WOOD, Circuit Judge.

Potash, a naturally occurring mineral used in agricultural fertilizers and other products, is produced and sold in a global market. In this case, the plaintiffs, U.S. companies that are direct and indirect purchasers of potash, accuse several global producers of price-fixing in violation of the U.S. antitrust laws. See 15 U.S.C. §§ 1 et seq. The district court denied the defendants' motion to dismiss the complaint, but it certified its ruling for interlocutory appeal under 28 U.S.C. § 1292(b). We agreed with that court's assessment of the importance of the issues presented and accepted the appeal. A panel of the court concluded that the complaint failed to meet the requirements of the Foreign Trade Antitrust Improvements Act of 1982 (FTAIA), 15 U.S.C. § 6a, and it thus voted to reverse. Minn-Chem, Inc. v. Agrium Inc., 657 F.3d 650 (7th Cir.2011). We then decided to rehear the case en banc. We hold first that the FTAIA's criteria relate to the merits of a claim, and not to the subject-matter jurisdiction of the court. We therefore overrule our earlier en banc decision in United Phosphorus, Ltd. v. Angus Chem. Co., 322 F.3d 942 (7th Cir.2003). We then address the applicable standards for antitrust cases involving import commerce and the restrictions imposed by the FTAIA. We conclude that the district court correctly ruled that the complaint does state a claim under the federal antitrust laws.

I

The district court's opinion details the critical facts alleged in the Complaint, see In re Potash Antitrust Litig., 667 F.Supp.2d 907, 915-19 (N.D.Ill.2009), but for convenience we briefly summarize them here. The term "potash" refers to mineral and chemical salts that are rich in potassium. It is mined from naturally occurring ore deposits and its primary use is in agricultural fertilizers, but it is also used in the production of such varied products as glass, ceramics, soaps, and animal feed supplements. Importantly for our later antitrust analysis, potash is a homogeneous commodity: One manufacturer's supply is interchangeable with another's. As a result, buyers choose among suppliers based largely on price. Markets for this type of product are especially vulnerable to price-fixing.

We focus our analysis on the Direct Purchaser Amended Consolidated Class Action Complaint (referred to here simply as the Complaint), because the complaint filed by the indirect potash purchasers focuses primarily on state law remedies (since indirect purchasers are not entitled to sue for damages under the federal antitrust laws, see Illinois Brick Co. v. Illinois, 431 U.S. 720, 729, 97 S.Ct. 2061, 52 L.Ed.2d 707 (1977)).1 The Complaint alleges that the world's potash reserves are confined to a handful of areas, with over half of global capacity located in just two regions — Canada and the former Soviet Union (in particular, Russia and Belarus). Commercially, the industry has been dominated by a small group of companies that market, sell, and distribute potash. The key actors are:

The Complaint alleges that as of 2008, these seven entities produced approximately 71% of the world's potash.

In 2008, the United States consumed 6.2 million tons of potash. Of that total, 5.3 million tons were imports, and PCS, Mosaic, Agrium, and Belarusian Potash (acting for both Uralkali and Belaruskali, its equal and joint owners) were responsible for the lion's share of those sales. Data for other years covered by the Complaint are comparable.

The total world market for potash, in which the United States is an important consumer (second only to China, Complaint ¶ 51), is allegedly under the thumb of a global cartel consisting primarily of the companies listed above. This cartel restrained global output of potash in order to inflate prices. The cartel members used a rolling strategy: They would first negotiate prices in Brazil, India, and China (Complaint ¶ 111), and then use those prices as benchmarks for sales to U.S. customers. (Complaint ¶¶ 117-121). For example, in May 2004, the cartel arranged for prices to increase by $20 per ton for some foreign customers; shortly thereafter, prices in the United States went up by precisely the same amount.

The cartel initiated a sustained and successful effort to drive prices up beginning in mid-2003; by 2008 potash prices had increased at least 600%. The plaintiffs assert that this increase cannot be explained by a significant uptick in demand, changes in the cost of production, or other changes in input costs. In fact, U.S. consumption of fertilizer, of which potash is a consistent part, remained relatively steady throughout the period covered by this case; demand declined somewhat in 2008 but then returned to normal levels in 2009.2 One might think that the decrease in demand in 2008 was because of the increase in price, but the slippage in demand did not build up over the entire Class Period and appears to have been only temporary, and is thus not correlated to potash price movements. Furthermore, the specific allegation in the Complaint that a $100 per ton increase in the price of potash adds only $0.03 to the production cost of a bushel of corn suggests that demand for potash is inelastic. Complaint ¶ 54. Prices for potash rose and stayed high, increasing even while fertilizer prices declined. Based on World Bank statistics, average fertilizer price indices rose from 1.0 to 2.2, and then fell back to 1.0 in 2008, while potash price indices started in 2008 at 1.0 and rose to 3.5 by the end of the year. Earnings by cartel members reinforce this picture of financial gain even in the face of waning demand: PCS posted first-quarter income figures in 2008 that tripled its previous-year figure, while Mosaic's earnings for that quarter were up more than tenfold over the year before.

The Complaint goes into detail about ways in which the defendants managed their collective output. (A cartel will always try to restrict output to the level where marginal cost equals marginal revenue, but in the real world, this normally requires constant adjustment.) For example, when global demand for potash declined in 2005, rather than decreasing its price, PCS announced that it was shutting down three of its mines in November and December 2005 for "inventory control purposes." Complaint ¶ 88. This action had the effect of removing 1.34 million tons of potash from the world market. At the same time, rather than jumping into the gap this drastic cutback created, Mosaic announced that it too was implementing temporary cutbacks that would remove an additional 200,000 tons from the market. These (allegedly) coordinated and deep reductions continued into 2006. In the first three months of that year, PCS reduced output from 2.4 million tons to 1.3 million tons, removing yet another 1.1 million tons from the market, or the equivalent of 32 weeks of mining. Uralkali reduced its output by 200,000 tons, and Belaruskali cut its exports back by 50%, or 250,000 tons. In the second quarter of that year, Silvinit followed suit with mine stoppages that removed about 100,000 tons from the market. Collectively, these three companies removed over half a million tons of potash from the market in early 2006. See Complaint ¶¶ 88-93. Their compatriots applauded the "discipline" of the former Soviet Union producers, "noting that many years earlier when demand for potash declined those same producers had sought to maintain volume over price and flooded the market with excess supply." Complaint ¶ 93.

China was a particular target of the cartel's efforts, given its importance as a consumer. The shortages created by Uralkali's and PCS's supply restrictions in the first half of 2006 induced China to accept an increase in the price of potash. Shortly thereafter, a similar price increase was implemented throughout the world. Complaint ¶ 95. Comparable actions took place in 2007, as the Complaint rehearses in detail. The plaintiffs assert that a number of the defendants had excess capacity throughout the period between 2003 and mid-2009 (which represents the Class Period defined in ¶ 1 of the Complaint). PCS, for instance, had a utilization rate of only 54% to 69%, and Uralkali bragged in December 2007 that it had the "ability to add significant capacity on the cheapest basis vs. global peers." Complaint ¶¶ 133-134. This pattern of restrained output made it possible for the cartel to maintain its inflated prices, but the excess capacity inevitably gave its members an easy opportunity to cheat, and so the group had to coordinate to ensure that its price control efforts were not undermined.

The Complaint also points to several ways in which the cartel members had the opportunity to cooperate, to conspire on future actions, and to monitor one another's actions for possible cheating. First, the major suppliers participated in joint ventures that facilitated coordination. PCS, Agrium, and Mosaic were joint venturers and equal shareholders in Canpotex Ltd., a Canadian company that sold, marketed, and distributed potash throughout the world excluding the United States. Through that vehicle, those three companies had access to one another's sensitive production and pricing information. Canpotex in turn entered into cooperative marketing agreements with the Russian and Belarusian entities. As part of those deals, Canpotex agreed to market Uralkali potash outside North America and Europe. For their part, the former Soviet producers coordinated their sales and marketing through Belarusian Potash. That joint venture, formed between Uralkali and Belaruskali in 2005, supplied 34% of the market for potash by the following year. Complaint ¶ 26. Silvinit has sought to join the venture, and one of its owners (with a 20% share) owns 60% of the stock of Uralkali.

Beyond the access created by these structural relations among the entities, there were other more immediate opportunities to collude. The defendants routinely held meetings during the Class Period and engaged in an exchange program through which senior executives from each visited the others' plants. These meetings gave the defendants an opportunity to exchange sensitive information. Critically, one such meeting of the key players at PCS, Canpotex, Mosaic, Uralkali, Belaruskali, and Silvinit — mostly at the presidential level — took place in October 2005. As we described above, in the very next month, November 2005, PCS and Mosaic announced significant production cutbacks; the others followed suit with additional supply reductions through the beginning of 2006.

In addition, all of the defendants are members of the International Fertilizer Industry Association and the Fertilizer Institute, and they regularly attended those trade organizations' conferences. During one such meeting in Turkey, in May 2007, the defendants announced an additional price increase.

The Complaint contains, in its 165 paragraphs, many more details, which we discuss as needed below. What we have said here, however, is enough to set the stage for the two legal issues before us: how the FTAIA should be interpreted, and whether the district court correctly allowed this case to go forward.

II

Whether this case can be entertained by a court in the United States turns on the global reach of the antitrust laws, and to a significant degree on the Foreign Trade Antitrust Improvements Act of 1982, 15 U.S.C. § 6a. Before delving into the FTAIA's requirements, however, we take this opportunity to revisit the question whether that law affects the subject-matter jurisdiction of the district court or if, on the other hand, it relates to the scope of coverage of the antitrust laws. Nine years ago, in United Phosphorus v. Angus Chemical, the en banc court concluded that the former interpretation was correct. 322 F.3d 942, 952 (7th Cir.2003). In so doing, we relied on the legislative history of the statute, the vocabulary used by a number of commentators, and a number of court decisions that used the word "jurisdiction" in describing the requirement that challenged conduct must affect interstate or import commerce in specified ways.

Since that decision, the Supreme Court has emphasized the need to draw a careful line between true jurisdictional limitations and other types of rules. Thus, in Morrison v. National Australia Bank Ltd., ___ U.S. ___, 130 S.Ct. 2869, 177 L.Ed.2d 535 (2010), which dealt with the securities laws, the Court squarely rejected the notion that the extraterritorial reach of § 10(b) of the Securities Exchange Act, 15 U.S.C. § 78j(b), raises a question of subject-matter jurisdiction. Id. at 2877. "[T]o ask what conduct § 10(b) reaches is to ask what conduct § 10(b) prohibits, which is a merits question. Subject-matter jurisdiction, by contrast, refers to a tribunal's power to hear a case." Id. (citing Union Pacific R. Co. v. Locomotive Eng'rs & Trainmen Gen. Comm. of Adjustment, ___ U.S. ___, 130 S.Ct. 584, 596, 175 L.Ed.2d 428 (2009); Arbaugh v. Y & H Corp., 546 U.S. 500, 514, 126 S.Ct. 1235, 163 L.Ed.2d 1097 (2006); United States v. Cotton, 535 U.S. 625, 630, 122 S.Ct. 1781, 152 L.Ed.2d 860 (2002)). The Court might have added to that list Reed Elsevier, Inc. v. Muchnick, ___ U.S. ___, 130 S.Ct. 1237, 1243, 176 L.Ed.2d 18 (2010), Kontrick v. Ryan, 540 U.S. 443, 455, 124 S.Ct. 906, 157 L.Ed.2d 867 (2004), and Steel Co. v. Citizens for a Better Environment, 523 U.S. 83, 89, 118 S.Ct. 1003, 140 L.Ed.2d 210 (1998). Even more recently, the Court restated this proposition in Henderson v. Shinseki, ___ U.S. ___, 131 S.Ct. 1197, 1202-03, 179 L.Ed.2d 159 (2011). Notably, what may have been thought a nascent idea at the time United Phosphorus was decided (as one can tell by the dates of decision in our list) has now become a firmly established principle of statutory construction.

The panel in the present case had no quarrel with the proposition that this recent string of decisions undermined the holding in United Phosphorus. 657 F.3d at 653. It commented that "[t]hese intervening developments suggest that United Phosphorus may be ripe for reconsideration," but it was hesitant to take that step on its own. The panel also observed that the same issue had recently come before the Third Circuit, which held that the FTAIA does not impose a jurisdictional limit but instead establishes an element of a Sherman Act claim. Id. at 659 n. 3 (citing Animal Sci. Prods., Inc. v. China Minmetals Corp., 654 F.3d 462 (3d Cir. 2011)). Indeed, the Animal Science opinion expressly approved the position of the United Phosphorus dissenters. 654 F.3d at 469 n. 8. We agree with the panel that this issue is indeed ripe for reconsideration and ought to be settled now.

The Supreme Court's decision in Morrison, we believe, provides all the guidance we need to conclude that, like § 10(b) of the Exchange Act, the FTAIA sets forth an element of an antitrust claim, not a jurisdictional limit on the power of the federal courts. As the Court put it, limitations on the extraterritorial reach of a statute describe what conduct the law purports to regulate and what lies outside its reach. The Supreme Court itself used much the same language with respect to the antitrust laws in its decision in F. Hoffmann-La Roche Ltd. v. Empagran S.A., 542 U.S. 155, 124 S.Ct. 2359, 159 L.Ed.2d 226 (2004), which dealt specifically with the FTAIA. The Court spoke, for example, of the FTAIA's "removing from the Sherman Act's reach" certain types of conduct, id. at 161, 124 S.Ct. 2359, and whether it was reasonable under the facts presented there "to apply this law to conduct that is significantly foreign," id. at 166, 124 S.Ct. 2359. Even if one thought the language in Empagran to be less than dispositive, we can now see no way to distinguish this case from Morrison.

We add briefly that the interpretation we adopt today — that the FTAIA spells out an element of a claim — is the one that is both more consistent with the language of the statute and sounder from a procedural standpoint. When Congress decides to strip the courts of subject-matter jurisdiction in a particular area, it speaks clearly. The FTAIA, however, never comes close to using the word "jurisdiction" or any commonly accepted synonym. Instead, it speaks of the "conduct" to which the Sherman Act (or the Federal Trade Commission Act) applies. This is the language of elements, not jurisdiction.

From a procedural standpoint, this means that a party who wishes to contest the propriety of an antitrust claim implicating foreign activities must, at the outset, use Federal Rule of Civil Procedure 12(b)(6), not Rule 12(b)(1). This is not a picky point that is of interest only to procedure buffs. Rather, this distinction affects how disputed facts are handled, and it determines when a party may raise the point. While "it is the burden of the party who seeks the exercise of jurisdiction in his favor clearly to allege facts demonstrating that he is a proper party to invoke judicial resolution," FW/PBS, Inc. v. City of Dallas, 493 U.S. 215, 231, 110 S.Ct. 596, 107 L.Ed.2d 603 (1990) (citations and quotation marks omitted), we "accept as true all of the allegations contained in a complaint," Ashcroft v. Iqbal, 556 U.S. 662, 678, 129 S.Ct. 1937, 173 L.Ed.2d 868 (2009) (citing Bell Atlantic Corp. v. Twombly, 550 U.S. 544, 127 S.Ct. 1955, 167 L.Ed.2d 929 (2007)) subject, of course, to the limitations articulated in those cases. Likewise, subject-matter jurisdiction must be secure at all times, regardless of whether the parties raise the issue, and no matter how much has been invested in a case. See, e.g., Cotton, 535 U.S. at 630, 122 S.Ct. 1781 (citing Louisville & Nashville R. Co. v. Mottley, 211 U.S. 149, 29 S.Ct. 42, 53 L.Ed. 126 (1908)). By contrast, a motion to dismiss for failure to state a claim may only be brought as late as trial. FED. R.CIV.P. 12(h)(2). Although this is a significant difference, we note that foreign connections of the kind at issue here are unlikely to be difficult to detect, and so we are confident that parties who want to argue that a particular claim fails the requirements of the FTAIA will be able to do so within these generous time limits.

III

Having established that the FTAIA relates to the merits of a claim, rather than the subject-matter jurisdiction of the court, we can now turn to the principal issues in this appeal. We consider first how the statute should be interpreted and then, on that understanding of the law, we decide whether the district court correctly found that the Complaint stated a claim that could go forward.

A

Although the FTAIA has been parsed in a number of judicial opinions, including notably Empagran, we think it important to begin with the language of the statute, in order to place our discussion of these decisions in context. We note that the 1982 legislation that we are examining actually amended both the Sherman Act, see 15 U.S.C. § 6a, and the Federal Trade Commission Act, see 15 U.S.C. § 45(a)(3), using identical language. That fact is important insofar as it underscores the generality of the issue we face: The statute applies not only to private actions, such as this one, but also to actions brought by the two federal agencies entrusted with the enforcement of the antitrust laws. Since it is the Sherman Act that applies to our case, however, from this point forward we cite only its provision. It reads as follows:

The opening phrase (sometimes referred to as a chapeau in international circles) reflects Congress's effort to indicate that the Sherman Act does not apply to every arrangement that literally can be said to involve trade or commerce with foreign nations. As the Supreme Court stressed in Empagran, the public recognition of this limitation was inspired largely by international comity. But, by inserting the parenthetical "other than import trade or import commerce" in the chapeau, Congress recognized that there was no need for this self-restraint with respect to imports, even though they represent part of the foreign commerce of the United States. Although some, including the Third Circuit in Animal Science, have referred to this as the "import exception," that is not an accurate description. Import trade and commerce are excluded at the outset from the coverage of the FTAIA in the same way that domestic interstate commerce is excluded. This means only that conduct in both domestic and import trade is subject to the Sherman Act's general requirements for effects on commerce, not to the special requirements spelled out in the FTAIA. Where the FTAIA does apply, it "remov[es] from the Sherman Act's reach... commercial activities taking place abroad, unless those activities adversely affect ... imports to the United States" Empagran, 542 U.S. at 161, 124 S.Ct. 2359. The Court's decision in Hartford Fire Ins. Co. v. California, 509 U.S. 764, 113 S.Ct. 2891, 125 L.Ed.2d 612 (1993), suggests a pragmatic reason for this distinction: The applicability of U.S. law to transactions in which a good or service is being sent directly into the United States, with no intermediate stops, is both fully predictable to foreign entities and necessary for the protection of U.S. consumers. Foreigners who want to earn money from the sale of goods or services in American markets should expect to have to comply with U.S. law.

Next, we come to the statute's treatment of non-import, non-domestic commerce. Empagran explained that the FTAIA handles that problem by "lay[ing] down a general rule placing all (nonimport) activity involving foreign commerce outside the Sherman Act's reach ... [and then] bring[ing] such conduct back within" the Act provided that it meets the two criteria provided. Id. at 162, 124 S.Ct. 2359 (emphasis in original). The first criterion dictates the kinds of effects that truly foreign commerce must have in the U.S. market. Conduct "involving trade or commerce ... with foreign nations" must have "a direct, substantial, and reasonably foreseeable effect" on either [A] U.S. domestic commerce (phrased awkwardly as "trade or commerce which is not trade or commerce with foreign nations") or U.S. import commerce, or [B] the export trade or commerce of a U.S. exporter. See § 6a(1). The export trade provision plays no part in our case, and so we do not address it further here. The second criterion, which was the focus of Empagran, is that the direct, substantial and foreseeable effect shown under subpart (1) must give rise to a substantive claim under the Sherman Act. The reason this was important in Empagran is that the plaintiffs there were foreign purchasers of allegedly price-fixed products that were sold in foreign markets. The Court held that their claims fell outside the scope of the Sherman Act. In our case, by contrast, the plaintiffs are all U.S. purchasers, and so the particular problem addressed in Empagran does not arise here.

Thus, before we can address the merits of the complaint, we must address two distinct questions of statutory interpretation. The first is how to define pure import commerce — that is, the kind of commerce that is not subject to the special rules created by the FTAIA. Second, we must explore the FTAIA's standards further and explain what it takes to show that foreign conduct has a direct, substantial, and reasonably foreseeable effect on U.S. domestic or import commerce.

1

There can be no question that the import commerce exclusion puts some of the conduct alleged in the Complaint outside the special rules created in the FTAIA for Sherman Act claims. The plaintiffs are U.S. entities that have purchased potash directly from members of the alleged cartel. The defendant members of the cartel are all located outside the United States. Those transactions that are directly between the plaintiff purchasers and the defendant cartel members are the import commerce of the United States in this sector.

The FTAIA does not require any special showing in order to bring these transactions back into the Sherman Act, as Empagran put it, because they were never removed from the statute. That does not mean, however, that plaintiffs are home free. Rather, we must still apply the rules governing import commerce for purposes of the antitrust laws. For several decades, the leading authority on this subject was Judge Learned Hand's opinion for the Second Circuit in United States v. Aluminum Co. of America, 148 F.2d 416, 444 (2d Cir.1945) (Alcoa ). There the court (sitting as a court of last resort because the Supreme Court lacked a quorum) held that the Sherman Act covers imports when actual and intended effects on U.S. commerce have been shown. In Hartford Fire, the Supreme Court confirmed this rule, stating that "the Sherman Act covers foreign conduct producing a substantial intended effect in the United States." 509 U.S. at 797, 113 S.Ct. 2891. The Third Circuit has suggested that this standard is met where "the defendants' conduct target[s] import goods or services." Animal Science, 654 F.3d at 470.

As noted, the Complaint before us alleges import transactions. Thus, the only outstanding question is whether this import trade has been substantially and intentionally affected by an anticompetitive arrangement (i.e., something that would violate the U.S. antitrust laws). There is nothing particularly "international" about that question. Effects on commerce are a part of every Sherman Act case. See, e.g., Hartford Fire, supra (import commerce); Summit Health, Ltd. v. Pinhas, 500 U.S. 322, 111 S.Ct. 1842, 114 L.Ed.2d 366 (1991) (interstate commerce). We address the adequacy of the Complaint under the Sherman Act in more detail below.

2

As we already have observed, trade involving only foreign sellers and domestic buyers (i.e., import trade) is not subject to the FTAIA's extra layer of protection against Sherman Act claims implicating foreign activities. Some of the activities alleged in the Complaint, however, may be best understood as sufficiently outside the arena of simple import transactions as to require application of the FTAIA. For example, Canpotex is the unified marketing and sales agent for Agrium, Mosaic and PCS in all markets except Canada and the United States, yet its actions are an important part of the alleged scheme to set inflated benchmark prices. Presumably, in order to avoid Illinois Brick's prohibition on "pass on" antitrust damages, 431 U.S. at 728, 97 S.Ct. 2061, the plaintiffs are seeking to hold firms like Canpotex jointly and severally liable for any damages the direct sellers might be ordered to pay, perhaps under a conspiracy theory. If this were an action by the Department of Justice or the Federal Trade Commission, we would not need to worry about Illinois Brick, but regardless of whether the case is brought by the government or in private litigation, it is essential to meet the criteria spelled out by the FTAIA. We thus take a closer look at what kind of conduct "involve[s] trade or commerce ... with foreign nations" and what showing is necessary to demonstrate "direct, substantial and reasonably foreseeable" effects on domestic [i.e., "not trade or commerce with foreign nations"] or import commerce.

The first question — whether the conduct alleged in this case "involves" foreign commerce — is readily answered. The Complaint alleges an international cartel in a commodity, and it asserts that the cartel succeeded in raising prices for direct U.S. purchasers of the product, potash. This alleged arrangement plainly involves foreign commerce, and so we move immediately to the second inquiry — the task of parsing the statute's central requirements. As Empagran put it, after excluding foreign activities from the scope of the Sherman Act, the FTAIA brings back into the statute's reach conduct that has a "direct, substantial, and reasonably foreseeable effect" on domestic or import commerce.

The potash cartel described in the Complaint is one for which the requirements of substantiality and foreseeability are easily met. There is little dispute that the Complaint has alleged substantial effects: The Complaint alleges that 5.3 million tons of potash were imported into the United States in 2008 alone, and the Complaint elsewhere asserts that the vast majority of these imports came from the defendants. From 2003 to 2008, the price of potash increased by over 600%. We do not need to belabor the point. These allegations easily satisfy the requirement to show substantial effects in the U.S. market. Wherever the floor may be, it is so far below these numbers that we do not worry about it here.

Foreseeability is equally straightforward. It is objectively foreseeable that an international cartel with a grip on 71% of the world's supply of a homogeneous commodity will charge supracompetitive prices, and in the absence of any evidence showing that arbitrage is impossible (and there is none here), those prices (net of shipping costs) will be uniform throughout the world. Higher prices cannot be divorced from reductions in supply, and so the effects alleged here are a rationally expected outcome of the conduct stated in the Complaint.

The question that has caused more discussion among various courts and commentators is what it takes to show "direct" effects. One school of thought, launched by the Ninth Circuit's split decision in United States v. LSL Biotechs., 379 F.3d 672 (9th Cir.2004), has borrowed the definition of the word "direct" that the Supreme Court adopted for a different statute, the Foreign Sovereign Immunities Act (FSIA), 28 U.S.C. § 1605(a)(2); see Republic of Argentina v. Weltover, Inc., 504 U.S. 607, 112 S.Ct. 2160, 119 L.Ed.2d 394 (1992). The word appears in the exception for foreign sovereign immunity that applies for commercial activity that takes place outside the territory of the United States when "that act causes a direct effect in the United States." In that setting, the Court held that an effect is "direct" if it "follows as an immediate consequence of the defendant's ... activity." Id. at 618, 112 S.Ct. 2160. The other school of thought has been articulated by the Department of Justice's Antitrust Division, which takes the position that, for FTAIA purposes, the term "direct" means only "a reasonably proximate causal nexus." Makan Delrahim, Drawing the Boundaries of the Sherman Act: Recent Developments in the Application of the Antitrust Laws to Foreign Conduct, 61 N.Y.U. ANN. SURV. AM. L. 415, 430 (2005) (remarks of the Deputy Assistant Attorney General); Brief for Appellant United States of America 38 in United States v. LSL Biotechs., supra, available at http://www.justice.gov/atr/ cases/f200200/200243.pdf (directness is a synonym for proximate cause).

In our view, the Ninth Circuit jumped too quickly to the assumption that the FSIA and the FTAIA use the word "direct" in the same way. Critically, the Supreme Court in Weltover reached its definition of "direct" for FSIA purposes only after refusing to import from the legislative history of that statute the notion that an effect is "direct" only if it is both "substantial" and "foreseeable." 504 U.S. at 617, 112 S.Ct. 2160. "[W]e reject," it said, "the suggestion that § 1605(a)(2) contains any unexpressed requirement of `substantiality' or `foreseeability.'" Id. at 618, 112 S.Ct. 2160. Only then did the Court endorse the appellate court's definition that an effect is "direct" if it follows "as an immediate consequence" of the defendant's activity. Id.

No one needs to read the words "substantial" and "foreseeable" into the FTAIA. Congress put them there, and in so doing, it signaled that the word "direct" used along with them had to be interpreted as part of an integrated phrase. Superimposing the idea of "immediate consequence" on top of the full phrase results in a stricter test than the complete text of the statute can bear. To demand a foreseeable, substantial, and "immediate" consequence on import or domestic commerce comes close to ignoring the fact that straightforward import commerce has already been excluded from the FTAIA's coverage.

We are persuaded that the Department of Justice's approach is more consistent with the language of the statute. The word "direct" addresses the classic concern about remoteness — a concern, incidently, that has been at the forefront of international antitrust law at least since Judge Hand wrote in Alcoa that "[w]e should not impute to Congress an intent to punish all whom its courts can catch, for conduct which has no consequences within the United States." 148 F.2d at 443; see also LSL Biotechs., 379 F.3d at 683-91 (Aldisert, J., dissenting) (tracing the history of the FTAIA's effects test through Alcoa ). Just as tort law cuts off recovery for those whose injuries are too remote from the cause of an injury, so does the FTAIA exclude from the Sherman Act foreign activities that are too remote from the ultimate effects on U.S. domestic or import commerce.

This understanding of the FTAIA should allay any concern that a foreign company that does any import business at all in the United States would violate the Sherman Act whenever it entered into a joint-selling arrangement overseas regardless of its impact on the American market. A number of safeguards exist to protect against that risk. If the hypothetical foreign company is engaged in direct import sales, it must naturally comply with U.S. law just as all of its domestic competitors do. If its foreign sales do not meet the threshold for "effects" on import or domestic commerce established by cases such as Hartford Fire and Summit Health, then, for those transactions, it has nothing to worry about. If the hypothetical foreign company is engaged in the kind of conduct outside the United States that the FTAIA addresses, then its actions can be reached only if there are direct, substantial, and reasonably foreseeable effects. This is a standard with teeth, as the many cases that have been dismissed for failing to meet those criteria attest. E.g., Turicentro, S.A. v. Am. Airlines, Inc., 303 F.3d 293 (3d Cir.2002); Carpet Grp. Int'l v. Oriental Rug Imps. Ass'n, 227 F.3d 62 (3d Cir.2000); McGlinchy v. Shell Chem. Co., 845 F.2d 802 (9th Cir.1988); Filetech S.A. v. France Telecom S.A., 212 F.Supp.2d 183 (S.D.N.Y.2001).

Empagran is consistent with the interpretation we adopt here. While it holds that the U.S. antitrust laws are not to be used for injury to foreign customers, it goes on to reaffirm the well-established principle that the U.S. antitrust laws reach foreign conduct that harms U.S. commerce:

Empagran, 542 U.S. at 165, 124 S.Ct. 2359. Finally, we note that § 6a(2) will protect many a foreign defendant. No matter what the quality of the foreign conduct, the statute will not cover it unless the plaintiff manages to state a claim under the Sherman Act. In this connection, we point out that a great many joint-selling arrangements are legal, efficiency-enhancing structures. See, e.g., Texaco Inc. v. Dagher, 547 U.S. 1, 126 S.Ct. 1276, 164 L.Ed.2d 1 (2006); Broadcast Music, Inc. v. Columbia Broadcasting Sys., Inc., 441 U.S. 1, 99 S.Ct. 1551, 60 L.Ed.2d 1 (1979).

B

Having described the requirements for both simple import commerce and the FTAIA, our final task is to measure the Complaint against these standards. In particular, we must decide whether the plaintiffs have plausibly alleged that the defendants' conduct took place either in import commerce and are thus subject to the more general rules of Hartford Fire for effects on commerce, or if they have in whole or in part described conduct subject to the FTAIA, and if so, whether the allegations describe direct, substantial, and foreseeable effects on domestic or import commerce.

1

In our view, much of the Complaint alleges straightforward import transactions. Under Hartford Fire the plaintiffs thus must allege that the conduct of the foreign cartel members was (1) meant to produce and (2) did in fact produce some substantial effect in the United States. See also Animal Science, 654 F.3d at 470 ("[T]he import trade or commerce [exclusion] requires that the defendants' conduct target import goods or services."). The Complaint contains ample material supporting both of those points.

The plaintiffs describe a tight-knit global cartel, similar to OPEC in its heyday, that restrained global output of potash so that prices throughout this homogeneous world market would remain artificially high. Just like the raisin producers in California in the famous state-action antitrust case, Parker v. Brown, 317 U.S. 341, 63 S.Ct. 307, 87 L.Ed. 315 (1943), who controlled 90% of the world market in raisins, the alleged cartel members here control a comparable share of the world market in potash. The purpose of this cartel was to inflate the profits of its members. Its alleged effect was substantial. The United States, according to the Complaint, is one of the two largest consumers of potash in the world, and approximately 85% of U.S. potash comes from overseas. From 2003 to 2008, the price of potash increased six-fold. The inference from these allegations is not just plausible but compelling that the cartel meant to, and did in fact, keep prices artificially high in the United States.

2

We turn next to an analysis of the conduct that falls outside the import exclusion to determine whether it may nevertheless be subject to the Sherman Act under the FTAIA. For example, the Complaint alleges that Canpotex, a Canadian entity that does not sell directly into the United States, restricted supply during a period of especially difficult price negotiations with China. This supply restriction compelled Chinese buyers to accept a price increase. Complaint ¶ 94. We assume for present purposes that none of this literally involved import trade. Our discussion, however, is rooted in the facts of this Complaint. In that connection, it is important to recall that the FTAIA itself demands that the facts of each case must be evaluated for compliance with its demands. We thus address only the situation before us, in which several members of the cartel sold directly into the United States and others allegedly worked with them in connection with those efforts. The question before us is thus whether the allegations in the plaintiffs' Complaint describe conduct that had a direct, substantial, and reasonably foreseeable effect on domestic or import commerce by, for example, setting a benchmark price intended to govern later U.S. sales.

As we noted above, the effects of the supply restriction on U.S. potash prices were foreseeable. So too were the effects of forcing foreign purchasers to accept higher prices in a commoditized and cartelized market: Either someone in the cartel would cheat, or a new entrant would begin to arbitrage its purchases, or, as the plaintiffs allege, the cartel would succeed in pushing prices up across all of its markets, including the United States. And, as we have explained, there is every reason to infer that any such effects in the U.S. potash market were substantial.

We turn to the question whether these effects are "direct," as we have defined the term. The plaintiffs allege that the defendants would first negotiate prices in Brazil, India, and China, and then they would use those prices for sales to U.S. customers. The alleged supply reductions led to price hikes in these foreign markets, and those increases showed up almost immediately in the prices of U.S. imports. The defendants do not suggest that the potash market is insulated from these effects by regulatory structures or other arrangements, and even if they did, that would be no reason to dismiss the Complaint outright. To the contrary, the plaintiffs have alleged that the cartel established benchmark prices in markets where it was relatively free to operate, and it then applied those prices to its U.S. sales. (Benchmark prices set in one market for general use are common: think, for instance of the London Interbank Offered Rate (LIBOR), in the credit market; the Brent Crude price, formally used for North Sea oil but in general use in oil markets; or even the Medicare Fee Schedule, which though technically only for Medicare reimbursements, has widespread effects on the healthcare market.) It is no stretch to say that the foreign supply restrictions, and the concomitant price increases forced upon the Chinese purchasers, were a direct — that is, proximate — cause of the subsequent price increases in the United States.

The allegations in the Complaint state a claim, as required by Federal Rule of Civil Procedure 8, and thus are enough to withstand a motion to dismiss under Rule 12(b)(6), as interpreted by the Supreme Court in Bell Atlantic Corp. v. Twombly, 550 U.S. 544, 127 S.Ct. 1955, 167 L.Ed.2d 929 (2007) and Ashcroft v. Iqbal, 556 U.S. 662, 129 S.Ct. 1937, 173 L.Ed.2d 868 (2009). The Complaint is not defeated by the defendants' contention that the alleged cartel was not efficacious. See In re High Fructose Corn Syrup Antitrust Litig., 295 F.3d 651, 656 (7th Cir.2002). We are also satisfied that the allegations suffice, at this stage, to support a plausible story of concerted action. See In re Text Messaging Antitrust Litig., 630 F.3d 622 (7th Cir. 2010). We stress, however, that our evaluation throughout has proceeded exclusively on the face of the Complaint. Nothing we have said should be understood as a prediction of the facts that may turn up in discovery, nor are we opining about the likely fate of any possible defenses. In particular, the defendants mentioned in their opposition to the petition for rehearing en banc that some of their actions were undertaken with the approval of foreign governments (e.g., Canada's). We express no opinion on either the contours or the likely success of any such argument. Similarly, we do not have before us any question about the court's personal jurisdiction over the various defendants. Cf. J. McIntyre Machinery, Ltd. v. Nicastro, ___ U.S. ___, 131 S.Ct. 2780, 180 L.Ed.2d 765 (2011). We are not faced with the question of whether the actions of the non-selling defendants, such as Canpotex, fall outside the substantive scope of Sherman Act § 1 (as opposed to the law's territorial reach), nor have the defendants argued that Congress as a matter of U.S. law has no constitutional power to enact laws with some extraterritorial effect. These or other theories may all be important to explore as the case goes forward, but they do not provide a reason to throw out the case on the grounds that the plaintiffs failed to show either that the challenged transactions occurred in import commerce or that they had a direct, substantial, and reasonably foreseeable effect on either the domestic or import commerce of the United States.

IV

Foreign cartels, especially those over natural resources that are scarce in the United States and that are traded in a unified international market, have often been the target of either governmental or private litigation. The host country for the cartel will often have no incentive to prosecute it. Canada and Russia, here (just like California in Parker), would logically be pleased to reap economic rents from other countries; their losses from higher prices for the potash used in their own fertilizers are more than made up by the gains from the cartel price their exporters collect. Export cartels are often exempt from a country's antitrust laws: the United States does just that, through its Webb-Pomerene Associations, see 15 U.S.C. §§ 61 et seq., and Export Trading Companies, see 15 U.S.C. §§ 4001 et seq. This case is actually the mirror image of the situation described in Empagran, where the foreign country whose consumers are hurt would have been the better enforcer. It is the U.S. authorities or private plaintiffs who have the incentive — and the right — to complain about overcharges paid as a result of the potash cartel, and whose interests will be sacrificed if the law is interpreted not to permit this kind of case.

The world market for potash is highly concentrated, and customers located in the United States account for a high percentage of sales. This is not a House-that-Jack-Built situation in which action in a foreign country filters through many layers and finally causes a few ripples in the United States. To the contrary: foreign sellers allegedly created a cartel, took steps outside the United States to drive the price up of a product that is wanted in the United States, and then (after succeeding in doing so) sold that product to U.S. customers. The payment of overcharges by those customers was objectively foreseeable, and the amount of commerce is plainly substantial. We AFFIRM the order of the district court denying the motion to dismiss for failure to state a claim.