Supreme Auto Transport, LLC v. Arcelor Mittal USA, Inc.

• Court: Court of Appeals for the Seventh Circuit
• Decided: September 6, 2018
• Docket: 17-2910
• Status: Published

Opinion

In the

United States Court of Appeals For the Seventh Circuit ____________________ No. 17-2910 SUPREME AUTO TRANSPORT, LLC, et al., Plaintiffs-Appellants, v.

ARCELOR MITTAL USA, INC., et al., Defendants-Appellees. ____________________

Appeal from the United States District Court for the Northern District of Illinois, Eastern Division. No. 08 C 5468—Manish S. Shah, Judge. ____________________

ARGUED FEBRUARY 16, 2018 — DECIDED SEPTEMBER 6, 2018 ____________________

Before WOOD, Chief Judge, and KANNE and ROVNER, Circuit Judges. WOOD, Chief Judge. This putative class action rests on alle- gations of a massive antitrust conspiracy within the domestic steel industry. Plaintiffs, indirect purchasers of steel, assert that eight U.S. steel producers colluded to slash output in an effort to drive up the price of steel nationwide. Years after their initial complaint, however, the plaintiffs transformed their theory of liability. The original complaint, filed in 2008, 2 No. 17-2910

charged that plaintiffs overpaid for steel sheets, rods, and tub- ing manufactured by the defendants at their steel mills. Eight years later, the plaintiffs amended their complaint, asserting instead that they overpaid for end-use consumer goods, such as vehicles, washing machines, and refrigerators, that were manufactured by third parties using steel. This new product definition greatly expanded the potential scope of the class. The district court dismissed the suit for two reasons. First, it determined that plaintiffs’ amended complaint is time- barred because it redefines “steel products” in a way that gives rise to an entirely different, and exponentially larger, universe of plaintiffs. Second, in the alternative, the court held that the amended complaint does not plausibly plead a causal connection between the alleged antitrust conspiracy and plaintiffs’ own injuries. Plaintiffs now appeal both rulings. We affirm. I According to the allegations in plaintiffs’ amended com- plaint, which we accept as true for the purposes of this appeal, defendant Arcelor Mittal is the largest integrated producer of steel in the United States, controlling 20–25 percent of total domestic raw steel capacity. Defendant U.S. Steel is the sec- ond largest integrated producer, controlling about 16 percent of total domestic raw steel capacity. Defendant Nuncor is the nation’s largest mini-mill producer, controlling 21 percent of total domestic raw steel capacity. Other major domestic steel producers include defendants Gerdau Ameristeel (10 per- cent), AK Steel (5 percent), Steel Dynamics (4 percent), IPSCO (2.5 percent), and Commercial Metals (2 percent). The remain- ing 15–20 percent of the steel market is controlled by firms No. 17-2910 3

that are not named defendants and are not alleged to have en- gaged in any anticompetitive behavior. It can be difficult to organize and maintain a price-fixing cartel in a market with many competitors—including firms that are not party to the conspiracy—but that is what plain- tiffs allege happened here. They theorize that sometime in early 2005, Arcelor Mittal organized a scheme to “improve in- dustry discipline” by cutting raw steel output in order to drive up prices and reap supracompetitive profits. They point to a statement by a Mittal executive at an industry meeting in March 2005 criticizing the industry’s traditional business model on the grounds that it “ensured that most producers would cut price before reducing volume.” The executive then called on his competitors to coordinate supply cuts: he asked them directly to “respond to market fluctuations” by cutting production at “marginal facilities” to avoid a “race to the bot- tom” in steel prices. A series of industry meetings followed in early-to-mid 2005; at those meetings, multiple defendants al- legedly cautioned the industry against oversupplying the market and urged everyone to “adjust … operating levels” to preserve high prices. According to plaintiffs, the result of all this “urging” and “cautioning” was a series of major production cuts in mid- 2005. Arcelor Mittal closed five of its twelve U.S. integrated blast furnaces and reduced production at U.S. mills to 55 per- cent of total capacity. U.S. Steel reduced output from 90 per- cent capacity in the first quarter of 2005 to 75 percent capacity in the second quarter, and closed at least two of its twelve do- mestic blast furnaces. Nuncor similarly reduced its output from 96 percent of capacity in 2004 to 79 percent capacity in 4 No. 17-2910

the second quarter of 2005. The smaller defendants took com- parable measures. After these cuts took effect, the price of a steel sheet increased by about 25 percent. The supply cuts al- legedly remained in effect until mid-to-late 2007. Plaintiffs insist that there is no procompetitive explanation for this huge reduction in output. They allege that from early 2005 to late 2007, “annual domestic demand for steel far ex- ceeded the United States production capacity of Defendants and other domestic producers,” and that as a result, “there was a shortage of steel in the United States market” through- out the relevant period. (We can assume that there was such a shortage. We observe, however, that neither the original nor amended complaint discusses input or energy costs, or what effect these costs might have had on steel prices. This is a trou- bling omission, especially given the fact that a separate sec- tion of plaintiffs’ complaint admits that the cost of energy, in- puts, and many other commodities—including “aluminum, copper, precious metals, resins, … plastic, … zinc and nickel”—increased substantially during the class period. The rising cost of inputs would provide an obvious innocent ex- planation for the increase in steel prices. Cf. Bell Atlantic Corp. v. Twombly, 550 U.S. 544 (2007). We will not address this pos- sibility because the defendants did not raise it on appeal. Nor will we address the impact of steel imports on the market— an issue that also has not been discussed by the parties and that would take us down a deep rabbit-hole.) In the end, the 2005 production cuts and subsequent in- creases in steel price led to two major antitrust class actions. The first suit, which we call the “direct-purchaser” suit, was filed in early September 2008, with Standard Iron Works as the lead plaintiff. Standard Iron filed that suit on behalf of a No. 17-2910 5

putative class. See Complaint at 1, 6, Standard Iron Works v. Arcelor Mittal, No. 08-cv-5214 (N.D. Ill. Sept. 12, 2008), ECF No. 1. The complaint defined class members as those who purchased “steel products” directly from the defendants. In turn, it defined “steel products” as: [A]ll products derived from raw steel … including, but not limited to, steel sheet and coil products; galvanized sheet and other galvanized and/or coated steel prod- ucts; tin mill products; steel slabs and plates; steel beams, blooms, rails, and other structural shapes; steel billets, bars, and rods; steel pipe and other tubular products; and all other products derived from raw steel. Id. at 6. The second suit, which we call the “indirect-purchaser” suit, is the one that gave rise to this appeal. It was brought by Supreme Auto in late September 2008, just weeks after the di- rect-purchaser action. The original complaint did not include a formal class definition. It simply alleged that Supreme Auto was injured when it “purchased several items of steel tubing [at an inflated price] indirectly from one or more of the de- fendants … for end use” and indicated that Supreme Auto was bringing suit on behalf of “all others similarly situated.” Supreme Auto’s original complaint included a definition of the types of steel products at issue. Its description mirrored the language from the direct-purchaser suit: “Steel products” means any consumer steel product in- cluding but not limited to produced flat steel sheets and coils; galvanized steel products; tin mill products; steel plates; steel beams, rails and other structural 6 No. 17-2910

shapes; steel bars and rods; steel wire and wire rod; steel pipes and other tubular products; and a variety of other products derived from raw steel. Supreme Auto stated that it was seeking relief under federal antitrust law, as well as under the antitrust, consumer protec- tion, and unjust enrichment laws of 28 states. The district court deferred proceedings on class certification in the indi- rect-purchaser suit until the court in the direct-purchaser suit decided whether to certify a class. After several years of delay, in September 2015 the district court certified a class in the direct-purchaser case for the sole purpose of determining whether defendants engaged in a conspiracy in violation of federal antitrust laws. See Standard Iron Works, 2015 WL 5304629, at *12 (Sept. 9, 2015). The court denied certification on the issues of impact and damages, holding that common questions there would not predominate over individual ones. Id. The direct-purchaser suit was settled shortly thereafter. In March 2016, defendants moved to dismiss the indirect- purchaser complaint. They argued that Supreme Auto failed to allege any injuries outside of its home state (Michigan), that it failed to allege that the steel it purchased was manufactured by any one of the named defendants, and that Supreme Auto’s complaint therefore failed to state a plausible claim for relief. Supreme Auto did not respond to defendants’ motion to dismiss. Instead, Supreme Auto—which was, at the time, the only named plaintiff—filed an amended complaint adding No. 17-2910 7

15 new plaintiffs from 11 states1 who purchased “one or more steel products containing steel purchased from one or more of the Defendants during the Class period.” Relying on juris- diction pursuant to the Class Action Fairness Act, 28 U.S.C. § 1332(d)(2), which requires only minimal diversity and an amount in excess of $5 million (both satisfied here), the amended complaint dropped the federal antitrust claims and alleged only state-law antitrust injuries in 21 states,2 con- sumer-protection injuries in 19 states,3 and unjust enrichment injuries in 48 states and the District of Columbia. It also changed the definition of “steel products” from the definition that appeared in the original complaint to the following: “Steel products” means any consumer steel product for end use and not for resale, including clothes wash- ers, clothes dryers, refrigerators, freezers, dishwashers, microwave ovens[,] regular ovens, automobiles, semi- tractor trailers, farm and construction equipment, room air conditioner units, hot water heaters, snow blowers, barbeque grills, lawn mowers, and reinforc- ing bars used in patios, driveways, swimming pools and sidewalks.

1 Arizona, California, Florida, Iowa, Kansas, Michigan, Montana, New

York, North Carolina, South Dakota, and Tennessee. 2 Arizona, California, District of Columbia, Iowa, Kansas, Maine, Michigan, Minnesota, Mississippi, Nebraska, Nevada, New Mexico, New York, North Carolina, North Dakota, South Dakota, Tennessee, Utah, Ver- mont, West Virginia, and Wisconsin. 3 Alaska, Arkansas, California, Colorado, Delaware, District of Co- lumbia, Florida, Idaho, Maine, Massachusetts, Michigan, Montana, Ne- braska, Nevada, New Hampshire, New York, North Carolina, Vermont, and Wisconsin. 8 No. 17-2910

Supreme Auto further announced that it was abandoning claims related to its initial injury, which arose out of its pur- chase of $171 worth of steel tubing. Going forward, the firm proclaimed, it would be claiming injury solely based on its purchase of two semi-truck trailers, each priced at over $115,000. Defendants again moved to dismiss the complaint, repeat- ing their earlier argument that plaintiffs’ injuries were too re- mote to establish a right to recover under either federal or state antitrust laws. They also argued that the claims set forth in the amended complaint were time-barred because they al- leged a new set of injuries. The district court agreed with de- fendants in both respects and granted the motion to dismiss with prejudice. Plaintiffs then appealed. II We begin by addressing the statute of limitations. Plain- tiffs concede that their amended complaint, filed in April 2016, falls outside all relevant limitations periods. The longest period for any of plaintiffs’ claims is six years, meaning that their claims expired, at the latest, sometime in the late sum- mer of 2014—six years after the alleged conspiracy ended, and a year and a half before plaintiffs filed their amended complaint. Thus, the amended complaint is untimely unless the plaintiffs can show that their claims were tolled or that the amendments to their new complaint relate back to the original one. An amended pleading relates back to the original if “the amendment asserts a claim or defense that arose out of the conduct, transaction, or occurrence set out—or attempted to be set out—in the original pleading.” FED. R. CIV. P. No. 17-2910 9

15(c)(1)(B). The central inquiry under Rule 15(c) is whether the original complaint “gave the defendant enough notice of the nature and scope of the plaintiff’s claim that he shouldn’t have been surprised by the amplification of the allegations of the original complaint in the amended one.” Santamarina v. Sears, Roebuck & Co., 466 F.3d 570, 573 (7th Cir. 2006) (citing Tiller v. Atlantic Coast Line R.R. Co., 323 U.S. 574, 581 (1945)). Even “significant” changes to a complaint or to a class defini- tion can relate back so long as the defendant had fair notice of the substance of the new allegations from the outset. See Knudsen v. Liberty Mut. Ins. Co., 411 F.3d 805, 806 (7th Cir. 2005). The district court held that the plaintiffs’ amended com- plaint did not satisfy the fair-notice standard, because the transactions at issue in the first complaint (the indirect pur- chase of steel pipes, tubing, and sheets) were wholly distinct from the transactions at issue in the amended complaint (the purchase of washing machines, cars, swimming pools, and the like). We agree. The original complaint defined “steel products” by illustrating the meaning of that term through a list of examples, such as steel sheets, rods, and tubing. All of the examples listed in plaintiffs’ definition are mill output— that is to say, they are steel products manufactured at steel plants. No reasonable defendant, upon reading the original definition, would have imagined that plaintiffs were in fact suing over the thousands of end-use household and commer- cial goods manufactured by third parties—a reading so broad that it would transform nearly every person in the country into a potential class member. Plaintiffs fight this conclusion by emphasizing that the original definition of steel products included “any” product 10 No. 17-2910

“derived from raw steel.” They also insist that the list was in- tended to be “illustrative, not exhaustive.” These points do not save the day. The word “derived,” as used in the context of product manufacturing, is susceptible to multiple interpre- tations. To say that a product is derived from steel could mean simply that the product has some steel ingredients. It could just as easily mean that the product is “formed” in its entirety out of steel. See OXFORD ENGLISH DICTIONARY ONLINE, http://www.oed.com/view/Entry/50613?redirectedFrom=de- rive#eid (last visited Sept. 5, 2018). And while we do not doubt that the list in the original complaint was intended to be illustrative rather than exhaustive, we must ask ourselves what general category the list was designed to illustrate. Here, the items that appeared in the original complaint—steel sheets, tubes, pipes, and rods—illustrate a category along the lines of “steel products manufactured at steel mills.” It cannot fairly be read as illustrating the much broader category of “all consumer goods that include any steel component.” As the Supreme Court often reminds us, the purpose of illustrative lists is to indicate that a definition includes only those objects that are similar to the objects in the list. See Begay v. United States, 553 U.S. 137, 142 (2015). No object in Supreme Auto’s original list bore any resemblance to the end-use consumer goods listed in the amended complaint. If the original defini- tion of steel products was intended to be all-encompassing, as plaintiffs now argue, then “it is hard to see why [the com- plaint] would have needed to include the examples at all.” Id.; cf. Circuit City Stores, Inc. v. Adams, 532 U.S. 105, 114 (2001) (explaining that, under the interpretative canon ejusdem gene- ris, where general words follow specific words in an enumer- No. 17-2910 11

ation, the general words are limited to the shared characteris- tics of the specific words, even though the general words may ordinarily have a much broader meaning). What’s more, Supreme Auto gave no indication during the first seven years of litigation that its suit included any prod- ucts other than mill output. Not only did its complaint and other filings refer only to mill output, but the definition of “steel products” in plaintiffs’ complaint appears to be lifted in large part from the complaint in the direct-purchasers’ suit, which had been filed less than a month earlier. All parties knew that the direct-purchaser litigation was exclusively about mill output, and so the definitional similarity between the two complaints further buttressed the already-intuitive inference that Supreme Auto’s suit was also about products made by the defendants at their steel mills. The first hint of a shift in Supreme Auto’s litigation strat- egy did not come until late 2015, when Supreme Auto sent subpoenas to third-party retailers such as Whirlpool and John Deere asking about their use of steel in washing machines and lawnmowers. Defendants complained about this expansive discovery tactic when they moved to dismiss the original complaint. They protested that plaintiffs’ suit now purported to encompass “not only tubes but ‘any consumer steel prod- uct,’ from refrigerators to washing machines to other con- sumer products made of steel among other constituent mate- rials.” On appeal, plaintiffs assert that this statement—which defendants wrote in 2016—constitutes an “admission” that defendants knew all along that the suit was really about this vast array of products with some steel in their make-up. We are not persuaded. Our review of the record, as well as our interrogation of both parties during oral argument, reveals no 12 No. 17-2910

indication that defendants had notice of plaintiffs’ new claims at any time before the expiration of the limitations period. Be- cause the original complaint did not give defendants fair no- tice of the nature and scope of the claims set out in the amended complaint, the amendments do not relate back un- der Rule 15(c). We also see no basis for tolling the statute of limitations. Tolling is not available under American Pipe & Construction Co. v. Utah, 414 U.S. 538 (1974), which suspends the applicable statute of limitations in class action suits to all members of the purported class, because the plaintiffs named in the amended complaint were not “asserted members of the class” defined in the original complaint and their claims were not encom- passed by the original suit. See id. at 553. If there were any doubt about this (and we have none), the Supreme Court’s recent decision in China Agritech, Inc. v. Resh, 138 S. Ct. 1800 (2018)—to the effect that upon denial of certification, a puta- tive class member may not commence a new class action be- yond the time allowed by the applicable statute of limita- tions—also supports our reasoning. The substantial prejudice to defendants—who had no reason to think in 2008 that they should preserve evidence relating to the gargantuan number of consumer products now at issue—further counsels against applying any form of equitable tolling. III Given that plaintiffs’ amended complaint is time-barred, we could end our analysis now without reaching the proxi- mate causation issue. But since the district court ruled in the alternative, and we recognize that we do not necessarily have the last word, we think it prudent also to say a word about this basis for the district court’s decision. No. 17-2910 13

Proximate causation is an essential element that plaintiffs must prove in order to succeed on any of their claims. The purpose of the proximate causation requirement—in both an- titrust and tort law—is to avoid speculative recovery by re- quiring a direct relation between the plaintiff’s injury and the defendant’s behavior. Holmes v. Sec. Inv’r Prot. Corp., 503 U.S. 258, 268 (1992). In the antitrust context, the proximate causation require- ment in the past has been termed “antitrust standing,” even though it has nothing to do with a plaintiff’s standing to sue under Article III of the U.S. Constitution (which requires only but-for causation, an injury-in-fact, and redressability). Greater Rockford Energy & Tech. Corp. v. Shell Oil Co., 998 F.2d 391, 395 & n.7 (7th Cir. 1993); see also Associated Gen. Contrac- tors of Cal. Inc. v. Cal. State Council of Carpenters, 459 U.S. 519, 535–43 (1983) (providing a six-factor test for determining whether the requirements of proximate causation are satisfied in an antitrust case). The Supreme Court clarified in Lexmark International, Inc. v. Static Control Components, Inc., 572 U.S. 118 (2014), in the context of the Lanham Act, that the considera- tions often referred under the rubric of “prudential standing” are nothing more or less than substantive questions about the coverage of a statute. Id. at 127–28. We will thus similarly es- chew the term “antitrust standing” and speak only of the re- quirements to bring a case under the particular statutes in- volved. Along with the ordinary requirement of proximate causa- tion, federal antitrust law imposes additional limits on recov- ery in suits for treble damages under the Clayton Act, 15 U.S.C. § 15. One of the most significant is the direct-pur- chaser requirement announced in Illinois Brick Co. v. Illinois, 14 No. 17-2910

431 U.S. 720, 729–30 (1977), which held that only direct pur- chasers of a product can sue the supplier for damages. Allow- ing private suits by purchasers further down the supply chain, the Court held, would risk duplicative awards. But Supreme Auto and its co-plaintiffs are not suing under the Clayton Act or any other federal statute (at least, not any more); rather, they are suing under the antitrust laws of nearly two dozen states. While most states model their anti- trust statutes and jurisprudence on federal law, they are un- der no obligation to do so. California v. ARC Am. Corp., 490 U.S. 93, 102 (1989). Relevant to this appeal, 21 of the states where plaintiffs claim injuries have enacted statutes that do not in- corporate Illinois Brick’s strict direct-purchaser requirement.4 In these states, downstream purchasers who pay inflated prices for consumer goods are not automatically barred from bringing antitrust suits against upstream price fixers. And there’s more, plaintiffs say. They argue that the so- called “Illinois-Brick repealer states” have replaced one per se rule with another and now allow all indirect-purchaser suits to go forward, no matter how far removed the purchasers are from the production process and no matter how speculative the damages award would be. We do not read these state laws

4See Ariz. Rev. Stat. Ann. §§ 44-1401, -1408; Cal. Bus. & Prof. Code § 16750(a); D.C. Code Ann. § 28-4509; Iowa Code §§ 553.12, .4; Kan. Stat. Ann. § 50-161(b); Me. Rev. Stat. Ann. 10 § 1104(1); Mich. Compiled Laws § 445.778(2); Minn. Stat. § 325D.57; Miss. Code Ann. § 75-21-9; Neb. Rev. Stat. Ann § 59-821; Nev. Rev. Stat. Ann. § 598A.210; N.M. Stat. Ann. § 57- 1-3(A), (C); New York Gen. Bus. Law § 340(6); N.C. Gen. Stat. § 75-16; N.D. Cent. Code § 51-08.1-08(3); S.D. Codified Laws Ann. §§ 37-1-14.3, -33; Tenn. Code Ann. § 47-25-106; Utah Code Ann. § 76-10-3109(1)(a); Vt. Stat. Ann. tit. 9 § 2465(b); W. Va. Code R. § 142-9-2; Wis. Stat. § 133.18(1)(a). No. 17-2910 15

so expansively. It is one thing to say that a state is willing to allow someone other than a direct purchaser to have the op- portunity to shoulder the burden of showing proximate cau- sation; it is quite another thing to say that the state has thrown both the direct-purchaser rule and proximate causation out the window. There are many suits that satisfy ordinary principles of proximate causation but nevertheless would be barred under federal law by Illinois Brick’s direct-purchaser requirement. This very case provides an example: many if not all Illinois- Brick repealer states would have allowed Supreme Auto’s original complaint to go forward. That first complaint alleged injury based on the purchase of steel rods and similar items from distributors who, in turn, had purchased those same items from the defendants. The original complaint involves an indirect purchase (and so would be barred by Illinois Brick at the federal level) where the alleged injury is still fairly trace- able to the defendant steel manufacturers. The amended complaint is a different story. It alleges that plaintiffs purchased steel only insofar as it was one among many components of other more complex products, all of which have gone through numerous manufacturing altera- tions and lines of distribution. In many of these products, steel is not even a primary or necessary ingredient. We cannot im- agine—and plaintiffs have not told us—how one might tackle the task of tracing the effect of an alleged overcharge on steel through the complex supply and production chains that gave rise to the consumer products at issue here. The district court thus appropriately ruled that the claims asserted here were too remote to support a claim under the different state laws plaintiffs invoked. 16 No. 17-2910

IV We express no opinion here on the ultimate question whether these defendants violated the federal antitrust laws. Plaintiffs cite many statements by steel-industry executives that sound suspiciously like invitations to fix prices and out- puts. Price-fixing agreements are per se illegal under the Sher- man Act. Both direct purchasers and the federal government are authorized to sue when they believe they have found such an agreement. See, e.g., 15 U.S.C. §§ 4 (government civil suits), 15 (private treble damage actions). But, to repeat, we do not have a federal case before us. Plaintiffs have pared their action down to a case asserting violations of state laws, and we have concluded that the district court correctly decided that they have not shown how their alleged injury could be traced to defendants’ conduct—a requirement that is just as essential under the state laws as it is under federal law. Also, and more straightforwardly, they failed to raise their current claims within the relevant limitations periods. For both these rea- sons, we AFFIRM the judgment of the district court.