Two United States companies that purchase rubber thread brought this private antitrust action, alleging a price-fixing conspiracy led by Southeast Asian producers of the thread. After an eight-day trial, the jury returned a special verdict, finding that although one or more of the producers engaged in a conspiracy to fix prices that was intended to affect United States commerce, that conspiracy had no "substantial effect" on this country's commerce. The district court then entered judgment on the verdict for the producers. The purchasers appeal, principally contending that the substantial-effect test applies only to "wholly" foreign conduct, and so does not govern this case because the rubber-thread conspiracy resulted in the sale of price-fixed goods directly into the United States. Because the conspiracy involved primarily foreign conduct, we hold that the district court did not abuse its discretion in applying the substantial-effect test. Accordingly, we affirm.

In 1997, Dee-K Enterprises, Incorporated, and Asheboro Elastics Corporation (collectively Dee-K), United States corporations that purchase rubber thread to make elastic fabric, brought this class action, alleging a conspiracy to fix the price of rubber thread in the United States, in violation of the Sherman Act. See 15 U.S.C.A. § 1 (West 1997). Rubber thread, also called extruded rubber thread or elastic rubber thread, and sometimes abbreviated as "ERT," is manufactured in Southeast Asia and used to make elastic fabric, bungee cords, toys, and other products.

Dee-K named as defendants nine Southeast Asian producers of rubber thread and some of their subsidiaries and distributors in the United States. Five of the producers are Malaysian companies: Heveafil Sendirian Berhad, Filmax Sendirian Berhad, Rubfil Sendirian Berhad, Rubberflex Sendirian Berhad, and Filati Lastex Sendirian Berhad. (The suffix "Sendirian Berhad," used in Malaysia and abbreviated "Sdn. Bhd.," translates as "private limited company.") Two are Indonesian: PT. Bakrie Rubber Industries and PT. Perkebunan III. Two are Thai: Longtex Rubber Industries Company, Limited, and Natural Rubber Thread Company, Limited. Dee-K also named as defendants the United States subsidiaries of three Malaysian producers (Rubfil USA, Incorporated, Flexfil Corporation of Rhode Island, Flexfil Corporation, and Filati Lastex Elastofibre USA, Incorporated) and two United States independent distributors used by other producers (Consortium International and JPS Elastomerics).

In its complaint, Dee-K alleged that the members of the class it sought to represent, domestic purchasers of rubber thread, paid "artificially high and non-competitive prices" for rubber thread, that they "were deprived of free and open competition in the market" for rubber thread, and that "competition among defendants" in the United States sale of rubber thread "was restrained." As to injury, Dee-K contended that "plaintiffs ... purchased substantial quantities of extruded rubber thread from defendants."

Dee-K originally filed this action in the Eastern District of Virginia. Following a number of early rulings not relevant to our disposition of this appeal, the district court determined that venue did not lie in Virginia and transferred the case to the Western District of North Carolina. See Dee-K Enters. v. Heveafil Sdn. Bhd., 985 F. Supp. 640 (E.D.Va.1997). Prior to trial, that court denied class certification. After most defendants settled, declined to appear, or were dismissed, the case against the five Malaysian producers and the United States subsidiary of one of them (Rubfil USA), none of whom now contest personal jurisdiction, see Dee-K Enters. v. Heveafil Sdn. Bhd., 174 F.R.D. 376 (E.D.Va.1997), proceeded to trial before a jury.

Dee-K introduced substantial evidence at trial of horizontal price fixing among the producers. This price fixing apparently originated at least in part in reaction to 1991 threats by the United States government to punish Southeast Asian rubber-thread producers for violating antitrust prohibitions against "dumping." "Dumping" occurs when a foreign producer injures a United States producer by selling a product in the United States at less than what would be "fair value" in the foreign producer's home market. See 19 U.S.C.A. § 1673 (West 1999) (authorizing an "antidumping duty"). The United States Department of Commerce may impose tariffs on dumpers to bring the United States price into line with the price in the producer's home market. See id. Thus, if a product sells for $1 in the home market, it warrants dumping duties if it sells for less than $1 in the United States. Of course, avoidance of dumping penalties in itself does not provide foreign producers with a license to fix prices in violation of United States antitrust laws. Although to avoid dumping a company must price goods at or above the price in its own home market, it may not agree with its competitors to fix prices, restricting the market movement of prices in the United States market. See Dee-K Enters. v. Heveafil Sdn. Bhd., 982 F. Supp. 1138, 1156 & n. 45 (E.D.Va.1997); see also United States v. Nippon Paper Indus., 62 F. Supp. 2d 173, 180 (D.Mass.1999) (noting that foreign companies threatened with anti-dumping provisions must "walk a fine line").

In December 1991, responding to the dumping accusations, officials of the Malaysian producers representing Heveafil, Rubfil, Rubberflex, and Filati Lastex met with a Malaysian government official and agreed to fix rubber-thread prices throughout the world. Later joined by other rubber-thread producers from Malaysia, Indonesia, and Thailand, they continued to meet for several years, at conventions and in other settings, to discuss and implement these and other efforts to fix prices. They met regularly between 1992 and 1995 — in Kuala Lumpur, in Columbo, in Bali, and in Penang. They never met in the United States.

Throughout the period during which they met to fix prices, the Malaysian producers sold their rubber thread around the world, distributing it to the United States market in three different ways. Heveafil and Filmax sold to the United States through a division of Heveafil based in the United States. Rubfil, Rubberflex, and Filati Lastex all sold to large United States customers directly and to smaller customers through wholly owned subsidiaries incorporated in the United States, all four of which were named as defendants. See Dee-K, 982 F.Supp. at 1142. The record does not disclose the United States share of the global market.

From 1991 to 1996, United States prices for rubber thread (adjusted for inflation using the producer price index) generally rose, with some decreases on various scales. Dee-K attributes these increases to the price-fixing conspiracy. The producers attribute them to an antidumping order entered by the United States Department of Commerce in 1992 that imposed a duty on rubber thread and to increases in the price of raw materials, particularly the price of latex.

At the conclusion of an eight-day trial, the district court submitted a special verdict form to the jury. The verdict form included two questions: (1) Was there "a conspiracy ... to fix the prices of extruded rubber thread, which was intended to have a substantial effect in the United States"? (2) If so, did "the conspiracy have a substantial effect in the United States"?

The jury answered the first question in the affirmative, finding a conspiracy to fix prices with the intent of affecting the United States. But the jury answered the second question in the negative, finding that the conspiracy did not have a substantial effect in the United States. In accordance with this verdict, the court entered judgment for the producers.

Dee-K moved for a new trial pursuant to Rule 59, arguing that the jury's verdict as to the lack of a substantial effect on United States commerce was contrary to the weight of the evidence. The district court denied the motion and simultaneously denied a late-filed Rule 50 motion for judgment as a matter of law on the substantial-effect question. Dee-K appeals, contending that the substantial-effect test (at least as stated in the jury instructions and special verdict) does not govern this case.1

In Hartford Fire Insurance Co. v. California, the Supreme Court noted that "it is well established that the Sherman Act applies to foreign conduct that was meant to produce and did in fact produce some substantial effect in the United States." 509 U.S. 764, 796, 113 S. Ct. 2891, 125 L. Ed. 2d 612 (1993) (citations omitted).

Dee-K argues that this statement in Hartford Fire does not govern the case at hand, primarily because the price-fixing conspiracy it proved does not constitute "foreign conduct." Dee-K contends that instead the jurisdictional test used in domestic antitrust cases controls here; in domestic cases a plaintiff need only demonstrate "that the defendant's activity is itself in interstate commerce, or ... that it has an effect on some other appreciable activity demonstrably in interstate commerce." See McLain v. Real Estate Bd. of New Orleans, 444 U.S. 232, 242, 100 S. Ct. 502, 62 L. Ed. 2d 441 (1980) (emphasis added and citation omitted). Accordingly, Dee-K asserts that the district court abused its discretion by applying the substantial-effect test from Hartford Fire. See Nelson v. Green Ford, Inc., 788 F.2d 205, 208-09 (4th Cir.1986) (noting that we review jury instructions for abuse of discretion); Tights, Inc. v. Acme McCrary Corp., 541 F.2d 1047, 1060 (4th Cir.1976) (noting that we review special verdict interrogatories for abuse of discretion).

The producers respond that Hartford Fire is "analytically indistinguishable" from Dee-K's case. Thus, they maintain that the district court correctly applied Hartford Fire when it required Dee-K to show a "substantial effect in the United States" before concluding that "the Sherman Act applies" to the "foreign conduct" that Dee-K alleged.

The mixture of foreign and domestic elements in this case makes its analysis challenging; either an entirely foreign or an entirely domestic conspiracy would present a comparatively easy jurisdictional question. If the conspiracy had involved participants with United States affiliations, acting only in the United States, and targeting a United States market, the jurisdiction of United States courts would be clear, without any proof of effect.2 At the other extreme, if everything about the conspiracy were foreign — if it were a conspiracy formed in Southeast Asia, by Southeast Asian persons and corporations, intended to affect only Southeast Asian markets, and affecting only those markets — the Sherman Act would not provide a United States court any jurisdiction to address it. See Matsushita Elec. Indus. v. Zenith Radio Corp., 475 U.S. 574, 582, 106 S. Ct. 1348, 89 L. Ed. 2d 538 (1986) ("American antitrust laws do not regulate the competitive conditions of other nations' economies.").

We earn our keep, of course, in the middle ground between such extremes — in cases such as this. Dee-K alleged (and proved to the jury) a largely foreign conspiracy with some domestic elements, aimed at a global market including, but certainly not limited to, a United States import market. Indeed, the conspiracy mixed foreign and domestic elements in several respects: it included many participants with foreign affiliations but a few who also had United States affiliations; acts that range from a series of conspiratorial meetings, all held abroad, to routine communications, a few with the United States; and a target market embracing dozens of nations including the United States. This sort of mixed fact pattern will probably become increasingly familiar as global economic links and assertions of trans-national jurisdiction increase. See generally Kenneth W. Dam, Extra-territoriality in an Age of Globalization: The Hartford Fire Case, 1993 Sup.Ct. Rev. 289, 290-92. But this is one of the few cases to date that involves an import market and presents such a mixed fact pattern, and neither the statutory scheme nor the case law provides clear guidance.

Although the Supreme Court noted in Hartford Fire, 509 U.S. at 796, 113 S. Ct. 2891, that the substantial-effect test applies to "foreign conduct," no antitrust statute defines "foreign conduct." Nor does any statute explicitly address any aspect of a case involving the effect of foreign conduct on United States import commerce, like that at issue here.

To be sure, Congress legislated in this area relatively recently, establishing a threshold jurisdictional standard for "conduct involving trade or commerce (other than import trade or import commerce) with foreign nations." Foreign Trade Antitrust Improvements Act (FTAIA) of 1982 § 402, Pub.L. 97-290, 96 Stat. 1246 (codified at 15 U.S.C.A. § 6a). According to that statutory standard, United States antitrust laws do not apply to such conduct unless it has "a direct, substantial and reasonably foreseeable effect" on United States commerce. Id. Because this case involves importation of foreign-made goods, however — conduct Congress expressly exempted from FTAIA coverage as "involving ... import trade or import commerce... with foreign nations," id. — the FTAIA standard obviously does not directly govern this case, even though it may constitute an effort to "clarify the application of United States antitrust laws to foreign conduct" in other circumstances.3 See Den Norske Stats Oljeselskap AS v. HeereMac Vof, 241 F.3d 420, 421 (5th Cir.2001); Hartford Fire, 509 U.S. at 796 n. 23, 113 S. Ct. 2891 (refusing to decide whether the FTAIA applies to the conduct alleged, an agreement between foreign and United States participants, formed abroad, to refuse to sell a service to United States consumers); see also id. (refusing to decide whether the FTAIA's "direct, substantial, and reasonably foreseeable effect" provision "amends existing law or merely codifies it"); Kruman v. Christie's Int'l P.L.C., 284 F.3d 384, 399 n. 5 (2d Cir.2002) (same). Nor do we find any guidance in the FTAIA as to what constitutes "foreign conduct."

We thus must rely on case law, which provides some, albeit limited, assistance. Only a few cases give any hint of how to decide whether conduct is foreign.

In cases that date from the early twentieth century, silence on this point is hardly surprising, because the Supreme Court then rejected any exercise of jurisdiction based only on acts committed abroad, on the theory that the law of the country where an act occurred should govern it. See American Banana Co. v. United Fruit Co., 213 U.S. 347, 356, 29 S. Ct. 511, 53 L. Ed. 826 (1909); IA Areeda & Hovenkamp § 272b at 350-51 (describing post-American Banana Supreme Court cases in which jurisdiction depended on finding "actions within the United States").

In 1945, a leading opinion by Judge Learned Hand, which the Supreme Court later endorsed, displaced this approach based on the location of the conduct, and shifted attention to the location of the conduct's actual or intended effect. See United States v. Aluminum Co. (Alcoa) of America, 148 F.2d 416 (2d Cir.1945); IA Areeda & Hovenkamp § 272c; I Wilbur L. Fugate, Foreign Commerce and the Antitrust Laws § 2.10 at 68-69 (5th ed.1996). The case, Alcoa, thus introduced a predecessor of Hartford Fire's effects test. Although Alcoa refocused the jurisdictional analysis on the location of the effects of alleged violations of the antitrust laws, it did not eliminate the location of the conduct from the inquiry entirely, since its test applied only to cases involving foreign conduct.4 See, e.g., Kruman, 284 F.3d at 393-94 ("Under [Alcoa's] ... `effects test,' foreign conduct was actionable under our antitrust laws if it was intended to affect domestic commerce and actually did so." (emphasis added)). Yet Alcoa does not discuss how to assess whether conduct is "foreign," rather than domestic and therefore subject to McLain's test.

In the decades after Alcoa, courts and commentators analyzed not how to assess whether conduct is foreign, but how to interpret and apply Alcoa's effects test. They disagreed as to whether antitrust jurisdiction over foreign conduct requires both "an effect" on United States commerce and an "intent to affect United States commerce," or just effect or just intent, and as to the magnitude of any effect. See, e.g., Matsushita, 475 U.S. at 582 n. 6, 106 S. Ct. 1348 (characterizing the Sherman Act as "reach[ing] conduct outside our borders, but only when the conduct has an effect on American commerce") (emphasis added); Restatement (Third) of Foreign Relations Law § 415 (concluding that the Sherman Act applies to a foreign-made agreement with "a principal purpose ... to interfere with the commerce of the United States" and "some effect on that commerce" and to other foreign-made agreements with a "substantial effect" on United States commerce if "the exercise of jurisdiction is not unreasonable"); Fugate § 2.12 at 82 (proposing "the `direct and substantial' [effect] test, plus an element of intent" where "U.S. jurisdiction is based upon acts or agreements abroad which are not in the flow of foreign commerce" (emphasis omitted)); id. § 2.5 at 56, § 2.8 at 62 (suggesting and citing authority for a requirement of a "substantial effect"); IA Areeda & Hovenkamp § 272a at 350, § 272f at 354 (describing a minimal requirement of "significant effects"); see also Den Norske Stats Oljeselskap, 241 F.3d at 423-24 & n. 12 (describing the "general [ ] disagree[ment]," "assorted tests," and "confusing and unsettled" federal case law governing "the extraterritorial reach of the antitrust laws" before (and after) Hartford Fire). In promulgating these variants on the effects test, however, the authorities do not explore how to define "foreign conduct" in deciding whether to apply an effects test at all.

Although Hartford Fire itself does not discuss how to define "foreign conduct" either, it gives us some guidance by characterizing certain conduct as foreign. The complaint in Hartford Fire alleged multiple, partially overlapping conspiracies among reinsurers and their insurers (that is, insurers of reinsurers, also known as providers of retrocessional insurance) to limit the kinds of policies written for general corporate coverage in the United States market. See 509 U.S. at 770-78, 113 S. Ct. 2891. Three counts specifically alleged that companies in London agreed among themselves to refuse to reinsure policies covering the United States market unless the policies contained certain restrictions on liability. Two of these alleged conspiracies involved only international participants, see id. at 776, 778 & n. 7, 113 S. Ct. 2891, but one involved both international and United States participants. See id. at 775-77, 113 S. Ct. 2891.

It was in its discussion of these three counts that the Hartford Fire Court stated that "it is well established by now that the Sherman Act applies to foreign conduct that was meant to produce and did in fact produce some substantial effect in the United States." 509 U.S. at 796, 113 S. Ct. 2891. The Court thus plainly characterized the three counts it was then considering (including the count involving conspirators based in the United States) as "foreign conduct," but did not further define the term. The Court did not explain whether or how it had weighed the national affiliation or affiliations of the violators, the location of their acts, the location of the target market, or some combination of these factors in determining that the counts involved "foreign conduct."5

Dee-K seeks to rely on the Hartford Fire Court's silence on these issues to argue that it need not satisfy the substantial-effect test described in Hartford Fire. Specifically, Dee-K maintains that, even if the Hartford Fire statement does set forth minimal requirements for antitrust jurisdiction over foreign conduct, those requirements govern only antitrust violations involving "wholly" foreign conduct — and do not govern foreign conspiracies that result in the sale of price-fixed goods directly into United States commerce. Meanwhile, the producers argue that Hartford Fire directly controls here, without acknowledging any distinctions between the two cases. They insist that the only relevant consideration in assessing whether conduct is foreign is the location of actual conspiratorial meetings. We ultimately reject both sides' rigid views of Hartford Fire and apply a more nuanced analysis.

We begin with Dee-K's contention that the standard set forth in Hartford Fire does not govern this case because the conduct at issue here is not foreign enough. Dee-K offers two arguments in support of this contention. Neither is persuasive.

First, Dee-K relies on the fact that two of our sister circuits have characterized Hartford Fire as involving "wholly foreign" conduct. See Carpet Group Int'l v. Oriental Rug Importers Ass'n, 227 F.3d 62, 75 (3d Cir.2000); United States v. Nippon Paper Indus., 109 F.3d 1, 9 (1st Cir.1997). Dee-K maintains, in line with this characterization, that Hartford Fire does not govern a conspiracy like that at hand, with some domestic elements, including direct sales into the United States and one or more participants based in the United States (here the United States subsidiaries and distributors of the producers). Dee-K does not specify whether it is the nationality of the participants involved in the alleged antitrust violation, the location of their acts, the nature of their target market, or some combination of these factors that must be foreign to render their conduct "wholly foreign."

The Carpet Group and Nippon Paper courts did not state or suggest the origin of the phrase "wholly foreign." A possible source for the phrase is the legislative history of the FTAIA, the 1982 statute that required "a direct, substantial and reasonably foreseeable effect" on United States commerce for antitrust jurisdiction over conduct involving non-import foreign commerce. See 15 U.S.C.A. § 6a(1); H.R.Rep. No. 97-686, at 10 (1982), reprinted in 1982 U.S.C.C.A.N. 2487, 2495. A House Report explained that the FTAIA covers "wholly foreign transactions as well as export transactions" but not "import transactions." Id. In the FTAIA's legislative history, then, "wholly foreign transactions" appear to be those conducted entirely outside the United States — transactions that have no point of contact with this country. See also In re Ins. Antitrust Litig., 723 F. Supp. 464, 486 (N.D.Cal.1989) (characterizing the FTAIA as applicable to "wholly foreign commerce") (subsequent history, culminating in Hartford Fire, 509 U.S. 764, 113 S. Ct. 2891, 125 L. Ed. 2d 612, omitted). If this legislative history indeed constitutes the source of the phrase "wholly foreign," the history itself makes plain that the phrase does not apply to cases like ours, which involve "import transactions." Dee-K derives little benefit, however, from this possible origin of the phrase, because based on the legislative history, "wholly foreign" cannot be an appropriate characterization of the transactions at issue in Hartford Fire itself — which involved a refusal to sell to United States consumers. Thus, if the FTAIA's legislative history is the source of the term "wholly foreign," a court certainly should not use the term in deciding whether to apply the Hartford Fire test.

Carpet Group and Nippon Paper's characterization of Hartford Fire as involving "wholly foreign" conduct may, however, simply have been a shorthand description of the case, for in neither Carpet Group nor Nippon Paper does the holding turn on this characterization or otherwise assist Dee-K. Indeed, the analysis of the Third Circuit in Carpet Group is directly contrary to that suggested by Dee-K. The Third Circuit did refuse to apply the Hartford Fire test to a conspiracy among United States and foreign entities to protect the role of distributors in the United States market for imported carpets. However, the Carpet Group court did not simply point to the existence of some domestic contacts and end its analysis, as Dee-K would have us do. Rather, before applying the McLain domestic-conduct jurisdictional standard, the Carpet Group court determined that the alleged conduct "deal[t] primarily with conduct in the United States," because of the United States location of most participants, several targets of their pressure, and several of their meetings. 227 F.3d at 64-68, 75. Thus, looking at participants' affiliations and the locations of both acts and targets, the court focused on whether the conduct at issue was "primarily" domestic or foreign, not whether it was "wholly" domestic or foreign.

Furthermore, and perhaps most importantly, characterization of a prior case by a lower court cannot revise or amend the case's language or facts. Hence, even if Carpet Group or Nippon Paper had held that Hartford Fire controlled only "wholly foreign" conduct, that could not change the language or facts of Hartford Fire itself. In Hartford Fire, although the district court had used the term "wholly foreign" in a related context, see In re Ins. Antitrust Litig., 723 F.Supp. at 486, the Supreme Court never described the conduct before it as "wholly" foreign. Moreover, some of the participants in one of the conspiracies alleged in Hartford Fire were entities based in the United States, see 509 U.S. at 775, 776, 795, 113 S. Ct. 2891, all of the London-based corporate defendants were "subsidiaries of American corporations," and at least one key meeting attended by London-based defendants occurred in New York City. See In re Ins. Antitrust Litig., 938 F.2d 919, 922-23, 929 (9th Cir.1991), rev'd on other grounds sub nom. Hartford Fire, 509 U.S. 764, 113 S. Ct. 2891, 125 L. Ed. 2d 612. Thus, Hartford Fire itself severely undercuts Dee-K's claim that the substantial-effect test applies only to "wholly" foreign conduct, and Dee-K has shown no other reason to limit Hartford Fire to such conduct. But see IA Areeda & Hovenkamp § 272e at 354 (hypothesizing that the Sherman Act might reach even conduct with "only remote[] and insignificant[]" effects on the United States and no intent to affect the United States if "American firms participate" or "some planning or other conduct" occurs here).

Dee-K's other foreign-conduct argument, although reiterated several times, fares no better. Dee-K repeatedly contends that whenever conspirators "sell price-fixed goods directly into U.S. commerce," the conspiracy must be regarded as domestic rather than foreign conduct, because these goods by definition are then "in" United States commerce. This single-factor test — under which even a conspiracy hatched and planned entirely on foreign soil by foreign entities would be domestic if it resulted in direct sales of price-fixed goods into a United States import market — comes perilously close to clear conflict with Hartford Fire. After all, there the Supreme Court characterized as "foreign conduct" a conspiracy to refuse to sell reinsurance policies directly into United States markets. The Hartford Fire conspiracy had a direct, anticompetitive impact on a United States import market — just like the direct anticompetitive impact that Dee-K attributes to the rubber-thread price-fixing conspiracy.

Dee-K apparently hopes to distinguish Hartford Fire simply because that case concerned a refusal to sell to, rather than a sale to, purchasers in the United States. We acknowledge that in some circumstances this factual distinction may have significance.6 When we consider actual harm to United States consumers, however, we see no distinction between the impact of refusing to sell and that of fixing the price of imported goods, both of which may seriously and directly harm United States consumers. Both situations involve harmful conduct affecting a transaction in which a United States participant takes place.

In refusal cases, courts have not hesitated to require antitrust plaintiffs to allege and prove a substantial effect on the United States. In addition to the Hartford Fire Court itself, the Ninth Circuit required such proof in a case involving allegations that competitors had divided up the right to sell directly to United States consumers, even though market division rings the same alarm bells as price fixing; indeed, both constitute per se Sherman Act violations. See Metro Indus., 82 F.3d at 841, 843-44.

Moreover, given complex global trade patterns of sale, resale, and distribution, Dee-K's formulation would grant United States courts jurisdiction over a great variety of foreign conduct. Cf. Den Norske Stats Oljeselskap, 241 F.3d 420 (rejecting the existence of any form of antitrust jurisdiction over a foreign conspiracy even though it assertedly indirectly raised prices in a United States import market, in part because the plaintiff was not a buyer or seller in that market). For example, in every case involving direct sales to the United States in which our antitrust laws condemn an activity per se, however foreign the conduct, United States courts would have jurisdiction without any showing whatsoever of an effect on United States commerce. See, e.g., Stephen Jay Photography, Ltd. v. Olan Mills, Inc., 903 F.2d 988, 995 (4th Cir.1990) (noting that explicit horizontal minimum price fixing is a per se violation of the Sherman Act). Yet the rationale for per se rules in cases addressing domestic conduct seems plainly "inapplicable to foreign restraints that ... pose very little danger to American commerce." IA Areeda & Hovenkamp § 273 at 370-71 (noting that "price fixing in a foreign country might have some but very little impact on United States commerce," and concluding that "appraising restraints abroad requires an assessment of effects on American foreign commerce").

Perhaps for this reason, courts have consistently required a showing of effect on United States commerce even in cases involving price fixing on imports. Before Hartford Fire, the Seventh Circuit considered a conspiracy among "twenty domestic and nine foreign corporations ... to fix the price of uranium in the world market" at meetings in five foreign countries and the United States. See In re Uranium Antitrust Litig., 617 F.2d 1248, 1254 (7th Cir. 1980). In that case the Seventh Circuit applied Alcoa's effects test, although the domestic participants greatly outnumbered the foreign participants, a meeting occurred in the United States, and the target market included the United States. See id. at 1253-54. Similarly, in Nippon Paper, on which Dee-K heavily relies, the First Circuit had no trouble treating a criminal conspiracy resulting in price fixing of goods imported into the United States as "foreign conduct" requiring proof of a substantial effect on United States commerce. Nippon Paper, 109 F.3d at 2.7

Finally, the Executive Branch has endorsed the view that plaintiffs alleging antitrust violations affecting imports into a United States market must still prove a substantial effect on United States commerce. See U.S. Dep't of Justice & Fed. Trade Comm'n, Antitrust Enforcement Guidelines for International Operations § 3.11 (1995), available at http://www.usdoj.gov/atr/public/guidelines/internat.htm (as of July 2, 2002) (advising that "[i]mports into the United States by definition affect the U.S. domestic market directly" but that they yet must "produce the requisite substantial effects," while commenting that effects are likely to be "clear" in such a case).

For these reasons, we reject Dee-K's contention that Hartford Fire never applies when a conspiracy involves price-fixed goods sold directly into United States commerce.

We likewise reject the producers' equally rigid view as to the applicability of Hartford Fire. The producers contend that in assessing whether "conduct" is "foreign" (and thus subject to the substantial-effect, rather than the domestic McLain test), we should consider only the location of acts that themselves independently violate the Sherman Act. According to the producers, "the Hartford Fire test ... is based solely upon the situs of a conspiratorial agreement." The producers thus turn Dee-K's argument back upon it: even if Hartford Fire does apply just to "wholly foreign conduct," the only relevant "conduct" is the conspiratorial meetings themselves. In this case, they argue, we should therefore focus only on the location of the meetings at which the conspirators agreed to fix prices — and all of those meetings were held outside the United States.

The producers' suggested rule eliminates any examination of the national affiliation of the participants in the conduct, the scope and location of the target market, and the location of all acts other than the actual conspiratorial acts. This rigidity troubles us. A noted commentary has warned against an overly rigid approach in this area of antitrust law: "to say that domestic per se rules are not necessarily and automatically applicable in the international context is not to say that an antitrust court needs to hesitate very long before condemning restraints with significant and obvious effects on United States commerce, and without any plausible purpose other than the suppression of competition with and in the United States." IA Areeda & Hovenkamp § 273 at 371. In short, the flexibility that permits a court not to apply domestic rules to a defendant engaging in foreign conduct should be balanced by a court's ability to allow a plaintiff to pursue foreign conduct that does harm United States interests — even if the defendants scrupulously entered into their illegal agreements only outside the United States.

The suggestion that we consider only the location of certain acts concerns us particularly given that ever since Alcoa, the Supreme Court's jurisdictional analysis has emphasized above all else the effects, i.e., the intended location, actual location, and magnitude of those effects. See, e.g., Kruman, 284 F.3d at 395. Quite simply, the Supreme Court has moved away from its earlier doctrine focused solely on the location of acts. We therefore doubt whether a jurisdictional rule focused exclusively on defining and locating a certain limited set of acts could possibly be appropriate.

The potential consequences of the rule that the producers champion only confirm our doubts. If, for example, executives of United States corporations that did business solely in the United States met abroad simply to consummate a price-fixing conspiracy that was carried out entirely within the United States, the producers' proposed rule would still require special proof of the conspiracy's effect before a United States court could assert jurisdiction.

We reject this overly narrow view. McLain might well apply to such a conspiracy, even though the acts that constituted the illegal agreement took place exclusively abroad. Conversely, as the Second Circuit has noted, meetings in the United States to negotiate "an agreement to fix prices in an overseas foreign market that had no effect on domestic commerce" do not yield antitrust jurisdiction in our courts. See Kruman, 284 F.3d at 395. The universe of factors relevant to our jurisdictional analysis must be greater than the location of the specific conspiratorial acts alone.

Instead of the parties' bright-line rules, we believe a court should properly engage in a more flexible and subtle inquiry. In determining which jurisdictional test (Hartford Fire or McLain) applies, a court should consider whether the participants, acts, targets, and effects involved in an asserted antitrust violation are primarily foreign or primarily domestic.

This inquiry will best accommodate the cases with mixed fact patterns, defying ready categorization as "foreign" or "domestic" conduct, which our increasingly global economy will undoubtedly produce. We cannot begin to foresee the scope or complexity of future transactions. To adopt the simplistic rules the parties favor might well yield unintended and unfortunate results.

Moreover, this area of antitrust law has historically been marked by change, and remains a subject of serious debate. Not only the courts, but also numerous commentators and even representatives of various other countries (as frequent amici), have presented differing views, focusing on various elements of conduct as crucial in addressing whether assertion of jurisdiction is appropriate. Given these concerns, we believe courts should remain able to consider the full range of factors that may appropriately affect the exercise of our antitrust jurisdiction in any given case.

We note that this approach echoes that of the Third Circuit in Carpet Group and finds support in several of the treatises. See Carpet Group, 227 F.3d at 75; Fugate § 2.12 at 80-82 (discussing the location of "acts and agreements," the location of their effects, and the nationality of participants in considering jurisdiction); Restatement (Third) of Foreign Relations Law § 415(2) (applying the same proposed jurisdictional test to "[a]ny agreement in restraint of United States trade that is made outside of the United States, and any conduct or agreement in restraint of such trade that is carried out predominantly outside of the United States" (emphasis added)); cf. Montreal Trading Ltd. v. Amax, Inc., 661 F.2d 864, 869 (10th Cir.1981) ("When the contacts with the United States are few, the effects upon American commerce minimal, and the foreign elements overwhelming, ... we do not accept jurisdiction.").

In this case, the bulk of the conduct related to the global conspiracy alleged in the complaint and proved at trial occurred abroad. As in Nippon Paper, the agreements here were all formed entirely outside the United States — in several other countries. The target of the conspiracy was a global market, and the participants monitored prices in many markets around the world. Although dozens of people participated in the meetings over the years, Dee-K can only point to two who held office in United States companies, and both of them also had important and in fact primary roles in Southeast Asian companies.

To be sure, the producers sold rubber thread in the United States to United States consumers. In particular, Heveafil sold through a United States division of the parent producer, directing pricing by communications from Southeast Asia, while Rubfil sold through a wholly owned United States subsidiary and co-defendant. In addition, a handful of faxes from the producers describe pricing in the United States. It further appears that a United States distributor was made aware, by a fax sent "FYI," of a group commitment not to hire an executive whom one producer had fired for his opposition to price increases.

But these links to the United States are mere drops in the sea of conduct that occurred in Southeast Asia (and around the world). See In re Uranium Antitrust Litig., 617 F.2d at 1254 (applying Alcoa's effects test in a price-fixing case even though twenty participants were domestic and only nine foreign and the United States was part of the target market). Just as the Third Circuit looked at a mass of conduct and concluded it was "primarily" domestic in Carpet Group, 227 F.3d at 75, so examining all of the conduct here, we can only conclude that it is primarily foreign. The handful of contacts with the United States that Dee-K cites cannot change the fact that the rubber-thread conspiracy was formulated and furthered at numerous meetings, all of which took place in Southeast Asia, with attendees, all of whom worked for Southeast Asian firms, and who directed their activity to the global market. Only a few of the participants had any United States affiliation, and each of those also had a primary Southeast Asian affiliation. In addition, although we have explained that we do not find this factor alone dispositive, we consider it significant that not one of the conspirators' many meetings took place in the United States. In light of all of these factors, we conclude that the price-fixing conspiracy Dee-K alleged and proved was primarily "foreign conduct" to which the Hartford Fire test properly applied.

In closing, we note that even when a conspiracy constitutes "foreign conduct," an antitrust plaintiff still has access to the federal courts to challenge it. The district court did not dismiss Dee-K's lawsuit because its claim involved foreign conduct. The court merely required Dee-K to prove that the foreign conduct had a substantial effect on United States commerce in order to recover. Other litigants who claim that foreign conduct, which violates our antitrust laws, has harmed them may also have their day in court, and the federal courts will provide redress for those who can show that the harm of which they complain had a substantial effect on our commerce. For Dee-K, that day in court has come and gone.

For the reasons given herein, the judgment of the district court is

AFFIRMED.