DONALD RUSSELL, Circuit Judge:

This diversity suit by seven individual claimants, suing on their own behalf and as the proposed class representatives of all injured Dalkon Shield claimants, seeks recovery against Aetna Casualty and Surety Company (Aetna) for injuries resulting from the use of an allegedly defective intrauterine device known as the Dalkon Shield. Aetna was neither the manufacturer nor the vendor of the device; it was the products liability insurance carrier of A.H. Robins Company, Inc. (Robins), the manufacturer and distributor of the device. It is the theory of the plaintiffs that Aetna's conduct, while acting in its role as insurance carrier, was such that it rendered itself liable as a joint tortfeasor with Robins for any injuries sustained by persons using the device. The plaintiffs sought class certification of the suit. During consideration whether to give final certification of the suit, the parties entered into a settlement of the action conditioned on certification. After a duly-noticed hearing, the District Court granted in separate orders final class certification of the action and approval of the settlement of the action so certified. The appeal challenges the two orders. Since the two orders from which the appeals are taken are intimately connected, the two appeals have been consolidated. We affirm both the class certification and the settlement orders.

I.

In the early 1960's, Dr. Hugh Davis, a gynecologist on the staff of the Johns Hopkins Hospital and the Johns Hopkins Medical School, developed an intrauterine device which has been generally described as a Dalkon Shield. In conjunction with some associates, Dr. Davis, in the mid 1960's, began to manufacture and sell on a small scale the Dalkon Shield. This limited operation continued until early 1970 when A.H. Robins Co., Inc. (Robins), which was a well-recognized and successful manufacturer and marketer of pharmaceutical products, acquired the exclusive right to manufacture and market the Dalkon Shield. In his findings in connection with this purchase by Robins of exclusive rights in the manufacture and sale of the Dalkon Shield, the District Judge said:

Prior to that time ( i.e., the purchase of the Dalkon Shield), Robins had never marketed any kind of a birth control product and had no gynecologist on its medical staff. Nevertheless, Robins did no further testing, readily accepted the figures of the inventor's testing, and aggressively promoted the device to the medical profession and, uniquely for such a device, to the general public.

Robins, however, did not begin the marketing of the device until early 1971. Whether as a result of the promotional efforts of Robins or not, the product received wide public acceptance, and its sale was brisk. Between 1971 and mid-1974 when sales were discontinued, some 2,200,000 Dalkon Shields were sold. These sales generated a gross revenue for Robins of $11,240,611.00 and a gross profit of $505,499.00.

The product, however, had hardly been introduced into the market before what the District Court herein described as "an extraordinary volume" of complaints of injuries suffered by reason of its use began to surface. By 1973, the complaints had reached such a level that Robins felt compelled to send its first "Dear Doctor" letter to the medical profession advising the profession of a septic abortion problem arising out of the use of the device along with a suggestion for reducing the risk of such problem. The complaints from users of the device, however, continued at such a pace that by mid-1974 Robins finally withdrew the product from the market. Robins did not, though, issue any product recall until September 1980, when it addressed its second "Dear Doctor" letter to the medical profession. In this letter it recommended that the devices be removed from all continuing users, even those with no manifest problems, and it noted, as summarized by the District Court, that "long-term users of all IUD's had a higher risk of exposure to actinomyces, a virulent strain of microorganism implicated as a cause of pelvic inflammatory disease." However, there was no realistic recall of the Dalkon Shield until 1984.

The early complaints of injury from the use of the device quickly generated lawsuits against Robins in various parts of the country. Aetna Casualty and Surety Company (Aetna), as the products liability insurance carrier, was obligated to defend the suits filed against Robins in that connection. The first action to come to trial against Robins charging injuries from the use of the Dalkon Shield arose in the state court in Kansas and resulted in a verdict in February 1975 in favor of the plaintiff in the amount of $85,000, including a $75,000 punitive damage award. This verdict was widely publicized and the filing of actions against Robins quickened. This multiplication of suits — such, for instance, as approximately two hundred at the same time in the Northern District of California and in the District of Maryland, — posed a serious problem of manageability for the courts before which the suits were pending.

Many of the courts referred the multiplicity of suits in their district for consolidated pre-trial proceedings under the direction of the Judicial Panel on Multi-District Litigation in the hope that such action might reduce the difficulties of disposing expeditiously of the mounting Dalkon Shield case burden. See In re A.H. Robins Co., Inc. "Dalkon Shield" IUD Products Liability Litigation, 406 F. Supp. 540 (J.P. M.D.L. 1975); 419 F. Supp. 710 (J.P.M.D.L. 1976); and 438 F. Supp. 942 (J.P.M.D.L. 1977). After the entry of a number of such orders of transfer, the Judicial Panel finally concluded that their previous orders accepting transfers had accomplished as much as could be achieved in reducing pretrial proceedings and it began vacating later transfer orders, returning the cases for further proceedings to the respective transferor courts. In re A.H. Robins Co., Inc. "Dalkon Shield" IUD Products Liability Litigation, 453 F. Supp. 108 (J.P.M.D.L. 1978); and 505 F. Supp. 221 (J.P.M.D.L. 1981). While these proceedings before the Judicial Panel had aided the problems of discovery in the Dalkon Shield cases, they did nothing to relieve the clogging of court calendars by the constantly increasing stream of Dalkon Shield cases to be tried, nor did they reduce substantially the trial time of the cases. It was manifest that other measures were required if this overloading of the courts with Dalkon Shield cases was to be relieved.

District Judge Miles Lord of the Minnesota District Court, to whom had been assigned in the late 1970's and early 1980's a large number of Dalkon Shield cases, determined to try his hand at expediting the disposition of such cases as were before him. He consolidated a large number of such suits and appointed a lead counsel to handle discovery. He prodded the counsel so appointed to proceed aggressively, and he named two masters to sift through all the material discovered and to prepare a report for counsel and the court based on all material discovered and any developed in their independent investigations. The masters did conduct considerable independent investigations. Thus, they reviewed carefully the relevant records in the files of Robins and inquired into the possible involvement of Aetna with Robins in the defense planning and activity. All of this discovery, along with that developed by lead counsel, was catalogued by lead counsel appointed by the Court. The suits in litigation before Judge Lord, however, were at this point settled for $38 million and no further proceedings were had in Judge Lord's court (and that of his successor). The lead counsel in this litigation before Judge Lord withdrew from any subsequent Dalkon Shield litigation but made available to counsel in other cases the discovery developed in that case.

II.

Other courts took another tack in an effort to manage expeditiously and fairly the mass of Dalkon Shield cases. They turned to the class action procedure provided by Rule 23, Fed.R.Civ.P. The great volume of cases which were inundating the court system and the similarity of the issues in all the cases, it was thought, provided a proper basis for class treatment both in the interests of the parties and of the courts. The first case in which this procedure was invoked was Rosenfeld v. A.H. Robins Co., Inc., 63 A.D.2d 11, 407 N.Y.S.2d 196 (N.Y.App.Div.), appeal dismissed, 46 N.Y.2d 731, 413 N.Y.S.2d 374, 385 N.E.2d 1301 (1978). In that case, the plaintiffs sought class certification under a New York statute modeled after Rule 23, and Robins, oddly enough, opposed such certification. The motion court denied certification and, on appeal, that decision was affirmed. The court, in affirming that decision, stated that:

In view of the complicated issues of fact which must be resolved on an individual basis, it is our opinion that common questions of law and fact do not predominate in this action and that the Special Terms correctly denied plaintiffs' motion for class certification.

407 N.Y.S. at 201.

The court added as the rationale for this conclusion that "[t]he difficulties involved in determining the causality of injuries of this sort cannot be overstated." Ibid. at 201. It also denied partial certification because the issues which would have been pertinent in such a certification were "so thoroughly intertwined with those which must be determined individually." Ibid. at 201. It is obvious from these comments that it was the problem of individual causality which motivated the court in its denial in this case of class certification.

The next stab at class certification arose in In re Northern District of California "Dalkon Shield" IUD Products Liability Litigation, 526 F. Supp. 887 (N.D.Cal. 1981) vacated, 693 F.2d 847 (9th Cir. 1982), cert. denied, 459 U.S. 1171, 103 S.Ct. 817, 74 L.Ed.2d 1015 (1983). In that case, the District Judge, confronted with a mass of Dalkon Shield cases pending for trial in his court, concluded that a class certification was an appropriate — in fact, the only — available procedure for expediting the fair disposition of such suits and preventing a waste of judicial resources in trying individually cases in which there would be an endless repetition at great length of many of the same facts. He accordingly proceeded on his own motion to certify a nationwide class action on the issue of punitive damages under Rule 23(b)(1)(B) and a statewide (California) class action on the issue of liability under Rule 23(b)(3). He supported his decision with a cogently reasoned opinion. However, on appeal, that decision was reversed. 693 F.2d 847 (9th Cir. 1982).

The Court of Appeals, in reversing the District Court's certification sought to divide the certification issue into two broad classes: One deals with the class (b)(3) certification of the issue of liability for the manufacture and sale of a defective product; the second relates to the District Court's certification under the "limited fund" doctrine of the claim for punitive damages under (b)(1)(B). On the first point, the Court of Appeals said that, despite the fact that the certification under (b)(3) was "limited to the issue of liability," "Robins' overall liability, under some of the theories [of liability], cannot be proved unless each plaintiff also proves that Robins' breach of duty proximately caused her particular injury." 693 F.2d at 853, 856. It recognized that Rule 23(c)(4)(A) authorizes and encourages separating issues for certification where economies of trial may be achieved, but dismissed the use of such authority in this case because "[t]he complexity of issues peculiar to individual claims militates against grouping all plaintiffs into a class for only part of their recovery" and "[t]he few issues that might be tried on a class basis in this case, balanced against issues [such as causation and damages] that must be tried individually, indicate that the time saved by a class action may be relatively insignificant. A few verdicts followed by settlements might be equally efficacious." Ibid. at 855, 856. It then added this somewhat contradictory sentence: "We do not preclude further consideration by the District Court of motions to certify a more limited class or subclasses under Rule 23(b)(3)," Ibid. at 856, seemingly suggesting that if the District Court would further break the issue down certification might be appropriate. It proceeded to add that the petitioner for class certification had failed to satisfy the "typicality" requirement of subsection (a)(3) or the "superiority" requirement of (b)(3). For this conclusion the Court of Appeals relied on the Advisory Committee's Note to (b)(3) and on La Mar v. H B Novelty and Loan Co., 489 F.2d 461 (9th Cir. 1973) and McDonnell Douglas Corp. v. U.S. District Court for the Central Division of California, et al., 523 F.2d 1083 (9th Cir. 1975), cert. denied, 425 U.S. 911, 96 S.Ct. 1506, 47 L.Ed.2d 761 (1976), two cases which held that the "flexible" language of Rule 23 itself had, in the interest of manageability, to be reined in with constraining limitations, which were not expressed in the Rule. The Court of Appeals appears to have found also that the (a)(4) requirement of "adequate representation" was not met. In that regard, it stated that "[n]o plaintiff has appeared in this appeal in support of class certification," "no plaintiff [had] accepted" the role of class representative, "none of the attorneys already involved [was] willing to serve as class counsel" and the District Court had been forced to appoint counsel to maintain the class action. 693 F.2d at 850-51. The appointed counsel had resigned and the replacement had not started to represent the class. 693 F.2d at 850-51. However, in concluding this phase of its opinion, the Court of Appeals ambiguously said:

We are not necessarily ruling out the class action tool as a means for expediting multi-party product liability actions in appropriate cases, but the combined difficulties overlapping each of the elements of Rule 23(a) preclude certification in this case. 693 F.2d at 851.

In reversing the certification of the punitive damages claim under (b)(1)(B), the Court of Appeals held apparently that to qualify for certification under the "limited fund" doctrine, the movant for certification must prove the fact that the fund is inadequate to pay all claims "inescapably," 693 F.2d at 851, and it found that the District Court had erred in finding to the contrary in entering an order of certification of the punitive damage claim under (b)(1)(B).

In 1984, Robins itself thereafter filed an action seeking class certification of the claim for punitive damages in all Dalkon Shield suits in the Eastern District of Virginia. In re Dalkon Shield Punitive Damages Litigation, 613 F. Supp. 1112 (E.D.Va. 1985). Certification was denied. The District Judge construed the decision by the Ninth Circuit Court of Appeals in In re Northern District of California Dalkon Shield Products Liability Litigation, supra, ( 693 F.2d 847), as having held that Rule 23(a) with its requirements of commonality, typicality and adequate representation were not met in the negligence and warranty claims for punitive damages against Robins and concluded that such decision barred under familiar principles of collateral estoppel this subsequent request for certification of the propriety of punitive damages awards in Dalkon Shield cases.

The final effort at maintaining a class action against Robins before the filing of the action which is the subject of this appeal occurred in early 1985. At that time counsel who later filed this action instituted in the District Court of Virginia a proposed class action against Robins. The complaint in that action included the same general allegations which had been made in earlier attempts at maintaining a class action in the Dalkon Shield litigation but added an important new averment. It alleged that at the time the action was filed, Robins was insolvent and that the value of outstanding claims, filed and unfiled, exceeded Robins' value, thereby exposing filing victims to economic prejudice at the hands of more timely filed claims. The action sought a gross judgment, the proceeds from which would be equitably distributed among Dalkon Shield claimants under an alternative dispute resolution process. Robins opposed the suit and denied insolvency. Though a hearing was had by the District Court on the motion for class certification, no order on that motion was issued prior to the filing by Robins of the Chapter 11 proceedings. The motion has remained dormant since.

III.

During the period of these proceedings before the Judicial Panel and the attempts at class certification as already detailed, the flood of suits against Robins by Dalkon Shield claimants had continued unabated. By 1985 "an average in excess of seventy (70) cases per week were being filed," usually seeking "both compensatory and punitive damages." Moreover, large verdicts were being returned against Robins. Illustrative of that was a case in the same court before which the first verdict in a Dalkon Shield case in mid-1975 was returned. In 1985, with the same counsel representing the plaintiff, a verdict of $9.1 million (of which $7.5 million was awarded as punitive damages) was returned against Robins. In short, beginning with an award of $85,000 in 1975, awards in the Dalkon Shield cases escalated by 1985 with verdicts such as this latest verdict to over $9 million. Such suits understandably were taking their toll on Robins' resources.

IV.

Under these circumstances, it was but natural that some Dalkon Shield claimants and their counsel would have become apprehensive of the ability of Robins to meet the liabilities being asserted in all the pending Dalkon Shield cases. Another "deep pocket" had to be sought and a number of the claimants focused on Aetna as this other "deep pocket." Various suits were begun prior to the filing or immediately after the filing by Robins of its Chapter 11 petition against Aetna charging liability on its part as a tortfeasor with Robins.

Typical of such suits was Bast, et al. v. A.H. Robins Co., Inc., 616 F. Supp. 333 (E.D.Wis. 1985). In that case, the Dalkon Shield plaintiffs sought, in the words of the Court, "to impose direct liability for their injuries upon Aetna Casualty and Surety Company, the A.H. Robins Company's insurance carrier," under a claim that Aetna had, by its actions, made itself a joint tortfeasor equally liable as Robins. They alleged virtually the same substantive claims as did the plaintiffs in the case before us, to wit, negligence, express and implied warranty, misrepresentation and fraud, all garnished with a RICO claim. The Court found the complaint flawed because the plaintiffs had failed to "aver the existence of a duty of care owed the plaintiffs by Aetna, and resultant injuries proximately caused by such breach." 616 F. Supp. at 334. It, therefore, dismissed the action.

A few weeks after Bast was dismissed, and about a week before Robins filed its Chapter 11 petition, a similar case came before the District Court in the other district of the same state. Campbell, et al. v. A.H. Robins Co., Inc., 615 F. Supp. 496 (W.D.Wis. 1985). The complaint in this action charged the same negligence, warranty, misrepresentation, conspiracy, fraud and RICO counts as had Bast. Aetna filed a motion to dismiss and the Court granted the motion. In dismissing the suit, the Court said that all the claims were "faulty for the simple reason that an insurer owes no duty of care to injured third parties under Wisconsin law." In the same vein, the court said that "there must be allegations of how a defendant may have breached such duties, and how the plaintiffs' injuries may have been caused by the defendant." 615 F. Supp. at 500 (citation omitted). It added too that "retrospective liability of co-conspirators does not operate to make the late-entering conspirator responsible for the already completed substantive offenses of his cohorts." Ibid. It dismissed the RICO claim as "deficient in two respects: the complaint does not allege a causal connection between the racketeering offenses attributable to Aetna and the injuries that plaintiffs suffered; and the complaint does not, and cannot, allege an injury to `business or property' as required by 18 U.S.C. § 1964(c)." Ibid. (footnote omitted).

Other suits against Aetna were filed seeking to charge it as a joint tortfeasor with Robins in the Dalkon Shield litigation. So far as we can ascertain, motions for dismissal in all such actions were filed by Aetna and, such as had been heard and ruled on, had been granted. See Bathurst v. Aetna Casualty and Surety Co., No. 85-C-0324 (E.D.Wis. 1985) and Dyes v. Aetna Casualty and Surety Co., No. C-85-3932A (N.D.Ga. 1985).

V.

In addition to all these attempts at class action certification, either total or partial, and attempted suits against Aetna, there had been a running dispute between Aetna and Robins on the extent of the coverage under Aetna's policy. This difference apparently existed from the very beginning of the Dalkon Shield litigation. In 1979 Aetna had filed a declaratory action to resolve some of these disputes and differences. In October, 1984, a settlement was agreed upon in this litigation after extensive negotiations. In a compromise settlement of the dispute, Aetna agreed, among other things, to provide a small additional coverage above its claimed limits of $300 million. By mid-1985, there was argument between Robins and Aetna over the amount of remaining liability, if any, under Aetna's policy available for payment of judgments entered in the Dalkon Shield litigation.

VI.

By this time, Robins had been frustrated in every effort to reduce the expenses of defending the individual suits being filed through the use of the class action device. The costs of defense, the disruption that all the individual trials was placing on its executive force, and the expense of discharging some of the judgments were putting great financial strain on Robins. Its funds had been so depleted by the suits that its unrestricted funds had been reduced to $5 million and "financial institutions were unwilling to lend it money." It had actually experienced considerable difficulty in collateralizing as a condition of appeal the judgment in the case where the plaintiffs had recovered in early 1985 a $9.1 million judgment. Faced with its obvious financial deterioration, a deterioration which gave every indication of accelerating, Robins felt its only avenue for paying equitably and fairly all the claims was through a Chapter 11 proceeding. So on August 22, 1985, Robins filed in the Eastern District Court of Virginia its Chapter 11 petition. By the time Robins filed this petition it and/or its insurance carrier had settled 9,238 claims for approximately $530,000,000. Despite these settlements, Robins, at the time it filed for relief, still faced over five thousand pending cases in state and federal court. In re A.H. Robins Co., Inc., supra, 89 B.R. at 557.

VII.

A few months after the Chapter 11 petition was filed, seven Dalkon Shield claimants, suing individually and as the class representative of all Dalkon Shield claimants, filed this action against Aetna alone to recover for the injuries allegedly suffered by them as a result of the use of the Dalkon Shield. The plaintiffs (hereafter often referred to as the Breland plaintiffs) asked at the outset of their complaint that the action be certified as a class suit on behalf of the whole class of Dalkon Shield claimants, which class they averred should be divided into two such classes identified as Class A and Class B. Class A members were those claimants who had complied with all the requirements for proof as a claimant in the Robins bankruptcy proceedings, and Class B members were those who had failed to so comply. Alleging compliance with the prerequisites of Rule 23, they asked that the action against Aetna be certified as a class action against Aetna on behalf of both Classes A and B under Rule 23(b)(1)(A) and (B), without opt-out rights on the part of any member of either Class. They proffered allegations of justification for this action. They then separated their substantive claims into two categories. The first was characterized by them as their "Limited Fund Contract Claim." This claim related to Aetna's liability under its Robins policy. In 1984, after several years of haggling between Robins and Aetna, an agreement, as we have already said, was reached which changed in some particulars Aetna's obligation under its policy to the prejudice of those entitled to avail themselves of rights under it. It was the premise of the Breland plaintiffs that the Dalkon Shield claimants were third-party beneficiaries under the policy and as such had a common, undivided interest in the value of the policy. They sought "vitiation" of this 1984 agreement between Robins and Aetna. This was the subject of the "limited fund" claim.

The second claim of the Breland plaintiffs was for recovery of damages against Aetna as a joint tortfeasor for all injuries sustained by the two Dalkon Shield Classes of claimants by reason of the use of the Dalkon Shield. The plaintiffs alleged as a basis for their charge against Aetna as a joint tortfeasor that from "the time the Defendant [Aetna] agreed to insure Robins and continuing up to the present, Defendant became an active participant in all stages of the development, testing, promotion, and marketing of the product as well as being inexorably involved in the palliation of the public and the medical profession." They followed with a statement of the same claims of liability under negligence, strict liability, express and implied warranty, fraud, and conspiracy plus the standard RICO averments that were becoming routine allegations in the suits filed by Dalkon Shield claimants against Robins.

In its answer to this complaint, Aetna admitted at the outset federal jurisdiction "over this action pursuant to 28 U.S.C. § 1334(b) and 11 U.S.C. § 105 because this action arises in and is related to In re A.H. Robins Company, Inc. Chapter 11 Case No. 85-01307-R," and expressly conceded the accuracy of plaintiffs' allegations in paragraph 26 of their [the plaintiffs'] complaint, which was in these words:

The class is severely prejudiced and adversely affected unless a single class action is permitted to be maintained in which class representatives under strict court guardianship are in a position to bind the class and Aetna in a unitary proceeding against Aetna. The equitable underpinnings for the certification of mandatory Rule 23(b)(1) class actions are satisfied under these extraordinary circumstances.

It added affirmatively in that connection "that Aetna [itself] would be severely prejudiced if forced to respond to multiple suits filed by class members. The application of incompatible standards to Aetna would impair Aetna's ability to engage in the consistent, continuing course of conduct relative to its duties and obligations as an insurer."

On the merits of the claims, Aetna denied specifically all allegations of the complaint charging liability on its part along with Robins in the Dalkon Shield cases as a joint tortfeasor, stating affirmatively that

Aetna's action and activities complained of by plaintiffs are based on Aetna's relationship with Robins as its insurer, and not on any direct conduct or contact between Aetna and any of the plaintiffs claiming injury from use of the Dalkon Shield. No action by Aetna caused any injury to any plaintiff as a matter of fact or of law.

It said in further explication of this defense that it was "neither a manufacturer nor a seller of the Dalkon Shield and, therefore, is not subject to strict liability."

VIII.

With the joinder of issues in the action, the plaintiffs moved promptly for class certification in accordance with the allegations of their complaint. The District Court heard such motion on October 30, 1986. In its order entered some weeks after the hearing, the court found that "the requirements for class action certification under Federal Rules of Civil Procedure 23 [had] been" met and it conditionally certified the action. However, consistent "with the Court's determination by [its] Order of November 4, 1986, to lift the stay in this action to permit Plaintiffs to proceed with discovery" and "to assure that the class will have the benefit of all such discovery," it made its certification "subject to further consideration in light of discovery developments, whether this action should continue to be maintained as a class action, and if so, in what form."

In the meantime, a number of Dalkon Shield claimants, by counsel, filed for leave to intervene at the hearing on the motion for final certification in order to oppose class certification. The District Court granted the right to intervene for the purpose of filing briefs in opposition to class certification as well as "to present oral argument before the Court prior to the disposition of the class certification motion before this Court."

While the discovery was continuing under the District Court's order for conditional certification, the Official Dalkon Shield Claimants' Committee filed a motion with the District Court for leave to file its own complaint against Aetna. That motion was granted and the Committee proceeded to file its complaint. The Committee's complaint followed substantially the Breland plaintiffs' complaint on its broad claim against Aetna but it injected into the controversy an entirely new issue of considerable complexity. This additional allegation was that "[a]s a result of the joint negligence of Aetna and Robins, Aetna is liable to Robins for contribution for its pro-rata share of those damages already paid by Robins to Dalkon Shield users." Aetna responded to this complaint, admitting the insurance policy issued by it to Robins and that Robins had "expended in excess of $200,000,000 pre-petition with respect to Dalkon Shield claims" but denying absolutely on a number of grounds any claim that it was a joint tortfeasor along with Robins in the Dalkon Shield case. It, too, added an affirmative claim of a right to contribution from Robins in the event it was found liable as a joint tortfeasor. This latter claim was stated in paragraph 20 of Aetna's answer to the Committee's complaint as follows:

In the event, however, Aetna is liable to Robins for contribution or for other form of relief (all such liability being expressly denied by Aetna), then, as to any amounts which Aetna is required to pay to Dalkon Shield claimants (other than under the policies of insurance issued by Aetna to Robins) on account of injuries related to the Dalkon Shield, Robins is liable to Aetna for contribution as to all such payments to the extent based on conduct by Aetna found sufficient to give rise to a right by Robins to contribution.

This action of the Committee was consolidated with the Breland suit without objection and was disposed of in the orders granting certification and approving the settlement, all as subsequently discussed.

IX.

As discovery on the certification issue proceeded, considerable inquiry on the part of counsel for the Breland plaintiffs into all dealings between Robins and Aetna, including the files of Robins and those of Aetna, was conducted. At the same time, discussions of a possible settlement between the plaintiffs in Breland and Aetna were engaged in by counsel for the parties. The parties recognized that the Breland case and the Robins reorganization would necessarily involve some coordination if a feasible settlement were to be achieved. Such a coordination would require as an essential step a final class certification in the Breland case. To expedite certification, Aetna stipulated of record that "if Breland were certified, and if claims resolution in Breland could be coordinated with the claims resolution process which would be required in the Robins reorganization proceedings, [it] would agree not to litigate separately any of the non-common issues of individual medical causation and individual amounts of damages." This stipulation left open either for trial or for settlement the issue whether Aetna could be held liable as a joint tortfeasor with Robins; all other issues — the non-common ones of individual causation and damages — were to be resolved along with the similar claims in the Robins reorganization under a claims resolution procedure for the disposition of such claims.

A settlement figure on the common issue of Aetna's possible liability as a tortfeasor was extensively discussed. The discussion ranged somewhere "between zero and $300 million in cash." However, it became "obvious," as the Breland counsel wrote Aetna's counsel, "that somebody is going to buy Robins," and, though the parties to the Breland suit had "agreed in principle to a solution," he (the Breland counsel) felt that any final attempt at settlement should "wait until we see what type of proposal is seriously made by a serious suitor that has a chance of becoming a feasible plan of reorganization prior to our finally resolving our lawsuit." When American Home Products Corporation (Home Products) later entered the picture as a "serious suitor" for Robins, the negotiations in the consolidated suits of the Breland plaintiffs and the Claimants' Committee and the discussions of a Plan of Reorganization of Robins, based primarily on an offer of American Home, began to merge and take shape, looking to a combined settlement of all Dalkon Shield litigation.

X.

While discussions had been proceeding between Aetna and the Breland plaintiffs, the parties in the bankruptcy reorganization of Robins had fortunately been working at the same time on a Plan of Reorganization. As a basis for any plan of reorganization it was necessary that an estimation be made of the unliquidated claims against the debtor under Section 502(c) of the Bankruptcy Code, which directs that "[t]here shall be estimated for purposes of allowance under this section — (1) any contingent or unliquidated claim, the fixing or liquidation of which, as the case may be, would unduly delay the administration of the case." An estimation under this section contemplates a full adjudication. The estimation of the unliquidated claims in this case was arrived at after testimony by expert witnesses, taken in a proceeding before the District Judge, where all parties were represented by counsel and had participated. The result of the hearing was an estimation made by the District Judge on the basis of the full record before him. The Dalkon Shield claims constituted substantially all the unliquidated claims involved in the estimation. The value of such unliquidated Dalkon Shield claims was fixed in the District Judge's estimation order at an "aggregate" dollar amount or value of $2.475 billion. The District Judge announced to all parties interested in the reorganization that no plan of reorganization would be considered which failed to provide this amount for the full payment of all Dalkon Shield claims.

At this point the American Home Products Corporation (American Home) made an offer that gave promise of enabling the debtor to meet this requirement of the District Judge. It proposed a merger of Robins with a subsidiary of American Home, which, with the assistance of its parent company, would as an incident of the merger provide, among other things, a sum, which, supplemented with some other anticipated contributions would, provide a fund of $2.475 billion for the full satisfaction of the unliquidated Dalkon Shield claims. One expected to make some contribution was Aetna. The reason for such inclusion was American Home's insistence that its contribution was contingent on a settlement by all Dalkon Shield claimants of their claims against any party. This insistence on American Home's part was a desire to assure the new company to be formed by the merger against any Dalkon shield liability. It may be that American Home felt that if Aetna were left out and Dalkon Shield claimants prevailed on its joint tortfeasor claim against Aetna, Aetna would acquire a right of contribution against Robins or its successor under the laws of a majority of the states. American Home apparently demanded that the new company, Robins and American Home, be protected against that risk. It accordingly became clear to the parties in the bankruptcy, as it had to the parties in the Breland case, that the negotiations for settlement of the suit of the Breland plaintiffs and of the Claimants' Committee had to be merged into a comprehensive procedure for the liquidation of all Dalkon Shield claims against Robins and Aetna. This made it essential that the settlement of the Breland case be tied to the Plan of Reorganization of Robins, and that is what the parties in the two proceedings did.

The procedure adopted to achieve this merger of the Breland action and the Plan of Reorganization of Robins began with the formulation of what was intended as the Final Plan of Reorganization of Robins. This Plan of Reorganization contemplated the creation of a Trust Fund to consist of $2.475 billion, to be funded primarily by payments made by American Home in connection with its acquisition by merger of Robins and by contributions of Aetna. Since it was manifest that Robins' liability in the Dalkon Shield litigation was direct while that of Aetna's liability as a possible joint tortfeasor was rather tenuous, if at all, the contribution of Robins through its merger with American Home's subsidiary and that of Aetna would vary markedly, and it did. The contribution of Aetna was to consist of a net of $75 million and of four policies of insurance, one of which was to be in the total amount of $250 million to be used "in the event there [were] not sufficient funds of the Trust (including investment income) to make further payments on claims" and the other to consist of an approved policy or policies of insurance in the amount of $100 million to pay Class B claimants whose claims were "ineligible for such recovery on a nonsubordinated basis." Aetna was to pay the premiums on all such insurance. Upon the approval of the Plan of Reorganization and the Certification and Settlement in the Breland suit, and the payments and transfers of insurance contracts to the Trust Fund required under the Plan and Settlement, the claims of all Dalkon Shield claimants against both Robins and Aetna were to be converted into claims solely against the Trust Fund established to pay in full their claims as proved under the procedure provided in the Plan, and Robins, Aetna and American Home and a successor company and the individuals contributing to the fund were to be released absolutely from any liability to Dalkon Shield claimants.

The Trust itself was to be operated by five trustees named in the Plan. The trustees were to create a Claims Resolution Facility which would resolve the liability due to each Dalkon Shield claimant. The issue of liability in each case would be resolved, if possible, by negotiations between the Facility acting for the Trust and the claimant and her attorney, if she had an attorney. If resolution of the claim could not be accomplished through negotiation, then the claim in every case would be resolved either by binding arbitration or by a jury trial at the option of the claimant. Should any claimant elect to settle her claim by suit, venue of such trial "[would] be unchanged by the Chapter 11 case" and the "right to a jury trial [would] be preserved," with "[a]ll claims and defenses . . . available to both sides in a trial." See Exhibit Vol. III, pp. 1952-53. Under Aetna's stipulation, already described, this provision meant that questions of individual causation and damages in both the Robins bankruptcy and in the Breland case would be resolved by the above proceedings.

XI.

Considerable testimony was taken — much of it before the District Court itself — on the propriety of certification and of the Settlement. The parties' in interest engaged in extensive cross-examination of the witnesses in the various proceedings. A full record of such proceedings is included in the Joint Appendix herein. Some of the records considered in the matter were made before the Claimants Committee's action was filed, but a great part was after that filing. In addition, the discovery developed as a result of the proceedings before Judge Lord as well as the report of the masters appointed by Judge Lord in that proceeding was made a part of the record. Moreover, the masters testified in person in connection with their investigations and reports. At the conclusion of the receipt of testimony and admission of exhibits, lengthy arguments of counsel both for and against certification and of the Settlement were heard by the District Court.

On the basis of the record and the arguments, oral and written, by the parties, the District Court made its findings and conclusions separately on both the motion for class certification and the motion to approve the Settlement. It is essential in reviewing these decisions to remember that we are reviewing two decisions, both made substantially on the same factual record but dealing separately with the different issues in the two appeals. However, before we discuss the two appeals on their merits, we must address a jurisdictional issue, which, if upheld, would require the dismissal of the entire action.

XII.

The appellants have raised a threshold claim that this action fails the jurisdictional test established for a diversity action in federal court. The jurisdictional flaw is said to be the failure of the action to meet the jurisdictional amount test under 28 U.S.C. § 1332. The appellants base this contention on Zahn v. International Paper Co., 414 U.S. 291, 301, 94 S.Ct. 505, 512, 38 L.Ed.2d 511 (1973), in which the Supreme Court said that "[e]ach plaintiff in a Rule 23(b)(3) class action . . . must satisfy the jurisdictional amount, and any plaintiff who does not must be dismissed from the case." We address this jurisdictional issue first, as we should.

Unquestionably, all the named plaintiffs in this case meet the jurisdictional amount test for federal diversity jurisdiction. Each plaintiff alleges a claim in excess of $10,000. Such an allegation was sufficient at the time to establish jurisdiction. St. Paul Mercury Indemnity Co. v. Red Cab Co., 303 U.S. 283, 288, 58 S.Ct. 586, 590, 82 L.Ed. 845 (1938). But it is the appellants' construction of Zahn that not only must the named plaintiffs in a proposed class action suit meet the jurisdictional amount but that all unnamed class claimants must likewise do so. Under this argument, it would be necessary to find on the record that the 300,000 Dalkon Shield claimants met the jurisdictional test. We do not understand, however, that the appellants assert that the plaintiffs herein must establish by direct proof that the claims of the unnamed class members meet the jurisdictional amount, nor would prevalent authority sustain such contention, if made. The courts which have considered the jurisdictional amount requirement in the mass tort class action context have ruled that the court will dismiss on this ground only if it can be said "to a legal certainty" that the claims of the unnamed claimants fail the jurisdictional amount test. This was the decision of the court in the first decision to address this point in a mass tort case, the decision which is generally regarded as the guiding authority in this connection. Payton v. Abbott Labs., 83 F.R.D. 382, 395 (D.Mass. 1979).

Payton, on its facts, has striking similarity with this case. It involved claims by a class of women suing for injuries caused by exposure to diethylstilbestrol (DES), manufactured and sold by the defendant pharmaceutical manufacturer. The defendant objected to class certification on the ground "that members of the plaintiff class [did] not satisfy that [the jurisdictional amount] minimum and that, in the context of this action, separating members of the class who meet the jurisdictional minimum from those who do not is impossible." The Court, in deciding the issue in that case, looked first to St. Paul Indemnity, which had said that "[i]t must appear to a legal certainty that the claim is really for less than the jurisdictional amount to justify dismissal" for want of the jurisdictional amount, 303 U.S. at 289, 58 S.Ct. at 590, and, applying that principle, denied the motion of the defendant to dismiss, saying

Plaintiffs' claimed damages are unliquidated and subject to a jury's evaluation of many subjective factors. I cannot now find to a legal certainty that the claim of any member of the plaintiff class is less than the jurisdictional amount. 83 F.R.D. at 395.

In In re No. District of Cal. "Dalkon Shield" IUD Products, 526 F. Supp. 887, 910-11 (N.D.Cal. 1981), in which there was, as we have said, a class certification of Dalkon Shield claimants, the District Court apparently sua sponte noted that it might be argued in denial of jurisdiction as do the appellants here that "each class member, named or unnamed" may not have a claim meeting "the amount in controversy requirement." The court, however, dismissed such possible jurisdictional objection. In so doing, it relied, as did Payton, on St. Paul Mercury Indemnity Co. v. Red Cab Co., 303 U.S. 283, 288, 58 S.Ct. 586, 590, 82 L.Ed. 845 (1938), with its rule requiring a finding of "legal certainty" for its dismissal on the jurisdictional issue. On appeal, the decision of the District Court on the merits was reversed, but the appellate court did not disapprove of the jurisdictional ruling made by the District Court. That court did not even notice the point in its decision, even though the District Court in its opinion had ruled on the point. In fact, it is not clear from the opinion that the appellants even raised the point. In any event, since this jurisdictional defect, if sound, would have made unnecessary consideration of the other points on appeal and should have been considered by the Court sua sponte if of merit, it seems safe to assume that the Court of Appeals agreed with this part of the District Court's ruling. See In re No. District of Cal. "Dalkon Shield" IUD Products, 693 F.2d 847 (9th Cir. 1982), cert. denied, 459 U.S. 1171, 103 S.Ct. 817, 74 L.Ed.2d 1015 (1983).

In In re Agent Orange Prod. Liability Litigation, 818 F.2d 145, 163 (2d Cir. 1987), cert. denied, ___ U.S. ___, 108 S.Ct. 695, 98 L.Ed.2d 648 (1988), which was a class action suit on behalf of all former servicemen claiming injuries as resulting from the use of Agent Orange by the military forces in Vietnam, the same question arose whether the plaintiffs class satisfied the standard established by Zahn. In answering this question, the court noted the rule in St. Paul Mercury Indemnity to which we have already referred. It applied that rule and dismissed the point with this statement:

Appellants do not argue that any class members made bad faith damage claims. Nor do they offer us any basis for determining whether such claims clearly are for less than the jurisdictional amount. Instead, they claim the District Court failed to carry out an obligation to police the damage claims. No such affirmative obligation exists, however, absent some apparent reason to make inquiry. Plaintiffs made what must be assumed to have been good faith allegations that each of them was entitled to at least $10,000 in damages. Defendants did not challenge the bona fides of these claims, and the District Court thus had no reason to inquire further.

The author of the comment, Federal Mass. Tort Class Actions: A Step Towards Equity and Efficiency, 47 Alb.L.Rev. 1180, 1189-90, n. 36 (1983), suggests as the proper rule to be applied in the mass tort situation that the action will be permitted to proceed against a jurisdictional amount challenge if the class plaintiffs themselves all meet the jurisdictional amount requirement since it "would be consistent with the rule that once federal jurisdiction in a class action has been fixed by the parties of record, subsequent intervenors do not need to meet the jurisdictional amount." Ibid., 1190 n. 36. This rule would not be inconsistent with the literal reading of Zahn, since that decision said simply that, if any class member's claim was less than the jurisdictional amount, that class member's claim should be dismissed. However, it is not necessary to go this far in order to dispose of this challenge to the maintenance of the action herein. Applied as it was by the court in Payton and the other cases, St. Paul clearly sustains jurisdiction here. Thus, there have been countless suits, both in federal and state courts, to recover for injuries caused by the use of the Dalkon Shield. So far as a hasty review of the reported cases has indicated, not one suit has been filed in which the plaintiff sought less than the then jurisdictional amount. In the face of such a record, it would appear to be indisputable that it cannot be said to a "legal certainty" that the jurisdictional amount herein was not satisfied. We accordingly dismiss appellants' claim of lack of jurisdictional amount herein under St. Paul Mercury Indemnity.

XIII.

We turn now to the challenge of the certification order. This order was issued in what is emerging to be one of the most difficult management problems confronting the courts today, particularly federal courts, i.e., that posed by the mass tort suit. Mass. tort suits are generally divided into two classes. Note, Federal Mass. Tort Class Actions: A Step Toward Equity and Efficiency, 47 Alb.L.Rev. 1180, 1183 (1983); 3 Newberg on Class Actions, § 17.21, pp. 394-47 (2d ed. 1985). The first is concerned with what is designated as the mass accident suit, in which a large number of persons are injured as a result of a single accident. A familiar instance of such a tort is an airplane crash or the failure of a building as in In re Federal Skywalk Cases, 680 F.2d 1175 (8th Cir. 1982), cert. denied, 459 U.S. 988, 103 S.Ct. 342, 74 L.Ed.2d 383 (1983). The other type of mass tort action, of which this case is an outstanding example, is one arising out of the sale, on a national or international market to thousands of persons, of an alleged defective product, from the use of which has caused many persons to have suffered injury. In some of these products liability cases, the number of persons involved may be enormous. The number of potential plaintiffs, in this case, for instance, are in the hundreds of thousands. In most instances, this type of mass tort will generate far more cases than the normal accident type.

Within recent years, the proliferation in the development and distribution of new products and remedies and the complaints of injuries from the use of these products have brought an accelerating avalanche of mass products liability suits primarily in federal courts which represents what is, as we have already observed, probably the most important and difficult management problem facing the federal court system today. Professor Seltzer has well summarized the size of the problem:

Mass. tort litigation in the 1980's has reached unprecedented levels. Thousands of personal injury lawsuits have been filed against manufacturers of such mass-marketed products as asbestos, formaldehyde, diethylstilbestrol (DES), Agent Orange, automobiles, tampons and intrauterine contraceptive devices [IUD's].

The Note, Class Certification in Mass. Accident Cases under Rule 23(b)(1), 96 Harv.L.Rev. 1143 (1983), is to the same effect:

Mass. accidents are a recurring phenomenon of modern technological society. It is not unusual for hundreds or thousands of injuries to be caused by a sudden event — an airplane crash or a structural building failure — or by a product defect in a widely used drug. Victims typically seek redress for such injuries by initiating individual suits. As a result, the judicial system becomes crowded with multiple claims concerning one basic course of conduct, with each victim seeking to maximize his own recovery and each defendant attempting to reduce his potential liability. As each party acts to further a narrow perception of his own best interests, the collective interests of the parties, society, and the judicial system may be ignored. Ibid. (footnote omitted)

The burden on the courts that these mass torts imposes and the tremendous public and private costs arising from the endless trials of the same issues in repeated court actions covering the same facts in great detail are illustrated by a recent analysis of only one aspect of the cost of such litigation to the parties in the Agent Orange mass tort litigation:

By way of illustration, in the Agent Orange litigation, solely with respect to the government contract defense, the defendants took more than 200 depositions of former or current government employees. Assume that the depositions were performed by outside counsel and that each deposition took two days. Assume that the defendants' counsel each spent one day preparing for each deposition and another day summarizing and reviewing. Assume that the law firms used teams of two attorneys with an hourly billing rate for each attorney of $150. The calculation begins with: 200 depositions X 4 days X 8 hours = 6400 hours. The next step is 6400 hours X 14 attorneys (2 for each defendant) X $150 per hour = a total cost for attorneys of $13,440,000 for deposition costs for this aspect of the suit, leaving out transcript fees, travel costs, and other expenses. This cost would be multiplied several times over in the taking of depositions of the plaintiffs, the defendants, their physicians, and expert witnesses.

Judge Rubin of the Fifth Circuit, in an article titled Mass. Torts and Litigation Disasters, 20 Ga.L.Rev. 429, 430 (1986) has said:

These mass tort claims have a number of similarities: they result in the filing of many suits; they produce high litigation costs; they are generally resolved only after great delay; they affect not only the litigants but other users of the court system; and their total human and economic costs affect all of society.

Judge Rubin then reviewed one example of such mass torts, i.e., the asbestos litigation and made this comment based on a Rand Corporation report:

Asbestos litigation [for instance] has resulted in far more expense than in recovery of damages for injured persons. A Rand Corporation study estimated that injured persons receive less than thirty-seven percent of the total amount spent on litigation. Almost two-thirds of the total expenditures are for attorneys' fees and other litigation expenses.

When account is also taken of the toll of such cases on the court system itself, it is evident that the proper functioning of the courts and the fair and efficient administration of justice for other litigants whose right to a judicial determination are inevitably delayed inordinately by the clogging of the court system by mass tort actions tried individually and the societal costs of the endless repetition of these suits in separate trials at substantial costs to the judicial system mean that a mechanism for deciding expeditiously, efficiently and relatively inexpensively these actions without the delays of individual suits is demanded. This demand has been well pointed up by one of the recent groups of mass tort suits. In In re Bendectin Products Liability Litigation, 102 F.R.D. 239, 240 (S.D.Ohio 1984), Judge Rubin, commenting on the waste of judicial time involved in individual disposition of one mass tort products case in this court estimated that "[f]or the identified . . . cases, a full trial on each would probably require at least 30 trial days. Two Bendectin cases have already been tried. One required two trials of 44 and 45 days each. The other required 20 trial days. Assuming 200 trial days available per year per Judge, disposition of the present Bendectin cases at the trial level alone might require 21,000 trial days or the equivalent of 105 Judge years, i.e., one Judge for 105 years or 105 Judges for one year." Ibid., at 240 n. 3.

There is no clearer statement of the critical situation in the operation of the judicial system created by the mass tort litigation explosion than the language in a recent article co-authored by the lead counsel for the appellants in this very case:

The public is crying out for better solutions to mass tort litigation than separate redundant trials in thousands of related cases. Why should hundreds, or even thousands, of trials take place in which the same issue of a defendant's liability is litigated over and over again? Why should some plaintiffs recover compensatory and punitive damages, while other plaintiffs in other trials recover much less, or nothing at all, even though the issues of liability and causation are the same? Why should a defendant be put to the expense of parading the same witnesses on the stand in hundreds of different trials to demonstrate over and over again the same evidence that its product was correctly designed and engineered? Is it fair that defendants are exposed to multiple punitive damage claims?

Panzer Patton, Utilizing the Class Action Device in Mass. Tort Litigation, 21 Tort Insurance Law Journal, 560, 561 (1986).

The class action, as this article suggests, is the manifest fair and expeditious procedure for disposing of the mass tort litigants. It has been for years the recognized procedure for dealing with cases which have a marked similarity to those flowing from mass torts. The class action was judicially developed first in the equity courts and later accepted and applied by the law courts. Drawing on English and state procedural decisions, the Supreme Court over a century ago stated in Smith v. Swormstedt, 57 U.S. (16 How.) 288, 303, 14 L.Ed. 942 (1853) and restated later in Hartford Life Ins. Co. v. IBS, 237 U.S. 662, 672, 35 S.Ct. 692, 59 L.Ed. 1165 (1915) and Hansberry v. Lee, 311 U.S. 32, 41-42, 61 S.Ct. 115, 118, 85 L.Ed. 22 (1940) the circumstances justifying the use of such procedure in federal court proceedings. In Hansberry, that court declared that the class action device was an "invention [or innovation] of equity [designed] to enable . . . [the court] to proceed to a decree in suits [or judgment] where the number of those interested in the subject of the litigation is so great that their joinder as parties in conformity to the usual rules of procedure is impracticable." To the same effect, see General Telephone Co. of Southwest v. Falcon, 457 U.S. 147, 155, 102 S.Ct. 2364, 2369, 72 L.Ed.2d 740 (1982). In Falcon, the Supreme Court said:

The class action device was designed as "an exception to the usual rule that litigation is conducted by and on behalf of the individual named parties only." Califano v. Yamasaki, 442 U.S. 682, 700-701 [ 99 S.Ct. 2545, 2557, 61 L.Ed.2d 176]. Class relief is "peculiarly appropriate" when the "issues involved are common to the class as a whole" and when they "turn on questions of law applicable in the same manner to each member of the class." Id., at 701 [ 99 S.Ct. at 2557].

The procedure to be followed by a federal court in such a case is now embodied in Rule 23 of the Federal Rules of Civil Procedure as adopted in 1938. The formulation of the practice in this Rule was intended to embody the standards and purposes substantially already established in the federal decisions beginning with Smith. Rule 23 was revised in 1966 in its present form but without change in its purpose. It would seem reasonable to assume that the Rule was intended to be applied with the same flexibility and in the same spirit of "invention" as earlier prompted the adoption of the procedure. We accordingly turn to a review of the Rule itself as well as of the decisions which have applied it in cases such as mass tort suits or similar suits.

XIV.

Under the Rule, certification of a class action, whether a mass tort action or not, requires that the action meet the requirements of a two-step test. As a first step, an action must satisfy all four of the prerequisites mandated by subsection (a) of the Rule. These requisites are: (1) numerosity of parties; (2) commonality of legal and factual issues; (3) typicality of the claims and defenses of the class representative; and (4) adequacy of representation. Assuming qualifications under (a) have been met, the next step demands that the action fit within at least one of the three categories of actions identified in subsection (b) of the Rule. An action may qualify under (b)(1), the first of such categories, "if individual adjudication of the controversy would prejudice either the party opposing the class, (b)(1)(A), or the class members themselves, (b)(1)(B)." Zimmerman v. Bell, 800 F.2d 386, 389 (4th Cir. 1986); Intern. Woodworkers v. Chesapeake Bay Plywood, 659 F.2d 1259, 1269 (4th Cir. 1981). Subsection (b)(2) covers suits for injunctive or declaratory relief. The final category, (b)(3), applies where there are common issues of law or fact and the class action device has superiority over any other available procedure for disposing fairly and efficiently of the controversy.

Though not specified in the Rule, establishment of a class action implicitly requires both that there be an identifiable class and that the plaintiff or plaintiffs be a member of such class. 7A C. Wright, A. Miller M. Kane, Federal Practice and Procedure: Civil 2d § 1760, pp. 115 et seq. (2d ed. 1986); Roman v. ESB, Inc., 550 F.2d 1343, 1348 (4th Cir. 1976). Moreover, often an action may qualify under both (b)(1) and (b)(3). In such a situation, certification is to be made under (b)(1). 7A C. Wright, A. Miller M. Kane, Federal Practice and Procedure: Civil 2d § 1772, 425 (2d ed. 1986); 3B Moore's Federal Practice, § 23.31[3], p. 236 (1987 ed.); Robertson v. National Basketball Ass'n., 556 F.2d 682, 685 (2d Cir. 1977); Green v. Occidental Petroleum Corp., 541 F.2d 1335, 1340 (9th Cir. 1976); Meyer v. Citizens Southern Nat'l Bank, 106 F.R.D. 356, 363 (D.Ga. 1985). The rationale for such decision is stated by Moore thus:

If an action can be maintained under (b)(1) and/or (b)(2), and also under (b)(3), the court should order that the suit be maintained as a class action under (b)(1) and/or (b)(2), rather than under (b)(3), so that the judgment will have res judicata effect as to all the class (since no member has the right to opt out in a (b)(1) or (b)(2) suit), thereby furthering policy underlying (b)(1) and (b)(2) class suits.

Too, if the action includes multiple claims, one or more of which might qualify as a certifiable class claim, the court may separate such claims from other claims in the action and certify them under the provisions of subsection (c)(4). See 1 Newberg on Class Actions, § 4.20, pp. 310-1 (2d ed. 1985). This subsection provides that "an action may be maintained as a class action as to particular issues only. . . . Two or more issues, for instance, may be represented in a single action." Among such separate issues, for instance, may be a "limited fund" in which all members of the class have a common interest: such a class is recognized in the Comments of the Advisory Committee as qualifying under subdivision (b)(1)(B). In re Bendectin Products Liability Litigation, 749 F.2d 300, 306 (6th Cir. 1984) ("The District Court was therefore not clearly erroneous as a matter of law to hold that a limited fund is a justification for a class action under a Rule 23(b)(1)(B)"); Coburn v. 4-R Corp., 77 F.R.D. 43, 45 (E.D.Ky. 1977); Hernandez v. Motor Vessel Skyward, 61 F.R.D. 558, 561, n. 8 (S.D.Fla. 1973, aff'd without opinion, 507 F.2d 1278 (1975). However, to repeat: each subclass must independently meet all the requirements of (a) and at least one of the categories specified in (b). Roby v. St. Louis Southwestern Ry. Co., 775 F.2d 959, 961 (8th Cir. 1985). Moreover, when an action is certified under (b)(1), whether (A) or (B) or under (b)(2), all party members become mandatory class members without any opt-out rights, Kyriazi v. Western Elec. Co., 647 F.2d 388, 393 (3d Cir. 1981); if certified under (b)(3), however, any class member may opt out. The burden of establishing that a case meets the requirements for class certification under the Rule rests on the party seeking certification. Davis v. Romney, 490 F.2d 1360, 1366 (3d Cir. 1974). The assessment required for a certification, though, is the responsibility of the District Court, which is to make its decision after "a rigorous analysis" of the particular facts of the case. The "District Court has wide discretion in deciding whether or not to certify a proposed class," and its decision may be reversed "only for abuse of discretion." Jenkins v. Raymark Industries, Inc., 782 F.2d 468, 471-72 (5th Cir. 1986).

XV.

After the Rule was approved, the early view was that Rule 23 should be given "a liberal rather than a restrictive interpretation." See Eisen v. Carlisle Jacquelin, 391 F.2d 555, 563 (2d Cir. 1968), rev'd on other grounds, 417 U.S. 156, 94 S.Ct. 2140, 40 L.Ed.2d 732 (1974); Moss v. Lane Co., 50 F.R.D. 122, 125 (W.D.Va. 1970), aff'd in pertinent part, 471 F.2d 853 (1973). The decisions on the proper standard of construction under this view were that "if there is to be an error made, let it be in favor and not against the maintenance of the class action," Esplin v. Hirschi, 402 F.2d 94, 99 (10th Cir. 1968), cert. denied, 394 U.S. 928, 89 S.Ct. 1194, 22 L.Ed.2d 459, (1969); Wright v. Stone Container Corp., 524 F.2d 1058, 1061-62 (8th Cir. 1975). But despite the clear mandate in the first Section of the Rules that such Rules "[should] be construed to secure the just, speedy, and inexpensive determination of every action," some courts decided to depart from this liberal construction of Rule 23 and to adopt a standard of construction of the basis of the standard of "strict scrutiny." Roby v. St. Louis Southwestern Ry. Co., supra, 775 F.2d at 961. The reason for this departure from the liberal and flexible construction of Rule 23 and for the adoption of the "strict scrutiny" standard was stated in La Mar v. H B Novelty Loan Company, 489 F.2d 461, 468 (9th Cir. 1973). In that case, the Court based its strict construction of the Rule on its "belief that restrictions on the flexible language of Rule 23 [was] a necessary contribution to the effort to avoid the intractable problems of massive class actions and to maintain a wholesome degree of difference between the judicial and administrative functions" (Italics added). It declared that such construction was one "of prudence and caution" in line with what it declared was the "tone of the Advisory Committee's Note" (referring to the Advisory Committee's Note on Rule 23(b)(3)). 489 F.2d 465, 466.

The "belief" that the class action procedure should be sharply reined in prompted the court which decided La Mar later to create two rules intended to accomplish this purpose of constraining the use of the class action device. It is important to remember that neither limitation, designed purposefully to hobble the use of the class action device, was said to be justified in the language of the Advisory Committee's Note; these limitations were judicially imposed to justify a constrained use of the Rule mistakenly adopted in the belief that the manageability of mass tort litigation was beyond the resources of the courts. The first of the limitations found expression in Green v. Occidental Petroleum Corp., 541 F.2d 1335, 1340 (9th Cir. 1976), where it was held that suits for damages were not appropriate for class action certification. It cited in support of this principle La Mar, supra, which in turn had relied in large part on this statement in the Advisory Committee's Note on (b)(3):

A "mass accident" resulting in injuries to numerous persons is ordinarily not appropriate for a class action because of the likelihood that significant questions, not only of damages but of liability and defenses of liability, would be present, affecting the individuals in different ways. In these circumstances an action conducted nominally as a class action would degenerate in practice into multiple lawsuits separately tried.

For its second limitation, it held in McDonnell Douglas Corp. v. U.S. District Ct., C.D. of California, 523 F.2d 1083 (9th Cir. 1975), that the fact that one judgment in favor of one plaintiff in a mass tort situation and a judgment against the plaintiff in another suit in the mass tort situation would not be the type of inconsistency referred to in Rule 23(b)(1)(A).

It is evident that these cases which have sought to circumscribe drastically the courts in the use of the class action device find their rationale in the Advisory Committee Note on (b)(3) of Rule 23 for their result. That Note was written in connection with the 1966 version of the Rule. Since that time, it has become recognized that the reluctance to use class actions in mass tort cases, as stated in cases such as La Mar and its progeny, represents an unnecessary limitation on the efficient case management in federal courts of mass torts. As a consequence of this changed attitude, commentators have expressed their dissatisfaction with the constraints which the Advisory Committee's Note and the decisions of the Ninth Circuit would place on the use of Rule 23 in the mass tort context. Thus, Newberg on Class Actions states:

Mass. torts in modern-day jurisprudence are taking a fresh look at the value of Rule 23 class actions. The economies of time, effort, and expense of the class device cut across categorical tort lines and ought not to be obscured by the narrow application of circumstances or by undue emphasis on traditional interests in one-to-one litigation.

I was an ex officio member of the Advisory Committee on Civil Rules when Rule 23 was amended, which came out with an Advisory Committee Note saying that mass torts are inappropriate for class certification. I thought then that was true. I am profoundly convinced now that that is untrue. Unless we can use the class action and devices built on the class action, our judicial system is simply not going to be able to cope with the challenge of the mass repetitive wrong that we see in this case and so many others that have been mentioned this morning and afternoon.

[Quoting from Statement by] Prof. Charles Alan Wright In Re: School Asbestos Litigation Master File 83-0268 (ED Pa) Class Action Argument, July 30, 1984, Tr. 106.

Mr. Newberg in an article published in Trial, February Issue 1986, at page 53, under the title Mass. Tort Class Actions, expressed in somewhat extended form the same view:

Mass. tort class actions are rapidly emerging as a way to handle claims resulting from negligent acts or defective or toxic products affecting groups of similar parties. The historical barriers to class actions in the negligence field are fast giving way. Why? The problems of trying to resolve personal injury claims of often tragic proportions on anything but an individual basis still exist, do they not? Yes, but. . . .

Two benchmarks herald the arrival of mass tort class actions. First, the emerging judicial acceptance of class suits involving mass torts is undoubtedly the result of cumulative effects of mass production, with its attendant imperfections, in the context of a growing population and a court system with finite growth dimensions. The other arises from a single event — the Johns-Manville bankruptcy proceedings.

Society relies on the tort system, as enforced by private damage actions, to prevent or deter mass accidents and torts, and to compensate their victims. As mass production increases, there will inevitably be an increase in personal and economic injuries arising from mass accidents and from the mass distribution of defective or toxic products. Lawsuits seeking redress for these injuries will also inevitably multiply.

Many courts are now abandoning their historical reluctance to certify mass tort class actions in light of what is often an overwhelming need to create an orderly, efficient means for adjudicating hundreds or thousands of related claims. It is not surprising that the Manual for Complex Litigation, Second, has a section dealing with class actions in mass disasters and other complex tort cases. More and more tribunals now recognize that a central judicial tool for coordinating many related tort claims is the class action.

In the Tentative Draft No. 1 issued by The American Law Institute Complex Litigation Project and prepared for consideration at the 1989 Annual ALI Meeting, the Committee on the Federal Intrasystem Consolidation, of which Professor Arthur Miller was the Reporter, declared (Pages 36-37):

Rule 23 provides for the adjudication of the claims or defenses of an entire class of similarly situated parties in a single action. The Federal Rule is intended to eliminate or reduce the threat of repetitive litigation, to prevent inconsistent resolution of similar cases, and to provide an effective means of redress for individuals whose claims are too small to make it economically viable to pursue them in independent actions. Despite Rule 23's ambitious goals, multiparty, multiforum cases often are not certified for class treatment because its requirements have been read quite restrictively by some federal courts. Large scale tort actions involving personal injuries rarely are certified. In the past, this may be due primarily to the federal courts' reliance on the statement in the 1966 Advisory Committee Note that

[a] "mass accident" resulting in injuries to numerous persons is ordinarily not appropriate for a class action because of the likelihood that significant questions, not only of damages but also of liability and defenses to liability, would be present, affecting the individuals in different ways. In these circumstances, an action conducted nominally as a class action would degenerate in practice into multiple lawsuits separately tried.

Fed.R.Civ.P. 23(b)(3), Advisory Committee Note (1966). Although this reasoning may be criticized as shortsighted, it nonetheless has been influential. In addition, concerns about how to handle individual issues and large numbers of claimants have served to restrict class certification in nationwide products liability cases, as well as in consumer, securities, and antitrust actions. Recent years, however, have seen some weakening in the resistance to the certification of mass tort class actions. E.g., Jenkins v. Raymark Indus., Inc., 782 F.2d 468, rehearing en banc denied, 785 F.2d 1034 (5th Cir. 1986); In re School Asbestos Litigation, 104 F.R.D. 422 (E.D.Pa. 1984), vacated in part, aff'd in part, 789 F.2d 996 (3d Cir. 1986), cert. denied, 479 U.S. 852, 107 S.Ct. 182, 93 L.Ed.2d 117 (1986). Nonetheless, the full procedural advantages of class actions have not been realized.

Among the cases which the courts were said to have "read the requirements of Rule 23 narrowly," the Draft identified In re Northern District of California Dalkon Shield IUD Prods. Liability Litigation, 693 F.2d 847, (9th Cir. 1982), cert. denied, 459 U.S. 1171, 103 S.Ct. 817, 74 L.Ed.2d 1015 (1983); La Mar v. H B Novelty Loan Co., 489 F.2d 461 (9th Cir. 1973); In re Federal Skywalk Cases, 680 F.2d 1175 (8th Cir.), cert. denied, 459 U.S. 988, 103 S.Ct. 342, 74 L.Ed.2d 383 (1982). Page 40 of the Draft. Later, the Draft observes (page 43):

Several commentators have decried the courts' narrow approach to the Rule, arguing it should be interpreted more broadly to achieve unitary adjudication.

It cites in support Professor Miller's article, An Overview of Federal Class Actions: Past, Present and Future 45 (Fed. Jud. Center 1977); 96 Harv.L.Rev. 1143 (1983) and 47 Alb.L.Rev. 1180, 1199 (1983), already discussed.

It will be observed that the principal reason assigned by the Advisory Committee for its cautionary comment on mass torts in the 1966 Notes was that individualized proof of damages in such a case was not practical in a class action format. The authors in 7B C. Wright, A. Miller M. Kane, Federal Practice and Procedure, § 1784, at 78-82 (2d ed. 1986), found this reasoning not persuasive, saying:

Rule 23 provides more than enough flexibility and room for judicial innovation so that the question of damages can be determined through individual trials on that issue pursuant to "equitable procedures" devised by the court.

. . . .

One possibility is to try the damage issue only once, making a single award for the class, and then develop an expeditious administrative means of dividing the lump sum among the class members. This approach has been employed in judicially approved settlements under Rule 23(e) and the courts could look for guidance to some of the procedures that have been developed in that context.

. . . .

Utilizing procedures to distribute a lump sum recovery is judicially economical since it eliminates the need for separate trials on damage issues. In a case involving an extremely large plaintiff class whose members only have small individual claims, the savings in time and expense for all will be considerable. Occasionally this will be achieved at the expense of some traditional procedural safeguards, most notably jury trial. Nonetheless, simplifying the process of establishing individual claims may be the only way of making it economically feasible for class members to come forward and assert their rights and some of the procedural patterns that are considered fundamental when the litigation involves the single plaintiff and a single defendant will have to be abandoned. (Footnotes omitted).

Moreover, Professor Arthur Miller in a perceptive article published in 1979, Of Frankenstein Monsters and Shining Knights: Myth, Reality, and the "Class Action Problem," 92 Harv.L.Rev. 664, warned against the tendency in some of the decisions to disregard the effect of the changing pattern of litigation on the application of the Rule and suggested an innovative application of the Rule dealing with this changing pattern of litigation as represented by the mass torts. He wrote:

It is important in understanding the class action debate to realize that the "big case" phenomenon transcends the class action. The "big case" is an inevitable byproduct of the mass character of contemporary American society and the complexity of today's substantive regulations. It is a problem that would confront us whether or not rule 23 existed. Indeed, it is becoming increasingly obvious that the traditional notion of civil litigation as merely bilateral private dispute resolution is outmoded. Since our conception of the roles of judges and advocates is based on this traditional view, the ferocious attack on the class action may reflect anxiety over the growing challenge to the model's immutability.

. . . .

Even in its current elaborated form, rule 23 really must be thought of as a procedural skeleton requiring fleshing out by judges and lawyers experimenting with it in an ever-increasing range of circumstances and in a variety of innovative ways. . . .

The procedural complexities that can emerge under rule 23 are extraordinarily variegated in character. Class actions tend therefore to be processed in a highly individualistic fashion, making the extrapolation of general propositions from judicial opinions difficult and overall evaluation risky; procedural techniques employed in one case may be ill suited in another context. Every rule 23 decision, therefore, must be viewed through the prism of its particular facts, some of which may not even appear in the court's opinion. 92 Harv.L.Rev. at 668, 677 (footnote omitted).

Another commentator has chided appellate courts for failing to recognize the seriousness of the mass tort problem and for failing to abandon their narrow approach to the use of the class action device in the handling of mass tort litigation. Mullenix, Class Resolution of the Mass-Tort Case: A Proposed Federal Procedure Act, 64 Tex.L.Rev. 1039 (1986). She said:

Confronted with the realities of this growing, nationwide mass-tort litigation crisis, an increasing number of judges and commentators have urged a more flexible application of Rule 23 in the mass-tort context in order to better serve the interests of justice. Yet these judicial and scholarly pleas continue to fall on deaf appellate ears as the circuit courts persist in construing narrowly the Rule 23 class action procedure. 64 Tex. L.Rev. at 1043 (footnotes omitted).

It is obvious that there is a movement towards a more liberal use of Rule 23 in the mass tort context. This new attitude has found expression in a number of recent cases which recognize that the mass tort phenomenon has created a pressing "[n]ecessity . . . to change and invent," Jenkins v. Raymark Indus. Inc., 782 F.2d 468, 473 (5th Cir. 1986), with the result that, as one court has correctly concluded, "the trend has been for courts to be more receptive to use of the class action in mass tort litigation," In re School Asbestos Litigation, 789 F.2d 996, 1009 (3d Cir.), cert. denied, 479 U.S. 852, 107 S.Ct. 182, 93 L.Ed.2d 117, 479 U.S. 915, 107 S.Ct. 318, 93 L.Ed.2d 291 (1986). This approach, with its skepticism on the current validity of the reasons suggested in the earlier 1966 Advisory Committee Notes on which the Ninth Circuit had rested its limited restrictions as stated in La Mar, Green and McDonnell Douglas, was explained in Jenkins:

Courts have usually avoided class actions in the mass accident or tort setting. Because of differences between individual plaintiffs on issues of liability and defenses of liability, as well as damages, it has been feared that separate trials would overshadow the common disposition for the class (citing Advisory Notes). The courts are now being forced to rethink the alternatives and priorities by the current volume of litigation and more frequent mass disasters (citing authority). If Congress leaves us to our own devices, we may be forced to abandon repetitive hearings and arguments for each claimant's attorney to the extent enjoyed by the profession in the past. Be that as time will tell, the decision at hand is driven in one direction by all the circumstances. Judge Parker's plan is clearly superior to the alternative of repeating, hundreds of times over, the litigation of the state of the art issues with, as that experienced judge says, "days of the same witnesses, exhibits and issues from trial to trial." 782 F.2d at 473.

The decision in Jenkins involved a mass asbestos case. In all asbestos cases, a primary defense had evolved about the "state of the art" at the time of the happening of the alleged tort. Evidence on that issue, the District Court found, "would vary little as to individual plaintiffs while consuming a major part of the time required for their trials." 782 F.2d at 471. The District Court concluded that "[c]onsiderable savings, both for the litigants and for the court, could thus be gained by resolving [this `state of the art' defense] and other defense-related questions, including product identification, product defectiveness, gross negligence and punitive damages, in one class trial" Ibid. (Italics added). It accordingly "certified the class as to the common questions, ordering them resolved for the class by a class action jury. The class jury would also decide all the individual issues in the class representatives' underlying suits; individual issues of the unnamed members would be resolved later in `mini-trials' of seven to ten plaintiffs." Ibid. The District Court said further that "[a]lthough the class action jury would evaluate the culpability of defendants' conduct for a possible punitive damage award, any such damages would be awarded only after class members had won or settled their individual cases." Ibid. Such were the circumstances under which the class certification was granted by the District Court in that case. The defendants on appeal challenged the certification under (b)(3). The Court of Appeals, after observing that "[t]he purpose of class actions is to conserve `the resources of both the courts and the parties by permitting an issue potentially affecting every [class member] to be litigated in an economical fashion" Ibid. at 471, (quoting from General Telephone Co. of Southwest v. Falcon, 457 U.S. 147, 155, 102 S.Ct. 2364, 2369, 72 L.Ed.2d 740 (1982), and recognizing the innovativeness of the District Court's procedure, affirmed that Court's action, saying that "in light of the magnitude of the problem and the need for innovative approaches, we find no abuse of discretion in this court's decision to try these cases by means of a Rule 23(b)(3) class suit." 782 F.2d at 475.

The School Asbestos case was a consolidated suit instituted by various school districts in several states to recover the costs of testing and removing asbestos material from their buildings. It is another decision in which the court engaged in "rethinking" the use of class actions in mass tort suits. In that case, the District Court certified a nationwide mandatory class for punitive damages and an opt-out class ( i.e. (b)(3)) for compensatory damages. On appeal, the certification of a mandatory class for punitive damages on the record then before the Court was vacated with the statement that "we hold open the possibility of a 23(b)(1)(B) punitive damage class in more appropriate circumstances," 789 F.2d at 1008, but the certification for compensatory damages under (b)(3) was affirmed. In approving the (b)(3) certification, the Court recognized, as we have observed, that "the trend has been for courts to be more receptive to use of the class action in mass tort litigation." 789 F.2d at 1009. After taking note of the Advisory Committee Note on the mass tort case, the Court then dismissed the modern applicability of such comment thus:

Although that statement continues to be repeated in case law, ( i.e., Advisory Note to (b)(3) of Rule 23) there is growing acceptance of the notion that some mass accident situations may be good candidates for class action treatment. An airplane crash, for instance, would present the same liability questions for each passenger, although the damages would depend on individual circumstances. Determination of the liability issues in one suit may represent a substantial savings in time and resources. Even if the action thereafter "degenerates" into a series of individual damage suits, the result nevertheless works an improvement over the situation in which the same separate suits require adjudication on liability using the same evidence over and over again. See Hernandez v. Motor Vessel Skyward, 61 F.R.D. 558 (S.D. Fla. 1973).

Reassessment of the utility of the class action in the mass tort area has come about, no doubt, because courts have realized that such an action need not resolve all issues in the litigation. See Fed.R.Civ.P. 23(c)(4)(A). If economies can be achieved by use of the class device, then its application must be given serious and sympathetic consideration. 789 F.2d at 1008-09.

After expressing agreement with the reasoning in Jenkins, the court in School Asbestos further commented:

Manageability is a practical problem, one with which a District Court generally has a greater degree of expertise and familiarity than does an appellate court (omitting citation). Hence, a District Court must necessarily enjoy wide discretion, and we are not inclined to reverse a certification before the District Judge has had an opportunity to put the matter to a test. 789 F.2d at 1011.

Finally, in upholding the (b)(3) certification, the Court concluded its opinion with this statement:

We acknowledge that our reluctance to vacate the (b)(3) certification is influenced by the highly unusual nature of asbestos litigation. The District Court has demonstrated a willingness to attempt to cope with an unprecedented situation in a somewhat novel fashion, and we do not wish to foreclose an approach that might offer some possibility of improvement over the methods employed to date. Ibid.

A highly publicized example of this new thinking is also provided by In re "Agent Orange" Product Liability Litigation, 100 F.R.D. 718 (E.D.N.Y. 1983). That was an action by a class of former military servicemen or their families seeking recovery of both compensatory and punitive damages for injuries suffered by them while serving in Vietnam as a result of the use by the military forces of a certain herbicide. A primary defense asserted by the manufacturers of the herbicide was the government contract defense. After notice and hearing, the District Court certified all issues, other than punitive damages for trial and disposition, under (b)(3); it certified the punitive damage issue under (b)(1)(B). After certification, the parties submitted a proposed settlement of the action. In granting such certification, the court noted the common issues involved in the government contract defense asserted by the defendants and in the claim of causation, the immense size of the plaintiff class, and the likelihood that certification might "encourage settlement of the litigation." 100 F.R. D. at 720-21. It found that in the previous cases where certification had been denied in this type of case, the courts had relied for their action on the Comment of the Advisory Committee already quoted, and it seems to have dismissed the Comment as applicable to the "mass accident" type of mass torts but not to "a mass products liability based upon a series of discrete events." 100 F.R.D. at 722. In response to the defendants' request that certification be confined to the issue of the government contract defense, the court found such proposal was "not a workable one," that it was impossible to try that issue without litigating the issue of causation, and that the common question in the case predominated over any issues affecting individual members. 100 F.R.D. at 723. In resolving the certification issue, it emphasized the importance of certification as related to settlement of the action:

. . . the court may not ignore the real world of dispute resolution. As already noted, a classwide finding of causation may serve to resolve the claims of individual members, in a way that determinations in individual cases would not, by enhancing the possibility of settlement among the parties and with the federal government.

It accordingly certified the case on all issues other than punitive damages under (b)(3) and on punitive damages under (b)(1)(B).

On appeal, the Court of Appeals affirmed. In re Diamond Shamrock Chemicals Co., 725 F.2d 858 (2d Cir.), cert. denied, 465 U.S. 1067, 104 S.Ct. 1417, 79 L.Ed.2d 743 (1984). On the (b)(3) certification, the court found that

it seems likely that some common issues, which stem from the unique fact that the alleged damage was caused by a product sold by private manufacturers under contract to the government for use in a war, can be disposed of in a single trial. The resolution of some of these issues in defendants' favor may end the litigation entirely. Moreover, since these issues may involve extensive documentary and testimonial evidence, Chief Judge Weinstein found that obviating a retrial in countless individual cases will lead to substantial economies in the use of judicial and private resources. 725 F.2d at 860-61.

It explicated this conclusion with the comment that "there [were] substantial grounds at this stage to support [the District Court's] conclusion that the common issues predominate and that a class action is the most efficient means of adjudicating them." Ibid. at 861. It found in particular that "[t]he unique common issues take the case out of the general rule that `[a] "mass accident" resulting in injuries to numerous persons is ordinarily not appropriate for a class action because of the likelihood that significant questions, not only of damages but of liability and defenses of liability, would be present, affecting the individuals in different ways.'" Ibid., quoting the Advisory Committee Note. Later, on appeal from final judgment, the grant of certification under (b)(3) was reaffirmed, though with some expressions of skepticism of the difficulties of managing expeditiously mass tort suits, finding that "class certification was justified under Rule 23(b)(3) due to the centrality of the military contractor defense. First, this defense is common to all of the plaintiffs' cases, and thus satisfies the commonality requirement of Rule 23(a)(2)." In re "Agent Orange" Products Liability Litigation, 818 F.2d 145, 166-67 (2d Cir. 1987), cert. denied, ___ U.S. ___, 108 S.Ct. 695, 98 L.Ed.2d 648 (1988). Because of this ruling, it found no need to consider at the time the question of the punitive damage claim.

One of the most instructive cases in this connection was cited by the appellants in support of their objection to class certification, Federal Skywalk Cases. The action in that case concerned claims arising out of the Skywalk disaster in the Hyatt Regency Hotel in Kansas City, Missouri on July 17, 1981. There the District Court had certified a Rule 23(b)(1) mandatory class action on the issues of liability for compensatory damages, liability for punitive damages and amount of punitive damages. On appeal, the Court of Appeals found that the mandatory class action was barred by the Anti-Injunction Statute, 28 U.S.C. § 2283, but in so ruling, it recognized the complex issues faced by the District Court and commended the District Court's "creative efforts," but concluded that the certification could not "stand on the facts before it." Federal Skywalk Cases, 680 F.2d 1175 (8th Cir. 1982), cert. denied, 459 U.S. 988, 103 S.Ct. 342, 74 L.Ed.2d 383 (1982). However, by limiting its opinion to the facts before it, the Court of Appeals left open the possibility that some future mandatory certification in the action might be better received. The subsequent developments in the Skywalk case appear in 97 F.R.D. 380 (W.D.Mo. 1983) with a comment on those late developments in an article co-authored by Judge Wright, who had written the earlier District Court opinion, and Joseph Colussi, titled The Successful Use of the Class Action Device in the Management of the Skywalks Mass. Tort Litigation, 52 UMKC L.Rev. 141, 141-43 (1984), in which they made this interesting comment on the denouement of the Skywalk litigation:

Several commentators have recently suggested that the appellate court decision in the Skywalk mass tort case vacating certification of a mandatory class action struck the death knell for the use of the class action procedure in the management of mass tort litigation. The manner in which the Skywalks litigation was managed and resolved in the wake of that appellate court decision, however, demonstrated that the opposite is true. What many observers of the Skywalks litigation have failed to recognize is that the litigation was ultimately managed and resolved through the use of the class action procedure. After the mandatory class action was vacated by the appellate court, two nearly identical voluntary class actions were certified and settled in the federal and state trial courts. The certification and settlement of those two voluntary class actions performed the same management function and permitted the claims of all litigants to be resolved in the same equitable and efficient manner that would have resulted if the mandatory class action had not been vacated.

The experience of the Skywalks litigation has made clear that the mandatory class action device is the only procedure presently available to state and federal courts which allows for the equitable and efficient management of mass tort litigation. Though the appellate court ruling in the Skywalks litigation did not prevent the eventual equitable and efficient management of that mass tort case through the use of the class action device, the ruling unwittingly sanctioned forum shopping and inequitably parceled the burden of mass tort litigation management between state and federal trial courts. The appellate court effectively placed the full responsibility of mass tort management on its state court brethren.

It may be definitely said — respecting the language of School Asbestos, supra — that the "trend" of the authorities is clearly in the direction of a more liberal approach to the certification of the mass tort action.

XVI.

It is to be noted that a number of the decisions we have discussed involved settlements of the proposed class action. This raises the question whether class certification for settlement purposes is permissible under Rule 23. The appellants argue flatly that certification for such purposes is impermissible. However, despite this statement of appellants, certification for such purposes finds strong support in a number of various Law Review comments. Nielson, Was the 1966 Advisory Committee Right?: Suggested Revisions of Rule 23 to Allow More Frequent Use of Class Actions in Mass. Tort Litigation, 25 Harv.J. on Legislation 461, 480 (1988), for instance, recognized a "trend" in favor of such certification:

In recent years, several federal judges have explicitly recognized the effect of class certification on the likelihood of prejudgment settlement in mass-tort suits, and have apparently allowed such recognition to influence their decision to certify class actions.

Professor Transgrud, though he took a somewhat skeptical view of mass-tort certifications generally, felt that certification "as a pretrial joinder device to facilitate group settlements [was] both proper and desirable," adding

Its use is proper because Federal Rule of Civil Procedure 23 provides that the court may certify a common question class action when it will prove "superior to other available methods for the fair and efficient adjudication of the controversy." A judicially supervised and approved class action settlement, like a judicially supervised trial, is a means of hearing and determining judicially, in other words, "adjudicating," the value of claims arising from a mass tort. As a result, if conditional certification of the case as a common question class action for settlement purposes would enhance the prospects for a group settlement, then Rule 23 authorizes certification.

Transgrud, Joinder Alternatives in Mass. Tort Litigation, 70 Cornell L.Rev. 779, 835 (1985).

Recent court decisions have also spoken approvingly of the class certification of mass-tort actions for purposes of settlement in conformity with these academic comments. These decisions confirm that the promotion of settlement may well be a factor in resolving the issue of certification. Thus, in Agent Orange, Chief Judge Weinstein wrote, ( 100 F.R.D. at 723):

Finally, the court may not ignore the real world of dispute resolution. As already noted, a classwide finding of causation may serve to resolve the claims of individual members, in a way that determinations in individual cases would not, by enhancing the possibility of settlement among the parties and with the federal government.

Similarly, in In re School Asbestos Litigation, 789 F.2d at 1009, the Court said:

Concentration of individual damage suits in one forum can lead to formidable problems, but the realities of litigation should not be overlooked in theoretical musings. Most tort cases settle, and the preliminary maneuverings in litigation today are designed as much, if not more, for settlement purposes than for trial. Settlements of class actions often result in savings for all concerned.

The only federal court decision which has actually granted class certification for settlement purposes in a mass-tort action is In re Bendectin, 102 F.R.D. 239, 240 (S.D. Ohio 1984). In that case, Judge Rubin, in granting certification for settlement purposes, said:

The Bendectin litigation is but one example of massive product liability lawsuits involving large numbers of plaintiffs, protracted trials and substantial litigation costs. The traditional court system is simply unequipped to handle such [mass tort] litigation in a conventional manner without materially depleting the judicial resources available for all other litigation. It is theoretically possible to assign sufficient judicial time to hear these cases promptly but only at the cost of further delay in an already overburdened system. The cost to the parties of litigating these cases under current procedures is such that few plaintiffs could afford the expense or the delay. Justice is not served by erecting tollgates at the courthouse door.

There is a solution. The resolution of disputes does not necessarily require trial. Within the judicial authority of this Court is a means whereby the parties might be assisted in reaching a prompt and equitable disposition of the entire problem. That solution involves limited use of Rule 23 of the Federal Rules of Civil Procedure. A class certification would enable any proposed settlement to be presented to all class members and by them either accepted or rejected.

On appeal, the certification was invalidated, 749 F.2d 300 (6th Cir. 1984). The ground for denial was that the action failed to qualify for certification under (b)(1) because of the limitation in the use of the class action decision stated in McDonnell Douglas, and under the "limited fund" doctrine. The Court, however, was careful to point out that it was not determining that class certification for settlement purposes of the mass tort in that case was impermissible. To emphasize this fact, the Court declared at p. 305, n. 10:

We do note that there is precedent for the proposition that a class can be certified for settlement purposes only. See, e.g., In re Beef Industry Antitrust Litigation, 607 F.2d 167 (5th Cir. 1979), cert. denied, 452 U.S. 905, 101 S.Ct. 3029, 69 L.Ed.2d 405 (1981). The Beef Industry case involved the certification of a temporary settlement class prior to certification of a class for trial. In this case, the District Judge certified a class for settlement purposes after having rejected the same class for trial purposes. The District Judge therefore implicitly held that the standards for certifying a class are different depending on whether the class is for settlement or whether it is for trial. Because we decide the case on other grounds, we do not consider whether this holding is correct.

As Bendectin had recognized, courts in cases involving numerous parties, though not mass-tort cases, had granted class certification for settlement purposes. Judge Wisdom in In re Beef Industry Antitrust Litigation, 607 F.2d 167 (5th Cir. 1979), cert. denied, 452 U.S. 905, 101 S.Ct. 3029, 69 L.Ed.2d 405 (1981), carefully considered all angles of the question of class certification to promote settlement and, in his convincing opinion, found certification under proper circumstances to be in order. That decision has been followed in other cases, perhaps the most notable one being Weinberger v. Kendrick, 698 F.2d 61, 72-73 (2d Cir. 1982), in which Judge Friendly, speaking for the court, said:

The hallmark of Rule 23 is the flexibility it affords to the courts to utilize the class device in a particular case to best serve the ends of justice for the affected parties and to promote judicial efficiencies. Temporary settlement classes have proved to be quite useful in resolving major class action disputes. While their use may still be controversial, most Courts have recognized their utility and have authorized the parties to seek to compromise their differences, including class action issues through this means.

See also In re Mid-Atlantic Toyota Antitrust Litigation, 564 F. Supp. 1379, 1388-90 (D.Md. 1983); In re First Commodity Corp. of Boston, etc., 119 F.R.D. 301, 306-08 (D.Mass. 1987).

Though, as we have said, these cases were not mass-tort suits, there seems to be no real reason why the precedent established by the cases, just cited, should not be equally applicable to the mass-tort. If not a ground for certification per se, certainly settlement should be a factor, and an important factor, to be considered when determining certification. That is all the District Court did in this case. Its action in considering this circumstance would appear to have been appropriate.

XVII.

In summary, we take it as the lessons to be gleaned from the authorities already cited and discussed to be (a) that the "trend" is once again to give Rule 23 a liberal rather than a restrictive construction, adopting a standard of flexibility in application which will in the particular case "best serve the ends of justice for the affected parties and . . . promote judicial efficiency"; (b) that the Advisory Committee's Note suggestion that suit for damages is "not appropriate" for class certification has proved unworkable and is now increasingly disregarded; (c) that the theory that the Rule should be constrained by establishing judicially, without support in the Rule itself, limitations on its use such as were stated in La Mar, Green, and McDonnell Douglas have been outdated by the increasing phenomenon of the mass products tort action and by the growing body of recent class action decisions and comments favoring class actions in the mass tort context; (d) that, in order to promote the use of the class device and to reduce the range of disputed issues, courts should take full advantage of the provision in subsection (c)(4) permitting class treatment of separate issues in the case and, if such separate issues predominate sufficiently ( i.e., is the central issue), to certify the entire controversy as in Agent Orange; and (e) that it is "proper" in determining certification to consider whether such certification will foster settlement of the case with advantage to the parties and with great saving in judicial time and services; and (f) that the mass tort action for damages may in a proper case be appropriate for class action, either partially on in whole.

We accordingly now turn to an attempt to apply these lessons to the facts of this case.

XVIII.

The overriding fact to be observed in attempting to apply these rules for class certification here is the uniqueness of this case. It is not simply a case by a numerous group or class suing one or more tortfeasors in one action to recover for a single set of wrongful acts. This is a case in which literally hundreds of thousands of claimants are involved as proposed class parties. It is a suit that breaks down into a number of separate and distinct issues, each of which is susceptible of independent proof, analysis and resolution and several of which arise out of the same factual background, pose a like legal issue and provide a basis for a separate certification. The first of these separate issues is one for the rescission or, as it is described in the complaint herein, a "vitiation" of a 1984 agreement between Robins and Aetna on the coverage afforded Robins under Aetna's policy. The differences between Robins and Aetna which this settlement disposed of, grew out of certain extra sums allegedly due "on account of defense costs and related items" over and above the stated face amount of the contract of insurance by Aetna to Robins. The agreement which settled the difference provided that Aetna would increase coverage to the extent of several hundreds of millions of dollars beyond the face limits of the policy. The Breland plaintiffs challenge the correctness of this settlement made by Robins with Aetna. They assert that they, as the representatives of the class of Dalkon Shield claimants, were third-party beneficiaries of the policy between Aetna and Robins and that they should have been involved in the settlement but were not and that their interest and the interest of all members of the class they seek to represent were prejudiced by such settlement. They wish to reopen the settlement and to seek apparently some additional sum in settlement from Aetna, thereby creating a fund in which all Dalkon Shield claimants would have a common interest. This is a classic type of "limited fund" embraced within the authorization of subsection (b)(1)(B) of Rule 23. The District Court properly certified the issue under (b)(1)(B) and (c)(4), and we do not understand the appellants to contest this ruling.

The second issue represents the point which is critical and in dispute in this appeal. That issue poses the question whether Aetna may be subjected to liability in the Dalkon Shield litigation as a joint tortfeasor. The theory on which the Breland plaintiffs premise this right against Aetna has as its basis the claim that Aetna went beyond its duty as the insurance carrier for Robins and became an "active participant in all stages of the development, testing, promoting, and marketing of the product [Dalkon Shield]. . . ." It alleged that Aetna's conduct in this regard made it a joint tortfeasor equally liable with Robins in all the Dalkon Shield suits. This contention goes to the very heart of the plaintiffs' case here and overhangs the other issues in the litigation. It occupies the same place in this litigation as the military contractor defense did in Agent Orange and as the "state of the art" did in Jenkins. If the Dalkon Shield plaintiffs were to be unable to establish that basic issue in their favor, there would be no Dalkon Shield liability for Aetna and no trial herein would be required on the defectiveness of the Dalkon Shield. Further, there would be no dispute for resolution herein over causation or damages in these cases. It was accordingly very much in the interests of both Aetna and Dalkon Shield claimants as well as in the public interest that this issue common to every Dalkon Shield case be resolved through a class certification rather than through an interminable procession of thousands of individual cases. If that issue could be disposed of in a class certified context, the courts might well be saved the trial anew of thousands of Dalkon Shield claims.

Whether Aetna is liable as a joint tortfeasor in Dalkon Shield litigation will depend on an analysis of a common nucleus of facts that will not vary in the individual cases. The bulk of these facts, it seems apparent from those in the record, will be documentary. Most of those facts on which the determination of joint tortfeasor on Aetna's part is to be made are detailed already in the record on appeal. Assuming no certification, that legal determination will be made in each individual case by the court before whom the individual case may be tried. Absent class certification, the risk of conflicting decisions on that issue could be anticipated in practically every individual case against Aetna that would be tried. That threat of conflicting legal standards and results could well create a chaotic situation in this case. There are over 300,000 claims and possibly the same number of suits (assuming there is no certification). It would be intolerable that whether liability could exist might vary from state to state on the legal standard that each particular court, hearing a Dalkon Shield case against Aetna, would apply to this common nucleus of relevant facts. It would thus be possible that one plaintiff, suing in State A, would be permitted to recover against Aetna as a joint tortfeasor but in sister State B, another plaintiff, equally entitled to relief, would not be permitted to recover. This is the very situation to which the lead attorney for the plaintiffs on this appeal, along with his co-author, referred in his article, already cited, when he said that in the mass tort context the public cry was:

Why should hundreds, or even thousands, of trials take place in which the same issue of a defendant's liability is litigated over and over again? Why should some plaintiffs recover compensatory and punitive damages, while other plaintiffs in other trials receive much less, or nothing at all, even though the issues of liability and causation are the same?

The writer in 96 Harvard Law Review at 1154, 1155 commented in the same vein:

Two goals of tort law should be predictability and uniformity. . . . But if various factfinders reach inconsistent conclusions about the same set of facts, the defendant (and others in similar circumstances) is left without any guidance concerning the legality of its conduct, which may serve important legitimate aims. . . . But defendants and plaintiffs alike have compelling interests in consistent liability determinations, interests that can and should be addressed. Therefore, certifying (b)(1)(A) classes to promote these interests, even in cases in which there is no danger of conflicting equitable decrees, is fully compatible with the language and policy of the rule [Rule 23].

There is another important circumstance. Under Aetna's stipulation, a certification and resolution of the issue of Aetna's liability as a tortfeasor would, if the decision were adverse to Aetna, result in a resolution of all individual issues of causality and damages through a Claims Resolution Facility, obviating an endless procession of separate individual suits with varying results (some possibly in the few cases where the claimant elected for a jury trial rather than arbitration under the procedure to be followed under the Claims Resolution procedure) in suits heard in courts all over the country.

When an issue central to all the claims can be resolved by a class certification for all claims in the Dalkon Shield context without requiring a trial of the underlying Dalkon Shield liability claims and can be done with fairness to all Dalkon Shield claimants against Aetna, as it could be under the plan for the resolution of the non-common issues, it would be unfair alike to the litigants and to the societal interest in maintaining an efficient court system to refuse class certification. The likelihood of conflicting decisions and varying decisions of liability based on the standards of law applied to identical facts by different courts provide the very conditions contemplated in Rule 23(b)(1)(A) for class certification herein.

There is still another circumstance that pleads for certification of the issue of Aetna's liability as a joint tortfeasor. If Aetna is forced to go to trial on every individual Dalkon Shield case on the issue whether it is a joint tortfeasor and assuming that some of the individuals prevail both on that issue and that on the issue of recovery in a few of such cases, Aetna would, under the rule prevailing in a majority of the States be entitled to contribution by Robins. Such claim for contribution could be a contingent, unliquidated claim entitled to treatment in the Robins bankruptcy. To dispose of that claim, the court having jurisdiction over the bankruptcy would be required to delay the bankruptcy proceedings until the right to and the amount of the claim of contribution could be exactly ascertained through an "estimation" as mandated under Section 502(c) of the Bankruptcy Code.

Further, it is possible that even if this issue of Aetna's liability as a tortfeasor were resolved in favor of Aetna's liability, the burden of Dalkon Shield cases on the courts would still have been rendered more manageable by the courts because of a certification. Aetna has stipulated in the record herein that, if this case is certified as a mandatory class action under (b)(1)(A), and "if the liability issue were resolved against Aetna in Breland unfavorable to Aetna," [i]t would "agree not to litigate separately any of the non-common issues of individual medical causation and individual counts of damages. Instead these issues, if reached, would be resolved for class members through the Claims Resolution Facility", which would be coordinated with the Claims Resolution procedure to be provided in the Robins reorganization proceedings. It is interesting that this procedure for disposing of the case is not novel; it follows largely the suggestion made in 7B C. Wright, A. Miller M. Kane at 80 and 81, already discussed herein, and, as formulated by the courts in Jenkins and Skywalk II. The process begins under the Wright-Miller-Kane model with a trial "only once" of the damage issue, making a single award for the class, and "with the development of an expeditious administrative means of dividing the lump sum among the class members." Similarly, in the Robins reorganization, the procedure begins as we have seen, with a judicially determined "estimation," made on the basis of a full hearing and after arguments of counsel, of a sum which would fully compensate every qualifying Dalkon Shield claimant for her injury. To accomplish that result, the Plan of Reorganization creates a Trust which in turn was to establish a claims resolution facility. Each Dalkon Shield claimant would file her claim with the Trust or facility, which, in turn, would following some discussions, make an offer of settlement. If the claimant accepted the settlement offer, that would constitute full and complete settlement of her claim against Robins, Aetna, American Home and all others. Should she not wish to accept, she has the option of deciding her claim by arbitration or by a jury trial at her option. If the claimant opts for a jury trial, the suit shall be against the Trust, the "[v]enue shall be unchanged by the chapter 11 case" and "[a]ll claims and defenses shall be available to both sides at a trial." That was the combined procedure to be followed for resolving all Dalkon Shield claims against both Robins and Aetna, as well as any other parties involved. This procedure accords with what the Supreme Court in General Telephone Co. v. Falcon, 457 U.S. at 155, 102 S.Ct. at 2369, declared to be the purpose of class actions, especially in the mass tort context, to ensure "the resources of both the courts and the parties by permitting an issue potentially affecting every [class member] to be litigated in an economical fashion." [citation omitted] As the Court in Jenkins said, "in the light of the magnitude of the problem and the need for innovative approaches," 782 F.2d at 475, we find no abuse of discretion on the part of the District Judge in his certification herein under these circumstances.

There is a final reason why certification is warranted in this case. As a result of the conditional certification made by the District Judge, the parties in the Breland action were enabled to develop a settlement of this controversy — a settlement which has been approved by approximately 95% of all voting claimants. As Professor Transgrud, in the article already cited and quoted has correctly stated, class certification "is proper because Federal Rules of Civil Procedure 23 provides that the court may certify a common question class action when it will prove `superior to other available methods for the fair and efficient adjudication of the controversy.'" See also Schuster, Precertification of Class Actions: Will California Follow the Federal Lead, 40 Hastings L.J. 863, 873 (1989).

The appellants, however, would find the certification invalid because it fails to provide an opt-out right in favor of claimants under the procedure established. They assert that Phillips Petroleum Co. v. Shutts, 472 U.S. 797, 105 S.Ct. 2965, 86 L.Ed.2d 628 (1985), proscribes any class certification, which does not provide the class members with the right to opt out. Shutts, however, has been the subject of considerable discussion by commentators. Many of the commentators take the view that Shutts held "that a state court can bind absent class plaintiffs to a judgment for money damages (only) if it provides the absent parties the minimal procedural protections of adequate representation, notice of the action, and an opportunity to opt out of the litigation." A view which would put in jeopardy non opt-out certification. Note, Phillips Petroleum Company v. Shutts: Procedural Due Process and Absent Class Members: Minimum Contacts Is Out — Is Individual Notice In?, 13 Hastings Const. L.Q., 817, 821 (1986); Note, Phillips Petroleum Company, v. Shutts: Multistate Plaintiff Class Actions: A Definite Forum, But Is It Proper?, 19 John Marshall L.Rev. 483, 485 (1986). The most careful and thoughtful comment on the decision, however, is in Miller Crump, Jurisdiction and Choice of Law in Multistate Class Actions after Phillips Petroleum Co. v. Shutts, 96 Yale L.J. 1 (1986). The authors are not entirely certain of the effect of the decision on the mandatory class certification with its no opt-out provision. The authors said that "[t]here is no neat and logical means of resolving the question whether mandatory actions survive Shutts. The answer depends upon the view one takes of Shutts itself and of the need for mandatory classes. It also depends upon the characteristics of the particular class action." Ibid. at 52 (footnote omitted). Professor Mark Weber, in an article titled Preclusion and Procedural Due Process in Rule 23(b)(2) Class Actions, 21 Univ. of Mich.J. of Law Reform 347 (1988), indicates a situation in which a mandatory class action which did not include an opt-out provision might qualify under Shutts. He wrote:

The problem of the Rule 23(b)(2) class action is that binding absent class members without giving them notice and the right to opt out violates due process. The problem is clear under the apparently applicable standards of Phillips Petroleum Co. v. Shutts. Indeed, the only way in which the binding Rule 23(b)(2) class action might be seen not to be a violation of due process is if one follows Mathews v. Eldridge and weighs the harm to the interests of absentees with the cost of the alternative safeguard of notice and the right to opt out. 21 Mich. J. of Law Reform at 394 (footnote omitted).

Fortunately, however, Shutts — if taken as requiring an opt-out provision in any class certification, and assuming without deciding, that opt-out could not be validated here under the Mathews v. Eldridge standard — is satisfied in this case. The requirement of an opt-out provision if proven, is based on principles of due process. Chief Justice Rehnquist made this quite clear in his opinion in Shutts. In this case the Trust created by the parties for the resolution of all Class A claims on individual causation and damages does not in express terms include an opt-out provision but in effect it does. The Plan gives every such class member the right to elect to have her claim settled in a trial with all the procedural rights normally attaching to a jury trial. That is everything that an express opt-out provision could give a class member if such right is required under due process.

Due process, the courts have often declared, "is a flexible concept," intended to ensure "fundamental fairness." Walters v. Nat. Ass'n. of Radiation Survivors, 473 U.S. 305, 320, 105 S.Ct. 3180, 3189, 87 L.Ed. 2d 220 (1985). See also Mathews v. Eldridge, 424 U.S. 319, 334, 96 S.Ct. 893, 902, 47 L.Ed.2d 18 (1976). It is not a matter of semantics, not a matter of whether the talismanic word "opt-out" is used but whether the right given exists in effect. Walters, 473 U.S. at 321, 105 S.Ct. at 3189. Certainly, the procedure provided every class member here fairly meets the standard of "fundamental fairness" which due process demands. It follows that Shutts, if applicable, is fully complied with in this case.

The appellants also posit that every class certification by a District Court has been reversed by the Court of Appeals and argue that such uniformity among the authorities should induce us to deny certification in this case. They list four decisions in support of this contention. In re Bendectin Products Liability Litigation, 749 F.2d 300 (6th Cir. 1984); In re Northern District of California, Dalkon Shield I.U. D. Products Liability Litigation, 693 F.2d 847 (9th Cir. 1982), cert. denied, 459 U.S. 1171, 103 S.Ct. 817, 74 L.Ed.2d 1015 (1983); In re Federal Skywalk Cases, 680 F.2d 1175 (8th Cir.), reversing 93 F.R.D. 415 (W.D.Mo. 1982), cert. denied, 459 U.S. 988, 103 S.Ct. 342, 74 L.Ed.2d 383 (1982); and In re Temple, 851 F.2d 1269 (11th Cir. 1988). We do not find that these cases support appellants' broad statement. For instance, Temple expressly disclaimed any intention to hold that "a products liability or mass accident suit could never be the proper subject of a class action." 851 F.2d at 1273 n. 7 (emphasis in original). It then cited Dalkon Shield, 693 F.2d at 853 (noting "that mass torts caused by a single transaction, i.e., a plane crash, might be proper subject of a class action"), and Jenkins v. Raymark Indus. Inc., 782 F.2d 468, 475 (5th Cir. 1986) (noting that the single issue of "state of art" defense certifiable in non-mandatory context), 851 F.2d at 1273 n. 7, as mass tort cases in which certification had been had. Actually, class action certification was denied in Temple because of lack of commonality on the part of the proposed class members; a contention not made in this case. 851 F.2d at 1273. In Bendectin, the first case cited by appellants, the Court of Appeals expressly refused to decide whether class certification for settlement purposes was permissible under Rule 23 and decided the case on the ground that the action did not qualify under (b)(1) because of the limitation placed on the use of Rule 23(b)(1)(A) in McDonnell Douglas, already discussed, and because it did not produce sufficient proof of a limited fund. 749 F.2d at 305-06. For reasons already given, we believe McDonnell Douglas, is outdated as a proper limitation on the use of (b)(1)(A). We have already discussed Dalkon Shield and pointed out the differences between that decision and this case. Properly construed, Dalkon Shield may be said to have gone off on lack of a proper class representative, according to Agent Orange, supra. Skywalk's final end as set forth in 97 F.R.D. 380 (W.D.Mo. 1983) is not noted by the appellants. The most significant error of the appellants, however, is that, in making the broad statement that no Court of Appeals had sustained a mass tort certification, they completely disregarded School Asbestos, Jenkins, and Agent Orange, as well as the later decision in Skywalk, all of which sustained a mass tort certification.

The appellants also assert that the Breland plaintiffs are not proper and adequate class representatives. This contention is based on the claim that there are claimants who object to class certification and the settlement which the Breland plaintiffs as class representatives submitted to the District Court for approval. Appellants' argument would deny class certification if any number of members of the class, however small, objected. Class certification cannot be defeated simply because a few claimants — actually, their attorneys — feel that they would fare better by pressing as quickly as they could their individual suits than they would if the suit were to be prosecuted for the benefit of all claimants. Certainly, in the mass tort situation, there is a substantial societal interest. The burden placed on the judicial system resulting from every products liability litigation entails great public and private expense, encumbers the court calendars, and prejudices other litigants by unusual delays in their litigation. As we have already said, judges in such cases are required to try case after case where the same documentary evidence is produced, where the same expert witnesses and their tests are exhaustively developed, where there is extended direct and cross-examination, and required, as by rote, to repeat the same instructions to a jury. To permit a small number of claimants who are seeking actually to promote their own interests largely at the expense of the other claimants to frustrate a class certification in such a case is unthinkable.

And what has been said here is equally applicable to the objection to the action of the class representative in submitting a proposed settlement to the District Court for approval. That proposed settlement was submitted to all the claimants and all claimants were given notice of a hearing on the proposed settlement where any claimant could raise such objection as she might have to the settlement. That procedure was followed by the objecting claimants. At the hearing, there was extensive testimony taken with attorneys for the objecting claimants participating and being heard. After the hearing the settlement was submitted to all the claimants and an overwhelming majority (almost 95%) of the Class A claimants and almost 98% of the Class B claimants voted in favor of the settlement. Should a class certification and a class settlement be dismissed simply because a minimal number of the claimants voting objected? Must the federal courts be forced to try individually thousands of individual cases presenting the same issues of liability and causation year after year, with an intolerable burden on the court system and the judiciary and with great delay in the resolution of the individual claims, when there is a ready remedy for expeditious adjudication and fair recovery available for the full 100% of claimants through a properly constructed class action when the objections are so limited? We think not, and we dismiss this objection to the class certification herein.

The appellants also find error in the certification in this case because the District Court, in finding compliance with (b)(1), failed to indicate whether it was certifying under (1)(A) or (1)(B). This objection is without merit. As Wright, Miller Kane points out, it is unimportant whether certification is under (A) or (B) "inasmuch as actions under clause (A) and clause (B) are treated in the same manner for purposes of the other portions of Rule 23 and nothing turns on which provision is held controlling." 7A C. Wright, A. Miller M. Kane, Federal Practice and Procedure § 1772, at 424-25 (2d ed. 1986).

In conclusion, we affirm the class certification herein. It is clear from the record that the determinative, critical issue in this case was whether Aetna was liable as a joint tortfeasor; it overshadows all other issues. Just as the military defense was central to the case in Agent Orange, so the question whether Aetna was a joint tortfeasor here was the critical issue common to all the cases against Aetna, and one which, if not established, would dispose of the entire litigation. See 818 F.2d at 166-67. Such an issue satisfies the commonality requirement of Rule 23(a)(2), as well as the other prerequisites of subsection (a). Moreover, the action qualifies under subsection (b)(1)(A) as well as (3). Manifestly the class action as a method of adjudicating that overshadowing issue is superior to any alternative form of adjudication. Further, the resolution of this issue, if required to be made in the thousands of individual suits in different courts, would involve the risk of varying and contradictory legal decisions on what was precisely the same factual record. The result would be chaotic. This is a situation which would unquestionably qualify under (b)(1)(A). It is of no moment that this action qualifies under (b)(3) or that the District Court in this case may have certified this action under (b)(3) instead of (b)(1). In fact, it is "very common for an action to fall under Rule 23(b)(1) and one of the other categories in Rule 23(b)." 7A C. Wright, A. Miller M. Kane, supra, at 425. Were we not satisfied that there was authority under (b)(1)(A) for certification, we could, it is true, affirm the decision of the District Court as permissible under (b)(3). We realize that it was under (b)(3) that Jenkins and School Asbestos made their class certification, and the District Judge could have followed those decisions. It is, however, of no moment that the District Judge did not grant certification on this ground; it would be appropriate for us to affirm the certification herein as a (b)(3) since the decision of the court below can be affirmed on any ground supported by the record, whether raised below or by the parties in the proceedings before us, if one concluded that this action could not be certified under (b)(1). Thigpen v. Roberts, 468 U.S. 27, 29-30, 104 S.Ct. 2916, 2918, 82 L.Ed.2d 23 (1984); Schweiker v. Hogan, 457 U.S. 569, 585, 102 S.Ct. 2597, 2607, 73 L.Ed.2d 227 (1982); Stern v. Merrill, Lynch, Pierce, Fenner Smith, 603 F.2d 1073, 1093 (4th Cir. 1979); Blackwelder v. Millman, 522 F.2d 766, 771 (4th Cir. 1975); Charley's Taxi Radio Dispatch v. SIDA of Hawaii, 810 F.2d 869, 874 (9th Cir. 1987); Colorado Flying Academy, Inc. v. United States, 724 F.2d 871, 880 (10th Cir. 1984), cert. denied, 476 U.S. 1182, 106 S.Ct. 2915, 91 L.Ed.2d 544 (1986). The basic difference between a mandatory certification under (b)(1)(A) and a certification under (b)(3) is that there is no opt-out requirement on the former but there is in the latter. As we have already pointed out, there is in practice, though not in express language, an opt-out provided by the right of a dissatisfied class claimant to choose a jury trial for resolution of her individual causation and damages claims. Accordingly, while the District Court might have certified this case under (b)(3), it acted in accordance with the general rule applicable in such a situation to certify under (b)(1)(A) ( See C. Wright, A. Miller, M. Kane, supra, § 1772, at 425, and other authorities cited on page 728 herein). For reasons already given, we are of opinion that the District Judge had the right to base his certification on (b)(1). We accordingly affirm the class certification as made herein by the District Court.

We turn now to consider the appeal from the order of the District Court approving the settlement herein.

XIX.

Under Flinn v. FMC Corporation, 528 F.2d 1169 (4th Cir. 1975), cert. denied, 424 U.S. 967, 96 S.Ct. 1462, 47 L.Ed.2d 734 (1976), a case cited by both parties as the guide for our decision on this phase of the appeal, the review of the decision of a District Court approving a proposed settlement of a class action is one for "clear abuse of discretion." 528 F.2d at 1172. Flinn declares that the most important fact to be considered in conducting such a review is "whether the trial court gave proper consideration to the strength of the plaintiffs' claim on the merits." Ibid. The reviewing court should consider in evaluating whether the trial court had given proper consideration to the strength of the plaintiffs case "the extent of discovery that [had] taken place, the stage of the proceedings, the want of collusion in the settlement, and the expense of counsel who [had] represented the plaintiffs in the negotiation" in the record before the trial court. Ibid. at 1173. The trial court should also give proper consideration to the attitude of the class members themselves to the settlement. It is not, however, to be assumed that a settlement is "unfair or unreasonable because a large number of class members oppose it" or that the settlement "may only amount to a fraction of the potential recovery." Ibid. It would have been in order for the trial court to grant to any objector to the settlement the right to be heard, to cross-examine witnesses or to submit evidence, but the trial court may limit the proceedings to "whatever is necessary to aid in reaching an informed, just and reasoned decision," Ibid. and specifically the trial court "should not `turn the settlement hearing into a trial or a rehearsal of the trial.'" Ultimately, the rule to be applied is that

[s]o long as the record before it is adequate to reach "an intelligent and objective opinion of the probabilities of ultimate success should the claim be litigated," and "form an educated estimate of the complexity, expense and likely duration of such litigation, . . . and all other factors relevant to a full and fair assessment of the wisdom of the proposed compromise," Ibid.

If that procedure is followed, a trial court has acted properly in determining whether to approve the proposed settlement or not. We think that the District Judge in this case followed those principles in reaching his decision. Under such circumstances, he did not abuse his discretion in approving the settlement.

It is important to recognize that this case is closely tied in with the bankruptcy of Robins. Both this case and the Robins bankruptcy were before the same District Judge. Under the proposed settlement, the resolution of individual causation and damages for Dalkon Shield claimants seeking recovery from Robins and Aetna is through the same mechanism. The Robins Plan of Reorganization and the Breland settlement are intended to provide full payment of all compensatory damages suffered by all Dalkon Shield claimants who have properly filed claims. The Plan creates a Trust Fund which will be sufficient to achieve this goal. Neither the appellants nor anyone else has asserted that such fund is insufficient. When the parties have funded the trust, Robins, Aetna and any successor corporations are relieved of any further liability, and this is proper since Dalkon Shield claimants will be paid by the Trust the full value of their claims. Aetna has made a minimal contribution to that fund. But that is not all that Aetna has done to assure the full payment of all Dalkon Shield claimants: It has provided an extra fund of $325 million to protect the Trust against any deficiency that may occur in the Trust Fund. That undertaking of Aetna is the "icing on the cake" for the Dalkon Shield claimants, who, when the Robins Plan of Reorganization and the Breland settlement are approved, can begin promptly to collect compensatory damages for their injuries. The Plan of Reorganization and the Breland settlement are interdependent. Failure of approval of either the Plan of Reorganization or the Breland settlement would derail hopelessly the carefully negotiated and crafted Plan and Settlement and leave the Dalkon Shield claimants to the vagaries, expenses and delay of further extended litigation. It is understandable that, under these circumstances, about 95% of the claimants who voted on the Settlement voted in favor. Such a strong showing of support for the Plan by those most interested should not be lightly disregarded.

Of course, this expression of the feelings of claimants cannot relieve the trial judge of his obligation to review and make his own informed and reasoned judgment on the fairness of the settlement. The most important consideration, as declared by Flinn, 528 F.2d at 1173, in reviewing a class action settlement, is whether the District Court gave proper consideration to the strength of the Breland claim on the merits. The District Judge, in reaching his decision in this case, had before him a record that had been accumulated over many years. He had the extensive discovery conducted before the Judicial Panel on Multi-District Litigation over a period of years. He had available the records developed as a result of the discovery conducted in the actions pending before Judge Lord in the District of Minnesota, as well as the investigations and reports of the two masters appointed by Judge Lord. Counsel for the Breland plaintiffs had thoroughly examined in Hartford Connecticut, the files of Aetna, and in Richmond Virginia, Robins' files. They had made copies of the relevant items in such files and collated them in various volumes. These examinations by the Breland plaintiffs covered a period of several months and entailed the services of a team of lawyers numbering at times as many as seven or ten. All this was brought to the attention of the District Court either by the filing of the requisite documents or papers themselves or by the testimony or representations during argument by Breland counsel. Some forty-two depositions were included in the record. The trial judge received exhaustive briefs and heard at length counsel on the legal aspects of the Breland plaintiffs' claim. He took the testimony of a number of witnesses in open court. The conclusion seems inescapable, in our opinion, that the District Judge conducted a most exacting inquiry into the strength of the Breland plaintiffs' case and stated an informed and reasoned conclusion on that issue.

It is the appellants' position, though, that the record will show that counsel for the class did not adequately investigate this case or conduct appropriate discovery. From the discovery already conducted before this suit was brought and the extensive investigation by counsel after this suit was instituted as already detailed, it appears to us, as it apparently did to the trial judge, the Breland Court had covered thoroughly the facts involved herein. The report of the investigations by class representatives' counsel was set forth in a "Memorandum Settlement" submitted by Mr. Friedberg and Mr. Meshbesher as counsel for the Breland plaintiffs. Later Mr. Friedberg and other counsel engaged with him in developing the Breland case and in formulation on behalf of the Breland class the settlement, testified in person and were extensively cross-examined by counsel for the appellants. In addition, depositions of other counsel were submitted in the record. All this supported the statements in the Memorandum submitted in support of the Settlement. Counsel for the Breland plaintiffs aggressively and at great expense appear to have investigated every possible avenue for relevant evidence that might support the Breland suit.

What we have discussed relates to the relevant factual record which counsel for the Breland plaintiffs zealously developed and analyzed and which was before the District Judge. We now look to the legal basis for the Breland suit against Aetna. There was no evidence that Aetna had any connection with the manufacture, labeling or marketing of the Dalkon Shield by Robins as developed either by Judge Lord's masters or by the Breland counsel (or, for that matter, the appellants). Nor is there any evidence that Aetna took any part in the decision not to recall the Dalkon Shield in 1974; that decision was made by Robins, apparently on the advice of its counsel. There was no evidence of the making of false misrepresentation of any kind to Dalkon Shield claimants by Aetna that involved any prejudice to such claimants and there was thus no basis for a fraud action against Aetna. This case is accordingly clearly distinguishable from West v. Western Cas. and Sur. Co., 846 F.2d 387 (7th Cir. 1988), in which the agent of the insurance carrier was found to have induced the plaintiff West by false representation to delay filing a suit which could have been prejudicial to the interests of the insurer. Aetna had no dealings with any Dalkon Shield claimants, assumed no duty to such claimants and could not be liable to Dalkon Shield claimants under accepted insurance law. This was the very ground for the decision in Bast and Campbell discussed on pages 715 and 716 herein, and similar cases. The District Court in this case reached the same conclusion.

The only actions as charged against Aetna in this regard related to conduct of counsel paid by Aetna, as it was required to do under its contract with Robins, to represent the latter in certain Dalkon Shield litigation. This contention raises the question whether an insurance carrier in such a case can be regarded in law to be the employer of the counsel, for whose actions the carrier may be liable under respondeat superior. It is universally declared that such counsel represents the insured and not the insurer. Repeated opinions issued by the American Bar Association [ABA], as illustrated by ABA Comm. on Ethics and Professional Responsibility, Informal Opinion 1476 (1981) declare: "When a liability insurer retains a lawyer to defend an insured, the insured is the lawyer's client." See also the following opinions in ABA/BNA Lawyers' Manual on Professional Conduct (1984): Connecticut, Informal Opinion 83-5, at 801:2059; Delaware Opinion 1981-1 at 801:2201; Michigan Opinion CI-866 at 801:4856. See also Point Pleasant Canoe Rental v. Tinicum TP., 110 F.R.D. 166, 170 (E.D.Pa. 1986); Gibson v. Western Fire Ins. Co., 210 Mont. 267, 682 P.2d 725, 736 (1984). The acts of counsel retained by Aetna to defend its insured Robins cannot be imputed to Aetna under any rule of respondeat superior.

It appears at another point in the appellants' catalogue of objections to the settlement to be the position of the appellants that Aetna knew of the deficiencies in the Dalkon Shield and failed to tell claimants of such. They would characterize this failure to be an actionable fraud. Such an argument assumes that an insurance carrier which, in reviewing with its insured a claim against the latter learns that its insured was at fault in some particular, owes a duty to advise the claimant of such dereliction on the part of its insured. Obviously, the carrier has no such duty and no court, so far as we know, has ever so held. See Nipper v. California Automobile Assigned Risk Plan, 19 Cal.3d 35, 136 Cal.Rptr. 854, 560 P.2d 743, 749 (1977) ("[w]e are aware of [no authority] which suggests that an insurer . . . either stands in a special relationship with the applicant or his potential victims, or alternately owes any affirmative duty of inquiry or disclosure regarding the applicant.")

The appellants, in their brief, refer critically to the alleged fact that certain damaging files of Robins were either destroyed or were planned to be destroyed. There is, however, no evidence, either direct or circumstantial, in the record to indicate that Aetna knew of or participated in such action. When it later did hear that there had been some discussion of the possible destruction of certain Dalkon Shield records by house counsel, it is apparent, from the record on which the appellants base this assertion, that Aetna promptly demanded as a condition of renewing its policy that the Robins official who had allegedly proposed such plan be removed from any connection with the Dalkon Shield files in litigation. It is thus plain that Aetna neither participated nor did it later do any act to condone such proposed or actual destruction of records. This record is simply devoid of any evidence of any wrongful conduct on Aetna's part or participation in this incident on the part of Robins (assuming, without deciding that there was such an incident).

Nor did the appellants offer any evidence or point to any fact in the record which established or indicated collusion between counsel for the Breland plaintiffs and Aetna. Breland counsel aggressively pursued this case. They did so at great expense. After this exhaustive review of the Robins and Aetna records, it is obvious that they recognized they had found no "smoking gun" but it was not because they had not tried relentlessly to discover one. The negotiations for settlement that later developed appear to have been conducted on the part of Breland counsel with complete fidelity to their responsibility.

Finally, appellants discount the settlement as a "nuisance settlement." It was a settlement that involved a potential liability on Aetna's part of $425 million. Considering the expense of defending individually all the Dalkon Shield claims, the settlement figure might be thought of as the "nuisance" value of the Breland plaintiffs' claim to Aetna which had to bear the expense of such litigation, but this in no way taints the settlement, when the record indicates clearly that the Breland plaintiffs had but a tenuous, if at all, chance of success on this claim. Further, the settlement, coupled with that offered in the Robins Reorganization Plan, assured every Breland plaintiff of full payment of her proven claim, with an added contingent provision supplied by Aetna to protect against any possible deficiency in the Trust. Moreover, a settlement involving the acceptance of a possible liability of $425 million is not an insignificant settlement. Moreover, in every settlement, the amount is directly related to the strength of the plaintiffs' case. This is indicated by the settlement of $1,800 million of claims in the San Juan hotel fire for $100 million. See note 41. It is difficult to see what more the Breland action, if continued, could have availed the Dalkon Shield claimants. Without further elaboration, we hold the District Court did not err in approving the settlement of the class action.

The judgments of the District Court granting class certification and approving the settlement are accordingly

AFFIRMED.

. . . . .

The historical background and policy of section 502(c) support a reading that would permit the bankruptcy court to estimate only the aggregate of claims. The old Bankruptcy Act permitted courts to decline to allow claims "not capable of liquidation or of reasonable estimation" or whose liquidation or estimation "would unduly delay the administration of the estate or any proceeding under this Act." The new requirement that the court estimate all claims is one of many reforms in the Bankruptcy Code that are intended to implement the Code's broad policy of affording the debtor the most complete relief and the freshest start feasible by disposing of all possible claims during the bankruptcy proceeding. Congress wished to eliminate the possibility that after the completion of reorganization the debtor would still be faced with the uncertainty of contingent debts that could ruin the financial stability achieved in the reorganization proceedings. Congress' goals would be achieved equally well by assigning a dollar value to the whole of the asbestos plaintiffs' claims as by assigning a dollar value to each individual claim. In either case, Manville would exit from the reorganization certain of all of its liabilities and able to carry on its business without the fear that pending or future asbestos-related claims would endanger its financial condition. (Italics added).