United States v. United States Sugar Corporation

PRECEDENTIAL UNITED STATES COURT OF APPEALS FOR THE THIRD CIRCUIT _______________ No. 22-2806 _______________ UNITED STATES OF AMERICA, Appellant v. UNITED STATES SUGAR CORPORATION; IMPERIAL SUGAR COMPANY; LOUIS DREYFUS COMPANY LLC; UNITED SUGARS CORPORATION _______________ On Appeal from the United States District Court for the District of Delaware (D.C. Civil No. 1:21-cv-01644) District Judge: Honorable Maryellen Noreika _______________ Argued: January 18, 2023 Before: AMBRO*, PORTER, and FREEMAN, Circuit Judges. * Judge Ambro assumed senior status on February 6, 2023. (Filed: July 13, 2023) _______________ Melissa Arbus Sherry [ARGUED] Amanda P. Reeves Lindsey S. Champlin David L. Johnson Charles S. Dameron Latham & Watkins LLP 555 Eleventh Street, NW Suite 1000 Washington, DC 20004 Lawrence E. Buterman Latham & Watkins LLP 1271 Avenue of the Americas New York, NY 10020 Christopher S. Yates Latham & Watkins LLP 505 Montgomery Street Suite 2000 San Francisco, CA 94111 Jack B. Blumenfeld Brian P. Egan Morris, Nichols, Arshit & Tunnell LLP 1201 North Market Street P.O. Box 1347 Wilmington, DE 19899 Counsel for Defendant-Appellee United States Sugar Corp. 2 Timothy G. Cameron Peter T. Barbur David R. Marriott Daniel K. Zach Michael K. Zaken Lindsey J. Timlin Hannah L. Dwyer Cravath, Swaine & Moore LLP Worldwide Plaza 825 Eighth Avenue New York, NY 10019 Amanda L. Wait Norton Rose Fulbright US LLP 799 9th Street, NW Suite 1000 Washington, DC 20001 Kelly E. Farnan Richards, Layton & Finger, P.A. 920 N. King Street Wilmington, DE 19801 Counsel for Defendant-Appellee Imperial Sugar Company and Louis Dreyfus Company LLC Peter J. Schwingler Stinson LLP 50 South Sixth Street Suite 2600 Minneapolis, MN 55402 3 Daniel K. Hogan Hogan McDaniel 1311 Delaware Avenue Wilmington, DE 19806 Counsel for Defendant-Appellee United Sugar Corporation Jonathan S. Kanter Doha Mekki Maggie Goodlander David B. Lawrence Daniel E. Haar Nikolai G. Levin Peter M. Bozzo [ARGUED] U.S. Department of Justice Antitrust Division 950 Pennsylvania Avenue, NW Room 3224 Washington, DC 20530-0001 Brian Hanna Jonathan Y. Mincer U.S. Department of Justice Antitrust Division 450 5th Street, NW Suite 800 Washington, DC 20530-0001 Counsel for Plaintiff-Appellant United States Lee Hepner 140 San Carlos Street 4 San Francisco, CA 94110 Katherine Van Dyck American Economic Liberties Project 2001 Pennsylvania Avenue NW Suite 540 Washington, DC 20006 Counsel for Amicus Appellant American Economic Liberties Project ______________ OPINION OF THE COURT ______________ PORTER, Circuit Judge. The government appeals the denial of its motion to permanently enjoin the acquisition of Imperial Sugar by United States Sugar Corporation. The District Court found that the government failed to identify the relevant product and geographic markets and thus failed to establish a prima facie case under Section 7 of the Clayton Act. 15 U.S.C. § 18. It concluded that the government overlooked the procompetitive effects of distributors in the market for refined sugar, erroneously lumped together heterogeneous wholesale customers, and defined the relevant geographic market without regard for the high mobility of sugar throughout the country. Because the District Court’s rejection of the government’s proposed product market is not clearly erroneous, we will affirm. I 5 Georgia-based Imperial Sugar Company has been in financial distress for years. It went bankrupt in 2001 and suffered a costly accident at its plant in 2008, prompting its owners to put it up for sale. Purchased by the Louis Dreyfus Company, Imperial has since received from Louis Dreyfus only a subsistence level of investment to keep its operation safe and environmentally sound. Imperial’s internal reports describe it as an “import-based, price-uncompetitive sugar refinery” that is “structurally uncompetitive” and suffers from a shrinking customer base, losing roughly ten percent of its customers from 2021 to 2022. For more than five years, Louis Dreyfus has been trying to sell it. Enter United States Sugar Corporation, a large Florida- based sugar refiner that agreed to purchase Imperial. The government contends that U.S. Sugar’s acquisition should be blocked because it would have anticompetitive effects in the market for refined sugar. The government alleges that the transaction would leave only two entities in control of 75% of refined sugar sales in the southeastern United States. It proffers an application of the hypothetical monopolist test (HMT) and argues that the results of that test demonstrate the validity of its proposed product and geographic markets. U.S. Sugar answers first that it does not even sell its own sugar but rather participates with three other producers in a Capper-Volstead agricultural cooperative, United Sugar, that markets and sells the firms’ output collectively but exercises no control over the quantities that its members produce.1 Even if operated at capacity, it argues, Imperial’s facility could 1 Capper-Volstead cooperatives are agricultural cooperatives exempted from certain antitrust scrutiny. See 7 U.S.C. §§ 291– 92. 6 produce only about seven percent of national output—an insufficient share either to invoke a per se presumption of anticompetitiveness or to merit an injunction under a rule-of- reason analysis. Second, U.S. Sugar argues that sugar distributors constitute a crucial competitive check on producer- refiners that would undermine any attempt to increase prices. This effect, it argues, goes unappreciated in the government’s HMT analysis and undermines the government’s product market definition. Finally, it argues, evidence of the high mobility of refined sugar throughout the country renders the government’s proposed regional markets vacuous and unrepresentative. After an expedited trial, the District Court denied the government’s plea for an injunction. It determined the credibility of the expert witnesses and carefully weighed the evidence. As to product market definition, the Court concluded that U.S. Sugar was right to extoll the effects of sugar distributors, who account for approximately 25% of sales of refined sugar in the U.S. It rejected the government’s proposed product market, concluding that any proposed product market must include sales of refined sugar sold by distributors if it is to be relevant. Turning to the proposed geographic market, the Court recounted considerable evidence presented at trial of sugar’s high geographic mobility and the ease with which producers and distributors could avail themselves of arbitrage by selling to out-of-region buyers. It concluded that the government’s analysis failed to account for this mobility, making its proposed markets too narrow to be relevant. The government timely appealed. II 7 The District Court had jurisdiction under 15 U.S.C. § 25 (“district courts of the United States are invested with jurisdiction to prevent and restrain violations of [the Clayton] Act”). It entered final judgment on September 23, 2022. The government filed a notice of appeal on September 26, 2022. This Court has jurisdiction under 28 U.S.C. § 1291. On appeal from a Rule 52 ruling, we review “findings of fact for clear error” and “conclusions of law de novo.” See FTC v. Penn State Hershey Med. Ctr., 838 F.3d 327, 335 (3d Cir. 2016) (“Hershey”).2 As for the specific issues on appeal, “the determination of a relevant market is composed of the articulation of a legal test which is then applied to the factual circumstances of each case.” Id. (quoting White & White, Inc. v. Am. Hosp. Supply Corp., 723 F.2d 495, 499 (6th Cir. 1983)). Thus, “while a district court’s conclusion concerning what constitutes the relevant market is subject to the clearly erroneous standard of review, the district court’s formulation of the market tests may be freely reviewed on appeal as a matter of law.” Worldwide Basketball & Sport Tours, Inc. v. Nat’l Collegiate Athletic Ass’n., 388 F.3d 955, 960 (6th Cir. 2004). A court’s conclusion concerning what constitutes the relevant market is a finding of fact that is “clearly erroneous” only if it is “completely devoid of minimum evidentiary support displaying some hue of credibility or bears no rational relationship to the supportive evidentiary data.” Berg Chilling Sys., Inc. v. Hull Corp., 369 F.3d 745, 754 (3d Cir. 2004). “[S]o 2 U.S. Sugar suggests, that, by arguing only in terms of plenary review, the government has forfeited a clear error challenge. But “[a] party cannot waive, concede, or abandon the applicable standard of review,” so we reject this contention. United States v. Escobar, 866 F.3d 333, 339 (5th Cir. 2017) (per curiam). 8 we review for clear error.” FTC v. Hackensack Meridian Health, Inc., 30 F.4th 160, 167 (3d Cir. 2022).3 However, “where a district court applies an incomplete economic analysis or an erroneous economic theory to those facts that make up the relevant geographic market, it has committed legal error subject to plenary review.” Hershey, 838 F.3d at 336. III A. The District Court did not clearly err in rejecting the government’s product market definition. 1. The relevant product market is the market for refined sugar. A claim arising under the Clayton Act, § 7, is evaluated under a three-part burden-shifting framework. Hackensack, 30 F.4th at 166. First, the government must establish a prima facie case that the merger is anticompetitive. Id. To do so, it must “propose the proper relevant market and . . . show that the effect of the merger in that market is likely to be anticompetitive.” Id. Second, the burden to produce evidence 3 Though the government characterizes the District Court’s product market holding as “legal error” and urges us to review the question de novo, to do so would create inconsistency: market definition inquiries such as this rely heavily upon the testimony of expert witnesses who are prohibited from rendering legal opinions. M.S. ex rel. Hall v. Susquehanna Twp. Sch. Dist., 969 F.3d 120, 129 (3d Cir. 2020) (noting that “an expert cannot testify to [a] legal conclusion”). Therefore, the history of Section 7 litigation and reliance upon expert witnesses necessitates an understanding of product market definition as a factual inquiry. 9 to rebut the government’s prima facie case shifts to the defendant. See, e.g., United States v. Baker Hughes Inc., 908 F.2d 981, 982–83 (D.C. Cir. 1990). Third, “[i]f the defendant successfully rebuts the presumption, the burden of producing additional evidence of anticompetitive effect shifts to the government, and merges with the ultimate burden of persuasion, which remains with the government at all times.” Id. at 983. This appeal concerns only the first prong of the first part of that analysis: identification of the relevant market. The government argues that, in defining a product market under Section 7, the District Court clearly erred by treating distributors as separate sources of refined sugar capable of undercutting efforts by a hypothetical monopolist to restrict output and increase price. The government contends that the HMT, the test “commonly used in antitrust actions to define the relevant market,” FTC v. Sanford Health, 926 F.3d 959, 963 (8th Cir. 2019), requires courts to adhere to a stratified model of the market in which refiners sell to distributors, which then sell to wholesalers, which then sell to retailers, which then sell to consumers. In this abstract, stratified model, a distributor would have no source of refined sugar beyond the refiners in its (properly defined) geographic market. Therefore, the government argues, a court cannot take cognizance of distributors or alternate sugar sources as potential checks on refiners’ market power. This case is somewhat atypical among product-market- definition disputes, as it is not a dispute about defining the product. Notably, in United States v. E.I. du Pont de Nemours & Co., the Supreme Court was pressed to answer whether 10 DuPont’s patent on cellophane created a monopoly or whether the relevant product market should be more broadly conceived of as “flexible wrapping materials,” of which cellophane was just one kind. 351 U.S. at 396–400. Similarly, in Erie Sand & Gravel Co. v. FTC, we were asked to decide whether “lake sand” used to make concrete was its own product or whether sand from a pit or sand from a bank could fairly be included with it. 291 F.2d 279, 281 (3d Cir. 1961). These cases typify the standard product-market-definition inquiry. Here, all are agreed: the product is refined sugar. The dispute is over defining the product market as either that for the sale of refined sugar or for the “production and sale” of refined sugar, Appellant’s Br. 2, the latter of which excludes parties who sell but do not produce—i.e., distributors and wholesalers. The government contends that a proper application of the HMT requires us to limit our focus to only those firms that both produce and sell refined sugar and to exclude sellers who do not themselves refine sugar. It reasons that even if, arguendo, distributors can bring sugar to market from other regions of the country or from overseas in response to higher local prices, all refined sugar must begin with a refiner, so under the HMT, distributors are customers, not suppliers, and should be treated as such under a proper application of the test. It therefore maintains that the District Court erred in considering distributors who could counteract monopolistic restrictions by releasing their own supplies. U.S. Sugar, in turn, citing our decision in Allen-Myland Inc. v. IBM Corp., 33 F.3d 194 (3d Cir. 1994), argues that where, as a matter of fact, independent distributors in this industry already are and will remain competitors, acknowledging them as such is entirely consistent with our case law. There, we addressed allegations against IBM of 11 unlawful tying in violation of the Sherman Act. Id. at 200. In defining the relevant market, the question arose whether leases of computer mainframes competed with computer manufacturers’ sales to end users. Id. at 202. Reversing the district court, we held that to the extent that leasing companies deal in used, non- IBM mainframes that have not already been counted in the sales market, these machines belong in the relevant market for large-scale mainframe computers. Unlike IBM, there is no allegation that the manufacturers of these computers possess the market power to control prices, much less that they would do so in concert with IBM. When these computers are placed in service by leasing companies, they provide an alternative that limits IBM’s power in the market. Id. at 203 (footnotes omitted). Allen-Myland thus makes clear that resellers may serve as competitive checks on a seller-manufacturer. The government offers a different interpretation. In its telling, the Allen-Myland Court “reversed because manufacturers’ market shares would already include new large-scale mainframe computers sold by manufacturers to leasing companies,” and to add the products in again when end-users leased them from leasing companies would lead to double-counting “because the leasing companies themselves ‘do nothing to increase the supply of new machines.’” Appellant’s Br. 26 (quoting Allen- Myland, 33 F.3d at 202) (cleaned up). While the government’s quoted language from the Allen-Myland opinion is accurate, it lacks context. The Allen- Myland Court did say that leasing companies themselves did 12 “nothing to increase the supply of new machines” when the machines in question were IBM machines. 33 F.3d at 202. But “to the extent that leasing companies deal in used, non-IBM mainframes that have not already been counted in the sales market,” “they provide an alternative that limits IBM’s power in the market.” Id. at 203. So when the product is defined by its source, e.g., “IBM computers” or “U.S. Sugar refined sugar,” then aftermarket sales or leases of that product “do nothing to increase the supply” thereof. Id. at 202. But when an ostensible downstream party—a leasing company or wholesaler—has alternative sources for the same kind of product, e.g., computers or refined sugar, and places them in the stream of commerce, they impose a competitive check upon that source’s power in the market. And this is more likely to be true the more homogeneous the product: customers generally place less value upon the question of who manufactured the product when the product is a commodity, like sugar, rather than a branded piece of technology, like a computer. See George Stigler, A Theory of Oligopoly, 72 JOURN. POL. ECON. 44, 49 (1964) (“From the viewpoint of any one buyer . . . [t]he costs of shifting among suppliers will be smaller the more homogeneous the goods[.]”). The government’s focus on “production and sale” is a red herring. The proper product market definition here is the market for refined sugar, much as it was the market for “flexible packaging material” in E.I. du Pont, 351 U.S. at 400, or for concrete-grade sand in Erie Sand & Gravel, 291 F.2d at 281. The District Court noted that the “Government introduced no evidence at trial that purchasers care whether their sugar supplier is a refiner producer, a marketing entity, a cooperative or a distributor.” J.A. 36, ¶ 85. So defining the product market to include production and sale is irrelevant to consumer 13 welfare and a purely self-serving description by the government. 2. The District Court’s analysis, based in practical indicia, was valid. There is some ambiguity regarding the extent to which the District Court relied upon HMT analysis in making its decision. It mentioned the HMT only once, in the section of its opinion analyzing the government’s proposed geographic market. In defining the product market, it instead focused on the “practical indicia” of “industry or public recognition of the submarket as a separate economic entity, the product’s peculiar characteristics and uses, unique production facilities, distinct customers, distinct prices, sensitivity to price changes, and specialized vendors.” J.A. 46–47 (quoting Brown Shoe Co., Inc. v. United States, 370 U.S. 294, 325 (1962)). And it properly defined its inquiry as one of interchangeability and cross-elasticity in order “to recognize where competition exists.” J.A. 47. The District Court did not err by considering facts on the ground rather than relying upon HMT analysis. Our precedent makes this clear. Cf. Hershey, 838 F.3d at 345 (“We are not suggesting that the hypothetical monopolist test is the only test that the district courts may use.”). The government cites no authority for its contrary view that “the hypothetical monopolist test . . . governs market definition.” Appellant’s Br. 22 (emphasis added). The District Court permissibly considered the “highly factual issue” of cross-elasticity of demand and the “[s]pecial characteristics of the relevant industry [that] may influence market definition.” Tunis Bros. Co., Inc. v. Ford Motor Co., 952 F.2d 715, 723 (3d Cir. 1991). 14 That factual inquiry was well-founded and not clearly erroneous. The District Court found that “[d]istributors have the ability to purchase large quantities of refined sugar from many different sources, including foreign competitors, and this allows distributors to price resales competitively.” J.A. 33, ¶ 79. It also found that “[d]istributors account for approximately 25% of sales of refined sugar in the U.S.,” id. at 34, ¶ 80, and noted that “[a]t trial, there were many examples of customers purchasing large quantities of sugar from distributors,” id., ¶ 81, with distributors “compet[ing] for sales to wholesale customers of all sizes, including large industrial customers,” J.A. 35, ¶ 83. It found that “distributors are the primary importers of refined [sugar] imports,” tending “to purchase the majority of foreign-produced refined sugar imports” in the United States, J.A. 33–34, ¶ 79. And U.S. Sugar’s expert witness, Dr. Hill, testified “that distributors buy from a variety of sources, which gives them independence and the ability to compete with refiners in the market.” Id. Based on sufficient evidence and weighing the testimony of the parties’ expert witnesses, the District Court thus rejected, in this industry, the government’s preferred rigid hierarchy of refiners, distributors, wholesalers, retailers, and consumers, with each only buying from the source above it. The District Court’s factual findings are extensive and carefully noted. It considered the government’s proposed market, the objections thereto, and other factors on which it was free to rely to inform its view of the situation. The government would prefer that the HMT be deemed to “govern” the field at Stage I and that its own construction of the HMT be accepted without scrutiny. The law does not require the 15 District Court to do either, and its decision as to product- market definition evinces no clear error. 3. The District Court was not required to accept a market encompassing “widely divergent customers.” The government contends that the District Court committed “legal error” by “requiring” the government to “further subdivide the market and differentiate between refined sugar sales to industrial customers and refined sugar sales to retail customers.” Appellant’s Br. 27 (cleaned up). It proposed a product market in which wholesale customers include industrial food and beverage manufacturers, retailers, food service companies, and distributors. J.A. 37, ¶ 87. The District Court found that, in advocating for this proposed market, the government’s expert failed to differentiate between refined sugar sales to industrial customers and those to retail customers and that he “made no attempt to consider whether industrial consumers have the same competitive alternatives as other customers.” Id. “Defining a relevant product market is . . . a factual question,” and the District Court treated it as such. Polypore Intern., Inc. v. FTC, 686 F.3d 1208, 1217 (11th Cir. 2012). It looked to the facts, considered expert witness testimony, made credibility determinations, and concluded that the government witness was less credible and his testimony less helpful than that of the defendant’s witness. This is not “a matter of law,” as the government styles it, but a matter of fact. Appellant’s Br. 27; see SmithKline Corp. v. Eli Lilly & Co., 575 F.2d 1056, 1062 (3d Cir. 1978). 16 The government asserts that “the Supreme Court has repeatedly confirmed[] there is no legal requirement that plaintiffs divide a broadly defined market into submarkets.” Appellant’s Br. 27. Its references meant to support the argument are not on point. Appellant’s Br. 28 (citing Brown Shoe, 370 U.S. at 327; United States v. Phillipsburg Nat’l Bank & Tr. Co., 399 U.S. 350, 360 (1970); United States v. Greater Buffalo Press, Inc., 402 U.S. 549, 553 (1971); United States v. Cont’l Can Co., 378 U.S. 441, 457–58 (1964)). In Brown Shoe, the Court held that a district court was not “required to employ finer[] distinctions” or pursue “[f]urther division” of the market where such “division does not aid [the court] in analyzing the effects of [a] merger.” 370 U.S. at 327. To say that finer distinctions are not “required” does not mean that they are prohibited. Id. The standard is what “aid[s]” a district court in analyzing the facts. Id. And the government’s other cases stand for the mere proposition that a court should not let the existence of submarkets persuade it to “disregard a broader line of commerce that has economic significance.” Phillipsburg Nat’l Bank, 399 U.S. at 360. Here, the District Court did not disregard the broader line of commerce in refined sugar; it simply determined that distinguishing between industrial and retail sales more accurately described the reality of the market. Contrary to the government’s portrayal, the District Court did not make any categorical statements about a need to always subdivide markets. It simply recognized that when defining a market, courts may draw distinctions as necessary to understand a merger’s effects on consumers. That factual determination was not clearly erroneous. To establish its prima facie case, the government must propose the proper relevant market, “includ[ing] both a product 17 market and a geographic market.” Hackensack, 30 F.4th at 166. As it has failed to articulate a relevant product market, we do not consider its argument as to a proposed geographic market. B. The District Court’s consideration of USDA responses as a remedy for antitrust harm was error, but it is not material. The District Court offered one further argument against the government’s case: that “even if U.S. Sugar’s acquisition of Imperial were likely to have any anticompetitive effects, the Court believes that the USDA has the ability to counteract those effects” by “increas[ing] the amount of low- or no-duty sugar that can be imported into the U.S.” J.A. 63. “Doing so,” it reasoned, “would increase the available sugar for sale in the U.S., thereby bringing prices back down.” Id. In particular, the Court highlighted the USDA’s “discretionary ability to increase the amounts imported under the TRQ [tariff-rate quota] system and U.S.-Mexico Suspension Agreements in order to maintain reasonable prices.” Id. The government is correct to describe this line of reasoning as improper. “Repeals of the antitrust laws by implication from a regulatory statute are strongly disfavored, and have only been found in cases of plain repugnancy between the antitrust and regulatory provisions.” United States v. Phila. Nat’l Bank, 374 U.S. 321, 350 (1963) (citations omitted). A finding of implied repeal “can be justified only by a convincing showing of clear repugnancy between the antitrust laws” and an alternative “regulatory system.” United States v. Nat’l Ass’n of Sec. Dealers, 422 U.S. 694, 719 (1975). And in this case, there is no argument presented that any statutory provision immunizes the sugar industry against antitrust challenge. The government’s simultaneous efforts to keep sugar prices high 18 through USDA policy and to lower them through antitrust suits may seem contradictory, but it is not unlawful for the government to pursue contradictory aims, and price supports do not create immunity from antitrust. That said, as the District Court’s reasoning on USDA price supports did not alter the outcome of its opinion, reversing it would not salvage the government’s case. The Court’s analysis of market definition stands unaffected by those portions of its opinion on USDA policy. * * * Defining a relevant market depends in equal parts upon defining the product market and the geographic market, and a failure to do either is dispositive. The District Court concluded that the government failed to define a relevant product market. Its analysis is not clearly erroneous. We will affirm. 19