No. 96-270


on writ of certiorari to the united states court of appeals for the third circuit

[June 25, 1997]

Justice Breyer, with whom Justice Stevens joins, concurring in part and dissenting in part.

Although I agree with the Court's basic holding that "settlement is relevant to a class certification," ante, at 25, I find several problems in its approach that lead me to a different conclusion. First, I believe that the need for settlement in this mass tort case, with hundreds of thousands of lawsuits, is greater than the Court's opinion suggests. Second, I would give more weight than would the majority to settlement related issues for purposes of determining whether common issues predominate. Third, I am uncertain about the Court's determination of adequacy of representation, and do not believe it appropriate for this Court to second guess the District Court on the matter without first having the Court of Appeals consider it. Fourth, I am uncertain about the tenor of an opinion that seems to suggest the settlement is unfair. And fifth, in the absence of further review by the Court of Appeals, I cannot accept the majority's suggestions that "notice" is inadequate.

These difficulties flow from the majority's review of what are highly fact based, complex, and difficult matters, matters that are inappropriate for initial review before this Court. The law gives broad leeway to district courts in making class certification decisions, and their judgments are to be reviewed by the Court of Appeals only for abuse of discretion. See Califano v. Yamasaki, 442 U.S. 682, 703 (1979). Indeed, the District Court's certification decision rests upon more than 300 findings of fact reached after five weeks of comprehensive hearings. Accordingly, I do not believe that we should in effect set aside the findings of the District Court. That court is far more familiar with the issues and litigants than is a court of appeals or are we, and therefore has "broad power and discretion . . . with respect to matters involving the certification" of class actions. Reiter v. Sonotone Corp., 442 U.S. 330, 345 (1979); cf. Cooter & Gell v. Hartmarx Corp., 496 U.S. 384, 402 (1990) (district court better situated to make fact dependent legal determinations in Rule 11 context).

I do not believe that we can rely upon the Court of Appeals' review of the District Court record, for that review, and its ultimate conclusions, are infected by a legal error. E.g., Georgine v. Amchem Products, Inc., 83 F. 3d 610, 626 (CA3 1996) (holding that "considered as a litigation class," the class cannot meet Rule 23's requirements) (emphasis added). There is no evidence that the Court of Appeals at any point considered the settlement as something that would help the class meet Rule 23. I find, moreover, the fact related issues presented here sufficiently close to warrant further detailed appellate court review under the correct legal standard. Cf. Reno v. Bossier Parish School Bd., 520 U. S. ___ (1997) (slip op., at 14). And I shall briefly explain why this is so.

First, I believe the majority understates the importance of settlement in this case. Between 13 and 21 million workers have been exposed to asbestos in the workplace--over the past 40 or 50 years--but the most severe instances of such exposure probably occurred three or four decades ago. See Report of The Judicial Conference Ad Hoc Committee on Asbestos Litigation, pp. 6-7 (Mar. 1991) (Judicial Conference Report); App. 781-782, 801; B. Castleman, Asbestos: Medical and Legal Aspects 787-788 (4th ed. 1996). This exposure has led to several hundred thousand lawsuits, about 15% of which involved claims for cancer and about 30% for asbestosis. See In re Joint Eastern and Southern Dist. Asbestos Litigation, 129 B. R. 710, 936-937 (E and SD N. Y. 1991) (Joint Litigation). About half of the suits have involved claims for pleural thickening and plaques--the harmfulness of which is apparently controversial. (One expert below testified that they "don't transform into cancer" and are not "predictor[s] of future disease," App. 781.) Some of those who suffer from the most serious injuries, however, have received little or no compensation. In re School Asbestos Litigation, 789 F. 2d 996, 1000 (CA3 1986); see also Edley & Weiler, Asbestos: A Multi Billion Dollar Crisis, 30 Harv. J. Legis. 383, 384, 393 (1993) ("[U]p to one half of asbestos claims are now being filed by people who have little or no physical impairment. Many of these claims produce substantial payments (and substantial costs) even though the individual litigants will never become impaired"). These lawsuits have taken up more than 6% of all federal civil filings in one recent year, and are subject to a delay that is twice that of other civil suits. Judicial Conference Report 7, 10-11.

Delays, high costs, and a random pattern of noncompensation led the Judicial Conference Ad Hoc Committee on Asbestos Litigation to transfer all federal asbestos personal injury cases to the Eastern District of Pennsylvania in an effort to bring about a fair and comprehensive settlement. It is worth considering a few of the Committee's comments. See Judicial Conference Report 2 (" `Decisions concerning thousands of deaths, millions of injuries, and billions of dollars are entangled in a litigation system whose strengths have increasingly been overshadowed by its weaknesses.' The ensuing five years have seen the picture worsen: increased filings, larger backlogs, higher costs, more bankruptcies and poorer prospects that judgments--if ever obtained--can be collected' ") (quoting Rand Corporation Institute for Civil Justice); id., at 13 ("The transaction costs associated with asbestos litigation are an unconscionable burden on the victims of asbestos disease," and citing Rand finding that "of each asbestos litigation dollar, 61 cents is consumed in transaction costs . . . . Only 39 cents were paid to the asbestos victims"); id., at 12 ("Delays also can increase transaction costs, especially the attorneys' fees paid by defendants at hourly rates. These costs reduce either the insurance fund or the company's assets, thereby reducing the funds available to pay pending and future claimants. By the end of the trial phase in [one case], at least seven defendants had declared bankruptcy (as a result of asbestos claims generally"); see also J. Weinstein, Individual Justice in Mass Tort Litigation 155 (1995); Edley & Weiler, supra, at 389-395.

Although the transfer of the federal asbestos cases did not produce a general settlement, it was intertwined with and led to a lengthy year long negotiation between the co chairs of the Plaintiff's Multi District Litigation Steering Committee (elected by the Plaintiff's Committee Members and approved by the District Court) and the 20 asbestos defendants who are before us here. Georgine v. Amchem Products, Inc., 157 F. R. D. 246, 266-267, (ED Pa. 1994); App. 660-662. These "protracted and vigorous" negotiations led to the present partial settlement, which will pay an estimated $1.3 billion and compensate perhaps 100,000 class members in the first 10 years. 157 F. R. D., at 268, 287. "The negotiations included a substantial exchange of information" between class counsel and the 20 defendant companies, including-confidential data" showing the defendants' historical settlement averages, numbers of claims filed and settled, and insurance resources. Id., at 267. "Virtually no provision" of the settlement "was not the subject of significant negotiation," and the settlement terms "changed substantially" during the negotiations. Ibid. In the end, the negotiations produced a settlement that, the District Court determined based on its detailed review of the process, was "the result of arms-length adversarial negotiations by extraordinarily competent and experienced attorneys." Id., at 335.

The District Court, when approving the settlement, concluded that it improved the plaintiffs' chances of compensation and reduced total legal fees and other transaction costs by a significant amount. Under the previous system, according to the court, "[t]he sickest of victims often go uncompensated for years while valuable funds go to others who remain unimpaired by their mild asbestos disease." Ibid. The court believed the settlement would create a compensation system that would make more money available for plaintiffs who later develop serious illnesses.

I mention this matter because it suggests that the settlement before us is unusual in terms of its importance, both to many potential plaintiffs and to defendants, and with respect to the time, effort, and expenditure that it reflects. All of which leads me to be reluctant to set aside the District Court's findings without more assurance than I have that they are wrong. I cannot obtain that assurance through comprehensive review of the record because that is properly the job of the Court of Appeals and that court, understandably, but as we now hold, mistakenly, believed that settlement was not a relevant (and, as I would say, important) consideration.

Second, the majority, in reviewing the District Court's determination that common "issues of fact and law predominate," says that the predominance "inquiry trains on the legal or factual questions that qualify each class member's case as a genuine controversy, questions that preexist any settlement." Ante, at 28 (footnote omitted). I find it difficult to interpret this sentence in a way that could lead me to the majority's conclusion. If the majority means that these pre-settlement questions are what matters, then how does it reconcile its statement with its basic conclusion that "settlement is relevant" to class certification, or with the numerous lower court authority that says that settlement is not only relevant, but important? See, e. g., In re A. H. Robins Co., 880 F. 2d 709, 740 (CA4), cert. denied sub nom. Anderson v. Aetna Casualty & Surety Co., 493 U.S. 959 (1989); In re Beef Industry Antitrust Litigation, 607 F. 2d 167, 177-178 (CA5 1979), cert. denied sub nom. Iowa Beef Processors, Inc. v. Meat Price Investigators Assn., 452 U.S. 905 (1981); 2 H. Newberg & A. Conte, Newberg on Class Actions §11.27, pp. 11-54 to 11-55 (3d ed. 1992).

Nor do I understand how one could decide whether common questions "predominate" in the abstract-- without looking at what is likely to be at issue in the proceedings that will ensue, namely, the settlement. Every group of human beings, after all, has some features in common, and some that differ. How can a court make a contextual judgment of the sort that Rule 23 requires without looking to what proceedings will follow? Such guideposts help it decide whether, in light of common concerns and differences, certification will achieve Rule 23's basic objective--"economies of time, effort, and expense." Advisory Committee's Notes on Fed. Rule Civ. Proc. 23(b)(3), 28 U. S. C. App., p. 697. As this Court has previously observed, "sometimes it may be necessary for the court to probe behind the pleadings before coming to rest on the certification question." General Telephone Co. of Southwest v. Falcon, 457 U.S. 147, 160 (1982); see also C. Wright, A.Miller, & M. Kane, 7B Federal Practice and Procedure §1785, p. 107, and n. 34 (1986). I am not saying that the "settlement counts only one way." Ante, at 26 n. 16. Rather, the settlement may simply "add a great deal of information to the court's inquiry and will often expose diverging interests or common issues that were not evident or clear from the complaint" and courts "can and should" look to it to enhance the "ability . . . to make informed certification decisions." In re Asbestos, 90 F. 3d 963, 975 (CA5 1996).

The majority may mean that the District Court gave too much weight to the settlement. But I am not certain how it can reach that conclusion. It cannot rely upon the Court of Appeals, for that court gave no positive weight at all to the settlement. Nor can it say that the District Court relied solely on "a common interest in a fair compromise," ante, at 29, for the District Court did not do so. Rather, it found the settlement relevant because it explained the importance of the class plaintiffs' common features and common interests. The court found predominance in part because:

"The members of the class have all been exposed to asbestos products supplied by the defendants and all share an interest in receiving prompt and fair compensation for their claims, while minimizing the risks and transaction costs inherent in the asbestos litigation process as it occurs presently in the tort system." 157 F. R. D., at 316.

The settlement is relevant because it means that these common features and interests are likely to be important in the proceeding that would ensue--a proceeding that would focus primarily upon whether or not the proposed settlement fairly and properly satisfied the interests class members had in common. That is to say, the settlement underscored the importance of (a) the common fact of exposure, (b) the common interest in receiving some compensation for certain rather than running a strong risk of no compensation, and (c) the common interest in avoiding large legal fees, other transaction costs, and delays. Ibid.

Of course, as the majority points out, there are also important differences among class members. Different plaintiffs were exposed to different products for different times; each has a distinct medical history and a different history of smoking; and many cases arise under the laws of different States. The relevant question, however, is how much these differences matter in respect to the legal proceedings that lie ahead. Many, if not all, toxic tort class actions involve plaintiffs with such differences. And the differences in state law are of diminished importance in respect to a proposed settlement in which the defendants have waived all defenses and agreed to compensate all those who were injured. Id., at 292.

These differences might warrant subclasses, though subclasses can have problems of their own. "There can be a cost in creating more distinct subgroups, each with its own representation. . . . [T]he more subclasses created, the more severe conflicts bubble to the surface and inhibit settlement. . . . The resources of defendants and, ultimately, the community must not be exhausted by protracted litigation." Weinstein, Individual Justice in Mass Tort Litigation, at 66. Or these differences may be too serious to permit an effort at group settlement. This kind of determination, as I have said, is one that the law commits to the discretion of the district court--reviewable for abuse of discretion by a court of appeals. I believe that we are far too distant from the litigation itself to reweigh the fact specific Rule 23 determinations and to find them erroneous without the benefit of the Court of Appeals first having restudied the matter with today's legal standard in mind.

Third, the majority concludes that the "representative parties" will not "fairly and adequately protect the interests of the class." Rule 23(a)(4). It finds a serious conflict between plaintiffs who are now injured and those who may be injured in the future because "for the currently injured, the critical goal is generous immediate payments," a goal that "tugs against the interest of exposure only plaintiffs in ensuring an ample, inflation protected fund for the future." Ante, at 32.

I agree that there is a serious problem, but it is a problem that often exists in toxic tort cases. See Weinstein, supra, at 64 (noting that conflict "between present and future claimants" "is almost always present in some form in mass tort cases because long latency periods are needed to discover injuries"); see also Judicial Conference Report 34-35 ("Because many of the defendants in these cases have limited assets that may be called upon to satisfy the judgments obtained under current common tort rules and remedies, there is a `real and present danger that the available assets will be exhausted before those later victims can seek compensation to which they are entitled' ") (citation omitted). And it is a problem that potentially exists whenever a single defendant injures several plaintiffs, for a settling plaintiff leaves fewer assets available for the others. With class actions, at least, plaintiffs have the consolation that a district court, thoroughly familiar with the facts, is charged with the responsibility of ensuring that the interests of no class members are sacrificed.

But this Court cannot easily safeguard such interests through review of a cold record. "What constitutes adequate representation is a question of fact that depends on the circumstances of each case." Wright, Miller, & Kane, 7A Federal Practice and Procedure, §1765, at 271. That is particularly so when, as here, there is an unusual baseline, namely, the " `real and present danger' " described by the Judicial Conference Report above. The majority's use of the lack of an inflation adjustment as evidence of inadequacy of representation for future plaintiffs, ante, at 32-33, is one example of this difficulty. An inflation adjustment might not be as valuable as the majority assumes if most plaintiffs are old and not worried about receiving compensation decades from now. There are, of course, strong arguments as to its value. But that disagreement is one that this Court is poorly situated to resolve.

Further, certain details of the settlement that are not discussed in the majority opinion suggest that the settlement may be of greater benefit to future plaintiffs than the majority suggests. The District Court concluded that future plaintiffs receive a "significant value" from the settlement due to variety of its items that benefit future plaintiffs, such as: (1) tolling the statute of limitations so that class members "will no longer be forced to file premature lawsuits or risk their claims being time-barred"; (2) waiver of defenses to liability; (3) payment of claims, if and when members become sick, pursuant to the settlement's compensation standards, which avoids "the uncertainties, long delays and high transaction costs [including attorney's fees] of the tort system"; (4) "some assurance that there will be funds available if and when they get sick," based on the finding that each defendant "has shown an ability to fund the payment of all qualifying claims" under the settlement; and (5) the right to additional compensation if cancer develops (many settlements for plaintiffs with noncancerous conditions bar such additional claims). 157 F. R. D., at 292. For these reasons, and others, the District Court found that the distinction between present and future plaintiffs was "illusory." 157 F. R. D., at 317-318.

I do not know whether or not the benefits are more or less valuable than an inflation adjustment. But I can certainly recognize an argument that they are. (To choose one more brief illustration, the majority chastises the settlement for extinguishing loss of consortium claims, ante, at 33, 34, but does not note that, as the District Court found, the "defendants' historical [settlement] averages, upon which the compensation values are based, include payments for loss of consortium claims, and, accordingly, the Compensation Schedule is not unfair for this ascribed reason," 157 F. R. D., at 278.) The difficulties inherent in both knowing and understanding the vast number of relevant individual fact based determinations here counsel heavily in favor of deference to district court decisionmaking in Rule 23 decisions. Or, at the least, making certain that appellate court review has taken place with the correct standard in mind.

Fourth, I am more agnostic than is the majority about the basic fairness of the settlement. Ante, at 31-34. The District Court's conclusions rested upon complicated factual findings that are not easily cast aside. It is helpful to consider some of them, such as its determination that the settlement provided "fair compensation . . . while reducing the delays and transaction costs endemic to the asbestos litigation process" and that "the proposed class action settlement is superior to other available methods for the fair and efficient resolution of the asbestos-related personal injury claims of class members." 157 F. R. D., at 316 (citation omitted); see also id., at 335 ("The inadequate tort system has demonstrated that the lawyers are well paid for their services but the victims are not receiving speedy and reasonably inexpensive resolution of their claims. Rather, the victims' recoveries are delayed, excessively reduced by transaction costs and relegated to the impersonal group trials and mass consolidations. The sickest of victims often go uncompensated for years while valuable funds go to others who remain unimpaired by their mild asbestos disease. Indeed, [these] unimpaired victims have, in many states, been forced to assert their claims prematurely or risk giving up all rights to future compensation for any future lung cancer or mesothelioma. The plan which this Court approves today will correct that unfair result for the class members and the . . . defendants"); id., at 279, 280 (settlement "will result in less delay for asbestos claimants than that experienced in the present tort system" and will "result in the CCR defendants paying more claims, at a faster rate, than they have ever paid before"); id., at 292; Edley & Weiler, 30 Harv. J. Legis., at 405, 407 (finding that "[t]here are several reasons to believe that this settlement secures important gains for both sides" and that they "firmly endorse the fairness and adequacy of this settlement"). Indeed, the settlement has been endorsed as fair and reasonable by the AFL-CIO (and its Building and Construction Trades Department), which represents a " `substantial percentage' " of class members, 157 F. R. D., at 325, and which has a role in monitoring implementation of the settlement, id., at 285. I do not intend to pass judgment upon the settlement's fairness, but I do believe that these matters would have to be explored in far greater depth before I could reach a conclusion about fairness. And that task, as I have said, is one for the Court of Appeals.

Finally, I believe it is up to the District Court, rather than this Court, to review the legal sufficiency of notice to members of the class. The District Court found that the plan to provide notice was implemented at a cost of millions of dollars and included hundreds of thousands of individual notices, a wide-ranging television and print campaign, and significant additional efforts by 35 international and national unions to notify their members. 157 F. R. D., at 312-313, 336. Every notice emphasized that an individual did not currently have to be sick to be a class member. And in the end, the District Court was "confident" that Rule 23 and due process requirements were satisfied because, as a result of this "extensive and expensive notice procedure," "over six million" individuals "received actual notice materials," and "millions more" were reached by the media campaign. Id., at 312, 333, 336. Although the majority, in principle, is reviewing a Court of Appeals' conclusion, it seems to me that its opinion might call into question the fact related determinations of the District Court. Ante, at 34. To the extent that it does so, I disagree, for such findings cannot be so quickly disregarded. And I do not think that our precedents permit this Court to do so. See Reiter, 442 U. S., at 345; Yamasaki, 442 U. S., at 703.

The issues in this case are complicated and difficult. The District Court might have been correct. Or not. Subclasses might be appropriate. Or not. I cannot tell. And I do not believe that this Court should be in the business of trying to make these fact based determinations. That is a job suited to the district courts in the first instance, and the courts of appeal on review. But there is no reason in this case to believe that the Court of Appeals conducted its prior review with an understanding that the settlement could have constituted a reasonably strong factor in favor of class certification. For this reason, I would provide the courts below with an opportunity to analyze the factual questions involved in certification by vacating the judgment, and remanding the case for further proceedings.