Federal Trade Commission v. H.J. Heinz Co.

Opinion for the court filed by Circuit Judge HENDERSON. KAREN LeCRAFT HENDERSON, Circuit Judge. On February 28, 2000 H.J. Heinz Company (Heinz) and Milnot Holding Corporation (Beech-Nut) entered into a merger agreement. The Federal Trade Commission (Commission or FTC) sought a preliminary injunction pursuant to section 13(b) of the Federal Trade Commission Act (FTCA), 15 U.S.C. § 53(b), to enjoin the consummation of the merger. The injunction was sought in aid of an FTC administrative proceeding which was subsequently instituted by complaint to challenge the merger as violative of, inter alia, section 7 of the Clayton Act, 15 U.S.C. § 18. The district court denied the preliminary injunction and the FTC appealed to this court. For the reasons set forth below, we reverse the district court and remand for entry of a preliminary injunction against Heinz and Beech-Nut. I. Background Four million infants in the United States consume 80 million cases of jarred baby food annually, representing a domestic market of $865 million to $1 billion. 1 FTC v. H.J. Heinz, Co., 116 F.Supp.2d 190, 192 (D.D.C.2000). The baby food market is dominated by three firms, Gerber Products Company (Gerber), Heinz and BeechNut. Gerber, the industry leader, enjoys a 65 per cent market share while Heinz and Beech-Nut come in second and third, with a 17.4 per cent and a 15.4 per cent share respectively. Id. The district court found that Gerber enjoys unparalleled brand recognition with a brand loyalty greater than any other product sold in the United States. Id. at 193. Gerber’s products are found in over 90 per cent of all American supermarkets. 2 By contrast, Heinz is sold in approximately 40 per cent of all supermarkets. Its sales are nationwide but concentrated *712 in northern New England, the Southeast and Deep South and the Midwest. Id. at 194. Despite its second-place domestic market share, Heinz is the largest producer of baby food in the world with $1 billion in sales worldwide. Its domestic baby food products with annual net sales of $103 million are manufactured at its Pittsburgh, Pennsylvania plant, which was updated in 1991 at a cost of $120 million. Id. at 192-93. The plant operates at 40 per cent of its production capacity and produces 12 million cases of baby food annually. Its baby food line includes about 130 SKUs (stock keeping units), that is, product varieties (e.g., strained carrots, apple sauce, etc.). Heinz lacks Gerber’s brand recognition; it markets itself as a “value brand” with a shelf price several cents below Gerber’s. Id. at 193. Beech-Nut has a market share (15.4%) comparable to that of Heinz (17.4%), with $138.7 million in annual sales of baby food, of which 72 per cent is jarred baby food. Its jarred baby food line consists of 128 SKUs. Beech-Nut manufactures all of its baby food in Canajoharie, New York at a manufacturing plant that was built in 1907 and began manufacturing baby food in 1931. Beech-Nut maintains price parity with Gerber, selling at about one penny less. It markets its product as a premium brand. Id. Consumers generally view its product as comparable in quality to Gerber’s. Id. Beech-Nut is carried in approximately 45 per cent of all grocery stores. Although its sales are nationwide, they are concentrated in New York, New Jersey, California and Florida. 3 Id. at 194. At the wholesale level Heinz and BeechNut both make lump-sum payments called “fixed trade spending” (also known as “slotting fees” or “pay-to-stay” arrangements) to grocery stores to obtain shelf placement. Id. at 197. Gerber, with its strong name recognition and brand loyalty, does not make such pay-to-stay payments. The other type of wholesale trade spending is “variable trade spending,” which typically consists of manufacturers’ discounts and allowances to supermarkets to create retail price differentials that entice the consumer to purchase their product instead of a competitor’s. Id. Under the terms of their merger agreement, Heinz would acquire 100 per cent of Beech-Nut’s voting securities for $185 million. Accordingly, they filed a Premer-ger Notification and Report Form with the FTC and the United States Department of Justice pursuant to the Hart-Scott-Rodino Antitrust Improvement Act of 1976, 15 U.S.C. § 18a. 4 On July 7, 2000 the FTC authorized this action for a preliminary injunction under section 13(b) of the FTCA and, on July 14, 2000, it filed a complaint and motion for preliminary injunction. The district court conducted a five-day hearing in late August and early September and heard final arguments on September 21, 2000. 'The record before the district court consisted of 1,267 exhibits, including 150 demonstrative exhibits, 32 depositions and 41 affida *713 vits. In addition, eleven witnesses testified. On October 18, 2000 the district court denied preliminary injunctive relief. The court concluded that it was “more probable than not that consummation of the Heinz/Beech-Nut merger will actually increase competition in jarred baby food in the United States.” H.J. Heinz, 116 F.Supp.2d at 200. The FTC appealed and sought injunctive relief pending appeal, which this court granted on November 8, 2000. On November 22, 2000 the FTC filed an administrative complaint against Heinz and Beech-Nut, charging that the proposed merger violates section 5 of the FTCA and, if consummated, would violate section 7 of the Clayton Act. In the Matter of H. J. Heinz, Docket No. 9295 (filed Nov. 22, 2000). II. Analysis A. Standard of Review We review a district court order denying preliminary injunctive relief for abuse of discretion, National Wildlife Fed’n v. Burford, 835 F.2d 305, 319 (D.C.Cir.1987), and will set aside the court’s factual findings only if they are “clearly erroneous.” Fed.R.Civ.P. 52(a); United States v. Marine Bancorporation, Inc., 418 U.S. 602, 615 n. 13, 94 S.Ct. 2856, 41 L.Ed.2d 978 (1974). If our review of the district court order “reveals that it rests on an erroneous premise as to the pertinent law, however, we must examine the decision in light of the legal principles we believe proper and sound.” Ambach v. Bell, 686 F.2d 974, 979 (D.C.Cir.1982). We apply die novo, review to the district court’s conclusions of law. See FTC v. National Tea Co., 603 F.2d 694, 696-98 (8th Cir.1979) (reviewing de novo proper standard of proof under section 13(b) of FTCA); cf. FTC v. Warner Communications Inc., 742 F.2d 1156, 1160 (9th Cir.1984) (per curiam) (finding as matter of law district court applied incorrect standard for section 7 violation). In deciding whether to grant preliminary injunctive relief under section 13(b), the court evaluates whether it is in the public interest to enjoin the proposed merger. See 15 U.S.C. § 53(b). B. Section 7 of the Clayton Act Section 7 of the Clayton Act prohibits acquisitions, including mergers, “where in any line of commerce or in any activity affecting commerce in any section of the country, the effect of such acquisition may be substantially to lessen competition, or to tend to create a monopoly.” 15 U.S.C. § 18; see United States v. Philadelphia Nat’l Bank, 374 U.S. 321, 355, 83 S.Ct. 1715, 10 L.Ed.2d 915 (1963) (“The statutory test is whether the effect of the merger ‘may be substantially to lessen competition’ ‘in any line of commerce in any section of the country.’”). The “Congress used the words ‘may be substantially to lessen competition’ (emphasis supplied), to indicate that its concern was with probabilities, not certainties.” Brown Shoe Co. v. United States, 370 U.S. 294, 323, 82 S.Ct. 1502, 8 L.Ed.2d 510 (1962) (emphasis original); see S.Rep. No. 1775, at 6 (1950), U.S.Code Cong. & Admin. News at 4293, 4298 (“The use of these words [“may be”] means that the bill, if enacted, would not apply to the mere possibility but only to the reasonable probability of the prescribed effect....”). “Merger enforcement, like other areas of antitrust, is directed at market power. It shares with the law of monopolization a degree of schizophrenia: an aversion to potent power that heightens risk of abuse; and tolerance of that degree of power required to attain economic benefits.” Lawrence A. Sullivan & Warren S. Grimes, The Law of Antitrust § 9.1, at 511 (2000). The Congress has empowered the FTC, inter alia, to weed out those mergers whose effect “may be substantially to lessen competition” from those that enhance competition. See H.R.Rep. No. 1142, at 18-19 (1914). *714 In section 13(b) of the FTCA, the Congress provided a mechanism whereby the FTC may seek preliminary injunctive relief preventing the merging parties from consummating the merger until the Commission has had an opportunity to investigate and, if necessary, adjudicate the matter. C. Section 13(b) of the Federal Trade Commission Act “Whenever the Commission has reason to believe that a corporation is violating, or is about to violate, Section 7 of the Clayton Act, the FTC may seek a preliminary injunction to prevent a merger pending the Commission’s administrative adjudication of the merger’s legality.” FTC v. Staples, Inc., 970 F.Supp. 1066, 1070 (D.D.C.1997); see 15 U.S.C. § 53(b). Section 13(b) provides for the grant of a preliminary injunction where such action would be in the public interest — as determined by a weighing of the equities and a consideration of the Commission’s likelihood of success on the merits. 15 U.S.C. § 53(b). 5 The Congress intended this standard to depart from what it regarded as the then-traditional equity standard, which it characterized as requiring the plaintiff to show: (1) irreparable damage, (2) probability of success on the merits and (3) a balance of equities favoring the plaintiff. H.R.Rep. No. 93-624, at 31 (1971). The Congress determined that the traditional standard was not “appropriate for the implementation of a Federal statute by an independent regulatory agency where the standards of the public interest measure the propriety and the need for injunc-tive relief.” Id. “The courts had evolved an approach to cases in which government agencies, acting to enforce a federal statute, sought interim relief. The agency, in such cases, was not held to the high thresholds applicable where private parties seek interim restraining orders.” FTC v. Weyerhaeuser Co., 665 F.2d 1072, 1082 (D.C.Cir.1981); see FTC v. Exxon Corp., 636 F.2d 1336, 1343 (D.C.Cir.1980) (“In enacting [Section 13(b) ], Congress further demonstrated its concern that injunctive relief be broadly available to the FTC by incorporating a unique ‘public interest’ standard in 15 U.S.C. § 53(b), rather than the more stringent, traditional ‘equity’ standard for injunctive relief.”). The FTC is not required to establish that the proposed merger would in fact violate section 7 of the Clayton Act. Staples, 970 F.Supp. at 1071; see FTC v. Food Town Stores, Inc., 539 F.2d 1339, 1342 (4th Cir.1976) (“The district court is not authorized to determine whether the antitrust laws have been or are about to be violated. That adjudicatory function is vested in the FTC in the first instance.”). We now consider the FTC’s likelihood of success and weigh the equities. Accord FTC v. Freeman Hosp., 69 F.3d 260, 267 (8th Cir.1995); FTC v. University Health, Inc., 938 F.2d 1206, 1217 (11th Cir.1991); Warner Communications, 742 F.2d at 1160. 1. Likelihood of Success To determine likelihood of success on the merits we measure the probability that, after an administrative hearing on the merits, the Commission will succeed in proving that the effect of the Heinz/BeechNut merger “may be substantially to lessen competition, or to tend to create a monopoly” in violation of section 7 of the Clayton Act. 15 U.S.C. § 18. This court and others have suggested that the standard for likelihood of success on the merits is met if the FTC “has raised questions going to the merits so serious, substantial, *715 difficult and doubtful as to make them fair ground for thorough investigation, study, deliberation and determination by the FTC in the first instance and ultimately by the Court of Appeals.” FTC v. Beatrice Foods Co., 587 F.2d 1225, 1229 (D.C.Cir.1978) (Appendix to Statement of Mac-Kinnon & Robb, JJ.) 6 ; Staples, 970 F.Supp. at 1071; Warner Communications, 742 F.2d at 1162 (quoting National Tea, 603 F.2d at 698); see University Health, 938 F.2d at 1218. This specific standard was articulated by the court below, see H.J. Heinz, 116 F.Supp.2d at 194, and it is a standard to which the appellees have not objected. In United States v. Baker Hughes Inc., 908 F.2d 981, 982-83 (D.C.Cir.1990), we explained the analytical approach by which the government establishes a section 7 violation. First the government must show that the merger would produce “a firm controlling an undue percentage share of the relevant market, and [would] result[] in a significant increase in the concentration of firms in that market.” Philadelphia Nat’l Bank, 374 U.S. at 363, 83 S.Ct. 1715. Such a showing establishes a “presumption” that the merger will substantially lessen competition. See Baker Hughes, 908 F.2d at 982. To rebut the presumption, the defendants must produce evidence that “show[s] that the market-share statistics [give] an inaccurate account of the [merger’s] probable effects on competition” in the relevant market. United States v. Citizens & S. Nat’l Bank, 422 U.S. 86, 120, 95 S.Ct. 2099, 45 L.Ed.2d 41 (1975). 7 “If the defendant successfully rebuts the presumption [of illegality], the burden of producing additional evidence of anticom-petitive effect shifts to the government, and merges with the ultimate burden of persuasion, which remains with the government at all times.” Baker Hughes Inc., 908 F.2d at 983; see also Kaiser Aluminum, 652 F.2d at 1340 & n. 12. Although Baker Hughes was decided at the merits stage as opposed to the preliminary injunctive relief stage, we can nonetheless use its analytical approach in evaluating the Commission’s showing of likelihood of success. Accordingly, we look at the FTC’s prima facie case and the defendants’ rebuttal evidence. a. Prima Facie Case Merger law “rests upon the theory that, where rivals are few, firms will be able to coordinate their behavior, either by overt collusion or implicit understanding, in order to restrict output and achieve profits above competitive levels.” FTC v. PPG Indus., 798 F.2d 1500, 1503 (D.C.Cir.1986). 8 Increases in concentration above certain levels are thought to “raise[ ] a likelihood of ‘interdependent anticompeti- *716 tive conduct.’ ” Id. (quoting General Dynamics, 415 U.S. at 497, 94 S.Ct. 1186); see FTC v. Elders Grain, 868 F.2d 901, 905 (7th Cir.1989). Market concentration, or the lack thereof, is often measured by the Herfindahl-Hirschmann Index (HHI). See Staples, 970 F.Supp. at 1081 n. 12. 9 Sufficiently large HHI figures establish the FTC’s prima facie case that a merger is anti-competitive. See Baker Hughes, 908 F.2d at 982-83 & n. 3; PPG, 798 F.2d at 1503. The district court found that the pre-merger HHI “score for the baby food industry is 4775” — indicative of a highly concentrated industry. 10 H.J. Heinz, 116 F.Supp.2d at 196; see PPG, 798 F.2d at 1503; Horizontal Merger Guidelines, supra, § 1.51. The merger of Heinz and Beech-Nut will increase the HHI by 510 points. This creates, by a wide margin, a presumption that the merger will lessen competition in the domestic jarred baby food market. See Horizontal Merger Guidelines, supra, § 1.51 (stating that HHI increase of more than 100 points, where post-merger HHI exceeds 1800, is “presumed ... likely to create or enhance market power or facilitate its exercise”); see also Baker Hughes, 908 F.2d at 982-83 & n. 3; PPG, 798 F.2d at 1503. 11 Here, *717 the FTC’s market concentration statistics 12 are bolstered by the indisputable fact that the merger will eliminate competition between the two merging parties at the wholesale level, where they are currently the only competitors for what the district court described as the “second position on the supermarket shelves.” H.J. Heinz, 116 F.Supp.2d at 196. Heinz’s own documents recognize the wholesale competition and anticipate that the merger will end it. JA 2680; see also JA 2185. Indeed, those documents disclose that Heinz considered three options to end the vigorous wholesale competition with Beech-Nut: two involved innovative measures while the third entailed the acquisition of Beech-Nut. JA 2184. Heinz chose the third, and least pro-competitive, of the options. Finally, the anticompetitive effect of the merger is further enhanced by high barriers to market entry. 13 The district court found that there had been no significant entries in the baby food market in decades and that new entry was “difficult and improbable.” H.J. Heinz, 116 F.Supp.2d at 196. This finding largely eliminates the possibility that the reduced competition caused by the merger will be ameliorated by new competition from outsiders and further strengthens the FTC’s case. See University Health, 938 F.2d at 1219 & n. 26. As far as we can determine, no court has ever approved a merger to duopoly under similar circumstances. *718 b. Rebuttal Arguments In response to the FTC’s prima facie showing, the appellees make three rebuttal arguments, which the district court accepted in reaching its conclusion that the merger was not likely to lessen competition substantially. For the reasons discussed below, these arguments fail and thus were not a proper basis for denying the FTC injunctive relief. 1. Extent of Pre-Merger Competition The appellees first contend, and the district court agreed, that Heinz and BeechNut do not really compete against each other at the retail level. Consumers do not regard the products of the two companies as substitutes, the appellees claim, and generally only one of the two brands is available on any given store’s shelves. Hence, they argue, there is little competitive loss from the merger. This argument has a number of flaws which render clearly erroneous the court’s finding that Heinz and Beech-Nut have not engaged in significant pre-merger competition. First, in accepting the appel-lees’ argument that Heinz and Beech-Nut do not compete, the district court failed to address the record evidence that the two do in fact price against each other, see, e.g., 8/31/2000 Tr. 247-48, and that, where both are present in the same areas, 14 they depress each other’s prices as well as those of Gerber even though they are virtually never all found in the same store. See, e.g., 8/30/2000 Tr. 147-48, 172; PX 531 at ¶ 8; PX 481 at ¶ 12; PX 479 at ¶ ¶ 6-7; PX 478 at ¶ 6; DX 14 at RP-110. This evidence undermines the district court’s factual finding. Second, the district court’s finding is inconsistent with its conclusion that there is a single, national market for jarred baby food in the United States. The Supreme Court has explained that “[t]he outer boundaries of a product market are determined by the reasonable interchangeability of use [by consumers] or the cross-elasticity of demand between the product itself and substitutes for it.” Brown Shoe, 370 U.S. at 325, 82 S.Ct. 1502; see also United States v. E.I. du Pont de Nemours & Co., 351 U.S. 377, 395, 76 S.Ct. 994, 100 L.Ed. 1264 (1956). 15 The definition of product market thus “focuses solely on demand substitution factors,” i.e., that consumers regard the products as substitutes. Horizontal Merger Guidelines, supra, § 1.0; Sullivan & Grimes, supra, § 11.2bl, at 579. By defining the relevant product market generically as jarred baby food, the district court concluded that in areas where Heinz’s and Beech-Nut’s products are both sold, consumers will switch between them in response to a “small but significant and nontransitory increase in price (SSNIP).” Horizontal Merger Guidelines, supra, § 1.11; H.J. Heinz, 116 F.Supp.2d at 195. The district court never explained this inherent inconsistency in its logic nor could counsel for the appellees explain it at oral argument. Third, and perhaps most important, the court’s conclusion concerning pre-merger competition does not take into account the indisputable fact that the merger will eliminate competition at ’ the wholesale level between the only two competitors for the *719 “second shelf’ position. Competition between Heinz and Beech-Nut to gain accounts at the wholesale level is fierce with each contest concluding in a winner-take-all result. JA 2680. The district court regarded this loss of competition as irrelevant because the FTC did not establish to its satisfaction that wholesale competition ultimately benefited consumers through lower retail prices. The district court concluded that fixed trade spending did not affect consumer prices and that “the FTC’s assertion that the proposed merger will affect variable trade spending levels and consumer prices is ... at best, inconclusive.” 16 H.J. Heinz, 116 F.Supp.2d at 197. Although the court noted the FTC’s examples of consumer benefit through cou-poning initiatives, the court held that it was “impossible to conclude with any certainty that the consumer benefit from such couponing initiatives would be lost in the merger.” Id. In rejecting the FTC’s argument regarding the loss of wholesale competition, the court committed two legal errors. First, as the appellees conceded at oral argument, no court has ever held that a reduction in competition for wholesale purchasers is not relevant unless the plaintiff can prove impact at the consumer level. Oral Arg. Tr. at 22, 28; see Hospital Corp. of Am. v. FTC, 807 F.2d 1381, 1389 (7th Cir.1986) (“Section 7 does not require proof that a merger or other acquisition has caused higher prices in the affected market. All that is necessary is that the merger create • an appreciable danger of [collusive practices] in the future. A predictive judgment, necessarily probabilistic and judgmental rather than demonstrable, is called for.”) (citation omitted). Second, it is, in any event, not the FTC’s burden to prove such an impact with “certainty.” To the contrary, the antitrust laws assume that a retailer faced with an increase in the cost of one of its inventory items “will try so far as competition allows to pass that cost on to its customers in the form of a higher price for its product.” In re Brand Name Prescription Drugs Antitrust Litig., 123 F.3d 599, 605 (7th Cir.1997), reh’g and suggestion for reh’g en banc denied (Oct. 8, 1997). Section 7 is, after all, concerned with probabilities, not certainties. United States v. El Paso Natural Gas Co., 376 U.S. 651, 658, 84 S.Ct. 1044, 12 L.Ed.2d 12 (1964); Brown Shoe, 370 U.S. at 323, 82 S.Ct. 1502; Baker Hughes, 908 F.2d at 984. 17 *720 2. Post-Merger Efficiencies The appellees’ second attempt to rebut the FTC’s prima facie showing is their contention that the anticompetitive effects of the merger will be offset by efficiencies resulting from the union of the two companies, efficiencies which they assert will be used to compete more effectively against Gerber. It is true that a merger’s primary benefit to the economy is its potential to generate efficiencies. See generally 4A Phillip E. Areeda, Herbert Hovenkamp & John L. Solow, Antitrust Law ¶ 970 at 22-25 (1998). As the Merger Guidelines now recognize, efficiencies “can enhance the merged firm’s ability and incentive to compete, which may result in lower prices, improved quality, or new products.” Horizontal Merger Guidelines, supra, § 4. Although the Supreme Court has not sanctioned the use of the efficiencies defense in a section 7 case, see Procter & Gamble Co., 386 U.S. at 580, 87 S.Ct. 1224, 18 the trend among lower courts is to recognize the defense. See, e.g., FTC v. Tenet Health Care Corp., 186 F.3d 1045, 1054 (8th Cir.1999), reh’g and reh’g en banc denied (Oct. 6, 1999); University Health, 938 F.2d at 1222; FTC v. Cardinal Health, Inc., 12 F.Supp.2d 34, 61 (D.D.C.1998); Staples, 970 F.Supp. at 1088-89; see also ABA Antitrust Section, Mergers and Acquisitions: Understanding the Antitrust Issues 152 (2000) (“The majority of courts have considered efficiencies as a means to rebut the government’s prima facie case that a merger will lead to restricted output or increased prices. These courts, however, generally have found inadequate proof of efficiencies to sustain a rebuttal of the government’s ease.”). In 1997 the Department of Justice and the FTC revised their Horizontal Merger Guidelines to recognize that “mergers have the potential to generate significant efficiencies by permitting a better utilization of existing assets, enabling the combined firm to achieve lower costs in producing a given quantity and quality than either firm could have achieved without the proposed transaction.” Horizontal Merger Guidelines, supra, § 4. Nevertheless, the high market concentration levels present in this case require, in rebuttal, proof of extraordinary efficiencies, which the appellees failed to supply. See University Health, 938 F.2d at 1223 (“[A] defendant who seeks to overcome a presumption that a proposed acquisition would substantially lessen competition must demonstrate that the intended acquisition would result in significant economies and that these economies ultimately would benefit competition and, hence, consumers.”); Horizontal Merger Guidelines, supra, § 4 (stating that “[e]fficiencies almost never justify a merger to monopoly or near-monopoly”); 4A Areeda, et al, Antitrust Law ¶ 971f, at 44 (requiring “extraordinary” efficiencies where the “HHI is well above 1800 and the HHI increase is *721 well above 100”). Moreover, given the high concentration levels, the court must undertake a rigorous analysis of the kinds of efficiencies being urged by the parties in order to ensure that those “efficiencies” represent more than mere speculation and promises about post-merger behavior. The district court did not undertake that analysis here. In support of its conclusion that post-merger efficiencies will outweigh the merger’s anticompetitive effects, the district court found that the consolidation of baby food production in Heinz’s underutilized Pittsburgh plant “will achieve substantial cost savings in salaries and operating costs.” H.J. Heinz, 116 F.Supp.2d at 199. The court also credited the appellees’ promise of improved product quality as a result of recipe consolidation. 19 The only cost reduction the court quantified as a percentage of pre-merger costs, however, was the so called “variable conversion cost”: the cost of processing the volume of baby food now processed by Beech-Nut. The court accepted the appellees’ claim that this cost would be reduced by 43% if the Beech-Nut production were shifted to Heinz’s plant, see JA 4619, a reduction the appellees’ expert characterized as “extraordinary.” The district court’s analysis falls short of the findings necessary for a successful efficiencies defense in the circumstances of this case. We mention only three of the most important deficiencies here. First, “variable conversion cost” is only a percentage of the total variable manufacturing cost. A large percentage reduction in only a small portion of the company’s overall variable manufacturing cost does not necessarily translate into a significant cost advantage to the merger. Thus, for cost reduction to be relevant, we must at least consider the percentage of Beech-Nut’s total variable manufacturing cost that would be reduced as a consequence of the merger. At oral argument, the appellees’ counsel agreed. Oral Arg. Tr. at 43. This correction immediately cuts the asserted efficiency gain in half since, according to the appellees’ evidence, using total variable manufacturing cost as the measure cuts the cost savings from 43% to 22.3%. See JA 4620. Second, the percentage reduction in Beech-Nut’s cost is still not the relevant figure. After the merger, the two entities will be combined, and to determine whether the merged entity will be a significantly more efficient competitor, cost reductions must be measured across the new entity’s combined production — not just across the pre-merger output of Beech-Nut. See 4A Areeda, et al, supra, ¶ 976d at 93-94. The district court, however, did not consider the cost reduction over the merged firm’s combined output. At oral argument the appellees’ counsel was unable to suggest a formula that could be used for determining that cost reduction. See Oral Arg. Tr. at 45-47. Finally, and as the district court recognized, the asserted efficiencies must be “merger-specific” to be cognizable as a defense. 20 H.J. Heinz, 116 F.Supp.2d at *722 198-99; see Horizontal Merger Guidelines, supra, § 4; 4A Areeda, et al., supra, ¶ 973, at 49-62. That is, they must be efficiencies that cannot be achieved by either company alone because, if they can, the merger’s asserted benefits can be achieved without the concomitant loss of a competí-' tor. See generally 4A Areeda, et al., supra, ¶ 973. Yet the district court never explained why Heinz could not achieve the kind of efficiencies urged without merger. As noted, the principal merger benefit asserted for Heinz is the acquisition of Beech-Nut’s better recipes, which will allegedly make its product more attractive and permit expanded sales at prices lower than those charged by Beech-Nut, which produces at an inefficient plant. Yet, neither the district court nor the appellees addressed the question whether Heinz could obtain the benefit of better recipes by investing more money in product development and promotion — say, by an amount less than the amount Heinz would spend to acquire Beech-Nut. At oral argument, Heinz’s counsel agreed that the taste of Heinz’s products was not so bad that no amount of money could improve the brand’s consumer appeal. Oral Arg. Tr. at 54. That being the case, the question is how much Heinz would have to spend to make its product equivalent to the BeechNut product and hence whether Heinz could achieve the efficiencies of merger without eliminating Beech-Nut as a competitor. The district court, however, undertook no inquiry in this regard. In short, the district court failed to make the kind of factual determinations necessary to render the appellees’ efficiency defense sufficiently concrete to offset the FTC’s prima facie showing. 3. Innovation The appellees claim next that the merger is required to enable Heinz to innovate, and thus to improve its competitive position against Gerber. Heinz and BeechNut asserted, and the district court found, that without the merger the two firms are unable to launch new products to compete with Gerber because they lack a sufficient shelf presence or ACV. See H.J. Heinz, 116 F.Supp.2d at 199-200. This kind of defense is often a speculative proposition. See 4A Areeda, et al., supra, ¶ 975g (noting “truly formidable” proof problems in determining innovation economies). In this case, given the old-economy nature of the industry as well as Heinz’s position as the world’s largest baby food manufacturer, it is a particularly difficult defense to prove. The court below accepted the ap-pellees’ argument principally on the basis of their expert’s testimony that new product launches are cost-effective only when a firm’s ACV is 70% or greater (Heinz’s is presently 40%; Beech-Nut’s is 45%). *723 That testimony, in turn, was based on a graph that plotted revenue against ACV. According to the expert, the graph showed that only four out of 27 new products launched in 1995 had been successful — all for companies with an ACV of 70% or greater. The chart, however, does not establish this proposition and the court’s consequent finding that the merger is necessary for innovation is thus unsupported and clearly erroneous. All the chart plotted was revenue against ACV and hence all it showed was the unsurprising fact that the greater a company’s ACV, the greater the revenue it received. Because the graph did not plot the profitability (or any measure of “cost-effectiveness”), there is no way to know whether the expert’s claim — that a 70% ACV is required for a launch to be “successful” in an economic sense — is true. 21 Moreover, the number of data points on the chart were few; they were limited to launches in a single year; and they involved launches of all new grocery products rather than of baby food alone. Assessing such data’s statistical significance in establishing the proposition at issue, ie., the necessity of 70% ACV penetration, is thus highly speculative. The district court did not even address the question of the data’s statistical significance and the appellees’ counsel could offer no help at oral argument. See Oral Arg. Tr. at 39 (“I’m not aware of the statistical significance of the underlying study.”). 22 In the absence of reliable and significant evidence that the merger will permit innovation that otherwise could not be accomplished, the district court had no basis to conclude that the FTC’s showing was rebutted by an innovation defense. Moreover, Heinz’s insistence on a 70-plus ACV before it brings a new product to market may be largely to persuade the court to recognize promotional economies as a defense. Heinz argues that to profitably launch a new product, it must have nationwide market penetration to recoup the money spent on advertising and promotion. It wants to spread advertising costs out among as many product units as possible, thereby lowering the advertising cost per unit. It does not want to “waste” promotional expenditures in markets where its products are not on the shelf or where they are on only a few shelves. For example, in a metropolitan area in which Heinz has a 75 per cent ACV, every dollar spent on advertising is two or three times more “effective” than in a market in which it has only a 25 per cent ACV. As one authority notes, however, “[t]he case for recognizing a defense based on promotional economies is relatively weak.” 4A Aree-da, et al., supra, ¶ 975f, at 77. The district court accepted Heinz’s claim that it could not introduce new products without at least a 70 per cent ACV because it would *724 be unable to adequately diffuse its advertising and promotional expenditures. But the court failed to determine whether substantial promotional scale economies exist now and, if they do, whether Heinz and Beech-Nut “for that reason operate at a substantial competitive disadvantage in the market or markets in which they sell” or whether there are effective alternatives to merger by which the disadvantage can be overcome. Id. at ¶ 975Í2, at 78. 4. Structural Barriers to Collusion In a footnote the district court dismissed the likelihood of collusion derived from the FTC’s market concentration data. “[Sjtructural market barriers to collusion” in the retail market for jarred baby food, the court said, rebut the normal presumption that increases in concentration will increase the likelihood of tacit collusion. H.J. Heinz, 116 F.Supp.2d at 198 n. 7. The court’s sole citation, however, was to testimony by the appellees’ expert, Jonathan B. Baker, a former Director of the Bureau of Economics at the FTC, who testified that in order to coordinate successfully, firms must solve “cartel problems” such as reaching a consensus on price and market share and deterring each other from deviating from that consensus by either lowering price or increasing production. He opined that after the merger the merged entity would want to expand its market share at Gerber’s expense, thereby decreasing the likelihood of consensus on price and market -share. 9/8/2000 Tr. 1010-1013. In his report, Baker elaborated on his theory, explaining that the efficiencies created by the merger will give the merged firm the ability and incentive to take on Gerber in price and product improvements. DX 617. He also predicted that policing and monitoring of any agreement would be more difficult than it is now, due in part to a time lag in the ability of one firm to detect price cuts by another. But the district court made no finding that any of these “cartel problems” are so much greater in the baby food industry than in other industries -that they rebut the normal presumption. In fact, Baker’s testimony about “time lag” is refuted by the record which reflects that supermarket prices are available from industry-wide scanner data within 4-8 weeks. See DX 617 at ¶ 86 (report of appellees’ expert Jonathan Baker); see also Oral Arg. Tr. at 30 (statement by appellees’ counsel that nothing in record reflects time lag is greater in baby food industry than in other industries). His testimony is further undermined by the record evidence of past price leadership in the baby food industry. 23 The combination of a concentrated market and barriers to entry is a recipe for price coordination. See University Health, 938 F.2d at 1218 n. 24 (“Significant market concentration makes it ‘easier for firms in the market to collude, expressly or tacitly, and thereby force price above or farther above the competitive level.’ ” (citation omitted)). “[WJhere rivals are few, firms will be able to coordinate their be *725 havior, either by overt collusion or implicit understanding, in order to restrict output and achieve profits above competitive levels.” PPG, 798 F.2d at 1503. The creation of a durable duopoly affords both the opportunity and incentive for both firms to coordinate to increase prices. The district court recognized this when it questioned Baker on whether the merged entity will, up to a point, expand its market share but “then [with Gerber will] find a nice equilibrium and they’ll all get along together.” 9/8/2000 Tr. 1014. Tacit coordination is feared by antitrust policy even more than express collusion, for tacit coordination, even when observed, cannot easily be controlled directly by the antitrust laws. It is a central object of merger policy to obstruct the creation or reinforcement by merger of such oligopolistic market structures in which tacit coordination can occur. 4 Phillip E. Areeda, Herbert Hovenkamp & John L. Solow, Antitrust Law ¶ 901b2, at 9 (rev. ed.1998). Because the district court failed to specify any “structural market barriers to collusion” that are unique to the baby food industry, its conclusion that the ordinary presumption of collusion in a merger to duopoly was rebutted is clearly erroneous. 24 ❖ * * # H* Although we recognize that, post-hearing, the FTC may accept the rebuttal arguments proffered by the appellees, including their efficiencies defense, and permit the merger to proceed, we conclude that the FTC succeeded in “raising] questions going to the merits so serious, substantial, difficult and doubtful as to make them fair ground for thorough investigation, study, deliberation and determination by the FTC.” Warner Communications, 742 F.2d at 1162. The FTC demonstrated that the merger to duopoly will increase the concentration in an already highly concentrated market; that entry barriers in the market make it unlikely that any anticompetitive effects will be avoided; that pre-merger competition is vigorous at the wholesale level nationwide and present at the retail level in some metropolitan areas; and that post-merger competition may be lessened substantially. These substantial questions have not been sufficiently answered by the appellees. As we said in Baker Hughes, “[t]he more compelling the pri-ma facie case, the more evidence the defendant must present to rebut it successfully.” 908 F.2d at 991. In concluding that the FTC failed to make the requisite showing, the district court erred in a number of respects. Regarding the contention of lack of pre-merger competition, it made a clearly erroneous factual finding and misunderstood the law with respect to the import of competition at the wholesale level. Regarding the proffered efficiencies defense, the court failed to make the kind of factual findings required to render that defense sufficiently concrete to rebut the government’s prima facie showing. Finally, as to the contention that the merger is necessary for innovation, the court clearly erred in relying on evidence that does not support its conclusion. Because the district court incorrectly assessed the merits of the ap-pellees’ rebuttal arguments, it improperly discounted the FTC’s showing of likelihood of success. *726 2. Weighing of the Equities Although the FTC’s showing of likelihood of success creates a presumption in favor of preliminary injunctive relief, we must still weigh the equities in order to decide whether enjoining the merger would be in the public interest. 15 U.S.C. § 53(b); see PPG, 798 F.2d at 1507; Weyerhaeuser, 665 F.2d at 1081-83. The principal public equity weighing in favor of issuance of preliminary injunctive relief is the public interest in effective enforcement of the antitrust laws. University Health, 938 F.2d at 1225. The Congress specifically had this public equity consideration in mind when it enacted section 13(b). See Food Town Stores, 539 F.2d at 1346 (Congress enacted section 13(b) to preserve status quo until FTC can perform its function). The district court found, and there is no dispute, that if the merger were allowed to proceed, subsequent administrative and judicial proceedings on the merits “will not matter” because Beech-Nut’s manufacturing facility “will be closed, the Beech-Nut distribution channels will be closed, the new label and recipes will be in place, and it will be impossible as a practical matter to undo the transaction.” H.J. Heinz, 116 F.Supp.2d at 201. Hence, if the merger were ultimately found to violate the Clayton Act, it would be impossible to recreate pre-merger competition. See Warner Communications, 742 F.2d at 1165 (“A denial of a preliminary injunction would preclude effective relief if the Commission ultimately prevails and divestiture is ordered.”). Section 13(b) itself embodies congressional recognition of the fact that divestiture is an inadequate and unsatisfactory remedy in a merger case, 119 Cong. Rec. 36612 (1973), a point that has been emphasized by the United States Supreme Court. See, e.g., FTC v. Dean Foods Co., 384 U.S. 597, 606 n. 5, 86 S.Ct. 1738, 16 L.Ed.2d 802 (1966) (“Administrative experience shows that the Commission’s inability to unscramble merged assets frequently prevents entry of an effective order of divestiture.”). On the other side of the ledger, the appellees claim that the injunction would deny consumers the proeompetitive advantages of the merger. See FTC v. Pharm-tech Research, Inc., 576 F.Supp. 294, 299 (D.D.C.1983) (explaining that public equities include “beneficial economic effects and proeompetitive advantages for consumers”). The district court found that if the merger were preliminarily enjoined, the injury to competition would also be irreversible, that is, the merger would be abandoned and could not be consummated if ultimately found lawful. By contrast to its first finding, however, for the latter conclusion the court relied not on the facts of this case but on our statement in Exxon that — as a general matter — temporarily blocking a tender offer is likely to end an attempted acquisition, “as a result of the short life-span of most tender offers.” Id. (quoting Exxon, 636 F.2d at 1343). In their brief in this court, the appellees offer nothing more to support the finding that the merger would never be consummated were an injunction to issue. Indeed, they devote only a single sentence, without any citation, to the point. The district court’s finding that an injunction would “kill this merger” is thus not a factual finding supported by record evidence. This case does not involve a short-lived tender offer as did the case cited by the court for its “kill the merger” conclusion. The appellees acknowledge that there is no alternative buyer for Beech-Nut and the court found that it is not a failing company but rather a “profitable and ongoing enterprise.” H.J. Heinz, 116 F.Supp.2d at 201 n. 9. If the merger makes economic sense now, the appellees have offered no reason why it would not do so later. Moreover, BeechNut’s principal assets of value to Heinz are, assertedly, its recipes and brand *727 name. Nothing in the record leads us to believe that both will not still exist when the FTC completes its work. It may be that Beech-Nut will have to sell its recipes to Heinz at a lower price than the price of today’s merger. But that is at best a “private” equity which does not affect our analysis of the impact on the market of the two options now before us and which has not in any event been urged by the appel-lees. 25 See id. In sum, weighing of the equities favors the FTC. If the merger is ultimately found to violate section 7 of the Clayton Act, it will be too late to preserve competition if no preliminary injunction has issued. On the other hand, if the merger is found not to lessen competition substantially, the efficiencies that the appellees urge can be reclaimed by a renewed transaction. Our conclusion with respect to the equities necessarily lightens the burden on the FTC to show likelihood of success on the merits, a burden which the FTC has met here. III. Conclusion It is important to emphasize the posture of this case. We do not decide whether the FTC will ultimately prove its case or whether the defendants’ claimed efficiencies will carry the day. 26 Our task is to review the district court’s order to determine whether, under section 13(b), preliminary injunctive relief would be in the public interest. We have considered the FTC’s likelihood of success on the merits. We have weighed the equities. We conclude that the FTC has raised serious and substantial questions. We also conclude that the public equities weigh in favor of preliminary injunctive relief and therefore that a preliminary injunction would be in the public interest. Accordingly, we reverse the district court’s denial of preliminary injunctive relief and remand the case for entry of a preliminary injunction pursuant to section 13(b) of the Federal Trade Commission Act. So ordered. Footnotes [1] The facts as set forth herein are based on the district court's factual findings and the record material submitted by the parties. [2] Product volume in retail stores throughout the country is measured by the product's All Commodity Volume (ACV). Gerber's near 100 per cent ACV is impressive because virtually all supermarkets stock at most two brands of baby food. In at least one area of the country as many as 80 per cent of supermarket retailers stock only Gerber. [3] Although Heinz and Beech-Nut introduced evidence showing that in areas that account for 80% of Beech-Nut sales, Heinz has a market share of about 2% and in areas that account for about 72% of Heinz sales, BeechNut's share is about 4%, the FTC introduced evidence that Heinz and Beech-Nut are locked in an intense battle at the wholesale level to gain (and maintain) position as the second brand on retail shelves. [4] Section 18a requires pre-merger notification for a merger in which either the acquiring or the acquired firm has total net sales or assets of at least $10 million and the other firm has annual sales or total assets of at least $100 million. The acquirer must have at least 15 per cent or $15 million worth of the target firm’s voting securities or assets. 15 U.S.C. § 18a(a). Filers must disclose specific financial and market data and pay a filing fee. [5] Section 13(b) of the FTCA provides that "[u]pon a proper showing that, weighing the equities and considering the Commission’s likelihood of ultimate success, such action would be in the public interest, ... a preliminary injunction may be granted." 15 U.S.C. § 53(b). [6] In Beatrice Foods, two members of the court, writing separately from a denial of en banc review, included the quoted language from an unpublished judgment and memorandum issued earlier in the litigation. [7] To rebut the defendants may rely on "[n]on-statistical evidence which casts doubt on the persuasive quality of the statistics to predict future anticompetitive consequences” such as "ease of entry into the market, the trend of the market either toward or away from concentration, and the continuation of active price competition.” Kaiser Aluminum & Chem. Corp. v. FTC, 652 F.2d 1324, 1341 (7th Cir.1981). In addition, the defendants may demonstrate unique economic circumstances that undermine the predictive value of the government’s statistics. See United States v. General Dynamics Corp., 415 U.S. 486, 506-10, 94 S.Ct. 1186, 39 L.Ed.2d 530 (1974) (fundamental changes in structure of coal market made market concentration statistics inaccurate predictors of anticompetitive effect); see also University Health, 938 F.2d at 1218. [8] A "horizontal merger” involves firms selling the same or similar products in a common geographical market. [9] "The FTC and the Department of Justice, as well as most economists, consider the measure superior to such cruder measures as the four-or eight-firm concentration ratios which merely sum up the market shares of the largest four or eight firms.” PPG, 798 F.2d at 1503. The Department of Justice and the FTC rely on the FIHI in evaluating proposed horizontal mergers. See United States Dep’t of Justice & Federal Trade Comm’n, Horizontal Merger Guidelines §§ 1.5, 1.51 (1992), as revised (1997). The HHI is calculated by totaling the squares of the market shares of every firm in the relevant market. For example, a market with ten firms having market shares of 20%, 17%, 13%, 12%, 10%, 10%, 8%, 5%, 3% and 2% has an HHI of 1304 (202 + 172 + 132 + 122 + i02 + 102 + 82 + 52 + 32 +22). If the firms with 13% and 5% market shares were to merge, the HHI would increase by 130 points, expressed by the formula 2ab, which is derived from (a+b)2 or a2 + 2ab + b2. Under the Merger Guidelines a market with a postmerger HHI above 1800 is considered "highly concentrated” and mergers that increase the HHI in such a market by over 50 points "potentially raise significant competitive concerns.” Id. at § 1.51. Mergers "producing an increase in the HHI of more than 100 points [in such markets] are [presumed] likely to create or enhance market power or facilitate its exercise.” Id. Although the Merger Guidelines are not binding on the court, they provide “a useful illustration of the application of the HHI.” PPG, 798 F.2d at 1503 n. 4. [10] To determine the HHI score the district court first had to define the relevant market. The court defined the product market as jarred baby food and the geographic market as the United States. HJ. Heinz, 116 F.Supp.2d at 195. The parties do not challenge the court's definition. [11] The FTC argues that this finding alone— that it is certain to establish a prima facie case — entitles it to preliminary injunctive relief under PPG. We disagree with the Com- ’ mission’s reading of PPG. In PPG, the Commission appealed the district court's denial of its request for a preliminary injunction to prevent PPG Industries, the world's largest producer of glass aircraft transparencies, from acquiring Swedlow, Inc., the world's largest manufacturer of acrylic aircraft transparencies. 798 F.2d at 1502. After defining the relevant market and determining market share, the district court found that the merger would significantly increase the concentration in an already highly concentrated market. It also "found high market-entry barriers that would prolong high market concentration.” Id. at 1503. On appeal, this court stated: “There is no doubt that the pre-and post-acquisition HHI’s and market shares found in this case entitle the Commission to some preliminary relief.” Id. This statement came, however, in the context of a case in which the appellants offered no rebuttal (other than the observation of rapid and continuing technological changes in the industry) to the presumption generated by the market concentration data on which the FTC based its prima facie showing. Id. at 1506. The court then noted the rule established in Weyerhaeuser that the FTC is entitled to a "presumption in *717 favor of a preliminary injunction when [it] establishes a strong likelihood of success on the merits.” Id. at 1507. [12] The Supreme Court has cautioned that statistics reflecting market share and concentration, while of great significance, are not conclusive indicators of anticompetitive effects. See General Dynamics, 415 U.S. at 498, 94 S.Ct. 1186; Brown Shoe, 370 U.S. at 322 n. 38, 82 S.Ct. 1502 ("Statistics reflecting the shares of the market controlled by the industry leaders and the parties to the merger are, of course, the primary index of market power; but only a further examination of the particular market — its structure, history and probable future — can provide the appropriate setting for judging the probable anticompetitive effect of the merger.”). In General Dynamics the Supreme Court held that the market share statistics the government used to seek divestiture of the merged firm were insufficient because, in failing to take into account the acquired firm’s long-term contractual commitments (coal contracts), the statistics overestimated the acquired firm’s ability to compete in the relevant market in the future. General Dynamics, 415 U.S. at 500-504, 94 S.Ct. 1186. [13] Barriers to entry are important in evaluating whether market concentration statistics accurately reflect the pre- and likely postmer-ger competitive picture. Cf. Baker Hughes, 908 F.2d at 987. If entiy barriers are low, the threat of outside entry can significantly alter the anticompetitive effects of the merger by deterring the remaining entities from colluding or exercising market power. See United States v. Falstaff Brewing Corp., 410 U.S. 526, 532-33, 93 S.Ct. 1096, 35 L.Ed.2d 475 (1973); Baker Hughes, 908 F.2d at 987 ("In the absence of significant barriers, a company probably cannot maintain supracompetitive pricing for any length of time.”); Horizontal Merger Guidelines, supra, § 3.0 ("A merger is not likely to create or enhance market power or to facilitate its exercise, if entry into the market is so easy that market participants, after the merger, either collectively or unilaterally could not profitably maintain a price increase above premerger levels.”). Low barriers to entry enable a potential competitor to deter anticompetitive behavior by firms within the market simply by its ability to enter the market. FTC v. Procter & Gamble Co., 386 U.S. 568, 581, 87 S.Ct. 1224, 18 L.Ed.2d 303 (1967) ("It is clear that the existence of Procter at the edge of the industry exerted considerable influence on the market.”). Existing firms know that if they collude or exercise market power to charge supracompetitive prices, entry by firms currently not competing in the market becomes likely, thereby increasing the pressure on them to act competitively. See Baker Hughes, 908 F.2d at 988; Byars v. Bluff City News Co., 609 F.2d 843, 851 n. 19 (6th Cir.1979). [14] There are at least ten metropolitan areas in which Heinz and Beech-Nut both have more than a 10 per cent market share and their combined share exceeds 35 per cent. PX 781 at Ex. IB. [15] Interchangeability of use and cross-elasticity of demand look to the availability of products that are similar in nature or use and the degree to which buyers are willing to substitute those similar products for one another. See E.I. du Pont de Nemours, 351 U.S. at 393, 76 S.Ct. 994. [16] Fixed trade spending consists of "slotting fees,” "pay-to-stay” arrangements, new store allowances and other payments to retailers in exchange for shelf space and desired product display. H.J. Heinz, 116 F.Supp.2d at 197. Variable trade spending includes payments to retailers tied to sales volume and intended to insure a specific sales volume and lower shelf price. Id. [17] Although the merger's effects on the wholesale market for baby food are important to a determination of whether the merger is likely to reduce competition in the baby food market overall, we reject the FTC’s argument here that the "wholesale competition” between Heinz and Beech-Nut is an entirely distinct "line of commerce” within the meaning of section 7 of the Clayton Act such that it must be analyzed independently from "retail competition.” The Congress amended section 7 in 1950 “to make the measure of anti-competitive acquisitions the extent to which they lessened competition ‘in any line of commerce,' rather than the extent to which they lessened competition 'between' the two companies.” Citizen Publishing Co. v. United States, 394 U.S. 131, 137 n. 3, 89 S.Ct. 927, 22 L.Ed.2d 148 (1969). Courts interpret "line of commerce” as synonymous with the relevant product market. See General Dynamics, 415 U.S. at 510, 94 S.Ct. 1186; Falstaff Brewing, 410 U.S. at 531-32, 93 S.Ct. 1096. The district court defined only one market — jarred baby food sold throughout the line of commerce in the United States. Thus, the proper “line of commerce” for analysis in this case is the overall market for jarred baby food, which includes both retail and wholesale levels. At this point in the proceedings, *720 the wholesale market cannot be separated out for analysis without regard to the merger’s effect on other levels of competition. [18] In Procter & Gamble Co., 386 U.S. at 580, 87 S.Ct. 1224, the Supreme Court stated that "[p]ossible economies cannot be used as a defense to illegality” in section 7 merger cases. The issue is, however, not a closed book. See Staples, 970 F.Supp. at 1088 (collecting cases). Areeda and Turner explain that “[i]n interpreting the Clorox language, moreover, observe that the court referred only to possible’ economies and to economies that 'may' result from mergers that lessen competition. To reject an economies defense based on mere possibilities does not mean that one should reject such a defense based on more convincing proof.” 4 Phillip Areeda & Donald Turner, Antitrust Law ¶ 941b, at 154 (1980). They conclude that "[t]he Court’s brief and unelaborated language [in Clorox\ cannot reasonably be taken as a definitive disposition of so important and complex an issue as the role of economies in analyzing legality of a merger.” Id. [19] In addition, the district court described Heinz’s distribution network as much more efficient than Beech-Nut’s. H.J. Heinz, 116 F.Supp.2d at 199. It failed to find, however, a significant diseconomy of scale in distribution from which either Heinz or Beech-Nut suffers. 4A Areeda, et al., supra, ¶ 975el, at 73. In other words, although Beech-Nut has an inefficient distribution system, it can make that system more efficient without merger. Heinz’s own efficient distribution network illustrates that a firm the size of Beech-Nut does not need to merge in order to attain an efficient distribution system. [20] The Horizontal Merger Guidelines explain that "merging firms must substantiate efficiency claims so that the Agency can verify by reasonable means the likelihood and magnitude of each asserted efficiency, how and when each would be achieved (and any costs of doing so), how each would enhance the *722 merged firm's ability and incentive to compete, and why each would be merger-specific. Efficiency claims will not be considered if they arc vague or speculative or otherwise cannot be verified by reasonable means.” Horizontal Merger Guidelines, supra, § 4. Regarding the types of efficiencies asserted here, the Guidelines state: The Agency has found that certain types of efficiencies are more likely to be cognizable and substantial than others. For example, efficiencies resulting from shifting production among facilities formerly owned separately, which enable the merging firms to reduce the marginal cost of production, are more likely to be susceptible to verification, merger-specific, and substantial, and are less likely to result from anticompetitive reductions in output. Other efficiencies, such as those relating to research and development, are potentially substantial but are generally less .susceptible to verification and may be the result of anticompetitive output reductions. Yet others, such as those relating to procurement, management, or capital cost are less likely to be merger-specific or substantial, or may not be cognizable for other reasons. Id. [21] For example, a 5 cent piece of bubble gum introduced with a 90% ACV could appear as a failure on the graph because of low revenue but nonetheless be profitable. On the other hand, a high priced grocery product introduced with the same ACV could generate a lot of revenue (and thus appear as a “success” on the graph) yet be unprofitable. [22] The graph evidence is also not useful unless we know the “sunk” costs in bringing the product to market and the manufacturer's fixed and variable costs in producing the product. Sunk costs are costs that have already been incurred such as research and development and promotional expenses, including brand name development. See Henry N. Butler, Economic Analysis for Lawyers 935 (1998). Fixed costs refer to those expenses that do not vary with output and will be incurred as long as the firm continues in business. Variable costs are those that change with the rate of output such as wages paid to workers and payments for raw materials. See id. at 920, 936; E. Thomas Sullivan & Jeffrey L. Harrison, Understanding Antitrust and its Economic Implications 19-21 (3d ed.1998). [23] In an oligopolistic market characterized by few producers, price leadership occurs when firms engage in interdependent pricing, setting their prices at a profit-maximizing, su-pracompetitive level by recognizing their shared economic interests with respect to price and output decisions. See Brooke Group Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209, 227, 113 S.Ct. 2578, 125 L.Ed.2d 168 (1993); see also Jesse W. Markham, The Nature and Significance of Price Leadership, 41 Amer. Econ. Rev. 891 (1951); Richard A. Posner, Oligopoly and the Antitrust Laws: A Suggested Approach, 21 Stan. L.Rev. 1562, 1582 (1969); Donald Arthur Washburn, Price Leadership, 64 Va. L.Rev. 691, 693-697 (1978). In a duopoly, a market with only two competitors, supracompetitive pricing at monopolistic levels is a danger. See Edward Hastings Chamberlin, The Theory of Monopolistic Competition: A Re-orientation of the Theory of Value 46-55 (8th ed.1962). [24] Contrary to the appellees' claims, nothing in Baker Hughes suggests otherwise. In that case, the sophisticated nature of the purchasers of the industry’s product and the "volatile and shifting” nature of each firm’s market share rendered the HHI figures an unreliable measure of concentration. See 908 F.2d at 986-87. No such circumstances exist in this case. [25] The district court noted that "[t]he parties have not stressed private equities” but the court nonetheless considered them. It concluded that while "the corporate interests of Heinz and Milnot and especially the interests of Dearborn Capital Partners LP, which presumably acquired Milnot through a leveraged buyout with the purpose and intent of selling its interest at a profit” were important to the private parties, they should not affect the outcome of the proceeding. H.J. Heinz, 116 F.Supp.2d at 200 n. 9. We agree. "While it is proper to consider private equities in deciding whether to enjoin a particular transaction, we must afford such concerns little weight, lest we undermine section 13(b)’s purpose of protecting the 'public-at-large, rather than the individual private competitors.’ ” University Health, 938 F.2d at 1225 (citation omitted); cf. Weyerhaeuser, 665 F.2d at 1083 ("Private equities do not outweigh effective enforcement of the antitrust laws. When the Commission demonstrates a likelihood of ultimate success, a countershowing of private equities alone would not suffice to justify denial of a preliminary injunction barring the merger.”). [26] “The most difficult mergers to assess may be those that combine both negative and positive effects; creating market power that increases the risk of oligopolistic pricing while at the same time creating efficiencies that reduce production or marketing costs.” Sullivan & Grimes, supra, § 9.1, at 511.