CARSTENSEN - FINAL 4/10/2008 12:09 PM FALSE POSITIVES IN IDENTIFYING LIABILITY FOR EXCLUSIONARY CONDUCT: CONCEPTUAL ERROR, BUSINESS REALITY, AND ASPEN PETER C. CARSTENSEN* The decisions in Aspen and Trinko point the law of monopolization in different directions. Aspen offered a nuanced standard that focused on balancing the risks to competition from exclusionary conduct against the potential legitimate business needs of a monopolist to engage in such behavior. Trinko, in contrast, celebrates the right of the monopolist to exploit the market by excluding competition in the interest of, apparently, encouraging technological innovation rewarded by monopoly profits. Moreover, the case expresses great concern about false-positive decisions finding violations when in fact none exist. The Trinko Court and supportive commentators assume that such decisions are common and very harmful to economic efficiency. This Paper argues that Aspen and the cases following it provide the better approach to exclusionary monopolistic conduct. Monopoly is economically undesirable from both static and dynamic perspectives. The concern for false positives rests on incorrect and implausible assumptions while false negatives in fact create a more serious risk to the competitive process. Hence, the law of monopolization should embrace the stricter scrutiny mandated by Aspen, or if there are real risks of inefficient results, monopoly law should return to its historic mission of dissipating the monopoly power itself by dissolution of the monopolist or some other remedy that would in fact eliminate the undesirable and unnecessary monopoly power. Introduction................................................................... 296 I. The Road to Aspen and on to Trinko ............................. 297 II. Efficiency and Monopoly........................................... 304 A. Monopoly in Economics and Public Policy ................ 304 B. The Case against Strict Standards to Govern Monopolistic Conduct ......................................... 307 C. The False Premises behind the Concern with False Positives ......................................................... 309 1. Proposition One: Courts Make Frequent False- Positive Mistakes in Antitrust Decisions ............... 310 2. Proposition Two: Antitrust Rules are Effective in * Professor of Law, University of Wisconsin Law School. A much earlier version of this Paper was presented at the Law and Society Annual Meeting in 1998. The comments at that meeting as well as those of Lindsay Hampton have contributed greatly to the development of this piece. I am also indebted to Andrew Wang for diligent research assistance. Errors, unfortunately, remain my responsibility. CARSTENSEN - FINAL 4/10/2008 12:09 PM 296 WISCONSIN LAW REVIEW Barring Efficient Conduct ................................ 315 3. Proposition Three: Existing Market Structures Are Natural and Efficient ...................................... 318 D. The Danger of False Negatives .............................. 321 E. The Falsification of the False-Positive Mantra ............ 321 III. False Positives, False Negatives, and the Future of Monopoly Law....................................................... 322 A. The State of the Law and Its Prospects ..................... 322 B. What about a Structural Response to Monopoly? ......... 328 Conclusion.................................................................... 329 INTRODUCTION If, as Mao is reported to have said, “there are many roads to socialism,” then it seems unlikely that there is only one road to economic efficiency in a market economy. The central argument of this Paper is that the fear of false positives (i.e., holding exclusionary conduct by a monopolist unlawful when it should not be so held) is ill suited to the long-run needs of a market economy. It rests on mythical beliefs about the deterministic character of both law and economics. These myths rest on perceptions that there is but a single road to economic efficiency, that most markets in America are so narrow and restricted that only a monopolist can serve them efficiently, and that legal rules forbidding specific conduct effectively preclude monopolists from achieving efficiency-enhancing goals altogether. From these same myths there comes a stubborn hostility to structural remedies that would either dissipate the monopoly power itself or dissolve the monopolist—a corporate death penalty. Yet if false positives were a genuine concern, given that monopoly itself is economically undesirable, structural remedies eliminating monopoly power but leaving the enterprise free to conduct its business without regulation would seem the preferable strategy. Indeed, in business there is great enthusiasm for buying and selling corporate assets. Aspen Skiing Co. v. Aspen Highlands Skiing Corp.1 and its recent descendants, such as United States v. Microsoft Corp.,2 Conwood Co. v. U.S. Tobacco Co.,3 LePage’s Inc. v. 3M,4 and United States v. Dentsply International, Inc.,5 take a more Maoist view of the road to efficient markets and so risk the occasional false positive to avoid the 1. 472 U.S. 585 (1985). 2. 253 F.3d 34 (D.C. Cir. 2001) (en banc) (per curiam). 3. 290 F.3d 768 (6th Cir. 2002). 4. 324 F.3d 141 (3d Cir. 2003) (en banc). 5. 399 F.3d 181 (3d Cir. 2005). CARSTENSEN - FINAL 4/10/2008 12:09 PM 2008:295 False Positives 297 more serious risk of a false negative. Verizon Communications Inc. v. Law Offices of Curtis V. Trinko, LLP,6 by its narrow and grudging interpretation of Aspen, which is based on a concern for false positives and a desire to protect entrenched monopoly, revives the economic determinism found in some pre-Aspen cases. Trinko’s vision rests on bad economics and a misplaced faith in a lawless marketplace. This Paper joins the debate over the scope and nature of interventions by antitrust law against exclusionary conduct by monopolists very much on the side of the Aspen perspective of critical analysis. Part I traces the doctrinal road that led to Aspen but has continued to Trinko. It offers an explanation for the doctrinal confusion that besets monopoly law. By recognizing the tension between the Aspen and Trinko perspectives on exclusion, it provides a motivation for the following two Parts. Part II discusses the myth of efficient monopoly in both static and dynamic terms. The indisputable fact is that monopoly is not desirable from either perspective. Hence, rules that undermine monopoly, in general, make a positive contribution to the long-run efficiency of the economy, even if they may have short-run costs. Consequently, the greater danger for economic efficiency is in false negatives (decisions allowing anticompetitive, exclusionary conduct by monopolists) rather than false positives (decisions mistakenly banning specific conduct used to achieve a legitimate goal). Part III examines the implications of false-positive and false- negative concerns in recent monopoly cases to highlight the continued tension in this area of the law. This Part also discusses the general question of remedy and takes the position that if there were a legitimate concern for false-positive decisions that unduly regulate the conduct of monopolists, then the proper remedy would be eliminating the monopoly power rather than allowing it to continue under some dysfunctional regulation. I. THE ROAD TO ASPEN AND ON TO TRINKO The evolution of doctrine governing monopolistic conduct has followed an uncertain path. The cases prior to the late 1960s were largely government challenges to the merits of long-term, durable monopolies. The legal standards that emerged from Standard Oil Co. of New Jersey v. United States,7 United States v. American Tobacco Co.,8 6. 540 U.S. 398 (2004). 7. 221 U.S. 1 (1911). 8. 221 U.S. 106 (1911). CARSTENSEN - FINAL 4/10/2008 12:09 PM 298 WISCONSIN LAW REVIEW United States v. U.S. Steel Corp.,9 United States v. Aluminum Co. of America,10 United States v. E. I. du Pont de Nemours & Co.,11 and United States v. Grinnell Corp.12 addressed the question of when the government should be entitled to a remedy that would dissipate the monopoly power of such firms. The role of conduct in these cases was never very well theorized. The ultimate statement in Grinnell was that the “willful acquisition or maintenance” of monopoly was sufficient for a violation.13 Such a nearly no-fault standard is rational if the goal is to address failure of the ordinary dynamic, competitive processes by which the “centrifugal and centripetal forces” of the market overcome monopoly.14 The core of these cases was proof of a durable, substantial, and remediable monopoly. The relevance of conduct might have been only to support the inference that the monopoly was avoidable. Only by “embrac[ing]” all opportunities to expand production or engaging in unnecessary exclusion of equally efficient competitors could such a remediable monopolist survive.15 The issue was not the legal merits of the conduct itself but the overall consequence for the market: a durable and substantial monopoly. The remedy was to dissipate that monopoly power and not regulate its use.16 However, alongside these major monopoly cases was another set that was not clearly differentiated and that concerned specific conduct, usually exclusionary, of monopolists. This set of cases included United States v. Griffith 17 and Lorain Journal Co. v. United States.18 The issues in these cases were easily distinguished from the major structural-monopoly cases. They focused on the merits of specific conduct having exclusionary effect for which there was no excuse in the eyes of the courts. Some of these cases became targets for the emerging Chicago School19 because it was possible to suggest alternative explanations for the conduct that would have provided an efficiency 9. 251 U.S. 417 (1920). 10. 148 F.2d 416 (2d Cir. 1945). 11. 351 U.S. 377 (1956). 12. 384 U.S. 563 (1966). 13. Id. at 570–71. 14. Standard Oil Co. of N.J. v. United States, 221 U.S. 1, 62 (1911). 15. Aluminum Co. of Am., 148 F.2d at 431. 16. Standard Oil, 221 U.S. at 77–82. 17. 334 U.S. 100 (1948). 18. 342 U.S. 143 (1951). 19. The Chicago School includes a group of antitrust scholars such as Judges Robert Bork and Richard Posner as well as economists such as Ward Bowman and Aaron Director. See, e.g., Richard A. Posner, The Chicago School of Antitrust Analysis, 127 U. PA. L. REV. 925 (1979); Herbert Hovenkamp, Antitrust Policy After Chicago, 84 MICH. L. REV. 213 (1985). CARSTENSEN - FINAL 4/10/2008 12:09 PM 2008:295 False Positives 299 justification.20 To the critics, these decisions seemed to suggest a one- sided presumption against any conduct having exclusionary effects regardless of its justification or efficiency implications.21 The open-ended language of the structural-monopoly cases also invited private plaintiffs to challenge any conduct by a monopolist that had an adverse effect on the claimant. The result was a series of cases, including claims by Telex and Berkey, that essentially sought a combination of damages and a regulatory injunction that would require the monopolist to hold a price- and product-disclosure umbrella over its inefficient rivals.22 Other cases claiming predatory-pricing injuries raised similar concerns because the prices in fact charged were above the price that would have been charged in a competitive market.23 In another instance, SCM sought to piggyback on the Federal Trade Commission’s (FTC) successful effort to break up Xerox’s copier monopoly by claiming damages for past injuries resulting from its prior exclusion from the market.24 These plaintiffs relied on the expansive language of the structural-monopoly cases to claim damages and injunctions. Notably absent from these cases was any effort to dissipate the monopoly power of the dominant firm. In fact, given the apparent competitive capacity of Telex and Berkey, it is unlikely that they could have survived in such a market. The lower court decisions reinforced the concerns of those who feared that monopoly law would paralyze dominant firms and result in serious economic losses.25 Indeed, on the basis of the lower-court decisions, there was understandable concern for the potential adverse effect of unbridled monopoly law on efficiency. Moreover, juries in Telex Corp. v. International Business Machines Corp.26 and Berkey Photo, Inc. v. 20. See, e.g., ROBERT H. BORK, THE ANTITRUST PARADOX (1978); Robert H. Bork, The Rule of Reason and the Per Se Concept: Price Fixing and Market Division, 74 YALE L.J. 775 (1965); Lester G. Telser, Why Should Manufacturers Want Fair Trade?, 3 J.L. & ECON. 86 (1960). 21. See, e.g., BORK, supra note 20; Aaron Director & Edward H. Levi, Law and the Future: Trade Regulation, 51 NW. U. L. REV. 281 (1956); Telser, supra note 20. 22. Telex Corp. v. Int’l Bus. Machs. Corp., 510 F.2d 894 (10th Cir. 1975); Berkey Photo, Inc. v. Eastman Kodak Co., 603 F.2d 263 (2d Cir. 1979). 23. For a discussion, see Peter C. Carstensen, Predatory Pricing in the Courts: Reflections on Two Decisions, 61 NOTRE DAME L. REV. 928 (1986). 24. SCM Corp. v. Xerox Corp., 645 F.2d 1195 (2d Cir. 1981). 25. See, e.g., Milton Handler & Richard M. Steuer, Attempts to Monopolize and No-Fault Monopolization, 129 U. PA. L. REV. 125 (1980); Janusz A. Ordover & Robert D. Willig, An Economic Definition of Predation: Pricing and Product Innovation, 91 YALE L.J. 8 (1981). 26. 510 F.2d 894. CARSTENSEN - FINAL 4/10/2008 12:09 PM 300 WISCONSIN LAW REVIEW Eastman Kodak Co.27 not only found unlawful monopolization but also awarded very substantial damages.28 Lost in the academic noise about these decisions were other cases of greater merit.29 Moreover, the courts of appeals, in somewhat-incoherent opinions, reversed the verdicts in Berkey and Telex while affirming the trial court’s rejection of SCM’s claims. These results were entirely plausible even if the articulation was not. The courts were striving for a way to differentiate between the structural cases that the government had pursued and these conduct-oriented challenges that sought simply to force the monopolist to share its winnings. They got little help from the academic commentators. For example, Professors Phillip Areeda and Donald F. Turner, deeply concerned about the trial-court decision in the Telex case, focused on the standards for predation rather than on the question of appropriate relief or the important functional distinctions between challenges to monopoly itself and abuses of an otherwise-lawful monopoly position.30 In addition, the scholars associated with the Chicago School generally expressed alarm at the degree to which courts were invited to second guess business decisions.31 If markets are perfect or nearly perfect, no monopoly will survive unless it rests on efficiency, and all conduct will be rational because it would be irrational to try to entrench or extend monopoly power given that the forces of the market would overwhelm it in any event. The FTC joined the effort to reformulate the duty of monopolists. The In re Borden, Inc. (ReaLemon) 32 and In re E. I. duPont de Nemours & Co. (Titanium Dioxide) 33 cases provided vehicles for an increasingly conservative agency to express its concerns and seek to articulate a more permissive standard for judging the conduct of monopolists. Again, the FTC did not differentiate between challenges to the monopoly itself and challenges to specific conduct by the monopolist. Thus, the courts, the agencies, and the academic observers all agreed that there was a serious problem with monopoly law because 27. 603 F.2d 263. 28. See Telex, 510 F.2d at 897–98; see also Berkey, 603 F.2d at 268. 29. See, e.g., Greyhound Computer Corp. v. Int’l Bus. Machs. Corp., 559 F.2d 488 (9th Cir. 1977). 30. See Phillip Areeda & Donald F. Turner, Predatory Pricing and Related Practices under Section 2 of the Sherman Act, 88 HARV. L. REV. 697 (1975). This article has spawned a vast literature that largely begs the question of whether price setting alone should ever be a basis for liability. See, e.g., Carstensen, supra note 23, at 969–70. 31. See, e.g., BORK, supra note 20. 32. 92 F.T.C. 669 (1978). 33. 96 F.T.C. 653 (1980). CARSTENSEN - FINAL 4/10/2008 12:09 PM 2008:295 False Positives 301 it exposed monopolists to unreasonable risks of liability for conduct that these observers regarded as economically beneficial. The United States Supreme Court’s decision in Aspen provided renewed focus on the merits of the exclusionary conduct of the monopolist. Grounded in Lorain Journal ’s condemnation of naked exclusion, the Aspen decision upheld a functional approach to the problem of exclusionary monopoly conduct that “unnecessarily excludes or handicaps competitors.”34 Unfortunately, the opinion did not set forth criteria to implement this overall standard. Instead, the Court focused on a fact-intensive, case-specific review of the conduct that resulted in the conclusion that the jury could reasonably have found that the refusal to permit any kind of combination-ticket sales was without any legitimate business purpose. Given the factual analysis, the Court was not obliged to consider whether, if there had been a legitimate purpose for some refusal, the monopolist had a duty to use the least restrictive means available to implement its lawful goal. This issue is significant because a monopolist can obtain a greater degree of freedom if it only needs to show that its goal was consistent with that of a nonmonopolist in a similar situation, regardless of the competitive effect of the specific means used to achieve the goal. A standard requiring the monopolist to avoid unnecessarily restrictive means to achieve its lawful goal, on the other hand, would impose a stronger constraint on the conduct at issue. Moreover, the opinion did not set forth any criteria to determine what objectives of a monopoly business were “legitimate.”35 The opinion also completed the conversion of monopoly law to a conduct model by its failure to distinguish the structural-monopoly decisions36 as ones involving distinct issues and remedies.37 They simply ceased being important parts of the canon of monopoly law. Instead, monopoly law after Aspen has come to focus on the specific conduct of the monopolist.38 Such a shifted perspective necessarily required different criteria than those relevant to deciding if a durable, but remediable, monopoly warranted judicial reorganization. Regrettably, the Aspen transformation did not distinguish between the 34. Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585, 597 (1985) (quoting the jury instructions); see also id. at 605 (considering whether the exclusionary conduct has “impaired competition in an unnecessarily restrictive way”). 35. See Aspen, 472 U.S. 585. 36. See supra notes 7–16 and accompanying text. 37. See Aspen, 472 U.S. 585. 38. See, e.g., United States v. Microsoft Corp., 87 F. Supp. 2d 30 (D.D.C. 2000), aff’d in part, 253 F.3d 34 (D.C. Cir. 2001). This is true even in the one post- Aspen government case with important structural implications. See id. CARSTENSEN - FINAL 4/10/2008 12:09 PM 302 WISCONSIN LAW REVIEW kinds of monopoly cases,39 although the statutory language invited such a differentiation.40 The Court followed with the Eastman Kodak Co. v. Image Technical Services., Inc.41 decision, which extended the Aspen analysis to a case where the source of market power included intellectual- property rights.42 Again the articulation of the rationale and standards left much to be desired, but the functional focus of the analysis was reasonably clear. The Court balanced these decisions with others, notably Brooke Group Ltd. v. Brown & Williamson Tobacco Corp.,43 which showed an unwillingness to probe deeply the reasons for discounting prices.44 The Aspen and Kodak decisions have, however, excited a great deal of concern among practitioners and some academics. The concern is that these cases make it too easy for plaintiffs to prevail when the dominant firm imposes some exclusionary burden on them. The speculation is that there might be an efficiency justification for the conduct. As intellectual property and mergers have created more highly concentrated markets, this concern has grown. The mantra is that there is a great danger of false-positive decisions that incorrectly restrict the freedom of action of dominant firms and cause untold economic losses. This concern is a recurring theme of those who oppose Aspen and the cases that apply its critical criteria to review exclusionary conduct.45 The recent Trinko opinion stands in apparently direct opposition to Aspen.46 As in Aspen, a monopolist excluded its competitors from access to essential resources for competition.47 Although Justice Antonin Scalia attempted in his opinion to distinguish the cases on their 39. See Aspen, 472 U.S. 585. 40. Basically, “monopolization” could have applied only to cases challenging the legality of the monopoly itself and seeking a remedy that would dissipate the monopoly power; “attempt to monopolize” could then provide standards for reviewing the merits of specific conduct that the plaintiff claimed excluded competition or unfairly exploited monopoly power. See Sherman Act § 2, 15 U.S.C. § 2 (2000). 41. 504 U.S. 451 (1992). 42. Id. 43. 509 U.S. 209 (1993). 44. Id. 45. See, e.g., Thomas C. Arthur, The Costly Quest for Perfect Competition: Kodak and Nonstructural Market Power, 69 N.Y.U. L. REV. 1 (1994); Daniel F. Spulber & Christopher S. Yoo, Mandating Access to Telecom and the Internet: the Hidden Side of Trinko, 107 COLUM. L. REV. 1822 (2007). 46. Compare Verizon Commc’ns Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398 (2004) (suggesting that failure to cooperate with competitors is not a valid claim under the Sherman Act), with Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585 (1985) (suggesting a monopolistic competitor’s failure to cooperate with competitors may violate the Sherman Act). 47. See Trinko, 540 U.S. at 403–05. CARSTENSEN - FINAL 4/10/2008 12:09 PM 2008:295 False Positives 303 facts (arguing, for example, Congress and the FCC had mandated the sale of the products and services in Trinko while in Aspen there was evidence that competitive firms had in fact sold the essential elements), the distinctions were not economically significant.48 More to the point, perhaps, the opinion focused on the existence of alternative remedies that directly addressed the abuse of monopoly power though administrative-agency oversight and control.49 This strand of the opinion is not inconsistent with an indirect recognition that the original goal of the Sherman Act’s prohibition of monopoly was to dissipate the monopoly power itself. But it is fair to say that the thrust of the opinion was not to return to the older view of monopolization as a structural problem. The Trinko opinion stressed the need to protect monopolists and their “dream[s]”50 from the risk of liability for conduct that affects their rivals: The opportunity to charge monopoly prices . . . is what attracts “business acumen” . . . ; it induces risk taking that produces innovation and economic growth. . . . Compelling such firms to share the source of their advantage . . . may lessen the incentive for the monopolist, the rival, or both to invest in those economically beneficial facilities.51 Trinko, therefore, stands in strong contrast to Aspen, which it characterized as being “at or near the outer boundary of § 2 liability.”52 The decision represents a return to the pre-Aspen cases with their deep concern for the rights and privileges of monopoly. 48. Id. at 409–10. The opinion acknowledged but then ignored the fact that Congress was seeking to transform the telephone industry from a regulated monopoly into an industry governed by competition. Id. at 415–16. To do this, Congress mandated that the incumbent monopolists must sell specific components for use by competitors rather than requiring the dissolution of the monopolies themselves. Id. at 402. The Court had previously expressed its concerns about some of the pricing policies that the FCC had put in place for these new products. See AT&T Corp. v. Iowa Utils. Bd., 525 U.S. 366 (1999). Given the lack of history of competition in the market for component services, the Court’s emphasis on the lack of experience with a market in component services seems disingenuous at best. See id. at 366. 49. Trinko, 540 U.S. at 413. Trinko involved the telecommunications industry, which is the subject of extensive statutory control. It is unclear whether Trinko, despite its rhetoric, will govern exclusionary-conduct cases generally or whether it sets generous standards for exclusionary conduct where direct regulation can and did respond to the competitive concerns. 50. Id. at 409. 51. Id. at 407–08. 52. Id. at 409. CARSTENSEN - FINAL 4/10/2008 12:09 PM 304 WISCONSIN LAW REVIEW To determine whether Aspen should survive, it is important to look at two questions. First, how valid is the concern for false positives in general? Second, what are the implications of false negatives for economic efficiency? The responses offered in the next Part provide a general economic-policy perspective on the question of whether it is important to protect monopolists from the risks of false-positive decisions. That analysis provides the basis for this Paper’s conclusion that Aspen provides the better approach to exclusionary conduct and may (as well as ought to) dominate Trinko’s promonopoly stance. II. EFFICIENCY AND MONOPOLY A. Monopoly in Economics and Public Policy Efficient production of goods and services is a fundamental concern for any economy. Of equal importance for the long-run growth of an economy is the maintenance of the dynamics in the systems that stimulate and support innovation and evolution. Of the two, in the long run the dynamics of markets are more important than achieving optimal static efficiency at any point in time. The choice between these goals is an issue only if there is an inconsistency between them. Indeed, although the Schumpeterian vision of creative destruction may pose such a conflict,53 it actually mandates a stricter view of monopolistic conduct that limits or delays the destruction of market power. It is textbook economics that monopoly is inconsistent with static efficiency because the monopolist will raise its price and reduce its output to extract economic rent. As a result, the economy experiences suboptimal production and allocative inefficiency. Rational monopolization includes a long-term perspective in which the future casts a “large shadow” over the present.54 The long run invites investments that increase the durability and strength of the underlying market position but that increase neither productive nor distributional efficiency. The monopolist will expend its resources to entrench and protect its continued ability to reap those profits.55 Specifically, it may 53. JOSEPH A. SCHUMPETER, CAPITALISM, SOCIALISM AND DEMOCRACY 81–86 (2d ed. 1947). 54. ROBERT AXELROD, THE EVOLUTION OF COOPERATION 174 (1984); see also Peter C. Carstensen, Vertical Restraints and the Schwinn Doctrine: Rules for the Creation and Dissipation of Economic Power, 26 CASE W. RES. L. REV. 771 (1976). 55. Judge Posner has shown that the rational monopolist will expend nearly all its monopoly profits in this way in order to protect its long-run monopoly position. Richard A. Posner, The Social Costs of Monopoly and Regulation, 83 J. POL. ECON. 807, 824–25 (1975). CARSTENSEN - FINAL 4/10/2008 12:09 PM 2008:295 False Positives 305 engage in activities that frustrate and deter both existing and potential competition. A law-abiding monopolist seeking a legitimate, efficiency- enhancing objective, having a choice among lawful strategies, would rationally choose the one that could accomplish its primary goal and also increase or preserve barriers to entry or reduce the potential for competitors encroaching on its market position. It would be rational to do this even if the strategy chosen involved somewhat greater short- term costs or sacrificed some of the efficiency gain, provided that the expected gains from the exclusionary effect offset those extra costs.56 Indeed, any other choice would be economically irrational and might even constitute a breach of fiduciary duty to the monopolist’s shareholders.57 The incentives of a rational monopolist, therefore, would seem to suggest that any conduct having exclusionary effect should be subject to strict scrutiny. A different complaint against monopoly is that it is wasteful and inefficient. The monopolist, entrenched behind its entry barriers, can expend resources on excessive executive pay, lavish offices, and acquisitions that bulk up the corporate enterprise without contributing any gains to the economy. The monopolist will have less incentive to seek out and develop lower-cost methods of production or expand the array of products available to consumers. The monopolist looks to its own interests and seeks to protect them in a context where it has a significant degree of discretion. This undesirable dimension of monopoly is captured in the concept of X-inefficiency.58 Reducing the barriers to entry creates a positive incentive for such monopolist to operate in a more efficient manner. 56. The rational monopolist seeks to maximize profits, not efficiency. Hence, it will balance the costs of increasing the strength or durability of its monopoly power against the costs of doing so. See Posner, supra note 55, at 824–26. When the monopolist can improve efficiency and also increase exclusion by its choice of conduct, in making a decision it will balance any increased costs of available exclusionary methods against the projected gains to its monopoly, net of any efficiency loss that might result from using a lower-cost but less exclusionary option. It would be irrational for a monopolist not to consider both efficiency and exclusionary gains in selecting among the ways to achieve an efficiency-enhancing goal. 57. The managers of a corporate enterprise owe a fiduciary duty to the shareholders to maximize their wealth. Hence, the failure to balance the gains from lawful exclusion of competition against any increased costs would seem to constitute a breach of that duty. In practice, the business-judgment rule probably shelters managers from direct liability, but managerial incentive structures that reward profitability with prestige and bonuses may well provide a more relevant alternative. 58. See HARVEY LEIBENSTEIN, GENERAL X-EFFICIENCY THEORY AND ECONOMIC DEVELOPMENT 17, 43–44, 160 (1978). CARSTENSEN - FINAL 4/10/2008 12:09 PM 306 WISCONSIN LAW REVIEW The short-run efficiency costs of monopoly are well known and well proven. But in some respects, the greater costs are in the dynamics of technological change and growth. A monopolist has no incentive to support technological innovation that could undermine its dominant position in the market. Moreover, having sunk investments in existing technology, it may well delay or refuse to pursue work on new technology until it has accounted for its past investments. In addition, when only one decision maker controls the evolution of technology, there is a real risk that its internal priorities and decision processes will induce it to pursue the wrong or less desirable paths.59 Competition in innovation, given the extreme uncertainty of what is likely to prove most desirable in some future world, is a long-run feature of vast importance to a dynamic economy. The famous aphorism that the greatest benefit of monopoly is the quiet life speaks strongly to this aspect of monopoly. The classic monopoly cases provide direct proof of these costs. Standard Oil used its monopoly power to keep prices higher and output lower as well as raise its rivals’ costs.60 Indeed, after the breakup of the monopoly, the history of antitrust is that the survivors sought repeatedly to resurrect market power through collusion and restrictive downstream agreements.61 More recently, the breakup of AT&T released an enormous amount of technology held back because it was inconsistent with that company’s economic self-interest.62 The transformation of the copying-equipment market after the dissolution of Xerox’s patent monopoly provides yet another example—new technology flourished and prices dropped dramatically.63 59. See Richard J. Gilbert & Steven C. Sunshine, Incorporating Dynamic Efficiency Concerns in Merger Analysis: The Use of Innovation Markets, 63 ANTITRUST L.J. 569, 570, 574–76, 579 (1995). 60. Donald J. Boudreaux & Burton W. Folsom, Microsoft and Standard Oil: Radical Lessons for Antitrust Reform, 44 ANTITRUST BULL. 555, 559 (1999) (noting that the price of kerosene in the 1890s was ten to twelve times greater than production cost); Elizabeth Granitz & Benjamin Klein, Monopolization by “Raising Rivals’ Costs”: The Standard Oil Case, 39 J.L. & ECON. 1, 8–10 (1996) (describing how Standard Oil collected a rebate from the railroads calculated on the basis of the quantity of oil shipped by its rivals). 61. See, e.g., Simpson v. Union Oil Co. of Cal., 377 U.S. 13 (1964); Standard Oil Co. of Cal. v. United States, 337 U.S. 293 (1949); United States v. Socony-Vacuum Oil Co., 310 U.S. 150 (1940). 62. See, e.g., MCI Commc’ns Corp. v. AT&T, 708 F.2d 1081, 1098 (7th Cir. 1983) (noting that AT&T blocked the introduction of long-distance microwave technology to protect its investment in land lines). 63. In re Xerox Corp., 86 F.T.C. 364 (1975); see Willard K. Tom, The 1975 Xerox Consent Decree: Ancient Artifacts and Current Tensions, 68 ANTITRUST L.J. 967 (2001). CARSTENSEN - FINAL 4/10/2008 12:09 PM 2008:295 False Positives 307 In short, there are powerful reasons for condemning monopoly. This fundamental fact is too often overlooked in contemporary policy discussions and judicial decision making. It is useful next to consider the reasons for this sympathy for the “malefactors of great wealth.” B. The Case against Strict Standards to Govern Monopolistic Conduct Although monopoly is undesirable, many antitrust academics and judges believe that public policy should be very cautious in upholding any challenge to such enterprises.64 But large firms controlling significant sectors of the economy invite direct government regulation and are more able to seek out and obtain government protection. A central theme of public-choice analysis is the rent seeking and protectionism provided by government process. Banks or car makers that are “too big to fail” get protection and subsidies. Yet many of those who subscribe to public-choice theories when it comes to government regulation, as do some antitrust scholars,65 seem unwilling to insist on more rigorous control over either the structure or conduct of firms with dominating market positions despite the public-choice analysis showing that these firms are likely to distort government actions.66 One justification for monopoly is the dynamic interest in technological progress. The Schumpeterian hypothesis is that monopoly profits are the great incentive for innovation.67 Over time, monopoly will arise with technological advances only to be replaced by new technology. This vision leaves a major unanswered question, How 64. See, e.g., Arthur, supra note 45, at 5. 65. Fred S. McChesney, Public Choice and the Chicago School of Antitrust, in 2 THE ENCYCLOPEDIA OF PUBLIC CHOICE 769 (Charles K. Rowley & Friedrich Schneider eds., 2003). Professor McChesney is one of the leading critics of how antitrust law is enforced. See, e.g., THE CAUSES AND CONSEQUENCES OF ANTITRUST: THE PUBLIC-CHOICE PERSPECTIVE (Fred S. McChesney & William F. Shughart II eds., 1994); Robert D. Tollison, Public Choice and Antitrust, 4 CATO J. 905, 910–11 (1985), available at http://www.cato.org/pubs/journal/cj4n3/cj4n3-12.pdf. 66. A question beyond the scope of this Paper is why public-choice scholars do not question the expansion of intellectual-property rights. This expansion by legislative and judicial fiat protects and entrenches select economic interests at the expense of the overall market and consumer welfare. Apparently, the term property mesmerizes the public-choice analysis such that its proponents are not able to get around the use of a misleading label that conceals a governmental allocation of economic rights. See ADAM B. JAFFE & JOSH LERNER, INNOVATION AND ITS DISCONTENTS (2004); Mark A. Lemley, Property, Intellectual Property, and Free Riding, 83 TEX. L. REV. 1031, 1032 (2005). 67. This belief is a significant basis for the Trinko opinion. See Verizon Commc’ns Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398, 407 (2004). CARSTENSEN - FINAL 4/10/2008 12:09 PM 308 WISCONSIN LAW REVIEW much monopoly, for how long, is essential to achieve this churning of the economy? The implicit conclusion of those who rely on this model to justify monopoly is that the reward must be very substantial.68 Hence, the legal system should be very reluctant to interfere with the conduct of the monopolist. Thus stated, the model seems naive and questionable. The goal of monopolists, once they have monopolies, is to preserve and protect themselves. Hence, it would seem that the advocates of this model should be deeply concerned with the scope and duration of intellectual- property rights that can frustrate innovation as well as with other conduct that creates significant barriers to entry. Such rights and conduct directly increase the costs for the next wave of technological advance. Yet it is hard to find in this literature recognition of the need to keep pathways open for new entry.69 The concern is largely for the already-successful innovator and its reward. One explanation for this protectionist approach to incumbent monopolists is a belief that such firms are the only entities that can or will innovate but will do so only if promised massive rewards for such creative destruction of their own prior market power. Such a belief stands Schumpeter’s model on its head and excuses entrenchment of existing monopoly. A second strand to the defense of monopoly focuses on the social and economic costs of conduct-oriented remedies themselves. If a court or agency acts as the reviewer of business decisions, it has the potential to stultify and rigidify conduct in ways that do not advance either static efficiency or dynamic progress. An example of this danger is the ReaLemon decision of the FTC. The FTC found that Borden sold ReaLemon at below-cost prices (i.e., predatory prices) in certain geographic markets in order to interfere with the growth of small competitors in the reconstituted-lemon-juice market. The remedy was a price-regulation order that commanded that Borden not price below the FTC’s definition of a reasonable price. Such regulatory control, even if well informed, could create serious burdens on the operation of a firm. It must always consider and perhaps even check with its government controller before offering discounts.70 The decree in the Microsoft case may present similar risks. The remedy involves the creation of a panel of experts who are to review Microsoft’s behavior and consider complaints from third parties. The recommendations of the panel on these issues come to the court for action. This has the potential to create continuing judicial review of a 68. Cf. id. at 407, 410–11. 69. Notable exceptions are JAFFE & LERNER, supra note 66, and Lemley, supra note 66. 70. In re Borden, Inc. (ReaLemon), 92 F.T.C. 669 (1978). CARSTENSEN - FINAL 4/10/2008 12:09 PM 2008:295 False Positives 309 wide range of business decisions. So far it is not apparent how intrusive these reviews will be, but the fact that Microsoft agreed to the process suggests that it did not foresee them creating serious obstacles to its ongoing business plans.71 The concern with intrusive regulation of corporate business decisions would appear to have some validity. However, a central question is whether such regulatory remedies are a common element of most monopoly-conduct cases. The more usual challenge is to a business practice that can be forbidden without requiring extensive oversight of the full range of business decisions within that enterprise. Indeed, as the United States Court of Appeals for the Ninth Circuit demonstrated in the Kodak case, it is possible to impose a limit on exclusionary conduct (e.g., denial of access to repair parts) without entering into price regulation by the simple expedient of requiring equal treatment of all buyers.72 C. The False Premises behind the Concern with False Positives Putting to one side the cynical suggestion that the concerns with false positives are advanced simply to protect allies without regard to the merits or consistency of the positions taken, it is possible to postulate the premises on which such positions would rest. There are two primary premises that rest on a third premise about remedy. First, antitrust cases are too complex for judges, so the judges err a significant part of the time by making false-positive decisions.73 Second, these errors result in substantial social costs because antitrust law is a very powerful and effective instrument in restricting efficient economic behavior.74 Third, the natural order to the economy reflected by the current structure of markets means that any restructuring of monopoly markets to eliminate monopoly power would result in substantial inefficiency.75 The implication of these three premises is that decision makers should be deeply concerned with the risk of false positives; reluctant to impose any significant constraints on market behavior by dominant firms; and seek rarely, if ever, the dissolution of dominant firms. Indeed, if these premises were correct, they would 71. United States v. Microsoft Corp., 231 F. Supp. 2d 144 (D.C. Cir. 2002) (decree). 72. Image Technical Servs., Inc. v. Eastman Kodak Co., 125 F.3d 1195, 1224–28 (9th Cir. 1997) (price-regulation issue). 73. Frank H. Easterbrook, The Limits of Antitrust, 63 TEX. L. REV. 1 (1984). 74. Arthur, supra note 45. 75. Easterbrook, supra note 73, at 39; Arthur, supra note 45, at 5. CARSTENSEN - FINAL 4/10/2008 12:09 PM 310 WISCONSIN LAW REVIEW constitute a basis to reject any but the most limited interventions in the market. The three propositions reflect a kind of deterministic view of economic organization similar to Marxist conceptions. As a result, they suffer from the same failure as Marxism to appreciate the many roads that exist within any real economic system to achieve efficiency and progress. In a nondeterministic world, each of the three premises is highly questionable as the following critical review demonstrates. 1. PROPOSITION ONE: COURTS MAKE FREQUENT FALSE-POSITIVE MISTAKES IN ANTITRUST DECISIONS The proposition asserts that courts not only err frequently in antitrust cases but that those errors tend to be false positives. Why courts should err primarily in this direction is not explained. If anything, scholars would predict that courts, as conservative institutions, would be more likely to provide false-negative decisions. “Proof” of the premise of false positives rests largely on hypotheticals based on some plausible economic theories. The analysis takes the form of offering an alternative, efficiency-based explanation for the conduct condemned in leading cases. The proof of error in the decision consists, first, of the demonstration that under some assumptions the alternative explanation could explain the conduct. Second, the analysis often offers a narrow anticompetitive explanation for the conduct, usually on the basis of the court’s language. The narrow anticompetitive explanation is then disproved by claims derived from economic theory. The refutation of the anticompetitive explanation is said to prove that the alternative efficiency explanation is the only possible one, thereby demonstrating that a false-positive decision has occurred. The weaknesses with this method are many.76 A central difficulty is that such analyses fail to consider the actual facts of the cases and whether those facts were consistent with an anticompetitive conclusion even if it was different from the way the court framed the issues. Monsanto v. Spray-Rite Service Corp.77 and United States v. Topco Associates, Inc.78 provide good examples. Both cases are 76. See Peter C. Carstensen & Bette Roth, The Per Se Legality of Some Naked Restraints: A [Re]conceptualization of the Antitrust Analysis of Cartelistic Organizations, 45 ANTITRUST BULL. 349 (2000) (discussing further the weaknesses with this approach to social science). 77. 465 U.S. 752 (1984). 78. 405 U.S. 596 (1972). CARSTENSEN - FINAL 4/10/2008 12:09 PM 2008:295 False Positives 311 strongly criticized as examples of false positives that condemned efficiency-enhancing conduct. Topco’s restriction on its members’ resale locations (i.e., exclusion of competition) is understood to have been an efficient way to deal with free-riding risks that would arise from new entrants being able to take advantage of the prior promotion of the Topco house brand. Similarly, in Monsanto, the assumption is that distributors of Monsanto’s herbicides were engaging in complex and costly selling activities that were vulnerable to free riding. Hence, Monsanto needed to exclude Spray-Rite—a price cutter—because it must have failed to provide necessary services. Neither free-riding story has any merit. Monsanto itself specifically disclaimed the free-rider argument and sought to excuse its actions as being unilateral.79 This was a rational litigation strategy because Spray-Rite’s counsel had built a record showing that there had been no complaints about its services.80 Moreover, customers of herbicides (i.e., farmers) are repeat players whose needs for information and other services will continue from year to year.81 Because the need for services is continuous, there is little or no risk of free riding by buyers. The distributor who is victimized once will not provide valuable services in the future. The buyer, having betrayed its former distributor for a one-time discount, will have to rely on its new supplier to provide needed services in the future. But, in fact, farmers do not rely primarily on the sellers of herbicides for advice or other presale service. They use the agricultural-extension agents provided by the states who offer better and more neutral information and related services.82 These facts suggest that some other, anticompetitive purpose must have motivated the elimination of an efficient, price-competitive distributor.83 Despite this readily available information, some stubbornly cling to their perceptions that Monsanto illustrates a false positive. 79. See Monsanto, 465 U.S. at 757. 80. Id. at 767. 81. Steven P. Schneider, Comment, A Functional Rule-of-Reason Analysis for Resale Price Maintenance and Its Application to Spray-Rite v. Monsanto, 1984 WIS. L. REV. 1205, 1262–63 (1984). 82. Id. at 1265–66. 83. Id. at 1268–69 (suggesting that Monsanto wanted to entrench its market position after its patent expired and so promoted a distributor cartel so that the distributors would have less incentive to promote generic products after the patent expired, even though having Spray-Rite as a vigorous price competitor seriously undermined this cartel); see also Peter C. Carstensen, The Competitive Dynamics of Distribution Restraints: The Efficiency Hypothesis versus the Rent-Seeking, Strategic Alternatives, 69 ANTITRUST L.J. 569, 604 (2001); cf. JTC Petroleum Co. v. Piasa Motor Fuels, Inc., 190 F.3d 775, 777–78 (7th Cir. 1999) (two-level cartel explained as a rational strategy to control prices in the highway-paving market). CARSTENSEN - FINAL 4/10/2008 12:09 PM 312 WISCONSIN LAW REVIEW Topco provides a second striking example of the same kind of myth-making. Topco members had total combined sales that were five to ten times larger than necessary to achieve efficient scale for acquiring house brands.84 Moreover, house brands should be good- quality, cost-effective products, but grocery stores do not spend money to advertise their house brands. The potential, therefore, that a grocery store obtaining Topco products could achieve any significant competitive advantage over its rivals is doubtful even in theory. In fact, despite the absolute condemnation of the territorial restraints in its opinion, the Supreme Court subsequently upheld, by an eight–one vote, a decree that allowed Topco to continue to restrict competition among its members provided it could demonstrate that there was an actual need to do so.85 For the ten-year life of that decree, Topco never once sought to impose any restraint on its members, many of whom came to compete with each other.86 Today, Topco remains a major distributor of house brands whose members frequently compete with each other.87 Thus, the free-riding hypothesis was and is totally false as a matter of fact with respect to Topco. Therefore, the decision is not an example of a false positive despite the myths that scholars and casebook editors perpetuate.88 Further, there is good basis to believe that the Court’s decisions in the 1960s and early 1970s concerning distribution restraints, often pictured as false positives, had more rationality than critics were prepared to accept.89 Although United States v. Arnold, Schwinn & Co.90 is usually described as imposing a per se prohibition on territorial allocations, the Court in fact permitted Schwinn to retain significant control over its distribution system provided it did not pass title to the intermediate distributors who were otherwise its agents acting as independent contractors.91 Thus, the Schwinn decision created no insurmountable barriers to efficient distribution.92 Finally, contrary to 84. United States v. Topco Assocs., Inc., 405 U.S. 596, 614 n.1 (1972) (Burger, C.J., dissenting). 85. See United States v. Topco Assocs., Inc., 414 U.S. 801 (1973). 86. Willard F. Mueller, The Sealy Restraints: Restrictions on Free Riding or Output?, 1989 WIS. L. REV. 1255, 1268–69 n.74. 87. See Topco, http://www.topco.com (last visited Feb. 25, 2008). 88. See generally Peter C. Carstensen & Harry First, Rambling Through Economic Theory: Topco’s Closer Look, in ANTITRUST STORIES 171 (Eleanor M. Fox & Daniel A. Crane eds., 2007) (discussing Topco in detail). 89. See Carstensen, supra note 83. 90. 388 U.S. 365 (1967). 91. Id.; see Carstensen, supra note 54, 826–31 (discussing the ambiguous functional analysis in the Schwinn case). 92. See Robert C. Keck, The Schwinn Case, 23 BUS. LAW. 669 (1968). CARSTENSEN - FINAL 4/10/2008 12:09 PM 2008:295 False Positives 313 popular impression, in seeking to reverse Schwinn, Sylvania did not offer a free-riding defense for its restraints, although, ironically, the plaintiff did.93 It only claimed to lack market power in the market for television sets.94 Of course, if that statement were true, its restraints would have been meaningless for either efficiency-enhancing or anticompetitive purposes. The Supreme Court, perhaps responding to compelling arguments from the briefs of the amici, failed to observe that the defendant’s own evidence did not support a free-riding claim.95 Thus, Sylvania is as likely to be a false negative as a false positive.96 Ironically, conventional antitrust scholarship regards the Board of Trade v. United States 97 decision as the leading example of a false negative, that is, a decision permitting an unjustified restraint of trade. That case involved an absolute restraint on price competition among members of the exchange in bidding to rural elevators for grain to be received at the exchange after the close of the exchange. A detailed analysis of the record and underlying historical materials shows that the restraints were in fact ancillary to the function of the exchange as a joint venture and specifically aimed at serious problems of free riding and opportunistic conduct.98 Hence, the decision upholding the price restraints was in fact correct. Other examples of false negatives include Paschall v. Kansas City Star Co.99 and NicSand, Inc. v. 3M Co.100 In Paschall, the newspaper refused to sell papers to independent delivery services and insisted that they become agents of the paper so that the paper could control both price and delivery services. Many antitrust scholars believe that this 93. Peter C. Carstensen, Annual Survey of Antitrust Developments 1976– 1977, 35 WASH. & LEE L. REV. 1, 19–21 (1978). 94. Id. at 19–20. 95. Id. at 24 n.139. 96. The same is true for Business Electronics Corp. v. Sharp Electronics Corp., 485 U.S. 717 (1988), where the explanation for eliminating competition among retailers might well be similar to that postulated in the Monsanto case. See supra note 83. Other cases are just irrational, such as State Oil Co. v. Khan, 522 U.S. 3 (1997), which rests on an assumption of myopic bilateral monopoly pricing in the context of retail gasoline markets without a shred of evidence to support such a theory. See Peter Carstensen & David Hart, Khaning the Court: How the Antitrust Establishment Obtained an Advisory Opinion Legalizing “Maximum” Price Fixing, 34 U. TOL. L. REV. 241 (2003). However, in that case, dismissing the antitrust claims may not have been wrong. 97. 246 U.S. 231 (1918). 98. See Peter C. Carstensen, The Content of the Hollow Core of Antitrust: The Chicago Board of Trade Case and the Meaning of the “Rule of Reason” in Restraint of Trade Analysis, in 15 RESEARCH IN LAW AND ECONOMICS 1 (Richard O. Zerbe Jr. & Victor P. Goldberg eds., 1992). 99. 727 F.2d 692 (8th Cir. 1984) (en banc). 100. 507 F.3d 442 (6th Cir. 2007). CARSTENSEN - FINAL 4/10/2008 12:09 PM 314 WISCONSIN LAW REVIEW kind of vertical integration is economically efficient because it eliminates a bilateral monopoly (e.g., monopoly paper and monopoly distributor) in which, if the two monopolists are myopic, the predicted result is that prices will exceed the optimal monopoly level with a consequent excessive reduction in output. The record in the Paschall case showed, however, that after the integration, more than 90 percent of the subscribers would have the same or higher prices and reduced services.101 This evidence established that, in fact, the bilateral monopoly was producing lower prices and greater output than an integrated monopolist, thus serving the ultimate interest of consumers. The majority nevertheless sided with the newspaper in its effort to claim its “full monopoly profit.”102 Indeed, the majority characterized the retailers as “robbing” the Star Company of its rightful monopoly profit.103 This is a classic case of a false-negative analysis. It empowers dominant, upstream firms to take control of their downstream distributors solely to achieve greater market exploitation. The United States Court of Appeals for the Sixth Circuit, sitting en banc, upheld the dismissal of NicSand’s complaint against 3M that had charged 3M with exclusionary conduct resulting in monopolization of the specialized market for abrasives used by do-it-yourselfers for car repairs.104 The allegations were that 3M had made very large upfront payments to the five or six key retailers of this product line in order to obtain multiyear exclusive-dealing contracts.105 The upfront payments were allegedly equal to or greater than the expected profits from the first year or more of sales.106 The effect on NicSand—a small, specialized manufacturer107—was to drive it from the market entirely and into bankruptcy.108 While one-year exclusive deals were the industry norm and included the buyout of the inventory of the producer being replaced, the allegations in the case were that 3M used its deep pockets to raise the amount of the upfront payment and extend the duration of exclusivity without any business justification except foreclosing competition.109 In combination, this created an insuperable 101. Paschall, 695 F.2d 322, 330 (8th Cir. 1982), aff’d en banc, 727 F.2d 692 (8th Cir. 1984). 102. Paschall, 727 F.2d at 703. 103. Id. 104. NicSand, 507 F.3d at 449, 459. 105. Id. at 447–49. 106. Id. at 453. 107. NicSand is an Ohio corporation, started in 1982, that produces a variety of products for automotive-body repair. Of these products, automotive-sandpaper sales accounted for half of the company’s total revenue in 1996. Id. at 447. 108. Id. at 448–49. 109. Id. at 448, 452–53. CARSTENSEN - FINAL 4/10/2008 12:09 PM 2008:295 False Positives 315 barrier to small competitors with limited financial resources, such as NicSand, and effectively foreclosed the market for future entry because of limited opportunity for efficient entry resulting from the long duration of the exclusive-dealing commitments of the major buyers.110 The majority, however, concluded that NicSand had not suffered antitrust injury as a result of 3M’s actions.111 This decision therefore authorizes deep-pocket competitors to engage in comparable practices where there is no explanation for the conduct other than exclusion. This is another example of a false-positive decision that expands the arsenal of dominant firms in their ongoing efforts to exclude smaller competitors. In sum, courts can and do make mistakes, but despite the complex nature of antitrust litigation, there is no reason to think that they have any special proclivity to err on the side of intervention (i.e., false positives) in such cases. In fact, given the inherently conservative nature of even the most progressive judges, there is a strong bias against finding problems even if the evidence might support such a conclusion. 2. PROPOSITION TWO: ANTITRUST RULES ARE EFFECTIVE IN BARRING EFFICIENT CONDUCT The fear of false positives in antitrust requires more than the existence of mistakes. The mistakes must have consequences. One consequence is the cost of litigation itself. If the results are often wrong, then the costs are unnecessary. Antitrust litigation is no more or less a source of concern in this area than is other complex and risky corporate litigation. The risk-averse firm wanting to avoid litigation costs will have an incentive to avoid the conduct that creates the risk of litigation. This deterrent effect is a plausible basis for concern only if the specific conduct foreclosed constitutes the only way to achieve an efficient and desirable business objective. Hence, the analysis must focus on the implications of foreclosing specific conduct options and cannot rely simply on the fact that risk-averse businesses will seek to avoid such options. Thus, the primary cost of error resides in the impact on future conduct. If an actor with monopoly power is denied the opportunity to engage in efficient production or distribution, then overall consumer welfare will suffer. This harm necessarily assumes that the prohibited conduct is the only practical way to achieve a desirable goal. If the same results can be achieved in other lawful ways, then there is no 110. Id. at 451; id. at 461–62 (Martin, J., dissenting). 111. Id. at 459. CARSTENSEN - FINAL 4/10/2008 12:09 PM 316 WISCONSIN LAW REVIEW long-run harm to the economic order. The central empirical question then becomes whether there is only one way to achieve economic efficiency. Two lines of analysis deserve consideration. First, it is useful to ask whether, in general, there is any basis to believe that there are specific methods of doing business that are inherently more efficient than all other options. The focus is not on basic units of production but rather on the ways in which a dominant firm controls access to inputs or outputs. Hence, the issue is often distribution tactics—the use of exclusive dealing, bundling, pricing, and other specific aspects of marketing or input purchase that have the effect of foreclosing competitors. In such contexts, the question for evaluation has to be narrowly framed to focus on the connection of the specific exclusionary conduct to the efficiency-enhancing objective. How necessary is that conduct to the underlying goal? In a Coasian world, it would be obvious that there would be a wide range of ways that either contract or ownership integration could achieve some specific goal. There would be little or no need to introduce any exclusionary feature except those that inhere in the primary transaction. In a real world of transaction costs and opportunistic behavior, there may be legitimate concerns about loyalty and responsiveness to the underlying efficiency-enhancing goal. But this does not justify the use of exclusionary conduct. Such conduct seeks to address a problem akin to an externality—an incentive to act contrary to the agreement. Once the efficiency-enhancing activity is identified, the risks of opportunistic behavior can be identified. At this point, the monopolist, like any other buyer or seller, can reframe the transaction to reduce or eliminate the risk of opportunistic conduct.112 Basically, exclusionary restraints are one tool, and usually a weak one, when major efficiency gains are possible but vulnerable to strategic misconduct. Indeed, whenever the gain is significant but vulnerable under existing relationships, it is very likely that the most efficient solution (i.e., lowest policing costs) is to reframe the transaction or activity to 112. An illustration comes from United States v. Sealy, Inc., 388 U.S. 350 (1967). In Sealy, in order to ensure the success of a joint venture, producing a differentiated brand of mattress advertising was important. The risk of free riding in this scenario is obvious—let another member support advertising and then take a free ride on that effort by selling to that member’s customers. Hence, Sealy could have used a territorial limitation and a contractual requirement for advertising. In fact, Sealy imposed a per-mattress charge on all its participants that went to the central headquarters for use to reimburse actual advertising expenses, whether incurred by a participating manufacturer or by a retailer. This strategy eliminated the risk of free riding on advertising and avoided the potentially intractable problems of policing each member’s direct- and indirect-advertising efforts. Mueller, supra note 86, at 1265, 1280, 1300–01; see also Carstensen, supra note 83, at 607–08. CARSTENSEN - FINAL 4/10/2008 12:09 PM 2008:295 False Positives 317 internalize the benefits to the parties in ways that do not facilitate strategic conduct. As a generalization, it should be possible for well-counseled monopolists to determine where they risk opportunistic behavior and find ways to internalize that risk by reconceptualizing the transaction. Unfortunately, this capacity also can be used to secure monopoly profits that might otherwise elude the firm. Hence, the competitive analysis should focus on why the activity was reordered. Was it to avoid opportunistic conduct that might interfere with legitimate business objectives (e.g., efficient production or distribution) or was it to capture more of the consumer surplus that might otherwise be lost to the monopolist because the existing system of production or distribution did not capture the monopoly profits as fully as the alternative? Thus, the general conclusion is that there are likely to be many ways to accomplish an efficiency-enhancing improvement in the production or distribution of goods and services. Given that monopoly is socially inefficient and undesirable but can be durable if allowed to protect itself from entry and competition, any exclusionary conduct should be subject to strict scrutiny to determine whether the exclusionary effects were avoidable. Second, the cases themselves identify the kinds of conduct that are of concern in terms of competitive effect, and the efficiency-enhancing qualities of that conduct can then be evaluated. In looking at the run of cases involving exclusionary, monopolistic behavior, it is hard to see that any desirable goals were obstructed at all. In Aspen, for example, there was no reason why Aspen Ski Co. could not have sold day tickets.113 If it had done so, consumers would have been free to make choices on the basis of their preferences, and Aspen Ski Co. would not have had to collaborate with Aspen Highlands further.114 Similarly, 3M could have discounted specific products in its line rather than bundling them into a package.115 Nothing in U.S. Tobacco’s behavior toward Conwood seems important to efficient distribution of chewing tobacco.116 Kodak remained free to set any price it wanted for its repair parts, both patented and unpatented, so long as it made them available to all buyers.117 Finally, Dentsply’s “justification [for its restraints on its dealers’ sale of competing products] was pretextual and did not 113. Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585, 608– 10 (1985). 114. Id. 115. LePage’s Inc. v. 3M, 324 F.3d 141, 163–64 (3d Cir. 2003) (en banc). 116. Conwood Co. v. U.S. Tobacco Co., 290 F.3d 768, 788 (6th Cir. 2002). 117. Image Technical Servs., Inc. v. Eastman Kodak Co., 125 F.3d 1195, 1224–28 (9th Cir. 1997). CARSTENSEN - FINAL 4/10/2008 12:09 PM 318 WISCONSIN LAW REVIEW excuse its exclusionary practices.”118 In short, the conduct interdicted in the cases has little to do with general efficient behavior. Moreover, the commands leave a great deal of discretion to the firms to find other ways to compete in the market. Competent counsel can read such decisions and render good advice to clients with substantial market positions regarding the kinds of practices that will be likely to raise serious antitrust concerns. Indeed, counsel do this all the time. Given the motives of dominant firms to exclude and entrench their position in the market, the burden on counsel is to balance such urges with the legitimate needs of production and distribution. 3. PROPOSITION THREE: EXISTING MARKET STRUCTURES ARE NATURAL AND EFFICIENT Assuming a concern for the harms resulting from monopoly and a belief that conduct-oriented interventions carry serious efficiency costs might lead to a policy that monopoly itself should be eliminated. Indeed, monopoly law originally took exactly that position. In the Standard Oil case, the Supreme Court was explicit: [T]he duty to enforce the statute requires the application of broader and more controlling remedies. . . . [These remedies will] neutralize the extension and continually operating force which the possession of the power unlawfully obtained has brought and will continue to bring about.119 The folklore is that structural remedies for monopoly have had undesirable results. While some corporate dissolutions were not optimal, it is hard to claim that the oil, cans, aluminum, copier, or telecommunications industries were made worse off as a result of remedies that had the effect of dissipating market power and encouraging the emergence of competition. Not all monopoly-dissipating remedies required corporate dissolution to achieve that goal. In copiers, Xerox divested control over the patents controlling the competing technology, which encouraged large-scale entry and rapid transformation of the industry.120 Similarly, in the case of tin cans, the successful remedy involved the divestiture of 118. United States v. Dentsply Int’l, Inc., 399 F.3d 181, 197 (3d Cir. 2005). 119. Standard Oil Co. of N.J. v. United States, 221 U.S. 1, 77–78 (1911). 120. Tom, supra note 63, at 967. CARSTENSEN - FINAL 4/10/2008 12:09 PM 2008:295 False Positives 319 patented can-closing machines that opened the market for cans to all competitors.121 Corporate reorganizations have also proven successful in some very important cases. In aluminum, the primary remedy was to separate ownership of two of the major North American aluminum producers, Alcoa and Aluminum Limited of Canada.122 The result was a major increase in competition in that market. The long-run impact of the dissolution of Standard Oil was to stimulate the growth of competition in that industry. Similarly, the dissolution of AT&T has contributed to the rapid growth of technological innovation in telecommunications—a direct result of competition. In these cases, the monopoly enterprise was so constituted that a major corporate reorganization was feasible.123 In general, in an economy on the scale of the United States, it is very rarely the case that productive or distribution economies would warrant a single dominant firm as a matter of efficiency. Most such firms have acquired that position through merger and fortuitous market events. Restructuring such firms does no violence to underlying efficiency. On the other hand, Microsoft posed a real problem with respect to identifying a way to dissipate monopoly power.124 One of the many curious tensions within the legal world is the belief that existing market structures are both natural and efficient and the contrary belief that corporate acquisitions and takeovers are desirable because they will move assets to a higher and better use. But if corporate takeovers or buyouts are economically desirable, then no given ownership pattern is inherently optimal. Indeed, the real business world features a great deal of churning of assets. Firms acquire and then divest assets. Wall Street investment bankers are equally capable of both buying and selling enterprises. This behavior shows that the business community is not discomforted by corporate death, marriage, or divorce. Given the flexibility of the market to structure and restructure corporate organizations, the law should not regard dissolution of enterprises as the analog of the death penalty. Rather, it 121. James W. McKie, The Decline of Monopoly in the Metal Container Industry, 45 AM. ECON. REV. 499 (1955). 122. United States v. Aluminum Co. of Am., 91 F. Supp. 333, 398, 418–19 (S.D.N.Y. 1950). 123. With respect to the Standard Oil case, see Standard Oil, 211 U.S. at 77– 82 (discussing remedy involving divestiture of stock interests); in the case of Aloca, the remedy decision is found at Aluminum Co., 91 F. Supp. 333 (D.N.Y. 1950) (ordering shareholders in the defendant company to dispose of their stock in the defendant or in Aluminum Company of Canada). 124. See Peter C. Carstensen, Remedying the Microsoft Monopoly: Monopoly Law, the Rights of Buyers, and the Enclosure Movement in Intellectual Property, 44 ANTITRUST BULL. 577, 580, 604–06, 608–10 (1999). CARSTENSEN - FINAL 4/10/2008 12:09 PM 320 WISCONSIN LAW REVIEW is a tool of the business world and can be used to further public policy just as it is used to further corporate ends. A few industries, such as gas or oil pipelines, electric transmission and distribution, and, perhaps, cable television and land-line-based telephone service may have inherently monopolistic structures. Similarly, network systems may create a strong tendency toward monopoly. The incentive to exploit this monopoly power arises from the ownership of such monopolies. That is, a single owner has an incentive to capture the potential rents from the monopoly position. The lesson from essential-facilities situations is that other forms of ownership can dissipate or eliminate the incentive to exploit that monopoly.125 Specifically, if the ownership of the potential monopoly bottleneck is dispersed among stakeholders who need to use the monopoly to get inputs or reach consumers or rely on a network platform to provide a market for individual products, then the collective interest is in eliminating the bottleneck by expanding to serve all of the demand that can pay the competitive price. Illustrations of this principle appear in the United States v. Terminal R.R. Ass’n,126 the response of farmers to local grain-elevator monopolies,127 and, most recently, in a legislatively mandated restructuring of electric-transmission ownership in Wisconsin.128 Thus, while some intractable cases may exist, in general dissipation of monopoly power by some structural remedy is a very feasible option. It demands careful analysis and may result in disruption of the market during transition periods. But if there were a serious concern for the risks of false positives with respect to conduct-oriented relief in monopoly cases, then, given that monopoly itself is undesirable, it would seem that remedies that eliminated the monopoly 125. An essential facility is one to which competitors require access in order to compete in other, related markets. See generally Spencer Weber Waller, Areeda, Epithets, and Essential Facilities, 2008 WIS. L. REV. 359. ____ 126. 224 U.S. 383, 397–98, 411 (1911) (forcing railroad-terminal association to allow all railroads to become shareholders and thus eliminating the incentives to discriminate or exploit). 127. See Carstensen, supra note 98, at 29 (describing how cooperative elevators were larger than those in private ownership because the owners, farmers seeking to market their grain, had no incentive to try to exploit themselves by limiting capacity). 128. The Wisconsin legislature required all transmission assets to be sold to a single operating company and allowed downstream buyers of electricity to invest in the company so that all stakeholders were able to participate. See WIS. STAT. § 196.485(1m)(b), (3)(c), (3m)(c)(3) (2005–06). The resulting enterprise has much more incentive to eliminate congestion in the system and ensure open access to all buyers and sellers. CARSTENSEN - FINAL 4/10/2008 12:09 PM 2008:295 False Positives 321 power and left business free to compete as it wished should be preferable to continued monopolistic inefficiency. D. The Danger of False Negatives A false negative is a decision that fails to find a violation where the conduct unreasonably and unnecessarily excludes either new or existing competition. Such a decision operationally is one that finds certain kinds of conduct to be lawful despite unjustified exclusionary effect. The cause for concern is that once the legal system has blessed an unjustified exclusionary practice, this may have great negative impact both in the specific industry and the general economy because such a ruling will empower other dominant firms to adopt the same strategy. Thus the error will replicate itself as the initial decision becomes precedent for rejecting challenges to similar conduct. A false negative is more likely to have significant, durable economic effect than a false positive. The difference in predicted outcome is that in the case of false positives, the actual goal of the conduct is to accomplish something other than an adverse effect on competition; hence the actor has, as discussed earlier, other ways to achieve the same substantively efficient goal. In the case of a false negative, the objective of the conduct at issue is to harm competition; hence approval of such conduct means that the dominant firm and its peers now have an additional weapon in their arsenal to achieve exclusion. This directly increases the probability that unnecessary exclusion will occur, entrenching the dominance of the market leader and prolonging its exploitation of consumers. This is not a claim that in specific contexts the victims of exclusionary conduct upheld via a false-negative decision cannot find a way around that conduct. Clearly, the dynamics of the market, including strategic interests of either customers or suppliers, can make it possible to evade the impact of an exclusionary practice. Rather, the claim here is probabilistic: the more the courts uphold such unjustified, exclusionary practices, the more likely it is that exclusion will be successful in both the short and intermediate run. Moreover, as such conduct becomes insulated from legal challenge, investors and bankers will be less willing to take the risks associated with entry into these markets. E. The Falsification of the False-Positive Mantra To sum up this discussion, four central points can be made. First, monopoly is inherently suspect in a market economy. This is particularly true of monopoly that engages in self-protective, CARSTENSEN - FINAL 4/10/2008 12:09 PM 322 WISCONSIN LAW REVIEW exclusionary conduct. Second, the interest of the monopolist in exclusion and the conservative nature of the judicial process in reviewing such conduct combine to create a more serious risk of false negatives than of false positives. Third, the risk to longer-term economic efficiency resulting from false positives in the context of exclusionary, monopolistic conduct is low because of the existence of multiple ways to achieve important economic efficiencies. But, fourth, false negatives are more likely to result in durable costs to both static and dynamic efficiency because of the incentive structure of monopolists and the inherent conservatism, reflected in stare decisis, of the judiciary. III. FALSE POSITIVES, FALSE NEGATIVES, AND THE FUTURE OF MONOPOLY LAW It is time to tie together the themes that underlie this Paper. First, a survey of the current law of monopolistic exclusion shows that the courts have fashioned from Aspen and Kodak standards for judging such conduct that run few risks of false positives but may also limit the danger of egregious false negatives. The question for the future is the extent to which the Trinko decision, with its renewed concern for maximizing monopoly profits and fear of false positives, will undermine this trend. Second, if monopoly is not itself desirable, but if conduct-oriented remedies have a high risk of false-positive outcomes that also threaten economic efficiency, perhaps it is worth reconsidering the merits of an alternative response to unlawful monopoly: structural relief. A. The State of the Law and Its Prospects Aspen establishes a standard for protecting competition and the potential for competition oriented toward consumer choice and ensuring its protection in the context of monopoly behavior.129 The screen it creates focused on the business justification for the conduct outside the monopoly itself. This test should have led in Paschall to the rejection of the Star Company’s elimination of its distributors since the record showed that the Star Company’s only objective was to increase its own monopoly profits. Similarly, the claims of NicSand should have proceeded because, again, the allegations were that the specific actions sought only to create and entrench a monopoly. Aspen imposes a more critical review of the economic rationality of specific conduct, but its 129. Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585, 605– 10 (1985). CARSTENSEN - FINAL 4/10/2008 12:09 PM 2008:295 False Positives 323 test is respectful of the right of the monopolist to exclude competition.130 Indeed, despite a passing reference to the potential that less restrictive alternatives are relevant,131 the opinion’s factual analysis emphasized “Ski Co.’s failure to offer any efficiency justification whatever for its pattern of conduct.”132 As such, it is comparable to the original standard for judging the legality of a restraint of trade that excluded from the application of section 1 of the Sherman Act any restraint ancillary to some legitimate transaction or joint venture.133 The Supreme Court carried this analysis forward in the Kodak case when it again examined the justifications for exclusionary conduct critically but without any final resolution.134 The Kodak standard is also ambiguous because of the role of patent rights as a possible justification for exclusionary conduct. Patent law expressly confers a right to exclude others from the production or use of the patented good or process. However, if the exclusion advances an economic interest of the patent holder other than the direct exploitation of the patent privilege itself, it is subject to challenge.135 The Ninth Circuit’s solution to the question of balancing these rights is not very satisfactory.136 A better explanation for the decision’s pretextual standard is to consider the role and function of the discriminatory refusal. Where a significant effect is to preclude competition in related but unpatented markets, then the refusal to deal should be outside patent law and subject to antitrust law.137 130. Id. at 600. 131. Id. at 605 (considering as relevant whether the monopolist has “impaired competition in an unnecessarily restrictive way”). 132. Id. at 608. 133. Despite many misreadings of United States v. Trans-Missouri Freight Ass’n, 166 U.S. 290 (1897), and United States v. Joint Traffic Ass’n, 171 U.S. 505 (1898), Justice Peckham was clear in both cases that the absolute reading of section 1 applied only to naked restraints of competition. The decisions in Anderson v. United States, 171 U.S. 604 (1898), and Hopkins v. United States, 171 U.S. 578 (1898), decided the same day as Joint Traffic confirm this. See also Nolen Ezra Clark, Antitrust Comes Full Circle: The Return to the Cartelization Standard, 38 VAND. L. REV. 1125 (1985). 134. Eastman Kodak Co. v. Image Technical Servs., Inc., 504 U.S. 451 (1992). 135. Walker Process Equip., Inc. v. Food Mach. & Chem. Corp., 382 U.S. 172, 175 (1965); Int’l Salt Co. v. United States, 332 U.S. 392, 395–96 (1947); United States v. Microsoft Corp., 253 F.3d 34 (D.C. Cir. 2001) (en banc) (per curiam). 136. See Image Technical Servs., Inc. v. Eastman Kodak Co., 125 F.3d 1195 (9th Cir. 1997). But see In re Indep. Serv. Orgs. Antitrust Litig., 203 F.3d 1322, 1329 (Fed. Cir. 2000) (criticizing the Ninth Circuit’s approach). 137. See Seungwoo Son, Selective Refusals to Sell Patented Goods: The Relationship between Patent Rights and Antitrust Law, 2002 U. ILL. J.L. TECH. & POL’Y 109. CARSTENSEN - FINAL 4/10/2008 12:09 PM 324 WISCONSIN LAW REVIEW Despite the questionable phrasing by the Ninth Circuit of its standard on remand, the thrust of the Supreme Court’s Kodak decision is that exclusionary use of monopoly power should be reviewed critically. Because of the posture of the case, the Court’s critical analysis focused only on the question of whether any of the justifications that Kodak offered were so indisputable that they warranted dismissing the plaintiff’s case on summary judgment. 138 The courts of appeals have drawn on Aspen and Kodak to develop a somewhat-more-critical stance for reviewing exclusionary conduct. The key cases are Microsoft, 3M, Conwood, and Dentsply. Microsoft propounds a set of criteria for judging the lawfulness of harmful conduct by a monopolist that makes clearer and more concrete the Aspen methodology and appears also to impose a somewhat-more- restrictive test for conduct having mixed outcomes.139 The Court adopted a four-step approach to evaluating conduct: First, the plaintiff must establish that the conduct at issue had an adverse effect on competition in general, both as a matter of theory and fact. Second, if the plaintiff has satisfied step one, the defendant must advance a nonmonopolistic business justification for the conduct, but such a justification can include protection of legitimate intellectual-property rights.140 Third, if the defendant satisfies step two, the plaintiff has the burden of proving that the justification is pretextual or that there is a less restrictive means to accomplish the legitimate goal.141 Fourth, even if the defendant prevails in step three, the decision maker is to weigh the social gain from the harmful conduct against its competitive costs and may find it unlawful despite there being no less restrictive option that would be equally effective to accomplish the legitimate business purpose.142 The first two steps seem consistent with Aspen. Step three follows from the initial statement of the rule of the appropriate standard in Aspen although the decision’s factual analysis emphasized the lack of any business justification. Step four imposes stricter standards than Aspen might have suggested. In particular, while rejecting the monopolist’s justification on the grounds that it is pretextual is consistent with Aspen, imposing a less- restrictive-alternative standard results in stricter scrutiny of the conduct of the monopolist than courts have heretofore employed. It is consistent 138. See Kodak, 504 U.S. 451. 139. Compare Microsoft, 253 F.3d at 58–59, with Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585, 610–11 (1985). 140. Microsoft, 253 F.3d at 58–59. 141. Id. at 59. 142. See id. CARSTENSEN - FINAL 4/10/2008 12:09 PM 2008:295 False Positives 325 with the analysis suggested in Part II of this Paper that monopolists have a clear strategic incentive to impose the maximum exclusionary restraint consistent with their legitimate business goal.143 The United States Court of Appeals for the District of Columbia Circuit’s standard would address this tendency but at the cost of more intrusion into business decisions. Finally, step four, with its authorization of balancing harms and benefits without an explicit standard for striking that balance, is the most problematic. Such a step invites the greatest risk of false outcomes—positive or negative. It is worth noting that in applying this framework, the D.C. Circuit never got past step two. It either found that there was no justification for the exclusionary conduct or that the justification was not disputed on its merits. Hence, it remains to be seen whether and how the final two steps of the framework will apply in practice. Conwood was, given the facts found by the trial court, an easy case of unlawful, predatory, and exclusionary conduct.144 This conduct included destroying Conwood’s racks and sales materials, misleading retailers about relative sales, and using exclusive contracts to reduce the availability of Conwood’s product.145 Given the facts, the number and vigorousness of objections to the decision signal the continued concern for false positives in the bar and the academic community. But on the basis of the record in the case as described by the court of appeals, any other outcome would have been a false negative. LePage’s may appear to be a more difficult case because it involved the use of bundled sales with discounts.146 Such practices are part of the general commercial-business world. Hence, the ruling has a broader implication than that of Conwood. The critics can say that the United States Court of Appeals for the Third Circuit’s standard for deciding whether bundling was unlawful did not provide much guidance to firms with large market shares in some specific product line. However, this exaggerates the potential uncertainty. Sellers are on notice that when a discounted bundle combines a set of product lines for which no competitor can offer a competing bundle, there is going to be an exclusionary effect. The seller can then identify its business reason for needing to make such bundled sales and ask whether it needs to use that approach to accomplish its legitimate, nonexclusionary goal. Consultation with counsel when initiating such a bundled sale strategy will allow risk-averse businesses to identify their legitimate reasons for 143. See supra Part II. 144. Conwood Co. v. U.S. Tobacco Co., 290 F.3d 768, 783 (6th Cir. 2002). 145. Id. 146. LePage’s Inc. v. 3M, 324 F.3d 141, 154–59 (3d Cir. 2003) (en banc). CARSTENSEN - FINAL 4/10/2008 12:09 PM 326 WISCONSIN LAW REVIEW the strategy and consider less harmful alternatives. In short, the LePage’s standard, as amorphous as it may be, creates little downside risk to consumer welfare or longer-term efficiency. Indeed, the Third Circuit applied its LePage’s analysis to reverse the trial court’s dismissal of the government case against Dentsply.147 Dentsply, the monopolist of artificial teeth, threatened to cut off its dealers, the primary source of supply for the dental laboratories, if they carried any teeth made by its competitors.148 Even the trial court had found that the exclusionary conditions lacked any business justification.149 The opinion on appeal concurred and reached the obvious conclusion that such conduct should be unlawful monopolization.150 Taken as a group, these four decisions reinforce the Aspen approach and reflect a more balanced concern for mistakes in antitrust law. To be sure, Conwood, LePage’s, and Microsoft were all decided prior to Trinko, but the Supreme Court refused to review LePage’s, one of the most controversial of these cases, when certiorari was sought after the Trinko decision.151 Moreover, the Third Circuit upheld the government’s challenge to Dentsply’s exclusionary conduct after Trinko, without even citing that decision. Nevertheless, the Trinko decision renews the concern for false positives and the disruptive effects of intervention in the decisions of monopolists as to how they will treat competitors.152 Under Aspen and Kodak, Verizon could only defend its conduct with proof that its failure to serve its customer/competitor had a business justification other than exclusion.153 Given the public record with respect to these actions, such a defense was implausible.154 The conduct violated both contractual rights of its customer and the statutory obligations imposed on the company. In upholding the lawfulness of Verizon’s conduct as a matter of antitrust law, the Court focused on the risks of cabining a monopolist within the mildly restrictive rules of Aspen and Kodak and concluded that such limits would result in serious risk of harm to economic efficiency, including loss of innovation. This view assumes the validity 147. United States v. Dentsply Int’l, Inc., 399 F.3d 181, 186–87, 197 (3d Cir. 2005). 148. Id. at 185. 149. Id. at 185. 150. Id. at 196–97. 151. 3M Co. v. LePage’s, Inc., 542 U.S. 953 (2004) (denying petition for writ of certiorari). 152. Verizon Commc’ns Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398, 407–08 (2004). 153. See supra notes 25–35 and accompanying text. 154. Trinko, 540 U.S. at 402–05. CARSTENSEN - FINAL 4/10/2008 12:09 PM 2008:295 False Positives 327 of the proposition that false positives are inherently very harmful and should be avoided at all costs. Worse, the Trinko opinion seems to accept on its merits the position that antitrust law should protect the monopolist’s ability to exploit its market power. Yet Verizon’s monopoly position arose from the dissolution of another unlawful monopolist. Like the Kansas City Star Company,155 Verizon has a corporate history tainted with unlawful monopoly and a congressional determination that competition should be brought to local-phone-service markets.156 If the Court’s decision is taken literally, then the Court has abandoned Aspen’s constraints on the use of monopoly power to exclude or destroy competition. The Court’s subsequent refusal to review the LePage’s decision, however, argues against such a broad reading. Trinko can be distinguished from Aspen and the cases following it because of the unique context of an industry in transition from regulated monopoly to one where Congress sought to create workable competition.157 That is not, however, the flavor of the text of the opinion. The reiterated concern for the efficiency costs of false negatives, the implicit belief that courts are maladapted to decide the merits of such claims, and Schumpeterian enthusiasm for the right to exploit monopoly as an instrument of dynamic economics resonate through the opinion without objection from the dissent. The future of monopoly law is much in doubt even as it becomes a more significant issue in a world increasingly inhibited by dominant firms and various forms of intellectual-property rights having monopolistic character. Despite the paucity of actual false positives in the real world, the myth of the highly inefficient false positive remains a strong force in legal discourse. Moreover, as the uncritical version of the Schumpeterian vision of monopoly as a good thing has emerged as the basis for defending all monopolies, the concern for protecting the right to full monopoly profits will strengthen the opposition to stricter rules against exclusionary conduct even where the only defense is that it increases monopoly profits. 155. The Kansas City Star Company was convicted in the 1950s of criminal violations of section 2. Kansas City Star Co. v. United States, 240 F.2d 643 (8th Cir. 1957). 156. Trinko, 540 U.S. at 402–05. 157. The decision can read as an expansive version of the primary-jurisdiction concept. See, e.g., Ricci v. Chicago Mercantile Exch., 409 U.S. 289 (1973). The concept called for allowing regulators having concurrent authority to deal with industry- specific conduct before a court examined the conduct under the antitrust laws. Id. at 304–05. Trinko would expand this approach to impose a bar on damages for past harm when the administrative agency has acted to eliminate the competitive concerns by forward-looking enforcement. CARSTENSEN - FINAL 4/10/2008 12:09 PM 328 WISCONSIN LAW REVIEW B. What about a Structural Response to Monopoly? If the kinds of conduct-oriented remedies now popular in antitrust were likely to stultify business and rigidify markets, courts might consider alternatives that would respond to the problems of durable monopoly without imposing restrictive rules on firms. Indeed, the Standard Oil decision embraced exactly such a strategy for dealing with monopoly. It did so because the risks to the dynamics of the market from excessive judicial regulation outweighed the potential costs involved in corporate reorganization. It may be time to revisit that Standard Oil approach. The basic argument is that to avoid regulatory decrees, the market needs to be restored to a workably competitive structure. In such a world, strategic, exclusionary conduct is much less likely because it is unlikely to yield significant gains for any firm. Thus, individual firms would not have to be concerned about the legality of ordinary competition or its putative exclusionary effect. By returning to structural remedies and limiting relief in monopoly cases brought by victims of exclusionary conduct to structural reforms of the market, the incentives to seek the kinds of inefficient conduct- oriented relief that are most likely to produce false-positive results would be eliminated. Firms would have to decide if they could survive in a more competitive environment. The infamous Telex 158 and Berkey 159 cases are examples of the pursuit of self-serving gain by firms that would be unlikely to exist in a competitive market. Both cases in fact produced significant evidence of exclusionary conduct as well as monopoly power. But the plaintiffs in different ways both sought to have the courts provide them with protection from the rigors of competitive markets. This is a misuse of antitrust law and explains why, ultimately, both cases entirely or largely failed. But prior to reaching those results, the legal system laid the groundwork for false- positive analyses of exclusionary conduct. It would unduly extend this Paper to engage in a detailed effort to articulate the standards for structural monopoly based on the strength, durability, and remediability of the monopoly situation. Suffice it here to suggest that if there were a legitimate concern that conduct-oriented decrees impose unreasonable risks to economic efficiency and desirable market dynamics, then because monopoly itself is bad, it may be a good idea to refocus monopoly law on its traditional mission of eliminating monopoly and restoring workably competitive markets. 158. Telex Corp. v. Int’l Bus. Machs. Corp., 510 F.2d 894 (10th Cir. 1975). 159. Berkey Photo, Inc. v. Eastman Kodak Co., 603 F.2d 263 (2d Cir. 1979). CARSTENSEN - FINAL 4/10/2008 12:09 PM 2008:295 False Positives 329 CONCLUSION The present state of antitrust law is remarkable. Despite the teaching of economic texts that monopoly imposes substantial efficiency costs, a statute based on that insight (i.e., the Sherman Act), and a legal history that shows the correctness of interventions that eliminate monopoly and restore competition, the current case law seems to favor the monopolist. Concern that courts might require a monopolist to find a less harmful, but perhaps slightly more costly, way to accomplish a legitimate objective is driving the law to embrace the vision that monopolists should be free to seek the full monopoly profit regardless of the manner in which they accomplish that goal. Moreover, if the expansive reading of Trinko is correct, monopolists can defend their conduct by expressly stating that they sought to increase monopoly profits at the expense of actual or potential competitors. This would reverse the more measured and balanced standards derived from Aspen and would reflect a serious failure to appreciate the relevance and implications of the economic models being deployed in these cases.