dissenting.
I agree with the majority that this court has jurisdiction of the appeal under 28 U.S.C. § 1291 and Rule 54(b), F.R.Civ.P., and that the plaintiffs have standing to sue non-lessors under §§ 4 and 16 of the Clayton Act. At that point, however, I part company with my brothers. In my view, the decision of the district court holding that plaintiffs have failed to state a claim under § 1 of the Sherman Act should be affirmed, as should its decision denying class certification.
I.
A.
My first, and most fundamental, disagreement with the majority concerns the role of pleading and discovery in this kind of litigation. The majority says that a “complaint is much too blunt an instrument with which to forge fundamental policies.” But, if a complaint cannot even be depended on to state the essential theory of the litigation, I do not know how a court can correctly resolve the particular case, must less forge fundamental policies. This is not a pro se case. Plaintiffs’ counsel are competent, experienced and, in fact, nationally renowned attorneys in the antitrust field. Nor is this a case that was hastily terminated by summary judgment. The original complaints were filed in 1971, the Bogosian complaint was amended in 1972, and both complaints were amended in 1973. Now, six years after the action was commenced, the majority remarks that it is “unwilling to speculate at this stage as to the plaintiffs’ theory.” In my view, this case has long since passed the stage where anyone concerned — parties, lawyers, or judges— should have to speculate as to the theory of the litigation.
The majority’s precise holding is that “the ruling that the specific allegation of interdependent consciously parallel action made here fails to state a claim should be vacated so that the issue can be decided, if necessary, after the relevant facts are fully developed.” This is jurisprudential anarchy. Although the majority purports to act in the interest of efficient judicial administration, I fail to see how that interest is served by allowing what probably will be massive discovery prior to deciding whether the basic theory of the action is legally viable. The relevant facts should be fully developed after it is determined whether the claim is legally sufficient, not while that issue is still in doubt. Moreover, until the theory of the case is settled, it will not be known which are the “relevant” facts. Facts are only relevant insofar as they support a valid legal theory.
A motion to dismiss or for summary judgment for failure to state a claim seeks to obviate the necessity for time-consuming and expensive discovery in cases where the facts are irrelevant because no legal claim has been stated. Requiring discovery as a predicate to deciding such a motion defeats the very purpose of the motion. Although the majority says that further factual development will help to test the legal theory, I do not think that is the case here. The question is whether an allegation of interdependent consciously parallel action states a Sherman Act claim. Either it does or it does not. That may be a sophisticated question, but it is a question of policy, not of fact. The kind of adjudicative facts which will be developed by discovery in a particular case concerning particular plaintiffs and defendants in a particular industry will be of no help whatsoever.
I accept the principle that summary judgment is to be sparingly granted in complex antitrust cases. Primarily, this applies to cases where a cognizable claim has been stated and the question is whether there are disputed material issues of fact. I do not understand the principle as undermining *458the utility of pre-trial motions to test the sufficiency of a complaint. The Federal Rules require that a complaint make a “short and plain statement of the claim showing that the pleader is entitled to relief.” F.R.Civ.P. 8(a).
Federal procedure relies on notice pleading rather than fact pleading, but at a minimum, as explained by the Advisory Committee on Civil Rules in October, 1955, the pleader is required “to disclose adequate information as the basis of his claim for relief”. 2A J. Moore, Federal Practice ¶ 8.01[3] (2d ed. 1974). Conley v. Gibson, 355 U.S. 41, 47, 78 S.Ct. 99, 103, 2 L.Ed.2d 80 (1957), teaches that “the defendant [be given] fair notice of what the plaintiff’s claim is and the grounds upon which it rests,’’ (emphasis supplied), and we have recently reiterated that the defendant is entitled to “fair notice of the claim asserted.” Joiner Systems, Inc. v. AVM Corp., 517 F.2d 45, 47 (3d Cir. 1975).
Universe Tankships, Inc. v. United States, 528 F.2d 73, 75-76 (3d Cir. 1975). In massive and protracted cases like this, the attorneys bear a particularly heavy responsibility to narrow and simplify the issues as soon as possible. Where, after six years, it is still possible to speculate as to the exact nature of the claim asserted, then I think it is time for the district court to do its best in grasping the essential intendment of the complaint and, if no claim is stated, to dismiss or grant summary judgment for failure to state a claim. That is exactly what the district court did, and I would affirm its action.
B.
My reading of the complaint differs from the majority’s. After two or three amendments by skilled counsel, I think it is time to take the words of the complaint for what they actually say. By amendment to their complaint and pursuant to what the district court termed a “deliberately employed strategy”, 393 F.Supp. 1046, 1048 n.5, plaintiffs’ counsel deleted their allegations of conspiracy and substituted allegations of “interdependent consciously parallel action”. As the majority correctly notes, plaintiffs were not required to plead evidence. They did not do so. They'did not allege that a contract, combination or conspiracy existed and that it would be proven by evidence of interdependent consciously parallel action. Rather, they alleged that defendants had engaged in interdependent consciously parallel action in violation of § 1 of the Sherman Act. Even construing paragraphs 13, 14, 16, 17 and 18 of the second amended complaint most favorably to the plaintiffs, it alleges that interdependent consciously parallel action is a “combination” for Sherman Act purposes. And that is just not so.
It is well settled, and indeed it is not here disputed, that conscious parallelism is not a violation of § 1 of the Sherman Act. Theater Enterprises, Inc. v. Paramount Film Distributing Corp., 346 U.S. 537, 74 S.Ct. 257, 98 L.Ed. 273 (1954); Klein v. American Luggage Works, Inc., 323 F.2d 787 (3d Cir. 1963); Delaware Valley Marine Supply Co. v. American Tobacco Co., 297 F.2d 199 (3d Cir. 1961), cert. denied, 369 U.S. 839, 82 S.Ct. 867, 7 L.Ed.2d 843 (1962). The leading case, Theater Enterprises, explains: *459346 U.S. at 540-41, 74 S.Ct. at 259. From this, it is clear to me that consciously parallel behavior is admissible as circumstantial evidence of a properly pleaded contract, combination or conspiracy. That is the point that was implicated in Venzie Corp. v. United States Mineral Products Co., 521 F.2d 1309 (3d Cir. 1975). In Venzie, we specifically noted that the plaintiffs “relied on . circumstantial evidence . as proof of agreement. Their evidence does not, however, include two elements generally considered critical in establishing conspiracy from evidence of parallel business behavior . . . ”. Id. at 1314. Venzie was an appeal from an order granting a defendants’ motion for judgment n. o. v. As such, it was our obligation to review the plaintiffs’ evidence, and we concluded that they had not met their burden of proving conspiracy. This, however, is not an appeal from a judgment n. o. v., it is an appeal from a summary judgment for failure to state a claim. We are faced not with a question of evidence, but with a question of pleading.
*458The crucial question is whether respondents’ conduct toward petitioner stemmed from independent decision or from an agreement, tacit or express. To be sure, business behavior is admissible circumstantial evidence from which the fact finder may infer agreement. Interstate Circuit, Inc. v. United States, 306 U.S. 208, 59 S.Ct. 467, 83 L.Ed. 610 (1939); United States v. Masonite Corp., 316 U.S. 265, 62 S.Ct. 1070, 86 L.Ed. 1461 (1942); United States v. Bausch & Lomb Optical Co., 321 U.S. 707, 64 S.Ct. 805, 88 L.Ed. 1024 (1944); American Tobacco Co. v. United States, 328 U.S. 781, 66 S.Ct. 1125, 90 L.Ed. 1575 (1946); United States v. Paramount Pictures, Inc., 334 U.S. 131, 68 S.Ct. 915, 92 L.Ed. 1260 (1948). But this Court has never held that proof of parallel business behavior conclusively establishes agreement or, phrased differently, that such behavior itself constitutes a Sherman Act offense.
*459As it is clear from Theater Enterprises that consciously parallel behavior may be circumstantial evidence of an agreement, so also is it clear that consciously parallel behavior alone does not violate the Sherman Act. And, therefore, an allegation of consciously parallel behavior, without more, would not state a Sherman Act claim. But, if an allegation of consciously parallel behavior is legally insufficient, I am unable to see how an additional allegation of interdependence cures the insufficiency.
If there is anything in this case that hearkens back to “magic words”, it is not, as the majority suggests, the requirement that a Sherman Act complaint allege concerted action of some kind. It is rather, the asserted power of the word “interdependent” to breathe vitality into a lifeless theory of recovery. Nowhere has the talismanic character of that word been explained to my satisfaction. Indeed, it seems to me that interdependence is implicit in the notion of conscious parallelism and that the added word is hardly more than a redundancy. In the usual situation of parallel business behavior, a businessman is conscious of what his competitor is doing and his action, or inaction, depends on what the competitor does. This is not a violation of the antitrust laws; it is, in fact, the essence of the competitive behavior that those laws seek to promote. Because his competitor takes the same attitude toward him, the two businessmen are mutually conscious of each other and their actions are “interdependent”. In a concentrated industry, such mutually conscious and interdependent conduct by several competitors may have anti-competitive effects. But it is not necessarily collusive, and I cannot understand how a proliferation of descriptive words changes the legal status of the conduct. Until there is a contract, combination or conspiracy, in restraint of trade, there is no § 1 violation. As the Supreme Court observed in Theater Enterprises: “Circumstantial evidence of consciously parallel behavior may have made heavy inroads into the traditional judicial attitude toward conspiracy; but ‘conscious parallelism’ has not yet read conspiracy out of the Sherman Act entirely.” 346 U.S. at 541, 74 S.Ct. at 259 (footnote omitted).
In their brief in this court, plaintiffs rely primarily on Wall Products Co. v. National Gypsum Co., 326 F.Supp. 295 (N.D.Cal. 1971), and Modern Home Institute, Inc. v. Hartford Accident and Indemnity Co., 513 F.2d 102 (2d Cir. 1975). Based on these decisions, they say “it is clear that [the courts] have uniformly held that proof of interdependent conscious parallel action, while not conclusive proof of a Section 1 Sherman Act violation, is sufficient to support a finding of violation.” Appellants’ Brief at 21. In my view, it is not necessary to agree or disagree with that reading of the cases in order to affirm the summary judgment here.
Like the majority, the plaintiffs have mistaken an issue of pleading for an issue of evidence. We are not presented with a situation where interdependent consciously parallel action — however that may differ from merely consciously parallel action— has been introduced as evidence of a prop*460erly pleaded contract, combination or conspiracy. Nor are we presented with the issue whether such evidence alone would be sufficient to support a finding of concerted action. Instead, we are presented with the pure question of law whether an allegation of interdependent consciously parallel action, without more, states a claim under § 1 of the Sherman Act. In my view, it does not. As no amount of discovery can cure a complaint that fails to state a claim, I would affirm the entry of summary judgment.
II.
Having said initially that it is “unwilling to speculate at this stage as to the plaintiffs’ theory,” the majority proceeds to “identify in greater depth the nature of the claims” and to conclude that the district court erred in denying class certification. For purposes of summary judgment, apparently, the exact theory of the case remains foggy to the majority, but for purposes of class action determination the detailed elements of the claim have become pellucid. If the exact nature of the legal theory is clear, then the case is ripe for a test of whether a claim has been stated. If the legal theory is not clear, then the question of class certification ought to abide the clarification of the issues.
This is not an ordinary class action. Plaintiffs do not seek to form a class of all Exxon dealers, or all Gulf dealers, or all Texaco dealers. They seek to form a truly titanic class of all present and former lessee gasoline dealers of 15 different oil companies. The only limitation on the class is that the dealers be located in areas having a population over 15,000. At the very least, the size and diversity of the asserted class raise serious questions about the propriety of class action treatment. Moreover, the members of the class are not consumers with individually small or de minimis claims. Nor are the class members persons who have been deprived of non-quantifiable civil rights. The class members are businessmen, more or less sophisticated, with significant and definite financial interests in the litigation. This is not, therefore, a case where it would be economically impractical to prosecute individual actions.
A.
The majority has concluded that the district court erred when it found that common questions would not predominate over individual ones, as required by Rule 23(b)(3). I disagree. Rather than paraphrasing, I prefer the exact language concerning our review of a predominance finding as stated in Katz:
If the district court has properly identified the issues common and diverse, we would undoubtedly defer in most instances to its conclusion as to predominance, since that requirement relates to the conservation of litigation effort, and the trial court’s judgment probably will be as good as ours. If the district court has applied the correct criteria to the facts of the case, then, it is fair to say that we will ordinarily defer to its exercise of discretion.
Katz v. Carte Blanche Corp., 496 F.2d 747, 756 (3d Cir.), cert. denied, 419 U.S. 885, 95 S.Ct. 152, 42 L.Ed.2d 125 (1974). The majority’s analysis, in my opinion, pays no more than lip service to the discretion which a district court must have in evaluating the propriety of class certification in such a massive proceeding as this.
The majority says that the district court erred in identifying the issues which are common to the class. This is not to say that the district court, in its thorough treatment, failed to identify any issues in the case. Rather, this is a euphemistic way of indicating that the majority disagrees with the district court’s decision to require individualized proof on certain issues. I believe that the judicial structuring of the proof, especially in a complicated proceeding like this one, lies necessarily and unalterably within the discretion of the district court. Not only do I find no abuse of that discretion, but, on most points, my own analysis would track the district court’s quite closely-
*461, Concerning the lease claim, the majority correctly states that “[t]here is no single lease provision which requires that the lessee purchase all gasoline from his lessor.” It goes on to say, also correctly, that it is a “constellation of lease provisions” which allegedly has the practical economic effect of tying the purchase of gasoline to the lease of the service station. Then, having admitted that whether the lessee realistically could sell other brands of gasoline is a “question of fact which can be developed by expert testimony concerning the relative costs and benefits of [installing the lessee’s own pumps and tanks],” the majority nevertheless concludes that the practical economic effect of the various lease provisions would present a common question, provable by common proof. Even if there were only one defendant oil company and only one form contract, the practical economic effect would vary from dealer to dealer, city to city, and region to region. It might, for example, be economically feasible for a large volume dealer in a large city to install his own pumps and tanks while it might not be feasible for a smaller dealer in a smaller city to do so. Here there are more than a dozen oil companies, with operations concentrated in different regions of the country, and there are more than 400 different forms of contracts and agreements. A fortiori, the practical economic effects of the agreements will present diverse questions.
The question of market power in the tying product, leaseholds, will similarly present diverse, not common, questions. The majority asserts that market power could be demonstrated by showing that “defendants controlled a majority of existing service stations and that because zoning restrictions and high capital costs make development of new stations difficult, the defendants have sufficient market dominance over existing stations to impose a tie-in.” I have at least two serious problems with this line of reasoning. To begin with, I have no doubt that plaintiffs could show that defendants control a majority of existing stations. This follows almost tautologically from the fact that plaintiffs have named as defendants most of the major companies in a concentrated industry. This element of proof can be satisfied in any reasonably concentrated industry simply by naming enough defendants. Secondly, I fail to see how zoning restrictions and capital costs could be established by common proof. Certainly zoning laws vary from city to city, and I have no doubt that the capital costs of developing a station do also.
The majority concedes that the quantum of damages would have to be individually established. But it finds “nothing in this record which persuades us” that the fact of damage would have to be individually proven, and directs the district court to reconsider the matter in light of the majority’s opinion. Although I agree with the majority that the propriety of proving the fact of damage on a common basis depends on the circumstances of the case, I find nothing in this record which persuades me that the fact of damage would not have to be individually proven. There is no reason why there might not be dealerships which suffered no damage at all as a result of defendants’ conduct, be it legal or illegal. As the majority observes, it is possible that “in certain areas the free market price would be no lower than the conspiratorially affected price.” For me, however, the issue is of little importance as I would not upset the district court’s predominance finding even if the fact of damage were a common question.
I do not dispute the fact that there may be common questions. For example, if a conspiracy had been alleged, that would present a common question subject to the same proof whether made on behalf of one plaintiff or a thousand. But the fundamental issue is not the existence of common questions, it is whether the district court abused its discretion in concluding that common questions would not predominate. This is not simply a matter of numbering the questions in the case, labelling them as common or diverse, and then counting up. It involves a sophisticated and necessarily judgmental appraisal of the future course of the litigation, as well as an evaluation of *462the most efficient means of proving the claims and the time that will be consumed by each aspect of the proof. The district court is in a position to do that; we are not.
In my opinion, the district court made no significant error in identifying the issues as common or diverse, it correctly applied the criteria of Rule 23 to the facts, and it clearly did not abuse its discretion in concluding that common questions would not predominate in this colossal but intricate proceeding.
B.
The majority has also concluded that the district court’s reasons for finding the class action not to be superior to other adjudicative vehicles were “erroneous”. Again, Katz articulates the standard of review of a finding on the issue of superiority: “[0]ur review looks first at whether the district court properly applied the relevant criteria to the facts of the case. If this has been done it is fair to say that we will ordinarily defer to its exercise of discretion.” 496 F.2d at 757.
In this case the district court not only made the ultimate finding of lack of superiority, it expressly applied the subsidiary criteria, which the rule suggests as pertinent, to the facts. It found a significant class member interest in individually controlling separate actions which, as I have indicated, would not necessarily be impractical. It found that disparate factual considerations made concentration of the litigation in one forum not desirable or advantageous. And it found that management of the action would present “enormous difficulties.” 62 F.R.D. at 140. We have previously observed that the question of manageability of a class is largely a factual question and that the district court has a “wide range of discretion” in the matter. Link v, Mercedes Benz, 550 F.2d 860, 864 (3d Cir. 1977) (plurality opinion); see Neely v. United States, 546 F.2d 1059, 1069 (3d Cir. 1976).
If, for some reason, this case falls outside of the Katz promise “ordinarily” to defer to the district court’s exercise of discretion, that reason is not stated by the majority. Nor am I aware of any reason why this case requires such a departure from settled precepts. I would hold that the district court properly applied the relevant criteria to the facts of the case, and that its finding on the question of superiority was well within the wide range of discretion which it had.
If the district court saw fit to reconsider its class action decision at a later point in the litigation, that would be within its power under Rule 23(c)(1). But I can see no basis at present for upsetting its denial of the class certification, and I would, therefore, affirm that denial.