Suits involving large damage claims inevitably spark intensive pretrial skirmishing, as the litigants bombard each other and the district court with a variety of motions. In this case, brought by the City of New York (City), which alone has a $12,000,000 claim, as a class action alleging that General Motors Corporation (GM) has violated the antitrust laws principally by monopolizing or attempting to monopolize the nationwide market for city buses, we face appeals by GM from interlocutory orders deciding two bitterly contested pretrial, although unrelated, motions. The first is the City's successful motion to permit the suit to proceed as a class action; the second, GM's unsuccessful motion to have the City's privately-retained counsel, George D. Reycraft, disqualified for breach of the ethical precepts embodied in Canon 9 of the Code of Professional Responsibility.1 After carefully applying the Cohen2 collateral order doctrine to separate the appealable from the non-appealable order, we dismiss the appeal from the court's order determining that this action may proceed as a class action because in the context of this case that order is not appealable. With respect to the motion to disqualify counsel, however, we conclude, without intending to suggest any actual impropriety on the part of Reycraft, that his disqualification is required to 'avoid even the appearance of professional impropriety.'3 Accordingly, the court's order denying disqualification of Reycraft is reversed.

The facts necessary to an understanding of our disposition of these appeals have been gleaned, in the main, from the complaint and from the affidavits filed by the parties in support of and in opposition to the respective motions at issue. They are, thankfully, rather straightforward and, in all material respects, undisputed.

On October 4, 1972, the City filed a complaint alleging that GM had violated Section 2 of the Sherman Act4 by attempting to monopolize and monopolizing 'trade and commerce in the manufacture and sale of city buses.' The complaint contained, as a second cause of action, the allegation that GM had breached Section 7 of the Clayton Act5 by acquiring, in 1925, a controlling interest in Yellow Truck & Coach Manufacturing Co. (Yellow Coach)-- an acquisition which pruportedly 'threatens substantially to lessen competition and to tend to create a monopoly in the manufacture and sale of buses within the United State . . ..' The action, furthermore, was commenced on behalf of a class consisting of 'all non-federal governmental units and instrumentalities in the United States which have purchased or have contributed to the purchase of city buses or city bus parts . . ..' The relief sought was, inter alia, for appropriate divestiture, treble damages, costs and attorneys' fees.

According to Reycraft's affidavit, filed in opposition to the disqualification motion, he was asked by the Office of the Corporation Counsel, sometime in July 1972, to assist in the preparation of the complaint. When approached by the Corporation Counsel, then J. Lee Rankin, Reycraft responded by informing Rankin of his prior and substantial involvement in an action brought by the United States against GM, under Section 2 of the Sherman Act, based on GM's alleged monopolization of a nation-wide market for the manufacture and sale of city and intercity buses. United States v. General Motors (No. 15816, E.D.Mich.1956) (1956 Bus case).

In his affidavit, Reycraft described his participation in the 1956 Bus case, and his work for the Antitrust Division of the Department of Justice, in these words:

I was employed as an attorney for the Antitrust Division of the Department of Justice from the end of December, 1952 through the end of December, 1962. From sometime during the middle of 1954 through the end of 1962 I was employed in the Washington Office of the Antitrust Division. My initial assignment in the Washington Office of the Antitrust Division in 1954 was as a trial attorney in the General Litigation Section.

One of my first assignments as a member of the General Litigation Section was to work on an investigation of alleged monopolization by General Motors of the city and intercity bus business. The chief counsel in that matter from at least 1954 until the case was settled by Consent Decree in 1965 was Walter D. Murphy. At no time was I in active charge of the case. That investigation culminated in the Complaint filed on July 6, 1956 which I signed and in the preparation of which I participated substantially.

In 1958, I became Chief of the Special Trial Section of the Antitrust Division and no longer had any direct or indirect involvement with the 1956 Bus case. Subsequently in 1961 I became Chief of Section Operations of the Antitrust Division and had technical responsibility for all matters within the Washington Office of the Antitrust Division, including the 1956 Bus case. I have no recollection of any active participation on my part in the 1956 Bus case from 1958 through the time I departed from the Antitrust Division in December of 1962. The case was in the charge of Walter D. Murphy from its inception and he continued in charge until the Consent Decree was entered on December 31, 1965.6

In light of his substantial involvement as an employee of the Department of Justice in a matter which, at the very least, was similar to the dispute for which his retention was sought, Reycraft initially consulted his partners in the firm of Cadwalader, Wickersham & Taft and, subsequently, requested the advice of the Antitrust Division on the applicability of the Federal conflict of interest statute.7 That statute, we note, is penal in nature and its prohibitory urles, only two in number, must therefore be specifically defined and strictly construed. With that in mind, the Justice Department had little difficulty in concluding that the statute placed no bar on Reycraft's employment by the City. Its response to Reycraft states, in pertinent part:

It is clear that section 207(b) (which applies for only one year after separation from government employ) has no bearing on your case. As for section 207(a) (which applies only where the United States is a party or has a direct and substantial interest in the matter), although it appears that you participated personally and substantially in the case brought by the United States against General Motors, the Antitrust Division advises us that the United States will not be a party to or have a direct and substantial interest in the private antitrust suit by the City of New York against General Motors. Therefore, section 207(a) has no application.

Accordingly, with Cadwalader's approval and the absence of any barrier posed by federal law, Reycraft agreed to represent the City on a contingent fee basis, a not infrequent arrangement in actions where recovery is at the same time uncertain but potentially great.

On February 22, 1973, the City moved before Judge Carter for a determination that its suit could proceed as a class action pursuant to Fed.R.Civ.P. 23(a) and 23(b)(3). GM responded by opposing the class determination and, in turn, moved for the disqualification of Reycraft. Argument on both motions was subsequently heard by the court.

The City represented to Judge Carter that the class for which it sought representation would consist of 200-300 readily identifiable non-federal governmental units. It submitted a preliminary list of 177 entities identified as of the date of argument.8 The City also informed the judge that it would bear the cost of notifying all class members. The argument in the district court on the disqualification motion, although quite thorough, as indeed it was in this Court, added no material facts other than those we have already related.

In August 1973, Judge Carter entered his order, and filed an accompanying memorandum opinion,9 granting the City's motion for class action status and denying GM's motion to disqualify Reycraft. Rejecting GM's twofold contention that the requirements of Rule 23(b)(3)10 had not been met, the district judge concluded: (1) 'that the common underlying issue of liability pursuant to an unlawful, nationwide monopoly predominates over any questions as to the varying nature or amount of damages;' and (2) that the class action mechanism was the superior method for resolving this controversy because 'it is . . . inconceivable that other governmental units will not pursue such claims (of monopolization) in the event that the class action motion is denied and the suit brought by the City of New York is, or appears likely to be, successful.'

Turning to the disqualification motion, the district court recognized that DR9-101(b) requires11 Reycraft's disqualification, in order to avoid 'even the appearance of impropriety,' if his participation in this action would constitute 'private employment in a matter in which he had substantial responsibility while he was a public employee.' Since it was virtually conceded that Reycraft had 'substantial responsibility' over the 1956 Bus case, the only questions which remained were whether his engagement to represent the City was 'private employment,' and whether the City's antitrust action was, for purposes of DR9-101(B), the same 'matter' as to the 1956 Bus case. Judge Carter answered both questions in the Negative.

We need not tarry over the question of our jurisdiction to review the district court's order denying disqualification, for our recent en banc decision in Silver Chrysler Plymouth, Inc. v. Chrysler Motors Corp., 496 F.2d 800 (2d Cir. 1974) (en banc) concluded that such an interlocutory order falls within the narrow confines of permissible appeals under the collateral order doctrine. Hence, it is appealable as a 'final' order pursuant to 28 U.S.C. 1291. See e.g. Cohen v. Beneficial Industrial Loan Corp., 337 U.S. 541, 69 S. Ct. 1221, 93 L. Ed. 1528 (1949).

It is equally plain, however, in accordance with the teachings of an even more recent decision of this Court, Kohn v. Royall, Koegel & Wells, 496 F.2d 1094 (2d Cir. 1974), that interlocutory orders granting class standing are not uniformly deserving of such exceptional treatment. In Kohn, we sought to provide some guideposts to aid in discerning requisite finality.12 Our product, distilled from previous decisions of this Court, was a three-pronged test for appealability of an order granting class standing:

(1) whether the class action determination is 'fundamental to the further conduct of the case;'

(2) whether review of that order is 'separable from the merits;'

(3) whether that order will cause 'irreparable harm to the defendant in terms of time and money spent in defending a huge class action . . .'

Kohn v. Royall, Koegel & Wells, supra, at 1098. See also Herbst v. International Telephone and Telegraph, 495 F.2d 1308 (2d Cir. 1974). Here, as in Kohn, none of these requirements has been met.

The fundamental-to-the-further-conduct requirement, we concluded in Kohn, is satisfied where 'the action's viability turns on the class action determination.' Kohn v. Royall, Koegel & Wells, supra, at 1099. Although a narrow definition, we believe that our interpretation of the 'fundamentality' requirement is not only consistent with the Supreme Court's application of that concept13 but necessary, as well, lest this exception, in conjunction with the collateral order doctrine, 'swallow the salutary 'final judgment' rule.' Weight Watchers of Phila. v. Weight Watchers Int., 455 F.2d 770, 773 (2d Cir. 1972). Since it is undisputed that the City with its $12,000,000 claim would continue this action were class standing denied, it is quite clear that the first element of the appealability equation is lacking in this case.

Nor is the second prong of the appealability test fulfilled for, as in Kohn, review of the court's class action determination would take us far into the merits of the City's antitrust claim that GM has monopolized or attemped to monopolize a nationwide market in the manufacture and sale of city buses. GM, for example, vigorously disputes the district court's conclusion that proof of monopolization or attempted monopolization will present common questions of law or fact that will predominate over individual proof of damages. To buttress this contention, GM argues that 'proof relating to market, competition, economic power, (and) relevant economic behavior . . .' will vary for each purported member of the class. See Brief for Appellant at 27-28. A mere recitation of these factors demonstrates their obvious relationship to the very issues critical to the success of the City's underlying antitrust claim. Thus, we are asked at this preliminary stage in the ligation to undertake the difficult though eventually unavoidable task of product and geographic market definition, simply to determine whether evidence submitted in the course of this definitional process will be common or individual in nature. The redundancy of effort and consequent waste of judicial resources must be apparent; accordingly, we shall not belabor the point.

Finally, we do not find in this case the spectre of irreparable harm, either in terms of costs of defending the action or in its judicial management, as we did in the huge class actions before us in Eisen v. Carlisle & Jacquelin, 479 F.2d 1005 (2d Cir. 1973), aff'd, 417 U.S. 156, 94 S. Ct. 2140, 40 L. Ed. 2d 732 (1974) (estimated identifiable class of 2,250,000) and Herbst v. International Telephone and Telegraph Corp., supra (estimated identifiable class of 16,000). Although the instant case is a Rule 23(b)(3) class action, the class here, tentatively estimated at 177, does not remotely approach the thousands or, indeed, millions encountered in Eisen and Herbst. Moreover, the procedural costs unique to the class action mechanism-- essentially the cost of notice-- will be assumed by the plaintiff City, as indeed, they now must be under Eisen. And, since the City's claim of nationwide monopolization will, even absent class action status, sustain broad discovery and evidentiary admissibility at trial as to alleged national predatory paractices, the incremental cost of defending practices, the incremental cost of defending should not be significant. See Kohn v. Royall, Koegel & Wells, supra, 496 F.2d at 1100-1101.

Despite our conclusion that nonappealability is mandated by Kohn GM urged at argument that the Supreme Court's recent decision in Eisen v. Carlisle & Jacquelin, supra, might well have broadened the parameters governing the appealability of orders granting class action status. To the contrary, we find in Eisen a reaffirmation of the exceptional circumstances required to justify departure from the 'final judgment' rule-- circumstances not present in this case.

The Supreme Court, in Eisen, once again examined the policy behind the Cohen collateral order doctrine and found it applicable to the single question it was deciding-- the propriety of the notice procedure adopted by the district court.14 This, we might add, included consideration of both the method of notice and the cost of notice. The Court emphasized that the notice ruling by the district court was not 'tentative, informal or incomplete,' citing Cohen, 337 U.S. at 546, 69 S. Ct. 1221, nor did review of the order's validity require consideration with the merits of the underlying cause of action, see Cohen v. Beneficial Industrial Loan Corp., supra, 337 U.S. at 546-547, 69 S. Ct. at 1225. Accordingly, giving the requirement of finality a 'practical rather than a technical construction,' id. at 546, 69 S.Ct. at 1226, the Court, in Eisen, struck the balance between "the inconvenience and costs of piecemeal review on the one hand and the danger of denying justice by delay on the other,' Dickinson v. Petroleum Conversion Corp., 338 U.S. 507, 511, 70 S. Ct. 322, 334, 94 L. Ed. 299 (1950),'15 by permitting an immediate appeal from 'the District Court's resolution of the class action notice problems in this case . . ..' Eisen v. Carlisle & Jacquelin, supra, 417 U.S. at 172, 94 S. Ct. at 2150.

Applying each of the considerations discussed in Eisen to the facts here, it is clear to us that the district court's order granting class action status in this case is not now appealable. Unlike the rulings in Cohen and Eisen, Judge Carter's determination to permit the City's action to proceed as a class action is very much 'tentative,' subject always to reconsideration as the cause of action unfolds.16 We noted in Green v. Wolf Corp., 406 F.2d 291, 298 (2d Cir. 1968), cert. denied, 395 U.S. 977, 89 S. Ct. 2131, 23 L. Ed. 2d 766 (1969):

Rule 23 now emphasizes the flexibility which a trial court exercises in the management of the action . . .. It is this flexibility which indeed enables us to view liberally claims which assert a right to a class action . . . at the early stages of the litigation . . ..

Thus, if the concern voiced by GM at argument respecting the potentially explosive class size17 or the possible confusion bred by the disparate proof required for a proper market analysis should become a reality, we are confident that the district court will act appropriately, limiting or even striking the class if necessary.

We consider it imperative, furthermore, to recognize the sharp distinction between the bases for attacking the class action determination in this case as contrasted with Eisen. In Eisen, as noted, the Supreme Court focused narrowly on the unprecedented and extraordinary notice procedure authorized by the district court. In this case, however, there is nothing unique about the notice procedure approved by the court below-- individualized notice is to be given to all identifiable class members, and at the City's expense. Rather, GM's attack on the class action determination rests, as we have indicated, on the twofold contention that the requirements of Rule 23(b)(3) are not met because: (1) common questions of law or fact do not predominate and (2) the class action is not 'superior to other available methods for the fair and efficient adjudication of the controversy.' Accordingly, we would review here not a finite and conclusive determination of judicial power-- e.g. the power to shift notice costs and forego individualized notice, as in Eisen, or the power to dispense with security, as in Cohen-- but a discretionary decision, the propriety of which will necessarily vary from case to case. That this distinction is of fundamental importance in the calculus of appealability was plainly acknowledged in Cohen itself. Mr. Justice Jackson stated:

But we do not mean that every order fixing security is subject to appeal. Here it is the right to security that presents a serious and unsettled question. If the right were admitted or clear and the order involved only an exercise of discretion as to the amount of security, a matter the statute makes subject to reconsideration from time to time, appealability would present a different question.

Cohen v. Beneficial Industrial Loan Corp., supra, 337 U.S. at 547, 69 S. Ct. at 1226. And see Weight Watchers of Phila. v. Weight Watchers Int., supra, 455 F.2d at 773.

A final distinction, which cannot be overlooked, is that in reviewing the legitimacy of the notice procedure authorized in Eisen, the Court was not called upon to consider any aspect of Eisen's underlying antitrust action against the odd-lot brokerage firms. Accommodating the notice procedure contemplated by Rule 23(c)(2) with the overriding considerations of due process is a task wholly divorced from review of the underlying merits of the action.

Although the interlocutory order granting class action status is thus clearly not appealable at this nascent stage in the litigation, GM argues that since we have jurisdiction to hear the appeal from the district court's order denying disqualification, we should exercise our discretionary power to consider this nonappealable order as well. To be sure, the doctrine of pendent jurisdiction at the appellate level has been utilized. See e.g. Deckert v. Independence Shares Corp., 311 U.S. 282, 61 S. Ct. 229, 85 L. Ed. 189 (1940); Green v. Wolf Corp., supra, Semmes Motors, Inc. v. Ford Motor Co., 429 F.2d 1197, 1201 (2d Cir. 1970); 9 J. Moore, Federal Practice P110.25(1) (2d ed. 1973). But, further analysis suggests the inappropriateness of exercising that power here.

The guiding principle to inform the discretionary application of pendent jurisdiction is whether review of the appealable order will involve consideration of factors relevant to the otherwise nonappealable order. See 9 J. Moore, supra, P110.25(1) at 272. In this case, it is transparently obvious that no such overlap exists between the order granting class action status and the order denying disqualification of counsel. 'In addition,' according to Professor Moore:

where an interlocutory order is properly appealed and review is sought of an incidental discretionary order, the appellate court should not review the discretionary order where there is no showing of an abuse of discretion.

Id. Professor Moore's observation suggests that he would look with disfavor upon the exercise of pendent jurisdiction in the instant case because the determination to grant class standing for a relatively small class (less than 200), where notice will be given to all identifiable class members and paid for by the plaintiff City, is clearly not a prima facie abuse of discretion. And, indeed, it would be particularly inappropriate for us to reach out to review this order, at this very early stage in the litigation, for, as we have already indicated, the district judge retains the flexibility to modify or even strike the class should further developments prove his initial determination erroneous.

In a final effort to gain review of the class action determination, GM has also followed the mandamus route in the event we should hold the class action order nonappealable. In denying the mandamus petition, suffice to say, as we did in Donlon Industries, Inc. v. Forte, 402 F.2d 935, 937 (2d Cir. 1968) and repeated in Weight Watchers of Phila. v. Weight Watchers Int., supra, 455 F.2d at 775:

we do not-- indeed may not-- issue mandamus with respect to orders resting in the district court's discretion, save in (the) most extraordinary circumstances not remotely presented here.

We turn now to GM's unsuccessful motion to disqualify the City's privately-retained counsel, George Reycraft. It is necessary that we begin our discussion by focusing again on the language of Canon 9 of the Code of Professional Responsibility:

A lawyer should avoid even the appearance of professional impropriety.

Providing a measure of specificity to this general caveat, DR9-101(B) commands:

A lawyer shall not accept private employment in a matter in which he had substantial responsibility while he was a public employee.

The purpose behind this plain interdiction is not difficult to discern. Indeed, the City recognizes its salutary goal, as stated by the ABA Comm. on Professional Ethics, Opinions, No. 37 (1931)18 to be:

(to avoid) the manifest possibility that . . . (a former Government lawyer's) action as a public legal official might be influenced (or open to the charge that it had been influenced) by the hope of later being employed privately to uphold or upset what he had done.

Id. at 124. Viewed in this light, the question before us is whether Reycraft's decision to represent the City on a contingent fee basis in an antitrust suit strikingly similar, though perhaps not identical in every respect, to an antitrust action brought over his signature by the Department of Justice would raise an 'appearance of impropriety,' as private employment 'to uphold . . . what he had done' as a Government lawyer. Unlike the court below, we are constrained to answer in the affirmative.

Before we commence our analysis, we would do well to recall the following description of our task:

We approach our task as a reviewing court in this case conscious of our responsibility to preserve a balance, delicate though it may be, between an individual's right to his own freely chosen counsel (we do not presume the City to have a lesser right) and the need to maintain the highest ethical standards of professional responsibility. This balance is essential if the public's trust in the integrity of the Bar is to be preserved.

Emle Industries, Inc. v. Patentex, Inc., 478 F.2d 562, 564-565 (2d Cir. 1973).

Indeed, the 'public's trust' is the raison d'etre for Canon 9's 'appearance-of-evil' doctrine. Now explicitly incorporated in the profession's ethical Code,19 this doctrine is directed at maintaining, in the public mind, a high regard for the legal profession. The standard it sets-- i.e. what creates an appearance of evil-- is largely a question of current ethical-legal mores. See Kaufman, The Former Government Attorney and the Canons of Professional Ethics, 70 Harv.L.Rev. 657, 660 (1957).

Nor can we overlook that the Code of Professional Responsibility is not designed for Holmes' proverbial 'bad man' who wants to know just how many corners he may cut, how close to the line he may play, without running into trouble with the law. Holmes, The Path of the Law, in Collected Legal Papers 170 (1920). Rather, it is drawn for the 'good man,' as a beacon to assist him in navigating an ethical course through the sometimes murky waters of professional conduct. Accordingly, without in the least even intimating that Reycraft himself was improperly influenced while in Government service, or that he is guilty of any actual impropriety in agreeing to represent the City here, we must act with scrupulous care to avoid any appearance of impropriety lest it taint both the public and private segments of the legal profession.

It is undisputed that Reycraft had 'substantial responsibility' in initiating the Government's Sherman 2 claim against GM for monopolizing or attempting to monopolize the nationwide market for city and intercity buses. Thus, we are left to determine whether the City's antitrust suit is the same 'matter' as the Government's action and whether Reycraft's contingent fee arrangement with the City constitutes 'private employment.'Directing our attention to the simpler question first, we are convinced beyond doubt that Reycraft's and, indeed, his firm's opportunity to earn a substantial fee for Reycraft's services is plainly 'private employment' under DR9-101(B). The district judge apparently grounded his contrary decision on the rationale that Reycraft 'has not changed sides'-- i.e. 'there is nothing antithetical in the postures of the two governments in the actions in question . . ..' But, as we have already noted, Opinion No. 37 of the ABA Commission on Professional Ethics unequivocally applies the ethical precepts of Canon 9 and DR9-101(B) irrespective of the side chosen in private practice.20 And see Allied Realty of St. Paul v. Exchange Nat. Bank of Chicago, 283 F. Supp. 464, 466 (D.Minn.1968). We believe, moreover, that this is as it should be for there lurks great potential for lucrative returns in following into private practice the course already charted with the aid of governmental resources. And, with such a large contingent fee at stake, we could hardly accept 'pro bono publico' as a proper characterization of Reycraft's work, simply because the keeper of the purse is the City of New York or other governmental entities in the class.

It is manifest also, from an examination of the respective complaints (see the appendix to this opinion), that the City's antitrust action is sufficiently similar to the 1956 Bus case to be the same 'matter' under DR9-101(B). Indeed, virtually every overt act of attempted monopolization alleged in the City's complaint is lifted in haec verba from the Justice Department complaint. We cite, merely by way of illustration, paragraphs appearing in both complaints alleging the withdrawal of more than 20 companies from bus manufacturing, the coincidence of directors on the boards of GM and another bus manufacturer, the Flxible Company, and GM's acquisition of a controlling stock interest in Yellow Coach in 1925.

To be sure, as the City urges, the four-year statute of limitations, embodied in 15 U.S.C. 15b, requires the City to focus on market conditions since 1968, some ten years after Reycraft ceased his involvement in the Bus case.21 But, an equally essential element in proving a violation of Section 2 of the Sherman Act is either an intent to monopolize or an abuse of monopoly power. See United States v. Grinnell Corp., 384 U.S. 563, 570-571, 86 S. Ct. 1698, 16 L. Ed. 2d 778 (1966); Coniglio v. Highwood Services, Inc., 495 F.2d 1286, 1293 (2d Cir. 1974); United States v. Aluminum Co. of America, 148 F.2d 416, 429-430 (2d Cir. 1945). Moreover, to decide the question whether GM is a passive recipient of monopoly power, a history of its operations will be imperative. See e.g. United States v. Aluminum Co. of America, supra (included an exhaustive study of Alcoa's operations from 1902 to the date of the lawsuit). Accordingly, at the very forefront of the City's case will be proof of alleged predatory practices amassed by the United States, with the substantial participation of Reycraft, when the Justice Department built its case against GM in 1956.

The addition of the Clayton Act claim, based solely on the same 1925 Yellow Coach acquisition which was part of the Sherman Act violation alleged by both the United States and the City, hardly alters the nuclear identity of these two suits.22 Both, after all, allege monopolization or attempted monopolization of the same product line23 -- city buses-- and, in the same geographic market-- the United States. The subtleties of differential proof will not obviate the 'appearance of impropriety' to an unsophisticated public. We opined in Emle:

Nowhere is Shakespeare's observation that 'there is nothing either good or bad but thinking makes it so,' more apt than in the realm of ethical considerations.

Emle Industries, Inc. v. Patentex, Inc., supra, 478 F.2d at 571.

The City maintains, in the end, that if we reverse the court below and disqualify Reycraft, we will chill the ardor for Government service by rendering worthless the experience gained in Government employ. Indeed, the author of this opinion is hardly unaware of this claim, for he has cautioned:

If the government service will tend to sterilize an attorney in too large an area of law for too long a time, or will prevent him from engaging in the practice of a technical specialty which he has devoted years in acquiring, and if that sterilization will spread to the firm with which he becomes associated, the sacrifice of entering government service will be too great for most men to make.

Kaufman, supra, 70 Harv.L.Rev. at 668. But, in that commentary, and the case upon which it was based (United States v. Standard Oil Co. (N.J.), 136 F. Supp. 345 (S.D.N.Y.1955)-- Esso Export Case), the accommodation between maintaining high ethical standards for former Government employees, on the one hand, and encouraging entry into Government service, on the other, was struck under far different circumstances. Unlike the instant case, in which Reycraft's 'substantial responsibility' in the Bus case is undisputed, the writer of this opinion concluded in Esso Export that the lawyer:

never investigated or passed upon the subject matter of the pending case . . . never rendered or had any specific duty to render any legal advice in relation to the regulations involved in the litigation.

Kaufman, supra, 70 Harv.L.Rev. at 664. More to the point, therefore, is another admonition voiced in that article:

If there was a likelihood that information pertaining to the pending matter reached the attorney, although he did not 'investigate' or 'pass upon' it, . . . there would undoubtedly be an appearance of evil if he were not disqualified.

Id. at 665.

Esso Export unquestionably presented a case for the cautious application of the 'appearance-of-evil doctrine,' because the former Government lawyer's connection with the matter at issue was the tenuous one of mere employment in the same Government agency. If, for example, Reycraft had not worked on the 1956 Bus case, but was simply a member of the Antitrust Division at that time, a case not unlike Esso Export would be before us. To the contrary, however, Reycraft not only participated in the Bus case, but he signed the complaint in that action and admittedly had 'substantial responsibility' in its investigatory and preparatory stages. Where the overlap of issues is so plain, and the involvement while in Government employ so direct, the resulting appearance of impropriety must be avoided through disqualification.

Accordingly, we dismiss the appeal from the order granting class action status, and reverse the court's order denying disqualification of Reycraft.

The following comparison of portions of the complaints filed by the United States in the 1956 Bus case and by the City in the instant case was included in the record below:

United States v. General Motors Corporation

(1956 Bus case)

15. Between 1925 and 1955 more than 20 manufacturers of buses withdrew from the bus manufacturing business. Among these companies are the following:

International Harvester Co.

Reo Motors, Inc.

Beaver Metropolitan Coaches, Inc.

Kalamazoo Coaches, Inc.

Transicoach, Inc.

Cub Industries, Inc.

16. After 1946, no concern not theretofore engaged in the manufacture of buses has commenced the manufacture of this product.

17. Defendant manufactured approximately 85 per cent of the new buses delivered in the United States during 1955. In that year it delivered 2,724 buses, having a value of approximately $55,000,000. Defendant has manufactured at least 65 per cent of the new buses delivered in the United States during each year from 1952 to 1955 inclusive.

20. The Flxible Company (hereinafter referred to as 'Flxible'), has been engaged in the manufacture of buses for more than twenty-five years . . .. For many years Charles F. Kettering has been the Chairman of the Board of Directors of Flxible and holder of more Flxible stock than any other stockholder of Flxible. During much of the same period of time he also was an officer and director of General Motors.

22. For many years continuously up to the date of the filing of this complaint, the defendant has been monopolizing the above-described trade and commerce in the manufacture and sale of buses in violation of Section 2 of the Sherman Act (15 U.S.C. 2). The defendant threatens to continue said offense and will continue it unless the relief hereinafter prayed for is granted.

23. Pursuant to and in effectuation of the aforesaid monopolization, the defendant directly and through its subsidiaries, among other things, has done the following:

(a) Acquired a controlling interest in one bus manufacturer . . .;

(b) Acquired the power to influence the policies of its principal existing competitor in the manufacture of intercity buses by having an officer and director of General Motors as Chairman of the Board of Directors and principal stockholder of that company;

(c) Acquired the power to control some bus operating companies, and influenced the policies of others by:

(1) purchasing their capital stock;

(2) having officers and employees of General Motors own their capital stock;

(3) having officers or directors of General Motors on their boards of directors;

(4) promoting their formation and assisting in their organization and expansion.

(d) Entered into contracts with bus operating companies which required them to purchase stated percentage of their requirements of buses from defendant;

(e) Induced some bus operating companies to purchase buses from defendant exclusively or on a preferential basis by:

(1) agreeing to refuse and refusing to sell buses to competitors of some of said companies;

(2) offering to sell and selling some of said companies buses and parts for buses at prices lower or on terms more favorable, or both, than the prices or terms offered by defendant to other bus operating companies; and

(3) making substantial loans of money to some of said companies.

(f) Offered to finance and financed the sale of buses through YMAC on terms which General Motors' competitors with more limited resources have been unable to meet;

(g) Induced officials of municipally-owned transit systems to adopt bus specifications for use in obtaining so-called competitive bids which prevented other bus manufacturers from competing;

(h) Refused to sell to other bus manufacturing companies and to their suppliers, various parts, including automatic transmissions and diesel engines, for use in the manufacture of buses;

(i) Entered into contracts under the terms of which defendant was granted the exclusive right to use various improvements (patented and unpatented), including 'air suspension' and automatic transmissions, in the design and manufacture of buses;

(j) Made surveys of bus operating companies for the purpose and with the effect of inducing said companies to agree to purchase or to purchase buses from defendant on an exclusionary basis;

26. The effects of the aforesaid offenses, among others, have been and are:

(a) To drive most of General Motors' competitors out of the bus manufacturing business;

(b) To curtail the supply of new buses readily available for purchase;

(c) To deprive bus operating companies and the bus-riding public of the benefits of competition in the manufacture of buses;

(d) To increase prices paid by bus operating companies . . . for some types of buses;

(e) To prevent other concerns from entering the bus manufacturing business;

City of New York v. General Motors Corporation

16. Between 1925 and 1971, more than 20 bus manufacturers or bus assemblers withdrew from business. Among these companies were the following:

International Harvester Co.

Reo Motors, Inc.

Beaver Metropolitan Coaches, Inc.

Kalamazoo Coaches, Inc.

Transicoach, Inc.

Cub Industries, Inc.

Mack Trucks, Inc.

Southern Coach Manufacturing Co., Inc.

17. After 1946, no concern not theretofore engaged in the manufacture of buses has commenced the manufacture of buses.

11. Since 1955, General Motors has manufactured over 80 percent of all buses manufactured in the United States. Since 1965, it has manufactured and sold between 85 percent to 100 percent of all city and intercity buses manufactured in the United States.

19. The Flxible Company ('Flxible') has been in business for over 40 years. For many years, Charles F. Kettering was Chairman of the Board of Directors of Flxible and the largest individual stockholder of Flxible; during much of the same period, Mr. Kettering was also an officer and director of General Motors.

21. For many years continuously up to the date of the filing of this complaint, defendant has violated Section 2 of the Sherman Act (15 U.S.C. 2) by attempting to monopolize and by monopolizing the abovedescribed trade and commerce in the manufacture and sale of city buses. The defendant threatens to continue said offenses and violations and will continue them unless the relief hereinafter prayed for is granted.

22. Pursuant to and in effectuation of the aforesaid attempt to monopolize and monopolization, the defendant directly and through its subsidiaries has done, among other things, the following:

(a) Acquired controlling stock interest in one bus manufacturer.

(d) Acquired the power to influence the policies of its principal existing competitor in the manufacture of intercity buses by having an officer and director of General Motors as Chairman of the Board of Directors and principal stockholder of that company.

(e) Acquired the power to control some bus operating companies, and influenced the policies of others by:

(1) purchasing their capital stock;

(2) having officers and employees of General Motors own their capital stock;

(3) having officers or directors of General Motors on their boards of directors;

(4) promoting their formation and assisting in their organization and expansion.

(f) Entered into contracts with bus operating companies which required them to purchase stated percentages of their requirements of buses from defendant.

(g) Induced some bus operating companies to purchase buses from defendant exclusively or on a preferential basis by:

(1) agreeing to refuse and refusing to sell buses to competitors of some of said companies;

(2) offering to sell and selling some of said companies buses and parts for buses at prices lower or on terms more favorable, or both, than the prices or terms offered by defendant to other bus operating companies; and

(3) making substantial loans of money to some of said companies.

(h) Offered to finance and financed the sale of buses on terms which General Motors' competitors with more limited resources have been unable to meet.

(i) Persuaded officials of municipally-owned transit systems to adopt bus specifications for use in obtaining so-called competitive bids which prevented other bus manufacturers from competing.

(j) Refused to sell other bus manufacturing companies, and to their suppliers, various parts, including automatic transmissions and diesel engines, for use in the manufacture of buses.

(k) Entered into contracts under the terms of which defendant was granted the exclusive right to use various improvements (patented and unpatented), including 'air suspension' and automatic transmissions, in the design and manufacture of buses.

(l) Made surveys of bus operating companies for the purpose and with the effect of inducing said companies to agree to purchase or to purchase buses from defendant on an exclusionary basis.

26. The effects of the aforesaid offenses and violations by defendant, among others, have been and are:

(a) To drive all of General Motors' competitors in the United States out of the new bus manufacturing business;

(c) To curtail the supply of new city buses readily available for purchase;

(e) To deprive bus operators, the bus-riding public, the plaintiff, and the other members of the class of the benefits of competition in the manufacture and sale of new City buses . . .;

(d) To cause and maintain excessive and noncompetitive prices to be paid by bus operators . . . for new city buses and bus parts;

(f) To prevent or discourage other concerns from entering into the bus manufacturing business in the United States;MANSFIELD, Circuit Judge (concurring):

Although I disagree in part with the criteria used by Chief Judge Kaufman for determining the appealability of orders granting class action status, I concur in all other aspects of his characteristically thoughtful and thorough opinion, including the result reached on the issue of appealability at the present posture of this case.

The formulation of principles for determining when an interlocutory order may be treated as an appealable 'final decision' under 28 U.S.C. 1291 has repeatedly been recognized by the Supreme Court to be an extremely difficult task. In Gillespie v. United States Steel Corp., 379 U.S. 148, 152, 85 S. Ct. 308, 311, 13 L. Ed. 2d 199 (1964), for instance, Justice Black stated:

'And our cases long have recognized that whether a ruling is 'final' within the meaning of 1291 is frequently so close a question that decision of that issue either way can be supported with equally forceful arguments, and that it is impossible to devise a formula to resolve all marginal cases coming within what might well be called the 'twilight zone' of finality. Because of this difficulty this Court has held that the requirement of finality is to be given a 'practical rather than a technical construction.' Cohen v. Beneficial Industrial Loan Corp., supra, 337 U.S. at 546 (69 S.Ct. at 1226).'

More recently in Eisen v. Carlisle & Jacquelin, 417 U.S. 156, 170, 94 S. Ct. 2140, 40 L. Ed. 2d 732 (1974) ('Eisen III'), Justice Powell commented:

'No verbal formula yet devised can explain prior finality decisions with unerring accuracy or provide an utterly reliable guide for the future. We know, of course, that 1291 does not limit appellate review to 'those final judgments which terminate an action . . .,' Cohen v. Beneficial Loan Corp., 337 U.S. 541, 545, 69 S. Ct. 1221, 1225, 93 L. Ed. 1528 (1949), but rather that the requirement of finality is to be given a 'practical rather than a technical construction.' Id., at 546 (, 69 S. Ct., at 1226). The inquiry requires some evaluation of the competing considerations underlying all questions of finality-- 'the inconvenience and costs of piecemeal review on the one hand and the danger of denying justice by delay on the other.' Dickinson v. Petroleum Conversion Corp., 338 U.S. 507, 511, 70 S. Ct. 322, 324, 94 L. Ed. 299 (1950).'

With the emphasis thus placed on 'practical' construction we are once again required to decide what test should be applied to determine the appealability of an order granting class action status under Rule 23(b)(3), F.R.Civ.P., see, e.g., Herbst v. International Telephone and Telegraph Corp., 495 F.2d 1308 (2d Cir. 1974); and Kohn v. Royall, Koegel & Wells, 496 F.2d 1094 (2d Cir. 1974), as distinguished from orders denying such certification, see, e.g., Eisen v. Carlisle & Jacquelin, 370 F.2d 119 (2d Cir. 1966), cert. denied, 386 U.S. 1035, 87 S. Ct. 1487, 18 L. Ed. 2d 598 (1967) ('Eisen I'); Green v. Wolf Corp., 406 F.2d 291, 295 n. 6 (2d Cir. 1968), cert. denied, 395 U.S. 977, 89 S. Ct. 2131, 23 L. Ed. 2d 766 (1969); Korn v. Franchard Corp., 443 F.2d 1301 (2d Cir. 1971); Caceres v. International Air Transport Assn., 422 F.2d 141 (2d Cir. 1970).

While the 'three-pronged' test1 embraced by Judge Kaufman sets forth relevant factors that should be considered in deciding whether a class action certification is appealable, not all of them strike me as mandatory conditions precedent to review. Moreover, I fear that too strict or mechanical an adherence to the test, coupled with the engrafting of excessively specific conditions on appealability, will lead to violation of the Supreme Court's direction that 1291 be given a 'practical rather than a technical construction.' Cohen v. Beneficial Industrial Loan Corp., 337 U.S. at 546, 69 S. Ct. at 1226.

Turning to the first of the three prongs, the requirement that the district court's order be 'fundamental to the further conduct of the case,' see United States v. General Motors Corp., 323 U.S. 373, 377, 65 S. Ct. 357, 359, 89 L. Ed. 311 (1945); Gillespie v. United States Steel Corp., 379 U.S. 148, 153, 85 S. Ct. 308, 13 L. Ed. 2d 199 (1964), I do not agree that an order granting class action status should be appealable only upon a showing that the action would otherwise have withered on the vine. Nothing in the origin or history of the 'fundamentality' phrase so indicates.2 Where class action status has been denied by the district court, the likelihood that the action will survive as an individual suit or collapse entirely should weigh heavily in the scales of appealability, as Judge Kaufman pointed out in forging the 'death knell' doctrine in Eisen I. However, where class certification has been granted by the district court it does not follow that merely because the suit would have continued as an individual action if class certification had been refused, review should be denied. The logic of such a 'reverse death-knell' doctrine, although it may have surface appeal to those dedicated to sheer symmetry, escapes me. Nor do In interpret Judge Friendly's reference in Korn v. Franchard Corp., 443 F.2d 1301, 1307 (2d Cir. 1971) to 'equality of treatment as between plaintiffs and defendants' as sponsoring it. Indeed he there voiced serious doubts as to the workability of the 'death-knell' doctrine.

Despite the fact that, if class action status is denied, a class suit may continue as a viable individual action, practical considerations may dictate the advisability in some instances of immediate review of an order granting class action status. One such consideration is the likelihood that because of the sheer size and complexity of the action, the added time, expense and effort needed to defend it as a class suit may force the defendant, despite the doubtful merit of the claims, to settle rather than to pursue the long and costly litigation route required for review of the class action certification. See West Virginia v. Chas. Pfizer & Co., 314 F. Supp. 710 (S.D.N.Y.1970), affd., 440 F.2d 1079 (2d Cir.), cert. denied, 404 U.S. 871, 92 S. Ct. 81, 30 L. Ed. 2d 115 (1971); Herbst v. International Telephone and Telegraph Corp., 495 F.2d 1308, 1312-1313 (2d Cir. 1974); American College of Trial Lawyers, Report and Recommendations of the Special Committee on Rule 23 of the Federal Rules of Civil Procedure 15-17 (1972). To be meaningful the review must be from an interlocutory order. Otherwise appellate review, although theoretically available years later after trial, becomes prohibitive as a practical matter, principally because of the time and expense factor. Despite the pendency of hundreds of Rule 23(b)(3) class suits, only a handful have been tried to judgment, much less thereafter reviewed by a court of appeals. American College of Trial Lawyers, Report, supra, 15. In the meantime, because most class damage suits prove to be behemoths in the world of litigation, the district court judge who has been saddled with actions that have erroneously been granted class status under Rule 23(b)(3) must assume onerous and time-consuming duties not characteristic of most other litigation. He must make more interlocutory determinations than in other actions. He must identify the class and determine what sub-classes, if any, are to be established. Green v. Wolf Corp., 406 F.2d 291, 299 (2d Cir. 1968), cert. denied, 395 U.S. 977, 89 S. Ct. 2131, 23 L. Ed. 2d 766 (1969). He must supervise the type and terms of the notice and other communications to be sent to members of the class. He must classify and handle queries from class members, determine whether and when filing of proofs of claim by members will be required. He must register opt-outs, supervise pretrial discovery, formulate pretrial orders, and handle a plethora of motions. He must consider proposed settlements and hold hearings with respect to them. He must preside at trial of common issues, if trial there is to be, preside at such later trials of any individual issues as may occur, and supervise the distribution of awards.

We should not allow this complex, time-consuming and expensive process to be launched without ample justification. As we said recently in Herbst, supra, 495 F.2d at 1313:

'We believe that in the exercise of our supervisory powers over the administration of justice in the district courts it is desirable for us to review orders authorizing class actions before the parties and the district courts expend large amounts of time and money in managing them. Candor compels us to add that as appellate judges we would be reluctant to hold that a class action had been improper after the district court and the parties had expended much time and resources although we might have had serious doubts if we had reviewed the question at the inception of the action. Judicial efficiency requires that appellate review be made before the parties and district courts have spent considerable time, effort, and money, on such actions.'

Our initial reaction, upon first being exposed eight years ago to Rule 23(b) (3), was to urge that it be construed liberally, see Eisen II, 391 F.2d 555, 563 (2d Cir. 1968) ('we hold that the new rule should be given a liberal rather than a restrictive interpretation') (Medina, C.J.) and Green v. Wolf Corp., 406 F.2d 291, 298 (2d Cir. 1968) (the 'guiding principle' should be that 'if there is to be an error made, let it be in favor of and not against the maintenance of the class action') (Kaufman, C.J., quoting from Esplin v. Hirschi, 402 F.2d 94, 99 (10th Cir. 1968)). The time may have come for reappraisal of this philosophy, which may prove to ave been somewhat overly enthusiastic. However, as long as the district courts in this Circuit are so advised, I question the wisdom, much less the practicality, of adopting a restrictive set of conditions for review. Such tests, especially if applied in a wooden-like fashion, may have the effect of imposing an unnecessary and staggering burden upon the district courts and of delaying justice until the central issue becomes academic. Although the Supreme Court's decision in Eisen III, 417 U.S. 156, 94 S. Ct. 2140, 40 L. Ed. 2d 732 (1974), has dealt a lethal blow to those consumer-type class suits in which the named plaintiffs cannot afford the cost of individual notice to the class, the courtroom doors remain open to numerous actions where the class is smaller than in Eisen and the members are identifiable. See e.g., Illinois v. Harper & Row Publishers, Inc., 301 F. Supp. 484 (N.D.Ill.1969). The present case is but one example. Rather than hold that an order granting class action status will be reviewed only if the action could not have been continued as an individual suit, I would dispense with the 'fundamentality' requirement as a condition for such review and use the simple test of weighing 'the inconvenience and costs of piecemeal review' against 'the danger of denying justice by delay.' See Dickinson v. Petroleum Conversion Corp., 338 U.S. 507, 511, 70 S. Ct. 322, 324, 94 L. Ed. 2d 299 (1950); Gillespie v. United States Steel Corp., supra, 379 U.S. pp. 152-153, 85 S. Ct. 308.3

Turning to the second requirement-- that a class certification, to be appealable, must be 'separable from the merits'-- strict application of this condition would preclude almost any review of a trial judge's determination that common questions 'predominate' or that a class action is 'superior' to other available methods for adjudicating the controversy. In the absence of any judicial or legislative definition of the term 'predominate' as it is used in Rule 23(b)(3), the trial judge, upon being faced with some claims and proof common to all class members (e.g. joint action on the part of defendants) and others that are individual to each class member (e.g., differing affirmative representations made to different members, materiality, or proximate cause), looks for some controlling principles of general application. See, e.g., Epstein v. Weiss, 50 F.R.D. 387, 393 (E.D.La.1970); Siegel v. Chicken Delight, Inc., 271 F. Supp. 722, 726 (N.D.Cal.1967). Guiding principles are also required to determine whether injunctive relief, administrative action, or some other remedy would be superior to a damage suit.

For these enormously important threshold decisions, which determine whether the class action genie is to be released from the bottle, the district courts need guidance. Instead, because so few class action determinations ever reach the point of appellate review, the trial judges are asked to 'fly by the seat of their pants.' Despite the passage of seven years since the adoption of Rule 23, except in one or two instances, see, e.g., Green v. Wolf Corp., 406 F.2d 291, 300-301 (2d Cir. 1968), cert. denied, 395 U.S. 977, 89 S. Ct. 2131, 23 L. Ed. 2d 766 (1969); Korn v. Franchard Corp., 456 F.2d 1206 (2d Cir. 1972), we have yet to provide any meaningful guidelines interpreting key terms used in the Rule, such as 'predominate' and 'superior' found in (b)(3) and 'as soon as practicable' in (c)(1). As a result, district judges, left to fend for themselves, have adopted varying individual standards. We cannot disregard the fact that some judges, possibly influenced by our early exhortations to take a liberal view of the Rule, have tended to 'jump the gun' by making an early determination in favor of a class certification without benefit of adequate data. Others, perhaps more aware of the consequences of unleashing the titanic class action weapon, have sought to avoid the harmful effects of an unwarranted or premature grant of class action status in a doubtful case by adopting the principle that 'it is preferable to proceed with the claims of those before the Court, without prejudice to subsequent class action treatment if, upon further development of the litigation, it appears appropriate.' Morris v. Burchard, 51 F.R.D. 530 at 536 (S.D.N.Y.1971); See also Schaffner v. Chemical Bank, 339 F. Supp. 329 (S.D.N.Y.1972).

Similarly some judges, at least where the plaintiff's individual claim is large enough to be prosecuted individually, have refused to grant class certification in the absence of evidence that additional duplicate suits have been instituted or that a multiplicity of suits is threatened. See, e.g., Caceres v. International Air Transport Ass'n, 46 F.R.D. 89 (S.D.N.Y.1969), appeal dismissed, 422 F.2d 141 (2d Cir. 1970); Free World Foreign Cars, Inc. v. Alfa Romeo, S.p.A., 55 F.R.D. 26 (S.D.N.Y.1972); City of New York v. International Pipe & Ceramics Corp., 44 F.R.D. 584 (S.D.N.Y.1968), appeal dismissed, 410 F.2d 295 (2d Cir. 1969). In contrast, others have indicated that a class action might be warranted in the absence of evidence of other existing or prospective litigation, see Fidelis Corp. v. Litton Industries, 293 F. Supp. 164, 171 (S.D.N.Y.1968).4 In the present case Judge Carter has granted class action status despite the fact that the City would continue its own individual suit for $12,000,000 and despite the fact that, although 17 years have elapsed since the institution of the Government suit upon which the present action is based, and 8 years have passed since the settlement of the Government's action by entry of a consent decree, no other governmental body has seen fit to institute a coattail action of its own, much less to join as a named co-plaintiff in the City's suit.

Such sharply conflicting views as to the conditions required for maintenance of an action as a class suit under Rule 23(b)(3) indicate that the decision in close cases may depend on the interpretation and philosophy of the judge rather than upon uniform principles established by appellate review. The uncertainties posed by Rule 23 should be resolved in a manner that will enable district judges to act with some degree of uniformity. For these reasons and to permit meaningful review and appellate formulation of critical guidelines, I would suggest that, in lieu of a 'separability from the merits' test (which is not mandated by Cohen v. Beneficial Industrial Loan Corp.), the district court be required, where it has granted class certification, to make definitive findings prior to trial on the issues of 'predominance' and 'superiority.' Review might then be permitted upon a prima facie showing that the findings are 'clearly erroneous.' Cf. Ecology Action v. AEC, 492 F.2d 998, 1001 (2d Cir. 1974). Such a condition would permit review without undue involvement in the merits or excessive examination of record proof. It would also allow the district judge sufficient time to determine prior to trial whether an action that has tentatively been accorded class action status should be decertified.5

With respect to the third requirement of the 'tripartite analysis,' i.e., that the defendant demonstrate that the order will cause 'irreparable harm to the defendant in terms of time and money spent in defending a huge class action,' I take no issue except to state that the additional judicial management necessitated by prosecution of a suit as a class rather than an individual action should also be taken into consideration. Satisfaction of this condition, however phrased, will turn on the facts of the particular case presented for review. If common questions of fact do not predominate over those affecting individual members, the scope of proof, and accordingly the time, expense and resources, required to defend a class suit should be substantially greater than that required to defend an individual action, and the judicial time and energy consumed should be proportionately greater. It therefore would be a rare case where the grant of class action status would not meet this test.

Applying the foregoing considerations to this case, I would at this point in time reach the same conclusion on the issue of appealability as that reached by Judge Kaufman. Of necessity Judge Carter's conclusions on the essential elements of 'predominance' and 'superiority' had to be extremely tentative. Having virtually no facts with respect to these essential issues, he based his decision principally upon the allegations of the complaint and his speculation, despite many years of experience to the contrary, that other government units would probably bring similar suits. Under the circumstances he would have been better advised, in my view, to withhold class certification until relevant facts could be developed through pretrial discovery.6 However, as Judge Kaufman accurately observes, nothing prevents the district court, upon presentation of evidence, from concluding at a later point that the action must be decertified as a class suit because common questions do not predominate. In view of that possibility, it should be recognized that we are not here confronted with a claim of a single price-fixing conspiracy or agreement in violation of 1 of the Sherman Act where, as with a charge of fraud in violation of the federal securities laws, liability can be established by proof of one set of facts common to all parties, such as a price agreement or fraudulent prospectus. The principal claim here is monopolization in violation of 2 of the Sherman Act, which presents a potential for widely varying questions of law and fact that may not, at this early stage in the proceedings have been fully appreciated.

The complaint, which is virtually a copy of a civil complaint filed by the Government years ago, charges monopolization on a nationwide scale and alleges a series of some 12 or more separate and distinct monopolization activities. GM has offered to show that the relevant facts and circumstances with respect to many members of the proposed class differ radically from those existing as to others, thus giving rise to what in effect is a series of separate relevant markets and different competitive conditions in different municipalities within the purported class. Pretrial discovery may disclose that non-federal governmental purchasers differ widely in their modes of purchase, use differing competitive bidding arrangements, purchase different types and sizes of buses ('city,' 'intercity,' and smaller sizes not manufactured by GM) and have available different competitive means of mass transportation (e.g., small buses, subways, trains, electric trolleys). If this proof should indicate that there was not one nationwide market as alleged in the complaint but a series of individual dissimilar relevant local markets, ther would be little purpose in permitting the suit to proceed on behalf of a nationwide class of municipalities. Indeed such a suit might prove to be a chaotic affair, joining a series of individualized issues that could be resolved more efficiently in individual suits. Common questions of fact would not predominate. Evidence of GM monopolization in Atlanta or Albuquerque, which have no subways, would not be probative on the question of whether there has been monopolization in New York, which not only has subways but other competitive means of transportation, or in some smaller city which may be in a position to switch its purchases at will from GM to a competing manufacturer of smaller buses. Proof that GM may have rigged bid specifications to exclude competitiors in one city would be of questionable relevance in proving monopolization in another.

Under the foregoing circumstances proof adduced in an individual suit by the City of New York would be but a fraction of that presented in a suit on behalf of almost 200 municipalities. A class action would not then achieve any appreciable savings in time, expense or energy for the parties or for the court, particularly when it is doubtful that many other municipalities would think well enough of the claims to institute individual suits on their own behalf.

For these reasons I would prefer to consider our own holding, like that of the district court, to be tentative in nature and without prejudice to consideration upon a future appeal after the district court, upon renewal of the motion by GM on the basis of the additional facts which it proposes to show, makes a more informed determination on the issue of class action status.