Section 2 Standards and Consumer Welfare: Some Lessons from the World of Merger Enforcement SECTION 2 STANDARDS AND CONSUMER WELFARE: SOME LESSONS FROM THE WORLD OF MERGER ENFORCEMENT David L. Meyer* I. Introduction ................................................................ 371 II. The Proposed "Consumer Welfare" Standard ........... 378 A. Lesson No. 1: In Practice, Neither the Law Nor Its Enforcement Functions as a True "Consumer Welfare Prescription" ....................... 382 B. Lesson No. 2: Speed, Certainty and Finality Are H ighly Valued ............................................... 385 C. Lesson No. 3: A Full-Blown Net Welfare Analysis is Costly and Time Consuming ............ 394 D. Lesson No. 4: Outcomes of a Net Welfare Analysis are Hard to Predict ............................... 396 E. Lesson No. 5: Enforcers and Courts are not Good at Micro-Managing Firm Behavior in Furtherance of Consumer Welfare (or Any O ther G oal) .......................................................... 407 III. C onclusion .................................................................. 408 I. INTRODUCTION This article will discuss the role of "consumer welfare" in judging the legality of single-firm conduct.' There is a * Deputy Assistant Attorney General for Civil Enforcement, Antitrust Division, U.S. Department of Justice. The views and opinions expressed herein are those of the author and do not necessarily represent the official position or policies of the U.S. Department of Justice. ' Discussion of "consumer welfare" in this article should not be read as suggesting a position in the debate over the distinction between, and potentially divergent enforcement implications of, "consumer surplus" and COLUMBIA BUSINESS LAW REVIEW debate raging in the antitrust world over the standards for judging whether the business decisions of individual firms with a significant degree of market power will run afoul of competition law principles-embodied in the United States in Section 2 of the Sherman Act 2-and thus potentially give rise to claims for damages by the firm's customers and rivals. A few core principles are now well established in the United States. First, unilateral conduct has long been immune from antitrust scrutiny when the firm is not yet a monopolist and the firm's small share of the market or other market conditions make it unlikely that monopoly power could be achieved.' Second, "bigness" (or dominance) is no longer equated with "badness." The concern of antitrust law is instead with the maintenance or acquisition of monopoly power rather than its mere possession or exercise. Firms with monopoly power are thus free to exercise that power by charging high prices and collecting supra-competitive returns.' Those returns are the rewards for success that has been achieved lawfully, often by developing products or services that meet an important and previously untapped "total surplus." See Ken Heyer, Welfare Standards and Merger Analysis: Why Not the Best?, 2 COMPETITION POL'Y INT'L 29 (2006) (advocating use of the total welfare standard); Dennis W. Carlton, Does Antitrust Need to be Modernized?, at 1-3, EAG Discussion Paper EAG 07-3 (January 2007) (same), copies available from U.S. Department of Justice Antitrust Division Economic Analysis Group, Suite 10,000, 600 E Street, N.W., Washington, DC 20530. Rather, this article takes as a given that Section 2 properly should address economic "welfare," however measured, and then proceeds to examine whether that aim is best achieved by a standard that hinges liability on an explicit inquiry whether particular conduct by a dominant firm will on balance expand or contract such welfare. 2 15 U.S.C. § 2 (2006) (prohibiting "monopoliz[ing], or attempt[ing] to monopolize, or combin[ing] or conspir[ing] with any other person or persons, to monopolize any part of the trade or commerce among the several States, or with foreign nations"). ' See, e.g., Spectrum Sports, Inc. v. McQuillan, 506 U.S. 447, 459 (1993) (holding that there can be no liability under § 2 absent proof of a "dangerous" probability that the defendant "would monopolize a particular market"). ' See Verizon Commc'ns Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398, 407 (2004). [Vol. 2007 No. 2:371] SECTION 2 STANDARDS AND CONSUMER WELFARE 373 consumer need. Moreover, high prices are now generally recognized as "procompetitive" acts, at least in the sense that they stimulate entry and investment and other competitive responses.5 These first two principles are quite important. They prevent the law from penalizing unilateral conduct that clearly will not harm the competitive process. They also reflect a sound recognition that we should not penalize firms whose competitive striving leads to extraordinary success in the marketplace and, even if we might desire to regulate the behavior of monopoly firms, courts (and antitrust enforcers) lack competence to regulate monopoly (or any other) pricing.' The boundaries charted by these principles provide important certainty to firms going about their business. Firms that lack any serious prospect of achieving monopoly power, or monopolists that merely wish to reap the economic rewards of their success by charging high prices, do not need to turn the reins of management over to antitrust lawyers or industrial organization economists to help them avoid the risks of lawsuits brought by rivals. But these principles leave unresolved the harder question of what a firm with monopoly power (or, perhaps more importantly, an innovator who wishes to become a monopolist) can do, without fear of antitrust liability, other than charge high prices. On this front, the law is far less clear. We do know that Section 2 does not prohibit all conduct that allows a firm to acquire monopoly power or that entrenches an existing monopolist's position. In the famous but not very useful words of Grinnell, Section 2 outlaws only the "willful" acquisition or maintenance of monopoly power, as distinguished from "growth or development as a consequence of a superior product, business acumen, or ' See id. (noting that "[tihe opportunity to charge monopoly prices... is what attracts 'business acumen' in the first place [and] induces risk taking that produces innovation and economic growth"). 6 See Trinko, 540 U.S. at 408 (requiring courts "to act as central planners, identifying the proper price, quantity, and other terms of dealing [is] a role for which they are ill suited"). historic accident."7 Other formulations are not more helpful, suggesting, for example, that firms may operate safely only if they engage in "competition on the merits."' But "competition on the merits," like most other phrases used in the Section 2 context, is an illusive concept. In modern debate, this requirement is often couched in terms of distinguishing permissible conduct from that which might be condemned as "predatory" or "exclusionary."' Making these distinctions has been a vexing problem for decades because of the legendary difficulty of separating "good" aggressive competition, which is often the hallmark of our rough-and-tumble competitive process, from "bad" aggressive competition that can sometimes stifle competitive forces. In the recent argument in the Weyerhaeuser case, Justice Breyer aptly captured the conundrum when he observed that courts must "decide whether people are hogging goods unnecessarily for bad purposes, or rather storing up nuts for winter for good purposes."10 The same problem arises again and again in other settings." The United States v. Grinnell Corp., 384 U.S. 563, 570-71 (1966). 8 See Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585, 605 n.32 (1985) (noting that, in order to be subject to challenge under § 2, conduct must "not only (1) tend(] to impair the opportunities of rivals, but also (2) either . . .not further competition on the merits or do[] so in an unnecessarily restrictive way."). See, e.g., id. at 602. 10 Transcript of oral argument at 31, Weyerhaeuser Co. v. Ross- Simmons Hardwood Lumber Co., No. 05-381 (U.S., Nov. 28, 2006). This concern was echoed in the Court's subsequent decision, which noted that "actions taken in a predatory bidding scheme are often "the very essence of competition." Weyerhaeuser Co. v. Ross-Simmons Hardwood Lumber Co., 127 S. Ct. 1069, 1077 (2007) (citations and internal quotation marks omitted). The Court held that the test for predatory pricing under Brooke Group Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209 (1993), also applies to claims of predatory over-bidding. See Weyerhaeuser, 127 S. Ct. at 1072. " See, e.g., Brooke Group, 509 U.S. at 224 ("Even if the ultimate effect of the cut is to induce or reestablish supracompetitive pricing, discouraging a price cut and forcing firms to maintain supracompetitive prices, thus depriving consumers of the benefits of lower prices in the interim, does not constitute sound antitrust policy."); Trinko, 540 U.S. at COL UMBIA B USINESS LA W RE VIE W [Vol. 2007 No. 2:371] SECTION 2 STANDARDS AND CONSUMER WELFARE 375 policies underlying the antitrust laws are well served when monopolists, no less than other firms, work hard to maintain or enhance the attractiveness of their products and services to consumers. The law should encourage them to achieve success in ways that benefit consumers, such as by adding new features or entirely new products, lowering the prices they charge for their products, expanding the markets for those products, and identifying and implementing better or more efficient ways of making those products or distributing them to consumers. It is desirable to encourage these efforts even if success for these monopolists might mean that rivals will suffer or may not emerge at all. At the same time, as economists often remind us, actions that monopolists take to achieve or sustain their success can harm competition, directly or indirectly-by driving efficient rivals out of the market, raising entry barriers and entrenching monopoly power, and removing short- or long-run impediments to higher prices, diminished quality or innovation, and reduced output. Therefore, ideal Section 2 liability rules would accomplish at least three things simultaneously. They should properly identify and prohibit "bad" actions that destroy or impede competition without sufficient justification or sufficient benefit to consumers. 2 At the same time, those rules should not penalize "good" actions that reflect innovative or efficiency-enhancing behavior by the monopolist. 3 Perhaps most importantly, they must provide reasonably clear guidance to firms and their antitrust counselors as to the scope of permissible competitive freedoms and the range of potential legal risks. Otherwise, uncertainty about the 414 ("applying the requirements of § 2 'can be difficult' because 'the means of illicit exclusion, like the means of legitimate competition, are myriad,"' and because "mistaken inferences and the resulting false condemnations 'are especially costly, because they chill the very conduct the antitrust laws are designed to protect') (quoting United States v. Microsoft Corp., 253 F.3d 34, 58 (C.A.D.C. 2001) (en banc) (per curiam) and Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 594 (1986)). 12 In the vernacular of economists, they should avoid "false negatives," so-called Type I errors. 13 They should thus avoid "false positives" or Type II errors. COLUMBIA BUSINESS LA W REVIEW application of the rules-including both the risk that courts might reach the wrong results and the high cost of having to defend misguided lawsuits brought by disgruntled rivals- would slacken the drive of all firms that have or might hope to achieve significant market shares. The offense of monopolization might usefully be thought of as a form of "tort," the body of law that is antitrust's ancestral heritage. 4 In the field of tort law, legal scholars have long recognized that the aim is to minimize both the costs of injuries and the costs of avoiding those injuries. 5 In Section 2, there is the added twist that "injury" to competitors is part and parcel of a robust marketplace. We do not want firms-even those with monopoly power- holding back out of fear that their marketplace conduct might accidentally cause harm to an "eggshell rival."16 The significant costs of uncertainty should not be underestimated. Standards that are hard to comprehend or apply can have consequences that reach far beyond the cases that are decided, correctly or incorrectly, using those standards. The Third Circuit's en banc decision in LePage's Inc. v. 3M" is perhaps a poster child in this regard. It has been widely criticized as offering no workable Section 2 standard, for placing undue reliance on harm to rivals stemming from 3M's rebate programs, and for discounting the consumer benefits associated with millions of dollars that 3M rebated to its customers. It seems certain that 14 See PHILLIP AREEDA & LouIs KAPLOW, ANTITRUST ANALYSIS: PROBLEMS, TEXT, CASES 131 (4th ed. 1988). 15 GUIDO CALABRESI, THE COSTS OF ACCIDENTS: A LEGAL AND ECONOMIC ANALYSIS 26 (1970). 6 Cf Brackett v. Peters, 11 F.3d 78, 81 (7th Cir. 1993) (noting "eggshell skull" rule in tort law that "the tortfeasor takes his victim as he finds him"); Vosburg v. Putney, 50 N.W. 403, 404 (Wis. 1891) (noting that "wrongdoer is liable for all injuries resulting directly from the wrongftl act, whether they could or could not have been foreseen by him"). 17 LePage's Inc. v. 3M, 324 F.3d 141 (3d Cir. 2003), cert. denied, 542 U.S. 953 (2004). 18 See, e.g., Ronald W. Davis, Five Ingredients in Search of a Monopoly Broth, THE ANTITRUST SOURCE at 1 (November 2004) (LePage's "creates practical problems" and "has much potential for mischief'); Timothy J. [Vol. 2007 No. 2:371] SECTION 2 STANDARDS AND CONSUMER WELFARE 377 many firms have acted less vigorously in the marketplace because, with LePage's and other cases like it writing the rules of Section 2, antitrust counselors have been reluctant to offer sufficient assurances that a proposed course of action will not lead to costly, distracting, and hard-to-dismiss litigation.19 A possible silver lining is the debate that LePage's (and other developments) has fueled over the proper legal standard. That debate has provided research and writing topics for countless economists and lawyers. It has been addressed by the Antitrust Modernization Commission.2 1 It Muris, Comments on Antitrust Law, Economics, and Bundled Discounts at 1 (comments submitted on behalf of the United States Telecom Association to the Antitrust Modernization Commission, July 15, 2005) (LePage's is "based on a poorly articulated theory and an incomplete record [and] could deter procompetitive behavior"), http://www.amc.gov/commission- hearings/pdf/Muris.pdf; see also David L. Meyer, LePage's II: The En Banc Third Circuit Revisits 3M's Bundled Discounts and Sees Unlawful "Exclusion" Instead of Above-Cost Pricing, THE ANTITRUST SOURCE at 8 (July 2003) ("Less clear is what pricing actions other than simple volume discounts will run afoul of LePage's."). 19 See, e.g., David Balto, LePage's v. 3M, NATL. L. J. 25 (Aug. 11, 2003) ("Firms typically act with great caution because antitrust litigation is immensely costly and protracted and the threat of treble damages liability can be daunting. LePage's is likely to increase the frequency of firms 'holding their competitive punches' and not competing aggressively- particularly on price."); James A. Keyte, LePage's v. 3M-More Questions Than Answers for the Lawful "Monopolist," 17 ANTITRUST MAG. 25 (Summer 2003) (arguing that the decision is "woefully vague as a guidepost for the everyday business behavior of monopolists"). 20 See, e.g., Antitrust Modernization Commission, Report and Recommendations at 81-83, 91-94 (April 2007) (AMC Report) (discussing various proposed tests for defining conduct that may be challenged under Section 2), available at http://www.amc.gov/report-recommendation/ chapterl.pdf; Antitrust Modernization Commission, Public Hearing on Exclusionary Conduct: Refusals to Deal and Bundling and Loyalty Discounts (Sep. 29, 2005), http://www.amc.gov/commission-hearings pdf/050929_ExclusConduct_Transcriptreform.pdf. The AMC Report does not propose that any one test be adopted for all Section 2 purposes, noting that "[tihus far, no consensus exists that any one test can suffice to assess all types of conduct that may be challenged under Section 2." AMC Report, supra, at 94. Nevertheless, the Commission does endorse the general principles that "appropriate legal rules should identify is the topic of an ongoing series of hearings on single-firm conduct sponsored by the Department of Justice's Antitrust Division and Federal Trade Commission (FTC).2 It is also mirrored in a broader, transatlantic (and ultimately global) dialogue over appropriate single-firm competition law rules. We can hope that this debate will eventually yield an approach to the law of monopolization that effectively protects consumer welfare by neither over-restricting (or over-deterring) pro-competitive behavior, nor being overly permissive of harmful conduct. II. THE PROPOSED "CONSUMER WELFARE" STANDARD One intriguing approach that has been proposed- perhaps developed in greatest detail by Professor Salop-is to test the legality of the unilateral behavior of firms with monopoly power by attempting to measure explicitly the unreasonably exclusionary conduct, without discouraging aggressive competition that benefits consumers or creating excessive litigation and compliance costs for businesses and problems of administrability for courts," and that "vigorous competition, the aggressive pursuit of business objectives, and the realization of efficiencies not available to competitors are generally not improper, even for a 'dominant' firm and even where competitors might be disadvantaged." Id. at 81-83. 21 See Press Release, U.S. Dep't of Justice, DOJ AND FTC to Hold Joint Public Hearings on Competition Policy Related to Single-Firm Conduct (Nov. 28, 2005), available at http://www.usdoj.gov/atr/public/ press-releases/2005/213369.htm. Additional information about these hearings can be found at http://www.usdoj.gov/atr/public/hearings/single_ firmlsfchearing.htm. 22 See, e.g., DG Competition discussion paper on the application of Article 82 of the Treaty to exclusionary abuses (issued Dec. 19, 2005), http://ec.europa.eu/comm/competition/antitrust/art82/discpaper2005.pdf; Int'l Competition Network website, http://www.intemationalcompetition network.org/index.php/en/working-groups/unilateral-conduct (noting that the ICN established a working group on Unilateral Conduct in May 2006 to "examine the challenges involved in addressing anticompetitive unilateral conduct of dominant firms and firms with market power, and to promote greater convergence and sound enforcement of laws governing unilateral conduct"). COL UMBIA B USINESS LA W RE VIE W [Vol. 2007 No. 2:371] SECTION 2 STANDARDS AND CONSUMER WELFARE 379 effect of such conduct on "net consumer welfare."23 This test is offered as a rival to the "profit sacrifice" and "no economic sense" tests, both of which focus on the objective economic consequences of the challenged conduct for the current or aspiring monopolist. 4 The "short-run profit sacrifice" test finds conduct to be exclusionary if the behavior represents a "temporary sacrifice of net revenues in the expectation of greater future gains." 5 Under the "no economic sense" test, "conduct is not exclusionary or predatory unless it would make no economic sense for the defendant but for the tendency to eliminate or lessen competition." 6 The "net consumer welfare" standard would instead focus the examination "directly on the anticompetitive effect of 2 See Steven C. Salop, Exclusionary Conduct, Effect on Consumers, and the Flawed Profit-Sacrifice Standard, 73 ANTITRUST L.J. 311 (2006); see also Andrew I. Gavil, Exclusionary Distribution Strategies by Dominant Firms: Striking a Better Balance, 72 ANTITRUST L.J. 3, 18-19 (2004) (noting that "cases involving both inefficiency and efficiency must be assessed through some kind of balancing process," and that "if the antitrust laws are to realize their potential as a 'consumer welfare prescription,' evidence of consumer harm should be given great weight in identifying conduct that violates the Sherman Act"). 24 See Salop, supra note 23, at 312-13; see also Gregory J. Werden, Identifying Exclusionary Conduct Under Section 2: The 'No Economic Sense' Test, 73 ANTITRUST L.J. 413, 415-29 (2006) (discussing the "profit sacrifice" and "no economic sense" tests); A. Douglas Melamed, Exclusionary Conduct Under the Antitrust Laws: Balancing, Sacrifice and Refusals to Deal, 20 BERKELEY TECH. L.J. 1247 (2005); A. Douglas Melamed, Exclusive Dealing Agreements and Other Exclusionary Conduct-Are There Unifying Principles?, 73 ANTITRUST L.J. 375 (2006). " Werden, supra note 24, at 422-23 (quoting Phillip Areeda & Donald F. Turner, Predatory Pricing and Related Practices Under Section 2 of the Sherman Act, 88 HARV. L. REV. 697, 698 (1975)). 26 Id. at 413 (quoting Brief for the United States and the Federal Trade Commission as Amici Curiae Supporting Petitioner at 15, Verizon Comme'ns Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398, 407 (2004) (No. 02-682), available at http://www.usdoj.gov/osg/briefs/2002/ 3mer/lami/2002-0682.mer.ami.pdf). A short run profit sacrifice is "neither necessary nor sufficient for conduct to be deemed exclusionary by the no economic sense test," and where the defendant's conduct entails a short- run profit sacrifice, "the no economic sense test further asks why it is rational to make that sacrifice." Id. at 424. exclusionary conduct on price and consumer welfare."27 Professor Salop would ask whether the expected anticompetitive harms from the conduct (in terms of price increases or output reductions) outweigh any expected pro- competitive benefits.28 He points out that his approach bears some resemblance to the rule-of-reason "balancing approach" outlined by the D.C. Circuit in United States v. Microsoft. 29 Under the standard articulated in that case and others that have adopted its reasoning, once a plaintiff establishes that a monopolist's conduct has harmed competition and not just competitors and the defendant has proffered a cognizable pro-competitive justification for the conduct, then the plaintiff "must demonstrate that the anticompetitive harm of the conduct outweighs the procompetitive benefit." 0 Proponents of the net consumer welfare approach contend that its major advantage is that, by asking the ultimate consumer welfare question, it would reduce both types of errors in the application of Section 2: where bad conduct is erroneously permitted (Type I errors), and where good 7 Salop, supra note 23, at 313-14. 28 See id. at 333-34. See id. (discussing United States v. Microsoft, 253 F.3d 34 (D.C. Cir. 2001)). 30 Microsoft, 253 F.3d at 59; see also Abbott Labs. v. Teva Pharm. USA, Inc., 432 F. Supp. 2d 408, 422 (D. Del. 2006) (applying the rule-of- reason balancing test in a case alleging that new product formulations for branded pharmaceutical blocked entry by generic drug competition). The Federal Trade Commission's decision in Rambus applies this standard- and rejects the "profit sacrifice" and "no economic sense" approaches-to allegedly deceptive conduct in the standards setting arena, where "conduct that reduces consumer welfare . . . happens to be inexpensive to execute, and therefore does not involve a significant profit sacrifice." Opinion of the Commission at 31, In re Rambus, Inc., FTC Dk. No. 9302 (Aug. 2, 2006), available at http://www.ftc.gov/os/adjpro/d9302/060802 commissionopinion.pdf. Ironically, the Commission concluded-using language eerily reminiscent of that used to describe antitrust's per se rules-that it could find Rambus's conduct unlawful under this standard without the need to examine evidence of anticompetitive effects, because "[dieceptive conduct that confers durable market power by its very essence harms competition, and claims that the offender has not yet behaved like a monopolist provide no shelter." Id. at 115. COL UMBIA B USINESS LA W RE VIE W [Vol. 2007 No. 2:371 ] SECTION 2 STANDARDS AND CONSUMER WELFARE 381 conduct is erroneously prohibited (Type II errors). After all, isn't maximizing consumer welfare the ultimate goal of antitrust, and won't we best achieve that goal by asking about the effect of a particular practice or conduct using the metric of direct interest?31 Professor Salop correctly observes that standards that key liability to something other than consumer welfare effects, ostensibly in the interest of certainty, will at least sometimes penalize beneficial conduct or, more likely, exonerate harmful conduct. Critics of a net welfare standard tend to emphasize the difficulties businesses and their counselors would have in determining ahead of time whether a proposed course of conduct would create Section 2 liability risks-with the attendant concern that such a standard would chill behavior by risk-averse companies." Proponents answer that this uncertainty could be minimized by applying the test using an ex ante perspective based on "information reasonably available at the time the conduct was undertaken."34 "' As Salop notes, "Antitrust law is said to be a 'consumer welfare prescription."' Salop, supra note 23, at 312 (quoting Reiter v. Sonotone Corp., 442 U.S. 330, 343 (1979)). 32 See Salop, supra note 23, at 346 ("Although the profit sacrifice test is generally more prone to false negatives, it sometimes can lead to false positives and condemn potentially exclusionary conduct that raises consumer welfare."). As Salop discusses, one scenario under which the profit sacrifice test could lead to a false negative would be where a design change increased a product's value more than the cost of the change but also led to higher entry barriers (due, for example, to greater incompatibility with rival products) that permitted the defendant to raise prices far higher than the increased value of the re-designed product. Here, the conduct would not be condemned because no profits were sacrificed, but in Salop's view the exclusionary effect of the design change would nevertheless have harmed competition. See id. at 345-46. Salop's approach to new product innovation by a monopolist is arguably mirrored in the treatment of product reformulations that were alleged to maintain the defendant's monopoly in Abbott Labs. See Abbott Labs, 432 F. Supp. 2d at 416-18, 420-23 (holding that the alleged benefits of reformulations of defendant's branded anti-cholesterol medication, which added a new HDL indication and removed the requirement that it be taken with food, had to be balanced against alleged anticompetitive harm). " See Werden, supra note 24, at 432-33. 14 See Salop, supra note 23, at 341-42. Moreover, they observe, we should not be overly hesitant to apply such a test given that the same consumer welfare standard is employed in many other realms of antitrust jurisprudence, most notably the enforcement of Section 7 of the Clayton Act with respect to mergers and acquisitions." It is surely true that the agencies' enforcement of Section 7 is guided by an assessment of net welfare effects (again, leaving for another day the debate over the distinction, if any, between consumer surplus and total surplus). 6 That is to say, we look at proposed transactions with an eye to predicting whether they likely will, on balance, cause significant harm to competition and consumer welfare. If a transaction likely will harm competition substantially, then the agencies will challenge the transaction-or negotiate a fix that allows the pro-competitive aspects to go forward while avoiding the anticompetitive harms. It is thus fair to ask why the same sort of inquiry should not be carried out to assess whether a monopolist's conduct violates Section 2 of the Sherman Act. Fortunately, the federal antitrust enforcement agencies have considerable experience applying a net consumer welfare test-and doing so from an ex ante perspective-in the merger arena. One might even say that the Antitrust Division and the FTC are the merger enforcement experts. I submit that our experience with merger enforcement offers at least five lessons that counsel caution in the use of a net consumer welfare standard to determine antitrust liability for unilateral conduct. A. Lesson No. 1: In Practice, Neither the Law Nor Its Enforcement Functions as a True "Consumer Welfare Prescription" My first observation is that merger law is not quite a "consumer welfare prescription." To be sure, the agencies' focus is on economic efficiency rather than other social interests, and we do not challenge mergers unless we are 3 See id. at 374. 36 See note 1, supra. COL UMBIA BUSINESS LA W RE VIE W [Vol. 2007 No. 2:371] SECTION 2 STANDARDS AND CONSUMER WELFARE 383 persuaded that they likely will harm consumer welfare. Because we apply this standard, there is no doubt that firms tend to steer clear of potentially anticompetitive transactions and, other things being equal, favor deals that are relatively more beneficial to consumers. In addition, in some narrow circumstances our analysis does examine whether alternatives available to the merging parties might enhance welfare more than their proposed transaction. But this analysis only takes place when we disallow merging parties from taking credit for efficiencies that are not "merger specific," in that they could practicably be achieved in a less anticompetitive way." But requiring efficiencies to be merger specific is quite far from saying that we seek to apply the law in a way that overtly maximizes consumer welfare. Market forces determine what transactions get proposed, and we review those to determine their effects on competition, if any. If the chosen transaction does not on balance harm consumer welfare, we do not block it even if pursuing it will rule out some other, possibly more pro-competitive, outcome. And we certainly do not enforce the law against firms that choose not to pursue, or begin pursuing but then decide to abandon, an available procompetitive deal. Instead we rely on the market to achieve the most efficient set of outcomes. In the merger enforcement arena, our role with respect to maximizing consumer welfare is more akin to an umpire calling balls and strikes than a pitcher choosing his best pitch for the situation. There is another respect in which our merger policy does not always seek to maximize welfare. When we evaluate horizontal transactions, we ask whether they will facilitate anticompetitive coordinated interaction or lead the merged firm unilaterally to increase price or reduce output. But we do not condemn transactions because they might lead to such striking cost reductions or breathtaking improvements in the firms' product offerings that rivals will perish, possibly " See Dept. of Justice and Federal Trade Comm'n, Horizontal Merger Guidelines § 4 (rev. 1997), available at http://www.usdoj.gov/atr/public/ guidelines/hmg.pdf. leading in turn to theoretical long-run adverse welfare consequences. We don't even ask this question. In cases involving a vertical component, we do consider whether the transaction might place vital inputs out of the reach of rivals, and thereby harm the competitive process, but we do not condemn such transactions just because vertical integration via merger will allow the firms to improve their efficiency and perhaps in the process limit their rivals' prospects for success. This experience with mergers counsels caution in the world of unilateral conduct. There may be little dispute about this. I suspect few would suggest that Section 2 is (or should be) violated when a firm merely fails to act in a way that maximizes consumer welfare. As the Supreme Court observed in Verizon Commc'ns Inc. v. Law Offices of Curtis V. Trinko, "the Sherman Act .. . does not give judges carte blanche to insist that a monopolist alter its way of doing business whenever some other approach might yield greater competition."38 But under an explicit "net consumer welfare" test, firms may perceive exactly this risk. If firms anticipate that one course of conduct likely would harm rivals or impede their entry-even if it might be efficient and beneficial to consumers in the short run-a consumer welfare test would give them strong incentives to pursue an alternative course of conduct that improved efficiency less and thereby gave rivals a better chance of survival. 9 This calculus might be expected even when the short-term benefits were clear and the longer-term harm to rivals far more speculative. 3 Verizon Commc'ns Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398, 415 (2004). " An illustration of how Section 2 risks can lead firms to compete less hard may be present in a recent case applying Section 2 and LePage's. In Applied Med. Res. Corp. v. Ethicon Inc., the defendant restructured its discount program in late 2003, after the LePage's decision, in a manner that might be interpreted as reflecting less aggressive discounting against those rivals whose product lines happened to be narrower than the defendant's. See Applied Med. Res. Corp. v. Ethicon Inc., No. SACV031329, 2006 WL 1381697 at *2 (C.D. Cal. Feb. 3, 2006). COL UMIA B USINESS LA W RE VIE W [Vol. 2007 No. 2:371] SECTION 2 STANDARDS AND CONSUMER WELFARE 385 Some of Professor Salop's illustrations applying his test to hypothetical conduct suggest exactly this result. One such illustration involves an investment that reduces the monopolist's costs, which the net consumer welfare test might condemn if rivals cannot match the cost reduction and so are driven from the market, leading to later price increases. 0 Leaving aside whether we really want to discourage such investments under any circumstances, consider this slight amendment to his scenario. Imagine a firm deciding how to equip a new plant made necessary by rising demand. The investment will be made, but should the firm buy a low-cost machine or a high-cost machine? The high-cost machine will ensure that rivals stand a good chance of survival; the low-cost machine might well cause them grievous harm. It is not clear what signals Professor Salop's test would send in this instance. In the merger arena, we know that an acquisition that would result in an efficient productive asset being added to the industry would not be condemned even if it might someday lead to higher prices because rivals cannot keep pace. Why should the result be different under Section 2? And if the argument is that an exception must be made to create a safe harbor for certain forms of investments and innovations, then the key point is acknowledged: firms need greater certainty than an explicit and unadorned consumer welfare test can provide. B. Lesson No. 2: Speed, Certainty and Finality Are Highly Valued In the merger enforcement world, the vast majority of transactions posing potential competitive concerns must be reported to the federal antitrust agencies under the Hart- Scott-Rodino Act (HSR Act), and the parties must give the agencies time to investigate before consummating their transactions.4' This pre-merger review process is highly 40 See Salop, supra note 23, at 339-41. 41 See 15 U.S.C. § 18(a) (2006) (requiring parties to certain acquisitions of voting securities or assets to notify both the Antitrust Division and the FTC before consummating the proposed transaction and beneficial from the perspective of effective antitrust enforcement since it allows for an orderly review before the "eggs get scrambled." The HSR process is beneficial from the perspective of most parties as well because it provides a high degree of certainty about the outcome of the "net consumer welfare" analysis. Fundamentally, the parties are able to ask the agencies up-front whether their transaction will be challenged and then, once the waiting period has expired without a challenge, they may consummate the transaction with great confidence that the reviewing agency will not later allege that the transaction violates Section 7.42 We know from our merger experience that certainty and finality are very valuable. Deals are typically made contingent on review and clearance by the antitrust agencies. Many deals would not close, or would not be proposed in the first place, if parties faced significant lingering doubt about whether antitrust enforcers or private plaintiffs might challenge them in the future. Usually the merger review process provides that confidence quite rapidly. A very high percentage of all reportable transactions are cleared without either agency even opening a preliminary investigation (PK).43 When the Antitrust Division does open a PI, we work hard during the to submit certain information to both agencies; after notification, parties must wait a specified time, usually thirty days, before the transaction can be consummated). 42 Business review letters provide another mechanism for seeking the Antitrust Division's enforcement intentions as to particular prospective conduct. See Antitrust Division Business Review Procedure, 28 CFR § 50.6, available at http://www.usdoj.gov/atr/public/busreview/201659c.htm. However, they are not generally viewed as practical for most transactions. See Ky P. Ewing, Jr., Thoughts on Seeking Business Reviews of Competitor Collaborations, 19 ANTITRUST 40, 42 (Summer 2005) ("[AIll else being equal, it is almost always better to rely on the advice of experienced antitrust counsel than it is to seek a Business Review Letter."). " See Background Information on the 2006 Amendments to the Merger Review Process Initiative at 3-4 (noting that in FY 2002-05, the agencies opened preliminary investigations for just 19% of transactions filed under HSR), available at http://www.usdoj.gov/atr/public/ 220241.pdf. COL UMBIA B USINESS LA W RE VIE W [Vol. 2007 No. 2:371] SECTION 2 STANDARDS AND CONSUMER WELFARE initial waiting period (usually the first thirty days after the HSR filing) to determine whether the competitive issues are serious enough to warrant a more thorough investigation. In only a small fraction of all reported transactions does the review extend beyond the first thirty days-via issuance of a "second request" or, in some cases, as a result of the parties' voluntary decision to "pull and refile."" Between 2002 and 2006, the Antitrust Division and FTC issued second requests in only approximately 3% of all reportable transactions." In the overwhelming majority of cases, therefore, the parties learn very quickly that the agencies do not intend to challenge their transaction. Even in the relatively few instances where we decide we must investigate transactions further, we strive to reach a conclusion as quickly and efficiently as possible. We know from experience that we will decide not to challenge most of the transactions for which we issue second requests. As a result, we try hard to conduct our second request investigations efficiently, so as to reduce the burden and delay for transactions that will not be challenged. This is the thrust of the Antitrust Division's Merger Review Process Initiative-announced in 2001 and updated again in December 2006 46 -pursuant to which we have narrowed the information sought by our typical requests for documents 4" Under the premerger notification procedures, administered by the FTC, an acquirer in a transaction may withdraw and refile its HSR notification once without having to pay a second filing fee, provided the transaction is substantively unchanged and the refiling is made within two business days of the withdrawal. See ABA Section of Antitrust Law, Premerger Notification Practice Manual 309 (3d ed. 2003). This process "grants the agencies additional time to review the transaction and may help the parties avoid the issuance of a Second Request." See id. at 327. " See also Background Information on the 2006 Amendments to the Merger Review Process Initiative, supra note 43, at 3-4 (stating that in FY 2000-05 the Antitrust Division issued second requests in approximately 32% of its preliminary investigations). 41 See Press Release, U.S. Dep't. of Justice, Antitrust Division Announces Amendments to its 2001 Merger Review Process Initiative (Dec. 15, 2006), available at http://www.usdoj.gov/atr/public/press- releases/2006/220302.pdf. 388 COLUMBIA BUSINESS LAW RE VIE W [Vol. 2007 and structured our investigations to focus our inquiry on those issues likely to be dispositive while deferring further analysis until it proves necessary.4 7 In several recent merger investigations, we focused our investigation on specific likely-dispositive issues (akin to a "quick look") based on internal company documents and other facts available to us; doing so has allowed us to close our investigation within a few months of issuing a second request.4" A salient feature of the merger review process, however, is that no matter how brief or extended the review process turns out to be, when the agencies look at a transaction and decline to challenge it, the transaction is exceedingly unlikely to confront significant antitrust hurdles. Very few " See Thomas 0. Barnett, Asst. Att'y Gen., U.S. Dep't. of Justice, Speech, Merger Review: A Quest For Efficiency (Jan. 25, 2007), http://www.usdoj.gov/atr/publicspeeches/221173.pdf; see also Background Information on the 2006 Amendments to the Merger Review Process Initiative, supra note 43, at 6-7 (noting that since the Merger Review Process Initiative was first adopted in 2001, the average length of merger investigations in which second requests were issued but no case was filed decreased from 248 days to 134 days, a 46% reduction). Among the issues that have recently proven dispositive during "quick look" second request investigations were: (a) the role of competitors whose likely expansion or entry was not reflected in available market share data; (b) the lack of significant competitive overlap between the firms despite their apparent shares of a more-broadly-defined market; and (c) the likelihood that the transaction would achieve significant merger- specific efficiencies that clearly outweighed any likely competitive harm. One such case was MediaNews Group Inc.'s acquisition of certain newspapers from The McClatchy Company. See Press Release, U.S. Dep't. of Justice, Statement of the Department of Justice Antitrust Division on its Decision to Close its Investigation of MediaNews Group Inc.'s Acquisition of the Contra Costa Times and San Jose Mercury News (July 31, 2006), available at http://www.usdoj.gov/atr/public/press- releases/2006/217465.pdf. After asking the parties to focus their initial production of materials on several issues that appeared potentially dispositive while deferring the need for full compliance with the second request, the Division closed its investigation upon finding that only a relatively small number of readers and advertisers viewed the parties' relevant newspapers as substitutes, that the parties would continue to face competition for the sale of newspapers and newspaper advertising, and that the parties had documented large, merger-specific cost savings in production and delivery systems. Id. at 1-2. No. 2:371] SECTION 2 STANDARDS AND CONSUMER WELFARE 389 transactions cleared by the agencies in the HSR process have subsequently been challenged by state enforcers or private parties. I would suggest that a principal reason for this phenomenon is that, in contrast to Section 2 law generally, competitors rarely have standing to complain about the harm they will suffer if their rivals are allowed to merge, even if they claim that in the long run consumers will be worse off without them. 49 Likewise, although nothing in the HSR Act prohibits the antitrust agencies from suing to challenge a transaction long after the waiting period has expired, as a practical matter that does not happen. No HSR-reported transaction in recent memory was cleared by the agencies and subsequently challenged by them based on a post-hoc reappraisal of net consumer welfare effects. The FTC's case against Chicago Bridge & Iron should not be viewed as an exception to this: although the waiting period had expired, the FTC was still investigating via formal process when the parties chose to consummate. 0 I similarly do not view as an exception situations in which the parties abused the HSR notification process, such as by withholding critical 41 "[T]he Supreme Court [has] emphasized that competitors often will benefit from a merger's anticompetitive effects and be injured only by its procompetitive effects, and it [has] held that standing should be strictly limited to claims of injury from anticompetitive effects." ABA SECTION OF ANTITRUST LAW, ANTITRUST LAw DEVELOPMENTS (FIFTH) 397 (2002), (citing Brunswick Corp. v. Pueblo Bowl-O-Mat, 429 U.S. 477 (1977) (in merger case, rejecting competitor's standing to assert damages claim where the alleged injury was the result of increased competition) and Cargill, Inc. v. Monfort of Colo., Inc., 479 U.S. 104 (1986) (rejecting competitor's standing to seek injunction in merger case based on prospect of increased post- merger competition, at least absent proper allegation that merged firm would engage in predatory pricing)). 'o See Opinion of the Commission at 1 n.2 and accompanying text, In re Chicago Bridge & Iron Co., N.V. (F.T.C. Dk. No. 9300, Jan. 6, 2005) (noting that transaction was closed "in the midst of the Commission's investigation"), available at http://www.ftc.gov/os/adjpro/d9300/050106 opionpublicrecordversion9300.pdf; AUJ Initial Decision at 1-2, In re Chicago Bridge & Iron Co., N.V. (F.T.C. Dk. No. 9300, Jun. 18, 2003) (noting that the parties had filed HSR notifications for the transaction), available at http://www.ftc.gov/os/2003/06/cbiid.pdf. COLUMBIA BUSINESS LA W REVIEW information that revealed serious competitive issues leading to a more in-depth investigation and a substantive challenge to the transaction.5' Nor am I counting those situations where we have conducted a post-consummation investigation of potential "gun-jumping."52 For transactions that do not meet the HSR thresholds and are thus outside the formal HSR review process, the same principles apply. When the agencies learn about a deal prior to consummation-sometimes because the parties value certainty and come forward-we try very hard to conduct a pre-consummation investigation. In the Division's recent investigations of non-reportable transactions, we have 51 See, e.g., Automatic Data Processing, Inc. (FTC Dk. No. 9282). In that case, the FTC challenged Automatic Data Processing's (ADP) acquisition of the assets of AutoInfo, Inc. (AutoInfo) months after the HSR waiting period expired because the FTC learned that ADP's HSR filing had omitted 4(c) documents demonstrating, inter alia, that the proposed acquisition "would create serious competitive concerns and that ADP believed that the Acquisition would give ADP a monopoly or virtual monopoly in several product markets." Complaint at IT 4, 6, Automatic Data Processing, Inc. (FTC Dk. No. 9282, Nov. 13, 1996), available at http://www.ftc.gov/os/1996/11/d9282cmp.pdf. ADP paid $2.97 million in civil penalties for the HSR violation, see id. at 8, and ADP ultimately agreed to a consent decree requiring divestitures and other remedies for the competitive harm caused by the underlying transaction. See Decision and Order (FTC Dk. No. 9282, Oct. 20, 1997), available at http://www.ftc.gov/os/1997/10/autoinfo.htm. 2 See, e.g., Complaint for Civil Penalties for Violation of Premerger Reporting Requirements of the Hart-Scott-Rodino Act at $ 4, 7, United States v. Qualcomm Inc., Civ. A. No. 1:06CV00672 (PLF) (D.D.C. April 13, 2006) (after waiting period expired in December 2005 and transaction closed in January 2006, the Department of Justice charged that Qualcomm "had effectively acquired Flarion's business before the expiration of the Section 7A waiting period through the Merger Agreement's requirements that Flarion obtain Qualcomm's consent before undertaking numerous competitive activities and through the parties' conduct, by which Flarion did not make even routine business decisions unless and until Qualcomm consented"), available at http://www.usdoj.gov/ atr/cases/f215600/215608.pdf; see also Final Judgment, United States v. Qualcomm Inc., Civ. A. No. 1:06CV00672 (PLF) (D.D.C. April 19, 2006) (judgment entered ordering Qualcomm and Flarion to pay civil penalty of $1.8 million), available at http://www.usdoj.gov/atr/cases/f216200/ 216249.pdf. [Vol. 2007 No. 2:371] SECTION2 STANDARDS AND CONSUMER WELFARE 391 managed to complete our review prior to the deal being consummated in the large majority of cases, often with the cooperation of the parties. However, it is not always possible to complete a pre- consummation review. Indeed, sometimes the parties work to complete their deal under the radar, perhaps hoping that we will be less likely to act once the transaction is consummated and the "eggs are scrambled."53 This strategy, however, will not prevent us from investigating transactions that appear likely to harm consumers. And when we conclude that such harm is likely or has already occurred, we will find a way to reverse the harm even if the parties' actions have made it difficult or impossible to interpose a simple and straightforward structural remedy. In those instances, we will relax our usual preferences for structural relief, and we will be creative about remedying the violation. The fact that we do enforce Section 7 in these settings, however, does not diminish the important lesson that certainty is valuable. In those cases where we have challenged consummated transactions, the mergers were quite clearly anticompetitive-typically mergers of 2-to-1 or 3-to-2 in highly concentrated markets, where the parties knew perfectly well that they were buying (or combining with) one of their closest rivals. To cite just a few examples, in the Antitrust Division's Southern Belle case, DFA's acquisition of rival Southern Belle dairy reduced the number of independent bidders for school milk contracts from two to one for forty-five school districts in eastern Kentucky, and from three bidders to two for fifty-five school districts in eastern Kentucky and Tennessee. 4 Two recent FTC matters are similar: in the Chicago Bridge case, the FTC alleged that See Assessing Part III Administrative Litigation: Interview with Timothy J. Muris, 20 ANTITRUST 6, 9-10 (2006) (noting that "in almost all cases" it is difficult to continue an investigation after a transaction is consummated, since "once the eggs are scrambled, it is usually over"). " See Amended Complaint at 3, United States v. Dairy Farmers of America, Inc., Civ. A. No. 03-206-KSF (E.D. Ky. Mar. 30, 2004), available at http://www.usdoj.gov/atr/cases/f218700/218766.htm. COLUMBIA BUSINESS LA W REVIEW the merger gave the firm a monopoly position in the market for two types of industrial storage tanks,55 and in its challenge to Hologic's 2005 purchase of the breast cancer screening and diagnosis business of Fischer Imaging, the FTC alleged that the merger had eliminated Hologic's only significant U.S. competitor in the prone stereotactic breast biopsy market. 6 The broader lesson is the more important one, however. A net consumer welfare test works in the merger arena in part because the pre-consummation investigation process itself provides certainty, and a full-blown examination of net consumer welfare effects is reserved for a small fraction of all cases. These features are not present in the Section 2 arena. There is no mechanism for the antitrust agencies to provide up-front "clearance" for the wide array of conduct engaged in by firms on a day-to-day basis. The net welfare examination would instead have to be conducted in court, in many cases long after the firm committed to its business strategy. It would often take place in private antitrust litigation directed, at least in the first instance, by the firm's rivals. Competitors would generally have standing to assert Section 2 claims against conduct that they perceive causes harm to them, and through them, the competitive process. 7 " See Complaint at 20-23, In re Chicago Bridge & Iron Co., N.V. (F.T.C. Dk. No. 9300, Oct. 25, 2001), available at http://www.ftc.gov/os/2001/10/ chicagobridgeadmincmp.htm. 56 See Complaint at 17, Hologic Inc. (F.T.C. Dk. No. C-4165, Jul. 7, 2006), available at http://www.ftc.gov/os/caselist/0510263/0510263 complaint.pdf. 57 See, e.g., Barton & Pittinos, Inc. v. SmithKline Beecham Corp., 118 F.3d 178, 184 (3d Cir. 1997) (test for standing is whether plaintiff was a customer or a competitor of companies that actually supplied the good or service); General Indus. Corp. v. Hartz Mountain Corp., 810 F.2d 795, 809 (8th Cir. 1987) ("[S]tanding to sue under the Sherman Act is limited to a 'consumer or competitor' that proximately suffers antitrust injury." (citing Associated Gen. Contractors v. Cal. State Council of Carpenters, 459 U.S. 519, 539 (1983))); see generally ABA SECTION OF ANTITRUST LAw, ANTITRUST LAW DEVELOPMENTS (FIFTH) 838-66 (2002) (noting that [Vol. 2007 No. 2:371] SECTION 2 STANDARDS AND CONSUMER WELFARE 393 Magnifying the uncertainty, in most cases firms making Section 2 predictions would need to guess at how a jury would apply a consumer welfare standard rather than attempting to gauge (as in the merger arena) how a neutral enforcement agency will view the transaction through a lens of economic efficiency. If one assumes as a predicate the usual background of any Section 2 case-i.e., very high market shares, entry barriers, and rivals that have in fact been seriously harmed (either by the defendant's conduct or by their own failures in the marketplace)-the "consumer welfare" inquiry would seem typically to raise factual questions not conducive to resolution on summary judgment. To grasp the uncertainty this could cause, consider for a moment the jury instructions recommended by the ABA in rule-of-reason cases, where liability is already based explicitly on the conduct's net effect on consumer welfare. Those instructions ask juries to ponder, inter alia, whether in light of a myriad of "rule-of-reason" factors the conduct has resulted in "a substantial harm to competition" that "substantially outweighs the competitive benefit."58 It would be understandable if firms making complex, unilateral business decisions had difficulty predicting the likely outcome of a jury's deliberation on such questions. As a result, the firm's risk assessment must take into account that it might be sued not just by an agency motivated to do the right thing, but by a rival motivated to impede the firm's competitive initiatives, even if they benefit consumers. competitors injured by allegedly predatory or exclusionary acts generally are among those who can sue for damages under Section 2). 68 ABA SECTION OF ANTITRUST LAW, MODEL JURY INSTRUCTIONS IN CIVIL ANTITRUST CASES, 2005 Edition at A-4 (Instruction 3A) (2005). Model Instruction 3B lists some of the factors the jury may consider in evaluating whether competitive harm outweighs competitive benefit, including: "the effect of the restraint on price, output, product quality and service; the purpose and nature of the restraint; the nature and structure of the relevant market, both before and after the restraint was imposed; the number of competitors in the relevant market and the level of competition among them, both before and after the restraint was imposed; and whether the defendant possesses 'market power."' Id. at A-7. COLUMBIA BUSINESS LAW REVIEW C. Lesson No. 3: A Full-Blown Net Welfare Analysis is Costly and Time Consuming Another lesson from the agencies' merger enforcement experience is that conducting an assessment of the net consumer welfare effects of a proposed transaction is hard work. Sometimes, of course, we can determine very quickly and easily that a transaction is unlikely to have significant adverse effects on consumer welfare. In the merger setting, that conclusion can be reached based on market conditions that rule out any likely effect, perhaps because the market is unconcentrated, the parties will lack market power even if combined, entry is easy, or the parties do not compete with one another much. Screens of this sort are very important because they allow agency staffs to rule out serious competitive problems in the initial waiting period based on information in the public domain, the parties' HSR filings, and the staffs often extensive industry experience. In the Section 2 world, structural screens of this sort are vital as well, but they only go so far. For those firms that do have monopoly power or a realistic prospect of achieving it, applying such screens does not begin to answer the question of whether particular conduct might be deemed unlawful as "willful monopolization." For the relatively small percentage of proposed mergers that still raise serious competitive concerns after a threshold structural review, the agencies devote tremendous effort to predicting the transaction's likely consumer welfare effects. In the small minority of cases where the Antitrust Division opens a PI,59 the investigative staff undertakes considerable work during the initial waiting period to analyze the transaction's likely effects. During this period, legal and economic staff (usually at least two or three lawyers and one or more Ph.D. economists) conduct a dozen or more interviews of potentially knowledgeable customers, competitors, and industry consultants. In addition to the 4(c) documents and other materials supplied by the parties " See supra note 43 and accompanying text. [Vol. 2007 No. 2:371] SECTION 2 STANDARDS AND CONSUMER WELFARE 395 as part of their HSR submission," we often seek voluntary production of additional documents and data from the parties,6 and where it is available we obtain data from other industry sources. When our investigations move to the next stage, the investigative effort is considerably greater. Even though we often can narrow our focus to a few dispositive issues, the burdens are usually extensive-both for us and for the parties. Even for transactions that are ultimately cleared, the typical second request investigation that is not resolved on a quick look can take several months,62 employ an expanded staff of lawyers and economists, and involve the production and review of hundreds and sometimes many thousands of boxes of documents from the parties, and often many more from third parties. This is in addition to econometric analyses, still more interviews, and often depositions of the parties and third parties. The merger setting properly supports this amount of effort. Our investment of enforcement resources is worthwhile when there is a reasonable basis for suspecting that a transaction will cause significant harm to consumer welfare. And from the parties' perspective, mergers tend to be "big deals" in the life of the firm. They are significant one-time events that can support-both in terms of time and 60 See Instructions, Antitrust Improvements Act Notification and Report Form for Certain Mergers and Acquisitions I 4(C) (calling for submission of, inter alia, "all studies, surveys, analyses and reports" prepared by or for any officers or directors "for the purpose of evaluating or analyzing the acquisition with respect to market shares, competition, competitors, markets, potential for sales growth or expansion into product or geographic markets"), http'//www.ftc.govlbc/hsr/P989316PMNRulesand FormallnterpretationsElectronicSubmissionofForms-Instructions.doc. 61 See U.S. Dept. of Justice Antitrust Div., Merger Review Process Initiative at 2 (noting that staff will request voluntary production of listed information "as soon as feasible during the initial waiting period"), http://www.usdoj.gov/atr/public/220237.pdf. 62 See Background Information on the 2006 Amendments to the Merger Review Process Initiative, supra note 43, at 6-7 (noting that in FY 2005 Division investigations in which second requests were issued and no case was filed lasted an average of 134 days from opening to closing). resources-the investment in an ex ante net consumer welfare analysis. That may not be so for many other kinds of day-to-day business decisions, and firms may shy away from making those decisions in order to avoid the burdens of a full-blown welfare analysis. D. Lesson No. 4: Outcomes of a Net Welfare Analysis are Hard to Predict Despite features of merger analysis that arguably facilitate predictions of likely net consumer welfare effects, outcomes can still be quite hard to predict and subject to reasonable debate. In the Section 2 environment, reliable ex ante predications are likely to be far harder. Consider first the Antitrust Division's merger experience. Outside of the structural safe harbors, there are very few easily-accessible facts that provide a sound basis for predicting whether the reviewing agency will conclude that a particular transaction is likely to harm consumer welfare. Concentration figures-even if they were easy to calculate- certainly offer no such metric. Numbers are just the starting point of our analysis,63 and it is easy to identify recent cases where concentration figures and other structural indicia point in potentially misleading directions. Consider two recent Antitrust Division investigations: Blackboard/WebCT and Exelon/PSE&G. Blackboard involved academic "course management systems" purchased by colleges and other schools.64 Public estimates pegged Blackboard's share of past sales at 45% and WebCT's share See Horizontal Merger Guidelines, supra note 37, at § 2.0 ("[Mlarket share and concentration data provide only the starting point for analyzing the competitive impact of a merger. Before determining whether to challenge a merger, the Agency also will assess the other market factors that pertain to competitive effects, as well as entry, efficiencies and failure."). ' See E. Armington, E. Emch & K. Heyer, The Year in Review: Economics at the Antitrust Division, 2005-06, 29 REv. INDUS. ORG. 305, 306 (2006). [Vol. 2007COL UAMIA B USINESS LA W RE VIE W No. 2:371] SECTION 2 STANDARDS AND CONSUMER WELFARE 397 at 35%,65 and their post-merger combined share of over 80% would have yielded a post-merger Herfindahl-Hirschman Index (HHI) value of well over 6400.66 To see whether the merging parties' high shares reflected a genuine market preference for these two established companies compared to more recent entrants, the Antitrust Division looked at whether customers saw the firms' products as next-best substitutes. Our investigation disclosed that these rivals had been quite successful in luring new customers, with recent win rates far higher than predicted by installed-base market shares.6" As a result, the Division decided not to challenge the transaction9.6 In the Exelon-PSEG transaction, the Division challenged the transaction despite much lower concentration. Exelon involved electricity generation in the eastern part of the PJM control area, principally New Jersey.70 Concentration was much lower than in Blackboard/WebCT, with HHIs in the 2100-2700 range.71 However, the Division's detailed examination of the way the market for wholesale electricity actually worked and the specific assets being acquired led it to conclude that the merger would enable the parties to raise 65 See Ellen McCarthy, District Software Firm Will Buy Biggest Rival, WASHINGTON POST, October 13, 2005, at D4. ' See E. Armington, et al., supra note 64, at 307 (market shares as reflected in the installed base were "extremely high" by traditional Merger Guidelines standards). As explained in the Horizontal Merger Guidelines, market concentration is a function of the number of firms in a market and their respective market shares, and the agencies use the HHI index of market concentration to assist in understanding market data. See Horizontal Merger Guidelines, supra note 37, at § 1.5. The HHI "is calculated by summing the squares of the individual market shares of all the participants," and an HHI above 1800 means that the market is considered "highly concentrated." Id. 67 See E. Armington et al., supra note 64, at 307. Id. at 307-08. 69 Id. at 308. 70 See Complaint at 16-29, United States v. Exelon Corp. and Pub. Serv. Enter. Group (D.D.C., filed June 22, 2006), available at http://www.usdoj.gov/atr/cases/f216700/216785.pdf. 71 Id. at 30-33. COLUMBIA BUSINESS LA W REVIEW prices during times of high demand.72 The PJM system operator runs a daily auction market that "clears" at a price necessary to bring forth just enough generation to meet demand, with the clearing price determined by the cost of the incremental generating units needed to supply the market.7 3 By withholding output from their "mid-merit" and "peaking" generating units, the costs of which would be close to market-clearing prices, the merged firm could have profited by selling electricity from its baseload units at much higher prices.74 The Division thus insisted upon divestitures targeted at the specific assets that would have enabled the merged firm to pursue such a unilateral strategy profitably. 5 There is no simplistic predictive tool capable of describing the Antitrust Division's enforcement pattern that flows from its focus, in each merger investigation, on the transaction's likely effect on consumer welfare. Although the agencies are good at what they do, we often find that our own early assessments are not borne out by the facts. This unpredictability works in both directions. Sometimes, as in the Blackboard/WebCT matter, we conclude that a transaction's effects will be less than initially expected. Other times closer analysis leads us in the opposite direction. Our investigation of Premdor's acquisition of International Paper's Masonite business unit was one such example. Premdor made doors and some doorskins; Masonite made doorskins, and Premdor was its largest customer. 6 On its 72 Id. at 34-36. 73 Id. at 19. 74 Id. at 34-36. 71 See Competitive Impact Statement at 12-16, United States v. Exelon Corp., Civ. A. No. 1:06CV01138 (D.D.C. Aug. 10, 2006), available at http://www.usdoj.gov/atr/cases/f217700/217717.htm. The proposed merger ultimately did not go forward, and the United States dismissed its complaint against the parties. See Plaintiffs Notice of Dismissal of Complaint at 1-2, United States v. Exelon Corp., Civ. A. No. 1:06CV01138 (D.D.C. Sept. 27, 2006), available at http://www.usdoj.gov/atr/cases/ f219000/219032.pdf. 76 See Complaint at 1, United States v. Premdor Inc., Civ. No. 01- 01696 (D.D.C. Aug. 3, 2001), available at http://www.usdoj.gov/atr/cases/ f8800/8893.htm. [Vol. 2007 No. 2:371] SECTION 2 STANDARDS AND CONSUMER WELFARE face, this transaction appeared largely vertical, and such transactions seldom pose concerns. On close analysis, however, we determined that the acquisition would increase the likelihood of coordination by doorskin makers in two ways: by eliminating the potential of disruption caused by Premdor's expansion into doorskins and by more closely aligning the cost structures of the remaining doorskin firms.77 This outcome would not have been easy to predict when the investigation first began. When the Division does reach a judgment about consumer welfare effects after months of analysis, the correctness of our conclusion is often the subject of great debate. Again, it cuts both ways. Sometimes we are criticized for seeing a problem that others do not believe is real. Both the Antitrust Division and FTC have lost cases recently because courts disagreed with our appraisal of the consumer welfare implications of proposed mergers.78 And our wisdom is also questioned when we decline to challenge some or all of a transaction. Witness the debate in the Tunney Act proceedings over VerizonlMCI and SBC/ATT,79 and the questions raised about our decision in Whirlpool/Maytag. ° " See id. at I1 34-37. The Division's concerns were addressed by the divestiture of one of Masonite's U.S. doorskin plants, and the transaction as modified was allowed to proceed. See Competitive Impact Statement at 4-5, 11-12, United States v. Premdor Inc., Civ. No. 1:01 Cv 01696 (D.D.C. Aug. 3, 2001), available at http://www.usdoj.gov/atr/cases/f9000/9017.htm. 78 See United States v. Oracle Corp., 331 F. Supp. 2d 1098, 1175-76 (N.D. Cal. 2004) (entering judgment for defendants in government merger challenge, where court found that government had failed to prove, inter alia, that anticompetitive effects were likely to result from the merger); Federal Trade Comm'n v. Arch Coal, Inc., 329 F. Supp. 2d 109, 115 (D.D.C. 2004) (denying preliminary injunction in government merger challenge where transactions "[did] not reduce the number of competitors and only modestly increase[d] the concentration in what has been a very competitive market" and where court found that plaintiffs had not met their burden to "show a likelihood that the challenged transactions will substantially lessen competition"). " See United States v. Verizon Commc'ns, Inc., C.A. No. 1:05CV02103 (EGS) (D.D.C., complaint and proposed final judgment filed Oct. 27, 2005); United States v. SBC Commc'ns, Inc., C.A. No. 1:05CV02102 (EGS) (D.D.C., complaint and proposed final judgment filed Oct. 27, 2005). COLUMBIA BUSINESS LAW REVIEW If the correctness of our judgments about consumer welfare effects is subject to some uncertainty, ex ante "consumer welfare" judgments by firms in the Section 2 context would seem more uncertain by several orders of magnitude. Several features of merger analysis that facilitate predictions about likely competitive effects are not present in the Section 2 setting. In the merger setting, we ordinarily have a discrete pre-transaction "competitive" benchmark, we are able to isolate the effects of a proposed transaction on price and output in the market relative to that benchmark, and we can make sound judgments about the welfare implications of transactions that generate consumer benefits in the near term. Under the Antitrust Procedures and Penalties Act, 15 U.S.C. §§ 16(b)-(h) (also known as the Tunney Act), a proposed consent decree that settles a civil antitrust case brought by the United States may not be entered until the proposed Final Judgment and Competitive Impact Statement are published and the public has at least sixty days to submit comments. See Plaintiff United States' Explanation of Consent Decree Procedures, United States v. Verizon Commc'ns, Inc., supra, available at http://www.usdoj.gov/ atr/cases/f212400/212434.pdf. Public comments were received from, inter alia, two associations of competitors and customers of the merging parties. See Plaintiff United States' Response to Public Comments at 13, 28, United States v. Verizon Commc'ns, Inc. and United States v. SBC Comm'ns, Inc., supra, available at http://www.usdoj.gov/atr/cases/ f215100/215174.pdf. An evidentiary hearing was held on November 30, 2006, see Judge Not Ready to Act Yet on Bell Mergers, COMMUNICATIONS DAILY at 3 (Dec. 1, 2006), and in March 2007, seventeen months after the proposed settlements were filed, the court entered the consent decrees as filed. See United Stated v. SBC Commc'ns, Inc., 2007 WL 1020746 (D.D.C. Mar. 29, 2007). "0 See Neal R. Stoll & Shepard Goldfein, Withdrawal of Unilateral Effects Theory From § 7 Arsenal?, N.Y.L.J. at 3 (Apr. 18, 2006) ("Given the high market shares of the two companies, the decision was a surprise to many antitrust practitioners."); Stephen Labaton, New View of Antitrust Law: See No Evil, Hear No Evil, N.Y. TIMES, May 5, 2006, at C5 (decision "left private antitrust practitioners . . .wondering whether there are ... any mergers that this administration would challenge"); see also Department of Justice Antitrust Division Statement on the Closing of its Investigation of Whirlpool's Acquisition of Maytag (Mar. 29, 2006), http://www.usdoj.gov/atr/public/press-releases/2006/215326.htm. [Vol. 2007 No. 2:371] SECTION 2 STANDARDS AND CONSUMER WELFARE 401 Carrying out a net consumer welfare analysis in the Section 2 setting is far more complex. Section 2 cases- including many of the hypothetical ones presented by Professor Salop-often involve actions that in the short run are plainly beneficial to consumer welfare. Most of the tools of potential "exclusion"-e.g., formal exclusive dealing arrangements, bundled rebates, loyalty discounts and myriad other practices-are also recognized as achieving efficiencies and price reductions."' The consumer welfare effects of potential concern-i.e., a reduction in competition caused by the demise or deterrence of rivals-will occur, if at all, indirectly and later in time. These features of Section 2 analysis magnify the uncertainty facing the firm considering what activities it may lawfully engage in, for at least three reasons: (1) Because the harmful impacts will not be immediate, firms carrying out ex ante evaluations of Section 2 risks must try to predict how the effects of their conduct will be judged long after the business decision is made. On that time horizon, however, numerous other events will already have taken place by the time the conduct's own effects are felt-including changes in taste, technology, and other general market trends; other conduct by the firm itself; and decisions by rivals, including perhaps failure brought about by those firms' own incompetence or bad business plan. The effects of the firm's conduct on price, output and other measures of consumer welfare will not be easy to isolate. Rivals will have failed, or their entry will have been deterred, and prices will remain high, but why? Most troubling, the monopolist itself likely will have engaged in a wide array of conduct-some perhaps obviously benign, other conduct less so-that tended to diminish rivals' business opportunities. 81 See, e.g., Herbert Hovenkamp, Discounts and Exclusion, 2006 UTAH L. REV. 857, 860-61 (2006) (noting that "discounting is a vertical practice that is presumptively procompetitive"); Andrew I. Gavil, Exclusionary Distribution Strategies by Dominant Firms: Striking a Better Balance, 72 ANTITRUST L.J. 3, 9-10 (2004) (noting that "the courts early on recognized the competitive utility of exclusive dealing"). Professor Salop's illustrations of the application of the consumer welfare test neatly isolate and measure the effects of particular decisions, but in practice firms usually could not be nearly so sure about the feasibility of testing those effects in isolation. 2 Citing LePage's and other cases, plaintiffs will allege a "monopoly broth" worthy of condemnation. And when that broth is analyzed and the consumer welfare balance is struck, jurors who perceive some subset of the firm's actions as "unfair" and find that those actions contributed to rivals' lack of success may well expect the firm to prove measurable benefits that outweigh the full extent of the firm's anticipated monopoly price levels. (2) Closely related to this first point, a Section 2 consumer welfare inquiry, unlike one carried out in the typical pre-consummation setting of merger enforcement, likely would not be based on a straightforward forward- looking comparison between the status quo and the predicted state of the world following a specific identifiable transaction. When the Section 2 case is brought some years after a course of conduct has been carried out, the consumer welfare inquiry would instead compare the conduct's observed effects (intermixed with the clutter of other marketplace events) against some set of hypothetical "but for" worlds in which the dominant firm behaved differently. Ex ante, firms making business decisions that might 82 See Salop, supra note 23, at 336-41. For example, Salop's analysis of a hypothetical monopolist's changes in design depends on an assumed dollar amount assigned to the value that future consumers will place on the product improvement. Cases where the balancing test has ostensibly been applied do not provide any comfort about the ease of making such predictions. For example, Werden points out that the Microsoft court did not attempt to apply its stated balancing test because the district court "had found the challenged conduct exclusionary under what is essentially the no economic sense test." Werden, supra note 24, at 430. ' See LePage's Inc. v. 3M, 324 F.3d 141, 173 (3d Cir. 2003) (Greenberg, J. dissenting) ("Notwithstanding the evidence which demonstrates that LePage's lost business for reasons that could not possibly be attributable to any unlawful conduct by 3M, it argues that 3M willfully maintained its monopoly through a 'monopoly broth' of anticompetitive and predatory conduct."). COL UMBIA B USINESS LA W RE VIE W [Vol. 2007 No. 2:371] SECTION 2 STANDARDS AND CONSUMER WELFARE 403 implicate Section 2 would need to assess the relative welfare effects of two alternative and fairly speculative hypothetical states of the world. They will know that the firm is moving past the status quo one way or the other-since no firm that sits still can expect to succeed for very long. Thus, when pondering how to answer the question whether consumer welfare will be made better off or worse off by a particular set of contemplated actions, a fair response might be "compared to what?" When they make their business decisions, therefore, firms will be facing greater uncertainty that one might expect in the "before-and-after" comparisons involved in merger analysis. Apart from the uncertainty involved, one outcome of this line of reasoning might be a conscious choice to choose a course that may be less beneficial for consumers but also less likely to harm rivals and lead to long-run Section 2 risks. (3) Finally, the indirect character of most potential adverse Section 2 welfare impacts also magnifies uncertainty: it is hard to predict how a course of conduct will impact rivals and equally difficult to know whether that impact will be viewed by a jury as "welfare enhancing" or "harmful to consumers." The first question is easy to frame: how far can the firm go without pushing its rivals over the brink, or making their entry impossible? If a firm will be penalized for the welfare effects that follow the rival's demise, it will want to know where the edge of the cliff is located. Leaving aside whether as a policy matter we want firms generally to "lay off' their rivals in order to keep competitors alive, how will the firm know how far it can push without pushing too far? In our merger experience, we find that competitors' assessments of their rivals' competitive capabilities and financial health are highly imperfect. Sometimes .they are accurate, but often they are very far off. Similarly, we have all seen the futility of attempting precise--or even ballpark-predictions of how successful a new business strategy will be. Some ground balls turn into inside-the-park home runs, just as some towering flies that leave the bat with promise end up as outs. COLUMBIA BUSINESS LAW REVIEW We sometimes must make these kinds of predictions in the merger setting, such as when we ask whether rivals are likely to reposition in response to a merger of particularly close competitors. Such questions are not easy to answer, but we know enough not to rely entirely on the predictions of the merging firms in making our assessment. We dig deeper, probing data that is unavailable to the merging firms, and gathering facts from customers, competitors and third parties to which the merging parties do not have access, for obvious reasons. Sometimes this information is quite important. Sometimes it convinces us that the transaction likely will cause harm, and sometimes it makes us more confident that harm is unlikely. In the merger setting, reliance on information unavailable to the parties is appropriate, since in most cases a near-binding decision will be made before the parties consummate. In the Section 2 setting, of course, a test that relied critically on questions that were so hard to answer or facts outside the parties' knowledge would be far more problematic. Proponents of a consumer welfare test are aware of this problem, and to address it they propose to examine the welfare effects based on information reasonably available to the firm ex ante. 4 But we might still question how juries will ultimately decide the issue when it later turns out that rivals were actually harmed and prices actually went up. Even if this uncertainty could be overcome-perhaps by focusing on the firm's reasonable expectation of the impact on equally efficient rivals-it would still leave us with the more difficult, and persistent, question raised in the Section 2 arena: is a course of conduct that causes grievous harm to rivals good or bad for consumers? In the merger setting, we know that mergers leading to higher prices or reduced output are bad. When our analysis concludes that such effects are likely, there is no mystery about the right enforcement outcome. And observed pre-merger market See Salop, supra note 23, at 341-42. [Vol. 2007 No. 2:371] SECTION 2 STANDARDS AND CONSUMER WELFARE 405 conditions provide us with a base for comparison that is relatively easy to measure and well documented. Pre-merger conditions may not always be perfectly competitive, but they present a more or less measurable set of conditions that provides a benchmark for assessing whether the merger will make consumers worse off. We can ask: will prices be higher or will quality or output be lower than before the merger? Our economists can look for natural experiments that help answer this question-looking across markets or at markets across time for evidence predicting whether higher concentration (or, more pertinently, a merger among two specific competitors) will lead to higher prices. The famous empirical analysis in Staples /Office Depot provides a good example.85 By looking at markets with different numbers of office superstores, the FTC's analysis demonstrated the extent to which competition among those superstores, and indeed between the merging parties in particular, led to lower prices than would have occurred absent that competition-the ultimate "consumer welfare" issue under review. 86 In the Section 2 setting, this kind of evidence would tend to be far less useful to a consumer welfare test. Some such evidence was available in the Antitrust Division's Section 2 challenge to American Airlines' pricing conduct in response to entry by rivals in DFW city pair markets.87 There was cross-sectional evidence indicating that prices were lower in city pairs where two or more airlines provided nonstop " Federal Trade Comm'n v. Staples, Inc., 970 F. Supp. 1066 (D.D.C. 1997). 86 See id. at 1075-76 (granting a preliminary injunction blocking the proposed merger between Staples and Office Depot based in part on "compelling evidence" of harm, including evidence showing that prices at Staples were 13% higher in geographic areas where it faced no office superstore competition than where it competed with Office Depot, Office Max or both). 87 See United States v. AMR Corp., 140 F. Supp. 2d 1141 (D. Kan. 2001) (granting summary judgment in Section 2 predatory conduct action where court found airline did not price below an appropriate measure of cost and there was no dangerous probability of recoupment), affd, United States v. AMR Corp., 335 F.3d 1109 (10th Cir. 2003). COLUMBIA BUSINESS LA W REVIEW service than in pairs where there was only one nonstop provider.18 This said a great deal about whether "nonstop" service was a proper product market, but both the government and the court understood that it contributed very little to the question whether American's behavior- which the Division argued led to there being only one nonstop provider in many of American's DFW city pairs- was appropriately condemned. 9 It is not enough that monopoly means higher prices-we could have guessed that easily. The law and sound policy requires that enforcement agencies and courts make judgments about the propriety of the steps taken to acquire or preserve that monopoly. Many other hard questions present themselves: if the firm expects prices to increase at some point, is that because it is exploiting its monopoly power; because consumers will regard its product as more valuable and be willing to pay more; because it is acting efficiently and expects to prevail in the competitive race to the long-run extinction of its rivals; or because it is engaging in some conduct that unreasonably inhibits the ability of rivals to provide a competitive constraint on the firm's pricing? Framing the question as one about "net consumer welfare effects" does not answer these questions. For good reason, we know not to intervene in competitive markets every time prices go up or output goes down. We should not think intervention is always the appropriate course in monopoly settings either. 88 The government had submitted evidence that American's price-cost margins in non-stop routes from DFW, in which it did not compete with Southwest Airlines or a low-cost carrier (LCC), ranged from 44.3% to 50.7% in 1994-98, as compared to price-cost margins of 9.7% to 20.5% in the same years in non-stop routes from DFW in which American competed with Southwest or an LCC. See AMR Corp., 140 F. Supp. 2d at 1150. 9 See id. at 1194-95 (holding that the government had to meet the Brooke Group test and prove both that American priced its product below an appropriate measure of cost and that it enjoyed a realistic prospect of recouping its losses by supra-competitive pricing, and that under that standard, the government's claims failed). [Vol. 2007 No. 2:371] SECTION 2 STANDARDS AND CONSUMER WELFARE 407 E. Lesson No. 5: Enforcers and Courts are not Good at Micro-Managing Firm Behavior in Furtherance of Consumer Welfare (or Any Other Goal) This is my final lesson, and it is almost too obvious. Our merger experience gives us an extensive understanding of the competencies of agencies and the courts in remedying anticompetitive harm. It is of course well known that the Antitrust Division has a strong preference for structural remedies.9 0 We know from decades of hard experience that we are not generally well equipped to oversee "behavioral" remedies. A big part of the reason is our inability to regulate in real time a firm's business decision-making. The same lessons flow from our experience with hold separate orders, which are designed to preserve the competitive status quo pending the completion-usually within a few months-of a required divestiture. Even in that relatively short time frame, the marketplace does not stand still. Decisions must be made about the "held separate" business in real time-whether to invest in X, whether to compete for Y, on what terms, and so on. We are acutely aware how ill-equipped we are to make or even second-guess those decisions. Instead we establish processes that ensure these decisions will be made independently in the best interest of the business-even though we might not always be thrilled with the potential implications of the decisions that get made. I draw two lessons from this experience. First, of course, we are right to prefer structural remedies. And it should come as no surprise that we are generally more reluctant to intervene when there is not a clean, structural remedy that will preserve or restore consumer welfare. In the Section 2 setting, where the law is enforced primarily by private parties, the plaintiffs objective will usually be to recover 90 See U.S. Dep't of Justice, Antitrust Division Policy Guide to Merger Remedies at 8 (Oct. 2004), available at http://www.usdoj.gov/atr/public/ guidelines/205108.pdf. damages on behalf of an alleged "victim" rather than to recreate competitive conditions. This is not an improper goal of antitrust enforcement, but its implications are important. Such a regime serves its consumer welfare aim indirectly, through deterrence of conduct deemed anticompetitive. Therefore, it is imperative that the regime send properly calibrated signals lest a concern about leaving one "illegally" harmed rival uncompensated lead to consumers as a whole being deprived of the dynamic benefits of vigorous competition. Second, there is no reason to think that courts of general jurisdiction are any better at making or overseeing business decisions than antitrust enforcers. Yet private lawsuits are no less a form of regulation than government antitrust enforcement decisions. If the enforcement of Section 2 is used to second-guess the routine unilateral business judgments of firms, subjecting those decisions to an exacting assessment of their net effects on consumer welfare, that form of regulation likely would be just as clumsy and inefficient as attempts by antitrust enforcers to run the companies themselves as trustees for the public interest. III. CONCLUSION There is surely a place for consumer welfare in Section 2. Certainly unilateral conduct should not be condemned if it does not pose a threat to consumer welfare. And there likely are certain forms of conduct that could sensibly be judged under Section 2 based on their net effects on consumer welfare. For example, where a merger (be it horizontal or vertical) creates or maintains monopoly power, or where a pervasive pattern of exclusive dealing relationships prevent challenges to a monopolist's dominance, it is quite tempting to assess the legality of such conduct under Section 2 by asking about its net consumer welfare effects. There may well be other situations where monopolists (or attempted monopolists) should not be free to use any means possible, no matter how destructive of rivals' ability to compete, so long as they are pursuing a profitable or otherwise legitimate objective. [Vol. 2007COL UMBIA B USINESS LA W RE VIE W No. 2:371] SECTION 2 STANDARDS AND CONSUMER WELFARE 409 These possibilities notwithstanding, the Antitrust Division's experience applying a "net consumer welfare" test in the merger setting counsels caution in applying the same explicit consumer welfare test to judge the legality of unilateral conduct. First, since we cannot duplicate the pre-clearance regime of the merger review process in the Section 2 setting, the legal standards themselves must provide greater certainty to firms making decisions in real time, even if the price of that certainty might be more false negatives. A net consumer welfare standard, however, is likely to present opportunities for greater uncertainty in the Section 2 setting than it does in the merger setting. Second, in the merger setting, there is typically better information with which to evaluate ex ante the net welfare effects than there will be in the vast majority of Section 2 settings. It is also easier to isolate the potential welfare effects of a proposed transaction because we have a pre- transaction competitive price as a benchmark. In the Section 2 setting, there is much greater ambiguity: often the short run effects and long run effects will diverge, and even if adverse price or quantity effects are expected, it is much harder to judge whether they were caused by "bad" behavior or good. Hinging the Section 2 outcome solely on an inquiry about net consumer welfare effects might end up begging a whole series of other, harder questions. Finally, in the merger setting, typically the Antitrust Division or the FTC is responsible for deciding-on a consumer welfare standard-whether to bring a case. In the Section 2 setting, private parties have the initiative, and if a case is brought it very often will be decided by a jury without antitrust or economic expertise. These features of the Section 2 setting magnify uncertainty, and as a result adopting a net consumer welfare test in the Section 2 setting likely would chill procompetitive conduct more frequently than it does in the merger setting.