This certified interlocutory appeal from a pretrial order of the district court raises an issue that currently is the subject of much debate: In an action for treble damages under the antitrust and antidumping laws, do the parties have a right to trial by jury without regard to the practical ability of a jury to decide the case properly?

This litigation began in the District of New Jersey with the complaint of National Union Electric Corp. (NUE). A corporate successor to the Emerson Radio Co., NUE was a major domestic producer of television receivers until February 1970. The following December, it filed the first complaint of this litigation, charging several of its Japanese competitors with violations of the antitrust laws and the laws governing competition in international trade. The complaint names as defendants the Mitsubishi Corp., which is a Japanese trading company, and seven Japanese television manufacturers: Matsushita Electric Industrial Co., Toshiba Corp., Hitachi, Ltd., Sharp Corp., Mitsubishi Electric Corp., Sanyo Electric Co., and Sony Corp. Nine subsidiaries of these companies also are named as defendants in NUE's action.1

NUE alleges that the defendants have sought to drive American television producers out of the American market by selling televisions at artificially depressed prices. Charging that defendants have maintained lower prices for televisions sold in the United States than for comparable televisions sold in Japan, NUE asserts violations of the 1916 Antidumping Act, 15 U.S.C. § 72 (1976). NUE further alleges that these dumping practices are part of a large conspiracy in which defendants have agreed among themselves and have acted in concert with over 90 coconspirators around the world to maintain artificially low prices for Japanese televisions sold in the United States. Defendants allegedly have facilitated and financed this scheme by fixing high prices for televisions sold in Japan, a practice made possible by concerted action and protection afforded by the Japanese government. NUE asserts that this conduct violates §§ 1 and 2 of the Sherman Act, 15 U.S.C. §§ 1, 2 (1976), and § 73 of the Wilson Tariff Act, 15 U.S.C. § 8 (1976), which applies the basic proscriptions of the Sherman Act to trade in imports.

NUE seeks treble damages for injuries sustained between 1966 and 1970. The remedy of treble damages is made available for the dumping claims by the Antidumping Act itself, 15 U.S.C. § 72 (1976), and for the Sherman and Wilson Tariff Act claims by § 4 of the Clayton Act, 15 U.S.C. § 15 (1976). NUE also seeks injunctive relief.

Zenith Radio Corp., a major domestic producer of consumer electronic products, filed the second complaint of this litigation in 1974 in the Eastern District of Pennsylvania. The Zenith complaint named all of the defendants of the NUE action, a few additional subsidiaries,2 and two American companies: Motorola, Inc., and Sears, Roebuck, and Co. Two of the common defendants, Sony Corp. and its American sales subsidiary, Sony Corp. of America, subsequently reached a complete settlement with Zenith. They are now defendants only in the NUE action.

The Zenith complaint repeats NUE's allegations of dumping, conspiracy, and intent to destroy domestic competition in the American market, but Zenith's allegations are broader in two respects. First, Zenith seeks damages for injuries sustained over a longer period, from 1968 through 1977, as opposed to a period of 1966 through 1970 in NUE's complaint. Second, Zenith's allegations cover not only televisions but also radios, phonographs, tape and audio equipment, and electronic components. In addition, Zenith charges defendants with discriminating in price among American purchasers, in violation of the Robinson-Patman Act, 15 U.S.C. § 13(a) (1976). Finally, Zenith asserts that the Matsushita and Sanyo defendants have violated § 7 of the Clayton Act, 15 U.S.C. § 18 (1976), by acquiring interests in domestic producers of consumer electronic products previously held by Motorola and Sears. Like NUE, Zenith prays for treble damages and injunctive relief.

A group of the Japanese defendants in the Zenith action filed two counterclaims. The first charges Zenith and its distributors throughout the United States with territorial allocations, horizontal and vertical price-fixing schemes, "key dealer preferences," and price discrimination, in violation of §§ 1 and 2 of the Sherman Act and the Robinson-Patman Act, 15 U.S.C. §§ 1, 2, 13(a) (1976). The second counterclaim charges Zenith and about 30 coconspirators with maintaining a program of sham litigation against Zenith's competitors. See Otter Tail Power Co. v. United States, 410 U.S. 366, 379-80, 93 S. Ct. 1022, 35 L. Ed. 2d 359 (1973).

Sears filed a separate counterclaim challenging Zenith's advertising claims that Zenith color televisions are manufactured in the United States. Sears claims that the advertisements create an impression that all components of Zenith's color televisions and other consumer electronic products are of American origin, when some components are manufactured abroad. Sears asserts a violation of § 43 of the Lanham Act, 15 U.S.C. § 1125 (1976), which prohibits false designations of origin, and prays for damages and injunctive relief.

Shortly after the filing of the Zenith action, the two suits were consolidated for pretrial proceedings in the Eastern District of Pennsylvania. In re Japanese Electronic Products Antitrust Litigation, 388 F. Supp. 565 (Jud.Pan.Mult.Lit.1975). Subsequently, the district court, on NUE's motion, consolidated the two suits for trial. The court denied the motion of the Sony defendants for a separate trial on NUE's claims against them. Zenith Radio Corp. v. Matsushita Electric Industrial Co., No. 75-2451 (E.D.Pa. June 6, 1979) (Pretrial Order No. 182).

Both NUE and Zenith made timely demands for jury trial. Fourteen of the defendants moved to strike the demands, arguing that the case is too large and complex for a jury.3 The district court denied their motion, concluding that the seventh amendment does not recognize the complexity of a lawsuit as a valid reason for denying a jury trial. The court explained its reasoning in a thorough and scholarly opinion, Zenith Radio Corp. v. Matsushita Electric Industrial Corp., 478 F. Supp. 889 (E.D.Pa.1979), and certified its order for interlocutory appeal under 28 U.S.C. § 1292(b) (1976). See 478 F. Supp. at 942-46. We have permitted the appeal to determine whether the district court's reason for denying appellants' motion was correct.

Appellants argue that the proof of the foregoing claims will be too burdensome and complicated for a jury. They have cited several dimensions of complexity.

The district court accepted one of appellants' basic contentions: the trial will be protracted. The court predicted that the trial would last a full year. It noted that the parties are nearing the end of discovery, which after nine years has produced millions of documents and over 100,000 pages of depositions. The court did not estimate how much of this evidence will be introduced at trial.

Beyond these observations of the district court, we have only the parties' divergent predictions of the proof that appellees' claims call for. We understand their primary disagreements to concern four general sources of complexity: proof of the Antidumping Act claims, proof of the alleged conspiracy, resolution of a number of financial issues, and understanding of several conceptually difficult legal and factual issues.

Under the Antidumping Act, appellees must prove that the defendants made sales of articles in the United States at a price lower than the price of "such articles" in Japan. 15 U.S.C. § 72 (1976).4 Appellants read the Act to permit price comparisons only for identical products sold in the two countries. During the relevant periods, defendants produced thousands of technically distinct models of the products covered by this litigation. They contend that to identify the products appropriate for price comparisons, the jury will have to review the technical features of thousands of different models and understand how differences between the models relate to cost of manufacture, product performance, and marketability. Appellees construe the Antidumping Act to permit price comparisons between functionally equivalent products, such as all portable color televisions with particular screen size and VHF-UHF channel selection. They contend that a jury could identify such functionally equivalent products without massive or highly technical proof.

The conspiracy charged in this suit is massive. Appellees allege that it has lasted for at least 30 years, involved almost 100 firms around the world, and affected international trade in several consumer electronic products. Appellants argue that litigation of the existence and operation of this conspiracy will produce an enormous amount of evidence for the jury to consider. They see further difficulties in the fact that the alleged conspiracy involved Japanese businessmen and that its operations included restraint of trade in Japanese markets. Appellants fear that a jury might not understand the evidence due to the difficulty of understanding business practices and market conditions in Japan. Appellees respond that proof of the conspiracy and its operations will be simple because the facts are well established in unambiguous documentation. Appellees foresee no difficulties due to allegations involving Japan. They characterize the alleged conspiracy as "classic," much like combinations ordinarily revealed in Sherman Act cases.

Some parts of the case will require the jury to resolve a series of financial issues. Appellants have highlighted three such parts. First, for the Antidumping Act claims, the jury will have to decide whether the price of an article sold in the United States is "substantially less than the actual market value or wholesale price" in Japan and whether a defendant has maintained differential pricing "commonly and systematically." 15 U.S.C. § 72 (1976). This inquiry may be complicated by several influences on prices that might have to be factored out before comparing prices, such as currency fluctuations and different marketing techniques in the two countries. Second, appellees allege that the conspirators disguised their artificially low prices in the United States by a series of complicated rebate schemes. Appellants say that the jury will be able to test this allegation only by reviewing the circumstances surrounding thousands of separate transactions. Third, appellees intend to show injury by proving that they lowered their own prices in response to defendants' artificially low prices and that they lost sales to defendants. These allegations will require evidence of appellees' transactions and may raise issues regarding appellees' pricing policies and marketing techniques and the quality of appellees' products.

Appellants contend that litigation of these three parts of the case will produce an enormous mass of financial documentation for the jury to work through. They also contend that the jury will need the assistance of substantial amounts of expert testimony on accounting, marketing, and other technical matters. Appellees reject this prediction, arguing that all the relevant financial evidence can be submitted neatly in computer printouts with accompanying summaries. They do not foresee great problems in the jury's understanding of the evidence.

Finally, appellants argue that the complexity of the suit will be compounded by the presence of some issues that conceptually are very difficult. The claims under both the Antidumping Act and § 2 of the Sherman Act will require proof of predatory intent. 15 U.S.C. § 72 (1976); United States v. Aluminum Co. of America, 148 F.2d 416 (2d Cir. 1945). On the § 2 claims and on Zenith's claims under § 7 of the Clayton Act, appellees will have to prove relevant product markets, relevant geographic markets, and market shares. See Brown Shoe Co. v. United States, 370 U.S. 294, 82 S. Ct. 1502, 8 L. Ed. 2d 510 (1962); United States v. Aluminum Co. of America, supra. Zenith's claims under the Robinson-Patman Act will raise issues of whether products sold to different customers are of a "like grade and quality" and whether any price differences are cost justified. 15 U.S.C. § 13(a) (1976).

The district court did not resolve these and other specific points of dispute between the parties, except to say, "We expect that the actual size of complexity of this litigation falls somewhere in between the two extremes portrayed by the parties." 478 F. Supp. at 899. The court added, "By any yardstick, this case is at least as large and complex as the others in which jury demands have been struck (on grounds of complexity)." Id. It then proceeded to the conclusion that the seventh amendment preserves the right to jury trial in this suit regardless of its complexity. Id. at 942. That construction of the seventh amendment is the focus of this appeal.

Appellees offer a statutory ground for affirming the district court. They assert that the Clayton Act grants a right to jury trial even if the extraordinary complexity of the suit renders the seventh amendment guarantee inapplicable. In other words, they argue for a statutory right to jury trial in suits where the seventh amendment might not guarantee the right. Because this court should avoid unnecessary issues of constitutional interpretation, we shall consider this argument first.

Appellees rely first on the language of the Clayton Act. Although the Act makes no mention of jury trials or of any trial procedures, appellees contend that the right is implicit in the language of § 4 that any person who is injured by a violation of the antitrust laws "may sue therefore . . . and shall recover threefold damages." 15 U.S.C. § 15 (1976).

We do not read this language to have the broad implications suggested by appellees. The only association between the remedy provided in the Clayton Act and jury trials is incidental: Prevailing rules of practice at the time of the enactment of the Clayton Act almost surely would have provided a jury trial in all suits for treble damages under the Act. Courts of equity considered treble damages punitive in nature and generally refused to award punitive relief.5 See Decorative Stone Co. v. Building Trades Council, 23 F.2d 426 (2d Cir.), cert. denied, 277 U.S. 594, 48 S. Ct. 530, 72 L. Ed. 1005 (1928). Consequently, suits for treble damages were confined to actions at law, and the seventh amendment was read to guarantee a right to jury trial. However, nothing in the Clayton Act suggests that Congress intended to write this incident of the procedural rules of 1914 into the statute. It is more likely that Congress would allow the courts to try suits for treble damages in accordance with subsequent developments in procedural rules derived from the Constitution and from court rules. Therefore, we do not read the authorization of suits for "threefold damages" to imply a right to jury trial.

Our reasoning is consistent with Lorillard v. Pons, 434 U.S. 575, 98 S. Ct. 866, 55 L. Ed. 2d 40 (1978), in which the Supreme Court found an implicit guarantee of jury trials in the section of the Age Discrimination in Employment Act authorizing private suits to recover lost wages. § 7(c), 29 U.S.C. § 626(c) (1976). The Court found "a significant indication of Congress' intent in its directive that the ADEA be enforced in accordance with the 'powers, remedies, and procedures ' of the FLSA (Fair Labor Standards Act)." 434 U.S. at 580, 98 S.Ct. at 869 (emphasis supplied by Supreme Court) (quoting § 7(b), 29 U.S.C. § 626(b)). Noting that the right to jury trial in private actions under the FLSA was well established, the Court read this directive as incorporating that right into the ADEA. Id. at 580-83, 98 S.Ct. at 869-71. The Clayton Act contains no similar directive to incorporate a specific enforcement scheme that includes jury trials. Indeed the lack of any reference to trial procedures in the Clayton Act, in contrast to the ADEA, strongly suggests that Congress never understood the provisions of the Act to prescribe any of the procedures to be followed in a private enforcement action.

As a second source of support for their statutory argument, appellees have cited several portions of the legislative histories of the antitrust laws indicating that members of the enacting congresses expected that suits for treble damages would be tried to juries. Appellees have cited passages from the debate on the Clayton Act's provisions for treble damage actions and from the debate on a similarly worded provision in the Sherman Act. Ch. 647, § 7, 26 Stat. 210 (1890).

During the discussion of the Clayton Act in the House of Representatives, some members objected to a provision that would make a court's findings in a government suit for an injunction conclusive in a subsequent private enforcement action. Cf. Clayton Act, § 5(a), 15 U.S.C. § 16(a) (1976) (enacted version established such findings as prima facie proof). The effect of this provision, they said, would be a denial of the right to jury trial in a subsequent suit for treble damages. However, the congressmen who spoke on the subject state plainly that this right to jury trial derives from the seventh amendment. Their remarks suggest no understanding of the proposed Clayton Act as establishing a statutory right to jury trial. 51 Cong.Rec. 9488-91 (1914) (remarks of Reps. Scott, Floyd, Volstead, and Green).

The Senate's debate on the earlier Sherman Act provision contains a few passages in which a senator mentions, in the course of discussing some other issue, that juries will hear treble damage actions. One senator pointed to the seventh amendment as the basis for his assumption that trials would be before juries. 21 Cong.Rec. 2643 (1890) (remarks of Sen. Gray). The others made no mention of the basis for their assumptions. Id. 1767, 3150 (remarks of Sen. George); id. 3149 (remarks of Sen. Morgen). These passing remarks in the Senate's debate do not allow an inference that Congress intended that the Sherman Act itself would guarantee jury trials where the seventh amendment would not.

In short, the legislative history indicates nothing more than the expectation of several congressmen that the seventh amendment generally would guarantee a right to jury trial in treble damage actions under the antitrust laws. We are unable to translate their expectations of the ordinary application of the seventh amendment into an intention to require jury trial by statute.

Appellees seek further support in a circuit court opinion in a treble damage action under the Sherman Act. Meeker v. Lehigh Valley R. R., 162 F. 354 (C.C.S.D.N.Y.1908). Appellees read this opinion as finding a jury trial requirement in the Sherman Act. They argue that Congress' enactment of the Sherman Act's treble damage suit provision in the Clayton Act constituted legislative approval of this interpretation. We are not persuaded. The entire discussion of the jury trial issue in the Meeker opinion consists of one sentence: "This is an action at law, and the parties are entitled to a jury trial." Id. at 357. We read this sentence not as an interpretation of the statute but as an application of the seventh amendment. See part IV infra. Even if our reading is incorrect, the status of Meeker as a case of statutory interpretation is so vague that we cannot conclude that Congress tacitly approved of it as an interpretation of the Sherman Act.

Finally, appellees argue that the Supreme Court recognized a statutory right to jury trial in Fleitmann v. Welsbach Street Lighting Co., 240 U.S. 27, 36 S. Ct. 233, 60 L. Ed. 505 (1916). A shareholder of a corporation allegedly injured by a Sherman Act violation brought a derivative suit, under the Sherman Act,6 seeking treble damages for the corporation. Because the shareholder could file a derivative action only in equity, this manner of proceeding denied the defendant a right to a jury trial. The Court held that the statute did not authorize a suit in equity:

(W)e agree with the courts below that when a penalty of triple damages is sought to be inflicted, the statute should not be read as attempting to authorize liability to be enforced otherwise than through the verdict of a jury in a court of common law. On the contrary it plainly provides the latter remedy and it provides no other.

Id. at 29, 36 S.Ct. at 234. The Court affirmed the dismissal of the suit.

We do not read Fleitmann to recognize a statutory right to jury trials in suits not subject to the seventh amendment. The Court's ruling is a restriction on the scope of relief authorized by the Sherman Act: the Act did not authorize suits in equity for treble damages. The reason for this narrow construction of the statute was to preserve defendants' right to a jury trial. However, having confined suits for treble damages to actions at law, the Court did not rule that the right to jury trial in these legal suits derived from the Sherman Act rather than from the seventh amendment. The opinion of the Court of Appeals in Fleitmann states: "(A)n action to recover treble damages under section 7 of the (Sherman Act) must be an action at law, where the defendants have the constitutional right to a jury trial." Fleitmann v. United Gas Improvement Co., 211 F. 103, 105 (2d Cir. 1914) (emphasis added). The Supreme Court's opinion adds nothing on this point except to express general agreement with the Court of Appeals. 240 U.S. at 29, 36 S. Ct. at 234. Therefore, in a suit for treble damages that is not triable at law and hence not subject to the seventh amendment guarantee of jury trial, the mandate of Fleitmann is not, as appellees suggest, to try the suit to a jury on the authority of the statute. It is to dismiss the suit. Cf. United Copper Securities Co. v. Amalgamated Copper Co., 244 U.S. 261, 37 S. Ct. 509, 61 L. Ed. 1119 (1917) (Sherman Act did not authorize trial of derivative actions at law and before a jury). Fleitmann does not recognize the statutory right for which appellees argue.

Our reading of Fleitmann raises an important collateral issue: If appellants are correct in arguing that the seventh amendment permits trial of this suit without a jury because of its extraordinary complexity, must the complaints be dismissed for failing to state a cause of action? Although the broad language in Fleitmann quoted above might seem to require that result, we conclude that dismissal is not necessarily mandated.

We think it fair to say that when the Supreme Court decided Fleitmann it assumed that the seventh amendment would guarantee a right to jury trial in all treble damage actions tried at law. At that time, none of the recognized circumstances in which a court of law could deny a jury trial would arise in these actions. See, e. g., Bauman v. Ross, 167 U.S. 548, 593, 17 S. Ct. 966, 983, 42 L. Ed. 270 (1897) (suits to recover compensation for property taken for public use); McElrath v. United States, 120 U.S. 426, 26 L. Ed. 189 (1880) (suits at law against the United States). Thus, the Court did not specifically face the issue of whether the Sherman Act permitted a suit for treble damages at law without a jury if the Constitution would not require a jury.

Appellants offer two grounds for ruling that the seventh amendment does not guarantee a jury trial in this case, one of which does not depend upon characterizing the suit as equitable. That is the argument based on the due process clause. See Parts VI & VII infra. If this limitation on the seventh amendment is valid, Fleitmann's rule against suits in equity would not require dismissal. Given our conclusion that the language and legislative history of the Clayton Act do not reveal any intention to prescribe the mode of trial in private treble damage actions, we see no reason why the Clayton Act would not permit trial of such a case to the court. In fact, it would be most irrational if antitrust defendants could escape liability entirely by making their illegal activities too difficult to allow trial by jury.

In summary, we are unable to find a statutory right to jury trial that applies where the seventh amendment does not require a jury trial.7 The language of the Clayton Act alone is insufficient to permit an inference that Congress intended to provide for such a right. The legislative histories of the Clayton and Sherman Acts reveal no evidence of such an intent. Finally, no judicial decision supports appellees' statutory argument. Therefore, we must turn to the parties' constitutional arguments.

The seventh amendment provides in relevant part: "In Suits at common law ... the right of trial by jury shall be preserved."8 The Supreme Court's basic tool of construction in interpreting the amendment has been history: "The right of trial by jury thus preserved is the right which existed under the English common law when the Amendment was adopted." Baltimore & Carolina Line, Inc. v. Redman, 295 U.S. 654, 657, 55 S. Ct. 890, 891, 79 L. Ed. 1636 (1935).9 At times, the Court has determined that some aspects of the seventh amendment right, such as the validity of certain jury control techniques, should reflect the specific historical dimensions of the common law of 1791, the year of the amendment's adoption. See, e. g., Dimick v. Schiedt, 293 U.S. 474, 55 S. Ct. 296, 79 L. Ed. 603 (1935) (the seventh amendment does not permit a court to increase amount of a jury verdict).

However, the Court has never relied on this static view of history to confine the seventh amendment guarantee to causes of action recognized by the common law of 1791. Beginning with an 1830 decision, Parsons v. Bedford, 28 U.S. (3 Pet.) 433, 446-47, 7 L. Ed. 732 (1830), the Court has held the guarantee applicable to almost any suit, including a suit based on a statutorily created cause of action, that falls within the federal courts' jurisdiction over suits at law as opposed to suits in equity or admiralty. See Curtis v. Loether, 415 U.S. 189, 193-94, 94 S. Ct. 1005, 1007-08, 39 L. Ed. 2d 260 (1974).

Usually, courts have categorized suits among law, equity, and admiralty on the basis of the presence of a subject matter or a remedy peculiarly associated with one of the three jurisdictional categories. Thus, a suit to establish title to real property is normally considered legal because of the special competence that courts of law have exercised in these matters. See Whitehead v. Shattuck, 138 U.S. 146, 11 S. Ct. 276, 34 L. Ed. 873 (1891). Where such common indicia of law, equity, or admiralty have not been present, courts have classified a suit by comparing it with the particular actions that the courts of common law, chancery, or admiralty historically recognized as within their respective jurisdictions. See Pernell v. Southall Realty, 416 U.S. 363, 94 S. Ct. 1723, 40 L. Ed. 2d 198 (1974); Damsky v. Zavatt, 289 F.2d 46 (2d Cir. 1961).

Suits for treble damages under the antitrust and antidumping laws, as a class, are plainly legal in nature. They seek relief in a form traditionally associated with courts of law: compensatory and punitive damages for injuries caused by legal wrongs. See Curtis v. Loether, 415 U.S. 189, 195-96, 94 S. Ct. 1005, 1008-09, 39 L. Ed. 2d 260 (1974). Indeed, prior cases have always assumed that the seventh amendment guarantees a jury trial in antitrust suits. See Beacon Theatres, Inc. v. Westover, 359 U.S. 500, 504, 79 S. Ct. 948, 953, 3 L. Ed. 2d 988 (1959); Columbia Metal Culvert Co. v. Kaiser Aluminum & Chemical Corp., 579 F.2d 20, 23 (3d Cir.), cert. denied, 439 U.S. 876, 99 S. Ct. 214, 58 L. Ed. 2d 190 (1978).

Appellants dispute none of the foregoing and concede that a right to jury trial normally exists in suits for treble damages under the antitrust and antidumping laws. They argue that the seventh amendment does not guarantee a right to jury trial when any particular lawsuit, because of its extraordinary complexity, is beyond the ability of a jury to decide.

For the sake of clarity, we should state our understanding of complexity in this context. A suit is too complex for a jury when circumstances render the jury unable to decide in a proper manner. The law presumes that a jury will find facts and reach a verdict by rational means. It does not contemplate scientific precision but does contemplate a resolution of each issue on the basis of a fair and reasonable assessment of the evidence and a fair and reasonable application of the relevant legal rules. See Schulz v. Pennsylvania R. R. Co., 350 U.S. 523, 526, 76 S. Ct. 608, 610, 100 L. Ed. 668 (1956). A suit might be excessively complex as a result of any set of circumstances which singly or in combination render a jury unable to decide in the foregoing rational manner. Examples of such circumstances are an exceptionally long trial period and conceptually difficult factual issues.

Some district courts have recognized complexity as a grounds for denying jury trial.10 On the other hand, the Ninth Circuit recently has held that the seventh amendment applies without regard to a lawsuit's size or complexity. In re U. S. Financial Securities Antitrust Litigation, 609 F.2d 411 (9th Cir. 1979), cert. denied sub nom., Gant v. Union Bank, --- U.S. ----, 100 S. Ct. 1866, 64 L. Ed. 2d 281 (1980).

The Supreme Court has supplied direct support for appellants' position only in a footnote to its opinion in Ross v. Bernhard :

As our cases indicate, the "legal" nature of an issue is determined by considering, first, the pre-merger custom with reference to such questions; second, the remedy sought; and, third, the practical abilities and limitations of juries.

396 U.S. 531, 538 n. 10, 90 S. Ct. 733, 738, 24 L. Ed. 2d 729 (1970). The third prong of the test plainly recognizes the significance, for purposes of the seventh amendment, of the possibility that a suit may be too complex for a jury. Its inclusion in the three prong test strongly suggests that jury trial might not be guaranteed in extraordinarily complex cases, even though the Parsons v. Bedford line of Supreme Court cases, reflected in the first two prongs, would read the seventh amendment as applying to the suit.

The district court made no use of the Ross footnote, finding it too brief to authorize a major departure from the traditional construction of the seventh amendment. 478 F. Supp. at 926-30. We also find it unlikely that the Supreme Court would have announced an important new application of the seventh amendment in so cursory a fashion. Yet, at the very least, the Court has left open the possibility that the "practical abilities and limitation of juries" may limit the range of suits subject to the seventh amendment and has read its prior seventh amendment decisions as not precluding such a ruling. With this understanding of Ross, we shall consider the merits of appellants' arguments for a complexity exception.

Appellants' first argument relies on historical analysis to advance the proposition that the fact of extraordinary complexity renders a suit equitable in nature. Although complexity is not commonly recognized as a defining feature of equity, appellants argue that by the time of the adoption of the seventh amendment the chancellor's jurisdiction had extended to any suit that he found too complex for a jury. They have submitted a large body of historical authorities to support this proposition. The brief of International Business Machines Corporation, amicus curiae in this case, provides some additional historical authorities to support several of appellants' arguments.

Most of these authorities are suits seeking relief in the form of an accounting between the parties.11 The chancellor's jurisdiction over accounting actions consisted of two general categories. First, he had exclusive jurisdiction over any suit in which the substantive rules of equity imposed a duty to render an account. Second, he had jurisdiction when the accounts between the parties were too numerous and complicated for a common-law jury to unravel. Part of the rationale for equitable relief in this latter group of cases was that the difficulty of the jury's task made relief at law inadequate. See generally H. McClintock, Handbook of the Principles of Equity §§ 200-202 (1948). Appellants analogize complex antitrust and antidumping suits to actions for equitable accountings in this second group.

This analogy rests upon an incomplete statement of the chancellor's jurisdiction over complex accountings. Not every form of monetary liability at law could be characterized as a set of accounts for purposes of this jurisdiction. The chancellor typically ordered an accounting for money owing between persons with a particular legal relationship, like partnership or agency. In some cases, the chancellor ordered an accounting where no such relationship existed but money was owed as a result of contractual liability or fraud. See generally 2 J. Story, Commentaries on Equity Jurisprudence §§ 598-697 (14th ed. 1948); Langdell, A Brief Survey of Equity Jurisdiction IV, 2 Harv.L.Rev. 241 (1889). We are aware of no case, however, in which a chancellor ordered an accounting in a suit involving nothing more than liability for money damages in trespass or tort. Several American courts have held that an equitable accounting is not available in these circumstances. See United States v. Bitter Root Development Co., 200 U.S. 451, 470-71, 478-79, 26 S. Ct. 318, 324-327, 50 L. Ed. 550 (1906); Conklin v. Bush, 8 Pa. 514, 516-17 (1848); see Annot., Accounting in Equity in Case of Tort, 53 A.L.R. 815 (1928).

Suits for treble damages under the antitrust and antidumping laws are similar in form to suits for damages in tort. See Northwestern Oil Co. v. Socony-Vacuum Oil Co., 138 F.2d 967, 970 (7th Cir. 1943), cert. denied, 321 U.S. 792, 64 S. Ct. 790, 88 L. Ed. 1081 (1944). They impose liability for money damages without requiring any legal or contractual relationship between the parties or a showing of fraud. Hence, the present lawsuit is most similar to actions in which an equitable accounting was unavailable. Appellants' analogy, therefore, fails.

Appellants also cite a variety of cases that do not involve accountings. They claim that in each of these cases the chancellor took jurisdiction because extraordinary complexity made the suit too complex for a common-law jury.

Many of the appellants' cases arose under England's merged system of law and equity created by the Supreme Court of Judicature Acts, 36 & 37 Vict., ch. 66 (1873), 38 & 39 Vict., ch. 77 (1875). The procedural rules adopted in conjunction with the 1875 Act made jury trials generally available upon timely notice by either party. Order XXXVI, Rules 2, 3. This right was subject to the exception that a court without a jury could try an issue "arising in any cause or matter which previously to the passing of the Act could, without any consent of the parties, be tried without a jury." Id. Rule 26.

Appellants look for support in cases applying this exception. They first cite a passage from Clarke v. Cookson, 2 Ch.D. 746, 747-48 (V.C.1876), which they say is a restatement of premerger equity:

This rule was framed expressly to meet cases which would under the old system have been tried in the Chancery Division, and which might be considered, by reason of involving a mixture of law and fact, or from great complexity, or otherwise, not capable of being conveniently tried before a jury.

See also Bordier v. Burrell, 5 Ch.D. 512, 514 (M.R.1877). This passage gives no support to appellants' argument. It identifies two separate prerequisites to a denial of a jury trial demand: trial of the matter in the court of chancery prior to the merger and complexity or other grounds for believing that a jury would be unsuitable. Its use of the conjunction "and" implies the opposite of appellants' argument: complexity alone was not a grounds for relief in equity.

Appellants cite two postmerger cases that rely on the passage quoted from Clarke v. Cookson. Wedderburn v. Pickering, 13 Ch.D. 769 (M.R.1879); Garling v. Royds, 25 Wis. 123 (V.C.1876). These cases confirm our reading of the passage. In each, a clear ground for equitable jurisdiction other than complexity was present. In Wedderburn v. Pickering, supra, the plaintiffs prayed for an injunction. The relief sought in Garling v. Royds, supra, was cancellation of a note allegedly obtained by fraud and misrepresentation.

IBM's brief cites premerger cases involving a limitation on the use of advisory juries in chancery. The chancellor had the discretion to submit to a common-law jury issues of fact arising in an equity case. However, the use of advisory juries remained subject to his discretion, and he declined to direct an issue when he considered it too difficult for a jury. Gartside v. Isherwood, 1 Bro.C.C. 558, 28 Eng.Rep. 1297 (Ch.1783); Gyles v. Wilcox, 2 Atk. 141, 26 Eng.Rep. 489 (Ch.1740). The advisory jury cases are irrelevant to our inquiry. The Chancellor's authority to direct an issue to a jury derived from his control over the method of finding facts in suits already within his jurisdiction. See generally Langbein, Fact Finding in the English Court of Chancery: A Rebuttal, 83 Yale L.J. 1620 (1974). These cases say nothing about the chancellor's authority to provide for nonjury trials in suits at common law.

Appellants also rely on a rather derogatory reference to juries in Blad v. Bamfield, 3 Swans. 604, 36 Eng.Rep. 992 (Ch.1674). Blad, a Dane, had seized the goods of several Englishmen for fishing in Icelandic waters. The Englishmen sued at law for trespass and trover. Blad sought an injunction against these suits on the ground that his seizure was authorized by a patent from the King of Denmark granting him a monopoly over Icelandic fishing rights. The parties disputed the validity of this patent under a peace treaty between England and Denmark. The chancellor found that he had jurisdiction over the matter, "as it had relation to articles of peace, all leagues and safe conducts being anciently enrolled in this court." Id. at 606, 36 Eng.Rep. at 992. After finding that Blad had proven the validity of his patent and the lawfulness of his seizure under Danish law, the chancellor granted the injunction. He added:

Now, after all this, to send it to a trial at law, where the Court must pretend to judge the validity of the king's letters patent in Denmark, or of the exposition and meaning of the articles of peace; or that a common jury should try whether the English have a right to trade in Iceland, is monstrous and absurd.

Id. at 607, 26 Eng.Rep. at 993. Contrary to appellants' assertion, the rationale for equitable relief in this case was not that the chancellor found the issue particularly unsuitable for a jury. It was that issues of international relations were not proper matters for determination in a trial at common law, either by a judge or a jury.

Finally, appellants rely on Clench v. Tomley, Cary 23, 21 Eng.Rep. 13 (Ch.1603), a case involving a dispute over title to land. Tomley had been in possession of the land for fifty years, but Clench asserted a chain of title back to an heir of Tomley's predecessor. Tomley responded with an allegation that the supposed heir was an illegitimate and therefore unable to inherit the land. His evidence consisted of "testimony of ancient witnesses speaking of sixty years before, and account books and other writings." The chancellor decided that Tomley should remain in possession and decreed that:

Clench should not have any trial at the common law for his right till he had shewed better matter in the Chancery, being a thing so long past; it rested not properly in notice de pais, but to be discerned by books and deeds, of which the Court was better able to judge then a jury of ploughmen.

Thus, the chancellor seems to have enjoined trial because he considered an issue inappropriate for jury determination.

The chancellor's reasoning is most unusual for the time at which he was speaking. He questioned the jury members' capabilities because they would not be able to rely on notice de pais their own knowledge of events. Instead, they would have to rely entirely on documentary evidence. However, the practice of submitting documentary evidence to a jury had been established for centuries. See 2 F. Pollock & F. Maitland, The History of English Law 628 & n.2 (Milsom ed. 1968). We are aware of no other case in which a chancellor enjoined a suit at common law because it would require the jury to examine documentary evidence or because the jury could not rely on notice de pais.

The novelty of this reasoning strongly supports either of two conclusions. First, the report of the case may be inaccurate or incomplete. The reports of decisions from this period are neither official nor fully reliable, see T. Plucknett, A Concise History of the Common Law 280-81, 693 (1956), and the report of Clench v. Tomley seems especially subject to question. Scholars who have studied the old English authorities on behalf of the parties appearing before this court have examined the records in Clench v. Tomley, and they have found the report to be a substantially incomplete statement of the case. Moreover, they disagree on whether complexity actually was the basis for equitable relief. Arnold, A Historical Inquiry into the Right To Trial by Jury in Complex Civil Litigation, 128 U.Pa.L.Rev. 829, 840-45 (1980); Campbell & Le Poidevin, Complex Cases and Jury Trials: A Reply to Professor Arnold, 128 U.Pa.L.Rev. 965, 974-85 (1980); Arnold, A Modest Replication to a Lengthy Discourse, 128 U.Pa.L.Rev. 986 (1980).12 Second, Clench v. Tomley may be an aberration, without precedent and subsequently disregarded. We can find no citation of the case in later chancery decisions or in major treatises on equity. Moreover, the chancellor who decided the case, Lord Ellesmere, proved in his later clash with Chief Justice Coke that he held relatively expansive views on the authority of a chancellor to intervene into legal matters. See 5 W. Holdsworth, A History of English Law 231-38 (1924).

Whether or not either of these conclusions is correct is a question that may interest historians; we need not decide it here. We note that appellants' only support for the authority of a chancellor to remove difficult issues from juries in suits at law is a single decision of dubious authority. With this meager support, we cannot conclude that complexity alone ever was an established basis of equitable jurisdiction.

IBM has presented the court with a related historical argument. This argument relies on historical research commissioned by IBM for use in this suit and other suits considering the same issue. Devlin, Note on the Suit at Common Law in England at the Time of the Seventh Amendment (1979).13

We understand the essential terms of this argument as follows. In England in 1791, the chancellor controlled the boundary between law and equity. He exercised this control with a measure of flexibility, removing suits from trial at common law when the procedures of law were inadequate to do justice in a particular case. The chancellor was aware of the limited capabilities of juries and, as evidenced by his practices in accounting and advisory jury cases, often viewed these limitations as an impediment to justice. It is not decisive that complexity alone never was established as a basis for equitable relief. If, in 1791, a suit arose at law that was too complex for a jury, the chancellor would have exercised his control over his jurisdiction to decide the case in chancery. In essence, the argument is a deduction of the likely reaction of the English chancellor to a hypothetical complex suit filed at law in 1791. The principal bases for the deduction are the chancellor's attitudes towards juries and his concern for justice and efficiency.

We are aware of no federal court decision that employs history in this manner. Many have determined the legal or equitable nature of a suit by comparing it with suits actually tried in courts of common law or equity. See, e. g., Pernell v. Southall Realty, 416 U.S. 363, 94 S. Ct. 1723, 40 L. Ed. 2d 198 (1974). None have looked to the chancellor's concern for justice and his perception of the common-law jury of 1791 in order to conclude that he would have decided a matter outside of the categories of suits specifically recognized as within that jurisdiction.

We choose not to pioneer in this use of history. If developments since 1791 have so changed the character of a suit at law to make trial of particular suits to a jury unjust, then perhaps the historically recognized boundary between law and equity should not govern the extent of the seventh amendment right. If so, then deviations from this approach to the seventh amendment should be based on the current policies and the present circumstances of the federal courts. We see no persuasive reason for incorporating into the seventh amendment the policies and probable actions of the English chancellor of 1791.VI.

Both appellants and IBM offer a second constitutional argument. They contend that the due process clause of the fifth amendment prohibits trial by jury of a suit that is too complex for a jury. They further contend that this due process limitation prevails over the seventh amendment's preservation of the right to jury trial.

Although no specific precedent exists for a finding a due process violation in the trial of any case to a jury,14 the principles that define the procedural requirements of due process would seem to impose some limitations on the range of cases that may be submitted to a jury. The primary value promoted by due process in factfinding procedures is "to minimize the risk of erroneous decisions." Greenholtz v. Inmates of the Nebraska Penal and Correctional Complex, 442 U.S. 1, 13, 99 S. Ct. 2100, 2107, 60 L. Ed. 2d 668 (1979). See also Mathews v. Eldridge, 424 U.S. 319, 335, 96 S. Ct. 893, 903, 47 L. Ed. 2d 18 (1976). A jury that cannot understand the evidence and the legal rules to be applied provides no reliable safeguard against erroneous decisions. Moreover, in the context of a completely adversary proceeding, like a civil trial, due process requires that "the decisionmaker's conclusion . . . rest solely on the legal rules and evidence adduced at the hearing." Goldberg v. Kelly, 397 U.S. 254, 271, 90 S. Ct. 1011, 1022, 25 L. Ed. 2d 287 (1970). Unless the jury can understand the legal rules and evidence, we cannot realistically expect that the jury will rest its decision on them.

As we have noted, the law presumes that a jury will decide rationally; it will resolve each disputed issue on the basis of a fair and reasonable assessment of the evidence and a fair and reasonable application of relevant legal rules. See Part IV supra. We conclude that due process precludes trial by jury when a jury is unable to perform this task with a reasonable understanding of the evidence and the legal rules.

If a particular lawsuit is so complex that a jury cannot satisfy this requirement of due process but is nonetheless an action at law, we face a conflict between the requirements of the fifth and seventh amendments. In this situation, we must balance the constitutionally protected interest, as they are implicated in this particular context, and reach the most reasonable accommodation between the two constitutional provisions. See Gannett Co. v. DePasquale, 443 U.S. 368, 99 S. Ct. 2898, 2911-12, 61 L. Ed. 2d 608 (1979); Nebraska Press Ass'n v. Stuart, 427 U.S. 539, 96 S. Ct. 2791, 49 L. Ed. 2d 683 (1976).

The due process objections to jury trial of a complex case implicate values of fundamental importance. If judicial decisions are not based on factual determinations bearing some reliable degree of accuracy, legal remedies will not be applied consistently with the purposes of the laws. There is a danger that jury verdicts will be erratic and completely unpredictable, which would be inconsistent with evenhanded justice. Finally, unless the jury can understand the evidence and the legal rules sufficiently to rest its decision on them, the objective of most rules of evidence and procedure in promoting a fair trial will be lost entirely. We believe that when a jury is unable to perform its decisionmaking task with a reasonable understanding of the evidence and legal rules, it undermines the ability of a district court to render basic justice.

The loss of the right to jury trial in a suit found too complex for a jury does not implicate the same fundamental concerns. The absence of a jury trial requirement in equitable and maritime actions indicates that federal courts can provide fair trials and can grant relief in accordance with the principles of basic justice without the aid of a jury. Moreover, the Supreme Court has consistently refused to rule that preservation of civil jury trial is an essential element of ordered liberty required of the states by the due process clause of the fourteenth amendment. Palko v. Connecticut, 302 U.S. 319, 324-25, 58 S. Ct. 149, 151, 82 L. Ed. 288 (1937); Hardware Dealers Mutual Fire Ins. Co. v. Glidden Co., 284 U.S. 151, 158, 52 S. Ct. 69, 71, 76 L. Ed. 214 (1931); Melancon v. McKeithen, 345 F. Supp. 1025 (E.D.La.) (three-judge court), aff'd mem. 409 U.S. 943, 93 S. Ct. 289, 34 L. Ed. 2d 214 (1972), 409 U.S. 1098, 93 S. Ct. 908, 24 L. Ed. 2d 679 (1972).

Appellees argue that the due process objection to jury trials carries less weight than the preservation of the right to jury trial because it concerns a "hypothetical prospect" of an improper jury verdict and would be applied "prospectively prior to trial." We find no merit in this argument. The due process objection does concern, to be sure, a possibility of an erroneous and erratic jury verdict that might not occur, but this possibility is anything but remote. If the jury is unable to understand the evidence and legal rules the possibility is substantial. Striking a jury trial demand in order to prevent this possibility is prospective relief. However, the procedural requirements of due process are by their very nature prospective: They are safeguards against the possibility of erroneous and arbitrary deprivations of liberty and property. This feature never has been thought to diminish their importance.

The district court asserted that the due process argument fails to account for the special benefits that juries bring to civil litigation. Because the jury is a representative of the community and can call upon the community's wisdom and values, the legal system has relied on it to perform two important functions. The first is "black box" decisionmaking. The jury issues a verdict without an opinion to explain or justify its decision. This feature allows juries to perform a type of "jury equity," modifying harsh results of law to conform to community values in cases where a judge would have to apply the law rigidly. The second function is to accord a greater measure of legitimacy to decisions that depend upon determinations of degree rather than of absolutes, such as whether particular conduct constitutes negligence. Certain decisions of this "line-drawing" nature seem less arbitrary when made by a representative body like the jury. 478 F. Supp. at 938-42.

In the context of a lawsuit of the complexity that we have posited, however, these features do not produce real benefits of substantial value. The function of "jury equity" may be legitimate when the jury actually modifies the law to conform to community values. However, when the jury is unable to determine the normal application of the law to the facts of a case and reaches a verdict on the basis of nothing more than its own determination of community wisdom and values, its operation is indistinguishable from arbitrary and unprincipled decisionmaking. Similarly, the "line-drawing" function is difficult to justify when the jury cannot understand the evidence or legal rules relevant to the issue of where to draw a line.

The district court also noted that preservation of the right to jury trial is important because the jury "provides a needed check on judicial power." 478 F. Supp. at 942. See also Higginbotham, Continuing the Dialogue: Civil Juries and the Allocation of Judicial Power, 56 Tex.L.Rev. 47, 58-60 (1977). A jury unable to understand the evidence and legal rules is hardly a reliable and effective check on judicial power. Our liberties are more secure when judicial decisionmakers proceed rationally, consistently with the law, and on the basis of evidence produced at trial. If the jury is unable to function in this manner, it has the capacity of becoming itself a tool of arbitrary and erratic judicial power.

Therefore, we find the most reasonable accommodation between the requirements of the fifth and seventh amendments to be a denial of jury trial when a jury will not be able to perform its task of rational decisionmaking with a reasonable understanding of the evidence and the relevant legal standards. In lawsuits of this complexity, the interests protected by this procedural rule of due process carry greater weight than the interests served by the constitutional guarantee of jury trial. Consequently, we shall not read the seventh amendment to guarantee the right to jury trial in these suits.

The district court devoted most of its discussions of appellants' due process argument not to factors relevant to the balancing of interests set out in the foregoing section but to a number of practical objections to the argument. We shall consider those objections in this section.

First, the district court challenged the premise that a case could exceed a jury's ability to decide rationally15 and asserted that a jury was at least as able as a judge, the only alternative factfinder, to decide complex cases. 478 F. Supp. at 935-36. The court noted that a jury possesses the wisdom, experience, and common sense of twelve persons. It has a greater effect than a judge in disciplining attorneys to present their cases clearly and concisely. Furthermore, its capabilities can be enhanced by special trial techniques like the preliminary charge and interim charges on the law contemplated by the district court in this case. Id. at 935. See generally Higginbotham, Continuing the Dialogue: Civil Juries and the Allocation of Judicial Power, 56 Tex.L.Rev. 47, 53-55 (1977); Kalven, The Dignity of the Civil Jury, 50 Va.L.Rev. 1055, 1066-67 (1964). On the basis of these observations, the court concluded that a jury "is brighter, more astute, and more perceptive than a single judge, even in a complex or technical case; at least it is not less so." 478 F. Supp. at 935.

Any assessment of a jury's ability to decide complex cases should include consideration not only of a jury's particular strengths and the possible enhancement of its capabilities but also of the particular constraints that operate on a jury in complex cases. The long time periods required for most complex cases are especially disabling for a jury. A long trial can interrupt the career and personal life of a jury member and thereby strain his commitment to the jury's task. The prospect of a long trial can also weed out many veniremen whose professional backgrounds qualify them for deciding a complex case but also prohibits them from lengthy jury service. See Note, The Right to an Incompetent Jury: Protracted Commercial Litigation and the Seventh Amendment, 10 Conn.L.Rev. 775, 776-83 (1978). Furthermore, a jury is likely to be unfamiliar with both the technical subject matter of a complex case and the process of civil litigation. The probability is not remote that a jury will become overwhelmed and confused by a mass of evidence and issues and will reach erroneous decisions. The reality of these difficulties that juries encounter in complex cases is underscored by the experience of some federal district judges who have found particular suits to have exceeded the practical abilities of a jury. See, e. g., ILC Peripherals Leasing Corp v. IBM, 458 F. Supp. 423, 447-48 (N.D.Cal.1978), appeal docketed, Nos. 78-3050, 78-3236 (9th Cir. Sept. 12 and Oct. 6, 1978).

Given that a jury has both particular strengths and weaknesses in deciding complex cases, we cannot conclude a priori that a jury is capable of deciding a suit of any degree of complexity. A litigant might prove that a particular suit is too complex for a jury. Because of the important due process rights implicated, a litigant should have the opportunity to make that showing.

A general presumption that a judge is capable of deciding an extraordinarily complex case, by contrast, is reasonable. A long trial would not greatly disrupt the professional and personal life of a judge and should not be significantly disabling. In fact, the judge's greater ability to allocate time to the task of deciding a complex case can be a major advantage in surmounting the difficulties posed by the suit. Although we cannot presume that a judge will be more intelligent than a jury or more familiar with technical subject matters, a judge will almost surely have substantial familiarity with the process of civil litigation, as a result of experience on the bench or in practice. This experience can enable him to digest a large amount of evidence and legal argument, segregate distinct issues and the portions of evidence relevant to each issue, assess the opinions of expert witnesses, and apply highly complex legal standards to the facts of the case. The judge's experience also can enable him to make better use of special trial techniques designed to help the factfinder in complex cases, like colloquies with expert witnesses. The requirement that a judge issue findings of fact and conclusions of law offsets the substantial tendency to overlook issues in order that a verdict might be reached in these difficult cases. Fed.R.Civ.P. 52(a). Finally, if after trial and during deliberation a judge finds himself confused on certain matters or unable to decide certain issues, he can reopen the trial for the purpose of obtaining clarification or additional evidence. Fed.R.Civ.P. 59(a).

A judge's abilities are, of course, not unbounded. It is conceivable that a case might be so complex that a judge could not decide it rationally and competently. However, the possibility of such a case cannot justify trial by jury, because the presence of a jury does not relieve the judge of the need to understand the issues disputed in a case and the relevance and strength of the evidence. Such an understanding is necessary for the court to rule on most of the important matters committed to its decision in a jury trial, including motions for summary judgment, directed verdict, and new trial and exclusions of evidence as irrelevant, cumulative, or prejudicial.

The lawsuit that exceeds the ability of a judge to decide rationally and competently would challenge the basic capacity of our system of civil litigation to decide lawsuits by any means. It would call for adjustments far more fundamental than an allocation of issues between judge and jury. Prudence compels us to defer consideration of these adjustments until we face a real possibility of such a lawsuit. In the meantime, the best course to follow is to presume the judge's ability to decide a complex case and to focus inquiry on the jury's ability.

The district court's second objection to appellants' due process argument was that the court can prevent an "irrational" verdict with its power to direct a verdict or to grant judgment n. o. v. 478 F. Supp. at 937-38. These devices enable the court to enter judgment against any party that has not submitted at least the minimal quantity of evidence necessary for a jury to decide reasonably in its favor. Patzig v. O'Neil, 577 F.2d 841, 846 (3d Cir. 1978). Because the evidence in support of any jury verdict will satisfy this standard, the district court concluded that "there is little or no room for an 'irrational' verdict." 478 F. Supp. at 938.

The district court presumed that its ability to ensure a verdict that is rational in the foregoing sense provides all of the elements of due process that a litigant may demand from the procedures of civil litigation. We disagree. The specific elements of due process that call for denial of jury trial in excessively complex cases are not provided.

Denial of a jury trial may be necessary to minimize the risk of erroneous decisions. Cf. Mathews v. Eldridge, 424 U.S. 319, 335, 96 S. Ct. 893, 903, 47 L. Ed. 2d 18 (1976) (administrative procedures sufficient to minimize risk of error). The district court's review of evidence on motions for directed verdict and judgment n. o. v. will not serve this purpose adequately. The court may not grant one of these motions if the evidence might reasonably support a verdict for either side. See Patzig v. O'Neil, supra. Thus, the court can only ensure that the jury will return one of a range of possible verdicts that the court finds reasonably but minimally supported by the evidence. Given that substantial property rights often are at stake in actions at law, we believe that due process requires a greater measure of reliability in the decisionmaking process. It requires some fair assurance that the jury's findings of fact and applications of legal rules are reasonably correct. When a jury is unable to understand the evidence and the legal rules, it cannot provide this measure of assurance.

Denial of a jury trial also may be necessary to ensure that the verdict rests on the applicable legal principles and the evidence presented at trial. The rules governing directed verdicts and judgments n. o. v. call for no inquiry into whether the jury will rest or has rested its verdict solely on the evidence and relevant rules of law. When a jury cannot understand them but nonetheless is charged with reaching a verdict, compliance with this requirement of due process cannot be expected.

In short, the district court cannot ensure that trial by jury will conform to the procedural requirements of due process by confining the jury to cases in which the evidence survives review under motions for directed verdict and judgment n. o. v. A denial of jury trial in excessively complex cases remains necessary.

Finally, the district court feared that the authority to strike jury trial demands on case-by-case determinations of complexity would lead to the long-run dilution of the right to jury trial. This fear results in part from the seemingly broad discretion that district court judges will exercise when determining whether a jury is capable of deciding a case and in part from the difficulty that courts will have in applying the due process standard. 478 F. Supp. at 931-33. See also Radial Lip Machine, Inc. v. International Carbide Corp., 76 F.R.D. 224, 227-28 (N.D.Ill.1977).

We do not believe that a due process limitation allows the district courts a substantial amount of discretion to deny jury trials. Because preservation of the right to jury trial remains a constitutionally protected interest, denials of jury trial on grounds of complexity should be confined to suits in which due process clearly requires a nonjury trial. This implies a high standard. It is not enough that trial to the court would be preferable. The complexity of a suit must be so great that it renders the suit beyond the ability of a jury to decide by rational means with a reasonable understanding of the evidence and applicable legal rules. Moreover, the district court should not deny a jury trial if by severance of multiple claims, thoughtful use of the procedures suggested in the Manual for Complex Litigation, or other methods the court can enhance a jury's capabilities or can reduce the complexity of a suit sufficiently to bring it within the ability of a jury to decide. Due process should allow denials of jury trials only in exceptional cases.

We recognize that in borderline cases the district courts may experience some difficulties in applying the standard, but we do not believe that these difficulties will be so great as to justify a complete abandonment of the attempt to conform jury trial practices to the requirements of due process. Most district judges have sufficient familiarity with jury trials, either as judges or as practitioners, to make an informed judgment on the extent of a jury's capabilities. They surely will find that most suits plainly are within a jury's capabilities. For the unusually complex case, the district court can take a fairly objective measure of its complexity by examining three factors which contribute to a jury's inability to understand the evidence and legal rules: first, the overall size of the suit, the primary indicia of which are the estimated length of trial, the amount of evidence to be introduced and the number of issues that will require individual consideration; second, the conceptual difficulties in the legal issues and the factual predicates to these issues, which are likely to be reflected in the amount of expert testimony to be submitted and the probable length and detail of jury instructions; and third, the difficulty of segregating distinct aspects of the case, as indicated by the number of separately disputed issues related to single transactions or items of proof.

We further recognize a relative lack of precision in this standard, but we do not believe that this problem threatens a dilution in the right to jury trial. We believe that district judges will apply the standard with a good faith concern for the general preservation of the right to jury trial. The University of Chicago Jury Project revealed that trial judges overwhelmingly support preservation of the civil jury. Kalven, The Dignity of the Civil Jury, 50 Va.L.Rev. 1055, 1072-74 (1964). Any erroneous denial of jury trial can be remedied before trial through a writ of mandamus from this court. See Beacon Theatres, Inc. v. Westover, 359 U.S. 500, 511 & n. 20, 79 S. Ct. 948, 957, 3 L. Ed. 2d 988 (1959). As an added safeguard, we will require that the district court make explicit findings on the dimensions of complexity when it denies a jury trial in an action at law on grounds of complexity.

In summary, the district court's practical objections to a due process limitation do not compel its rejection. However, the concern for preservation of the right to jury trial should guide its application. In suits at law, a court should deny jury trial on due process grounds only in exceptional cases when the court, after careful inquiry into the factors contributing to complexity, determines that a jury would be unable to understand the case and decide it rationally. Before any such denial, due consideration should be given to the particular strengths of the jury in deciding complex cases, to the possible use of special trial techniques to increase a jury's capabilities, and to methods of reducing the suit's complexity.

We turn to the disposition of this appeal. The district court concluded that "the complexity of the case before us is not a constitutionally permissible reason for striking the plaintiffs' jury demands." 478 F. Supp. at 942. We are constrained to disagree. The appellants' assertions of extraordinary complexity are relevant to the issue of whether the trial of this case to a jury would violate due process and therefore would be beyond the guarantee of the seventh amendment.

Appellees argue that the district court already has considered and rejected the substance of appellants' assertions of extraordinary complexity. They point to the following passage from the court's opinion:

At the heart of both the plaintiffs' claims and the defendants' counterclaims are allegations of conspiracy and predatory intent. These allegations, if legally sufficient, will call for just the kind of judgments of fact that are traditionally within a jury's competence and for which a jury's unique abilities are especially valued. Moreover, we believe that with proper and frequent judicial guidance (we contemplate a preliminary charge and periodic interim charges on the law) as well as thoughtful organization of evidentiary presentation by counsel, the jury will be able to understand and deal intelligently with all the facts and issues.

478 F.Supp. at 935-36.

Although the district court's purpose for including this passage in its opinion is not entirely clear, we do not read it as a ruling on the issue of a jury's ability to decide this particular case. The remarks appear in the course of the court's argument that a jury, in general, is better qualified than a judge to decide complex cases. Id. at 935-36. The passage seems to illustrate or emphasize the court's proposition. In the passage's only reference to the elements of this particular case, regarding the allegations of conspiracy and predatory intent, the court appears to be making a categorical assertion that juries always will be capable of deciding issues of the type traditionally raised in suits at law. In the previous section, we held that the district court should not presume that a jury always will be capable of deciding issues in suits at law. Rather, the court should consider circumstances of a particular case when deciding whether the jury is capable of deciding it with sufficient understanding. The quoted passage does not indicate that the district court has examined the size and difficulties of this particular lawsuit and has found a jury capable.

Other portions of the district court's opinion indicate that the court has not ruled on the complexity of this suit. At the beginning, the court reviewed the arguments of each party on whether or not the suit is too complex for a jury and then expressly refrained from resolving their dispute: "(R)esolution of these disputes as to the degree of complexity of this case is unnecessary in light of our discussion of the legal issues." Id. at 899 (emphasis in original). At the end of the opinion, the court states its conclusions without any mention of the complexity of this lawsuit. Id., at 942.

Thus, the district court has not ruled on whether this particular lawsuit is too complex for a jury to understand and decide rationally. The court relied entirely on its construction of the seventh amendment and the due process clause. We shall vacate the court's order on the basis of our previous discussion and shall leave for consideration on remand the issue of the complexity of this lawsuit.

Finally, we turn to the matters raised in the separate brief of Sony Corp. and Sony Corp. of America. The Sony companies argued that circumstances make them especially vulnerable to an erroneous jury verdict. They also argue that a propensity of two of NUE's attorneys to make remarks prejudicial to Japanese makes a jury trial unsuitable for all of the Japanese defendants.

The Sony companies' fear of special vulnerability results from their rather limited involvement in a large, consolidated lawsuit. They are named as defendants only in the NUE action, having reached a complete settlement with Zenith. Zenith's claims are considerably broader than NUE's, covering more products, longer periods of time, and additional legal issues. Even within NUE's claims, the Sony companies argue that they were not involved in many of the illegal activities charged and that much of NUE's evidence does not implicate them. In essence, the Sony companies fear that a jury will be unable to give separate consideration to the claims against them and will be unable to segregate the evidence admissible against them.

Each litigant has an individual claim to the protections of due process. Therefore, the Sony companies may argue that special circumstances render a jury incapable of deciding the claims against them, even if the other defendants cannot make similar arguments. The district court seems to have considered and rejected the substance of the Sony companies' arguments as part of its denial of the Sony motion for a separate trial under Fed.R.Civ.P. 42(b).16 However, the district court is not precluded from reconsidering its decision in light of our ruling on the due process limitations to trial by jury. We shall leave these arguments for consideration on remand.

The charge that NUE's attorneys have shown a proclivity for prejudicial remarks unfortunately is well documented in the record. However, we do not agree that the likelihood that counsel will attempt to appeal to a jury's biases is a valid reason for striking a demand for jury trial. The district court can remedy this threat to a fair trial by other means, including careful supervision of the trial, reminders to the offending attorneys of their ethical responsibilities,17 and, if necessary, a grant of a new trial. See Minneapolis, St. P. & S. Ste. M. Ry. v. Moquin, 283 U.S. 520, 51 S. Ct. 501, 75 L. Ed. 1243 (1931); Draper v. Airco, Inc., 580 F.2d 91, 95 (3d Cir. 1978). Should this case be tried to a jury, we are confident that the able district judge will ensure that appeals to racial bias and wartime prejudices do not influence the jury.

The order of the district court will be vacated and the case remanded for proceedings consistent with this opinion.

GIBBONS, Circuit Judge, dissenting.

I dissent in this case with a good deal less confidence in my position than in most cases in which the majority has failed to persuade me. Chief Judge Seitz's excellent exposition of the difficult issue of jury comprehension of complex litigation is highly persuasive. Still, in the end I conclude that the majority has substituted for an express guarantee in the Bill of Rights a rule of district court discretion that in practice will be virtually unreviewable, and therefore largely unfettered. Between Judge Becker's opinion in the district court and that of Chief Judge Seitz here, enough has been written that an extended separate statement is hardly appropriate. I will therefore only summarize the areas of agreement and disagreement between us.

First, I agree that this case is complex. Three objective manifestations of complexity are present: the predicted length of trial; the multiplicity of factual issues which may have to be resolved; and the conceptual difficulty of the governing law. However, all three manifestations of complexity are for the most part products of the liberal joinder rules of the Federal Rules of Civil Procedure and of the district court's ruling consolidating two multi-count cases for trial. Neither the liberal joinder rules nor the rule permitting consolidation for trial are required by any provision of the Constitution. The seventh amendment guarantees a jury trial of any separate claim for relief which would have been tried to a jury at common law. It does not guarantee that a single jury will decide multiple separate claims. If by virtue of joinder and consolidation a case becomes too complex for a single jury to handle, the remedy mandated by the seventh amendment is separate juries, as at common law. Thus we should not even consider the constitutional issue which the majority undertakes to decide in the form in which it has been presented here, because we are considering not the actual constitutional issue, but a hypothetical construction of a series of procedural rulings. As I see it, the only seventh amendment issue is whether on any separate claim for relief by one plaintiff against one defendant the case would be too complex for decision by a jury. A plaintiff insisting on a jury trial obviously must forego the economic and procedural advantages of liberal joinder of claims and of parties if such joinder makes the case too complex for a single jury. A defendant insisting on a jury trial must, I think, be in a position to insist that the plaintiff break down his case to the separate components that it would have presented at common law. Breaking the case down to its separate components obviously would produce inefficiencies in the judicial process. But the provisions of the Bill of Rights which limit the way in which the federal courts conduct their business are designed to promote values other than efficiency. The inefficiencies which those provisions impose on the system are a small price to pay for the vindication of the values which the Bill of Rights advances.1

It is axiomatic that constitutional law pronouncements should be avoided unless the order appealed from clearly requires that they be made.2 I see the constitutional issue involved in this case in a much narrower context than does the majority. That issue would only be properly presented to this court if a single claim for relief against a separate defendant would be too complex for jury consideration. It is not presented by the order appealed from, which contemplates the trial of multiple claims by separate plaintiffs against multiple defendants. Indeed, by considering the constitutional implications of the district court's order, we decide a false issue, namely, whether by virtue of the joinder and consolidation permitted by the Federal Rules Enabling Act, 28 U.S.C. § 2072 (1976), a case can be made so complex that the constitutional right to jury trial, protected by the seventh amendment, must yield. The majority purports to resolve the narrower issue of whether any single case properly tried to a single jury can be so complex that the seventh amendment right may be overridden. But that question is not presented on this record. It is not ripe for decision because it is presented hypothetically. Our acceptance of an interlocutory appeal under 28 U.S.C. § 1292(b) does not, moreover, decide what issues the panel must consider.3 The trial court never determined whether after the case is broken down into the separate components in which it would be presented at common law it would or would not be too complex. Therefore, I would vacate the order granting leave to take an interlocutory appeal and wait for a case which actually presents the seventh amendment complexity issue in what I believe to be its actual rather than its hypothetical dimensions.

While my position with respect to ripeness might well dictate silence as to the merits, it seems likely that further review will be sought. An expression of my views on the merits, in the quite conceivable event that the Supreme Court disagrees with my ripeness analysis, may facilitate such review. If I were to reach the merits, then, my position on the issues discussed by Chief Judge Seitz would be as follows:

A. I would join in Part III, holding that there is no applicable statutory guarantee of jury trial.

B. I would join in Part IV, holding that the famous and enigmatic footnote in Ross v. Bernhard, 396 U.S. 531, 538 n.10, 90 S. Ct. 733, 738, 24 L. Ed. 729 (1970), is not dispositive.

C. I would join in Subparts A, B and C of Part V, which reject the contention that we can find guidance for the solution of the seventh amendment-complexity problem in eighteenth century precedents.

D. I do join in Parts VIII and IX of the majority opinion.

E. I do not join in Parts VI and VII. If the issue were properly ripe for decision I would be prepared to hold that I cannot conceive of a case in which what would be a separate claim for relief at common law, sufficiently comprehensible to a trial judge to satisfy due process, would be too complex for trial to a jury. There may be such a case, but it is inconceivable to me that it could be recognized as such in the absence of a trial record.

I have, however, a more serious disagreement with the majority. The court has now authorized pretrial denials of demands for jury trial in suits at common law "on due process grounds . . . in exceptional cases when the court, after careful inquiry into the factors contributing to complexity, determines that a jury would be unable to understand the case and decide it rationally." Majority opinion at 1089. The majority opinion attempts to objectify the factors that bear upon complexity, but in the end the factors which are identified will permit the exercise of trial court discretion. I fear that the exercise of that discretion will sometimes be influenced by unarticulated sympathies for or hostilities toward the underlying policies sought to be advanced in the lawsuit. Trial court discretion, moreover, in any practical sense will be completely unreviewable. The majority suggests that an erroneous denial of jury trial can be remedied before trial by a writ of mandamus. Majority opinion at 43. That remedy is hardly adequate, since formalistic application of the factors mentioned by the majority will result in a holding that the trial court acted within the bounds of its permissible discretion. In the event of a post-trial appeal on the ground that a demand for jury trial was erroneously denied, the pressure on an appellate tribunal not to order retrial of an otherwise error-free trial in a complex case will, I suspect, inevitably be irresistible. I would be less uncomfortable with the majority's rule if interlocutory review were available as a matter of right.4

Part of my difficulty with the majority's position probably results from a perception of the nature of the judicial process and the role of juries in that process. It is often said that the judicial process involves the search for objective truth. We have no real assurance, however, of objective truth whether the trial is to the court or to a jury. The judicial process can do no more than legitimize the imposition of sanctions by requiring that some minimum standards of fair play, which we call due process, are adhered to. In this legitimizing process, the seventh amendment is not a useless appendage to the Bill of Rights, but an important resource in maintaining the authority of the rule of law. In the process of gaining public acceptance for the imposition of sanctions, the role of the jury is highly significant. The jury is a sort of ad hoc parliament convened from the citizenry at large to lend respectability and authority to the process. Judges are often prone to believe that they, alone, can bear the full weight of this legitimizing function. I doubt that they can. Any erosion of citizen participation in the sanctioning system is in the long run likely, in my view, to result in a reduction in the moral authority that supports the process.

In light, therefore, of the important functions served by the seventh amendment's protection of the right to a trial by jury, I would hold that there is no case in which properly separated claims for relief cognizable at common law would be so complex that trial by jury would amount to a violation of due process.