Full opinion text
TJOFLAT, Circuit Judge: The appellate saga of the Corrugated Container antitrust litigation continues. In this chapter, we consider two district court orders approving settlements between plaintiff-class representatives and twenty-four of the thirty-seven defendants. The appellants in this appeal include two groups of dissident plaintiffs who argue that the settlements are invalid and must be set aside, and that even if they are valid, the formula by which they are to be distributed must be modified; a third group of plaintiffs who ask us to void those parts of the settlement that prevent them from pursuing state law remedies based on the same operative facts that underlie the settled federal claims; a fourth group of plaintiffs (comprised of one of the first two groups and the third group) who claim the notice informing the class of the settlement was defective; and various defendants who ask that we delay either approval of the settlements or distribution of the settlement proceeds until there is final judicial clarification of the rights of non-settling defendants to bring actions for contribution against the settling defendants. None of these arguments persuades us that the settlements must be set aside or modified; we are, however, in agreement with certain objectors that the district court’s findings are insufficiently detailed to allow us to determine whether the district court abused its discretion in concluding that the terms of the settlements, and the plan to distribute the settlement proceeds to class members, are fair, reasonable and adequate. We therefore remand the case to the district court for more detailed findings explaining its approval of the settlements. We retain jurisdiction during this limited remand. I The Facts We begin with a description of this litigation’s general history. In 1976, the government convened a grand jury to investigate possible criminal antitrust violations occurring in the corrugated container industry. Several purchasers of corrugated containers and corrugated sheets subsequently filed treble damages actions against thirty-seven defendants under section 4 of the Clayton Act, 15 U.S.C. § 15 (1976). These cases, most of which were class actions, were consolidated in the Southern District of Texas in the fall of 1977. On December 6, 1977, the District Court ordered the formation of a plaintiffs’ steering committee consisting of thirteen of the lawyers for plaintiffs in the consolidated actions. The district court order vested the committee with broad authority to control the continuing conduct of plaintiffs’ cases; this authority included the power to enter into settlement discussions with interested defendants. On January 25, 1978, the grand jury returned indictments against fourteen of the defendants. The indictments charged that these defendants (and several individuals who had been employed by the defendants) conspired to fix corrugated container and sheet prices between 1960 and 1974. In addition to the consolidated antitrust actions described, a group of container purchasers filed a state-law antitrust action in the Court of Common Pleas for Spartan-burg County, South Carolina. The defendants in this action were also defendants in the consolidated cases. The state-court plaintiffs in these actions resisted efforts to remove their case to federal court, where it would have been transferred to the Southern District of Texas for consolidation. After successfully resisting removal, the plaintiffs moved the South Carolina court to enjoin the defendants from entering a settlement with the federal plaintiffs that released the South Carolina claims. This mo: tion was granted. In March 1978, plaintiffs in forty of the pending consolidated cases filed a single complaint (in the suit now before us) on behalf of all purchasers of corrugated container and sheet products. Soon thereafter, a motion for certification of a class was filed. The motion had not been ruled upon when, on July 28, 1978, Great Northern Packaging Corporation (Great Northern) and Huron Packaging Corporation (Huron), purchasers of sheets, filed a separate antitrust class action in the United States District Court for the Southern District of Texas against the defendants named in the unified class action. The plaintiff class in this new suit, however, was definitionally limited to sheet plants, that is, sheet purchasers who fabricate containers from their purchases. According to Great Northern and Huron, container purchasers rather than sheet plants had filed, and were prosecuting, the earlier consolidated actions. Here we digress to consider the differences between the two purported classes. To do this, we must first look at the corrugated container industry. There are three steps in the manufacture of corrugated containers. In the first step, wood pulp is converted into containerboard. In the second step, the containerboard is confected into corrugated sheets. And, finally, in the third step, the sheets are fabricated into containers. The defendants in this appeal perform all three of these functions; that is, they begin with wood pulp and wind up with corrugated boxes. They are thus fully integrated operations. The defendants sell their finished corrugated boxes to purchasers ranging from “ma and pa” grocery stores to such multinational corporations as Xerox Corporation. This group of purchasers composes the class denominated as container purchasers. The defendants also sell corrugated sheets to independent enterprises that fabricate and sell their own containers. These corrugated sheet purchasers comprise the sheet plant class. It should be noted that many container purchasers purchase some sheets, and, similarly, many sheet plants will purchase finished containers. The’various actions that had been consolidated, with one exception, had been filed by container purchasers. Most of these actions, however, were class actions, and the class definitions were arguably broad enough to include sheet plants. In any event, the plaintiffs’ steering committee that was formed pursuant to the district court’s December 6, 1977, order was composed entirely of lawyers for container purchasers. During the summer of 1978, and prior to class certification, the steering committee entered into settlement discussions with St. Regis Paper Corporation, which had not been indicted in the criminal case. These discussions culminated in St. Regis agreeing to pay $1.7 million, which represented $428,-000 for each percentage point of the corrugated container market (market point) controlled by St. Regis, to settle the claims against it. St. Regis also made certain concessions and agreed, orally, to cooperate in plaintiffs’ discovery effort. Later that summer, plaintiffs concluded a second settlement, with International Paper, which had been indicted in the criminal case. International Paper agreed to pay $8.3 million, which reflected $1 million for each percentage point of the corrugated market controlled by the company. At the time the steering committee negotiated the aforementioned settlements, it had undertaken no discovery going to the merits of the case. (Such discovery was effectively precluded by restrictions imposed by the district court during the pend-ency of the. criminal prosecution.) The steering committee had available to it certain information, however, which, according to the settlement proponents, was adequate for it to make an informed judgment concerning the value of the claims. On September 6, 1978, after the first two settlements had been concluded, the district court entered an order certifying a single class in the consolidated litigation. Great Northern believed the parameters of the class were imprecisely drawn and, in particular, ambiguous concerning whether sheet plants were included in the class. In response to these concerns, Great Northern, on October 6, 1978, filed a motion to clarify the class definition by excluding sheet plants, or, alternatively, to create a subclass for sheet plants. The court took this motion under advisement. Then, on October 26, 1978, without ruling on Great Northern’s motion, the district court designated representatives of the class it had already certified. These class representatives included Atlas Container Corporation, which represented itself as a sheet plant. The class representatives immediately embarked on an exploration of further settlement possibilities. This led, in December, to settlements with seven additional unindicted defendants. These settlements ranged from $2 million per market point to $2.75 million. In addition, a misdemeanor indictee settled for $3.5 million per point, and a felony indictee for $4.5 million. On December 26, 1978, the district court responded to Great Northern’s motion by certifying two subclasses, one of container purchasers»and the other of sheet plants. The court named Great Northern as one representative of the sheet plant subclass; Atlas Container, which previously had been designated a class representative, was also named a sheet plant representative. With the exception of Atlas Container, the representatives of the container purchaser subclass were the plaintiffs who had been designated on October 26,1978 to represent the initial unitary class. Also on December 26, the two subclasses declared moratoriums on further negotiations. The container purchasers lifted their moratorium on January 5,1979, and throughout the month conducted settlement discussions, but strictly on behalf of their subclass. This created a problem for defendants, who desired settlements that would release them from the claims of both subclasses. This problem was compounded by the container subclass’s per-point dollar demands, which were in excess of what the single class had received in the December settlements. Defendants refused to negotiate on these terms, at least without the sheet plants’ participation in the discussions. The sheet plants, however, were unwilling to discuss settlement pending resolution of their investigation into damages suffered by their class. In order to break the resulting deadlock between the container class and defendants, a compromise was reached. The compromise was that a portion of any negotiated settlement would be available to the sheet class if the sheet class applied to participate in it, thus providing the defendants a possibility that the settlement they negotiated would release both subclasses’ claims. Eleven settlements, ranging from $1.5 million to $4 million per point, were negotiated on these terms. Late in January, the two classes negotiated additional settlements with two defendants who claimed an inability to pay large amounts in settlement because of borderline solvency. The plaintiff classes retained two accounting firms to investigate these hardship claims. After the firms reported that the claims were founded in fact, the plaintiffs settled at rates lower than those used in earlier settlements. The criminal trial immediately followed these settlements, and no further settlements were reached while the trial was in progress. In April, 1979, the jury returned a verdict of not guilty against the defendants who stood trial. (Several of the indicted defendants had previously pleaded nolo contendere.) On April 16, 1979, the plaintiff class representatives filed a motion for preliminary approval of the settlements. Several members of each of the two subclasses opposed the motion. Among the objections were that the settlements were contaminated by conflicts of interest; that the dollar amounts were inadequate; that even if the post-class certification settlements were adequate in amount, the pre-certification settlements with St. Regis and International Paper were clearly inadequate; that the negotiating attorneys lacked data necessary to evaluate the settlements; and that the objectors were denied the right to explore the settlement negotiations through discovery. By order dated May 30, 1979, the district court rejected these objections as obstacles to preliminary settlement approval, and held that “these settlements are within the range of possible approval and that notice of them should be given to the class members.” In Re Corrugated Container Antitrust Litigation, 1979-1 Trade Reg.Rep.(CCH) 162, 690 at 77,881 (S.D.Tex. 1979). The class representatives thereafter sent a single notice to the class members, advising them they were potential members of a class and that settlements with 24 of the defendants had been reached and preliminarily approved by the court. The notice indicated that the total value of the settlements was $300,000,000. Recipients of the notice were given options of participating, of opting out of the class, or of participating in the class but objecting to the settlements. The sheet plant and South Carolina objectors contended that the notice was defective because it omitted relevant information concerning the settlements and because separate notices should have been used to advise the class of the pendency of the actions and of the proposed settlements. In any event, the vast majority of potential class members chose to participate in the class, and only a very small percentage of the class members objected to the settlements. The court, on December 3 and 4, 1979, presided over a hearing on whether final approval should be given to the settlements. At this hearing, the objections to the settlements raised, at the preliminary approval stage were renewed. In addition, new objections were raised. Several defendants objected to approval of the settlements pri- or to final judicial resolution of the validity of the claims for contribution that certain non-settling defendants had raised against them; other defendants requested that the court approve the settlements, but delay distribution of the proceeds until judicial resolution of the contribution issue, thus preserving the assets until such time as a right of rescission might arise. The group of plaintiffs that had filed an antitrust action in South Carolina objected to a settlement provision releasing state law claims. Some sheet plants objected to the manner in which the proceeds were to be divided between the subclasses; some container purchasers made similar objections. The district court, nonetheless, approved the settlements and the objectors have brought this appeal. We discuss these various objections in the following sections of this opinion. Before beginning this discussion we add, parenthetically, that since the district court’s approval of these settlements, twelve of the then thirteen non-settling defendants settled their lawsuits. The plaintiffs proceeded against the thirteenth defendant, Mead Corporation, in a jury trial. The jury found that Mead had conspired to fix prices in the corrugated container industry with eighteen of the settling defendants. II Arguments that the District Court Abused Its Discretion in Approving These Settlements Rule 23(e) of the Rules of Civil Procedure requires that a settlement or compromise of a class action be approved by the district court. The district court below, in its 31st pretrial order, found the twenty-four settlements involved in this appeal to be fair, reasonable and adequate and, therefore, approved them. Two groups of plaintiffs, one comprised of certain members of the sheet subclass and the other of members of the container purchaser subclass, contend that the district court’s findings and its concomitant approval of the settlements were erroneous. We are, accordingly, asked to set aside the court’s approval. The objecting plaintiff groups, between them, advance three reasons why the district court should not have given these settlements its imprimatur. The first two reasons do not relate directly to the terms of the settlement, but rather to the manner in which those terms were negotiated by the class representatives. Specifically, the objecting plaintiffs contend that the settlements should be set aside because they were negotiated by attorneys who (1) attempted to represent both subclasses as a single client, a representation that, because of the inherent conflicts between the subclasses, could not adequately be undertaken; and (2) lacked sufficient data to evaluate the settlement value of their cases against each of the defendants. We hold that neither of these objections provides grounds for reversing the district court’s approval. The dissident plaintiffs' third objection is to the adequacy of the terms of each of these settlements. The plaintiffs argue that on the record before us, we must find that the district court abused its discretion in holding that the settlements were adequate. We are unable to evaluate this contention, however, because the district court’s findings and conclusions on adequacy were insufficiently detailed to inform us of why the court acted as it did. Although we are reluctant to delay further ultimate judgment on these settlements, we are, under these circumstances, compelled to remand to the district court for findings of fact sufficient for us to determine whether its approval of the settlements was a proper exercise of discretion. We will retain jurisdiction of this appeal during the limited remand. The above-summarized holdings are discussed below. A. Standard of Review Rule 23(e) provides no standard by which a court is to consider the settlement of a class action; rather, the rule states only that “[a] class action shall not be dismissed or compromised without the approval of the court .... ” Decisional law, however, provides us with a general measuring rod for considering settlements: In determining whether to approve a proposed settlement, the cardinal rule is that the District Court must find that the settlement is fair, adequate and reasonable. . .. Cotton v. Hinton, 559 F.2d 1326, 1330 (5th Cir. 1977). See also: Young v. Katz, 447 F.2d 431 (5th Cir. 1971). Approval of a settlement under this standard is not to be upset unless “the trial court clearly abused its discretion.” Young v. Katz, supra at 432. Thus, our appellate function respecting the various arguments presented on this appeal is a limited one, especially in light of the strong judicial policy favoring settlement of disputes. United States v. City of Miami, Florida, 614 F.2d 1322, 1344 (5th Cir. 1980). B. Conflict of Interest We turn to the dissidents’ first argument: that the district court should have disapproved these settlements because they were negotiated by attorneys with conflicts of interest that adversely affected their representation of one or both classes. As might be expected, the two classes have somewhat different views on which class was hurt and how the hurt was inflicted. The sheet plant dissidents argue that the pre-January 5, 1979, settlements, covering both classes, were negotiated by attorneys representing container purchasers only, and that those attorneys subordinated the interest of sheet plants to the interests of their clients. The container purchaser dissidents present a more subtle argument. They believe that the attempt to negotiate the pre-January 5 settlements on behalf of both subclasses, with their allegedly diverging and irreconcilable interests, made it impossible for negotiating counsel to “have represented the interests of both container purchasers and sheet plants adequately in settlement negotiations.” Brief of Plaintiff-Appellant Container Purchasers, at 23. Moreover, the dissident container purchasers also challenge the post-January 5 settlements, arguing that the container purchasers’ negotiation of a set-aside fund to be allocated among the sheet plants upon their application polluted those settlements with conflicts of interest. 1. Pre-January 5 Settlements We can assume, for purposes of considering the conflicts argument, that the two subclasses had significantly diverging interests and that an attorney could not adequately represent all these interests throughout the litigation. This does not mean that the settlements are necessarily void, however, because even “irregular settlement negotiations may ... form the basis for a judicially acceptable class action settlement ... if the record clearly indicates that representation of the class during the negotiations was adequate and that the settlement itself is fair.” In Re General Motors Corp. Engine Interchange Litigation, 594 F.2d 1106, 1131-32 (7th Cir.), cert. denied, 444 U.S. 870, 100 S.Ct. 146, 62 L.Ed.2d 95 (1979) (emphasis supplied). The question of fairness of the settlement terms is considered in a separate section of this opinion. We consider, in this section, the question of adequacy of representation. Recognizing that the allegedly differing interests of the two subclasses might have conflicted at some point in the litigation, we think the question before the district court was whether these interests conflicted at the point of settlement negotiation, and thus deprived either class of the vigorous and unqualified advocacy in settlement negotiations to which both were entitled. In deciding whether the settlements resulted from proper advocacy, we must inquire, first, whether the general interests of the subclasses respecting the settlements were the same and amenable to being achieved by unified representation; and, second, whether any specific features of the settlement sacrificed the interests of one class in favor of the interests of the other. The way we have framed our first inquiry indicates our agreement with the district court that “so long as all class members are united in asserting a common right, such as achieving the maximum possible recovery for the class, the class interests are not antagonistic for representation purposes.” In Re Corrugated Container Antitrust Litigation, 1980-1 Trade Reg. Rep. (CCH) If 63,163 at 77, 788 n.10 (S.D.Tex. 1979). We think it clear that the primary settlement goal of each class was to cause defendants to agree to pay substantial compensation in exchange for the most limited possible release from their obligations and potential liabilities as parties to the litigation. We have carefully reviewed the objectors’ appellate briefs, and their submissions to the district court in opposition to the settlements, and have not been pointed to anything suggesting that other significant subclasses’ settlement interests existed. Neither has there been a suggestion that two sets of negotiators leading to two sets of settlements might have better achieved the common aims of the two subclasses. On the contrary, in these circumstances, logic dictates that one set of negotiators, with the authority to release defendants from all claims, would be in a better bargaining position than negotiators with authority to compromise only part of the action. We next consider whether the negotiators, despite the uniformity of the two subclasses’ settlement interests, somehow sacrificed, advertently or inadvertently, the interests of one subclass for those of the other. While objectors from each class argue that the dollar values of the settlements were inadequate, neither class suggests that this inadequacy resulted from a tradeoff of its rights in order to realize the interests of the other class. In fact, the only suggested illustration of potential class conflict evolving into actual prejudice concerns the settlement with St. Regis. The sheet-class objectors contend that this settlement should have been larger because St. Regis, while a small container manufacturer, is the nation’s third largest manufacturer of corrugated sheets. To understand the problem the sheet plant objectors perceive, we must turn to the negotiating history of the settlements. The attorneys who negotiated the settlements believed a favorable psychological climate would be created by offering discounts to early settlors. The earliest settlor was St. Regis, which settled for $428,-000 per market point, the market points representing combined sheet and container sales. This was at less than one tenth the rate of the last of the settlements being considered in this appeal. The question here, then, is whether the negotiation of an early and substantially discounted settlement with a defendant whose market share is significant with respect only to sheets is prejudicial to the sheet subclass. On reflection, we think that it is not. In the first place, the settlement with St. Reg-is was only the first settlement in the steering committee’s strategy of demanding successively higher payments for each settlement. There is no suggestion that the steering committee consciously embarked on a course of negotiating discounted settlements with defendants only if their market shares were disproportionately large with respect to sheets. Neither is there any suggestion that the steering committee attempted to negotiate early, heavily discounted settlements for both subclasses, but confined the benefits of later, more valuable settlements to the container class. Since the sheet subclass is to share in both early and later settlements on the same basis — percentage of all sheet sales to all corrugated sales by all defendants — we do not see how this was inherently unfair to the sheet subclass. Moreover, at the time these settlements were negotiated, members of both subclasses, objectors and proponents alike, believed that all defendants were jointly and severally liable, see e. g., Beltz Travel Service, Inc. v. International Air Transport Association, 620 F.2d 1360, 1367 (9th Cir. 1980); State of Washington v. American Pipe & Construction Co., 280 F.Supp. 802, 804 (S.D. Cal.), cert. denied, 393 U.S. 842, 89 S.Ct. 122, 21 L.Ed.2d 113 (1968), and that nonsettling defendants, if ultimately found to be liable at trial, would be responsible for the total galaxy of damages caused by all participants in the conspiracy. The only expected reduction in damages would have been for the actual amounts paid in settlement. See Brief of Class Plaintiffs-Appellees, at 21. This being the expectation, it is difficult to understand how the St. Regis settlement adversely affected the interests of the sheet subclass, since even a defendant with a small sheet market share would be liable for all damages. But even assuming a theory of liability under which non-settling defendants were liable only for their own market shares, the sheet plant subclass would be harmed vis-a-vis the container subclass only if the settling defendants’ collective share of the sheet market was larger than their collective share of the container market. By isolating the St. Reg-is settlement for attack, the dissident sheet plaintiffs seek to ignore this fact. In short, we hold for the reasons stated that conflicts of interest have not infected the settlement proceedings leading to the pre-January 5 settlements. 2. Post-January 5 Settlements On December 26, 1978, the district court certified the sheet plant subclass. On the same date, and in apparent reaction to the subclass certification, both subclasses declared a moratorium on further negotiations. The container class resumed settlement talks on January 5, 1979; the sheet subclass did not, apparently deciding to seek further economic data first. This divergence in strategies produced a problem for the container class, since defendants were reluctant to settle at substantial dollar amounts for a release covering only one of the subclasses. To induce defendants to consider settlement on this basis, the container purchasers developed a novel approach. A portion of each settlement would be set aside for the sheet plant subclass, but only in the event the sheet plants applied to participate in these particular settlements. If the sheet plants decided not to participate, the entire amount paid by each settling defendant would inure to the exclusive benefit of the container purchasers. Moreover, the amount of the set-aside fund was to be a cap on the sheet plant subclass recovery; the sheet plants were to negotiate the actual dollar amount with the representatives of the container purchasers. In all, eleven settlements were negotiated on this basis. The amount paid by each of these settling defendants was in excess of that paid by defendants who had earlier settled. The container-purchaser subclass contends that the set-aside fund strategem engaged the plaintiff representatives in conflicts that destroyed their ability to negotiate fairly. The sheet plant objectors’ argument concerning these settlements is the subject of another appeal; on this appeal, only the container purchaser subclass arguments are before us. The container class contends that the set-aside fund “is a unilateral unbargained for giveaway of monies properly belonging to container purchasers, and [therefore] further evidence of a conflict of interests and lack of adequate representation in settlement negotiations.” See Brief of Plaintiff-Appellant Container Purchasers at 30. The logic of this position eludes us. By the time of these settlements, the container purchasers were grouped into a separate subclass and the negotiators after January 5 were negotiating for the container purchaser subclass only. When faced with the unwillingness of defendants to pay large sums of money to obtain a release from only one of the subclasses, the container purchasers agreed to earmark part of each settlement for possible allocation to the sheet plants. The position that defendants would have been willing to pay to the container purchasers an amount in excess of the total settlement less the set-aside fund, with no hope of binding the sheet plant subclass, is untenable; it is bereft of logical or factual support. While it is true that the container-purchaser representatives could have waited until the sheet-plant subclass was ready to participate in negotiations, this would have meant delaying negotiations until after the criminal trial was underway or, more likely, until it had concluded. The container purchasers believed that the eve of trial was an opportune time to bargain, and the only way to bring defendants to the bargaining table appeared to be to offer to create the set-aside fund. Since the container-purchaser objectors do not challenge the wisdom of bargaining on the eve of trial, we see no argument that the decision to negotiate the set-aside fund damaged the container purchasers. We point out that in finding that the container-purchasers’ negotiators adequately represented the container-purchaser interests, we do not pass judgment on the sheet-plant dissidents’ objections to the post-January 5 settlements. C. Lack of Discovery The dissident sheet plant plaintiffs next contend that the settlements should be disapproved because they were negotiated prior to probing discovery. Plaintiffs’ theory appears to be that representation in the settlement process is necessarily inadequate unless informed by the process of discovery. It is, in effect, argued that without discovery, the class representatives were not in a position of equality with negotiators for the defendants. From this we are asked to conclude that settlements resulting from this putative inequality of knowledge must be, as a matter of law, inadequate. We disagree. Initially, we note that notwithstanding the status of discovery, plaintiffs’ negotiators had access to a plethora of information regarding the facts of their case. In Cotton v. Hinton, 559 F.2d 1326, 1332 (5th Cir. 1977), a case in which settlement was reached despite only slight formal discovery, we said the following: It is true that very little formal discovery was conducted and that there is no voluminous record in the case. However, the lack of such does not compel the conclusion that insufficient discovery was conducted. At the outset, we consider this an appropriate occasion to express our concern over the common belief held by many litigators that a great amount of formal discovery must be conducted in every case. Thus, we are not compelled to hold that formal discovery was a necessary ticket to the bargaining table. Because the plaintiffs did have access to information, this case cannot be characterized as an instance of the unscrupulous leading the blind. Even assuming there was an imbalance of information between the defendants and the plaintiffs at the bargaining table, this would not in itself invalidate the settlements. We think in a case such as this, the trial court may legitimately presume that counsel’s judgment “that they had achieved the desired quantum of information necessary to achieve a settlement,” id. at 1332, is reliable. Of course, if the record points unmistakably toward the conclusion that the settlement was the product of uneducated guesswork, a court may be acting within its discretion in disapproving the agreement without ever considering whether the agreement’s terms are adequate. But since counsel had access to data, that is not our case. In general, we think a settlement should stand or fall on the adequacy of its terms. In a very real sense, a review of the terms provides a check on counsel’s evaluation of the sufficiency of his working knowledge. If the terms are fair, the court may reasonably conclude that counsel did perform adequately. We stress that this is not to say that a district court may approve a settlement whose adequacy is insufficiently documented at the settlement approval hearing. In fact, where it appears that counsel was relying on judgment bolstered by only a small amount of information, the district court should be especially thorough in its review of the fairness and adequacy of the settlement terms. D. Adequacy of the Settlement Terms The first two objections to the settlements were attacks on the adequacy of the lawyers who negotiated the settlement terms. As a matter of theory, these attacks presupposed that a settlement negotiated by inadequate representatives should not be approved because adequate representatives could have negotiated more favorable terms. The objectors’ third contention, however, is that the actual terms of all but the “hardship” settlements are so inadequate in dollar amount that the district court abused its discretion in approving them. Having disposed of the objectors’ first two arguments, we turn to a discussion of whether the settlement terms are fair, adequate and reasonable. As we have suggested earlier in this opinion, the district court’s most important function in reviewing compromises of class actions is its consideration of the settlement terms. See Cotton v. Hinton, 559 F.2d at 1330. It is, ultimately, in the settlement terms that the class representatives’ judgment and the adequacy of their representation is either vindicated or found wanting. If the terms themselves are fair, reasonable and adequate, the district court may fairly assume that they were negotiated by competent and adequate counsel; in such cases, whether another team of negotiators might have accomplished a better settlement is a matter equally comprised of conjecture and irrelevance. But all this, of course, begs the real question, which is how to determine whether the settlement terms are in fact adequate. Case law provides us with general ground rules: “The settlement terms should be compared with the likely rewards the class would have received following a successful trial.” Cotton v. Hinton, 559 F.2d at 1330. And: “the strength of the case for plaintiffs [must be] balanced against the amount offered in settlement.” West Virginia v. Chas. Pfizer & Co., 440 F.2d 1079, 1085 (2d Cir.) (quoting State of West Virginia v. Chas. Pfizer & Co., 314 F.Supp. 710, 740 (S.D.N.Y.1970), cert. denied, 404 U.S. 871, 92 S.Ct. 81, 30 L.Ed. 115 (1971). See also City of Detroit v. Grinnell Corp., 495 F.2d 448, 455 (2d Cir. 1974). We think this requires a three-step process. First, the district court must evaluate the likelihood that plaintiffs would prevail at trial. Second, the district court must establish a range of possible recovery that plaintiffs would realize if they prevailed at trial. And third, guided by its findings on plaintiffs’ likelihood of prevailing on the merits and such other factors as may be relevant, the district court must establish, in effect, the point on, or if appropriate, below, the range of possible recovery at which a settlement is fair and adequate. We note that this type of evaluation is not and cannot involve a trial on the merits. “[T]he very uncertainty of outcome in litigation, as well as the avoidance of wasteful litigation and expense, lay behind the Congressional infusion of a power to compromise. This is a recognition of the policy of the law generally to encourage settlements. This could hardly be achieved if the test on hearing for approval meant establishing success or failure to a certainty.” Florida Trailer & Equipment Co. v. Deal, 284 F.2d 567, 571 (5th Cir. 1960). Nor could it be achieved if the range of possible recovery had to be charted with precision. But the district court judge must “undertake an analysis of the facts and the law relevant to the proposed compromise,” and he must “support his conclusions by memorandum.” Cotton v. Hinton, 559 F.2d at 1330. “A ‘mere boiler-plate approval phrased in appropriate language but unsupported by evaluation of the facts or analysis of the law’ will not suffice.” Id. (quoting Protective Committee v. Anderson, 390 U.S. 414, 434, 88 S.Ct. 1157, 1168, 20 L.Ed.2d 1 (1968). This is because “[a]n appellate court ... must have a basis for judging the exercise of the trial judge’s discretion.” Id. Thus, we begin our analysis by reviewing the district court’s opinion to determine whether its approval of the settlements was based on adequate and careful analysis of “the facts of the case in relation to the relevant principles of applicable law...” Id. at 1331. Because we conclude that it was not, we must also determine whether, on the record before it, the district court, in the exercise of its discretion, was compelled to either approve or disapprove the settlements. After undertaking this review, we conclude that we must remand to the district court to prepare new findings; if necessary, to take new evidence; and if compelled to do so, to reach new conclusions. We turn to the specific reasons for our decision. As we have already indicated, the district court must establish the range of possible damages that could be recovered at trial, and, then, by evaluating the likelihood of prevailing at trial and other relevant factors, determine whether the settlement is pegged at a point in the range that is fair to the plaintiff settlors. In a case such as this, where there are objectors, the court is aided in its task; the proponents can be expected to present evidence and arguments suggesting that the settlements are within “a range of reasonableness” and the objectors will do the same for the contrary position. By weighing the competing evidence and evaluating the legal arguments, we think the court should be able to reach a just conclusion. It is for this reason that the court can generally fulfill its responsibilities by “examinpng] the settlement[s] in light of the objections raised and [by] set[ting] forth on the record a reasoned response to the objections including findings of fact and conclusions of law necessary to support the response.” Cotton v. Hinton, 559 F.2d at 1331. We will consider, then, each of the objectors’ attacks on the settlement, and the court’s resolution of them. Our initial discussion is broken down into three parts: one on the range of possible recovery, one on the likelihood of prevailing on the merits, and one on the court’s exercise of discretion. We then proceed to explain why we think remand is necessary and what we think the court’s task on remand is. 1. Damages The plaintiff proponents of the settlement presented the affidavit of an economist, Richard Hoyt, that estimated that damages to the plaintiff class were between two hundred and eight hundred million dollars for the years 1972-1976, the four-year period not barred by the statute of limitations. It seems clear that the court regarded this range of damages as the benchmark of possible recovery at trial. At the settlement hearing, and on this appeal, the objectors attacked Hoyt’s conclusions on two grounds. First, the sheet plant objectors contended that the period he selected to estimate damages should have been enlarged to reflect the entire period of overcharge since plaintiffs might have been able to establish that the defendants engaged in a pattern of fraudulent concealment, which would have tolled the statute of limitations. See In Re Beef Industry Antitrust Litigation, 600 F.2d 1148, 1169 (5th Cir. 1979), cert. denied, — U.S.-, 101 S.Ct. 280, 66 L.Ed.2d 137 (1980); Westinghouse Electric Corp. v. Burlington, 351 F.2d 762, 764 (D.C. Cir. 1965); Rinzler v. Westinghouse Electric Corp., 333 F.2d 719 (5th Cir. 1964). Second, objectors from both the sheet-plant and container-purchaser subclasses, through affidavits of their economists, argued that Hoyt’s methodology was defective and resulted in an estimate of damages that was far too low. The objectors urged the district court to accept their own estimates, which were substantially greater than Hoyt’s, as providing an accurate range of possible recovery. We consider first the argument that the court should have considered damages suffered in years prior to 1973. The proponents of the settlements argue that some courts have looked only to years within the statute of limitations to determine whether a settlement is reasonable. See, e. g., Detroit v. Grinneil Corp., 495 F.2d 448, 460 (2d Cir. 1974); In re Anthracite Coal Antitrust Litigation, 79 F.R.D. 707, 714 (M.D.Pa. 1978) , modified sub nom. Colonial Fuel Co. v. Blue Coal Corp., 612 F.2d 571 (3d Cir. 1979) . The objectors, however, counter that these courts were not confronted with strong arguments that the statute of limitations was tolled by defendants’ fraudulent conduct. They argue that in the case before us, evidence of fraudulent concealment does exist, which means that damages should be measured from 1960, the beginning of the conspiracy. The sheet-plant objectors are correct that the law does not limit the court to the four-year period within the statute of limitations in setting the range of possible damages recoverable after trial. The Grinneil case, relied upon by the proponents, explains that consideration of possibly time-barred years was inappropriate there because fraudulent concealment would not, on the facts alleged, extend the period for claims. Id., 495 F.2d at 461. Further, we have not been referred to any case suggesting that damages should, for settlement discussions, be confined to the statutory period in the face of a strong fraudulent concealment argument. We think that when objectors seek to include years outside the statute of limitations in the damage computation period, the court should evaluate the fraudulent concealment argument. It should also consider the extra problems of proving a price-fixing conspiracy for the longer period and the additional difficulties of proof in assessing damages. If the court finds the chances of substantial recovery for the added years to be insignificant, it may confine its consideration of damages to the non-time-barred years. If the district court concludes that the plaintiffs have a significant chance to recover substantial amounts for years outside the statute of limitations, it should refuse to approve a settlement until it is given some estimate of damages for that period, or at least an acceptable explanation why the estimate for the period within the statute of limitations is a fair measure of damages for both periods. Here the district court failed to explain why it only considered damages for the years within the statute of limitations. Its decision on this matter is an essential element of its overall decision to approve these settlements, but since the court has not provided us the benefit of its thinking, we are unable to review it. The objectors’ second concern is with the methodology Hoyt employed in estimating damages. The objectors argue that the best means of estimating damages in a price-fixing conspiracy is a before-and-after price comparison. Hoyt rejected this method, however, for several reasons, among them (1) that he would have to rely on industry-wide prices, which he perceived as undesirable because the conspirators comprised only part of the industry; (2) he believed that price information for the period before 1973 would be unavailable. In the place of the before-and-after price comparison, Hoyt believed a methodology based on excess profits. Using this approach, he computed the percentage of defendants’ sales receipts that constituted profit for the conspiracy period (1972-1976) and a non-conspiracy period (1977-1978) and then compared the two figures. Hoyt believed that the amount by which the profits for the conspiracy years exceeded the profits for non-conspiracy years constituted an excess profit percentage. This figure, when applied to total sales, closely approximated the actual damages suffered. The objectors believe this methodology is defective for several reasons, primarily the following two: (1) profits are affected by external facts that vary in impact from year to year; (2) costs during the conspiracy period would be higher since industries engaged in price-fixing are less cost-efficient than they would be during a non-conspiratorial period. The dissident plaintiffs also object to the years compared in Hoyt’s model, claiming that the conspiratorial period — 1973-1976 — was a recessionary period in which profits were low, while the nonconspiratory period — 1977-1978 — was a period of economic expansion, in which profits were high. They also object that Hoyt’s profit figures were derived from industry sources, without any significant check on their accuracy. The sheet objectors claim that Hoyt made no attempt to determine if their damages should be measured separately from those of the container purchasers. Finally, both sets of objectors take issue with proponents’ assertion that the task of preparing a more traditional model based on prices rather than profits could not have been successfully undertaken. The district court does not discuss these criticisms of the Hoyt study, nor does it explain why it found the Hoyt conclusions more accurate than those of the objectors’ experts, who pegged damages at substantially higher levels. In fact, the court’s only comment on the economic validity of the Hoyt analysis suggests that it perceived no challenge to the validity, or at least the utility, of the analysis. Although there have been some assertions that damages are greater than proponents’ expert estimates, no one has asserted that more money could have been or could now be obtained from the defendants .... In re Corrugated Container Litigation, [1980-1] Trade Reg.Rep. (CCH) 1163,163 at 77,788 n. 10. Objectors’ challenges to the Hoyt analysis are implicitly tied to their belief that they could have negotiated better settlement terms. Even if the court is correct, however, that more money could not have been obtained, there is no reason to approve an otherwise inadequate settlement solely because it was the best offer defendants were willing to make. Given the importance of the question of potential damages, the district court was compelled to consider the validity of the estimate on which it relied. If the court did find the estimate to be valid, it was bound to explain why. Thus, we must remand on this issue as well. Finally, we note that the district court, in passing on these settlements, considered them, for the most part, as a single settlement rather than as twenty-four separate agreements. The propriety of doing this is considered in the subsection on the court’s exercise of discretion, but two observations concerning damages are appropriate here. First, because each defendant’s liability, if any, would be joint and several, his potential liability is the total amount of damages, less any amounts already paid to the plaintiff class. This means, theoretically, that the plaintiffs could obtain the same amount from each defendant. Accordingly, a single computation of a range of possible recovery suffices for each defendant. The second observation, however, is that in determining the amount of damages actually, rather than theoretically, recoverable from a particular defendant, the question of that defendant’s financial condition may become relevant. Except in the case of the hardship settlements, the district court has not made findings about the extent to which these defendants could afford to pay more than they did. We consider this point, too, in the subsection on the court’s exercise of discretion. 2. Risks of Litigation The objectors claim that the court incorrectly evaluated the strength of their case on the merits, overemphasizing the risks of litigation. The district court stated its findings and conclusions on the risks of litigation in its preliminary approval of the settlement: There are many risks for plaintiffs in this litigation; the recent acquittal of all who went to trial in the cases brought by the United States serve [sic] to emphasize some of the problems, and there are others that apply peculiarly to those civil cases, such as class certification problems and the proposed Illinois Brick legislation. Plaintiffs’ ultimate success is by no means assured. In re Corrugated Container Antitrust Litigation, [1979-1] Trade Reg.Rep. (CCH) 162,690 at 77,884. The objectors contest the court’s evaluation of litigation risks for three reasons. First, they claim that the court should not have relied on the criminal acquittals, since the standard of proof in the criminal trials was greater than it would be in the civil litigation. Second, they argue that the court’s findings were insufficiently detailed, and conclusory, amounting to mere boilerplate. Third, they claim that the court failed to explain its conclusion that the risks of litigation justified a settlement at the $300-million mark on Hoyt’s $200-to-800 million range of possibilities. Turning to the first objection, the dissident plaintiffs argue that “a criminal acquittal has absolutely no tendency to make any contested fact in a subsequent civil proceeding more or less probable,” and that “giving ‘considerable weight’ to the criminal acquittals ... resulted in a distortion of the settlement equation.” Brief for Plaintiff-Appellant Container Purchasers at 58-59. If the objectors are contending that the district court should not have considered the acquittals at all, they are in conflict with the Manual For Complex Litigation, section 1.46, which states that the court should consider success or lack of success in a prior criminal prosecution. 1 Moore’s Federal Practice § 1.46 n. 100 (2d ed. 1980). The Manual’s position is sensible. A court may more safely predict that a civil litigant will prevail on the merits if there has been a successful criminal prosecution in a related matter than it could had there been no criminal trial, or, certainly, had there been an acquittal. This being so, a case in which defendants were acquitted has less settlement value than one in which convictions were obtained. Here the court drew from the acquittal only a single inference: “ultimate success is by no means assured.” In re Corrugated Container Antitrust Litigation, [1979-1] Trade Reg.Rep. (CCH) II 62,690 at 77,884. This was entirely proper. The objectors’ second contention is that the court failed to evaluate the strength of plaintiffs’ case on the merits. This simply is not true; as the quotation from the court’s preliminary approval suggests, the court noted that the risks of class decertification, of Illinois Brick legislation, and the problems of proof suggested by the criminal acquittals combined to deprive the class of guaranteed ultimate success. In noting this, the district court said enough, though it certainly could and probably should have said considerably more. We hope that on the limited remand we order today, the district court will explain more completely how it evaluated the strength of plaintiffs’ case on the merits. Particularly, the district court may wish to cite those portions of the record below, and, if applicable, the record of the related criminal proceedings, on which it relied. One final point must be made. The district court concluded that plaintiffs’ ultimate victory on the merits is not assured. This conclusion, however, applies to the plaintiffs’ case against all the defendants collectively rather than individually. In all probability, the plaintiffs’ case is stronger against some defendants than others. Because the district court treated the settlements in the aggregate, it had no occasion to consider the strength of plaintiffs’ case against individual defendants. As we explain in the next section, it may have been necessary for the court to consider, to some extent, the strength of plaintiffs’ cases against individual settling defendants. 3. Exercise of Discretion We have outlined a three-step process for a district court to use in determining whether it should approve a settlement’s terms as fair, reasonable and adequate. In the first two steps, which we have described above, the court must decide two key issues: the range of possible recovery if plaintiffs prevail on the merits and the likelihood that plaintiffs will, in fact, prevail. In the third step of the process, the court, in its discretion, must decide if the settlements are reasonable in light of these determinations. In doing this, the court is not confined to the mechanistic process of comparing the settlement to the estimated recovery times a multiplier derived from the likelihood of prevailing on the merits. The court should also be guided by other factors, the relevancy of which will vary from case to case. Some of the factors that may be relevant in a particular case are “the complexity, expense and likely duration of the litigation ...; the reaction of the class to the settlement; [and] the stage of the proceedings .... ” City of Detroit v. Grinnell Corp., 495 F.2d at 463 (2d Cir. 1974). After the court identifies these factors and explains their relevance to the settlement, it should proceed to explain why it is either approving or disapproving the settlement. Here, the district court, in addition to its findings concerning damages and the likelihood of success, identified two factors that it felt bore on the reasonableness of the settlements. The first factor was that the trial would be long, complex and expensive. The court, however, offered little to indicate how this affected the reasonableness of the settlement. While the prospect of a long, complex and expensive trial militates in favor of settlement, virtually all class actions will result in long, complex and expensive trials. The question is whether the likelihood of an especially long and complex trial is enough in a particular case to warrant a substantial reduction in what the class might otherwise receive in settlement. The district court, however, said nothing to indicate why the prospect of a long trial had a particular impact on this case. The district court also found that the number of objectors to the settlement and the number of opt-outs was insignificant; from this it concluded that there was great support for the settlements, which is certainly a factor favoring approval. But further explanation was required, for a low level of vociferous objection is not necessarily synonymous with jubilant support. In many class actions, the vast majority of class members lack the resources either to object to the settlement or to opt out of the class and litigate their individual cases. The Corrugated Container Litigation, however, includes many large corporations who were sufficiently endowed to opt out or object to the settlements if they chose to do so. If the numerically small group of objectors was comprised largely of such corporations, perhaps the court’s finding of support was improperly predicated. Based on what the court has told us, however, we cannot say. Thus, we must conclude that the court’s findings on factors other than possible recovery and the likelihood of success were not detailed enough for us to review. In its conclusion, father than explain why it approved the settlements, the court only recited a list of the findings it had made concerning the settlements. While an overall list of findings may be sufficient for review when a court has made detailed and careful explanations as it conducted its consideration of each factor affecting its exercise of discretion, this was not the case here. One last point must be made. As we indicated earlier, the district court appears to have approved these settlements in the aggregate. Ordinarily, an individually negotiated settlement with one of many defendants in a case should be approved or disapproved as an individual settlement and not as part of a package. The package of settlements the district court approved includes early settlements negotiated at a fraction of the value of later settlements. Settlements early and late may both be reasonable, particularly if the district court finds, as it did, that early settlements were needed to induce other defendants to the bargaining table. But this does not excuse a settlement at an inadequate amount. The district court should have made some findings concerning why each individual settlement was reasonable, or at least explained why it was unnecessary to do so. 4. Task on Remand Because the district judge has failed to support his approval of the settlements by adequate evaluation of the facts and analysis of the law, we lack a basis for reviewing his discretion. Cotton v. Hinton, 559 F.2d at 1330. Thus, were we at this juncture to affirm the approval of the settlements, we would not be reviewing the district court’s exercise of discretion but, rather, exercising our own discretion on the basis of the record before us. This is not our function, and we therefore must remand to the district court. The remand is a limited one to allow the court better to explain why it approved the settlements. If the district court, after reviewing this opinion, believes its previous approval of the settlements cannot be justified under the applicable legal principles, it may petition this court to vacate its order approving the settlements and to remand the case for further proceedings on the settlements, and, if necessary, the merits of plaintiffs’ claims. Ill Allocation of the Settlements The sheet-plant and container-purchaser objectors contend that even if the settlements are fair, reasonable and adequate, the formula by which the pre-January 5, 1979, and the “hardship” settlements are to be distributed is not. The formula is a simple one: each class member will file a claim based on its purchases between 1960 and 1978 and “all allowed claims will be totalled and the sum divided into the total [settlement] fund (after deduction of fees and costs), and the quotient will be multiplied by each claimant’s total allowed claim to determine his recovery.” In re Corrugated Container Litigation, [1980-1] Trade Reg.Rep. (CCH) 163,163 at 77,791. There are, before us, two specific objections to this formula. First, the sheet-plant and container purchasers argue that neither the court nor the subclass representatives had any economic data to justify equal, dollar-for-dollar treatment of container purchases and corrugated sheet purchases. Second, the sheet-plant objectors contend that claims based on purchases before 1973 — the cut-off point under the statute of limitations — should not have been treated on par with claims based on post-1972 purchases.