to standardize their products. Standardization can take a great variety of forms ranging from specifying product design in order to ensure that all number two flat head wood screws are alike to requiring that all graduates of accredited law schools have had similar training. From the perspective of monopsony, a difficult problem arises when the buyers agree on the actual composition of their product. For example, in National Macaroni Manufacturers, the defendants agreed on the percentage of various types of wheat to use in the production of macaroni. Although that agreement seems unlikely to have any procompetitive effects, there are instances in which standardization may result in substantial efficiencies. Product standardization can be an efficient method of communicating that reduces the costs to consumers of searching for and evaluating products.22 For example, an industry-wide agreement among toy manufacturers not to use paint containing toxic substances or that child care workers must meet certain educational requirements before they are hired can reduce the search and evaluation costs incurred by parents. The problem with this type of standardization, See Herbert Hovenkamp, Economics and Federal Antitrust Law 101 (St. Paul, Minn.: West Publishing 1985).

22

118

Cooperative Buying Efforts

however, is that it can be seen as a horizontal agreement to restrict the use of certain inputs similar to that in National Macaroni Manufacturers. The classification of a horizontal agreement not to purchase is made especially difficult because an agreement to use none of a particular input, without more, hardly benefits the cooperative. The agreement may reduce the demand and the price for the disfavored input, but the agreeing parties are not favored. Thus, unless the agreement is part of a scheme designed to force the sellers onto an all-or-none supply curve, the motivation seems unlikely to be anticompetitive. Moreover, the same action may increase the demand for the favored input with the possibility that price will increase. Indeed the market power of that input seller vis-à-vis the buyers may increase. This would seem to warrant cutting a wide path for such agreements. On the other hand, as National Macaroni Manufacturers and the all-or-none supply analysis suggest, such agreements can be anticompetitive. Moreover, the standardization can reduce the choices available to consumers.23 Finally, input standardization is a step toward production cost uniformity and, consequently, output price uniformity. This, of course, poses no problem if the firms are competing and not fixing prices since the uniform prices will be competitive. In spite of the inherent dangers of horizontal restraints and the likelihood that the agreeing buyers are in the best position to offer a procompetitive justification, the burden of proving anticompetitive effects should fall on the plaintiff. Application of the ancillary restraints doctrine suggests that the practice be evaluated under the rule of reason if the defendants can make a colorable argument that the horizontal agreement restricting purchases is necessary to achieve some kind of efficiency. The focus of this initial inquiry is both on the assertion that there is some advantage for consumers of product standardization and that such standardization would not come about through the evolution of free-market forces. For example, the case would have to be made that the standardized product was not simply a superior product that all producers would eventually have to produce in order to remain competitive. This suggests that an element of the defendants’ justification would have to be that, in the absence of an agreement, the beneficial result would be undermined by “free-riders.” This, too, can be good or bad. Ordinarily, the presumption is that more choice is better than less choice. Presumably, however, parents will not be dismayed at being unable to buy toys for their children that are coated with lead-based paint.

23

Antitrust Responses to Cooperative Buying Efforts

119

5.3.2.3  Northwest Wholesale Stationers and U.S. Department of Justice Guidelines Further guidance and support for the approaches suggested here can be found in the Supreme Court’s reasoning in its most important case involving a buying cooperative. In Northwest Wholesale Stationers, Inc. v. Pacific Stationery and Printing Co.,24 the Court examined the expulsion of a cooperative buying group member. The Court adopted the view that the expulsion was a type of boycott not subject to the per se standard. In particular, the Court noted the possible procompetitive effects of buying cooperatives and reasoned that, without market power or exclusive access to a necessary input, the expulsion was unlikely to be anticompetitive. Although the Court applied a slightly looser standard than the relaxed per se treatment proposed here, the application of the ancillary restraints analysis to price fixing and product standardization by buyers seems consistent with Northwest Wholesale Stationers. As the Court specifically noted, Northwest Wholesale Stationers did not involve a challenge to the pricing practices of the cooperative. Moreover, the Court specifically held open the possibility that the most dangerous forms of boycotts, as in the case of price fixing, are still per se unlawful.25 The different approach where a party challenges its exclusion from a cooperative, as opposed to the actual practice of horizontally fixing prices, can be traced to the fact that the Court’s view of group boycotts, unlike its view of price fixing, has never been completely clear. This lack of clarity can be explained by the varying forms and purposes of boycotts.26 Finally, the approach outlined here reflects the policy of the Antitrust Division of the Department of Justice as announced in its 1997 Merger Guidelines.27 The comparison of cooperative buying in particular to mergers is especially appropriate since the decision to engage in cooperative buying or product standardization is a contractual or functional form of merger. In the Merger Guidelines, the Department of Justice describes the circumstances under which it will consider possible efficiencies resulting 472 U.S. 284 (1985). Id. at 294. See also, Superior Court Trial Lawyers v. FTC, 493 U.S. 411 (1990). 26 472 U.S. at 298, 294. See, generally, L. Sullivan, Antitrust 229 (St. Paul, Minn.:  West Publishing 1977); E. Thomas Sullivan & Jeffrey L. Harrison, Understanding Antitrust and Its Economic Implications, 5th ed. 107 (New York: LexisNexis 2009). See also the Court’s discussion in St. Paul Fire and Marine Insurance Co. v. Barry, 438 U.S. 531 (1978). 27 Revised Section 4 Horizontal Merger Guidelines Issued by the U.S. Department of Justice and the Federal Trade Commission (April 8, 1997), http://www.usdoj.gov/atr/public/ guidelines/horiz_book/4.html. 24 25

120

Cooperative Buying Efforts

from a merger in making a decision about whether to challenge the merger. Departmental policy with respect to the possibility of an “efficiency defense,” as reflected in the Guidelines, is that the merging parties: [M]ust substantiate efficiency claims so that the Agency can verify by reasonable means the likelihood and magnitude of each asserted efficiency, how and when each would be achieved (and any costs of doing so), how each would enhance the merged firm’s ability and incentive to compete, and why each would be mergerspecific. Efficiency claims will not be considered if they are vague or speculative or otherwise cannot be verified by reasonable means.28

In addition, the Department of Justice will “consider only those efficiencies likely to be accomplished with the proposed merger and unlikely to be accomplished in the absence of either the proposed merger or another means having comparable anticompetitive effects …” and reject any efficiency justification if the same efficiencies can “reasonably be achieved by the parties through other means.”29 Certainly, the policy has a clear ancillary restraints flavor in that possible efficiencies are only relevant if unattainable with less-restrictive but realistic alternatives. In theory, at least, there should be little difference in the eventual outcome under the ancillary restraints approach as applied here and a rule of reason treatment. The chief advantage of the ancillary restraints approach is that it requires the party with superior knowledge to argue persuasively the existence of beneficial effects and that a horizontal agreement on price or inputs is necessary to achieve those effects. Of course, theory rarely conforms to reality, and burden shifting, no matter how slight, will affect judicial outcomes and ultimately business practice. In the context of cooperative buying and input restrictions, the risk is that procompetitive cooperative efforts will be discouraged. This risk, of course, applies to the use of any per se rule regarding business practices. Obviously, the question of whether the risks of overdeterrence under the ancillary restraints approach will offset the advantages it offers cannot be answered with certainty, but the danger is slight. First, it is not difficult for buyers involved in procompetitive activities to escape per se treatment under the approach outlined here. Second, horizontal price agreements are inherently Id. Id. For a discussion of the “efficiency defense” in merger cases generally, see Alan A. Fisher & Robert Lande, Efficiency Considerations in Merger Enforcement, 71 Cal. L. Rev. 1580 (1983); Herbert Hovenkamp, Merger Actions for Damages, 35 Hastings L. J. 937 (1984); Robert Pitofsky, Efficiencies in Defense of Mergers: Two Years Out, 7 Geo. Mason L. Rev. 485 (1999); Malcolm B. Coates, Efficiencies in Merger Analysis: An Institutional View, 13 Supreme Ct. Econ. Rev. 189 (2005).

28 29

Antitrust Responses to Cooperative Buying Efforts

121

dangerous whether the parties are buyers or sellers. Third, the cooperating buyers are apt to be competing sellers, and an agreement on input prices or amounts can be a convenient step toward agreement on output prices. 5.3.2.4  A Structural Dilemma Before concluding this discussion of the antitrust treatment of horizontal agreement among buyers, it is important to note a structural dilemma that is encountered in much of antitrust. Consider again the case of Topco, a democratically run wholesale cooperative.30 Topco’s decisions are the collective decisions of its members, which are its owners. For example, assume Topco provides private label products, an efficient procompetitive activity that should be encouraged. Now, one or more Topco members might recognize that by agreeing among themselves to reduce the quantities they collectively purchase, they could pay lower prices for some items. But that exercise of collusive monopsony power is not inextricably intertwined with those activities necessary to achieve the procompetitive efficiencies. As a result, the collusion on price would not be accorded rule of reason treatment; rather, it would be treated as a per se violation of the Sherman Act. In contrast, consider a wholesaler-organized cooperative such as the Independent Grocers Alliance (IGA).31 IGA signs up independent grocers to be members of its cooperative by providing the same services that Topco provides to its members, but IGA is not democratically run. Instead, IGA provides leadership and drives the organization. Due to IGA’s success, it prospers along with its members as they become increasingly effective competitors for the major chains. Because IGA accounts for a large share of the total purchases of particular items, it may begin to exercise its latent monopsony power. This exercise also is not inextricably intertwined with its procompetitive activities, but it is not a result of collusion. This puts one in the uncomfortable position of concluding that this use of monopsony power is not a violation of the Sherman Act. It is presumptively lawful in the same sense that a large seller’s use of selling power is not an antitrust violation. This dilemma, which is a common problem in determining antitrust policy, arises because the economic effects are the same in the two cases.32 See L. Stern & A. El-Ansary, Marketing Channels 329 (Englewood Cliffs, N.J.: Prentice Hall 1984). 31 Id. at 327. 32 This problem seems to be at the heart of judicial uncertainty about the proper interpretation of Section 2 of the Sherman Act. See, e.g., United States v. Aluminum Corp. of America, 148 F.2d 416 (1945). The Supreme Court has noted this problem. See Copperweld Corp. v. Independence Tube Corp., 467 U.S. 752, 774–5 (1984). 30

122

Cooperative Buying Efforts

The result is the drawing of a distinction in the absence of an economically meaningful difference. Moreover, it raises the possibility that clever businessmen may avoid prosecution simply by organizing their operations to disguise collusion. An important issue, therefore, is whether the standard proposed here would encourage structural change toward relatively immune “single entity” buying organizations. But this outcome is an unlikely consequence. Any effort by a group of buyers to form a “separate” entity would likely be defined as a “contract” under Section 1 of the Sherman Act. The new entity, therefore, would not escape antitrust scrutiny. Consequently, the reorganization would not result in per se legality for the buying efforts. At most, the organizers could escape per se treatment in favor of rule of reason treatment by making a colorable argument that procompetitive effects were only obtainable through their reorganization “contract.”

5.4  Concluding Remarks There is no doubt that cooperative buying can have socially beneficial results. In many instances, cooperation leads to gains in productive efficiency that redound to the benefit of consumers in the form of lower prices. But the cooperative buying can also result in collusive monopsony, which has deleterious consequences for social welfare. One focus of this chapter has been to provide an economic analysis of these two possible outcomes. In addition, we have provided a test to determine which of the two outcomes has materialized. Given the possibilities of either procompetitive or anticompetitive welfare effects, a per se standard is clearly inappropriate. Accordingly, the courts should employ a rule of reason analysis in dealing with cooperative buying. We have proposed a three-prong test for reasonableness that should assist in a rule of reason inquiry.

SI X

Bilateral Monopoly

6.1 Introduction From time to time in antitrust law, there has been either an express or an implicit recognition of what might be termed a “countervailing market power defense.”1 This is somewhat anomalous since such a defense is a version of the argument, rejected early in Sherman Act jurisprudence, that price fixing resulting in “reasonable” prices should be regarded as lawful.2 In the context of monopsony, the possibility of a countervailing market power defense requires a closer look at the phenomenon of bilateral monopoly. The issue, in its most pristine form, is whether collusion among buyers would be justified when they are faced by a monopolist on the selling side of the market. Alternatively, the same issue could arise in the context of a collusive monopoly facing a monopsonist.3 In theory, a full-blown version of such a defense would permit collusion designed to equal but not exceed the power of parties on the opposite side of the market even when The idea of countervailing power was advanced most persuasively by John K. Galbraith, American Capitalism: The Concept of Countervailing Power (Boston: Houghton Mifflin 1952). 2 An eloquent rejection of reasonableness as a defense in price-fixing cases is provided by Judge Taft’s opinion in United States v. Addyston Pipe & Steel Co., 85 F. 271 (6th Cir. 1898): 1



[I]t has been earnestly pressed upon us that the prices at which the cast-iron pipe was sold … were reasonable … . We do not think the issue an important one, because … we do not think that at common law there is any question of reasonableness open to the courts with reference to such a contract. Its tendency was certainly to give defendants the power to charge unreasonable prices … 

  See also R.C. Dick Geothermal Corp. v. Thermogenics, Inc., 890 F.2d 139, 167 (9th Cir. 1989) (dissent); Superior Court Trial Lawyers Assn. v. F.T.C. 855 F.2d 226 (D.C. Cir. 1988) reversed F.T.C. v. Superior Court Trial Lawyers Association, 110 S. Ct. 708 (1990). 3 This case is examined extensively in Chapter 10.

123

124

Bilateral Monopoly

that power fell short of monopolistic or monopsonistic levels. This chapter begins with three illustrations of instances in which antitrust law has been shaped by recognition of countervailing market power. Next, the theory of bilateral monopoly is examined. The analysis indicates that under special circumstances collusive monopsony (or monopoly) may actually lead to greater economic efficiency. The possibility of effectively recognizing this theoretical possibility in the application of the antitrust laws is then explored.

6.2  Countervailing Market Power in Antitrust Some measure of a countervailing market power rationale can be found in the 1933 case, Appalachian Coals v. United States,4 in which 137 producers of bituminous coal combined to create an exclusive joint selling agency. The central objective of the agency was to sell the coal at the “best price obtainable.”5 One focus of the joint effort was to control the supply of socalled “distress coal” – a byproduct of the production of the more desirable sizes and shapes of coal. The Supreme Court reversed a lower court holding that the actions of the producers violated Sections 1 and 2 of the Sherman Act. In what was obviously a Depression Era opinion, the Court started with the premise that “[t]he mere fact that the parties to an agreement eliminate competition between themselves is not enough to condemn it.”6 The Court noted both the “deplorable” conditions of the industry and the presence of “organized buying agencies and large customers” possessing “concentrated buying power” and creating a “buyers’ market.”7 Although hardly resting its opinion on any express application of a countervailing market power theory, industry conditions and particularly the imbalance of market power clearly influenced the Court’s decision. More on point in the context of this study on monopsony is the exemption afforded multiemployer bargaining units. This exemption seems to have evolved from the express statutory exemption from the antitrust laws provided for labor unions in the Clayton Act and the Norris-Laguardia Act. This statutory exemption would be rather meaningless if it were interpreted narrowly so as not to encompass the collective bargaining agreements 288 U.S. 344 (1933). George J. Stigler, The Economic Effects of the Antitrust Laws, J. L. & Econ. 9 (1966), pointed out that using a joint sales agency is a very efficient form of collusion because cheating is so difficult. 6 Supra note 4. 7 Id. 4 5

Countervailing Market Power in Antitrust

125

between unions and nonlabor entities. Thus, there is a nonstatutory exemption for the agreement, subject to certain limitations, between the union and management.8 Furthermore, this exemption now clearly extends to “joint employer preparation and bargaining in the context of a formal multi-employer bargaining unit.”9 Although the exemption is derived from the statutory antitrust exemption for unions, it now seems to have an as yet undefined independent basis since it has been held that the exemption continues for employers even when the union and management have reached an “impasse”.10 The judicial rationale for this extension is largely based on the fact that efforts to amend the Taft-Hartley Act to outlaw multiemployer bargaining were unsuccessful.11 There is a practical basis for the exemption in that it probably lowers the transaction costs of reaching agreements in industries in which there are a great number of employers. In addition, there is a distributive goal associated with evening out the balance of bargaining or market power between the union and management. A third illustration of the recognition of countervailing market power can be found in Balmoral Cinema v. Allied Artists Pictures Corp.,12 a case dealing with the commonly analyzed matter of motion picture “split” agreements. The split agreement involves motion picture exhibitors who agree not to compete against each other for the rights to exhibit specific films. In fact, they “split” the market with one exhibitor having the exclusive rights to “bid” on a particular film while its partner in the split has the right to bid on another film. From the standpoint of monopsony, such an understanding confers buying power on the exhibitor with the exclusive bidding rights. In Balmoral, the United States Court of Appeals for the Sixth Circuit held that such an agreement was not per se unlawful. In so holding, the court expressly relied on a monopsony-based countervailing market power rationale. It reasoned that “[e]xhibitors, as purchasers of films, may be justified on combating the market power of film suppliers by group action. Such action may lower prices to moviegoers at the box office and serve rather than undermine consumer welfare.”13 This reasoning is generally wrong, but may not be in the presence of monopoly power. See Connell Constr. Co. v. Plumbers & Steamfitters Local No. 100, 421 U.S. 616, 622 (1973). Phillip Areeda & Herbert Hovenkamp, Antitrust Law (Boston:  Little, Brown, 1991 supplement). 10 See Powell v. NFL, 888 F.2d 559 (8th Cir. 1989). 11 For a discussion, see NLRB v. Truck Drivers Local Union 449, 353 U.S. 87 (1957). 12 885 F. 2d 313 (6th Cir. 1989). 13 Id. at 316–17. 8 9

126

Bilateral Monopoly

The entire issue of collusive monopsony can also be viewed in light of the general treatment of buying cooperatives. As discussed in Chapter 5, a buying cooperative that exercises market power can create the types of inefficiencies about which the antitrust laws are concerned. In fact, the Supreme Court recognized this possibility in Northwest Wholesale Stationers v. Pacific Stationery and Printing14 in which it listed market power as one of the determinants of whether a buying cooperative would be held to close antitrust scrutiny. The more difficult question is whether a buying cooperative that possesses and uses market power should be able to use the existence of market power on the selling side of the market as a defense. The theory of bilateral monopoly adds to our understanding here.

6.3  The Bilateral Monopoly Muddle A bilateral monopoly exists when a single supplier of a well-defined product has only one customer. In this setting, the single customer is a monopsonist, whereas the single supplier is a monopolist. The monopolist’s effort to maximize profit by restricting output and raising price is confronted by the monopsonist’s effort to maximize profit by restricting purchases and lowering price. As a consequence, there is something of a problem in identifying the equilibrium price and quantity. This problem is exemplified by the erroneous treatment that was found in most intermediate textbooks.15 First, consider the usual monopoly problem. A single seller with marginal cost, labeled MC in Figure 6.1, faces a large number of buyers with no market power and a negatively sloped demand curve, labeled D. The marginal revenue associated with the demand curve is denoted by MR. Profit maximization will lead the monopolist to sell x1 units of output at a price of w1. Next, consider the usual monopsony problem. A single buyer faces a large number of sellers with no market power and a positively sloped supply curve, which is labeled MCx in Figure 6.1. The marginal factor cost associated with the supply is denoted as MFC. Ordinary profit-maximizing behavior of the monopsonist will lead to purchases of x2 units of output and a price of w2. The analytical muddle arises when one combines these two results and attempts to infer the bilateral monopoly solution. In attempting to deal with the bilateral monopoly market structure, many analysts alternately assume that one trading partner and then the other behaves as a perfect competitor would behave. If the input producer were 472 U.S. 284 (1985). See Roger D. Blair, David L. Kaserman, & Richard E. Romano, A Pedagogical Treatment of Bilateral Monopoly, 55 S. Econ. J. 831 (1989).

14 15

The Correct Solution

127

Price

MFC w1 MC

w2

D MR x1

0

x2

Quantity

Figure 6.1.  Bilateral Monopoly:1.

to behave competitively, then its supply curve would correspond to MC. In this case, the final good producer would arrive at the standard monopsony solution, buying x2 units of the intermediate product where marginal factor cost equals demand and paying w2 per unit. In contrast, if the final good producer were to act as a perfect competitor in its hiring decision, then the input producer would exercise its monopoly power in supplying the input. In this case, we have the input monopoly solution at x1 and w1, which is determined by the equality of MR and MC. Some have concluded erroneously that these two outcomes set the bounds on the equilibrium price-quantity combination, which is indeterminate because it is reached by bargaining. This exercise, however, is counterintuitive. If a firm is a monopolist or a monopsonist, there is no reason why it should behave competitively. Consequently, the results that follow from an assumption of competitive behavior on one side of the market are useless in analyzing bilateral monopoly.

6.4  The Correct Solution The theory of bilateral monopoly has a rich history that can be traced to the writings of Cournot and Menger.16 Over the 150 years or so that this Although Augustin Cournot, Recherches sur les Principes Mathematiques de la Theirie des Richesses (1838) (Nathaniel T. Bacon provided the English translation: Researches into the

16

128

Bilateral Monopoly

problem has been under consideration, economists have struggled to identify the correct solution. In spite of the fact that Bowley17 offered the theoretically correct solution in 1928, some problems persist because the usual marginal analysis fails us in these circumstances. Bowley recognized that there is a profit motive for cooperation to emerge between the trading partners under conditions of bilateral monopoly. This profit incentive arises because joint profits are not maximized at either of the two solutions described in the preceding section. Such cooperation may take the extreme form of vertical integration,18 or it may come about through the bargaining process. For the latter, it is important to realize that the negotiation that takes place must involve both price and quantity if joint profits are to be maximized. Machlup and Taber made this point and speculated that a failure to recognize this essential difference between bilateral monopoly and all other market structures created much of the analytical confusion.19

6.4.1  A Simple Model of Bilateral Monopoly Suppose that the final good (Q) is produced according to a production function:

Q = Q( x , y )

(6.1)

where x is the intermediate good traded under bilateral monopoly conditions and y is another input that is supplied competitively at a constant price of py. The total cost of producing the intermediate good x is given by C(x). If the intermediate good monopolist and its sole customer were vertically integrated, the profit function of the integrated firm would be

Π = P(Q)Q(x,y) − C( x ) − p y y



(6.2)

Mathematical Principles of the Theory of Wealth (1927)) and Carl Menger, Grundsätze der Volkwirtschaftslehre (1871), did not consider bilateral monopoly explicitly, they did treat the closely related topic of isolated exchange between two individuals. For an excellent examination of the historical development of the theory of bilateral monopoly, see Fritz Machlup & Martha Taber, Bilateral Monopoly, Successive Monopoly, and Vertical Integration, 27 Economica 101 (1960). 17 For an interesting development of the correct analysis, consult A. L. Bowley, Bilateral Monopoly, 25 Econ. J. 651 (1928). 18 The merger solution was mentioned by Milton Friedman, Price Theory 191 (Chicago:  Aldine Publishing 1976) and by George J. Stigler, The Theory of Price, 4th ed. 216 (New York: Macmillan 1987). In addition, see Henrick Horn & Asher Wolinsky, Bilateral Monopoly and Incentives for Merger, 19 RAND J. Econ. 408 (1988). 19 See Machlup & Taber, supra note 16.

The Correct Solution

129

where P = P(Q) is the final output demand function. The firm will maximize its profits by selecting quantities of x and y such that the first partial derivatives of (6.2) vanish:

and

∂Π dP ∂Q dC = (P + Q ) − =0 ∂x dQ ∂x dx

(6.3)

∂Π dP ∂Q = (P + Q ) − py = 0 (6.4) ∂y dQ ∂y In other words, integrated profits are maximized where the marginal revenue products of the inputs equal their respective marginal costs. It should be noted that this is the only level at which the joint profits are maximized. Now suppose that the bilateral monopolists had not vertically integrated. These firms will conduct arms-length negotiations on the price of x(w) and the quantity of x. But these firms are not assumed to be myopic. Bowley pointed out that under these conditions, the negotiation process will inexorably lead to precisely the same employment decisions as those that a vertically integrated firm would adopt. The same joint profit-maximizing quantities of inputs x and y will be exchanged. This, of course, will lead to the same quantity of the final good. Thus, a determinate quantity of the intermediate good will be transacted. Moreover, the price and quantity of the final good will also be determinate. The only thing that remains indeterminate is the price of x, but this is not allocatively significant because this price does not act as a rationing device. The price of x is just a means of dividing the jointly maximized profit. To see these results graphically, we employ a special case of the model presented earlier. Suppose that one unit of an intermediate good (x) is transformed into one unit of final good (Q). The demand for the final good is downward sloping, and the demand for the intermediate good is derived from the demand in the final good market. Assuming that the average cost of transforming (CT) one unit of input x into one unit of output Q is constant, we represent the average net revenue as a function of the quantity of x employed and label it DQ – CT in Figure 6.2. With monopoly in the sale of Q, the hypothetical derived demand for x is the curve marginal to DQ x– CT, which is labeled Dx in Figure 6.2. In other words, Dx is the net marginal revenue product of the input x.20 Now, MRx is the curve marginal to Dx and represents the marginal revenue associated with selling this Since the producer of the final output has monopsony power, Dx is not the derived demand curve because a monopsonist technically does not have a demand curve. Since this curve

20

Bilateral Monopoly

130 Price & Cost

MFCx MCx wu3 w1 ACx w2 wd3 DQ - CT 0

x3

x1 x2

MRx

Quantity MRPx = Dx

Figure 6.2.  Bilateral Monopoly: 2.

intermediate good to a final good producer that has monopoly power, but no monopsony power. Turning to the cost curves, ACx denotes the upstream monopolist’s average cost of producing input x, and MCx is marginal cost. If the supplier of x were to behave as a pure competitor, MCx would correspond to its supply curve.21 Then, if the downstream monopsonist were hiring this input from such a competitor, MFCx would be the marginal factor cost of the input. From equation (6.3), we know that the vertically integrated firm would maximize its profits at a quantity of x3 where the marginal cost of producing x (MCx) equals the marginal revenue product of x (MRPx). In essence, the sum of producer and consumer surplus is maximized at that quantity. In the absence of vertical integration (i.e., in a case of bilateral monopoly), the quantity x3 will maximize the joint profits. One would expect the parties to agree on this quantity of the input. The price of the input is indeterminate. It should be noted that price is not a rationing device under conditions of represents the same thing, we follow Friedman, supra note 18, at 190, and refer to this as the “hypothetical demand curve.” 21 Since the producer of the intermediate good x is a monopolist, it does not have a supply curve, but we can think of this as a hypothetical supply curve.

A Formula Price Contract Solution

131

bilateral monopoly. Instead, it is just a means of dividing the profit between the two parties. There are limits on the indeterminacy of the input price. In Figure 6.2, these are shown as wu3 and wd3. If the supplier completely dominates the bargaining process, it will extract all of the profit. The price will be wu3, and the buyer will have zero profits. In contrast, if the buyer completely dominates the bargaining process, it will extract all of the profit by driving the price down to wd3. These, of course, are the two extremes: The price cannot be higher than wu3 nor can it be lower than wd3; otherwise someone’s profits will be negative.22 If neither party can completely dominate the bargaining process, the actual price will lie between wu3 and wd3 and the parties will each have some positive share of the profit.23

6.5 A Formula Price Contract Solution24 As we saw in the preceding section, bilateral monopoly creates a need for repeated, protracted, and costly negotiations over the price and quantity of the intermediate product. Under these circumstances, vertical integration is attractive because it eliminates these costs and, therefore, increases the combined profits of the two firms. Although these enhanced profits provide a powerful incentive for vertical integration, ownership integration may not be an attractive alternative for reasons that we examine here.

6.5.1  Difficulties with Vertical Integration Ownership integration across two successive stages of production is not without cost.25 First, the integrated operation may experience managerial A more complete derivation of these outcomes is contained in the Appendix to this chapter. 23 In Figure 6.2, note that the monopolist is worse off at x3 and wu3 than it would be at x2 and w2. Similarly, the monopolist is worse off at x3 and wd3 than it would be at x1 and w1. As a consequence, it is natural to ask why either firm would be interested in a move to x3 if such a move drives profits to zero. But neither firm will behave competitively and, therefore, neither (x2, w2) nor (x1, w1) is a feasible outcome. As a result, the pure monopoly and pure monopsony solutions are not in the feasible set. Neither of these firms has these solutions as an option. 24 This section digresses from the primary focus of whether there should be a “bilateral monopoly” defense. The reader may skip to Section 6.6 in order to rejoin the primary theme of the chapter. This section depends heavily on Roger D. Blair & David L. Kaserman, A Note on Bilateral Monopoly and Formula Price Contracts, 77 Am. Econ. Rev. 460 (1987). 25 Ronald H. Coase, The Nature of the Firm, 4 Economica 386 (1937), examined this in some detail. This classic paper continues to be worth reading for its insights. 22

132

Bilateral Monopoly

diseconomies as spans of control are extended in the face of bounded rationality on the part of decision makers.26 Second, capital costs may increase if lenders and investors cannot be convinced that sufficient expertise exists for the efficient operation of both stages simultaneously.27 Finally, even in the absence of the preceding diseconomies, successful integration must entail considerable negotiation over the price of the acquired firm and the various ancillary terms that accompany the usual merger agreement. Consequently, it is not surprising that firms confronted with the problem of bilateral monopoly might seek contractual alternatives to ownership integration.

6.5.2  Vertical Integration by Contract The economic literature on contractual forms of vertical integration has tended to focus primarily on the variable proportions model in which an upstream monopolist sells its output to a competitive downstream industry.28 As a result, this literature has not satisfactorily treated the case in which bilateral monopoly exists. Historically, the only contractual alternatives that were found in the literature were either (1) a long-term contract or (2) a series of short-term contracts that specify both the price and the quantity of the intermediate good to be exchanged in the case of bilateral monopoly. But these particular contractual arrangements are distinctly unappealing alternatives to ownership integration for two reasons. First, a long-term contract on both price and quantity fails to allow the necessary adjustments to occur when changes in cost or final output demand are experienced. Second, a series of short-term contracts requires repeated negotiations in the face of small numbers bargaining and clear opportunistic incentives. As a result of these various problems, economists have tended to conclude that ownership integration is typically the preferred solution in most cases of bilateral monopoly. At the same time, various authors have recognized that the proximity of the potential transactional relationship that exists between the producer and the user of an intermediate product lies on a continuum, with pure spot market exchange and common ownership providing the extremes and the For a useful analysis, consult Oliver E. Williamson, Markets and Hierarchies:  Some Elementary Considerations, 63 Am. Econ. Rev. 316 (1973). 27 On this point, see Oliver E. Williamson, The Economics of Antitrust:  Transaction Cost Considerations, 122 U. Pa. L. Rev. 1439 (1974). 28 For an extensive (albeit somewhat dated) survey of this literature, see Roger D. Blair & David L. Kaserman, Law and Economics of Vertical Integration and Control (New York: Academic Press 1983). In addition, see Chapters 14–21 in Roger D. Blair & David L. Kaserman, Antitrust Economics, 2d ed. (New York: Oxford University Press 2009). 26

A Formula Price Contract Solution

133

myriad contractual and noncontractual (or relational) agreements falling somewhere in the middle.29 Blair and Kaserman have proposed a formula price contract with several desirable properties. First, it leads each firm to pursue independent profit-maximizing policies that will result in maximum joint profits. Second, it allocates a specified share of these maximized joint profits to each party to the contract. Finally, it automatically adjusts to changes in production costs and final output demand.

6.5.3  Formula Price Contracts The purpose of the formula price is to share the maximized profit between the monopolist and the monopsonist. Suppose that each unit of the final product (Q) requires one unit of the intermediate good (X), which is subject to the bilateral monopoly. In the absence of vertical integration, the upstream monopolist’s profit function (ΠU) is Π U = w(X)X − C x (X)X



(6.5)

where w(X) is the price of the intermediate good and Cx(X) is the average cost of producing X. The downstream monopolist’s profit function will be Π D = P(Q)Q − w(X)x − C T (Q)Q



(6.6)

where P(Q) is the price of the final good and CT(Q) is the average cost of transforming one unit of X into one unit of Q. It is useful to examine the profit function if the firms were vertically integrated. If these two firms were vertically integrated, the profit function of the combined operation would be Π I = P(Q)Q − C x (X)X − C T (Q)Q



(6.7)

Absent vertical integration, the firms would like to sign a long-term agreement that will generate combined profits equal to the maximized value equation (6.7), which we denote as Π*I, and assign shares of these profits equal to α and 1 – α to the upstream firm and downstream firm, respectively, where 0 ≤ α ≤ 1. Obviously, such an assignment of profit shares must occur through the price of the intermediate product that is determined by the contract. For important contributions to this recognition, see Benjamin Klein, Robert G. Crawford, & Armen A. Alchian, Vertical Integration, Appropriable Rents, and the Competitive Contracting Process, 21 J. L. & Econ. 297 (1978).

29

134

Bilateral Monopoly

Setting ΠU= αΠI and solving for w, we obtain

w = α(P − C T ) + (1 − α)C x

(6.8)

Thus, if the upstream monopolist can ensure that the price of the intermediate product is determined by final output price and production costs at both stages through the formula given in equation (6.8), it will be assured of receiving α of the profits available to a vertically integrated monopolist, We can rearrange (6.8) as

w = C x + α(P − C T − C x )

(6.9)

Thus, the formula price for the intermediate product is equal to the average cost of producing the product plus α times the optimal integrated monopoly mark-up over cost at the final good stage. By substituting equation (6.8) into equation (6.6), we find that Π D = PQ − [α(P − C T ) + (1 − α)C x ]X − C TQ

= (1 − α)Π I

(6.10)

That is, the formula price contract described in equation (6.8) also automatically assigns 1– α of the integrated profits to the downstream firm. Therefore, under the terms of the formula price contract of equation (6.8), ΠU = αΠI and ΠD = (1– α)ΠI. As a result, independent profit maximization by the two nonintegrated firms will lead to combined profits of Π*I. The bilateral monopoly contract negotiation problems are greatly reduced because such negotiation can focus entirely on the value of α. All changes in demand and cost are accommodated automatically in the formula. Thus, the formula price contract provides an attractive alternative to ownership integration in the presence of bilateral monopoly.

6.5.4  Performance Characteristics The formula price contract appears to provide a reasonably close substitute for vertical integration. It does this by largely reconciling the transaction cost differences between long-term and short-term contracts. That is, it allows a degree of flexibility that is equal to or greater than that obtained with a series of short-term contracts while still defining long-term supply and demand obligations that do not require periodic renegotiation. Such contracts exhibit three desirable performance characteristics. First, they facilitate the negotiation process by focusing attention on a single parameter, α. With each firm automatically driven to produce the joint

A Formula Price Contract Solution

135

profit-maximizing level of output, there is no need to specify the price and output of the intermediate product. The parties to the contract need only settle on mutually agreeable shares of the resulting maximized profits. At least as much negotiation would be required for vertical integration to occur. Second, the formula price contract further facilitates negotiation by economizing on the information needed for contract specification. Again, because of the incentive structure that automatically leads firms to produce the joint profit-maximizing output, the parties to the contract do not require information on final product demand in order to specify the terms of the contract. That is, the contract can be negotiated without specific knowledge of P(Q). In this one respect, the transaction costs associated with contractual integration would appear to be less than the transaction cost involved in achieving ownership integration. For the latter to occur, the parties must be able to agree on a purchase price of one of the firms. Since this purchase price incorporates the capitalized value of the profit stream of the acquired firm, successful negotiation requires information on the final output demand not only at the time the sale is negotiated but in all future periods as well. With impacted information and clear incentives for strategic misrepresentation by both parties, such negotiation is likely to be costly or, in the extreme, impossible. Thus, from the point of view of negotiation costs alone, firms may be encouraged to select the formula price contract over vertical ownership integration. Third, once the formula price contract is in effect, any changes in final output demand or in production costs at either stage will be reflected appropriately in the new optimal prices and outputs fostered by the contract. That is, the contract will automatically accommodate changes in P(Q), CT(Q), or Cx(Q), encouraging the parties to the contract to adjust to the new Π*I solution. Moreover, the firms’ shares in this new Π*I will remain at the agreed on values of α and 1 – α. No renegotiation is required as a result of dynamic changes in final product demand or in costs.30 In this respect, the formula price contract appears to operate in a fashion that is identical to the ownership integration alternative.31 This is not to say that renegotiation of the profit shares themselves will never be necessary, but that fluctuations in cost and demand conditions, which may occur frequently, do not require any renegotiation. 31 It is a sad fact of life that nothing seems to be perfect. The problem with the formula price contract is that the monopolist has a postcontractual incentive to misrepresent Cx(X), whereas the monopsonist has an incentive to misrepresent CT(Q). Since strategic misrepresentation can increase profit at the expense of the other firm, there will have to be some means of auditing claims of cost changes, which will be reflected in the transaction cost. This issue is examined analytically by Blair & Kaserman, supra note 24. 30

136

Bilateral Monopoly

6.6 Limitations of Countervailing Power Considerations From the standpoint of social welfare, the structural condition of bilateral monopoly is preferable to either a monopoly seller dealing with competitive buyers or a monopsony dealing with competitive suppliers. If rival buyers join forces in response to monopoly or rival sellers collaborate in response to monopsony, the quantity sold increases, which will increase the quantity of the final good. This result benefits consumer and, therefore, it is consistent with the most fundamental goal of the antitrust.32 There are, however, limitations in considering countervailing power as a justification for collaboration.33 These limitations fall into three categories. First, the process of discovering and agreeing to a joint profit-maximizing solution can mean that substantial transaction costs are incurred. Second, collusion for the purpose of simulating a monopsony creates an increased risk that the colluding firms will cooperate as sellers. Finally, the result of such an approach cannot be accurately predicted when considering its application to oligopoly markets.

6.6.1  Transaction Costs Collaboration can result in substantial transaction costs,34 which must be added to the usual cost of production and distribution. As a result, these added costs reduce the returns to collaboration. In this context, transaction costs can be seen as falling into two categories. First, even in the case of a pure monopolist negotiating with a pure monopsonist, there are costs associated with reaching an agreement on how to split the joint maximum joint profit. Second, when one moves to the possibility of permitting buyers to collude in order to simulate the condition of pure monopsony, there are also costs associated with forming the group and arriving at a consensus Although there has been considerable debate, most commentators agree that the promotion of consumer welfare is the most important goal of the antitrust laws. 33 Some of the practical problems of determining when such collaboration might be appropriate are examined in Chapter 10, which deals with proposals to allow physicians to bargain collectively with health insurers. 34 The general economic importance of transaction costs can be traced to Ronald Coase, The Nature of the Firm, 4 Economica 386 (1937). For a synthesis of the ensuing literature, see Oliver Williamson, Transaction Cost Economics, in Handbook of Industrial Organization, Richard Schmalensee & Robert Willig, eds. 135–82 (New York: Elsevier Science Pub. Co. 1989). The importance of transaction costs to antitrust analysis has been emphasized most thoroughly Williamson, supra note 27. 32

Limitations of Countervailing Power Considerations

137

on a common course of action. There is the very real possibility that these costs will more than offset the profits that would flow from the exercise of countervailing power. In that event, one would expect the collaborative effort to fall apart. To understand the first type of transaction cost, assume that there exists a bilateral monopoly. This does not mean that the parties will necessarily be aware of the economic consequences. Thus, there are likely to be costs associated with discovering that the market structure is such that the best each party can reasonably hope for is some portion of a jointly maximized profit. The firms experiencing this process of recognition are in a situation similar to the well-known prisoner’s dilemma.35 Each may first attempt to adopt a strategy that would only maximize its profits if the other firm “cooperated” and acted as though it were in a competitive industry. Of course, unlike the classic prisoner’s dilemma, there is no absolute bar to communications between the firms. Thus, the recognition of their mutual predicament and opportunity may come without the “signaling” required in solutions to the typical prisoner’s dilemma construct.36 Nevertheless, the recognition that profit can only be maximized through a joint profit-maximizing effort can be a costly process. Even if the parties recognize that theirs is truly a bilateral monopoly and that the “best” strategy is one that involves maximizing joint profits, they still must arrive at a decision about what the joint maximizing level of output will be. Moreover, and perhaps more importantly, they must also decide how to divide the jointly maximized profit. Since price determines this distribution, agreement on price is the most fundamental question the firms will be required to resolve. In other words, while the joint profitmaximizing level of output for the bilateral monopoly is not indeterminate, arriving at an agreement to jointly maximize profits would almost certainly also require that the parties agree on how the profit is to be allocated.37 And this division is, in fact, indeterminate in that there is no unique result that See, generally, R. Duncan Luce & Howard Raiffa, Games and Decisions (New York: John Wiley 1957); Anatol Rapoport & Albert Chammah, Prisoner’s Dilemma:  A Study in Conflict and Cooperation (Ann Arbor: University of Michigan Press 1965). 36 See, generally, Robert Axelrod, The Evolution of Cooperation (New York:  Basic Books 1984). 37 Paul Samuelson, The Monopolistic Competition Revolution, in III The Collected Scientific Papers of Paul Samuelson, R. Merton, ed. 35 (Cambridge, MA: MIT Press 1972). observes:   [T]he rational self-interest of each of the two free wills does not necessitate that there will emerge, even in the most idealized game theoretic situation, a Pareto-optimal solution that maximizes the sum of the two opponents’ profits, in advance of and without regard to how that maximized profit is to be divided up among them. 35

138

Bilateral Monopoly

is driven by price theory. This invites the type of strategic behavior that can be a serious obstacle to the agreement to jointly maximize profits. The transaction costs discussed previously would be incurred even if the market structure were already one of bilateral monopoly. The primary question for antitrust purposes, however, is whether there should be something like a bilateral monopoly defense whereby competing buyers could join together in order to bargain with a monopolist, thus simulating a bilateral monopoly. In this instance, another class of transaction costs would have to be considered. They are of two types. First, there are costs associated with the relatively mundane matters of coordinating the group’s efforts. More troublesome is the process of determining how each party will fare under the joint profit-maximizing agreement with the seller. Each of the potentially cooperating buyers will have some power to undermine the effort of the group if it does not get what it considers its fair share of the increased profit from the cooperative effort. Again, the dangers of strategic behavior and holdouts emerge. All of these costs are impediments to reaching an agreement to maximize joint profits. This does not mean, however, that these costs are an antitrust concern. More specifically, the possibility that these costs may be incurred is not a valid objection to a “bilateral monopoly defense.” This is not to say that every effort will be justified by the benefits or that firms will not make poor judgments about the difficulty of realizing joint profit maximization. Over the long run, however, these costs will not be incurred unless the overall benefit of achieving the joint profit-maximizing agreement offsets them.

6.6.2  The Risk of Seller Collusion A further cost that can be offered as a reason for not permitting a countervailing market power defense is that parties who buy the same inputs very often then sell their output in competition with each other. For example, suppose that a widget monopolist is selling to a competitively structured gadget industry. As to buyers of widgets, a buying cooperative will provide countervailing buying power. In isolation, that will yield positive social welfare consequences. If this is offset by less competition among the gadget manufacturers as sellers, the net social welfare of permitting their collusion as buyers could be negative. There are actually two versions of the risks of allowing the buyers to collude. The first is the simple fact that the legitimized collusion on buying efforts creates an opportunity for agreements on their selling price. This generalized fear of horizontal efforts that may evolve into agreements that

Limitations of Countervailing Power Considerations

139

violate the antitrust laws has been reflected in a number of Supreme Court opinions.38 The possibilities are not a simple matter of placing the proverbial “cookie jar” too close to the hungry culprits. The “approved” collusion would lower the costs of colluding on output price. For example, since the parties are permitted to gather for the purpose of determining a uniform purchase price, it would be more difficult to detect when they had crossed over to at least a tacit agreement on selling price. This decreased likelihood of detection lowers the risk associated with the price-fixing collusion. Second, there may be something approaching economies of scale in collusive activities. Thus, the costs of gathering together and deciding on a common plan could be spread over plans associated with both buying and selling. Thus, both the risk of detection would seem to decrease and the cost per agreement could be seen as declining, both of which would increase the likelihood of such agreements. It is important to note that the possibility that these buyers may become price-fixing sellers can quickly offset the gains made possible by allowing the monopolist and the collusive monopsony to determine the joint profit-maximizing level of output. Once the firms begin to act as a collusive mono­psony, they will raise prices and restrict output and consequently demand less of the input. The increase in consumer welfare, which was the justification for permitting the buyers to cooperate in the first place, could be quickly eliminated. More subtle is the selling price stability that may result if the parties do not agree on selling price but price individually while knowing the price paid by competitors for one or more inputs. Of course, it is always possible that input cost uniformity will lead to uniformity in the pricing of outputs. And, in industries in which production techniques and input mixes are standardized, one would not expect a great variety in price. This would be true even in highly competitive markets. On the other hand, if the market is oligopolistic and these buyers cooperate, the knowledge that they all have paid the same amount for a major input can reduce competitive pressure. An extension of the idea of collusion to create bilateral monopoly would be to permit a merger of all the competitors. Campbell makes the case that competitive sellers should be permitted to merge to monopoly if they confront a pure monopsonist.39 He argues that such a merger would result in bilateral monopoly, which is more efficient than the prior market structure. The same, of course, can be said for merging to monopsony to confront a See, e.g., Ariz. v. Maricopa County Med. Soc’y, 457 U.S. 332 (1982); Am. Column & Lumber Co. v. United States, 257 U.S. 553 (1936). 39 Tom Campbell, Bilateral Monopoly in Mergers, 74 Antitrust L. J. 521 (2007). 38

140

Bilateral Monopoly

monopoly. In either event, the emergence of countervailing power proves efficiency. At least in the case of merging to monopoly, Campbell argues that the merger would be consistent with merger policy as it enhances consumer welfare. In the specific circumstances analyzed, Campbell’s results are correct. It is, however, a risky policy choice. For example, whatever gains are produced by a merger to monopsony will be lost should there be new entrants on the selling side of the market so the monopsonist no longer faces a monopolist. In addition, the formation of a monopsonist may increase concentration in the markets in which the monopsonist sells its output and, thus, offset some or all of the gains. Moreover, since these losses would be the result of single-firm price increases or parallel pricing, an antitrust response would be unlikely.40 Baker, Farrell, and Shapiro join the discussion in a number of ways.41 First, they note that Campbell’s analysis is empirically irrelevant. They then conduct their own analysis of dubious empirical relevance. Based on this, they argue that their more complete theoretical analysis leads to results that are different from Campbell’s.42 In addition, they examine instances of bilateral monopoly and conclude that the results predicted by Campbell are not found. Finally, they challenge Campbell’s interpretation of merger policy under Section 7 of the Clayton Act. Neither Campbell nor Baker, Farrell, and Shapiro affect our analysis and conclusions. With the possible exception of a few relatively rare situations, promoting competition is always better than accommodating monopoly power or monopsony power by permitting consolidation on the other side of the market. To get a feel for the problem, consider the case of Microsoft with its dominance in the operating system market. Would we want to encourage all PC manufacturers to merge to deal with Microsoft? What happens if competition breaks out in the operating system market? There would no longer be an operating system monopolist, but there would still be a monopsonist that also would be a monopolist in the PC market.

6.6.3  Oligopoly and Oligopsony In the extreme case of bilateral monopoly, we know what the welfare consequences are. In cases involving substantial concentration (i.e., oligopoly In principle, the merger could be challenged after the fact if market conditions change, but such actions are rare. 41 Jonathan B. Baler, Joseph Farrell, & Carl Shapiro, Merger to Monopoly to Serve a Single Buyer: Comment, 75 Antitrust L. J. 641 (2008). 42 Id. at 639. 40

Concluding Remarks

141

or oligopsony), it is not so clear that the formation of countervailing power is desirable. This ambiguity follows from the lack of a unified theory of oligopoly.43 Since we cannot be sure a priori what the welfare effects of oligopoly are, it is not possible to say what the consequences of countervailing oligopsony power will be. For example, suppose there were a market occupied by four large sellers with a 25 percent market share each and twenty buyers with 5 percent each. In this example, one might conclude that it would make sense to permit the existence of four equally sized buying cooperatives. But one cannot conclude that consumer welfare would be enhanced by such a decision. The problem is that the range of competitive behavior for the four firms extends from monopoly pricing to competitive pricing. Without identifying the very specific behavioral characteristics of an industry and verifying that these will not change in the near term, one would be unable to predict the outcome of the sanctioned oligopsony.

6.7  Concluding Remarks In this chapter, we have dropped the assumption that the monopsonist faces competing sellers. Instead, there is monopsony on the buying side and monopoly on the selling side, that is to say, we examined bilateral monopoly. Under these circumstances, the usual marginal analysis fails us. The final outcome depends on bargaining over how to split the surplus that is available in the market. Interestingly, the bilateral monopoly situation is more efficient than either the usual monopoly solution or the usual monopsony solution. In spite of this theoretical result we urge caution regarding any policy that would encourage the formation of monopoly to deal with a monopsonist or the formation of a monopsonist to deal with a monopolist. The gains in welfare that the formation of countervailing power may yield may easily be lost if that power spills over into other markets. Usually, it is better to promote competition than to promote countervailing power.

For two excellent surveys that display the current state of knowledge, see Carl Shapiro, Theories of Oligopoly Behavior, and Alexis Jacquemin & Margaret E. Slade, Cartels, Collusion, and Horizontal Merger, in Handbook of Industrial Organization, Richard Schmalensee & Robert Willig, eds. 329 (Elsevier Science Pub. Co. 1989), at 329 and 415, respectively.

43

Bilateral Monopoly

142

APPENDIX Price Determination in Bilateral Monopoly

The range of price indeterminacy can be established a bit more analytically. In the absence of vertical integration, the upstream monopolist’s profit function is given by πu = wx − C(x)



(A.1)

and the downstream monopsonist’s profit function is given by

πd = P(Q(x, y))Q(x, y) − wx − p y y

(A.2)



Initially, suppose that the upstream firm dominates in the negotiation process. By definition, this means that πu will be maximized. But such maximization is subject to the constraint that the downstream firm participates in the exchange, which obviously requires that the downstream firm’s profit be no less than zero. At the same time, if the upstream firm dominates the outcome, the downstream firm will be allowed to earn no more than zero profit. Therefore, upstream firm domination leads to the constrained optimization problem Max x, y,px s.t .

πu

π d = 0.

Setting equation (A.2) equal to zero and solving for w yields

w=

P(Q(x, y))Q(x, y) − p y y x

(A.3)

Substituting equation (A.3) into (A.1) yields the unconstrained maximand

πu = P(Q(x, y))Q(x, y) − C(x) − p y y



(A.4)

which will be recognized as the profit function of a vertically integrated firm. Now, it is clear that maximizing (A.4) will result in equations (6.3) and (6.4) in the text (i.e., a maximization of joint profits). From equation (A.3), domination by the upstream firm in the negotiation process will lead to an intermediate product price equal to the average net revenue of the intermediate product. This price exhausts the downstream firm’s profits.

Price Determination in Bilateral Monopoly

143

Alternatively, suppose that the downstream firm dominates the negotiation process. In that case, the downstream firm faces the constrained optimization problem Max x,y,px s.t.

πd πu =0

Setting equation (A.1) equal to zero and solving for w yields w=



C(x) x

(A.5)

Substituting equation (A.5) into equation (A.2) yields the unconstrained maximand πd = P(Q(x, y))Q(x, y) − C(x) − p y y





(A.6)

which will also be recognized as the profit function of a vertically integrated firm. Once again, maximizing (A.6) will result in satisfying equations (6.3) and (6.4) and maximizing joint profits. In this case, however, the intermediate product price is set equal to the upstream firm’s average cost, which exhausts this firm’s profits. The Nondomination Solutions. Now, suppose that neither party dominates the negotiation process. Bilateral negotiation should lead to settlement at some point on the contract curve.44 In the bilateral monopoly case, the contract curve will be the locus of points of tangency between the isoprofit curves of the upstream and the downstream firms in w, x space. Taking the total differential of equation (A.1) setting dπu = 0, and solving for dw/dx, we obtain the slope of the upstream firm’s isoprofit curves dw dx



dΠ u = 0

dC −w = dx x

(A.7)

Assuming increasing marginal costs of producing x, equation (A.7) implies that the monopolist’s isoprofit curves will be U-shaped in Figure 6A.1, with higher curves denoting higher levels of profit. The minimum point on each isoprofit curve intersects the marginal cost curve. At points above marginal See, e.g., P. R. G. Layard & A. A. Walters, Microeconomic Theory 244 (New York: McGraw Hill 1978).

44

Bilateral Monopoly

144 w

MCX Πu3 Πd1=0 Πu2 Πd2 Πu1 Πd3 MRPX 0

x3

X

Figure 6A.1.  Bilateral Monopoly with Nondominant Conditions.

cost, we can see in (A.7) that w > dC/dx, which makes the numerator negative and, therefore, makes the slope of the isoprofit curve negative. The reverse is true for points below the marginal cost curve. Taking the total differentiation of equation (A.2), setting dπd = 0 and ∂πd/∂y = 0,45 solving for dw/dx, we obtain the slope of the downstream firm’s isoprofit curves dP  ∂ Q  =  P+Q − w  dQ  ∂ x

dw | dx

(A.8)

dπ d = 0 ∂π u ∂y



=0

With a declining net marginal revenue product curve,46 equation (A.8) implies that the isoprofit curves of the monopsonist will have an inverted We must set ∂πd/∂y = 0 because the monopsonist will optimize over input y given the negotiated solutions for x and w. 46 By “net marginal revenue product of x,” we mean the additional revenue generated by an additional unit of x given that y is adjusted optimally. Mathematically, 45

MRPx = (P + Q

dp

)

∂p

dQ ∂Q

is always evaluated where (6.4) is satisfied.

Price Determination in Bilateral Monopoly

145

U-shape in Figure 6A.1 with lower curves representing higher levels of profit. The maximum point on each isoprofit curve intersects the net marginal revenue product curve. At points below the marginal revenue product, (P + Qdp/dQ)∂Q/∂x > w, which makes the numerator positive and, therefore, makes the slope of the isoprofit curve positive. The reverse is true for points above the net marginal revenue product. The contract curve for this bargaining problem is found where dw | dx

= dπ u = 0

dw | dx

dπ d = 0 ∂π d = 0 ∂y

Substituting from equations (A.7) and (A.8) we have dC dp ∂Q = (P + Q ) dx dQ ∂y at all points along the contract curve. Thus, negotiation must lead to the maximization of joint profits if the firms are to be on the contract curve. Since the points of tangency must be at the joint profit-maximizing quantity of x3, the contract curve must be vertical. The core of the negotiation problem is that portion of the contract curve between wu3 and wd3. The intermediate product price is indeterminate within this range. Where it falls will determine how the maximized joint profits are split between the upstream and downstream firms. In other words, the price of the intermediate product is not a rationing device. Instead, it is a means of dividing the joint profits. Since this profit split has no allocative impact, it is only of distributive significance that the relative shares are indeterminate. The output of the intermediate product, however, is determinate as is the price of the final product and its output.

SE V EN

Monopsony and Antitrust Enforcement

7.1 Introduction The antitrust laws proscribe some business conduct that is in the individual self-interest of those with an opportunity to exercise market power as sellers or as buyers. One cannot expect the would-be monopolist or monopsonist to willingly surrender the potential profits that accompany the exercise of their market power. As a result, it is unlikely that enacting the antitrust laws would have accomplished much without an active enforcement effort and sanctions to go along with it. In this chapter, we turn our attention to antitrust enforcement and an analysis of the sanctions for antitrust violations. Enforcement of the antitrust laws is divided among federal antitrust enforcement agencies, the attorneys general of the individual states, and private plaintiffs. The level of enforcement by these parties can vary significantly from year to year. Moreover, when public authorities are involved, it is possible for violators to incur criminal penalties. For example, presently the maximum corporate fine is $100 million, the maximum individual fine is $1 million, and the maximum prison sentence is ten years. Regardless of the party bringing the case, the basic economics pertaining to liability does not change. What has become increasingly important, however, is the ability of private parties to bring actions. In other words, can they even raise the issue of whether an activity is unlawful? On this matter there are important considerations and this chapter is principally devoted to those matters.

7.2  Private Enforcement Section 4 of the Clayton Act provides an economic incentive for private enforcement. This provision allows anyone who has been injured by an 146

Private Enforcement

147

antitrust violation to sue in federal court and recover treble damages plus the costs of the suit, including a reasonable attorney’s fee: Sect. 4. … any person who shall be injured in his business or property by reason of anything forbidden in the antitrust laws may sue therefore … and shall recover threefold the damages by him sustained, and the cost of suit, including a reasonable attorney’s fee.1

Under Section 16 of the Clayton Act, a private party may also seek injunctive relief. The “private attorney general” feature of the antitrust laws was bolstered in 1976 by the Hart-Scott-Rodino Antitrust Improvement Act, which added a patens patriae provision to the Clayton Act.2 Under that provision, a state attorney general may file an action under the Sherman Act on behalf of the residents of his or her state. In these instances, the remedies available are the same as those available in a suit by an individual, but any monetary recovery is reduced by the amount recovered in other private actions. The impact of judicial interpretation of Sections 4 and 16 has had mixed effects on their viability as adjuncts to public enforcement.3 Specifically, there are complex and confusing issues regarding a potential plaintiff ’s standing to sue. Standing requirements serve mainly to limit the classes of plaintiffs who can sue. Most commentators seem to feel that developments in private suits through the 1960s tended to strengthen the impact of the antitrust prohibition of price fixing. But this consensus is not based on empirical evidence of deterrence at work. Rather, it is based on the conclusion that the private plaintiff ’s chances of a successful suit had greatly improved. Judicial developments have impacted private antitrust actions in different ways. From 1890 until 1940, very few private actions were initiated and even fewer were successful.4 Things did not get much better for private 15 U.S.C. §15 (1988). 15 U.S.C. §15c. For an analysis of parens partiae, see Roger D. Blair, The Sherman Act and the Incentive to Collude, 17 Antitrust Bull. 433 (1972). 3 The reviews on this are mixed. Some feel that the treble damage remedy in practice does not work as well as one might expect on a priori grounds. See, e.g., Russell Parker, Treble Damage Actions – A Financial Deterrent to Antitrust Violations?, 16 Antitrust Bulletin 483 (1971); and Malcolm Wheeler, Ántitrust Treble-Damage Actions: Do They Work?, 61 Cal. L. Rev. 1319 (1973). Others are concerned that private suits can be used to inhibit competition; see Richard Posner, Antitrust Law: An Economic Perspective 288 (Chicago: University of Chicago Press, 1976); Frank Easterbrook, The Limits of Antitrust, 63 Tex. L. Rev. 1 (1984); and William Baumol & Janusz A. Ordover, Use of Antitrust to Subvert Competition, 28 J. L. & Econ. 247 (1985). 4 From 1890 to 1940, only 174 private damage cases went to trial, and the plaintiffs prevailed in only 13. 1 2

148

Monopsony and Antitrust Enforcement

plaintiffs in the decade of the 1950s.5 During the 1960s, however, things changed dramatically for several reasons: (1) The Supreme Court reduced the procedural hurdles for private plaintiffs; (2) the Supreme Court broadened liability; and (3) there were changes in the rules governing class action suits. In fact, the number of private actions filed increased from 228 in 1960 to over 1,500 in 1977.6 Table 7.1 reveals what happened in the aftermath of the 1970s. It shows that the upsurge in private actions was over by 1980, and this was followed by a decline in both private and government actions. Since then, the number of cases brought by the government enforcement agencies has dropped precipitously, while private cases dropped initially but have undergone a recent upsurge. Even though it would be speculative to attempt to identify the precise causes of this trend, it may be more than coincidence that during the late 1970s and early 1980s there was a dramatic swing in the Supreme Court’s approach to antitrust law.7 This change had both substantive and procedural dimensions that have made it more difficult for private parties to prevail. The substantive changes were examined earlier in Chapter 2; thus, we focus on the procedural changes here. The most important procedural developments in the area of private enforcement have to do with the related issues of antitrust “standing” and “injury.” The antitrust standing requirement limits the universe of potential plaintiffs. The antitrust injury requirement limits the types of compen­ sable harms. Although the Supreme Court initially viewed it as analytically distinct, it is now safe to view the injury requirement as a component of the overall standing inquiry.8 Conceptually, the questions of standing and injury are threshold matters that must be satisfied before the plaintiff can proceed to the issue of liability. Thus, it is important to note that they are separate from the actual determination of the amount of damage suffered by the plaintiff. From 1952 to 1958, plaintiffs prevailed in only 20 of 144 cases. See Milton Handler, Harlan Blake, Robert Pitovsky, & Harvey Goldschmid, Trade Regulation, 3d ed. 109 (Westbury, NY: Foundation Press 1990). 7 In an effort to remedy a glaring empirical gap in our knowledge of private antitrust enforcement, the Georgetown project was undertaken. This effort involved the collection of data from the dockets of all antitrust cases filed in five selected districts during the 1973–83 period. Many of the results are presented in Lawrence J. White, Private Antitrust Litigation: New Evidence, New Learning (Cambridge, MA: MIT Press 1988). 8 The Supreme Court adopted this view in Cargill, Inc. v. Monfort of Colorado, Inc., 479 U.S. 104 (1986). 5 6

Private Enforcement

149

Table 7.1.  Antitrust cases filed in the U.S. district courts, 1980–2007 Year

Private cases

Government cases

1980

1,457

78

1981

1,282

142

1982

1.037

111

1983

1,192

95

1984

1,100

101

1985

1,052

90

1986

838

84

1987

758

100

1988

654

98

1989

639

99

1990

452

90

1991

650

93

1992

481

85

1993

638

86

1994

658

71

1995

744

75

1996

647

73

1997

570

62

1998

548

57

1999

608

76

200

811

90

2001

707

44

2002

806

44

2003

729

44

2004

731

33

2005

796

47

2006

967

37

2007

1,018

36

Source:  Andrew E. Abere, Trends in Private Antitrust Litigation: 1980–2004, Princeton Economic Group, http://www.econgroup.com/peg_news_view. asp?newid=27&pageno=.

150

Monopsony and Antitrust Enforcement

Questions of antitrust standing and injury require particular attention in the context of monopsony because all of the important case law addressing this topic concerns actions on the selling side of the market.9 From the perspective of competition among sellers, lower prices are the hoped for end result. Consequently, a rough rule of thumb has developed that harm resulting from lower prices does not satisfy the antitrust injury requirement. Of course, as illustrated in Chapters 3 and 4, lower prices may be indicative of collusive monopsony. Thus, in the monopsony context, one must resist applying the standing and injury requirements in a simplistic fashion. The critical antitrust standing and injury cases are summarized later in this chapter. The implications of these cases in typical monopsony cases are then discussed. Finally, the question of the appropriate measure of damages in monopsony cases is examined.

7.2.1  Antitrust Injury The leading antitrust injury case is Brunswick Corp. v. Pueblo Bowl-O-Matic, Inc.10 In the late 1950s, Brunswick’s sales of bowling lanes, automatic pinsetters, and related equipment soared along with interest in bowling generally. Most of Brunswick’s sales were on credit to the bowling alley operator. When the popularity of bowling waned in the 1960s, Brunswick’s sales fell, and it was faced with defaulting bowling centers. Selling or leasing repossessed equipment met with limited success and Brunswick was in great financial difficulty. As a result, it began to acquire and operate some of the failing bowling centers in order to salvage as much as it could financially. Pueblo filed suit claiming that the acquisitions violated Section 7 of the Clayton Act, which prohibits mergers that may have an anticompetitive effect. The theory was that Brunswick’s size enabled it to reduce competition by driving small rivals out of business. A jury agreed that Brunswick’s acquisitions violated Section 7. Pueblo then claimed damages on the grounds that if Brunswick had allowed the failing bowling centers to close, then Pueblo profits would have been higher. In other words, except for Brunswick’s competition, Pueblo would have had a larger share of the total market and the increased profits that would have accompanied that larger share. The Supreme Court agreed that Pueblo’s profits were lower than they would have been absent Brunswick’s illegal merger. But it ruled that Pueblo See, generally, Roger D. Blair & Jeffrey L. Harrison, Rethinking Antitrust Injury, 42 Vand. L. Rev. 1539 (1989). 10 429 U.S. 477 (1977). 9

Private Enforcement

151

had not suffered an injury that was compensable under the antitrust laws. Since the antitrust laws were enacted to protect competition, not competitors, it would have been bizarre to compensate Pueblo for the profits it would have earned had competition been reduced. The Court provided some further guidance on the nature of antitrust injury: Plaintiffs must prove antitrust injury, which is to say injury of the type the antitrust laws were intended to prevent and that flows from that which makes defendants’ acts unlawful. The injury should reflect the anticompetitive effect either of the violation or of anticompetitive acts made possible by the violation.11

Thus, the antitrust injury doctrine appears quite simple to apply in practice.12 First, one must determine what the anticompetitive effects of a particular violation are. Second, one must infer the logical consequences of those anticompetitive effects. If a plaintiff has been injured in his property or person due to the anticompetitive effects of an antitrust violation, then he would have suffered antitrust injury under the Brunswick rule. Any injury that is not a consequence of the anticompetitive effects of an antitrust violation would not be antitrust injury. Such an injury, therefore, would not be compensable under the remedial provisions of Section 4 of the Clayton Act.13 More recently, the antitrust injury issue was raised in Atlantic Richfield Co. v. USA Petroleum Co.14 Atlantic Richfield (ARCO) was an integrated oil company that sold gasoline directly to consumers through its companyowned stations and indirectly through ARCO-brand retail dealers. USA Petroleum (USA) was a nonintegrated, independent retailer of gasoline. USA claimed that ARCO conspired with the ARCO-brand dealers to retail gasoline at artificially low and uncompetitive prices. This agreement was Id. at 489. For an economic analysis of antitrust injury, see William H. Page, Antitrust Damages and Economic Efficiency: An Approach to Antitrust Injury, 47 U. Chi. L. Rev. 467 (1980); William H. Page, The Scope of Liability for Antitrust Violations, 37 Stan. L. Rev. 1445 (1985); William H. Page, The Chicago School and the Evolution of Antitrust:  Characterization, Antitrust Injury, and Evidentiary Sufficiency, 75 Va. L. Rev. 1221 (1989); and Blair & Harrison, supra note 9. For a recent summary, see John E. Lopatka, Antitrust Injury and Causation, in Issues in Competition Law and Policy, W. D. Collins, ed. (Chicago: American Bar Association 2008). 13 Application of this two-step process is not always easy. Consider the question of antitrust injury in Blue Shield of Virginia v. McCready, 457 U.S. 465 (1982). 14 118 S.Ct. 1884 (1970). For an analysis of the Arco decision, see Roger D. Blair & Gordon L. Lang, Albrecht After Arco: Maximum Resale Price Fixing Moves Toward the Rule of Reason, 44 Vand. L. Rev. 1007 (1991). 11 12

152

Monopsony and Antitrust Enforcement

in violation of Section 1 of the Sherman Act under the so-called Albrecht rule, which prohibits vertically imposed maximum resale prices.15 USA argued that the effect of the conspiracy was an expansion of ARCO sales at the expense of USA and other independents. By the time that the Supreme Court heard the case, USA had conceded that it could not prove that ARCO’s prices were predatory. Nonetheless, USA argued that the prices were driving independents out of the market. The Court technically left intact the per se illegal status of vertical maximum price fixing, but it held that the plaintiff had not suffered antitrust injury. It reasoned that the plaintiff ’s actual complaint was that it was unable to raise prices to the level that would exist in a less competitive context.

7.2.2  Antitrust Standing The antitrust standing requirement can be viewed most profitably as having two dimensions. First, the universe of possible plaintiffs in price-fixing cases is narrowed to those who are direct purchasers. Second, in other types of cases, the range of possible plaintiffs is narrowed by the difficulty of assessing damages, the risk of multiple liability, and how direct the harm to the party is. The direct injury question was addressed in Hanover Shoe, Inc. v. United Shoe Machinery Corp.,16 a 1968 case in which the Supreme Court was faced with the question of whether a defendant could resort to the so-called pass-on defense. Hanover Shoe challenged United Shoe Machinery’s policy of only leasing its machines as an instrument of illegal monopolization.17 Hanover Shoe alleged that its lease payments to United Shoe Machinery were inflated due to United Shoe’s monopolization of the shoe machine industry. As a result, Hanover Shoe claimed that it was entitled to recover the difference between its rental payments and what it would have paid if United Shoe Machinery had sold the machines.18 The defendant responded The Albrecht rule has been analyzed thoroughly and found wanting. See, e.g., Rogert D. Blair & David L. Kaserman, The Albrecht Rule and Consumer Welfare: An Economic Analysis, 33 U. Fla. L. Rev. 461 (1981). The Albrecht rule is inconsistent with virtually every goal of antitrust; see Roger D. Blair & James M. Fesmire, Maximum Price Fixing and the Goals of Antitrust, 37 Syracuse L. Rev. 43 (1986). Finally, in State Oil Co. v. Kahn, 522 U.S. 3 (1997), the Supreme Court overruled Albrecht. For an analysis, see Roger D. Blair & John E. Lopatka, The Albrecht Rule After Kahn: Death Becomes Her, 74 Notre Dame Lawyer 123 (1998). 16 392 U.S. 481 (1968). 17 This notion has been challenged in several places. See, e.g., Roger D. Blair & Jill Boylston Herndon, A Note on Hanover Shoe, 43 Antitrust Bull. 365 (1997). 18 392 U.S. 481. In principle this difference should be zero. 15

Private Enforcement

153

by pointing out that Hanover Shoe did not really suffer any damages because it had passed on any higher costs to its customers in the form of higher shoe prices. Thus, the Supreme Court had to decide whether United Shoe Machinery would be liable for the full amount of the overcharge that Hanover Shoe had allegedly incurred due to its antitrust violation even though Hanover Shoe had passed on part of the overcharge to its own customers. The Court rejected the pass-on defense responding to both an interest in forcing the defendant to disgorge the full amount of the illegally obtained revenue19 and the impracticality of apportioning the damages.20 The Court carved out an exception for those cases in which the plaintiff was reselling under a preexisting cost-plus contract.21 Having rejected the defensive use of the pass-on theory, the Court in Illinois Brick v. Illinois22 considered the offensive use of the pass-on theory and, thus, more squarely faced the question of how the universe of possible plaintiffs was to be shaped. The plaintiffs in Illinois Brick were the state of Illinois and local government entities who claimed that brick manufacturers were engaged in price fixing. The government agencies purchased brick indirectly by contracting with general contractors who, in turn, contracted with masonry contractors who actually purchased the brick. The argument was that the inflated price of brick was ultimately passed on to the final purchasers. It is important to note that the issue was not whether the plaintiff had suffered antitrust injury. Clearly, any overcharge paid by the plaintiff would constitute the type of harm the antitrust laws were designed to prevent. The Court’s approach was policy oriented; it identified several reasons for excluding indirect purchasers from the pool of potential plaintiffs. The primary rationales offered by the Court for excluding indirect purchasers were (1) the risk of multiple liability and (2) the complexity of adjudication likely to result from efforts to apportion an overcharge among various levels of direct and indirect purchasers. The Court noted that, although economic theory offers methods for making the apportionment, such apportionment was not judicially manageable.23 Id. at 494. Id. at 493. 21 Id. at 494. For an interesting analysis, see Herbert Hovenkamp, The Indirect Purchaser and Cost-Plus Sales, 103 Harv. L. Rev. 1717 (1990). See generally, Roger D. Blair & Jeffrey L. Harrison, Reexamining the Role of Illinois Brick in Modern Antiturst Standing Analysis, 68 Geo. Wash. L. Rev. 1 (1999). 22 431 U.S. 720 (1977). 23 For opposing views, see William Landes & Richard Posner, Should Indirect Purchasers Have Standing to Sue Under the Antitrust Laws? 46 U. Chi. L. Rev. 602 (1979); and Robert 19 20

154

Monopsony and Antitrust Enforcement

The Court addressed antitrust standing more generally in 1982. This time, the focus was different from the relatively simple direct purchaser question. In Blue Shield of Virginia v. McCready,24 the plaintiff ’s employer provided the plaintiff with health insurance. The contract provided that the plaintiff would be reimbursed for treatment by a psychiatrist but not for treatment by a psychologist unless the psychologist was supervised by a physician. The plaintiff alleged that the reimbursement of psychologists was excluded from coverage as a result of an anticompetitive pact between Blue Cross and the Neuropsychiatric Society of Virginia. The refusal of Blue Cross to provide reimbursement for the services of psychologists caused the actual injury. The Court distinguished its analysis from that in Illinois Brick, indicating that the question of multiple liability was absent in McCready. Instead, the Court framed the McCready standing issue in terms of how remote the plaintiff may be before falling outside the scope of the coverage in Section 4. Analogizing the question to one of determining “proximate cause,” the Court announced that it would look: (1) to the physical and economic nexus between the alleged violation and the harm to the plaintiff, and (2) to the relationship of the injury alleged with those forms of injury about which Congress was likely to have been concerned in making defendant’s conduct unlawful and in providing a private remedy under §4.25

The first question apparently dealt with the issue of antitrust standing, while the second focused on a Brunswick-type antitrust injury question. With respect to the issue of remoteness, the Court emphasized that the plaintiff, although neither a competitor nor literally a consumer, was a foreseeable and necessary victim of the reimbursement agreement. The harm, therefore, was neither so fortuitous nor so incidental as to fall outside the protection of Section 4. As for antitrust injury, the defendant invoked Brunswick by arguing that the plaintiff had not paid elevated rates for psychiatric services, the market supposedly benefiting from the agreement. Nor had the plaintiff claimed that her psychologist’s bills were higher than they would have been but for the agreement. Instead, her complaint was that her costs for psychological treatment were not covered under the health insurance plan. Thus, she Harris & Lawrence Sullivan, Passing on the Monopoly Overcharge: A Comprehensive Policy Analysis, 128 U. Pa. L. Rev. 269 (1979). 24 457 U.S. 465 (1982). 25 Id. at 478.

Private Enforcement

155

had not suffered the type of injury generally associated with an anticompetitive effort. The Court, however, ruled that those potential damages did not exhaust the Section 4 possibilities. In this instance, the plaintiff could seek the services of a psychiatrist, in which case the anticompetitive impact would have been felt by psychologists, or she could make use of a psychologist and forgo reimbursement. The plaintiff chose the latter route, leading to a higher net cost of psychological services. The Court held these higher net costs to be “inextricably intertwined with the injury the conspirators sought to inflict on psychologists.”26 In the following term, the Court attempted to clarify McCready and offered its most comprehensive examination of antitrust standing. In Associated General Contractors v. California State Council of Carpenters,27 the plaintiff union charged that a multiemployer builder’s association coerced members and nonmember landowners and subcontractors to hire non-union labor. The union claimed $25 million in damages as a result of the effort designed to “weaken, destroy, and restrain the trade of certain contractors.” The union did not couch its complaint specifically in terms of lower demand for union labor and resultant lower wages or financial losses to the union itself. The Court generally viewed the union as an independent business-like entity rather than as a group of sellers of labor. Moreover, the Court viewed the market in which the coercion was alleged to have taken place as being made up of building contractors and subcontractors. The Court applied a series of steps in determining that the union did not have standing. The Court first examined the nature of the plaintiff ’s alleged injury. Here, the Court applied Brunswick and apparently incorporated an antitrust injury requirement into the general question of antitrust standing. With respect to the injury, the Court indicated that the union was neither a consumer nor a competitor in the relevant market. In addition, it reasoned that as an economic matter there was no clear indication what the impact of increased competition in the relevant market would be on the union membership. In effect, the union failed to link its injury to the antitrust goal of increasing competition. The Court also examined the directness of the alleged injury. According to the Court, the initial impact of the alleged violation would fall on those contractors, landowners, and subcontractors who were coerced. The Court Id. at 484. 459 U.S. 519 (1983). See, generally, Note, More Trouble with Treble: The Effects of McCready and Associate General Contractors on the Antitrust Standing Opinions of the Federal Courts of Appeals, 10 J. Corp. L. 463 (1985).

26 27

156

Monopsony and Antitrust Enforcement

indicated that the presence of a group of direct victims made it unlikely that a serious antitrust violation would go undetected. This presence of direct victims diminished the need to permit the union to act as a private attorney general. Finally, the Court invoked the policy of Illinois Brick and explained that to permit the union to proceed could result in multiple liability and a complex process of apportioning damages.

7.3 Applying the Standards Some general observations on the issue of antitrust standing and injury are possible even though Supreme Court guidance remains vague. A party must satisfy three basic tests. First, the plaintiff must allege an injury that has flowed from the anticompetitive effects of an illegal activity. Second, the plaintiff, at least in a price-fixing context, must be a direct purchaser. Third, the plaintiff must not be affected too remotely. Judgment on this third factor will depend on the presence of more directly affected plaintiffs who normally would be expected to bring suit, the risk of multiple liability, and the complexity of the damage calculation. Whether these tests are viewed as components of a general test for antitrust standing or as separate tests of injury, directness and antitrust standing is functionally inconsequential.28 Consequently, unless it is essential to the analysis, the term “standing” is used in this chapter to denote the general concept inclusive of antitrust injury. The possibly confusing issue that the antitrust injury requirement presents in the context of monopsony can be seen most clearly by focusing on ARCO. Just like the plaintiff in ARCO, the seller who has been injured by the use of monopsony power will be claiming that it was harmed because it was unable to raise prices. Yet, in the case of monopsony, the lower prices do result in the type of harm forbidden by the antitrust laws. Specifically, the harm is the result of decreased competition among buyers. More importantly, the idea that if buyers pay less for inputs, then their customers will pay lower prices is simply wrong in the monopsony context. Thus, it is critical that courts look beyond the initial impact on price when addressing the question of antitrust injury. In the pages that follow, the antitrust standing and injury requirements are examined in the context of the substantive offenses illustrated in Chapter 4. There are a few instances in which the issues have been raised specifically in the context of monopsony. Phillip Areeda & Herbert Hovenkamp, Antitrust Law 371 (Boston: Little, Brown 1986).

28

Applying the Standards

157

7.3.1  Collusive Monopsony Collusive monopsony most obviously occurs when buyers agree among themselves on what price they will pay. On occasion, collusion is used to allocate markets or to foreclose rival buyers. We examine each in turn. 7.3.1.1  Price Fixing The potential plaintiffs in the case of a price-fixing collusive monopsony are sellers that have sold at prices below those that would have prevailed but for the monopsonistic restrictions and customers of the cartel who pay prices that would be lower but for the collusive monopsony’s restriction on input purchases. A final possibility, suggested by Hanover Shoe and Illinois Buick, would be an indirect seller who sells under a preexisting “price minus” contract to a buyer who resells to a collusive monopsony. In the case of direct sellers, it is tempting to conclude that they have suffered no antitrust injury because their real complaint is that they were unable to raise prices, even though the antitrust laws are not intended to enable a seller to raise prices. This is where a proper interpretation of Brunswick is crucial. Again, the Court found against Pueblo not because Pueblo preferred an interpretation of the antitrust laws that would permit it to raise prices. Instead, the sticking point was that the activity – a merger – of which the plaintiff complained, and which led to the harm it suffered, had increased, not decreased competition. Quite the opposite is true in the case of sellers to a collusive monopsony. The harm suffered is a direct result of a collusive activity that decreases competition among the buyers. Moreover, this injury is typically associated with decreases in the cartel’s output and increases in consumer prices. It seems just as clear that the sellers to the collusive monopsony should be regarded as meeting the additional components necessary to establish antitrust standing. Clearly, there can be no more direct victim; the sellers to a monopsony hold a position that is directly analogous to the position of the buyers from a price-fixing cartel.29 As we will see in the analysis of customers of the colluding firms, a finding that sellers to the collusive monopsony have suffered antitrust injury does not create a risk of multiple liability. An interesting question of antitrust injury and standing is presented when there is a dominant group of buyers and a fringe of competitive buyers. When the group exercises its buying power and depresses the price below the competitive level, its suppliers suffer antitrust injury. But how about the competitive fringe’s suppliers? They receive a price below the competitive level and are injured as a result. We would argue that this is antitrust

29

158

Monopsony and Antitrust Enforcement

The case of firms buying from a collusive monopsony is less clear and has to be one of the most curious in antitrust if not in economic theory generally. As we have already demonstrated, these buyers pay more even though the firms from which they make their purchases have employed monopsony power to obtain lower prices.30 Stated more accurately (and in a manner that proves that truth is often stranger than fiction), these buyers pay more because the firms they buy from have paid less for the input. Although the actual economics appears thoroughly counterintuitive, it seems clear that the harm suffered by these consumers flows from the unlawful collusion. Thus, they have also suffered antitrust injury. Whether they should be deemed to have antitrust standing is more difficult to determine. First, although they purchase from offending firms, they are not the parties to which the prohibited activity is directed. In effect, as far as the issue of remoteness is concerned, they seem to fall conceptually somewhere between the indirect purchasers of Illinois Brick and the sellers to the collusive monopsony. Multiple liability would not seem to be as great a concern. In Illinois Brick, the Court assumed that a price-fixing conspiracy would cause harm by an amount equal to the overcharge resulting from the conspiracy. The amount, under the analysis of the Court, was in essence a “pie” from which each successive purchaser would take a “slice.” Under the reasoning of the Court, to do anything other than apportion this fixed amount meant a risk of multiple liability. The case of price-fixing collusive monopsony is not analogous. The damage to the most direct victims – the suppliers – is distinct from the losses to the ultimate purchasers. The suppliers suffer a loss in revenue due to the decreased price. Ultimate purchasers suffer from higher prices resulting from restrictions in output that result from curtailed input purchases. Thus, there is no apportionment problem in the case of damages due to collusive monopsony and no problem of multiple liability. This is not to say that the damage calculation itself would be simple. In essence, one would have to determine what the price of the final output would have been if the colluding firms had paid more for the input,31 that is, if they had not restricted the purchase of the input. injury, and these suppliers should have standing to sue the collusive group. For a contrary view, see Note, Monopsonistic Price Fixing and Umbrella Pricing as a Theory of Antitrust Standing: A New View of Illinois Brick, 50 Cin. L. Rev. 52 (1981). For a view consistent with ours, see Areeda & Hovenkamp, supra note 28, at 337. 30 This explanation is provided in Chapter 3. 31 Traditionally, courts have been lenient with respect to the degree of certainty required when proving the amount of damages in an antitrust case. See, Bigelow v. RKO Pictures, Inc., 327 U.S. 251 (1946).

Applying the Standards

159

Although the factors may weigh in favor of finding that buyers from members of a collusive monopsony do have antitrust standing, the case for such a position is not beyond question. Moreover, it seems important to note that granting standing to this group of plaintiffs would be tantamount to deciding that the suppliers of inputs to price-fixing sellers would have standing. The case law indicates that such suits are rare if not nonexistent. Thus, our suggestion that buyers from a collusive monopsonist may have standing cuts against current antitrust realities. The introduction of a defense based on cooperative buying does little to change the analysis. One can understand why by once again referring to the three cooperative buying outcomes introduced in Chapter 5. In the first and third cases, the defendants’ actions involved the use of monopsony power. In the first case, it is without any offsetting benefits; in the latter case, it is accompanied by procompetitive efficiencies. The existence of antitrust injury and standing in the first instance is self-evident. Even in the third case, where a rule of reason analysis may result in a finding that the joint purchasing was procompetitive on balance, the plaintiff ’s harm is the product of the use of monopsony power. Any gains that were present also do not accrue to the plaintiff. Thus, the plaintiff may have technically suffered antitrust injury sufficient to pass standing requirements. On the other hand, the rule of reason analysis will reveal that there was no actual violation. This leaves the second possibility, cooperative buying resulting only in procompetitive effects. Obviously, these agreements do not result in antitrust injury to sellers. In fact, there is no “good faith” or common sense basis for a plaintiff to allege that it was harmed due to the benefits accruing to buyers as a result of cooperative buying. It does not seem unreasonable to expect that the possibility of Rule 11 sanctions would greatly reduce the likelihood of an action based on such ill-founded grounds.32 Rule 11, Federal Rules of Civil Procedure. Rule 11 provides for sanctions for frivolous claims. Since 1983, when the Federal Rule of Civil Procedure 11 was amended, Rule 11 sanctions have become a reality. This has not been without controversy. See, for a sampling, Note, Insuring Rule 11 Sanctions, 88 Mich. L. Rev. 334 (1989); Note, Plausible Pleadings:  Developing Standards for Rule 11 Sanctions, 100 Harv. L. Rev. 630 (1987), Note, The Dynamics of Rule 11: Preventing Frivolous Litigation by Demanding Professional Responsibility, 61 N.Y.U. L. Rev. 300 (1986); Lawrence M. Grosberg, Illusion and Reality in Regulating Lawyer Performance:  Rethinking Rule 11, 32 Vill. L. Rev. 575 (1987); and William W. Schwarzer, Rule 11 Revisited, 101 Harv. L. Rev. 1013 (1988). For the concerns of practitioners, see Tamar Lewin, A Legal Curb Raises Hackles, N.Y. Times, Oct. 2, 1986, at D1. For an argument that Rule 11 sanctions are too lenient in antitrust cases, see Martin B. Louis, Intercepting and Discouraging Doubtful Litigation: A Golden Anniversary View of Pleading, Summary Judgment, and Rule 11 Sanctions Under the Federal Rules of Civil Procedures, 67 N.C. L. Rev. 1023 (1989).

32

160

Monopsony and Antitrust Enforcement

A possible final class of victims of a price-fixing collusive monopsony would be indirect sellers. These would be the firms selling to the direct victims of the collusive monopsony. It will be recalled that Hanover Shoe and Illinois Brick left open the possibility that indirect buyers could avoid the direct purchasers dimension of antitrust standing if they were purchasing under preexisting, cost-plus contracts.33 The logic was that such arrangements avoided the difficult overcharge apportionment problem normally encountered in calculating the indirect purchaser’s damages. This only works, however, if the quantity is also fixed. Thus, the exception requires (1) that the contract preexist the infraction, (2) that the contract be specified as cost-plus some additional amount, and (3) that the quantity not vary as a result of the price change.34 At least in theory, the same rationale would apply to an indirect seller who sells under a preexisting, cost-plus contract.35 In these instances, too, it would be an easy matter to determine the undercharge damages. An attempt to employ this theory can be found in the Beef litigation in which cattle feeders who sold to meat packers claimed that retail grocery chains fixed the price at which they would pay the meat packers for the meat and that these depressed prices were passed along to the feeders.36 The feeders did not claim that they were victims of a preexisting, cost-plus contract, per se, but that the procedures used by the downstream firms were the functional equivalent. Under this “functional equivalency” test, the standard applied by the court was that the “impact of the retail chains’ price changes upon the pricing decisions of the [meat] packers be determined in advance without regard to the interactions of supply and demand.”37 Although the court recognized the reverse cost-plus theory and even a functional equivalent, the defendants were granted summary judgment. The analysis was applied at least in a partially correct manner more recently in Addamax v. Open Software Foundation, Inc. (OSF).38 In that case, the plaintiff competed in selling technological components to Open Software, which was attempting to develop a universally available computer 35 36

Ill. Brick v. Illinois, 431 U.S. at 736. See Kansas v. UtiliCorp United Inc., 497 U.S. 199 (1990). See generally Areeda & Hovenkamp, supra note 28, at 406–8. In re Beef Indus. Antitrust Litig., 542 F.2d 1122 (N.D. Tex. 1982), aff ’d, 710 F.2d 216 (5th Cir. 1983), cert. denied, 465 U.S. 1052 (1984). Feeders purchase the cattle from ranchers and then fatten and sell the cattle to packers. 37 In re Beef Indus. Antitrust Litig., 600 F.2d 1148, 1165 (5th Cir. 1979). This decision reversed an earlier unfiled opinion in which the theory of the cattle feeders was rejected on the pleadings. 38 888 F.Supp. 274 (D.C. Mass. 1995). 33 34

Applying the Standards

161

operating system. It was critical for each seller of technology to have its product as part of the system since it was likely that OSF would have market power in the downstream market. Addamax argued that OSF had used its monopsony power as a buyer to depress prices. The court noted that the mere fact that prices were depressed did not mean that consumers would be better off. In this context, in particular, the fact that OSF possessed power in the downstream market meant that it was unlikely that the depressed prices would accrue to the benefit of downstream purchasers.39 In short, the fact that plaintiff was forced to charge lower prices did not mean that they were unable to meet the requirement of antitrust injury.40 The issue of collusive monopsony was also addressed in Dyer v. Conoco, Inc.,41 in which a group of petroleum companies, under pressure from a local municipality, devised a buyout plan to purchase residential properties that had been adversely affected by the operation of those companies. The land was best suited for hydrocarbon storage. The companies essentially developed a formula for what would be paid to each landowner. The landowners claimed that they were victims of price fixing since the oil companies did not compete as buyers for each residential property, thereby causing the price paid to be lower. The claim was rejected on the basis of an antitrust standing analysis. According to the court, “In weighing the nature of the alleged injury, we evaluate the plaintiffs’ harm, the alleged wrongdoing by the defendants, and the relationship between them.”42 Further, “In order to establish antitrust injury, the plaintiffs must show some economic loss or harm as a result of the alleged conspiracy.”43 The court held that the plaintiffs lacked antitrust standing because they had not incurred an antitrust injury. It based this outcome on the fact that the property no longer had residential value, that its most valuable use was for hydrocarbon storage, and that no facilities had been purchased for that activity for six years or developed for fifteen years. In short, the court appeared to find no connection between the conspiracy and any harm suffered by the plaintiffs. Although the outcome of the case may have been correct, it is not clear that the concepts of antitrust injury and standing were applied correctly. In particular, the antitrust injury question is not about whether a plaintiff was Id. at 280. This, of course, is only true when the price reductions are due to monopsony power. 40 The reservation about whether the court was completely correct stems from a lack of clarity with respect to whether monopoly power in the downstream market was necessary for the outcome. 41 1995 WL 103233 (5th Cir. 1995). 42 Id. at 5. 43 Id. at 7. 39

162

Monopsony and Antitrust Enforcement

harmed, but whether the type of harm –whether it occurred or not – is the type the antitrust laws were designed to prevent. Here, the plaintiffs alleged conspiracy among buyers that caused prices to be lower than they would have been but for the conspiracy. Although the plaintiffs may have been incorrect as a factual matter, the injury they identified is the type of injury the antitrust laws seek to prevent. A more difficult and potentially more important issue arose in Sony Electronics, Inc. v. Soundview Technologies.44 Soundview claimed that Sony and others were fixing the royalties to be paid for the use of patented V-chips to be installed in television sets. The defendants claimed there was no antitrust injury because, contrary to the conventional monopsony model, the ultimate outcome would not be lower output and higher prices for consumers. Put differently, the defendants had no interest in manufacturing and selling fewer televisions and, thus, the practice was more consistent with an all-or-none model in which a monopsonist demands lower prices, but will not accept lower quantities.45 Soundview countered with the argument that even if the short-run effect may not be to lower output, the longer-run impact of collusively set prices would be lower levels of innovation and eventual decline in the number of suppliers. The court noted that it would not be in defendants’ favor to reduce the level of competition among their suppliers by setting prices at artificially low levels. Nevertheless, it declined to dismiss the complaint, reasoning that the theories involved were too complex to assess on a very limited record. This case suggests an issue yet to be fully resolved: Can a collusive monopsony successfully invoke an antitrust injury defense by demonstrating that, while suppliers may be injured, harm to consumers is unlikely? The issue is interesting. Absent the alleged conspiracy, the competitive price and quantity would prevail. The conspiracy reduces price but not quantity. The impact is purely distributional – all of the producer surplus is converted into consumer surplus, but the total consumer and producer surplus remains the same. The question becomes whether the loss suffered by the sellers because they were denied the benefits of competitive markets is compensable under the antitrust laws. 7.3.1.2  Market Division Aside from horizontal price fixing, there are a variety of ways in which a collusive monopsony may use its power. As in the case of a collusive monopoly, 157 F.Supp. 180 (D.Conn. 2001). For a discussion of the all-or-none model, see Chapter 3.

44 45

Applying the Standards

163

the colluding buyers may opt for a horizontal market division. The goal here is obviously to segment the market so that each party has monopsony power within its exclusive market. The standing analysis in these instances would track that found in the case of price fixing. 7.3.1.3  Boycotts A group boycott by buyers alters the analysis slightly by introducing the possibility of an additional plaintiff. For example, a group of buyers could attempt to exclude a competing buyer by threatening to stop purchasing from suppliers that sell to the competitor. The coerced ultimate objective would be to depress prices for inputs or increase them for the output of the boycotting firms. The suppliers and, possibly, the customers of the boycotting collusive monopsony would have standing. In addition, the excluded firm would seem to fit quite easily within the scope of those with antitrust standing. The presence of an additional and directly affected party may weaken the argument that the customers of the monopsony should be regarded as having standing.

7.3.2  Monopsonization 7.3.2.1  Nonprice Efforts Since the label of monopolization has not been applied to condemn the use of market power by a single firm to affect prices,46 the question of whether a seller to a monopsony has antitrust standing to challenge a monopsony buyer on the basis of the prices received is only of theoretical interest. Other legally recognized abuses of monopsony power, however, produce a predictable array of parties who would likely be regarded as having antitrust standing. For example, in the 1959 case of Klor’s v. Broadway Hale Stores,47 the defendant, a retail competitor of the plaintiff, used its monopsony power to deny the plaintiff access to appliances for resale. Although the Court did not engage in anything approaching an antitrust standing analysis, there is little doubt that the plaintiff would pass any modern standing test.48 See Kartell v. Blue Shield of Mass., 749 F.2d 922 (1st cir. 1984), cert. denied, 471 U.S. 1029 (1985). 47 359 U.S. 207 (1959). 48 In Klor’s, the antitrust issue involved the exclusion of a competitor that had been selling at discounted prices. The elimination of Klor’s would raise prices to consumers. Thus, the harm to Klor’s resulted in harm to consumers. 46

164

Monopsony and Antitrust Enforcement

Standing in the context of a monopsonization claim was addressed directly in White Mule Co. v. ATC Leasing Co.49 White Mule sold equipment used in the delivery of trucks and claimed that ATC Leasing, its principal buyer, was attempting to monopsonize the market for the purchase of equipment and to monopolize the market for delivery of trucks. It accused ATC Leasing of a variety of activities, but the central theory was that ATC Leasing used its power as a buyer to exclude competitors in the truck delivery market. More specifically, ATC Leasing made threats to White Mule, including threats to enforce fraudulently obtained patents, to dissuade White Mule from selling equipment to ATC Leasing’s competitors. The court found that White Mule lacked standing with respect to the delivery market. It reasoned that White Mule’s losses were secondary to those directly affected – competitors in the truck delivery market.50 In short, the injury was too remote. With respect to the monopsonization claim, the principal complaint was that ATC Leasing attempted to limit While Mule’s ability to sell to other buyers. In effect, ATC Leasing demanded exclusivity. The court again ruled that White Mule lacked standing, reasoning that White Mule’s complaint was that it lost sales to the defendant when it resisted the defendant’s demands and that the value of the company had declined. According to the court, these did not constitute injuries connected to a violation of the antitrust laws but were, instead, the consequences of hard bargaining.51 Instead, the appropriate plaintiffs were those in competition with ATC Leasing on the buying side of the market. 7.3.2.2  Predatory Pricing More difficult is the question of whom, if anyone, should have standing to challenge predatory pricing by a monopsony. To a limited extent, this question can be answered by reference to the U.S. Supreme Court’s recent examination of monopsony predation in Weyerhaeuser Co. v. RossSimmons Hardwood Lumber Co.52 As noted in Chapter 4, the alleged substantive offense involved the strategy of offering prices for inputs in excess of the input’s marginal revenue product and absorbing all inputs available at that price. When competing firms that were unable to obtain inputs left 540 F. Supp.2d 869 (N.D. Ohio 2008). Id. at 887–8. 51 Interestingly, the court indicated that had ATC Leasing been successful and pushed White Mule to an all-or-nothing supply curve, the holding on antitrust would be different. Id. at 890. 52 127 U.S. 1069 (2007). This case has been analyzed in Roger D. Blair & John E. Lopatka, Predatory Buying and the Antitrust Laws, Utah L. Rev. 415 (2008). 49 50

Applying the Standards

165

the market, the losses incurred due to the predator’s overpayment would be recouped by a sustained decrease in the prices paid for the inputs.53 The sellers to a firm that has successfully excluded rival buyers and then used its monopsony power to force prices lower would likely have antitrust standing. Even this conclusion raises the complicating issue of whether the defendant could offset the underpayment damages by the amount of overpayment received by these plaintiffs during the period of predation.54 The more difficult question, however, concerns the rival buyers and has three forms.55 There are three types of potential plaintiffs: the rival that was unable to stay in the market, the firm that has not yet left the market but claims it may if the predation continues, and the firm that survives an unsuccessful predatory effort. The obvious hurdle in all instances is that the plaintiff ’s actual complaint is that, due to the predation, it was unable to buy inputs at lower prices. In the monopsony context, this argument is as unappealing as the argument that the competitors of a monopolist engaged in predatory pricing have standing based on an inability to raise their output prices. The best guidance on this question comes from the selling-side predatory case, Brooke Group Ltd v. Brown and Williamson Tobacco Co.56 and the more recent buying-side case, Weyerhaeuser. In Brooke Group, the Court considered a predatory pricing case by a competing seller who had not left the market. The Court did not directly address the issue of standing but announced a substantive standard that can be used to inform the standing question. In an indirect fashion, it defined what an “antitrust injury” is in a predatory pricing context and, consequently, any potential plaintiff must link its harm to that injury. The standard announced for the substantive violation was a period of pricing below cost and the likelihood of successful recoupment if or when monopoly status was achieved. The inference from that case is that a rival buyer would have to demonstrate that its harm is directly traceable to above marginal revenue product pricing and that the defendant was likely to recoup the losses suffered during the period of predation by buying at low prices. In Weyerhaeuser, the selling-side case, the plaintiff claimed to have been forced from the market. Here again the Court did not address the Whether this strategy makes economic sense depends on the net present value of the entire effort. The present value of the losses incurred in driving out rival buyers must be weighed against the present value of the future gains. If the net present value is positive, then the strategy makes economic sense; otherwise, it does not. 54 For a similar argument in the context of monopoly, see William H. Page, The Scope of Liability for Antitrust Violations, 37 Stan. L. Rev. 1445 (1985). 55 For a general discussion of the matter in the more traditional monopoly context, see Blair & Harrison, supra note 9, at 1561–5. 56 509 U.S. 209 (2003). 53

166

Monopsony and Antitrust Enforcement

issues of standing and injury directly. It did, however, adopt the standards announced in Brooke Group in the selling-side context. Thus, according to the Court, the plaintiff would be required to demonstrate below-cost pricing of outputs and a dangerous probability of recouping losses by exercising monopsony power.57 Together these cases suggest answers for the three possibilities described earlier. First, from Weyerhaeuser itself, it seems clear that a rival buyer who is forced from the market has suffered an antitrust injury and does have antitrust standing if the claim is that the harm flowed from predatorily increased input prices that meant output prices were below cost. Similarly, from Brooke Group, it seems likely that a rival buyer who has not yet left the market but claims the harm it is suffering as (1) a result of increased input prices (2) that mean the defendant is selling at below-cost prices and (3) that there is a dangerous probability that it will leave the market meets the standing and injury requirements. The final case involving the potential plaintiff who has successfully withstood a failed effort at predation is less clear. This plaintiff will be unlikely to demonstrate that the defendant achieved monopsony power or that it is likely to. While these are technically elements of the substantive offense, what it means is that the survivor of attempted monopsonization may be deemed not to have suffered an antitrust injury. It is not clear, however, that this approach is fully consistent with economic theory. While there is a natural inclination to focus more on allocative inefficiencies occurring after monopsonization of the market has occurred, the period of predation is also marked by allocative inefficiency. In other words, the artificially high prices offered for the input will attract resources into its production even though they are more efficiently utilized in the production of other inputs. In addition, this intermediate period of allocative inefficiency and the harm done to rival buyers are necessary elements of the successful predatory effort. In short, these firms must suffer precisely this type of harm in order for the defendant to achieve its goal.58

7.4  Mergers The case of horizontal merger or “merger to monopsony” can be analyzed along the lines described earlier. From the point of view of sellers of inputs and buyers of the monopsonist’s output, the merger presents the same Id. For a different point of view in the context of monopoly, see Frank Easterbrook, Predatory Strategies and Counterstrategies, 48 U. Chi. L. Rev. 263 (1981).

57 58

Price Discrimination

167

competitive dangers as collusive monopsony. Thus, the sellers should have antitrust standing while the buyers do not. The analysis in Chapter 4 suggests that there is considerable doubt about the possibility that a private plaintiff could prevail on the substantive issues raised in the context of a vertical merger. As a practical matter, this makes the issue of antitrust standing relatively unimportant. On the other hand, the issue of standing is conceptually independent and could be raised with respect to economic harm associated with vertical foreclosure, both actual and potential. In the case of actual vertical foreclosure, the plaintiff ’s theory would be that sources of supply that were available are no longer available and that the acquisition of inputs is more expensive. In essence, the merger would be portrayed as the structural equivalent of an exclusive dealing arrangement. The foreclosed firm, using a rationale similar to that used by the target of a boycott, would seem to have antitrust standing. The customers of the vertically integrated firm arguably have standing only if the vertical merger results in monopoly power and higher prices. For the reasons discussed in Chapter 4, it seems more likely that actual economic harm may result from potential foreclosure than from actual foreclosure. Interestingly, the issue of antitrust standing is more difficult. The harm to potential entrants would be that their entry was delayed due to the necessity of being vertically integrated. Lost profit due to the risk and expense of entry at multiple stages is clearly an injury. If the vertical merger is a Section 7 violation, this lost profit is conceptually distinct from the loss suffered by Pueblo. That is, the loss would not be due to competition and, therefore, it may well be a “harm” the antitrust laws are designed to prevent. The case for consumer standing is somewhat awkward. It can only be seen as a generalized appeal for more market participants and the lower prices that would result.

7.5  Price Discrimination The type of price discrimination involving monopsony power that is most likely to raise antitrust questions is the result of the use of that power to force a single seller to sell to the monopsonist at a lower price than available to other buyers.59 The potential plaintiffs in such an instance would seem In Chapter 4, the possibility that a monopsonist would buy from different firms at different prices was discussed. This particular practice is unlikely to give rise to a substantive basis for antitrust liability unless it became part of an attempt to monopolize.

59

168

Monopsony and Antitrust Enforcement

to be the seller who is forced to give in to the monopsonist and buyers who resell in competition with the monopsonist. The seller who has acquiesced to the buyer’s demand for a more favorable price and who has no other basis for a claim may be regarded as having antitrust standing. As a general matter, the claim that the seller would like to charge a higher price would not seem to satisfy the antitrust injury component of antitrust standing. In this instance, however, the lower price charged is a necessary step in the buyer’s effort to damage its competitors. The problem for the seller is that buyers only violate the Robinson-Patman Act when they induce violations by sellers. Thus, the seller would be in the uncomfortable position of proving, as a threshold matter, that the concession it had made was a violation of the Act. The competitor of a firm receiving the concession is likely to have standing to press a claim under the Robinson-Patman Act but, as discussed in Chapter 4, faces a myriad of practical and legal obstacles in its effort to recover. For example, the monopsonist, as a defendant, will only be liable if it has induced a seller to violate the Robinson-Patman Act. This means that it would be able to use any defenses that would have been available to the seller. If the seller’s lower price to the monopsonist was the result of a “good faith” effort to meet competition, the monopsonist would not have induced a violation and would not have violated the Act itself.60

7.6  Concluding Remarks and Measuring Antitrust Damages The process of actually proving the damages the party is entitled to recover is a very different one than determining whether that plaintiff has antitrust standing.61 That is not to say, however, that the matters of actual damage and standing are unrelated. Even though standing is a threshold concern, the antitrust injury component can be viewed as an assessment of the plaintiff ’s theory of how it has been injured. This is the same injury that is later quantified in order to determine the amount of harm. In comparison with the burden of proving the fact of damage, courts traditionally have applied a very relaxed standard to the level of certainty with which the amount of damages must be shown.62 The reasoning behind See E. Thomas Sullivan & Jeffrey L. Harrison, Understanding Antitrust and Its Economic Implications, 5th ed. 334 (New York: LexisNexis 2009). 61 For an extensive analysis of antitrust damages, see Areeda et al., Antitrust Law §§ 390–9 (New York: Aspen Publishing 2007). 62 See Jeffrey L. Harrison, The Lost Profits Measure of Damages in Price Enhancement Cases, 64 Minn. L. Rev. 751 (1990). 60

Concluding Remarks and Measuring Antitrust Damages

169

this policy is that proof of antitrust damages can be complex, and an exacting standard of proof would result in many wrongdoers escaping penalty.63 Damages may be estimated as a matter of just and reasonable inference, but reasonable inference is neither guesswork nor rank speculation. Juries cannot be asked to speculate.64 In many cases, antitrust harm ultimately manifests itself as a loss of profits and, in theory, lost profit is the proper measure of damages in those cases. In addition to lost profits, an “overcharge” methodology is frequently employed to calculate damages. The overcharge theory seems to have evolved as an accommodation to the difficulties in some cases of proving the amount of lost profit.65 Plaintiffs who have made purchases from sellers who have colluded to fix prices most often use the overcharge method. Thus, buyers from the colluding sellers are entitled to the difference between the price actually paid and the price that would have been available but for the collusion.66 The overcharge measure can also be adapted for use in other cases in which an increase in price is the most obvious market effect (e.g., horizontal market division and refusals to deal). In the context of collusive monopsony, the perspective is reversed; the proper measure of damages would be the amount by which prices were depressed below what would have been available in the absence of the collusion. The actual total recovery would be equal to this amount multiplied by the quantity actually sold. This is illustrated graphically in Figure 7.1. Demand (D) and supply (S) intersect to determine the competitive price for the input (w1) and the quantity purchased (Q1). Once the collusive monopsony is formed, the price is decreased to w2 and the quantity purchased falls to Q2. The damages incurred due to undercharges on units actually sold are equal to the rectangular area, w1abw2. This measure of damages was adopted in the leading collusive monopsony case, American Crystal Sugar v. Mandeville Island Farms,67 in which See Eastman Kodak Co. v. S. Photo Materials Co., 273 U.S. 359 (1927). See Roger D. Blair & William H. Page, ‘Speculative’ Antitrust Damages, 70 Wash. L. Rev. 423 (1995). 65 See Harrison, supra note 62. 66 The price that would have been available must be estimated with some care. For a useful examination of the statistical techniques that can be employed, see Michael O. Finkelstein & Hans Levenbach, Regression Estimates of Damages in Price-Fixing Cases, 45 L. Contemp. Probs. 145 (1983). 67 195 F.2d 622 (9th Cir.), cert. denied, 343 U.S. 957 (1952). In an earlier, better-known opinion, the U.S. Supreme Court had reversed a ruling that the agreement at issue did not 63 64

Monopsony and Antitrust Enforcement

170 Price

MFC

S

w1 w2

d

a b

c D 0

Q2

Q1

Quantity

Figure 7.1.  Monopsony Damages.

growers of sugar beets challenged a pricing formula that was agreed on by three buyers. According to the court, the proper measure of damages was “the difference, trebled, between the amounts actually realized by the appellees during the critical crop years from the sale of the beets to Crystal, and what they would have realized during the period had the unlawful price combination not been in existence.”68 As already suggested, the true damage suffered is equal to the lost profit associated with the change in price. This too can be illustrated using Figure  7.1. The producer surplus prior to price fixing was equal to area cw1d. This is reduced to area cw2b by the collusive monopsony. The area w1dbw2, which is the difference in the areas of these triangles, would be the actual economic harm suffered by the plaintiff. The difference between the damages recognized by the courts, area w1abw2, and the true damage, area w1bdw2, is rather easy to measure in principle. It is the triangle, dab, which measures the producer surplus that is lost because Q1 – Q2 units of the input were neither produced nor sold. In practice, however, one would have to estimate the size of the producer surplus lost because output was involve interstate commerce. Mandeville Island Farms v. Am. Crystal Sugar, 334 U.S. 219 (1948). 68 195 F.2d at 625.

Concluding Remarks and Measuring Antitrust Damages

171

not produced. This is generally difficult to do, and courts typically do not require it. In a great variety of antitrust offenses, the harm may lead to the eventual exclusion of a competitor or prevent the entry of a new competitor. Whether exclusionary conduct is by buyers or sellers, the only available measure of damages is lost profit. Efforts to show lost profit, particularly when the firm has little or no history in an industry, can be difficult to ascertain. In most instances, this uncertainty will not prevent a plaintiff from at least having its damage claim considered by the trier of fact,69 but there is a very real possibility that the damage estimate will be deemed speculative. For a general discussion, see Areeda et al., supra note 61 and Blair & Page, supra note 64.

69

EIG H T

Monopsony in Action Agricultural Markets

8.1 Introduction This chapter considers the use and possible misuse of monopsony power in the context of agricultural markets. In this regard, we are primarily concerned with structural monopsony. Problems associated with collusion are examined in the next chapter.1 The typical scenario that we address involves relatively small farms selling to large processors of agricultural products. Whether such a structure is economically inevitable is not clear, but it is the likely result of economies of scale in processing that may not exist at the production stage. In any event, there can be little doubt that there has been significant consolidation on the buying side of agricultural markets.2 This has clearly resulted in distributive effects to which the antitrust laws are unlikely to respond. This does not mean that there are no negative allocative effects. In fact, every private antitrust plaintiff is concerned with a distributive impact rather than allocative efficiency. Part of the role of modern antitrust analysis is to distinguish cases with only distributive effects from those that have both. This chapter explores this distributive/allocative idea Recent cases involving buyer collusion in agricultural markets include Knevelbaard Dairies v. Kraft Foods, Inc., 232 F.3d 979 (9th Cir. 2000); In re Southeastern Milk Antitrust Litigation, 555 F.Supp.2d 934 (2008); Pease v. Jasper Wyman & Son, 2002 WL 1974081 (Me. Super. 2002). 2 See, generally, Note, Challenging Concentration of Control in the American Meat Industry, 117 Harv. L. Rev. 2643 (2004); Peter C. Carstensen, Concentration and the Destruction of Competition in Agricultural Markets: The Case for a Change in Public Policy, 2000 Wis. L. Rev. 531 (2000); Roger A. McEowen, Peter C. Carstensen, & Neil E. Harl, The 2002 Senate Farm Bill:  The Ban on Packer Ownership of Livestock, 7 Drake J. Agric. L. 268 (2002). Moreover, the issue is hardly a new one, see Robert F. Lanzillotti, The Superior Market Power of Food Processing and Agricultural Supply Firms – Its Relation to the Farm Problem, 42 J. Farm Econ. 1228 (1960). Increases in buyer concentration have been especially significant in meat packing, where the four firm concentration for cattle increased from 20 percent to 69 percent between 1980 and 2000. See McEowen et al., at 269. 1

172

The Monopsony Problem

173

in some detail and examines agriculture-specific legislation that is designed to address monopsony conditions. It suggests that a case-by-case analysis of distributive and allocative impacts is warranted.

8.2  The Monopsony Problem The focus of concerns about monopsony power in agriculture is typically on the actual producers of livestock or farm crops. These producers buy inputs – feed, seed, fertilizer, and so on – from one set of suppliers and typically sell to processors. In some instances, these producers buy from sellers of inputs that have monopoly power and sell their output to processors who have monopsony power. The focus here, of course, is on the second part of this analysis. The usual description that raises concerns is one in which producers are “locked in” or “captive” and cannot realistically respond, at least in the short run, to lower prices offered by buyers.3 One example is found in the poultry industry. Growers, who actually raise the chickens, make substantial investments in poultry houses in order to grow poultry for integrators, those who process the chickens for resale. Economist Robert Taylor describes the eventual outcome in these terms: [T]he system can best be described as feudal, with master (integrator) and servant (contract producer). Houses require a huge investment. Four to six houses, which are a full-time job for one person, cost from $500,000 to $1,000,000. Poultry houses generally have a 20–30 year economic life. Poultry houses have no practical alternative use; without a contract, the houses have essentially no salvage value. The integrator owns the birds and feed, and fully controls the breed, quality of chicks, feed deliveries and quality of feed, timing of chick delivery and time at which birds are processed. The pay system for growers has bonuses for performance, but the integrator determines in large part the ranking of the grower and fully controls computation of performance. Economists call this a tournament pay system, but it is closer to a lottery and, at the whim of the integrator, can be a rigged lottery.4 There is no precise definition of what it means to be locked-in, but typically the idea is that a firm has no feasible option and therefore can be exploited. For example, a farmer may make substantial investment that is “sunk” – cannot be recovered should he abandon farming. As long as the price offered for his output exceeds average variable cost, he will continue to produce even though his profits are negative. 4 C. Robert Taylor, The Many Faces of Power in the Food System, http://www.ftc.gov/bc/ mergerenforce/presentations/040217taylor.pdf (last visited April 27, 2009). As one would expect, the characteristics may vary from product to product, but the existence of mono­ psony power seems, if not uniform, at least typical. See Statement of Professor Peter Carstensen, Single Buyer Markets in Agriculture, Senate Judiciary Committee Hearings, October 30, 2003. 3

174

Agricultural Markets

Using this description as a general model, the critical question is whether it is something about which antitrust officials should be concerned. The possibility that the situation is not one to which the antitrust laws are designed to respond may be unsettling to many.5 In this regard, there are three factors to keep in mind. First, as noted in Chapter 2, the antitrust laws are not responsive to mere exercises of market power by monopolists to raise prices or by monopsonists to lower prices.6 Second, given this, the proper time to address the development of buying power is at the time it emerges. In other words, if an antitrust response was warranted, it would have been in the form of prohibiting mergers that increase monopsony power.7 Finally, as Professor Taylor notes in his article, the outcome may be that growers are pushed onto their all-or-none supply curves. At least in the short run, this has only distributive consequences.8 Put differently, there may be little or no efficiency-based reason to invoke the antitrust laws. Despite these general propositions, there are specific issues to consider.

8.3  Monopsony Power and Contract Power Part of the problem of addressing the issue of monopsony power in agricultural markets is that “locked-in” is a term of art. In a sense, every buyer or seller faced with market power is to some extent locked-in,9 meaning they This is not to say that the situation described here may not raise public policy concerns that could be addressed through other legal avenues. For example, the distributive issue can be raised under the contract law doctrine of unconscionability. See Been v. O.K. Indus., 495 F.3d 1217 (10th Cir. 2007). 6 Of course, the antitrust response would be different if the selling or buying price were determined collusively or if the power were used for purposes of tying or reciprocal dealing. 7 This path was taken in the late 1990s when the Justice Department investigated the acquisition of Continental Grain by Cargill based on the belief that it would “substantially lessen competition for purchases of corn, soybeans and wheat. …” See Complaint, U.S. v. Cargill, Inc., Civil No. 99-1875, (D.D.C. 1999). The merger was later approved on the condition that the parties divest themselves of a number of grain elevators, in order to avoid increases in buying power. 8 See discussion in Chapter 3. At least in theory, the consequences could be more than distributive if an especially aggressive pricing policy resulted in prices below average total cost. In the long run, one would expect exit from the industry by producers. One ironic aspect of the antitrust analysis is that a group of firms acting collusively to force sellers onto their all-or-nothing supply curve would clearly violate the antitrust laws, although a single firm almost certainly would not be in violation. Recent cases involving buyer collusion in agricultural markets include Knevelbaard Dairies v. Kraft Foods, Inc., 232 F.3d 979 (9th Cir. 2000); In re Southeastern Milk Antitrust Litigation, 555 F.Supp.2d 934 (2008); Pease v. Jasper Wyman & Son, 2002 WL 1974081 (Me. Super. 2002). 9 See discussion, supra note 2. 5

Monopsony Power and Contract Power

175

are unable to respond to prices changes. In the most typical agricultural scenario, the sellers have made substantial investments in fixed inputs that are not easily transferrable to other uses. It is not precisely this investment – a sunk cost – that locks in the supplier, but the fact that any effort to respond to low prices by switching to alternative buyers will mean substantial new investment in plant and equipment. This switching cost makes the producer vulnerable to being pushed onto its all-or-none supply curve. The producer is more likely to take what is offered at least in the short run.10 The fact that switching costs are a source of monopsony power does not resolve the issue of whether an antitrust response is warranted. Instead, it raises one of the more perplexing problems in antitrust. Simply put, the question is: When is the perceived market power the type to which the antitrust laws should respond? The concept of being locked in and its implications for market power analysis have been explored extensively in cases involving the selling side of the market. The analysis there applies equally to agricultural buyers. The most well-known case is Eastman Kodak Co. v. Image Technical Services.11 There, Kodak sold photocopiers as well as parts and service. It instituted a policy of selling parts only to owners of machines who made their own repairs or who purchased Kodak’s service. Technically, the case raised the issue of whether Kodak illegally tied its own service to the sale of Kodak parts thereby effectively eliminating to buyers of Kodak the option of using independent repair services. In effect, having made the investment in the machines, so the argument went, they were locked in to whatever changes or demands Kodak made. The case turned on whether Kodak possessed market power in the market for repair parts. Kodak argued that it did not because any action that would disadvantage consumers in the parts market would extend to the market for the machines themselves. In that market Kodak did not have market power. Put differently, without power in the market for the machines, Kodak argued it could not exercise power in the parts market. In effect, Kodak argued that it could not treat current machine owners as captives. The Supreme Court disagreed with Kodak and noted that Kodak announced its policy change after buyers had purchased the machines and In the short run, the producer’s supply curve remains the marginal cost curve above the average variable cost curve, and the producer may exhibit some short-run flexibility or elasticity. The extreme example of this would be if having been pushed to the all-or-none curve, the sellers opt for nothing. 11 504 U.S. 451 (1992). See, generally, Herbert Hovenkamp, Market Power in Aftermarkets: Antitrust Policy and the Kodak Case, 40 UCLA L. Rev. 1447 (1993). 10

176

Agricultural Markets

that buyers wishing to avoid Kodak’s power – to, in effect, use independent service providers – would have to switch to different producers. Switching was, however, a costly proposition.12 This meant that an owner of a Kodak copier was forced to compare the price of a Kodak part with the cost of switching to another brand of machine – the only way of avoiding Kodak’s higher price. To the extent this comparison favored remaining with Kodak, the buyer can be said to be locked in.13 The net effect is similar to the seller raising price after the initial transaction. It is useful to compare Kodak with Queen City Pizza, Inc. v. Domino’s Pizza, Inc.,14 in which the plaintiffs unsuccessfully attempted to apply the reasoning of the Kodak Court. In Queen City, franchisees entered into contracts with Domino’s that required them to purchase “ingredients, supplies, and materials” from Domino’s or an approved supplier. The franchisees argued that Domino’s possessed monopoly power by virtue of the fact that Domino’s could raise its prices for these supplies while the franchisees were precluded from buying those supplies from lowercost alternative sources. In effect, they were locked in because they could not switch to other franchisors or types of investment. In distinguishing Kodak, the court noted that purchasers of Kodak’s machines fell victim to a change in policy after the initial purchase was made. The change was not foreseeable at the time. In contrast, Domino’s franchisees contracted into a situation in which they were expressly told that they would be required to buy from Domino’s. In a sense, they had consented to a possible future exercise of monopsony power. And, in theory at least, they were compensated, ex ante, for the risk they accepted in the form of a lower price for the franchise itself. There is actually no bright line between Kodak and Queen City because both deal with the elusive notions of consent and foreseeability. The franchisees in Queen City consented to the possibility of future increases in prices. Similarly, the purchasers in Kodak can be said to have “known” that they did not control the future availability of parts and service. This, in a sense, they had consented as well. The Court also noted the difficulty of life-cycle pricing in which a buyer would determine the costs of the machine over its useful life. If the cost of determining the actual or true cost of a good is high enough, buyers will not make the calculation and will enter into the bargain with imperfect information about cost. This imperfection with regard to cost can itself be a source of market power. 13 Switching costs are pervasive and can include anything from switching to a different grocery store to a new barber. In all cases, they weigh against switching, but often they are low enough that the switch occurs anyway. 14 124 F.3d 430 (3rd Cir. 1997). 12

Monopsony Power and Contract Power

177

In the context of agricultural markets, the producer who invests in fixed inputs without a long-term contract at assured prices for its output is the monopsony version of Kodak and Queen City. Since it is foreseeable to the producer that it is investing resources for the specific purpose of producing agricultural outputs and that it is, therefore, highly dependent on future prices paid by buyers, these cases would seem to fall closer to Queen City than to Kodak. There are, however, two complications that make a direct application of Kodak and Queen City less straightforward. In both of those cases, the defendants sold the product that led to the lock-in. In a general sense, a prior contractual relationship led to the dependence. In the agricultural context, it does not appear that the locked-in producers buy their equipment from future buyers of the output with the understanding that those buyers will also determine the price paid for the output. In these instances, regardless of the source of the market power possessed by buyers, an effort to use that power to force prices down is not something to which the antitrust laws are likely to respond. The second complicating factor pushes the analysis in the opposite direction. In the agriculture/monopsony version of the lock-in cases, there are often accusations that producers relied on promises by the buyers that the engagement would be long term with prices determined by a preset formula. When this is the case, contract issues arise with respect to misrepresentation and breach. And, the market power resulting from misrepresentation may create the lack of foreseeability to which the Court responded in Kodak. Still, when the outcome is no more than the insistence on lower prices and movement to an all-or-none supply curve, the role of antitrust law is unclear.15 The analysis comes into focus when one considers a recent case, Been v. O.K. Industries.16 There, poultry growers brought a class action against O.K. Industries, the only integrator/buyer in the market. In order to contract with O.K. in the first place, growers were required to obtain financing and to build chicken houses to O.K.’s specifications. The cost of one chicken house was up to $160,000. Under the terms of the contract, O.K. supplied all inputs, including the chicks. It only obligated itself to supply one flock of chicks to a buyer. The grow-out period was seven weeks. In short, growers If Kodak and Queen City Pizza are based on the question of whether buyers would foresee or anticipate changes in the future, it is not clear that the buyer who responds to a misrepresentation is not also locked-in. 16 495 F.3d 1217 (10th Cir. 2007). The case was brought under the Packers and Stockyards Act, which is discussed later along with the court’s decision. 15

178

Agricultural Markets

committed a $160,000 investment for the promise that O.K. may replace flocks “from time to time” or, evidently, not at all.17 The plaintiff/growers complained of a variety of contract terms including the formula for determining payment and the fact that O.K. had decreased the number of chicks it supplied each year. As in Kodak and Queen City, there was no evidence that O.K. possessed market power over the plaintiffs until the initial investment was made. On the other hand, any effort to switch to another line of production was likely to be prohibitively expensive and the court did find that O.K. possessed monopsony power once the investments were made. The correct outcome of the case from an economic perspective is not clear. First, unlike Kodak and Queen City, there was no claim that would be analogous to tying. Instead, most of the claims concerned ways in which the compensation to growers would be decreased.18 The distinction between tying and simply lowering the price paid may not be significant from the seller’s point of view since both result in a lower net profit.19 Tying, however, has the additional impact of foreclosing other competitors – something to which the antitrust law do respond.20 Second, unlike Kodak, it is not clear whether or not the practices complained of were included as terms of the contract initially. In short, it was not clear that there was a change in practice that the processors had not, in effect, agreed to. Finally, even if no tying was involved and the only impact was on the prices paid for growing services, this does not mean the effects are distributive only. These factors lead to a number of possibilities. One, of course, is that there has been a breach of contract by O.K. That was not, however, part of the claim.21 A second possibility is that O.K. did use its power to force prices lower. This possibility has three versions. One is that this is not the type of exercise of power to which the antitrust laws respond. In other words, the case could be viewed as being similar to Queen City. The underlying idea would be that when entering the contract, the growers were compensated in advance for O.K.s future actions, and those actions are merely efforts by O.K. to enjoy the benefit of the bargain it made when it did not possess monopsony power. Id. at 1223. For example, O.K. allegedly delivered dead chicks to the growers, which the growers were required to pay for. 19 It may, however, have allocative consequences as discussed later. 20 This is discussed more fully in the next section. 21 Plaintiff did assert, however, that the contract was unconscionable. 17 18

Tying and Reciprocal Dealing

179

A second version is that the growers were forced onto their all-or-none supply curves. If so, there would be no decrease in production and the impact would be strictly distributive. A final version follows the analysis found in Chapter 3. Lower prices will mean lower quantities supplied and ultimately higher prices and lower quantities offered for sale in the output market. Obviously, this final version means allocative losses and at least raises the type of economic concerns antitrust law is designed to address. Interesting, while negative allocative effects are a possibility, it is unlikely that the antitrust law would be employed because all three versions involve single-firm, price-affecting behavior.22

8.4  Tying and Reciprocal Dealing Although there is no antitrust response to the simple claim that one is locked into selling at prices that are too low, the analysis changes when the focus is on the use of market power to achieve other ends. When tying and reciprocal dealing are involved, the use of the power can have an impact on competition in other markets. In the case of both reciprocal dealing and tying, it is still necessary to address the Kodak/Queen City Pizza issue of whether the defendant buyer possesses the type of power the antitrust laws are designed to address.23 In other words, the question of whether the producer/plaintiff more or less willingly contracted into monopsony conditions must be considered. Assuming they did not and that the possessor of that power goes further than demanding favorable prices, a different analysis emerges. One possibility is tying. In the monopsony context, this would mean agreeing to purchase one product from sellers only if the seller also sells a second product to the buyer. In this instance, the buyer might be the only customer for the first product and instead of using its monopsony power to lower the price, it uses that power to force the sale of a second product that the producer would prefer to offer on the open market. The tying scenario seems unlikely, but a similar use of power is found in reciprocal dealing.24 Reciprocal dealing means that one buyer will buy from a seller only if that seller will buy something from the first buyer. To raise As discussed later, the Packers and Stockyards Act may be employed to address single-firm price-affecting practices. 23 In more general terms, the type question goes to whether market imperfections can lead to market power the antitrust laws prohibit. This proposition is discussed most clearly by Justice Scalia in Kodak, 504 U.S. at 494. See, generally, Jeffrey L. Harrison, An Instrumental Approach to Market Power and Antitrust Policy, 59 SMU L. Rev. 1673 (2006). 24 For a detailed explanation of reciprocal dealing, see Chapter 19 in Roger D. Blair & David L. Kasserman, Antitrust Economics, 2d ed. (New York: Oxford University Press 2009). 22

180

Agricultural Markets

antitrust concerns, the practice must be both conditional and coercive.25 Using the poultry example, described previously, a processor of chickens or turkeys may purchase a producer’s growing services only if the producer purchases a number of inputs from the processor.26 For example, the producer could be required to purchase everything from chicks or poults to feed, medication, and litter. The impact would be on competition in the market for the items the producer is “forced” to buy as a consequence of its dependence on the processor as a buyer.27 The actual contract arrangement can take an even more convoluted form. For example, the buyer may purchase the grown-out poultry at a preset price claiming that it is “giving” the supplies to the producer. Of course, the price paid for the grown-out poultry will be lower, to allow for the cost of these supplies, and the economic impact is indistinguishable from the more straightforward example.28 Whether it is the unlikely possibility of tying or the more likely possibility of reciprocal dealing, antitrust law may be responsive. The theory for doing so is best understood by first focusing on the more typical selling side of the market. In the following examples, X is the product in which the firm has buying or selling power and Y is the product, the sale or purchase of which is tied. In the case of tying, the impact is on the possibility that more efficient sellers of the Y are foreclosed. Whether this is a realistic fear is uncertain. The firm involved in the tying would have an incentive to obtain Y at the lowest cost even if that means outsourcing production to other more efficient suppliers.29 In that case, one could hardly argue that Reciprocal dealing can be tricky, and it is present everywhere. At one level, it can simply mean that one person does something to “pay back” another. For example, a life insurance salesman buys haircuts, auto repairs, accounting services, and dental services from clients. This is reciprocal but not coercive. 26 See, e.g., Been v. O.K. Indus. Inc., 495 F.3d 1217 (10th Cir. 2007). 27 This may not always be the case. In Spartan Grain v. Mill, 581 F.2d 419 (5th Cir. 1979), a leading reciprocal dealing case, there was conditioning, but the arrangement was mutually beneficial. Poultry farmers already had hen houses, but no business. Spartan offered to buy hatching eggs from the farmers if they would agree to buy feed from Spartan. Here Spartan was creating a book of business for its animal feed by agreeing to provide a market for the customers output. The result appears to be procompetitive, as outputs expanded. This arrangement resulted in an antitrust suit for distributive reasons. Neither Spartan’s competitors in the feed market, nor Ayer’s competitors in the hatching egg market complained. The arrangement created a surplus, and Ayers was dissatisfied with his share. Unfortunately, the trial court confused reciprocity with tying. 28 In the context of tying and reciprocal dealing, it is useful to keep in mind that under some theories, these practices actually produce little or no harm. See Roger D. Blair & David L. Kasserman, supra note 24, at Chapters 18 & 19. 29 In fact, the X producer will not maximize its profits if it fails to purchase Y from others if they are more efficient producers of Y. 25

Tying and Reciprocal Dealing

181

more efficient firms have been foreclosed. The outcome is more distributive in nature than anything else. In a selling-side reciprocal-dealing case, the firm with monopoly power requires the buyer to sell an item or items that the buyer would prefer to sell to other buyers. In these instances, foreclosure means that buyers placing a higher value on Y are unable to obtain them. The value will be higher if these alternative buyers can use Y more efficiently for the purpose of producing profit. Again, whether this misallocation and foreclosure actually occurs is an empirical question. Unless it engages in predation, it is not clear why the reciprocal-dealing firm would not resell Y to firms with more valued uses. In this case there would be no foreclosures. Similarly, for there to be foreclosure, the buyer would have to purchase all of the Y produced. Consequently, sellers of Y may just increase their output. When the analysis is flipped to monopsony and the focus is tying, the foreclosure occurs in the same manner as the reciprocal-dealing firm with monopoly power. The firm with monopsony power in the market for X requires the seller to also sell Y. Firm’s valuing Y more than the monopsonist are arguably foreclosed. Again, though, the question is why the monopsony buyer would not then resell to those who place a higher value on Y. Under a monopsony reciprocal-dealing theory, the monopsonist purchases product X only under the condition that the seller purchase product Y from the monopsonist. Here the foreclosure issue mirrors that found in the typical selling-side tying case. The question, as in the seller-side tying case, is whether this results in the foreclosure of more efficient producers of Y. In fact, in all cases of either monopoly or monopsony tying or reciprocal dealing, there are arguments that the feared foreclosure will not occur. As already suggested, the possibilities of tying and reciprocal dealing do not eliminate the need for the Kodak/Queen City analysis. This sets up three relatively distinct possibilities with respect to antitrust enforcement. First are those cases in which a firm possesses legally obtained monopsony power from a source unrelated to its own prior contact with the buyer. For example, in the typical grow-out contract example described earlier, the seller may be selling to one processor and now finds that processor has been replaced in the market by a new one. The second processor has monopsony power, and a critical question for determining the role of antitrust law is whether the power of buyer two is used to depress price as opposed to requiring reciprocal dealing. Obviously, in the former case, no response is likely, and in the latter case, the usual tying or reciprocal-dealing analysis is appropriate. The second category of cases involves instances in which a firm has mono­psony power by virtue of a contract the seller willingly entered into

182

Agricultural Markets

and which includes terms that are favorable to the buyer and that were foreseeable by the seller. Whether the power is used to depress prices or engage in what would otherwise be regarded as tying or reciprocal dealing, the buyer is properly viewed as having consented to the use of the power. This is comparable to the Queen City scenario. A third possibility exists when a firm enters into a short-term contract with a buyer that requires – whether or not at the buyer’s insistence –the seller to make significant investment specific to one line of production. When the contract period ends, the buyer will possess monopsony power. Use of that power to depress prices does not ordinarily require an antitrust response. On the other hand, use of the power to force tying or reciprocal dealing does raise an antitrust response. Although all of these possibilities and the consequences conform to current antitrust law, it is not clear that the distinctions warrant different treatment. This is because all the cases have an overriding similarity. In each case, investments are made without any contractual assurance that the investment will turn out to be profitable. Put differently, whether the eventual advantage-taker is the initial buyer or a subsequent buyer and whether leverage is used to affect price or some other term, the producer’s position is the result of market imperfections that the firm possessing market power did not create. The exception to this would be when the firm’s power is a result of a deliberate effort to mislead the seller, but even here it is possible that a contract law solution is more appropriate.

8.5 Alternative Approaches to Agricultural Buying Power Legislation other than the antitrust laws has been used to benefit producers who may find themselves selling to firms with buying power.30 For example, in recent years, there have been efforts to pass legislation that would prohibit packers from ownership of livestock for more than a prescribed period prior to slaughter. The logic of the prohibition is that the packers eventually pay for the owned or controlled cattle at a price that is determined by a formula. If they pay more for cattle not owned or controlled, that affects the formula price. In other words, the marginal cost of independently raised livestock is multiplied because of the impact of the formula, and this discourages increases in price.31 See The Agricultural, Conservation and Rural Enhancement Act of 2001, S. 1731, 107th Cong. (2002). 31 See Taylor, supra note 4. 30

Alternative Approaches to Agricultural Buying Power

183

The most ambitious effort to equalize the power between buyers and seller is the Capper-Volstead Act of 1922. The Act was designed to exempt farmer cooperatives from the antitrust laws to a limited extent.32 Under the Act: [P]ersons engaged in the production of agricultural products as farmers, planters, ranchmen, dairymen, nut or fruit growers may act together in associations, corporate or otherwise, with or without capital stock, in collectively processing, preparing for market, handling, and marketing in interstate and foreign commerce, such products of persons so engaged.33

The antitrust exemption applies to cooperatives composed of producers who combine to market their output. It does not extend to monopolization,34 price discrimination, coercion of members or nonmembers, or efforts to restrict output.35 In short, the Act does not address monopsony power directly but does allow the development of market power on the selling side of the market, thus setting up at least the possibility of bilateral monopoly.36 The theory of bilateral monopoly suggests that the Act may address the concerns expressed about distributive outcomes but even this is not free from doubt. One relatively recent investigator concluded that the “special tools,” including the Capper-Volstead Act, designed to address power imbalances in the agricultural sector “seem not to have been used especially well ….”37 The impact on efficiency and consumer welfare is perhaps even less certain.38 Legislation that has also been used to address perceived imbalances in agricultural markets is the Packers & Stockyards Act39 of 1920. See, generally, James B. Dean & Thomas Earl Geu, The Uniform Limited Cooperative Association Act, 13 Drake J. Agric. L. 63 (2008); Amber Brady, Post-Smithfield and Hazeltine: An Evaluation of the Capper-Volstead Act as an Alternative Means of Marketing Power for Producers, 10 Drake J. Agric. L. 331 (2005). 33 7 U.S.C. §§291, 292 (2004). 34 This is not the same under the antitrust laws as simply possessing a dominant market share. 35 See U.S. Department of Agriculture, Rural Business and Cooperative Service, Understanding Capper-Volstead (April 1995), http://www.rurdev.usda.gov/rbs/pub/cir35.pdf. See also, Maryland and Virginia Milk Producers Ass’n v. U.S. 362 U.S. 458 (1960). 36 The economics of bilateral monopoly are discussed in Chapter 6. 37 Richard J. Sexton, Industrialization and Consolidation in the U.S. Food Sector: Implications for Competition and Welfare, 82 Am. J. Agric. Econ. 1087 (2000). 38 For one view on this which is discussed in Chapter 9, see Roger G. Noll, ‘Buyer Power’ and Economic Policy, 72 Antitrust L. J. 589 (2005). 39 7 U.S.C. §192 (2004). For an analysis of the Act, see Note, Challenging Concentration of Control of the American Meat Industry, supra note 2, at 2657. 32

184

Agricultural Markets

Section 213 of the Act lists a number of types of prohibited conduct. For example: It shall be unlawful for any packer or swine contractor with respect to livestock, meats, meat food products, or livestock products in unmanufactured form, or for any live poultry dealer with respect to live poultry, to: (a) Engage in or use any unfair, unjustly discriminatory, or deceptive practice or device; or … (c) Sell or otherwise transfer to or for any other packer, swine contractor, or any live poultry dealer, or buy or otherwise receive from or for any other packer, swine contractor, or any live poultry dealer, any article for the purpose or with the effect of apportioning the supply between any such persons, if such apportionment has the tendency or effect of restraining commerce or of creating a monopoly; or (d)  Sell or otherwise transfer to or for any other person, or buy or otherwise receive from or for any other person, any article for the purpose or with the effect of manipulating or controlling prices, or of creating a monopoly in the acquisition of, buying, selling, or dealing in, any article, or of restraining commerce; or (e) Engage in any course of business or do any act for the purpose or with the effect of manipulating or controlling prices, or of creating a monopoly in the acquisition of, buying, selling, or dealing in, any article, or of restraining commerce; or ….40

The important thing to note is that some sections prohibit activities that affect competition but subsection (a) does not. In addition, to put the importance of interpretations of the Act in context, it should also be noted that until 1970, the buying side of the meat industry remained relatively unconcentrated with sellers offering their output at auctions attended by sufficient buyers to create competitive conditions. Mergers, principally in the 1980s, led to concentration in the industry greatly reducing the number of buyers41 and leading to the monopsony conditions found today. In light of buyer-side consolidation, the issue has arisen of whether the Packers and Stockyards Act can be used to address purely distributive matters, as suggested by the foregoing subsection (a) above or whether it should be reconciled more generally with the goals of the antitrust laws 7 USC §§181–3 (2004), et seq. The increased concentration is attributed to the Justice Department’s and the Grain Inspection, Packers and Stockyard Administration’s failure during the Reagan administration to oppose these mergers. See Eric Schlosser, Fast Food Nation: The Dark Side of the All-American Meal 137 (New York: Perennial 2002).

40 41

Alternative Approaches to Agricultural Buying Power

185

as suggested by subsections (c) and (d). The question was considered in Pickett v. Tyson Fresh Meats, Inc.42 The context was cattle raised exclusively for slaughter. Some of these cattle were purchased on what is called the “cash” market meaning that the cattle are raised by ranchers who then sell the grown cattle through a traditional one-on-one bargaining process. The cattle are delivered seven days after the agreement. The cash market was the way nearly all cattle were sold until the 1980s. At that point, packers began to enter into marketing agreements that pegged the price of any purchase at the average cash price for the week prior to the purchase.43 The cattle are delivered within a two-week period with the actual day picked by the packer. The plaintiffs44 were producers selling in the cash market. Their claim was that Tyson used the marketing agreements to force the cash price lower. It did this, according to the plaintiffs, by using marketing agreements that essentially lowered the demand for cattle on the cash market. This meant lower prices on the cash market and, ultimately, also for those cattle sold under marketing agreements. At trial, the jury found for the plaintiffs and awarded $1.3 billion in damages. The trial judge reversed the judgment and found for Tyson. The case turned on an important issue of interpretation with respect to the Packers and Stockyards Act.45 Plaintiffs claimed that they had met their burden under subsection (a) of the Act by demonstrating that the cash and overall price of cattle was lower as a result of the marketing arrangements. Tyson argued that the Act required more. Specifically, their argument was that the Act required a demonstration that there had been an adverse effect on competition. The argument was one that imported into subsection (a) the requirements of other subsections and the more general requirement that the antitrust laws exist to protect competition but not necessarily individual competitors. The Eleventh Circuit Court of Appeals adopted Tyson’s reasoning and its argument that the practice was actually procompetitive in that it permitted a reliable and stable supply of cattle46 and lowered transaction costs.47 44 45

420 F.3d 1272 (11th Cir. 2005). The price is then adjusted based on the quality of the actual cattle delivered. The case was a class action. See, generally, Christopher M. Bass, More Than a Mirror:  The Packers and Stockyards Act, Antitrust Law, and the Injury to Competition Requirement, 12 Drake J. Agric. L. 423 (2007). 46 This meant Tyson could operate its processing plants at a high and constant level. 47 420 F.3d at 1281–6. In so holding, the court joined the Seventh, Ninth, and Eighth Circuit Courts of Appeal. 42 43

186

Agricultural Markets

A different position was adopted by the Court of Appeals for the Fifth Circuit in Wheeler v. Pilgrim’s Pride Corp.,48 a case dealing with growers and processors of poultry. The fact pattern was somewhat different in that the processor owned the chickens from the outset and the growers were supplied with the chicks, feed, and other supplies necessary to raise the chickens at which time they were “returned” to the processor. Defendant had a different arrangement with another grower under which that grower was compensated at a higher rate. In some respects, from a theoretical perspective, the defendant was acting as a price-discriminating monopsonist as described in Chapter 4. The Fifth Circuit reasoned that the drafters of the legislation had specifically expressed concerns about competition in some sections of the Packers and Stockyards Act and not in others. Consequently, it rejected the view that plaintiffs attempting to recover under subsection (a) were required to demonstrate that the practice in question harmed competition.49 Consequently, at this point there is a disagreement among the circuits on whether a practice must harm competition in order to violate the Packers and Stockyards Act. Even in those circuits that have adopted the view expressed by the Eleventh Circuit in Tyson, there may be an important role for the Packers and Stockyards Act to play. As Chapter 3 describes, unilateral actions by firms with monopsony power, like Tyson, can have negative allocative effects. Lower prices paid by processors may mean a lower quantity supplied and higher prices to consumers as a result of restricted output. On the other hand, unilateral conduct that simply forces prices down will not lead to antitrust liability. This suggests that, even with the requirement that a harm to competition be shown, there is an economically sensible role for the Packers and Stockyards Act. Indeed, this was the holding in Been v. O.K. Industries, which was discussed earlier. There the court held that the Act did require a showing of economic harm but held that the Act was to be applied more broadly than the antitrust laws.50 Specifically, the court indicated that it would focus on the use of monopsony power even if legally obtained. Plaintiffs were only required to demonstrate that O.K depressed prices paid and resold to consumers at higher prices.51 536 F.3d 455 (5th Cir. 455 2008). The court did not, however, indicate whether the practice in question violated subsection (a). For a lower court case with a similar outcome, see Schumacher v. Tyson Fresh Meats, Inc., 434 Supp.2d 748 (D. S.D. 2006). 50 495 F.3d 1231. 51 Id. at 1234. 48 49

Concluding Remarks

187

8.6  Concluding Remarks To some extent, the discussion of agricultural monopsony issues reflects the discussions that took place forty years ago. At that time, courts debated whether antitrust law was to be guided by broad populist goals that meant preserving small economic units (largely distributive concerns) or by allocative efficiency and consumer welfare. In large part, the arguments in many agricultural contexts seem to be distributive in nature in that they are motivated by a desire to protect the relatively small farmers or ranchers from the larger processor/buyers. This, in itself, does not mean the issues are without importance with respect to modern antitrust goals. All antitrust plaintiffs are ultimately motivated by distributive matters. The key is whether in the process of addressing those matters, positive allocative effects can be obtained. What this requires is a case-by-case analysis consistent with that employed in more familiar selling-side cases. Thus, the central questions are (1) whether sellers have willingly entered into arrangements that permit buyers to exercise buying power about which they later complain and (2) whether the practices result in purchases and prices that are lower than would exist in competitive markets or whether buying power is used to foreclose competition.

NINE

Monopsony in Action The NCAA

9.1 Introduction As noted in Chapter 1, one of the more prominent areas in which monopsony issues arise is in the market for athletes. Professional leagues, such as the National Football League and the National Basketball Association, use reverse order drafts to avoid competing among teams for entry-level players. The teams have agreed on eligibility requirements, free agency, and other terms of employment.1 Our focus here is not on professional athletes but on amateur athletes “employed” by various colleges and universities. As we will see, they operate under the auspices of the NCAA in monopsonistic ways.

9.2 Introduction to the NCAA2 The NCAA has been characterized as a cartel, that is, “a combination of independent commercial or industrial enterprises designed to limit competition.”3 This is unfair since one might object that the NCAA’s members are institutions of higher learning rather than “commercial or industrial enterprises.” But this objection is misplaced. As we will see, this is an accurate characterization, especially as it relates to the labor market for They are permitted to do this because of the nonstatutory labor exemption to the antitrust laws, which permits anticompetitive agreements among competitors if those agreements are also part of a collective bargaining agreement. See Jeffrey L. Harrison, Brown v. Pro Football, Inc.: The Labor Exemption, Antitrust Standing and Distributive Outcomes, Antitrust Bull. 565 (1997). 2 Some of the following is adapted from Chapter 19 in Roger D. Blair, Sports Economics (New York: Cambridge University Press 2006). 3 Webster’s New Ninth Collegiate Dictionary (Springfield, MA:  Merriam-Webster 1988). For a detailed study, see Arthur A. Fleisher, Brian L. Goff, & Robert D. Tollison, The National Collegiate Athletic Association: A Study in Cartel Behavior (Chicago: University of Chicago Press 1992). 1

188

Collusive Monopsony

189

student-athletes and coaches. More precisely, the NCAA acts as a buyer cartel or a collusive monopsony. Under the auspices of the NCAA, the member institutions collude to reduce competition among themselves in an effort to reduce costs and thereby generate more “profit” for their athletic programs.4 Examples of collusive monopsony can be found throughout the economy, but the NCAA is a particularly good example. The NCAA’s members collude on two key inputs in the production of athletic competition: the student-athletes themselves and their coaches. With respect to athletes, the agreement restricts quantities by placing a ceiling on the number of scholarships that a school may award in each sport. In addition, in the name of amateurism, the compensation of these athletes is limited to room, board, tuition, books, and incidentals.5 Bonuses for winning conference championships are limited to relatively inexpensive rings or watches. These restraints have not gone unchallenged. In re NCAA I-A Walk-on Football Players Litigation6 involved a challenge to the limitation on quantity. In White v. NCAA,7 a class of student-athletes complained about being undercompensated. With respect to coaches, the number employed in each sport is limited by an NCAA Bylaw. In this connection, we will examine the Hennessey v. NCAA litigation.8 The compensation of coaches is typically unconstrained, although there have been attempts to restrict the earnings of some coaches. In this regard, we will examine the Law v. NCAA litigation,9 which led the NCAA to rescind its rule restricting the earnings of some assistant coaches. In this chapter, we briefly review the economic theory of collusive mono­ psony, which was developed in Chapter 4. Then, we explain how the NCAA deals with the problems inherent in organizing a cartel, coordinating the efforts of the members, and enforcing the terms of their agreement.

9.3  Collusive Monopsony When buyers behave independently, the forces of supply and demand will typically lead to a price and quantity that maximizes social welfare. The NCAA members are not-for-profit entities, but the surplus that they generate in some parts of their athletic programs are profit in the usual sense. Generally, these profits are dissipated through expenditures in training facilities, coaches’ salaries, subsidies for “nonrevenue” sports, and the like. 5 For an interesting analysis, see Lawrence M. Kahn, Markets:  Cartel Behavior and Amateurism in College Sports, 21 J. Econ. Persp. 209 (2007). 6 2006 WL 1207915 (W.D. Wash. 2006). 7 No. CV06-0999 (C.D. Cal. 2008). 8 564 F. 2d 1136 (5th Cir. 1977). 9 902 F. Supp. 1394 (D. Kan. 1995). 4

The NCAA

190 Wage

MFC

S

W1 W2 D = Σ MRPi 0 L2

L1

Labor

Figure 9.1.  Monopsony Labor Market.

Figure  9.1 shows the market for labor (athletes or coaches) services L, which is an input in producing athletic contests. The independent decisions of buyers and sellers of labor will lead to a wage of w1 and a quantity of L1. Although the schools enjoy some profits at competitive wages, they can do better. If they agree among themselves and use their collective buying power to depress price, they can increase their profits at the expense of the athletes and coaches, and to the detriment of society. If the formerly independent schools pool their demands and coordinate their purchases, they can exploit whatever monopsony power they possess. In order to maximize their collective profits, they will restrict employment to L2 and thereby depress the wage to w2, which is below the competitive wage. All cartels face problems of organization, reaching collective agreements, implementing those agreements, and enforcing them. The NCAA is particularly well suited to acting like a collusive monopsony because of its structure and its ability to enforce its agreements.

9.4  The NCAA and Collusive Monopsony The NCAA behaves like a collusive monopsony in acquiring two crucial inputs:  student-athletes and coaches.10 As a result, it faces the usual The NCAA has been so successful that “the NCAA is the clear and deserving winner of the first annual prize for best [monopsony] in America.” Robert J. Barro, Let’s Play Monopoly,

10

The NCAA and Collusive Monopsony

191

problems confronted by all buyer cartels: deciding on payments, imposing hiring quotas, limiting nonprice competition, sharing the resulting profits, coordinating activities, and deterring cheating.11 The structure of the NCAA is designed to deal with all of these problems. The NCAA’s durability, resilience, and enormous success is proof of its ability to adjust as necessary to cope with the changing needs of its member institutions. Before intercollegiate athletics became big business, there was a need for uniform rules regarding the games being played, especially football, which was particularly brutal. As a result, a precursor to the NCAA was formed in 1905. Between then and now, the NCAA has evolved in size, scope, and power.12 Now that intercollegiate sports, particularly football and men’s basketball, have become big business, a good deal of the NCAA’s attention is centered on business matters: revenues, costs, and profits.

9.4.1  Organizing Limits on Prices and Quantities The NCAA members – some 1,000 educational institutions – meet annually to vote on a variety of matters. If necessary, special sessions can be called between the annual meetings. There are various committees that advise the NCAA members at large on specific issues of interest. Once agreement among the members has been reached, the policies of the NCAA are set out in bylaws that are binding on all members. Among these bylaws are those that deal with student-athletes and with coaches. There are rules governing the compensation of student-athletes. Generally, athletes may not receive more than a “full grant-in-aid,” which is defined as “financial aid that consists of tuition and fees, room and board, and required course-related books.”13 In addition to the limits on compensation, the NCAA limits the number of grants-in-aid that can be awarded in each sport.14 These limitations are summarized in Table 9.1.



11



12 13



14

Wall Street J., Aug. 27, 1991. The Antitrust & Trade Reg. Rep., Jan. 30, 1997, reported that the NCAA “continues in many regards to operate as a cartel … .” For an examination of cartels and their problems, see Roger D. Blair & David L. Kaserman, Antitrust Economics, 2d ed. (New York: Oxford University Press 2009), Chapter 9. For a more detailed historical account, see Fleisher et al., supra note 3. 2006–2007 NCAA Division I Operating Manual, article 15.02.5.1. The specific definition of each component of a full grant-in-aid is contained in article 15.2. It has been estimated that a “full grant-in-aid” is some $2,500 short of the actual cost. This obviously causes problems for student-athletes whose families cannot provide supplemental financial support. See Tom Farrey, NCAA Might Face Damages in Hundreds of Millions, ESPN.com (Feb. 20, 2006), http://sports.espn.go.com/ncaa/news/story?id=2337810. The limitations for men’s and women’s basketball are found in the NCAA Manual, at article 15.5.4; the limit for football is in article 15.5.5. For the other sports, see article 15.5.3.

The NCAA

192

Table 9.1.  Maximum number of grants-in-aid Men’s sports

Number of GIA

Baseball

11.7

Basketball

13.0

Cross country/track

12.6

Fencing

4.5

Football

85.0

Golf

4.5

Gymnastics

6.3

Ice hockey

18.0

Lacrosse

12.6

Rifle

3.6

Skiing

6.3

Soccer

9.9

Swimming and diving

9.9

Tennis

4.5

Volleyball

4.5

Water polo

4.5

Wrestling

9.9

Women/s sports

Number of GIA

Archery

5

Badminton

6

Basketball

15

Bowling

5

Cross country/track

18

Equestrian

15

Fencing Field hockey Golf

5 12 6

Gymnastics

12

Ice hockey

18

Lacrosse

12

Rowing

20

Rugby

12

Skiing

7

Soccer

12

Softball

12

The NCAA and Collusive Monopsony

Women/s sports

Number of GIA

Squash

12

Swimming and diving

14

Synchronized swimming Team handball Tennis

193

5 10 8

Volleyball

12

Water polo

8

Source:  2006–2007 NCAA Division I Operating Manual.

As one might expect, there are detailed rules regarding how one counts the student-athletes in the various sports and how to deal with those who participate in two sports. These details are important in closing loopholes. There are, for example, eighty-five grants-in-aid available to football players. One way around this limitation would be to recruit some football players and put them on track scholarships. The rules prevent this subterfuge. If an athlete plays football, his scholarship counts against the limit of eightyfive in football irrespective of how the school would like to record that player’s scholarship. There are also detailed rules regarding recruiting: number of official campus visits, length of stay, permissible activities during a visit, and text messaging and other contacts, among others. All of this is fully consistent with cartel theory. When competition on price and quantity has been restrained, it is imperative that competition on other dimensions also be restrained or the collusive profits will be dissipated. If restrictions on prices and quantities are to be fully effective, a cartel must prevent cheating on the agreement.15 The old saying, “there is no honor among thieves,” certainly applies to members of a cartel. Consequently, indirect means of skirting the agreed on restrictions must be foreclosed. If a cartel does not limit other ways of competing, costs will rise, and cartel profits will tend to disappear. The costs of nonprice competition can easily get out of hand and erode an athletic program’s profits. In competing for the best athletes, schools can offer lavish training facilities, plush athletic dorms, upgraded training tables, extensive educational counseling and tutoring, frequent campus visits with first-class plane tickets, and so on. As a program spends ever more money to attract the best players, its cartel profits fall. As a result, it is in the collective interest of the NCAA’s members For a detailed examination see Brad R. Humpreys & Jane E. Ruseski, Monitoring Cartel Behavior and Stability: Evidence from NCAA Football, 75 S. Econ. J. 720 (2009).

15

194

The NCAA

to implement rules that curtail such expenditures. They have been somewhat successful in this effort. As one might suspect, the limitations on the number of coaches by sport are fairly comprehensive for football, but they are less detailed for the other sports. A “countable” coach is anyone who “participates (in any manner) in the coaching of the intercollegiate team in practice, games, or organized activities directly related to that sport ….”16 For Division I-A football, there is a limit of one head coach, nine assistant coaches, and two graduate assistant coaches. In addition, an institution may employ a strength and conditioning coach who does not count against the number of football coaches. The limitation on the number of coaches for the other sports are summarized in Table 9.2.

9.4.2  Revenue Sharing In simple, textbook cartel models, there is no real need to worry about revenue sharing. When all the cartel members are identical, the restrictions on output or employment can usually be spread proportionately. As a result, each member just keeps the excess profit that its own operation generates, and equal profit sharing results. In the real world, of course, things are more complicated because college athletic programs are not identical, and this may pose problems in sharing the cartel’s spoils. In intercollegiate athletics, gate receipts are shared with opposing teams, bowl revenues are shared within conferences, television revenues are shared, and the NCAA basketball tournament revenues are shared. This revenue sharing is not complete. For example, individual institutions keep private donations, trademark licensing fees, concession revenues, and most of the gate receipts (net of payments to visiting teams), which means that some cartel members are going to be much better off financially than others. This financial imbalance cannot be eliminated entirely because there is a danger that the most powerful athletic programs could defect from the NCAA. In fact, the revenue-sharing formula has evolved over time to prevent the defection of a coalition of the most powerful athletic programs.

9.4.3  Sanctions for Cheating The NCAA has developed sanctions to deal with cheating on the cartel rules by member institutions. For example, when the NCAA found the University Operating Manual, supra note 13, at article 11.7.1.1.1.

16

The NCAA and Collusive Monopsony

195

Table 9.2.  Maximum number of coaches** Men’s sports

Number of coaches

Baseball

3

Basketball

4

Cross country/track

3

Fencing

2

Football*

12

Golf

2

Gymnastics

3

Ice hockey

3

Lacrosse

3

Rifle

2

Skiing

2

Soccer

3

Swimming

2

Swimming and diving

3

Tennis

2

Volleyball

3

Water polo

2

Wrestling

3

Women’s sports

Number of coaches

Archery

2

Badminton

2

Basketball

4

Bowling

2

Cross country/track

3

Equestrian

3

Fencing

2

Field hockey

3

Golf

2

Gymnastics

3

Ice hockey

3

Lacrosse

3

Rifle

2

Rowing

3

Rugby

3

(continued)

The NCAA

196

Table 9.2.  (continued) Women’s sports

Number of coaches

Skiing

2

Soccer

3

Softball

3

Squash

2

Swimming

2

Swimming and diving

3

Synchronized swimming

2

Team handball

2

Tennis

2

Volleyball

3

Water polo

2

*  In football, two graduate assistant coaches are also permitted. **  In addition to the numbers listed, a weight or strength coach may also be employed. Source:  2006–2007 NCAA Division I Operating Manual. Each entry shows the number of head and assistant coaches.

of Washington guilty of recruiting violations, it imposed substantial penalties: (1) a two-year ban on bowl participation, (2) a one-year ban on receiving television revenue, (3) a reduction in football scholarships for two years, (4) a reduction in the number of permissible football recruiting visits for two years, and (5) a two-year probation. The purpose of such sanctions is to deter cheating by making it unprofitable to cheat. Being able to impose such penalties on cartel members that cheat gives the NCAA an enormous advantage over most cartels that have limited enforcement options.

9.5 Antitrust Challenges to the NCAA Monopsony We examine several antitrust cases that demonstrate the NCAA’s ­behavior as a collusive monopsonist. These cases also illustrate the judiciary’s failure to understand some fundamental economic concepts regarding monopsony. Before beginning, it is useful to note that as a general matter, regardless of the activities of the NCAA and its members, courts have tended to assess those activities under the rule of reason. The rationale

Antitrust Challenges to the NCAA Monopsony

197

for this is that the NCAA, in order to offer a competitive product in the form of collegiate athletics, must have leeway to enact some rules that are designed to ensure rough parity between teams. These rules do ­restrict competition in the market, but without parity, so the theory goes, games would be lopsided and less attractive to both live and broadcast audiences.17 This is, in effect, an application of what is called the “ancillary restraints” doctrine. In each of the cases discussed in this section, the underlying antitrust question should be whether the restraint in question is necessary in order for the NCAA to compete with other sellers of ­entertainment products.

9.5.1  Walk-on Athletes: In re NCAA I-A Walk-on Football Players Litigation One of the most recent efforts to curb the NCAA’s use of monopsony power came in the form of a class action challenge to its limits on the number of football players under scholarship.18 Walk-on players (those not paid) argued that, but for the limit, they would have received scholarships. Along with the cap on what may be paid to scholarship players, the limit on the number of players limits a school’s labor costs and limits the means of competing for players. In terms of a pure antitrust analysis, the question would be whether increasing the number of scholarships would jeopardize the quality of the product sold by the NCAA, which is competition on the field. The answer to this is not clear. A walk-on player at one school may be a scholarship player at another. Limiting the number of scholarships may prevent some larger or wealthier schools from stockpiling players. On the other hand, walk-on players who excel are often subsequently offered scholarships, and a school would have little interest in stockpiling those who do not excel. Thus, it is not clear what the ultimate damage to competition in the sports entertainment market would be. These questions were not addressed, however, because the NCAA was able to prevail with an argument that the class of plaintiffs should not be certified.19 One of the requirements for class certification is that the NCAA v. Board of Regents, 468 U.S. 85 (1984). In re NCAA I-A Walk-on Football Players Litigation, 2006 WL 1207915 (W.D. Wash. 2006). 19 For a class to receive certification, a court must find that (1) the class is so numerous that joinder of all members is impractical, (2) there are questions of law and fact common to all members of the class, and (3) the claims of the representative plaintiffs are typical of 17 18

198

The NCAA

plaintiffs bringing the case in the name of the class be able to fairly and adequately protect the interests of the class. The court noted that the fact that an individual was a walk-on does not mean that that player would have received a scholarship “but for” the NCAA rule. This fact set up an internal conflict among the representatives of the class and other class members. For example, if it were determined that a school, in the absence of the regulation, would have offered twenty additional scholarships, the class members would then compete to determine which of them would have been awarded the scholarships. Although the court did not certify the class, the underlying antitrust issue remains undecided. Specifically, the NCAA avoided the requirement of showing that limiting the number of scholarships is necessary for the production of college football as a competitive product. In fact, to date it has avoided any kind of empirical showing that its use of monopsony power increases competition in the sport entertainment market.

9.5.2  Undercompensation: White v. NCAA In White,20 the issue was not the number of scholarships, but the amount paid per scholarship to football and basketball players since February 2002. The plaintiffs claimed that collusion among NCAA members led to a grant-in-aid (GIA) cap that was less than the actual cost of attending college. Consequently, they sought the difference between the actual costs of attending college and the GIA cap. The plaintiffs, perhaps wisely, avoided the bigger issue of whether players should be paid at levels that would prevail in the absence of any collusion at all. Had the issue been framed in this matter, it is quite likely the NCAA would have prevailed under the ancillary restraints analysis described previously. Although the class was certified, as with the limitation on the number of scholarships, a judicial decision on whether the cap violated the antitrust laws was avoided. This time, though, it was the result of a settlement between the NCAA and the class. Although denying any wrongdoing, the NCAA agreed to make available a fund totaling $218 million for the purpose of aiding student-athletes with “demonstrated financial and/or academic needs.” In addition, a $10 million fund was established to reimburse former players for a variety of expenses to be incurred in the future in connection with career development. the class members and the representative parties will fairly and adequately protect the interests of the class. 20 No. CV06–0999 (C.D. Cal. 2008).

Antitrust Challenges to the NCAA Monopsony

199

9.5.3  Numbers of Coaches: Hennessey v. NCAA At a special session in 1975, the NCAA members agreed to limit the maximum number of assistant football and basketball coaches.21 This restriction applied only to Division I schools. As a result of the restriction, Hennessey was demoted to part-time status after being an assistant coach for sixteen years. His compensation was reduced by 90 percent from $20,000 to $2,100. This cost saving for the university was a crippling financial blow to Hennessey. In response, Hennessey filed an antitrust suit alleging that the NCAA’s agreement on the number of coaches violated Section 1 of the Sherman Act. The court found that the NCAA Bylaw limiting the number of coaches resulted from an agreement among the member institutions that obviously restrained trade because they agreed not to compete by hiring more coaches. But the question was whether the agreement constituted an unreasonable restraint of trade. In this connection, the court noted an absence of specific intent to injure the named plaintiffs or other assistant coaches, either individually or as a group. This observation reveals a lack of understanding of supply. If the supply function of assistant coaches has a positive slope, an agreement to reduce the number of assistant coaches hired has two effects. First, some coaches will lose their jobs entirely or be demoted as Hennessey was. Second, there is movement along the supply curve to a lower wage for those who remain employed. As a consequence, the record may be devoid of any evidence that this was the unabashed intent of the restriction, but it nonetheless is true that the restraint would have an adverse effect on the community of assistant coaches. The fact that the NCAA members apparently were concerned about the welfare of those whom they knew would be hurt seems to have influenced the court.22 To see how silly this is, one need only ponder whether similar expressions of concern would get price fixers in the output market off the hook. The court clearly understood the nature and purpose of the NCAA’s endeavor: “Bylaw 12–1 was … intended to be an ‘economy measure.’ In this sense, it was both in design and effect one having commercial impact.”23 What then were the redeeming virtues of the restraint that saved it from being deemed “unreasonable”? Hennessey v. NCAA, 564 F. 2d 1136 (5th Cir. 1977). “There was, to be sure, an awareness by the members of the association that there would be some adversely affected by the Bylaw, but the attitude expressed was one of concern, not indifference.” Hennessey, 564 F. 2d at 1153. In this court, at least, a good attitude saves bad deeds. 23 Id. 21 22

200

The NCAA

Apparently, some member institutions were experiencing financial difficulty. Those colleges with athletically and economically more successful programs were seen as taking (unfair) advantage of their success by expanding their programs. This sounds very much like more output or higher quality output is a bad thing. The acknowledged concern was that the less successful programs were having a hard time catching up to and keeping up with the more successful programs. Success had to be harnessed so all could enjoy being mediocre. This is not to say that competitive balance is undesirable, but the goal should be to raise quality to a common level. It should not be to reduce the quality of superior programs so inferior programs can compete. The NCAA has had a measure of success in peddling the line that it is concerned with preserving “amateurism.” This, of course, is a sham. Everything about Division I football and basketball is big business. The only amateurs (arguably) are the players. As for Hennessey and his fellow coaches, restricting their number was deemed necessary to preserve “the competitive and amateur nature of the programs.”24 Incredibly, the court found that “the fundamental objective [of the bylaw] was to preserve and foster competition in intercollegiate athletics by curtailing … potentially monopolistic practices by the more powerful and to reorient the programs into their traditional role as amateur sports operating as part of the educational process.”25 But there was no proof that this restraint would save intercollegiate athletics. And the court did not demand any such proof. One would think that the NCAA should have had to shoulder the burden of proving that a plainly anticompetitive restraint would have redeeming features that result in net social benefits. It should not have been allowed to point to some vague ideal of preserving amateurism. But this court was “of the view admittedly bordering on speculation that the Bylaw will be of value in achieving the ends sought … .”26 In effect, the court was willing to trade Hennessey’s undeniable injury for some pie-in-the-sky benefit. Finally, the court fashioned a “ruinous competition” defense for the NCAA members. The restraint, according to the court, may have beneficial effects for the assistant coaches by preserving employment opportunities. Without the restraint – and the consequent cost saving – financially strapped athletic programs might have shut down. Thus, the restraint would have the long-run effect of increasing the number of potential employers above the free-market level. Id. Id. 26 Id. 24 25

Antitrust Challenges to the NCAA Monopsony

201

9.5.4  Coaches’ Compensation: Law v. NCAA In spite of a few methodological slips along the way, Judge Vratil got to the correct result in Law v. National Collegiate Athletic Association27 and sharply criticized the Hennessey case in the process. This case involved a collusive effort on the part of the NCAA and its members to reduce the salaries paid to certain assistant coaches. The explicit purpose of the resultant NCAA Bylaw 11.02.3 was to stabilize and depress the compensation of the coaches designated as restricted-earnings coaches. At least since 1940, such interference with the price mechanism has been illegal per se.28 In this case, however, the court engaged in a rule of reason analysis to determine whether the restraint was unreasonable after considering all of the anticompetitive and procompetitive effects. Apparently, a large number of NCAA members were still experiencing financial difficulties of one sort or another in the late 1980s. In 1989, the NCAA formed a Cost Reduction Committee to explore alternative ways of reducing the costs of the intercollegiate athletic programs. The goal was to reduce cost without sacrificing competitive balance or curtailing access to higher education by student-athletes. At the same time that some schools were eliminating certain nonrevenue sports, there were pressures to spend even larger sums to recruit the most talented athletes and coaches. To do otherwise would have meant becoming less competitive on the field. The NCAA decided that a collaborative effort to reduce costs was necessary. Of course, each school could have acted unilaterally to reduce its costs in any number of ways – reducing the number of sports offered, the number of scholarships, the value of a scholarship, the number of coaches, the salaries of the coaches, travel budgets, equipment budgets, and the like. But a unilateral cost reduction effort, if unmatched by rival schools, would have created an uneven playing field. The cost cutter would have become less competitive on the field than its rivals that did not reduce costs, and competitive balance would have been impaired. As a school began to lose on the field, spectator interest would wane, and revenues would fall thereby worsening the school’s financial position. The net result might not be improved financial viability if revenues fell by more than the cost savings. 902 F. Supp. 1394 (1995). In United States v. Socony-Vacuum Oil Co., 310 U.S. 150 (1940), the Supreme Court held that “[a]ny combination which tampers with price structures is engaged in an unlawful activity.” Later, the Court pointed out that this applied to depressing prices as well as raising them.

27 28

202

The NCAA

The NCAA’s Cost Reduction Committee decided that cost savings could be realized by reducing the number of coaches in all Division I sports. In addition, in every sport other than football, one coach would be designated a restricted-earnings coach. A restricted-earnings coach’s compensation could not exceed $12,000 during the academic year and $4,000 during the summer. Some of the coaches who were affected had been earning $60,000 to $70,000 annually. The adverse impact on these coaches was undeniable and obvious for all to see. The court began its analysis by noting that an agreement fixing the maximum salaries that will be paid to one class of employees attracts antitrust attention. In fact, in most circumstances, one would expect such an agreement to be a per se violation of Section 1 of the Sherman Act. But, as noted earlier, the NCAA enjoys rule of reason treatment because the NCAA and its members sell competition: athletic contests between rival schools. Thus, the court conceded that the NCAA plays a vital role in making intercollegiate athletic events available to the public and, therefore, its restraints should be examined for their reasonableness. At the same time, however, the court expressed its belief “that the Supreme Court [did not intend] to give the NCAA carte blanche in imposing restraints of trade on its member institutions or other parties because of its role in the marketplace. There can be no doubt that the NCAA is subject to the antitrust laws … .”29 It is clear on its face that the restricted-earnings coach rule had an adverse anticompetitive effect in the market. Some of the restricted-earnings coaches had earned salaries of $60,000 to $70,000 prior to the rule,30 which reduced their maximum compensation to $12,000 to $16,000. Presumably, the higher salaries resulted from market forces that were blunted by the agreement. Thus, the adverse anticompetitive impact of the rule was apparent to the court: Because the Restricted Earnings Coach Rule specifically prohibits the free operation of a market responsive to demand and is thus inconsistent with the Sherman Act’s mandates, it is not necessary for the Court to undertake an extensive market analysis to determine that the rule has had an anticompetitive effect on the market for coaching services.31

The NCAA responded by arguing that the rule was necessary to maintain “a level playing field in the sports arena, retaining and fostering the spirit 902 F. Supp. at 1404. Id. at 1405. 31 Id. at 1395. 29 30

Antitrust Challenges to the NCAA Monopsony

203

of amateurism … and protecting NCAA member institutions from self-imposed, ruinous cost increases.”32 In part, the NCAA pointed to its success in Hennessey, which we may recall challenged an NCAA bylaw restricting the number of football and basketball coaches that member institutions could employ. There, the court found that the plaintiffs failed to prove that the intent was to hurt them. The court found that the NCAA’s motivation was more noble: “to preserve and foster competition in intercollegiate athletics … and to reorient the programs into their traditional role as amateur sports operating as part of the educational process.”33 But it is the athlete – not the coach – who is supposed to be an amateur. It is not apparent that the rule at issue in Law would level any playing field or contribute to competitive balance. It would, of course, tend to weed out competent, experienced coaches and tend to weaken the quality of the coaching staff. This would tend to reduce the quality of the athletic performance. Since this is what the NCAA sells, the rule would reduce the quality of the output and thereby reduce social welfare. In Law, the court found that the restricted-earnings coach rule may be intended to further some legitimate goals of the NCAA, but the NCAA failed to do more than assert that the rule would, in fact, further these goals. No evidence was presented to support the NCAA’s claims. Given this finding, the court did not have to consider whether the rule was the least restrictive way of achieving the stated goals. Accordingly, the court found in favor of the plaintiffs.34 From an economic perspective, the Law and Hennessey decisions are ­inconsistent. According to Hennessey, the NCAA members can agree to ­reduce the number of coaches hired, which will have a predictable effect on coaches’ salaries. According to Law, the NCAA members cannot agree on a salary reduction, which would also have a predictable effect on coaches’ salaries. The better reasoning here is that of the Law court. The Hennessey court relied to a large extent the NCAA’s purported efforts to preserve amateurism. According the to the Law court, however: [T]he Hennessey court placed undue emphasis on the NCAA’s stated good intentions – not only in light of the Supreme Court’s clear articulation of the proper antitrust analysis [announced after Hennessey] but also in light of clearly established precedent existing at the time. … Although intent is relevant in helping a court Id. at 1406. 564 F. 2d at 1153. 34 The jury awarded over $22 million to 1,900 restricted-earnings coaches. This sum was automatically tripled to about $67 million plus the lawyer’s fees. The member schools had to chip in to pay the tab for their ill-fated efforts to control costs. 32 33

204

The NCAA

interpret facts and predict consequences in this type of case, ‘a good intention will [not] save and otherwise objectionable regulation … .’35

9.6  Concluding Remarks In this chapter, we have examined the cartel behavior of the NCAA. Our focus has naturally been on the NCAA’s monopsonistic behavior. Unlike most cartels, the NCAA has the organization to coordinate its member’s activities. Its bylaws formalize the agreements that it reaches – violating the bylaws (i.e., cheating on the cartel agreement) can be punished severely by the NCAA, which enhances cartel stability. As the cases demonstrate, the NCAA is not immune to antitrust challenge, but it is subject to the rule of reason in instances in which other groups of firms would be subject to the per se rule. This is because at least some anticompetitive measures are necessary to produce a viable product that will compete in the entertainment market. 902 F.Supp. at 1408, quoting U.S. v. Chicago Board of Trade, 246 U.S. 231, 238 (1918).

35

T EN

Monopsony in Action Physician Collective Bargaining: Monopoly or Bilateral Monopoly

10.1 Introduction Managed care, a source of significant buying power, has had a substantial impact on markets in which health care providers operate. For example, buying power in the market for pharmaceuticals has led to discounts to the more powerful buyers.1 Hospital buying cooperatives can force down the price of essential supplies. Most prominently, however, physicians have seen their reimbursement rates fall and their practice autonomy erode. This final chapter in the monopsony in action series focuses on physicians and their responses to buying-side pressures. The topic is especially interesting because different state’s laws may affect this relationship differently, leading to varying economic outcomes. In some instances, physicians have tried to join forces in order to negotiate collectively as a way of gaining some bargaining power. These efforts, however, have been thwarted by the antitrust laws, which forbid collaborative efforts that restrain trade.2 There is, of course, a specific antitrust exemption for labor unions,3 but most physicians do not qualify for the labor exemption because they are not employees.4 Instead, they are independent 1



2



3

4



This chapter draws from Roger D. Blair & Kristine L. Coffin, Physician Collective Bargaining, State Legislation, and the State Action Doctrine, 26 Cardozo L. Rev. 1731 (2005). See Frederick M. Scherer, How US Antitrust Can Go Astray: The Brand Name Prescription Drug Litigation, 4 Int’l J. Econ. Bus. 239 (1997); Jeffrey L. Harrison. The Brand Name Prescription Drug Litigation: Comments on Scherer, 4 Int’l J. Econ. Bus. 265 (1997). See Chapter 2 for a brief examination of Section 1 of the Sherman Act. Section 6 of the Clayton Act, 15 U.S.C. §17, originally provided the basic exemption for labor. Due to narrow interpretations by the Supreme Court (e.g., Duplex Printing Press Co. v. Deering, 254 U.S. 443 (1921)), Congress passed the Norris-La Guardia Act, 29 U.S.C. §§ 101–30, to reassert labor’s broad exemption from antitrust scrutiny. Although some 43.3 percent of all physicians are, in fact, employed and therefore eligible for the labor exemption, only 6 percent of all physicians are actually unionized. See Edward B. Hirschfeld, Physicians, Unions, and Antitrust, 32 J. Health L. 43 (1999).

205

206

Monopoly or Bilateral Monopoly

professionals who are largely self-employed.5 When physician groups are not horizontally integrated and have been formed simply to bargain with health insurers over reimbursement issues, they have been challenged by the antitrust authorities.6 In an effort to gain some bargaining power without running afoul of the antitrust laws, physicians have sought legislative protection. At the federal level these legislative efforts have failed.7 At the state level, however, physicians have won a few battles. Alaska, New Jersey, and Texas have all passed legislation that authorizes collective negotiations by physicians under some circumstances, which will be examined later.8 The Federal Trade Commission has uniformly opposed these legislative efforts on the grounds that they will increase the costs of health care services.9 In some cases, the FTC’s concerns are well founded, but in others, the states have taken the right course. The important issue is defining the circumstances when it is beneficial to offset monopsony power. When a state authorizes conduct that would otherwise violate federal antitrust law, its legislation must satisfy the state action doctrine, which can be summarized succinctly as follows: “there can be no [antitrust] immunity without (1) adequate and active public supervision and (2) a clear state purpose to displace competition.”10 This chapter analyzes the Alaska, New When first examining the application of the antitrust laws to professions, there was some suggestion by the U.S. Supreme Court that they would be subjected to a different analysis than other businesses. This has not developed in any substantial way. See E. Thomas Sullivan & Jeffrey L. Harrison, Understanding Antitrust and Its Economic Implications, 5th ed. 130 (New York: LexisNexis 2009); Jeffrey L. Harrison, Price Fixing, the Professions and Ancillary Restraints: Coping with Maricopa County, U. Ill. L. Rev. 925 (1982). 6 For example, the Federal Trade Commission complained that North Texas Specialty Physicians was engaged in fixing reimbursement rates among its ostensibly competing members. See In the Matter of North Texas Specialty Physicians, FTC File No. 021-0075 (Sept. 17, 2003), http://www.ftc.gov/os/adjpro/d9312/index.shtm. For similar complaints, see In re Memorial Hermann Health Network Providers, FTC File No. 031–0001 (Nov. 25, 2003), http://www.ftc.gov/os/caselist/0310001/0310001.shtm and In re Piedmont Health Alliance, Inc., FTC File 021-0119 (Aug. 11, 2004), http://www.ftc.gov/os/adjpro/d9314/ index.shtm. 7 For an extensive analysis of the proposed federal legislation, see Roger D. Blair & Jill Boylston Herndon, Physician Cooperative Bargaining Ventures: An Economic Analysis, 71 Antitrust L. J. 989 (2004). 8 Alaska, Senate Bill 0037; New Jersey, Assembly Bill 2169; Texas, Senate Bill 1468. Several others have seriously considered similar measures; see, e.g., Washington House Bill 2360 and District of Columbia Bill 13-333. 9 See, e.g., the FTC’s letter to the Texas Legislature on Senate Bill 1468 (May 13, 1999), http://www.ftc.gov/be/v990009.htm. 10 Keith N. Hylton, Antitrust Law:  Economic Theory and Common Law Evolution 372 (Boston, MA: Cambridge University Press 2003). 5

Economic Rationale of the State Legislation

207

Jersey, and Texas statutes and concludes that they satisfy the requirements of the state action doctrine, albeit in an unusual fashion. In addition, it proposes a methodology for assessment of monopsony conditions in the health care context.

10.2 Economic Rationale of the State Legislation Legislatures in Alaska, New Jersey, and Texas have now made it possible for physicians to bargain collectively.11 The economic rationale for the state legislation is the alleged monopsonistic exploitation of physicians by the managed care plans. Physicians allege that health insurers do not buy as competitors would purchase, but as monopsonists would. This means setting reimbursement rates below the competitive level, which leads to fewer services being provided to patients. Alternatively, it could lead to reduced quality as physicians may spend less time with each patient. The state legislatures, however, have been convinced that health plans enjoy at least some degree of monopsony power in some geographic markets. Monopsony power simply means that the supply of physician services is positively sloped.12 The health insurer then maximizes its profit by stopping short of the point where supply and demand are equal. As described in Chapter 3, the monopsonist health insurer will expand purchases of physician services until marginal factor cost equals the demand. It is clear that profit maximization leads the monopsonist to employ fewer units of service and thereby pay a lower price, which is precisely what the physicians claim. In addition, profit maximization leads to fewer health care services being provided to patients. The public policy concern is over the adverse effect on social welfare. If monopsony is, in fact, the actual market structure, then the physician complaints about reduced reimbursement rates and the legislative concerns about the availability of health care services are well founded. As described in Chapter 3, it is important to keep in mind that the reduced reimbursement rates do not result in lower prices charged by a monopsony insurer. In most instances, a reduction in an input price reduces the employer’s marginal cost and thereby leads to an expansion in the optimal output. In the New Jersey, for example, found that: “Inadequate reimbursement and other unfair payment terms offered by carriers adversely affect the quality of patient care and access to care by reducing the resources that physicians and dentists can devote to patient care and decreasing the time that physicians and dentists are able to spend with their patients.” 12 More formally, if the price is an increasing function of the quantity purchased, then w = w(q) and dw(q)/dq > 0. As a result, if the dominant buyer reduces q, then w will fall. This control over price through purchase decisions is the essence of monopsony power. 11

208

Monopoly or Bilateral Monopoly

case of monopsony, however, this is not what happens. Monopsony power causes the marginal cost curve to shift upward and the average cost curve to shift downward. Thus, the monopsonist earns more profit because average cost declines, but finds it optimal to curtail input purchases and, therefore, reduce output.13

10.3  The Effect of Cooperative Bargaining When the market for health care services is competitive on the supply side and monopsonistic on the buying side, cooperative bargaining by the health care providers will have beneficial results for consumers. Indeed, the health care providers will behave as a cartel, or a collusive monopoly, but the economic results will be positive. The consequent market structure is a bilateral monopoly (i.e., (collusive) monopoly on the selling side and monopsony on the buying side). The Alaska, New Jersey, and Texas legislation authorizes physicians to behave as a collusive monopoly and thereby create a bilateral monopoly market structure when monopsony is present. Since two wrongs seldom make a right, the legislation may seem counterintuitive, but introducing monopoly to countervail monopsony actually improves matters in terms of social welfare.14 The economic analysis of bilateral monopoly was examined in Chapter 6. In this setting, there will be cooperation between the monopolist and the monopsonist that leads to larger profits than are available absent the cooperation. As a result, cooperation can improve the lot of both parties. Interestingly, this cooperation is also good for the patients.

10.4  Reaction of the Federal Trade Commission The Federal Trade Commission, in reviewing the state legislation, expressed grave concerns regarding the potential economic effects of the bills. In its review of the New Jersey bill, for example, the FTC warned that the exemption from the antitrust laws would potentially increase the cost of receiving health care, while limiting the availability of health care services and ­reducing the quality of the health care being provided.15 Since These results are shown in more detail in Roger D. Blair & Richard E. Romano, Collusive Monopsony in Theory and Practice: The NCAA, 42 Antitrust Bull. 681, 686–9 (1997). 14 It would be even better to dissolve the monopsony, but if that is not feasible, introducing monopoly does improve social welfare. 15 Richard Feinstein, Federal Trade Commission. Letter to Chairwoman Charlotte Vandervalk Re. New Jersey Assembly Bill 2169, May 10, 2001. 13

Reaction of the Federal Trade Commission

209

these statutes would allow price fixing, the FTC warned that price fixing among competing physicians would, in fact, occur to the detriment of the patient. According to the FTC, the states’ misconception about the need for physician bargaining power stems from a faulty analogy to labor unions generally. The mechanic’s labor union, the FTC points out, is allowed to bargain for better wages, but this means it is not used to ensure better quality of mechanic services. It is necessary for the government to impose other standards to make sure that mechanic services are of high quality. In the same way, the FTC reasons, collective bargaining by physicians would not provide better health care quality for consumers, but only allow for the better payment of physicians. The FTC is not entirely correct. First, it is true that physicians are interested in better payment, but as we have just seen, that will lead to improved access and availability in the presence of monopsony. To the extent that physicians spend more time with each patient at higher reimbursement rates, quality may also improve. To be fair, however, it is true that the state statutes do not directly address the quality issue.16 In its review of the Alaska bill, the FTC warned that legislation leading to price fixing by physicians would reduce the quality of patient care and the benefits received while increasing prices. The FTC asserted that allowing collective bargaining would result in giving the physicians monopoly power, which, of course, is correct. It claimed that such joint negotiations are not necessary because the antitrust laws allow currently for physician groups that have already been established to discuss fee and non-fee issues.17 This is somewhat disingenuous, however, for two reasons. First, it only applies to practice groups that are horizontally integrated as a result of practice efficiencies. Second, even in that event, if the group comprises too large a percentage of all physicians, it will be challenged.18 But the FTC ignores these issues and simply asserts that physicians do not need group representation to negotiate effectively with health care plans. If, however, the health insurers enjoy monopsony power, the FTC is simply wrong:  An independent provider cannot bargain effectively. It is also possible that when physicians feel pressure to spend less time with patients, they are more likely to prescribe expensive diagnostic tests or practice defensive medicine. This too can raise the cost of medical care more generally. 17 Ted R. Cruz, Federal Trade Commission. The Threat of Consumer Harm Resulting from Physicians Collective Bargaining Under Alaska Senate Bill 37 (March 22, 2002), sec. III, http://www.ftc.gov/be/hilites/cruz020322.shtm. 18 See 1996 Department of Justice and Federal Trade Commission Enforcement Policy in Health Care, Statement 8, http://www.usdoj.gov/atr/public/guidelines/0000.htm# CONTNUM_61. Accessed date: April 19, 2009. 16

210

Monopoly or Bilateral Monopoly

The FTC is also concerned about the basis for inferring the presence of monopsony power. For example, the FTC contends that the Alaska statute does not provide sufficient checks to ensure against the physicians being allowed to form a cartel when one is not warranted (i.e., when the health plan has no monopsony power). According to the FTC, a risk involved in implementing the Alaska bill is that a health benefit plan may feel pressured to negotiate with the physician group and allow the physicians to set the prices. Faced with higher costs, the health benefit plan would then be compelled to pass the higher prices on to the consumer.19 The FTC also notes that the Alaska bill, while restricting the number of physicians who are allowed to collaborate, does not specify the types of physicians. This gives rise to a concern that all physicians with a certain specialty in a local market could jointly negotiate as a collusive monopoly. This would give the health benefit plan no choice but to work with them. With the attorney general as the only check on this possible monopoly formation, the FTC does not believe that he or she can render a satisfactory decision on whether the proposed agreement is fair.20 There are two related reasons for this view. First, the FTC does not believe that the Alaska bill requires that enough information be provided to the attorney general’s office. Second, the FTC does not believe that the bill affords enough times for the attorney general to process the information and conduct its own analysis.21 The FTC concludes from this that the attorney general could not provide appropriate supervision over these contracts, but as we argue later, the attorney general does not need to supervise in the sense of managing the process. The Texas bill, according to the FTC, has much the same problem. The FTC states in its review that the bill does not contain adequate standards for the attorney general to decide what constitutes substantial market power. The FTC also warns of the effects on cost that certain nonfee issues could impose. For example, the New Jersey statute characterizes clinical practice guidelines as a nonfee issue. But the increase in the cost of implementing the improved clinical practice guidelines could be very significant to the physician, the health care plan, and the consumer. There are no criteria in the legislation for the attorney general to use in evaluating the contracts that must be approved. Without having some guidance as to what should be allowed and what could promote unwarranted monopoly among In the absence of monopsony power, this is the correct economic analysis. These criticisms are well taken, but they can be remedied without much trouble. In Section 10.7, we outline the economic analysis that an attorney general should do to identify the circumstances that warrant antitrust immunity. 21 These may be valid concerns, but if experience shows that there are problems, the legislation can be amended to correct any deficiency. 19 20

State Action Doctrine

211

physicians, the FTC warns against using such vague legislation that would result in more expensive health care for the consumer.

10.5 Economic Rationale of FTC Concerns The FTC’s economic analysis of the limited exemptions contained in the Alaska, New Jersey, and Texas legislation is correct if the underlying assumptions of the FTC’s analysis are appropriate. In the FTC’s view, the market for health care services is competitive – on the buying side as well as on the selling side. When this is the case, the free play of market forces will lead to the competitive price and quantity of health care services. In the FTC’s world, the equality of supply and demand results in a price and output that maximizes social welfare. If physicians are given the freedom to bargain collectively with competing health plans and other buyers, they will be able to command the collusive profit-maximizing price and supply the corresponding quantity. This solution involves a reduction in social welfare since price rises and quantity falls. Thus, if the FTC is correct about the underlying market structure, then its concerns are well founded. If the underlying assumptions are incorrect, however, then the FTC’s advice to the state legislators is flawed. In any event, this is an empirical question, and the answer may vary from market to market within each state. It is the responsibility of the attorney general to limit collective negotiations to those markets where monopsony is a problem.

10.6 State Action Doctrine The state action doctrine22 permits states to legislate immunities from the federal antitrust statutes, but only under fairly stringent circumstances. It is a matter of some dispute whether the Alaska, New Jersey, and Texas statutes satisfy the requirements of the state action doctrine. The FTC contends that the statutes do not provide for adequate supervision. The New Jersey legislature, at least, has a decidedly different view: It is the intention of the Legislature to authorize independent physicians and dentists to jointly negotiate with carriers and to qualify such joint negotiations and related joint activities for the State-action exemption to the federal antitrust laws through the articulated State policy and active supervision provided under this act.23 For a compact survey, see Herbert Hovenkamp, Federal Antitrust Policy:  The Law of Competition and Its Practice, 2d ed. 721 (Minneapolis, MN: West Publishing Co. 1999). 23 Legislative finding (n) of the New Jersey statute. 22

212

Monopoly or Bilateral Monopoly

From an economic perspective, the states have satisfied at least the spirit of the state action doctrine.

10.6.1  Background In Parker v. Brown,24 the Supreme Court explained that the federal antitrust laws did not completely preempt state sovereignty in all circumstances.25 But state action that permits or mandates conduct that would be a clear violation of the Sherman Act is not necessarily permissible. There is an understandable concern that a state statute may “authorize unsupervised private conduct that impairs competition.”26 In Midcal,27 the Supreme Court explained that two requirements must be met for federal antitrust immunity. First, it must be unambiguous that the state fully intended to replace competition in the market with a regulatory scheme. Accordingly, the Court requires clear articulation and an affirmative expression of the state’s intent to regulate economic activity rather than rely on competitive market forces.28 The second requirement demands adequate and active supervision by the state. This requirement is designed to ensure that the results of the antitrust immunity are consistent with the state’s goals in creating the immunity in the first place.

10.6.2  Clear Articulation Alaska, New Jersey, and Texas have clearly articulated their competitive concerns regarding the buying power of some large health benefit plans in some geographic areas. Each state has explicitly recognized that collective negotiations may be anticompetitive in some (or even most) instances, but there are economic circumstances in which collective negotiations will have procompetitive consequences.29 It is only under those circumstances that collective negotiations are permitted. The legislative findings in the Texas 317 U.S. 341 (1943). If the authorized conduct would otherwise be a clear violation, the Court imposes restrictions on permissible legislation. If the statute authorizes conduct that would not be a per se violation of the Sherman Act, then the Court usually finds that it is not preempted by the federal law. See Hovenkamp, supra note 23, at 723. 26 Hovenkamp, supra note 23, at 723. 27 Cal. Retail Liquor Dealers Ass’n v. Midcal Aluminum, Inc., 445 U.S. 97 (1980). 28 Id. at 105. 29 New Jersey Assembly Bill 2169, C.52:17B-196; Alaska Senate Bill No. 37, Sec. 23.50.010 (b); Texas Senate Bill No. 1468, Art. 29.01. 24 25

State Action Doctrine

213

statute illustrate an understanding of the economic analysis that we presented earlier: The legislature finds that joint negotiation by competing physicians of certain terms and conditions of contracts with health plans will result in procompetitive effects in the absence of any express or implied threat of retaliatory joint action, such as a boycott or strike, by physicians. Although the legislature finds that joint negotiations over fee-related terms may in some circumstances yield anticompetitive effects, it also recognizes that there are instances in which health plans dominate the market to such a degree that fair negotiations between physicians and the plan are unobtainable absent any joint action on behalf of physicians. In these instances, health plans have the ability to virtually dictate the terms of the contracts they offer physicians. Consequently, the legislature finds it appropriate and necessary to authorize joint negotiations on fee-related and other issues where it determines that such imbalances exist.30

Thus, the legislature clearly contemplated a limited departure from competition among physicians to offset the buying power of large managed care plans. The articulation could hardly be clearer than this.

10.6.3  Active State Supervision When a state substitutes regulation for compitive market forces, it usually establishes a regulatory agency to protect the public interest. Most familiar are the public service (or utility) commissions that regulate natural moetnopolies – electric power transmission, local telephone service, natural gas transmission, and the like. In the case of large managed care plans, they are subject to the usual insurance regulations regarding reserves, premiums charged, contract forms, and the like, but the states have not attempted to regulate the prices that insurers pay for inputs, which include the fees they offer to health care providers. In order to ensure the reasonableness of reimbursement rates, the states have authorized collective negotiations between physicians as a group and large managed care plans. In a sense, the states are not replacing market forces with regulation. Instead, they are altering the market structure to obtain the benefits of competition that are unavailable in the presence of monopsony power. As we have already seen, collective negotiations among independent physicians can be socially desirable or undesirable depending upon the underlying market structure. Thus, active Article 29.01 of the Texas Senate Bill 1468. The language in the Alaska bill is nearly identical, Alaska Senate Bill No. 37, Sec. 23.50.010. Although the language of the New Jersey statute is somewhat different, the ideas are precisely the same. See Legislative finding (m) of the New Jersey statute.

30

214

Monopoly or Bilateral Monopoly

state supervision remains critical, but it does not take the form of traditional rate regulation. Instead, it is primarily in identifying those markets where collective negotiations are warranted and will improve social welfare by restoring competitive balance at the bargaining table. In those markets, public utility style regulation is unnecessary because economic forces will lead the parties to the socially optimal outcome. Each state relies upon its attorney general to supervise collective negotiations to protect the public interest. For its part, the FTC objects that this is not adequate state supervision. The FTC further contends that the regulatory scheme is not consistent with the standards set out by the relevant precedents. But the FTC apparently has the model of public utility regulation in mind. This is neither what the statutes envision nor what is necessary once the market structure is correctly identified. The precise role of the attorney general varies somewhat from state to state. In one important respect, however, the attorney general’s role is identical across states: ensuring that the underlying market structure warrants collective bargaining (i.e., that a competitive imbalance exists between independent physicians and managed care plans). The attorney general acts like a gatekeeper in the sense that he or she limits collective action to markets in which the health benefit plan has monopsony power. For example, the Texas bill requires prior approval of the attorney general, who must first determine that the health benefit plan has a substantial market share that is leading to unfair contracts.31 If such market power is found, then the bill allows for collective bargaining over fees for services, reimbursement factors, discounted prices for services, and the dollar amount of payment to physicians from health benefit plans.32 As we have seen, if the insurer does enjoy monopsony power, collective action by the physicians will lead to an improvement in social welfare, increased availability of health care services, and lower insurance premiums. In contrast, if the health benefit plan does not have monopsony power, collective bargaining makes things worse:  social welfare and availability decline while insurance premiums rise. Thus, it is critical that the attorney general get it right when assessing the power (or lack thereof) of the health benefit plan. If he or she gets this right, then there is no need to micromanage the bargaining process, and active state supervision of the public utility type is neither necessary nor desirable. If he or she gets it wrong, no amount of micromanaging will Texas Senate Bill No. 1468, Art. 29.09 (b). Texas SB 1468, Article 29.05, 29.06. Texas lists sixteen other nonfee issues that physicians may collectively bargain over without such limitations. Texas SB 1468, Article 29.04.

31 32

Diagnosing Monopsony Power

215

correct matters.33 Thus, it is imperative that the attorney general has full information and follows a sound analytical process in deciding whether to authorize collective negotiations.

10.6.4  Information Available When a group of physicians wants to negotiate collectively with a managed care plan, the physicians must select a joint bargaining representative. This representative is required to file a petition with the attorney general seeking approval. The petition must identify the physicians who will be in the group along with their business addresses. It must also reveal the percentage of all physicians in the insurance carrier’s service area that are in the group, the identity of the carrier, the subject matter of the proposed negotiations, the expected benefits of the joint negotiations, and the anticipated effects on the quality and availability of health care services.34 In addition, the attorney general may request further information or data. To aid the attorney general in analyzing the market structure on the buying side, New Jersey provides for the collection of market share data for each carrier in the state.35 The attorney general must act on the proposed group’s petition within 30 days.36 New Jersey’s statute instructs the attorney general to approve the petition if he or she finds that the proposed benefits outweigh the costs of any reduction in competition that joint negotiations may involve.37 As we have already seen, if the market structure warrants joint negotiations, there will be only benefits and no costs. Once again, identifying the market structure is vital, but we shall see that this is not just a matter of market shares.

10.7  Diagnosing Monopsony Power Since the presence of monopsony power is crucial for the authorization of joint negotiations, a methodology for diagnosing monopsony power is It should be apparent that we are not arguing that active state supervision is neither necessary nor important. On the contrary, active state supervision is crucial, but it occurs before the antitrust immunity is granted by authorizing collective negotiations. 34 New Jersey Assembly Bill 2169, C.52: 17B – 202; Alaska Senate Bill No. 37, Sec. 23.50.050 (d); Texas Senate Bill No. 1468, Art. 29.08. 35 New Jersey Senate Bill 2169, C.52: 17B – 199 (b); Alaska Senate Bill No. 37, Sec. 23.50.020 (d); Texas Senate Bill No. 1468, Art. 29.08. 36 The time limit is the same in Texas, but is 60 days in Alaska. See Texas Senate Bill No. 1468, Art. 29.09. Alaska Senate Bill No. 37, Sec. 23.50.020 (e). 37 Id. C.52: 17B – 203 (b); Alaska Senate Bill No. 37, Sec. 23.50.020 (f); Texas Senate Bill No. 1468, Art. 29.09 (b). 33

216

Monopoly or Bilateral Monopoly

critical. The following assesses the guidance provided by the state statutes. Then an economic road map for carrying out this vital step is offered.

10.7.1  Statutory Guidance In determining whether sufficient monopsony power exists, the states have given their attorneys general most of the responsibility and precious little guidance. In New Jersey, the attorney general, along with the commissioner of banking and insurance and the commissioner of health and senior services, must find that a carrier has substantial market power for any collaboration among physicians to be authorized. But the statute defines neither “market power” nor “substantial.” The Alaska bill provides a specific percentage of the physicians that may collaborate when the market power of the health benefit plan is small. It states that if the health benefit plan accounts for only 5 percent of the market, then no more than 30 percent of the physicians from that geographic region may collaborate. The attorney general, however, still has the authority to change these percentages if it is deemed necessary. The central directive in this bill is that collective negotiations should take place when there is an imbalance in bargaining power such that the health benefit plan is able to virtually dictate the terms of the contracts to the physicians. The Texas legislation provides the most flexibility. The Texas attorney general has complete discretion in determining whether the carrier has sufficient market power for collective bargaining to be permitted. The statute does advise that the attorney general should find that a health benefit plan’s terms and conditions have had a negative effect on patient care before authorizing collective negotiations. This, of course, can arise if the terms are so unfavorable that the availability of care is curtailed due to physicians exiting the market. In addition, it can occur if the health insurer unduly intrudes upon the physician’s discretion in treating his or her patients. But it does not occur simply because coverage in the health insurance policy is limited. Insurers generally provide financial benefits for procedures that they are paid to cover. If a particular policy does not cover a specific procedure, the insurer should not be expected to provide reimbursement for that procedure. This is not to say that disputes over the interpretation of the contract never arise or that the insurer is right when they do arise. Our point is simply that patients are often disappointed (and occasionally outraged) when they discover that their policy coverage is not as extensive as they would like it to be. But that is a matter of what was purchased and not a matter of market power.

Diagnosing Monopsony Power

217

The bottom line is that the state statutes instruct the attorney general to limit collective negotiations to those markets where monopsony power threatens the availability and quality of health care. No doubt, there could have been more specific guidance in these statutes, but the attorneys general presumably know that they will need some economic advice to make sure that they get this part right.38

10.7.2  Monopsony and Monopsony Power Since monopsony is a market structure in which there is a single (or a dominant) buyer, the first step in diagnosing monopsony power is to define the relevant product (or service) market and the relevant geographic market. The second step is to adopt a measure of monopsony power and assess the extent of monopsony power formed in the market in question. As we shall see, this can be a complicated process. 10.7.2.1  Analysis of the Relevant Product (or Service) Market The definition of the relevant product market should encompass all reasonable substitutes.39 In health care, the attorney general will want to focus on a relevant product market definition that relates to the group petitioning for permission to negotiate collectively. If the group offers a broad array of physician services, it may be appropriate to adopt the cluster of physician services as the relevant product.40 If the group is comprised of surgeons or psychiatrists, it would make more sense to adopt a narrow definition. The goal is to identify reasonable substitutable physician services so the attorney general can assess monopsony power over the relevant set of physician services. Moreover, this aspect of the legislation strengthens the case for concluding that the states knew what they were doing when they passed the legislation. The statutes do not provide blanket antitrust immunity; instead, they limit immunity to those markets where the attorney general finds a problem. 39 See materials, supra Chapter 4, Section 4.2. 40 Physician services are a collection, or cluster, of services that are not readily substitutable on the demand side. For example, patients cannot substitute vascular surgery for an appendectomy or for a diagnosis of flu-like symptoms. Nonetheless, the cluster of physician services includes all three. For the most part, this is not much of a problem because health benefit plans buy a wide array of physician services. For an excellent analysis of cluster markets, see Ian Ayres, Rationalizing Antitrust Cluster Markets, 95 Yale L. J. 109 (1985). 38

218

Monopoly or Bilateral Monopoly

10.7.2.2  Analysis of the Relevant Geographic Market In analyzing the relevant geographic market, the inquiry is largely the same:  identifying reasonable substitutes. The relevant geographic market includes all locations that are reasonably substitutable from the customer’s perspective. If price were to rise in a particular location, the question is where else could the buyer go for the product in question. For the attorney general, the definition of the relevant geographic market is a fact-specific inquiry that must be addressed on a case-by-case basis. The economic principles are the same, but the facts are bound to be different and, therefore, require individual analysis. Some of the statutory language suggests that the health benefit plan’s service area might be relevant. As an economic matter, however, using the insurance carrier’s service area as the relevant geographic market is not apt to provide sound results. An insurer will confine its area of coverage largely on the basis of demand for its insurance policies. This, however, may not coincide with the market for health care provider services. The attorney general should address the market definition exercise using principles that have been widely applied in antitrust cases. 10.7.2.3  Measuring Monopsony Power Once the attorney general has settled on a relevant market, he or she can turn to an assessment of monopsony power in that market. Again, absent monopsony power, it would be inappropriate to authorize collective negotiations. As explained in Chapter 3, monopsony power is the power that a dominant buyer has to control the price it pays by adjusting its purchases. The index of the dominant firm’s buying power (BPI) is a function of the dominant buyer’s share of all purchases (s), the elasticity of demand of the other buyers (η), and the elasticity of the sellers’ supply (ε): BPI =

s ε + η (1 − s )

Market share, which the courts rely on heavily, can be relatively easy to measure in practice. Its importance is also easy to understand intuitively. Monopsony power emanates from controlling the total quantity purchased and, therefore, the more important the large buyer, the more power he or she will have.

Diagnosing Monopsony Power

219

The elasticity of supply, which is a measure of the relative responsiveness of the quantity supplied to changes in price,41 is important (perhaps crucially so) in determining buying power. The greater the elasticity of supply is, the smaller the BPI is.42 Intuitively, this makes sense:  The greater the quantity reduction necessary to achieve any specific reduction in price is, the smaller a buyer’s ability to influence price is. In the limit, if the supply curve is horizontal, the elasticity will be infinite, and even a pure monopsonist (a single buyer) will have no buying power at all. Mathematically, this fact is apparent in the BPI expression because ε will be infinite for a horizontal supply curve, but the economic reason is that a reduction in the quantity purchased will not reduce price because price is not affected by quantity. This consideration may be crucial in practice when there is excess capacity on the supply side. Excess capacity means that additional output can be supplied without any increase in price. This is often the case in hospitals with low occupancy rates – vacant beds can be filled without an increase in price. For physicians, the elasticity of supply may be infinite if the physician’s time is not fully booked. In other cases, the elasticity of supply may be substantial albeit not infinite. If so, monopsony power will be negligible. Finally, the demand elasticity43 of other buyers is also important. The greater this elasticity is, the smaller the BPI is.44 Intuitively, this makes sense because when the dominant buyer restricts its purchases and thereby depresses price, other buyers expand their purchases and offset the initial reduction in quantity. As this demand becomes more elastic, the quantity response of other buyers increases, and the net effect on total quantity decreases, ensuring, therefore, a smaller net effect on price. In health care, the demand tends to be rather inelastic for two reasons: First, most physician services are necessities for the patient’s health and, second, most people have health insurance, which makes them less price sensitive. The elasticity of supply is given by ε = (dq/dp)(p/q), or the percentage change in quantity supplied divided by the percentage change in price. See, e.g., Michael Parkin, Economics, 5th ed. 98 (Boston, MA: Addison Wesley – Pearson 2000). 42 Analytically, the effect of a change in the elasticity of supply is given by the partial derivative of the BPI with respect to ε; ∂BPI / ∂ε < 0 . 43 The demand elasticity, which can be written as η = −( dQ/dP)(P/Q),  is the percentage change in the quantity demanded divided by the percentage change in the price. See Parkin, supra note 42 at 88. 44 Analytically, we find the effect of a change in the demand elasticity from the partial derivative of BPI with respect to η:  ∂BPI / ∂η < 0 . 41

220

Monopoly or Bilateral Monopoly

10.7.2.4  Importance of Entry Barriers Many courts have recognized that market power is fleeting in the absence of entry barriers. The economic reasoning is straightforward. When a buyer exercises market power, its profits rise. These profits, in turn, will attract entry of new buyers in the pursuit of higher returns. This entry will continue until competition is restored unless there is some barrier to the entry of these new firms. Thus, the attorney general must determine whether there are significant entry barriers that insulate whatever monopsony power exists. In the case of large managed care plans, entry barriers may be substantial because conditions must be ripe for entry into the health insurance market in order to provide competing buyers of health care services. But entry is not impossible. As reimbursement rates are depressed, other insurers will face lower claims costs. This reduces the costs of providing health insurance and will lead to premium reductions, which will amount to entry as new subscribers are enrolled. If there are firms that are authorized to conduct business in a state, but they are not currently doing so, entry is a matter of marketing the insurance policies. 10.7.2.5  Implications of the BPI The Alaska, New Jersey, and Texas statutes charge the attorney general with the responsibility for deciding whether the market structure warrants collective negotiations. As a result, the attorney general will need the time and the resources necessary to estimate the demand and supply elasticities required for a precise calculation of the BPI. If time and/or resources are short, there may be a tendency to rely too heavily on market shares. This would be unfortunate because market shares alone can be decidedly misleading. For example, suppose that a dominant buyer accounts for 50 percent of all purchases (s = 0.5) and the supply and demand elasticities are both equal to one (ε = 1.0 and η = 1.0). After substituting these numerical values, we find that the BPI equals 0.33. That is, the dominant buyer will be able to reduce price some 33 percent below the competitive level by restricting its purchases to profit-maximizing levels. But suppose that the supply elasticity is four (ε = 4.0) and the demand elasticity is two (η = 2.0). The same 50 percent market share then gives rise to only a 10 percent reduction in price below the competitive level. As this example illustrates, relying solely on market share data provides decidedly imperfect information regarding the extent of monopsony power. Incorrect inferences can be drawn, and serious policy blunders may result.

Diagnosing Monopsony Power

221

It should be clear that small market shares will not support a presumption of monopsony power. If a health benefit plan accounted for only 5 percent of all purchases of health care services, its influence on reimbursement rates would be minimal. Again, suppose that ε = 1.0 and η = 1.0. The BPI would then be equal to just over 3 percent if a health benefit plan had a market share of 5 percent. Larger elasticities of supply and demand would reduce the BPI even further.

10.7.3  Calculating Market Shares Even estimating the health insurer’s market share may be more complicated than it would appear at first blush. When the attorney general is determining whether the market structure warrants collective negotiations, he or she must consider all of the demands placed on the suppliers of health care services. This can be a bit complicated to estimate as the following numerical example illustrates. Suppose that an insurer accounts for 75 percent of all lives covered by private insurance. This would seem to meet conventional judicial thresholds for market power – at least on the selling side.45 Subject to the reservations expressed earlier regarding sole reliance on market shares,46 this may be correct in the private insurance market, but it is not correct on the buying side. In Florida, for example, the population was about 16 million in 2000.47 Of those, about 2.8 million were covered by Medicare, and another 2.0 million were covered by Medicaid. Roughly, 2.6 million were uninsured. Thus, the percent of the total population covered by the dominant health insurer is only 40.3 percent.48 Since all of the people in Florida demand health care services to one extent or another, a more meaningful market share would be around 40.3 percent rather than 75 percent. This, of course, is still a substantial number, but it is a far cry from 75 See United States v. E. I. duPont de Nemours & Co., 351 U.S. 377, 391 (1956). The precise number varies from case to case. See ABA Section of Antitrust Law, Antitrust Law Developments, 5th ed. 234–6 (Chicago, IL: American Bar Association 2002), for a brief survey with ample citations and examples. 46 Recall that market power depends on supply and demand elasticities as well as market share. Sole reliance on market share may yield spurious inferences. See discussion in Chapter 3. 47 This figure, as well as the rest of this numerical example, is based on more precise numbers contained in Susan S. Floyd, ed., Florida Statistical Abstract (Gainesville:  University of Florida 2003). The precise numbers are unimportant for the example in the text. 48 Since the total population is 16.0 million and the Medicare/Medicaid count is 4.8 million, we are left with 11.2 million. Of these, some 2.6 million are uninsured. Thus, the insured population is 8.6 million. The dominant insurer is assumed to have 75 percent of 8.6 million or 6.45 million. Now, 6.45 million is 40.3 percent of the total population. 45

222

Monopoly or Bilateral Monopoly

percent. Whether this market share is sufficient to warrant collective negotiations depends upon the values of the supply and demand elasticities.

10.8  Postnegotiation Review After the carrier and the physicians’ representative have reached an agreement, the representative must submit the proposal to the attorney general for approval. There is then a limited time, which varies by state, that the attorney general has to review the proposed contract and determine whether it is appropriate or unfair.49 There are no guidelines given in any of the bills stating how these contracts should be evaluated for their fairness, but the attorney general will know the purpose of the statute and presumably will use good judgment. At this point, fairness is a peculiar objective. Presumably, authorizing joint negotiations means that the attorney general has created a bilateral monopoly situation. The outcome of the bargaining process will lead to the social optimum. Given the existence of monopsony, nothing can be fairer for the consumer. Thus, fairness may refer to the split between the physician group and the insurer. But there is no economic reason for this to be of public policy concern.50

10.9  Concluding Remarks The legislatures in Alaska, New Jersey, and Texas were persuaded that access to high-quality health care services at reasonable prices was threatened by the monopsony power of large managed care plans. In response, they enacted legislation that charged the attorney general with the responsibility for authorizing joint negotiations among health care providers when the market structure indicated that it was warranted. The chapter has shown the market conditions that must be present for such authorizations to have the intended effect. It also provides an analytical approach to determining whether those market conditions are present. If the attorneys general follow the suggested approach, the active state supervision will occur ex ante and ex post monitoring can be confined to determining whether Texas SB 1468 Article 29.09; New Jersey AB 2169, 8; Alaska SB0037g sec. 23.50.020 (e). The third-party representative is responsible for paying a reasonable fee to the attorney general’s office that should be approximately equal to the cost of regulating and overseeing the collective bargaining proceedings. 50 One reason why the split may unfairly favor the managed care plan is that the health care providers cannot go on strike. See New Jersey Assembly Bill 2169; Alaska Senate Bill No. 37, Sec. 23.50.010 (a); Texas Senate Bill No. 1468, Art. 29.01. 49

Concluding Remarks

223

market conditions have changed such that joint negotiations are no longer necessary. The FTC’s assertions to the contrary notwithstanding, there is nothing inherently anticompetitive about the state legislation. If the attorney general does his or her job correctly, the results should be positive. But it is imperative that the attorney general does, in fact, do a good job in limiting collective negotiations to markets where monopsony power is present.

ELEVEN

Final Comments

11.1 Introduction Some twenty years ago, we began to incorporate monopsony into the general millieu of antitrust law and economics. Between then and now, we have continued that effort. In the preceding pages, we have attempted to summarize our work and that of others who have engaged in the same endeavor. The judicial understanding of and attitudes toward monopsony have become clearer in that time. Nevertheless, some uncertainties still remain. They can be traced to several issues discussed here. They provide the starting points for further research.

11.2  The Evolving Nature of Antitrust Law Antitrust law itself is constantly evolving in something akin to the common law tradition. This means that the true status of what one might think are well-settled rules is uncertain. A good example is provided by United States v. Topco,1 the 1972 decision ruling that horizontal market division is per se unlawful. In that case, the defendants had a plausible procompetitive justification for market division that might well have survived a rule of reason analysis. If the case had been decided ten years later, it probably would have been subject to such a searching analysis.2 Thus, while many antitrust scholars think Topco was wrongly decided, the Supreme Court has not said so and, in fact, relied on it in a more recent opinion.3 Nevertheless, Topco 405 U.S. 596 (1972). For a case adopting such a view, see Polk Brothers Inc. v. Forest City Enterprises, 776 F.2d 185 (7th Cir. 1985). See generally Martin Louis, Restraints Ancillary to Joint Ventures and Licensing Agreements: Do Sealy and Topco Logically Survive Sylvania and Broadcast Music? 66 Va. L. Rev. 879 (1980). 3 Palmer v. BRG of Ga., 111 S. Ct. 401 (1990). 1 2

224

Issues of Antitrust Injury and Standing

225

involves the type of practice that may receive different treatment as antitrust evolves.4 A second example is provided by vertical price restraints. Until recently, vertical price restraints were per se unlawful. Minimum resale price fixing (or resale price maintenance) had been illegal since the Supreme Court decided Dr. Miles in 1911.5 But that per se rule was overturned by the Supreme Court’s Leegin6 decision in 2007. A similar fate awaited the Albrecht7 rule, which outlawed maximum resale price setting. In its Khan8 decision, the Court overrule Albrecht and made maximum price fixing a rule of reason offense. In these instances, well-settled rules were overruled due to the weight of scholarly criticism. A third, and final, example is tying. As noted in Chapter 2, tying is, at least nominally, a per se offense. There is a great deal of scholarly and judicial resistance to this position based on (1) questions of whether tying is really likely to lead to allocative inefficiency and (2) the existence of procompetitive effects of some tying arrangements. The Court’s unbridled hostility toward tying was abandoned in the Hyde9 decision. When the Court turned its attention to the tying allegation in Kodak,10 many believed that the per se rule would be scrapped. But the Court reaffirmed the per se status of tying to the disappointment of many antitrust scholars. The problem, in sum, is that the economic revolution in antitrust is not, at least officially, complete. For the purposes at hand, this means the treatment of monospony power may change as antitrust law evolves. This evolutionary character of antitrust law provides ample opportunity for research on substantive antitrust doctrine.

11.3 Issues of Antitrust Injury and Standing A second area that is ripe for further research involves antitrust injury and standing. As the antitrust treatment of monopsony develops, questions regarding antitrust standing and injury must be addressed from a different point of view than that usually encountered. Recall that antitrust injury It can be argued that it has already. See Polk Bros. Inc., v. Forest City Enter., 776 F.2d 185 (7th Cir. 1985). 5 Dr. Miles Medical Co. v. John D. Park & Sons, 220 U.S. 373 (1911). 6 Leegin Creative Leather Prod. v. PKS, Inc., 551 U.S. 877 (2007). 7 Albrecht v. The Herald, 390 U.S. 145 (1968). 8 State Oil v. Khan, 522 U.S. 3 (1997). 9 Jefferson Parish Hosp. v. Hyde, 466 U.S. 2 (1984). 10 Eastman Kodak Co. v. Image Technical Serv. Co., 504 U.S. 451 (1992). 4

226

Final Comments

limits the scope of compensable losses, while antitrust standing limits those who are eligible to pursue private damages. As discussed in Chapter 7, courts are beginning to assess what these limitations on private actions mean when approached from the monopsony perspective. The impact of this analysis is important because raising or lowering the number of possible plaintiffs could have an impact on a firm’s decision with respect to alternative competitive strategies. One standing/injury issue to be resolved, as discussed in Chapter 7, concerns the array of possible plaintiffs in a collusive monopsony case. The obvious best candidates are those who sell to the colluding firms since they suffer the undercharge. We also noted the possibility of including buyers from the colluding firms. The logic is that the colluding firms buy less of the input and, therefore, must produce less of their output. The collusion does not simply restrict purchases but quickly begins to look like an agreement to limit output. This would be the case whether the colluding firms are competing sellers or not. But when they are competing sellers, the horizontal agreement on input price is, by implication, also an agreement to limit the output with the ultimate consequence being an increase in output prices and a consequent reduction in consumer welfare. This, of course, fits squarely with the conventional prohibition of price fixing by sellers. The research question then is whether buyers from the collusive monopsony should have standing to sue. How one answers this question should account for the impact on optimal deterrence but should not ignore the compensatory role of private damage actions. Given the extent of the economic ripples caused by the collusion to fix input prices, the question of whether purchasers from colluding buyers should have antitrust standing is complex. First, as we noted in Chapter 7, granting antitrust standing to this group of plaintiffs would be tantamount to finding that sellers of inputs to a collusive monopoly should have antitrust standing. This, as far as we know, has never been the case. Arguably, the analysis could stop there, but that avoids the substantive issue; perhaps suppliers to colluding sellers should have antitrust standing. Again the issue is how this would affect optimal deterrence and just compensation. The reasoning offered by the Supreme Court in its series of standing and antitrust injury cases does not provide a satisfactory answer. For example, it does appear that the customers of the colluding buyers have suffered antitrust injury. That is, the harm they suffer is a direct consequence of the anticompetitive acts of the buyers. Moreover, their injury can hardly be deemed indirect. They are direct purchasers from the collusive monopsony. Also, their harm is very different from the harm sustained by the sellers of inputs.

Issues of Antitrust Injury and Standing

227

The sellers suffer a loss in profits (producer surplus) due to the reduced purchases, while the consumers pay supra-competitive prices and suffer a loss in consumer surplus. The point is that the practical problems associated with permitting suits by indirect purchasers do not seem to come into play. One might counter these observations by concluding that these potential plaintiffs were only harmed as an incidental matter or were not “foreseeable” victims. But these reasons, while they might ultimately be the ones offered, really mask the actual substantive policy concerns. These concerns focus on the compensatory and deterrent functions of antitrust damages suits and the appropriate level of private enforcement. This should be the focus of the research effort. This is an issue fraught with confusion. For one thing, adherence to much of the Supreme Court’s guidance cuts in favor of finding that these buyers do have antitrust standing. First, in Illinois Brick, the Court viewed itself as choosing between compensating those who were victims of anticompetitive activity and deterring potential colluders.11 But in the case of customers of colluding buyers, the effect of allowing standing would serve both the compensation and the deterrence aims of damage awards. This raises the second concern. If antitrust standing were broadly defined, is there a risk of overdeterring price fixing?12 At first, the question seems nonsensical. How can there be an efficient level of something that is a per se offense? In fact, one can argue that the cost–benefit analysis that goes into the typical economic analysis of crime13 has been applied to price fixing with the conclusion that all price fixing is inefficient. All of these factors suggest that standing for buyers could hardly do anything other than enhance the benefits of private enforcement. As it turns out, though, there are complications. Labeling an activity “price fixing,” or at least the type of price fixing the antitrust laws should prohibit, is not a clear-cut exercise. Some activities that are literally “price fixing” actually may be procompetitive and, therefore, should be found lawful. In effect, there is uncertainty regarding the legality of some business practices. Increasing the pool of potential plaintiffs may discourage 431 U.S. 720, 746 (1977). The possibility that there is an efficient level of an activity that may or may not be illegal is an important matter in the economic analysis of law. It supposes that there are different sanctions for different activities. In antitrust, the sanction is pre set at treble damages for all illegal activities. There is no necessary connection, however, between trebling damages and achieving efficient outcomes. 13 See, e.g., Mitchell Polinsky, An Introduction to Law and Economics, 2d ed. 75 (Boston: Little Brown 1991); Richard Posner, The Economic Theory of Criminal Law, 85 Colum. L. Rev. 1193 (1985). 11 12

228

Final Comments

firms from engaging in activities that may be procompentitive, but fall within this gray area. Overdeterrence is, therefore, possible and undesirable. For example, when one switches the label from “price fixing” to “cooperative buying,” the emphasis changes. Cooperative buying is price fixing but, as demonstrated here, when the parties do not have market power, it may have beneficial consequences. Yet, members of a buying cooperative face some probability that they will be treated as run-of-themill price fixers. An expanded standing definition increases this probability and may deter desirable efforts at cooperative buying. Second, even when the cooperative buyers do have market power, the overall result of their “price fixing” efforts may be socially beneficial. In fact, our analysis suggests that in some instances, it may make economic sense to recognize expressly the beneficial social welfare effects of a “sanctioned” bilateral monopoly. Here again, the risk to those who would enter into efficiencyproducing agreements on price would increase with increases in the number of potential plaintiffs. Of course, none of this resolves the issue of whether buyers from a collusive monopsony should have standing or whether sellers of inputs to collluding sellers should have standing. Hopefully, further research will provide an answer that is consistent with sound economic and legal reasoning.

11.4  Cooperative Buying and Bilateral Monopoly The relationship between cooperative buying and bilateral monopoly is another area in which further work would be welcome. It is important to note that these two issues are not necessarily intertwined, but they may be. In Chapter 5, we discussed the fact that “cooperative buying” may be just a euphemism for collusive monopsony, but it might not be. The term “cooperative buying” is associated with images of small retailers or wholesalers combining in order to take advantage of economies of scale in warehousing, transportation, or promotion. Collusive monopsony, on the other hand, tends to be viewed as the buyer-side version of traditional, per se unlawful price fixing. In both cases, however, competitors must combine to agree on prices. We have suggested that the key to distinguishing the lawful from the unlawful agreements – however labeled – lies in the response by sellers. Sellers confronted by an efficiency-producing cooperative-buying effort are unlikely to complain. The buyers who have combined to take advantage of real economies expand their demand for the input, and both they and the sellers are better off.

Cooperative Buying and Bilateral Monopoly

229

In contrast, buyers who combine to exert monopsony power will not be simply increasing their demand for the seller’s output. Instead, they will be attempting to set a lower price. A lower price, however, means a lower quantity of the input purchased, a decrease in output, and higher prices for buyers of the final good. Sellers and consumers are worse off, and one would expect the sellers to react. Because it is not always evident what the overall effect of joint purchasing will be, we have suggested that an ancillary restraints approach be used in which buyers must come forth with a rationale linking their behavior to increases in consumer welfare in order to escape per se treatment. To us, at least, this seems to make good economic sense whenever the sellers are in a relatively competitive market. Here is where some futher research would be useful. The problem that is left open here is what sorts of beneficial efforts should be viewed as permitting this movement to rule of reason treatment. One possibility is when competing buyers confront a monopoly seller. This is where bilateral monopoly must be considered. More specifically, should some version of a countervailing market power or bilateral monopoly defense be available to the colluding buyers when confronting a monopsonist? Economic theory indicates that output and consumer welfare are higher with bilateral monopoly than with either of the other market structures. This follows from the fact that the bilateral monopoly’s joint profit-maximizing level of output will exceed the level of output when the monopolist deals with a number of powerless buyers or when a monopsonist deals with a number of powerless sellers. Presumably, subject to a number of transaction costs and the dangers of permitting competing buyers to collude, allowing a number of small, relatively powerless buyers (or sellers) to combine in order to create a simulated bilateral monopoly would also increase output and consumer welfare. The problems, however, are twofold. First, whether by express agreement or by subtle parallel behavior, “permission” to collude as buyers creates a substantial danger that collusion as sellers will also occur. The new result could be that the advantages of the bilateral monopoly structure may be completely offset by anticompetitive effects in the output market. This is an area for future research. The second problem is one of determining how far to extend the bilateral monopoly defense. We have noted that this is complicated and unpredictable even under the best of circumstances – collusion that creates a structure tantamount to a true bilateral monopoly. Obviously, this best case scenario will rarely, if ever, arise. Moreover, it would involve an express recognition

230

Final Comments

of a defense for price fixing based on the reasonableness of the outcome. The complexity, lack of predictability, and implications for similar sellingside defenses cause us to lean away from allowing such a defense. On the other hand, as noted in Chapter 6, courts have taken countervailing power into consideration already,14 and express recognition of such a defense, after more study, may be appropriate. Research efforts should identify the circumstances that justify the defense.

11.5  The Pervasiveness of All-or-None Supply The monopsony model illustrated in Chapter 3 demostrates that the use of monopsony power leads to allocative losses. An exception exists when the monopsonist is able to force sellers onto their all-or-none supply curve. In other words, sellers normally react to lower prices by restricting output. The all-or-none model assumes that the monopsonist has sufficient power to reduce price without reducing quantity, leaving the seller with the choice of producing at the premonopsony level or selling nothing. This model of monopsony raises some interesting questions for further research. This distinction between the conventional model and the all-or-none model is important. If sellers are unable to reduce output in light of lower prices, then it follows that the output of the monopsonist will not decline either. Since output remains the same, there is no negative allocative effect. In addition, consumers can expect to pay the same prices. All that has happened is that the monopsonist has transferred income from sellers to itself. At least as far as modern antitrust policy is concerned, there is little reason to suggest a response to all-or-none market conditions. In fact, sellers in this instance may be viewed as not having suffered an antitrust injury.15 This general conclusion, however, is not without complications. First, even if price setting by a monopsonist were actionable,16 the question arises of whether it is a simple matter to separate all-or-none cases from those that conform to the standard model. The question of whether it is easy to distinguish the two cases deserves further consideration. Second, even though the all-or-none model indicates that there will be no allocative impact, this may exist principally in the short run. Sellers who are consistently offered prices below short-run average total cost will eventually exit the market. Although one would not expect a monopsonist to price in a way that leads See Appalachian Coals v. United States, 288 U.S. 344 (1933); Balmoral Cinema v. Allied Artists Pictures Corp., 885 F.2d 313 (6th Cir. 1989). 15 This would certainly be the case if the monopsony is lawful. 16 This seems possible only with respect to some agricultural markets. See Chapter 8. 14

Single-Firm Behavior

231

to such an outcome, the possibility of error or excessively aggressive pricing exits. This error in judgment, however, is not the sort of thing the antitrust laws were intended to address.

11.6 Single-Firm Behavior A standard element of United States antitrust law is that single firm – even monopoly – pricing is not a practice to which the law responds. There are basically two reasons for this. First, once a court reacts to what it regards as excessive prices, it must answer the second question, which is what price is appropriate. In effect, courts become regulatory bodies, something for which they are ill-suited. Second, the existence of monopoly profit is, in theory, a problem that remedies itself. Supra-competitive profits attract new entrants, supply increases, and prices fall. The same analysis applies to monopsony. A court deciding that a monoponist is paying prices that are too low must decide what the price should be. Again, complications arise that courts are generally unable to address in a rational and fully informed way. So too, a monopsonist paying a price for an input that is well below that input’s marginal revenue product would seem to attract the attention of new buyers. Those buyers would switch to production techniques or to products that make use of the input. Again, at least in the long run, in the absence of barriers (which could raise antitrust concerns), the problem should resolve itself. Nevertheless, especially in agricultural sectors of the economy, there is pressure, in the form of existing and proposed legislation, to respond directly to prices offered by firms with monopsony power. The question for futher reseach is whether efforts to do so can or even should be reconciled with modern antitrust interpretations.

Bibliography

Abere, Andrew E. Trends in Private Antitrust Litigation: 1980–2004, Princeton Economic Group, http://www.econgroup.com/peg_news_view.asp?newid=27&pageno=. Adams, Walter. Competition, Monopoly and Countervailing Power, 67 Q. J. Econ. 469 (November 1953). Antitrust Law Developments, 5th ed. (Chicago: American Bar Association 2002). Areeda, Phillip, & Herbert Hovenkamp. Antitrust Law (Boston: Little, Brown 1986).   Antitrust Law (Boston: Little, Brown 1991 supplement). Areeda, Phillip, Herbert Hovenkamp, Roger D. Blair, & Christine Piette Durrance. Antitrust Law (New York: Aspen Publishing 2007). Areeda, Phillip, & Louis Kaplow. Antitrust Analysis (Boston: Little, Brown 1988). Areeda, Phillip, & Donald Turner. Predatory Pricing and Related Practices Under Section 2 of the Sherman Act, 88 Harv. L. Rev. 687 (1975).   Antitrust Law, vol. 3 (Boston: Little, Brown 1978).   Antitrust Law, vol. 4 (Boston: Little, Brown 1980). Arthur, Thomas C. Farewell to the Sea of Doubt: Jettisoning the Constitutional Sherman Act, 74 Cal. L. Rev. 263 (1986). Axelrod, Robert. The Evolution of Cooperation (New York: Basic Books 1984). Ayres, Ian. Rationalizing Antitrust Cluster Markets, 95 Yale L. J. 109 (1985).   How Cartels Punish: A Structural Theory of Self-Enforcing Collusion, 87 Colum. L. Rev. 295 (1987). Bacon, Nathaniel T. Researches into the Mathematical Principles of the Theory of Wealth (New York: Augustus M. Kelley 1927). Baker, Tyler A. The Supreme Court and the Per Se Tying Rule: Cutting the Gordian Knot, 66 Va. L. Rev. 1235 (1980). Baldwin, Lee, R. Marshall, & J. F. Richards. Collusion at Forest Service Timber Sales, 105 J. Pol. Econ. 657 (1997). Baler, Jonathan B., Joseph Farrell, & Carl Shapiro. Merger to Monopoly to Serve a Single Buyer: Comment, 75 Antitrust L. J. 641 (2008). Barro, Robert J. Let’s Play Monopoly, Wall Street J., Aug. 27, 1991. Bass, Christopher M. More Than a Mirror:  The Packers and Stockyards Act, Antitrust Law, and the Injury to Competition Requirement, 12 Drake J. Agric. L. 423 (2007). Baumol. William, & Janusz A. Ordover. Use of Antitrust to Subvert Competition, 28 J. L. & Econ. 247 (1985). 233

234

Bibliography

Becker, Gary. Crime and Punishment:  An Economic Approach 86 J. Pol. Econ. 169 (1968).   The NCAA: A Cartel in Sheepskin Clothing, Business Week, Sept. 14, 1987. Berg, Sanford, & J. Tschirhart. Natural Monopoly Regulation (New York:  Cambridge University Press 1988). Bernheim, B. Douglas, & Michael D. Winston, Exclusive Dealing, 106 J. Pol. Econ. 64 (1998). Blair, Roger D. The Sherman Act and the Incentive to Collude, 17 Antitrust Bull. 433 (1972).   A Suggestion for Improved Antitrust Enforcement, 30 Antitrust Bull. 433 (1985).   Sports Economics (New York: Cambridge University Press 2008). Blair, Roger D., & Kristine L. Coffin. Physician Collective Bargaining, State Legislation, and the State Action Doctrine, 26 Cardozo L. Rev. 1731 (2005). Blair, Roger D., & Christine Piette Durrance. The Economics of Monopsony, in Issues in Competition Law and Policy, W. Dale Collins, ed. (Chicago:  American Bar Association 2008). Blair, Roger D., & James M. Fesmire. Maximum Price Fixing and the Goals of Antitrust, 37 Syracuse L. Rev. 43 (1986). Blair, Roger D., & Jeffrey Finci. The Individual Coercion Doctrine and Tying Arrangements: An Economic Analysis, 10 Fla. St. U. L. Rev. 531 (1983). Blair, Roger D., & Jeffrey L. Harrison. Rethinking Antitrust Injury, 42 Vand. L. Rev. 1539 (1989).   Antitrust Policy and Monopsony, 76 Cornell L. Rev. 297 (1991).   Cooperative Buying, Monopsony Power and Antitrust Policy, 86 Nw. U. L. Rev. 331 (1992).   Reexamining the Role of Illinois Brick in Modern Antitrust Standing Analysis, 68 Geo. Wash. L. Rev. 1 (1999). Blair, Roger D., & Jill Boylston Herndon. A Note on Hanover Shoe, 43 Antitrust Bull. 365 (1997).   Physician Cooperative Bargaining Ventures: An Economic Analysis, 71 Antitrust L. J. 989 (2004). Blair, Roger D., & David L. Kaserman. The Albrecht Rule and Consumer Welfare:  An Economic Analysis, 33 U. Fla. L. Rev. 461 (1981).   Law and Economics of Vertical Integration and Control (New York: Academic Press 1983).   Antitrust Economics (Homewood, Ill.: Richard D. Irwin 1985).   A Note on Bilateral Monopoly and Formula Price Contracts, 77 Am. Econ. Rev. 460 (1987).   Antitrust Economics, 2d ed. (New York: Oxford University Press 2009). Blair, Roger D., David L. Kaserman, & Richard E. Romano. A Pedagogical Treatment of Bilateral Monopoly, 55 S. Econ. J. 831 (1989). Blair, Roger D., & Larry Kenny. Microeconomics with Business Applications (New York: John Wiley 1987). Blair, Roger D., & Gordon L. Lang. Albrecht after Arco: Maximum Resale Price Fixing Moves Toward the Rule of Reason, 44 Vand. L. Rev. 1007 (1991). Blair, Roger D., & John E. Lopatka. The Albrecht Rule After Kahn: Death Becomes Her, 74 Notre Dame Lawyer 123 (1998).

Bibliography

235

  Predatory Buying and the Antitrust Law, Utah L. Rev. 415 (2008). Blair, Roger D., & William H. Page. ‘Speculative’ Antitrust Damages, 70 Wash. L. Rev. 423 (1995). Blair, Roger D., & Richard E. Romano. Distinguishing Participants from Nonparticipants in a Price-Fixing Conspiracy: Liability and Damages, 28 Am. Bus. L. J. 33 (1990).   Collusive Monopsony in Theory and Practice:  The NCAA. 42 Antitrust Bull. 681 (1997). Blair, Roger D., & Carolyn D. Schafer. Antitrust Law and Evolutionary Models of Legal Change, 40 U. Fla L. Rev. 379 (1988). Bork, Robert. Vertical Restraints: Schwinn Overruled, 171 Sup. Ct. Rev. 1977 (1978).   The Antitrust Paradox (New York: Basic Books 1978). Bowley, A. L. Bilateral Monopoly, 25 Econ. J. 651 (1928). Bowman, Ward. Tying Arrangements and the Leverage Problem, 67 Yale L. J. 19 (1957). Brady, Amber. Post-Smithfield and Hazeltine: An Evaluation of the Capper-Volstead Act as an Alternative Means of Marketing Power for Producers, 10 Drake J. Agric. L. 331 (2005). Brodley, Joseph, & George Hay. Predatory Pricing: Competing Theories and the Evolution of Legal Standards, 66 Cornell L. Rev. 738 (1981). Burnstein, M. L. A Theory of Full-Line Forcing, 55 Northeastern U. L. Rev. 62 (1960). Campbell, Tom. Bilateral Monopoly in Mergers, 74 Antitrust L. J. 521 (2007). Campus Conspirators?, New York Times, Aug. 18, 1989, at A30. Carlton, Dennis W. A General Analysis of Exclusionary Conduct and Refusals to Deal – Why Aspen and Kodak Are Misguided, 68 Antitrust L. J. 659 (2001). Carstensen, Peter C. Concentration and the Destruction of Competition in Agricultural Markets: The Case for a Change in Public Policy, 2000 Wis. L. Rev. 531 (2000). Carter, John R. From Peckham to White: Economic Welfare and the Rule of Reason, 25 Antitrust Bull. 95 (1980). Cassady, Ralph Jr. Auctions and Auctioneering (Berkeley, CA: Univ. of Cal. Press 1967). Coase, Ronald H. The Nature of the Firm, 4 Economica 386 (1937). Coates, Malcolm B. Efficiencies in Merger Analysis: An Institutional View, 13 Supreme Ct. Econ. Rev. 189 (2005). Cooper, Edward. Attempts and Monopolization: A Mildly Expansionary Answer to the Prophylactic Riddle of Section Two, 72 Mich. L. Rev. 357 (1974). Cournot, Augustin. Recherches sur les Principes Mathematiques de la Theirie des Richesses (1838). Cox, Meg. At Many Auctions, Illegal Bidding Thrives as a Longtime Practice Among Dealers, Wall Street J., Feb. 19, 1988. Cruz, Ted R. Federal Trade Commission. The Threat of Consumer Harm Resulting from Physicians Collective Bargaining Under Alaska Senate Bill 37 (March 22, 2002), http://www.ftc.gov/be/hilites/cruz020322.shtm. Dean, James B., & Thomas Earl Geu. The Uniform Limited Cooperative Association Act, 13 Drake J. Agric. L. 63 (2008). Dewey, Donald. The Common Law Background of Antitrust Policy, 41 Va. L. Rev. 759 (1955). Division Seeks Documents From Colleges in Probe of Financial Aid and Tuition, 57 Antitrust Trade Reg. Rep. 278 (1989).

236

Bibliography

Easterbrook, Frank. Predatory Strategies and Counterstrategies, 48 U. Chi. L. Rev. 263 (1981).   The Limits of Antitrust, 63 Tex. L. Rev. 1 (1984). Edelman, Mark. Has Collusion Returned to Baseball? Analyzing Whether a Concerted Increase in Free Agent Player Supply Would Violate Baseball’s Collusion Clause, 24 Loy. Ent. L. Rev. 159 (2004). Elhauge, Einer. Defining Better Monopolization Standards, 56 Stan. L. Rev. 253 (2003). Elzinga, Kenneth G. Predatory Pricing: The Case of the Gunpowder Trust, 13 J. L. & Econ. 223 (1970).   New Developments on the Cartel Front, 29 Antitrust Bull. 3 (1984).   Collusive Predation:  Matsuchita v. Zenith, in J. Kwoka & L. White, The Antitrust Revolution (Glenview, Ill.: Scott, Foresman 1989). Elzinga, Kenneth G., & William Breit. The Antitrust Penalties (New Haven, Conn.: Yale University Press 1976). Elzinga, Kenneth G., & Thomas Hogarty. The Problem of Geographic Market Delineation in Antimerger Suits, 18 Antitrust Bull. 45 (1973). Elzinga, Kenneth G., & David Mills. Testing for Predation: Is Recoupment Feasible? 34 Antitrust Bull. 869 (1989). Farrey, Tom. NCAA Might Face Damages in Hundreds of Millions (Feb. 20, 2006), http:// sports.espn.go.com/ncaa/news/story?id=2337810. Fellner, William. Competition Among the Few (New York: Alfred A. Knopf 1949). Ferguson, C. E. The Neoclassical Theory of Production and Distribution (New York: Cambridge University Press 1971). Finkelstein, Michael O., & Hans Levenback. Regression Estimates of Damages in PriceFixing Cases, 45 L. Contemp. Probs. 145 (1983). Fisher, Alan A., & Robert Lande. Efficiency Considerations in Merger Enforcement, 71 Cal. L. Rev. 1580 (1983). Fisher, Alan, & Richard Sciacca. An Economic Analysis of Vertical Merger Enforcement Policy, 6 Res. L. & Econ. 1 (1984). Fisher, Franklin. Multiple Regression in Legal Proceedings, 80 Colum. L. Rev. 702 (1980). Fleisher, Arthur A., Brian L. Goff, & Robert D. Tollison. The National Collegiate Athletic Association:  A Study in Cartel Behavior (Chicago:  University of Chicago Press 1992). Floyd, Susan S., ed. Florida Statistical Abstract (Gainesville: University of Florida 2003). Forchheimer, Karl. Theoretisches zum unvollstandigen Monopole, Schmollers Jahrbuch 1 (1908). Freudenheim, Milt. Pressure Grows for Curbs on Prices of Prescriptions, New York Times, May 11, 1991. Friedman, Milton. Price Theory (Chicago: Aldine Publishing 1976).   How to Sell Government Securities, Wall Street J., Aug. 28, 1991. Friedman, Richard D. Antitrust Analysis and Bilateral Monopoly,1986 Wis. L. Rev. 873 (1986). Galbraith, John K. American Capitalism:  The Concept of Countervailing Power (Boston: Houghton Mifflin 1952). Gould, John, & Edward Lazear. Microeconomic Theory, 6th ed. (Homewood, Ill.: Richard D. Irwin 1989). Graham, Daniel A., & Robert C. Marshall. Collusive Bidder Behavior at Single Object Second-Price and English Auctions, 95 J. Pol. Econ. 1217 (1987).

Bibliography

237

Green, Edward, & Robert Porter. Noncooperative Collusion Under Imperfect Price Information, 52 Econometrica 87 (1984). Grosberg, Lawrence M. Illusion and Reality in Regulating Lawyer Performance: Rethinking Rule 11, 32 Vill. L. Rev. 575 (1987). Hall, Christopher D. A Dutch Auction Information Exchange, 32 J. L. & Econ. 195 (1989). Handler, Milton, Harlan Blake, Robert Pitovsky, & Harvey Goldschmid. Trade Regulation, 3d ed. (Westbury, N.Y.: Foundation Press 1990). Hansen, Hugh. Robinson-Patman Law: A Review and Analysis, 5 Fordham L. Rev. 1113 (1983). Harris, Robert, & Lawrence Sullivan. Passing on the Monopoly Overcharge:  A Comprehensive Policy Analysis, 128 U. Pa. L. Rev. 269 (1979). Harrison, Jeffrey L. The Lost Profits Measure of Damages in Price Enhancement Cases, 64 Minn. L. Rev. 751 (1980).   Price Fixing, the Professions and Ancillary Restraints: Coping with Maricopa County, U. Ill. L. Rev. 925 (1982).   The Brand Name Prescription Drug Litigation: Comments on Scherer, 4 Int’l J. Econ. Bus. 265 (1997).   Brown v. Pro Football, Inc.: The Labor Exemption, Antitrust Standing and Distributive Outcomes, Antitrust Bull. 565 (1997).   The Messenger Model: Don’t Ask, Don’t Tell? 71 Antitrust L. J. 1017 (2004).   An Instrumental Approach to Market Power and Antitrust Policy, 59 SMU L. Rev. 1673 (2006). Hay, George. Oligopoly, Shared Monopoly, and Antitrust Law, 67 Cornell L. Rev. 439 (1982). Hay, George, & Daniel Kelley. An Empirical Survey of Price Fixing Conspiracies, 17 J. L. & Econ. 13 (April 1974). Hendricks, Ken, Robert Porter, & Gunfu Tan. Bidding Rings and the Winner’s Curse, 39 RAND J. Econ. 1018 (2008). Herman, Tom. A Primer on the Treasury Market:  How the Government Sells Its Debt, Wall Street J., Aug. 19, 1991. Herndon, Jill B. Health Insurer Monopsony Power: All-Or-None Model, 21 J. Health Econ. 197 (2002). Hicks, Sir John. A Revision of Demand Theory (Oxford: Oxford University Press 1956). Hirschfeld, Edward B. Physicians, Unions, and Antitrust, 32 J. Health L. 43 (1999). Horn, Henrick, & Asher Wolinsky. Bilateral Monopoly and Incentives for Merger, 19 RAND J. Econ. 408 (1988). Hovenkamp, Herbert. Distributive Justice and the Antitrust Laws, 51 Geo. Wash. L. Rev. 1 (1982).   Merger Actions for Damages, 35 Hastings L. J. 937 (1984).   Economics and Federal Antitrust Law (St. Paul, Minn.: West Publishing 1985).   The Indirect Purchaser and Cost-Plus Sales, 103 Harv. L. Rev. 1717 (1990).   Market Power in Aftermarkets: Antitrust Policy and the Kodak Case, 40 UCLA L. Rev. 1447 (1993).   Federal Antitrust Policy:  The Law of Competition and Its Practice, 2d ed. (Minneapolis: West Publishing Co. 1999). Hoyt, Richard C., Dale C. Dahl, & Stuart D. Gibson. Comprehensive Models for Assessing Lost Profits to Antitrust Plaintiffs, 60 Minn. L. Rev. 1233 (1976).

238

Bibliography

Humpreys Brad R., & Jane E. Ruseski. Monitoring Cartel Behavior and Stability: Evidence from NCAA Football, 75 S. Econ. J. 720 (2009). Hylton, Keith N. The Law and Economics of Monopolization Standards, in Antitrust Law & Economics, Keith N. Hylton ed. (Boston: Edward Elgar Publishing, 2010).   Antitrust Law:  Economic Theory and Common Law Evolution (Boston:  Cambridge University Press 2003). Jacobson, Jonathan. Exclusive Dealing, Foreclosure, and Consumer Harm, 70 Antitrust L. J. 311 (2002). Jacobson, Jonathan, & Gary Dorman. Joint Purchasing, Monopsony, and Antitrust, 36 Antitrust Bull. 1 (1991). Jacquemin, Alexis, & Margaret E. Slade. Cartels, Collusion, and Horizontal Merger, in Richard Schmalensee & Robert Willig, Handbook of Industrial Organization (Chicago: Elsevier Science Pub. Co. 1989). Jegadeesh, Narasimhan. Treasury Auction Bids and the Salomon Squeeze, XLVIII J. Fin. 1403 (1993). Johnston, David. Price-Fixing Inquiry at 20 Elite Colleges, New York Times, 10 Aug. 1989. Joskow, Paul, & Alan Klevorick. A Framework for Analyzing Predatory Pricing Policy, 89 Yale L. J. 213 (1979). Kahn, Lawrence M. Markets:  Cartel Behavior and Amateurism in College Sports, 21 J. Econ. Persp. 209 (2007). Kaplow, Louis. Extension of Monopoly Power Through Leverage, 85 Colum. L. Rev. 515 (1985). Katz, Michael L., & Harvey S. Rosen. Microeconomics (Homewood, Ill.: Richard D. Irwin 1991). Kirkwood, John B. Controlling Above-Cost Predation:  An Alternative to Weyerhaeuser and Brooke Group, 53 Antitrust Bull. 369 (2008). Kinnucan, Henry, & Gregory Sullivan. Monopsonistic Food Processing and Farm Prices:  The Case of the West Alabama Catfish Industry, S. J. Agric. Econ. 15 (1986). Klein, Benjamin, Robert G. Crawford, & Armen A. Alchian. Vertical Integration, Appropriable Rents, and the Competitive Contracting Process, 21 J. L. & Econ. 297 (1978). Kmenta, Jan. Elements of Econometrics (New York: Macmillan 1971). Koller, Roland. The Limitations of Local Price-Cutting as a Barrier to Entry, 64 J. Pol. Econ. 329 (1956).   The Myth of Predatory Pricing:  An Empirical Study, 5 Antitrust L. & Econ. Preview 105 (1971). Kramer, Victor. The Supreme Court and Tying Arrangements:  Antitrust as History, 69 Minn. L. Rev. 1013 (1985). Krattenmaker, Thomas G., Robert H. Lande, & Steven C. Salop. Monopoly Power and Market Power in Antitrust Law, 76 Geo. L. J. 241 (1987). Lancaster, Hal. Baseball Owners Conspire to Shoot Down Market for Free Agents, Arbitrator Rules, Wall Street J., Sept. 22, 1987. Lande, Robert. Wealth Transfers as the Original and Primary Concern of Antitrust: The Efficiency Interpretation Challenged, 34 Hastings L. J. 65 (1982). Landes, William, & Richard Posner. Should Indirect Purchasers Have Standing to Sue Under the Antitrust Laws? 46 U. Chi. L. Rev. 602 (1979).

Bibliography

239

  Market Power in Antitrust Cases, 94 Harv. L. Rev. 937 (1981). Lanzillotti, Robert F. The Superior Market Power of Food Processing and Agricultural Supply Firms  – Its Relation to the Farm Problem, 42 J. Farm Econ. 1228 (Dec. 1960). Layard, Richard G., & A. A. Walters. Microeconomic Theory (New York: McGraw-Hill 1978). Leeman, Wayne. The Limitations of Local Price-Cutting as a Barrier to Entry, 65 J. Pol. Econ. 329 (1956). Lehn, Kenneth. Property Rights Risk Sharing and Player Disability in Major League Baseball, 25 J. L. & Econ. 343 (1982). Lerner, Abba. The Concept of Monopoly and the Measurement of Monopoly Power, 1 Rev. Econ. Stud. 157 (1934). Leslie, Connie, & Sue Hutchinson. An Ivy League Cartel, Newsweek, Aug. 21, 1989. Lewin, Tamar. A Legal Curb Raises Hackles, New York Times, Oct. 2, 1986, at D1. Liebeler, Wesley. Whither Predatory Pricing? From Areeda and Turner to Matsushita, 61 Notre Dame L. Rev. 1052 (1986). Liman, Arthur. The Paper Label Sentences: Critique, 86 Yale L. J. 630 (1977). Lockhart, William B., & Howard R. Sacks. The Relevance of Economic Factors in Determining Whether Exclusive Arrangements Violate Section 3 of the Clayton Act, 65 Harv. L. Rev. 913 (1952). Lopatka, John E. Antitrust Injury and Causation, in Issues in Competition Law and Policy, W. D. Collins, ed. (Chicago: American Bar Association 2008). Louis, Martin. Restraints Ancillary to Joint Ventures and Licensing Agreement: Do Sealy and Topco Logically Survive Sylvania and Broadcast Music? 66 Va. L. Rev. 879 (1980).   Intercepting and Discouraging Doubtful Litigation:  A Golden Anniversary View of Pleading, Summary Judgment, and Rule 11 Sanctions Under the Federal Rules of Civil Procedures, 67 N.C. L. Rev. 1023 (1989). Luce, R. Duncan, & Howard Raiffa. Games and Decisions (New York:  John Wiley 1957). Machlup, Fritz, & Martha Taber. Bilateral Monopoly, Successive Monopoly, and Vertical Integration, 27 Economica 101 (1960). Marvel, Howard P. Exclusive Dealing, 25 J. L. & Econ. 1 (1982). McCall, Charles. Predatory Pricing: An Economic and Legal Analysis, 1 Antitrust Bull. 32 (1987). McCoy, Charles. The Paramount Cases:  Golden Anniversary in a Rapidly Changing Marketplace,2 Antitrust 32 (1988). McEowen, Roger A., Peter C. Carstensen, & Neil E. Harl. The 2002 Senate Farm Bill: The Ban on Packer Ownership of Livestock, 7 Drake J. Agric. L. 268 (2002). McGee, John S. Predatory Pricing:  The Standard Oil (N.J.) Case, 1 J. L. & Econ. 137 (1958).   Predatory Pricing Revisited, 23 J. L. & Econ. 289 (1980). McGee, John, & Lowell Bassett. Vertical Integration Revisited, 19 J. L. & Econ. 17 (1976). Menger, Carl. Grundsätze der Volkswirtschaftslehre (1871). Milgrom, Paul, Auctions and Bidding: A Primer, 3 J. Econ. Persp. 3 (1989). MLB, Players Union Settle Potential Collusion Claims, http://sports.espn.go.com /mlb/ news/story?id=2652091 (last visited Sept. 8, 2008).

240

Bibliography

NCAA Settles Antitrust Lawsuit; Will Create $218 Million Fund for Athletes, University Business, Jan. 31, 2008, http://www.universitybusiness.com/newssummary.a spx?news=yes&postid=15221 (last visited Oct. 14, 2008). Noll, Roger G. ‘Buyer Power’ and Economic Policy, 72 Antitrust L. J. 589 (2005). Note. Bidding for the Best and the Poorest, New York Times, Feb. 19, 1990. Note. Challenging Concentration of Control in the American Meat Industry, 117 Harv. L. Rev. 2643 (2004). Note. Division Seeks Documents From Colleges in Probe of Financial Aid and Tuition, 57 Antitrust & Trade Reg. Rep. 278 (1989). Note. The Dynamics of Rule 11: Preventing Frivolous Litigation by Demanding Professional Responsibility, 61 N.Y.U. L. Rev. 300 (1986). Note. Insuring Rule 11 Sanctions, 88 Mich. L. Rev. 334 (1989). Note. Monopsonistic Price Fixing and Umbrella Pricing as a Theory of Antitrust Standing: A New View of Illinois Brick, 50 Cin. L. Rev. 52 (1981). Note. More Trouble with Treble: The Effects of McCready and Associate General Contractors on the Antitrust Standing Opinions of the Federal Courts of Appeals, 10 J. Corp. L. 463 (1985). Note. Plausible Pleadings: Developing Standards for Rule 11 Sanctions, 100 Harv. L. Rev. 630 (1987). Note. A Reevaluation of Reciprocal Dealing Under the Federal Antitrust Laws: Spartan Grain v. Ayers, 11 Loy. U. Chi. L. J. 577 (1980). Orr, Daniel, & Paul MacAvoy. Price Strategies to Promote Cartel Stability, 32 Economica 186 (1965). Osborne, D. K. Cartel Problems, 66 Am. Econ. Rev. 835 (1976). Pae, Peter. Boycott Threat Spurs American Express to Rethink Fees, Wall Street J., Mar. 28, 1991.   Today’s Special: Cut in American Express Fees, Wall Street J., Apr. 22, 1991. Page, William H. Antitrust Damages and Economic Efficiency: An Approach to Antitrust Injury, 47 U. Chi. L. Rev. 467 (1980).   The Scope of Liability for Antitrust Violations, 37 Stan. L. Rev. 1445 (1985).   The Chicago School and the Evolution of Antitrust: Characterization, Antitrust Injury, and Evidentiary Sufficiency, 75 Va. L. Rev. 1221 (1989). Parker, Russell. Treble Damage Actions – A Financial Deterrent to Antitrust Violations?, 16 Antitrust Bull. 483 (1971). Paul, Ron. Introducing the Quality Health Care Coalition Act (Aug. 8, 2007) http://www. reasontofreedom.com/introducing_the_quality_health_care_coalition_act_by_ us_rep_ron_paul.html (last visited Oct. 21, 2008). Parkin, Michael. Economics, 5th ed. (London: Addison Wesley–Pearson 2000). Perry, Martin. Vertical Integration: The Monopsony Case, 68 Am. Econ. Rev. 561 (1978). Philips, Louis. The Economics of Imperfect Information (New York: Cambridge University Press 1988). Pitofsky, Robert. The Sylvania Case: Antitrust Analysis of Non-Price Vertical Restrictions, 78 Colum. L. Rev. 1 (1978).   Efficiencies in Defense of Mergers: Two Years Out, 7 Geo. Mason L. Rev. 485 (1999). Polinsky, Mitchell. An Introduction to Law and Economics, 2d ed. (Boston: Little, Brown 1991). Posner, Richard A. A Statistical Study of Antitrust Enforcement, 13 J. L. & Econ. 365 (1970).

Bibliography

241

  The Social Cost of Monopoly and Regulation, 83 J. Pol. Econ. 807 (1975).   The Economic Theory of Criminal Law, 85 Colum. L. Rev. 1193 (1985).   Antitrust Law:  An Economic Perspective (Chicago:  University of Chicago Press 1976).   The Robinson-Patman Act:  Federal Regulation of Price Differences (Washington, D.C.: American Enterprise Institute 1976). Posner, Richard, & Frank Easterbrook. Antitrust:  Cases, Economic Notes, and Other Materials (St. Paul, Minn.: West Publishing 1981). Preston, R & John McMillan. Bidding Rings, 82 Am. Econ. Rev. 578 (1992). Rapoport, Anatol, & Albert Chammah. Prisoner’s Dilemma:  A Study in Conflict and Cooperation (Ann Arbor: University of Michigan Press 1965). Rasmusen, Eric. Games and Information (New York: Basil Blackwell 1989). Rasmusen, Eric, J. Mark Ramseyer, & John S. Wiley, Jr. Naked Exclusion, 8 Am. Econ. Rev. 1137 (1991). Renfrew, Charles. The Paper Label Sentences: An Evaluation, 86 Yale L. J. 590 (1977). Roberts, Johnnie L. FTC Probes American Express Restaurant Fee Revolt, Wall Street J., Apr. 26, 1991. Robinson, Joan. The Economics of Imperfect Competition 293 (London:  Macmillan & Co. 1933). Rosen, Sherwin. Comment, in Empirical Studies in Health Economics, Herbert Klarman, ed. (Baltimore: Johns Hopkins University Press 1970). Rovner, Jack. Monopsony Power in Health Care Markets: Must the Big Buyer Beware Hard Bargaining? 18 Loy. U. Chi. L. J. 857 (1987). Rowe, Frederick. Price Discrimination Under the Robinson-Patman Act (Boston: Little Brown 1962). Rule, Charles. Antique Dealers Face Price Fixing Charges, 53 Antitrust Trade Reg. Rep. 117 (1987). Salop, Steven. Measuring the Ease of Entry, 31 Antitrust Bull. 551 (1986). Samuelson, Paul. Foundations of Economic Analysis (Cambridge, Mass.:  Harvard University Press 1947).   The Monopolistic Competition Revolution, in III The Collected Scientific Papers of Paul Samuelson, R. Merton, ed. (Cambridge, Mass.: MIT Press 1972). Saving, Thomas R. Concentration Rations and the Degree of Monopoly, 11 Int’l Econ. Rev. 139 (1970). Scherer, Frederick M. Industrial Market Structure and Economic Performance (Chicago: Rand McNally 1980).   How US Antitrust Can Go Astray: The Brand Name Prescription Drug Litigation, 4 Int’l J. Econ. Bus. 239 (1997). Scherer, Frederick M., & David Ross. Industrial Market Structure and Economic Performance, 3d ed. (Boston: Houghton Mifflin 1990). Schlosser, Eric. Fast Food Nation:  The Dark Side of the All-American Meal (Evansville: Perennial 2002). Schwarzer, William W. Rule 11 Revisited, 101 Harv. L. Rev. 1013 (1988). Sexton, Richard J. Industrialization and Consolidation in the U.S. Food Sector: Implications for Competition and Welfare, 82 Am. J. Agric. Econ. 1087 (2000). Shapiro, Carl. Theories of Oligopoly Behavior, in Handbook of Industrial Organization, Richard Schmalensee & Robert Willig, eds. (Chicago:  Elsevier Science Pub. Co. 1989).

242

Bibliography

Siconolfi, Michael, Michael R. Sesit, & Constance Mitchell. Hidden Bonds, Wall Street J., Aug. 19, 1991. Slawson, W David. A New Concept of Competition:  Reanalyzing Tie-In Doctrine After Hyde, 30 Antitrust Bull. 257 (1985) Smith, Claire. Arbitrator Finds 3d Case of Baseball Collusion, New York Times, Jul. 19, 1990. Spulber, D. Regulation and Markets (Cambridge, Mass.: MIT Press 1989). Stern, L., & A. El-Ansary. Marketing Channels (Englewood Cliffs, N.J.:  Prentice Hall 1984). Steuer, Richard M. Exclusive Dealing After Jefferson Parish, 54 Antitrust L. J. 1229 (1986). Stigler, George J. Notes on the Theory of Duopoly, 48 J. Pol. Econ. 521 (1940).   A Theory of Oligopoly, 72 J. Pol. Econ. 44 (February 1964).   The Economic Effects of the Antitrust Laws, J. L. & Econ. 9 (1966).   The Theory of Price, 4th ed. (New York: Macmillan 1987). Stigler, George J., & Robert A. Sherwin. The Extent of the Market, 28 J. L. & Econ. 555 (1985). Sullivan, Daniel. Monopsony Power in the Market for Nurses, 32 J. L. & Econ. 135 (1989). Sullivan, E. Thomas, & Jeffrey L. Harrison. Understanding Antitrust and Its Economic Implications, 5th ed. (New York: LexisNexis 2009). Sullivan E. Thomas, & Herbert Hovenkamp. Antitrust Law, Policy and Procedure, 5th ed. (Charlottesville, Va.: LexisNexis 2007). Sullivan, Lawrence. Antitrust (St. Paul, Minn.: West Publishing 1976). Sullivan, Lawrence, & John Wiley, Recent Antitrust Developments: Defining the Scope of Exemptions, Expanding Coverage and Refining the Rule of Reason, 27 UCLA L. Rev. 265 (1979). Taylor, C. Robert. Buyer Power Litigation in Agriculture: Pickett v. Tyson Fresh Meats, Inc., 53 Antitrust Bull. 455 (2008).   The Many Faces of Power in the Food System, http://www.ftc.gov/bc/mergerenforce/ presentations/040217taylor.pdf (last visited April 27, 2009). Telser, Lester. A Theory of Self-Enforcing Agreements, 53 J. Bus. 27 (1980). Thorelli, Hans. The Federal Antitrust Policy (Baltimore: Johns Hopkins University Press 1954). Tullock, Gordon. The Welfare Costs of Tariffs, Monopolies, and Theft, 5 W. Econ. J. 224 (1967). Varian, Hal. Intermediate Microeconomics (New York: W. W. Norton 1987). Vecsey, George. Bring Back Collusion in Baseball, New York Times, Dec. 13, 1989. Vernon, John, & Daniel Graham. Profitability of Monopolization by Vertical Integration, 79 J. Pol. Econ. 924 (1971). Walters, K., & J. Stephen. Reciprocity Reexamined: The Consolidated Foods Case, 29 J. L. & Econ. 423 (1986). Webster’s New Ninth Collegiate Dictionary (Springfield, Mass.:  Merriam-Webster 1988). Wheeler, Malcolm. Ántitrust Treble-Damage Actions:  Do They Work?, 61 Cal. L. Rev. 1319 (1973).

Bibliography

243

White, Barbara. Different Rules for Different Markets: Conduct and Context in Antitrust Law and Economics, 41 Duke L. J. 1045 (1992). White, Lawrence J. Private Antitrust Litigation: New Evidence, New Learning (Cambridge, Mass.: MIT Press 1988). Williamson, Oliver. The Economics of Antitrust: Transaction Cost Considerations, 122 U. Pa. L. Rev. 1439 (1974).   Economics as an Antitrust Defense:  The Welfare Tradeoffs, 58 Am. Econ. Rev. 18 (1968).   Markets and Hierarchies:  Some Elementary Considerations, 63 Am. Econ. Rev. 316 (1973).   Predatory Pricing: A Strategic and Welfare Analysis, 87 Yale L. J. 204 (1977).   Transaction Cost Economics in Handbook of Industrial Organization, Richard Schmalensee & Robert Willig, eds. (Chicago: Elsevier Science Pub. Co. 1989). Wu, S. Y. The Effects of Vertical Integration on Price and Output, 2 W. Econ. J. 117 (1964).

Index

Alcoa, 20 All-or-none Supply Cases, viii, 83 all-or-none supply curve, 83, 118, 175, 177, 230 ancillary restraints doctrine, 17, 113, 116, 118 antitrust injury, 28, 79, 148, 150, 151, 153, 154, 155, 156, 157, 158, 159, 161, 162, 165, 166, 168, 225, 226, 230 Antitrust Responses to Cooperative Buying Efforts, ix, 113 antitrust standing, xiii, 81, 148, 150, 152, 154, 155, 156, 157, 158, 159, 160, 161, 163, 165, 166, 167, 168, 225, 226, 227 Aspen Skiing Co. v. Aspen Highlands Skiinq Corp, 21 auctions, 2, 3, 5, 33, 50, 87, 184 bid rigging, viii, 12, 33, 34, 50, 81, 86, 87, 88 bilateral monopoly, xiii, 15, 123, 126, 127, 128, 130, 131, 132, 133, 134, 136, 137, 138, 139, 140, 141, 143, 183, 208, 222, 228, 229 Blue Shield of Virginia v. McCready, 154 BMI, 114, 115 BPI, viii, xi, 54, 58, 59, 60, 61, 63, 64, 66, 67, 218, 219, 220, 221 Broadcast Music Inc., v. Columbia Broadcasting System (BMI), 114 Brook Group Ltd. v. Brown & Williamson Tobacco Co, 73

Brown Shoe Co. v. United States, 28 Brunswick Corp. v. Pueblo Bowl-O-Matic, Inc, 150 buying cooperative, xiii, 106, 109, 110, 111, 112, 116, 117, 119, 126, 138, 228 buying power, 7, 11, 12, 24, 29, 36, 37, 38, 41, 42, 53, 54, 55, 57, 58, 59, 60, 61, 63, 64, 65, 68, 71, 78, 83, 89, 101, 104, 106, 107, 109, 112, 116, 117, 124, 125, 138, 157, 174, 182, 187, 190, 205, 212, 213, 218, 219 buying power index, 53, 54 Capper-Volstead Act, 183, 235 Clayton Act, vii, 16, 22, 23, 24, 26, 27, 29, 37, 65, 90, 91, 95, 99, 124, 140, 146, 147, 150, 151, 205, 239 collusive monopsony, xiii, 14, 35, 41, 68, 69, 78, 79, 85, 124, 139, 150, 157, 158, 160, 167, 189, 226 Collusive Monopsony, vii, viii, x, 11, 14, 30, 48, 49, 157, 189, 190, 208, 235 Consumer surplus, 43 Continental T.V. v. G.T.E. Sylvania, 19 cooperative buying, 14, 60, 106, 107, 108, 109, 110, 112, 115, 116, 119, 120, 122, 159, 228. See buying cooperatives countervailing market power defense, 123, 138

245

246

Index

damages, 9, 147, 150, 152, 153, 155, 156, 158, 160, 165, 168, 169, 170, 171, 185, 191, 226, 227, 236 direct purchasers, 152, 160, 226 Eastman Kodak Co. v. Image Technical Services, 175 entry barriers, 28, 220 exclusive dealing, 22, 24, 26, 27, 37, 91, 167 financial aid, 1, 6, 7, 191 Friedman, Milton, 5, 44, 83, 128 group boycott, 18, 45, 163 health care industry, 12 horizontal market division, 85, 86, 107, 113, 163, 169, 224 horizontal price fixing, 29, 30, 39, 107, 109, 162 Inelastic Supply and Perishable Commodities, 79 Jefferson Parish Hospital v. Hyde, 25 Kartell v. Blue Shield of Massachusetts, 13, 71, 85 Klor’s v. Broadway-Hale Stores, Inc, 36, 74 Managed care, 205 Mandeville Island Farms v. American Crystal Sugar Co, 31 marginal factor cost, 43, 44, 46, 47, 54, 55, 56, 96, 110, 126, 127, 130, 207 market share, 21, 22, 28, 53, 57, 58, 60, 61, 62, 63, 64, 65, 66, 67, 94, 141, 183, 214, 215, 220, 221 Measuring Buying Power, viii, 53 Merger Guidelines, 29, 53, 66, 119 mergers, 12, 14, 22, 27, 28, 29, 30, 40, 53, 55, 61, 93, 98, 109, 119, 150, 174, 184 monopolization, 20, 21, 29, 61, 68, 88, 152, 163, 183

monopsony, i, xiii, xiv, 1, 3, 10, 11, 12, 13, 14, 15, 16, 23, 24, 27, 29, 30, 31, 35, 36, 37, 38, 39, 40, 41, 42, 44, 45, 46, 47, 48, 50, 51, 52, 53, 54, 55, 59, 62, 63, 64, 66, 67, 68, 69, 70, 71, 72, 74, 75, 76, 77, 78, 79, 82, 83, 85, 86, 88, 89, 90, 91, 92, 93, 94, 95, 98, 99, 102, 104, 105, 106, 107, 109, 110, 111, 112, 116, 117, 121, 122, 123, 124, 125, 126, 127, 129, 130, 131, 136, 139, 140, 141, 150, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 169, 170, 172, 173, 174, 176, 177, 178, 179, 181, 182, 183, 184, 186, 187, 189, 190, 196, 197, 198, 205, 206, 207, 208, 209, 210, 211, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 228, 229, 230, 231 agricultural markets, 230, 11, 12, 85, 172, 174, 177, 183 leverage, 12, 35, 37, 91, 182 NCAA, 8, 11, 49, 200 predatory pricing, 164 Monopsony and Unexploited Scale Economies, ix, 101 monopsony power, xiii, 2, 14, 24, 29, 30, 35, 38, 41, 47, 53, 67, 68, 71, 74, 88, 89, 92, 94, 111, 112, 121, 156, 158, 159, 163, 173, 175, 179, 181, 186, 210, 214, 215, 217, 218, 230 Monopsony Power and Contract Power, x, 174 National Collegiate Athletic Association v. Board of Regents, 19 NCAA, x, 1, 10, 11, 15, 19, 20, 33, 35, 49, 114, 115, 188, 189, 190, 191, 193, 194, 196, 197, 198, 199, 200, 201, 202, 203, 204, 208, 234, 235, 236, 238, 240 Northwest Wholesale Stationers, Inc. v. Pacific Stationery and Printing Co, 119 oligopoly, 48, 136, 140 Overdeterrence, 228

Index Packers and Stockyards Act, 177, 179, 183, 184, 185, 186, 233 perishable commodities, 81 physicians, 12, 13, 15, 72, 85, 115, 136, 205, 206, 207, 208, 209, 210, 211, 213, 214, 215, 216, 219, 222 Posner, Richard, 239, 1, 21, 23, 30, 45, 47, 65, 147, 153, 227, 238 predatory bidding, 39, 76, 77 price discrimination, 14, 16, 22, 23, 73, 90, 99, 100, 102, 103, 167, 183 price fixing, 1, 18, 19, 29, 30, 31, 32, 33, 51, 78, 79, 106, 113, 114, 115, 116, 119, 123, 139, 147, 152, 153, 156, 157, 158, 159, 160, 161, 163, 170, 209, 225, 226, 227, 228, 230 private enforcement, 15, 146, 148, 227 Producer surplus, 44 reciprocal dealing, 27, 37, 174, 179, 180, 181, 182 Robinson Patman Act, 23, 90 rule of reason, 18, 19, 20, 26, 62, 79, 113, 114, 115, 116, 117, 118, 120, 121, 122, 159, 197, 201, 202, 204, 224, 225, 229

247

Sherman Act, vii, 13, 16, 17, 19, 20, 22, 27, 29, 30, 37, 38, 39, 68, 69, 71, 72, 75, 78, 85, 88, 89, 90, 91, 95, 107, 114, 121, 122, 123, 124, 147, 152, 199, 202, 205, 212, 233, 234 state action doctrine, 206, 211, 212 Tullock, Gordon, 45 transaction cost, 137 transaction costs, 99, 106, 107, 125, 135, 136, 138, 185, 229 tying, 18, 22, 24, 25, 26, 27, 28, 37, 89, 90, 91, 92, 113, 174, 178, 179, 180, 181, 182, 225 United States v. Socony Vacuum, 31 United States v. Topco, 18, 85, 224 Welfare Consequences of Collusion, 82 Welfare Effects of Monopsony, viii, 43 Weyerhaeuser, 1, 39, 45, 76, 77, 100, 164, 165, 166, 238 Weyerhaeuser Co. v. Ross-Simmons Hardwood Lumber Co, 39, 76, 164