skip navigation

ORTIZ V. FIBREBOARD CORP. (97-1704) 527 U.S. 815 (1999)
134 F.3d 668, reversed and remanded.
[ Souter ]
[ Rehnquist ]
[ Breyer ]
HTML version
PDF version
HTML version
PDF version
HTML version
PDF version
HTML version
PDF version

Breyer, J., dissenting


No. 97—1704



[June 23, 1999]

    Justice Breyer, with whom Justice Stevens joins, dissenting.

    This case involves a settlement of an estimated 186,000 potential future asbestos claims against a single company, Fibreboard, for approximately $1.535 billion. The District Court, in approving the settlement, made 446 factual findings, on the basis of which it concluded that the settlement was equitable, that the potential claimants had been well represented, and that the distinctions drawn among different categories of claimants were reasonable. 162 F. R. D. 505 (1995); App. to Pet. for Cert. 248a—468a. The Court of Appeals, dividing 2 to 1, held that the settlement was lawful. 134 F.3d 668 (CA5 1998). I would not set aside the Court of Appeals’ judgment as the majority does. Accordingly, I dissent.



    Four special background circumstances underlie this settlement and help to explain the reasonableness and consequent lawfulness of the relevant District Court
determinations. First, as the majority points out, the settlement comprises part of an “elephantine mass of asbestos cases,” which “defies customary judicial administration.” Ante, at 1. An estimated 13 to 21 million workers have been exposed to asbestos. See Report of the Judicial Conference Ad Hoc Committee on Asbestos Litigation 6—7 (Mar. 1991) (hereinafter Judicial Conference Report). Eight years ago the Judicial Conference spoke of the mass of related cases having “reached critical dimensions,” threatening “a disaster of major proportions.” Id., at 2. In the Eastern District of Texas, for example, one out of every three civil cases filed in 1990 was an asbestos case. See id., at 8. In the past decade nearly 80,000 new federal asbestos cases have been filed; more than 10,000 new federal asbestos cases were filed last year. See U.S. District Courts Civil Cases Commenced by Nature of Suit, Administrative Office of the Courts Statistics (Table C2—A) (Dec. 31, 1994—1998) (hereinafter AO Statistics).

    The Judicial Conference found that asbestos cases on average take almost twice as long as other lawsuits to resolve. See Judicial Conference Report 10—11. Judge Parker, the experienced trial judge who approved this settlement, noted in one 3,000-member asbestos class action over which he presided that 448 of the original class members had died while the litigation was pending. Cimino v. Raymark Industries, Inc., 751 F. Supp. 649, 651 (ED Tex. 1990). And yet, Judge Parker went on to state, if the district court could close “thirty cases a month, it would [still] take six and one-half years to try these cases and [due to new filings] there would be pending over 5,000 untouched cases” at the end of that time. Id., at 652. His subsequent efforts to accelerate final decision or settlement through the use of sample cases produced a highly complex trial (133 trial days, more than 500 witnesses, half a million pages of documents) that eventually closed only about 160 cases because efforts to extrapolate from the sample proved fruitless. See Cimino v. Raymark Industries, Inc., 151 F.3d 297, 335 (CA5 1998). The consequence is not only delay but also attorney’s fees and other “transaction costs” that are unusually high, to the point where, of each dollar that asbestos defendants pay, those costs consume an estimated 61 cents, with only 39 cents going to victims. See Judicial Conference Report 13.

    Second, an individual asbestos case is a tort case, of a kind that courts, not legislatures, ordinarily will resolve. It is the number of these cases, not their nature, that creates the special judicial problem. The judiciary cannot treat the problem as entirely one of legislative failure, as if it were caused, say, by a poorly drafted statute. Thus, when “calls for national legislation” go unanswered, ante, at 1, judges can and should search aggressively for ways, within the framework of existing law, to avoid delay and expense so great as to bring about a massive denial of justice.

    Third, in that search the district courts may take advantage of experience that appellate courts do not have. Judge Parker, for example, has written of “a disparity of appreciation for the magnitude of the problem,” growing out of the difference between the trial courts’ “daily involvement with asbestos litigation” and the appellate courts’ “limited” exposure to such litigation in infrequent appeals. Cimino, 751 F. Supp., at 651.

    Fourth, the alternative to class-action settlement is not a fair opportunity for each potential plaintiff to have his or her own day in court. Unusually high litigation costs, unusually long delays, and limitations upon the total amount of resources available for payment, together mean that most potential plaintiffs may not have a realistic alternative. And Federal Rule of Civil Procedure 23 was designed to address situations in which the historical model of individual actions would not, for practical reasons, work. See generally Advisory Committee’s Notes on Fed. Rule Civ. Proc. 23, 28 U.S.C. App., p. 696 (discussing, in relation to Rule 23(b)(1)(B), instances in which individual judgments, “while not technically concluding the other members, might do so as a practical matter”).

    For these reasons, I cannot easily find a legal answer to the problems this case raises by referring, as does the majority, to “our ‘deep-rooted historic tradition that everyone should have his own day in court.’ Ante, at 28 (citation omitted). Instead, in these circumstances, I believe our Court should allow a district court full authority to exercise every bit of discretionary power that the law provides. See generally Califano v. Yamasaki, 442 U.S. 682, 703 (1979) (“[M]ost issues arising under Rule 23 . . . [are] committed in the first instance to the discretion of the district court”); Reiter v. Sonotone Corp., 442 U.S. 330, 345 (1979) (district courts have “broad power and discretion . . . with respect to matters involving the certification” of class actions). And, in doing so, the Court should prove extremely reluctant to overturn a fact-specific or circumstance-specific exercise of that discretion, where a court of appeals has found it lawful. Cf. Universal Camera Corp. v. NLRB, 340 U.S. 474, 490—491 (1951) (Supreme Court will rarely overturn appellate court review of agency fact-finding). This cautionary principle of review leads me to an ultimate conclusion different from that of the majority.


    The case before us involves a class of individuals (and their families) exposed to asbestos manufactured by Fibreboard who, for the most part, had not yet sued or settled with Fibreboard as of August 1993. The negotiating parties estimated that Fibreboard faced approximately 186,000 of these future claims. See App. to Pet. for Cert. 321a; cf. AO Statistics, Table C2—A (total number of all civil cases filed in federal district courts in 1998 was 252,994). Although the District Court was unable to give a precise figure, see App. to Pet. for Cert. 356a—357a, there is no doubt that a realistic assessment of the value of these claims far exceeds Fibreboard’s total net worth.

    But, as of 1993, one potentially short-lived additional asset promised potential claimants a greater recovery. That asset consisted of two insurance policies, one issued by Continental Casualty, the other by Pacific Indemnity. If the policies were valid (i.e., if they covered most of the relevant claims), they were worth several billion dollars; but if they were invalid, this asset was worth nothing. At that time, a separate case brought by Fibreboard against the insurance companies in California state court seemed likely to resolve the value of the policies in the near future. That separate litigation had a settlement value for the insurance companies. At the time the parties were negotiating, prior to the California court’s decision, the insurance policies were worth, as the majority puts it, the value of “unlimited policy coverage” (i.e., perhaps the insurance companies’ entire net worth) “discounted by the risk that Fibreboard would ultimately lose the coverage dispute litigation.” Ante, at 33.

    The insurance companies offered to settle with both Fibreboard and those persons with claims against Fibreboard (who might have tried to sue the insurance companies directly). The settlement negotiations came to a head in August 1993, just as a California state appeals court was poised to decide the validity of the insurance policies. This fact meant speed was important, for the California court could well decide that the policies were worth nothing. It also meant that it was important to certify a non opt-out class of Fibreboard plaintiffs. If the class that entered into the settlement were an opt-out class, then members of that class could wait to see what the California court did. If the California court found the policies valid (hence worth many billions of dollars), they would opt out of the class and sue for everything they could get; if the California court found the policies invalid (and worth nothing), they would stick with the settlement. The insurance companies would gain little from that kind of settlement, and they would not agree to it. See In re Asbestos Litigation, 90 F.3d 963, 970 (CA5 1996).

    After eight days of hearings, the District Court found that the insurance policies plus Fibreboard’s net worth amounted to a “limited fund,” valued at $1.77 billion (the amount the insurance companies were willing to contribute to the settlement plus Fibreboard’s value). See App. to Pet. for Cert. 492a. The court entered detailed factual findings. See generally 162 F. R. D., at 518—519. It certified a “non opt-out” class. And the court approved the parties’ Global Settlement Agreement. The Global Settlement Agreement allows those exposed to asbestos (and their families) to assert their Fibreboard claims against a fund that it creates. It does not limit recoveries for particular types of claims, but allows for individual determinations of damages based on all historically relevant individual factors and circumstances. See 90 F.3d, at 976. It contains spendthrift provisions designed to limit the total payouts for any particular year, and a requirement that the claimants with the most serious injuries be paid first in any year in which there is a shortfall. It also permits an individual who wishes to retain his right to bring an ordinary action in court to opt out of the arrangement (albeit after mediation and nonbinding arbitration), but sets a ceiling of $500,000 upon the recovery obtained by any person who does so. See generally 162 F. R. D., at 518—519.

    The question here is whether the court’s certification of the class under Rule 23(b)(1)(B) violates the law. The majority seems to limit its holding (though not its discussion) to that question, and so I limit the focus of my dissent to the Rule 23(b)(1)(B) issues as well.


    The District Court certified a class consisting primarily of individuals (and their families) who had been exposed to Fibreboard’s asbestos but who had not yet made claims. See ante, at 6—7, and n. 5. It did so under the authority of Federal Rule of Civil Procedure 23(b)(1)(B), which, by analogy to pre-Rules “limited fund” cases, permits certification of a non opt-out class where

“the prosecution of separate actions by or against individual members of the class would create a risk of … adjudications with respect to individual members of the class which would as a practical matter be dispositive of the interests of the other members not parties to the adjudications or substantially impair or impede their ability to protect their interests.”

The majority thinks this class could not be certified under Rule 23(b)(1)(B). I, on the contrary, think it could.

    The case falls within the Rule’s language as long as there was a significant “risk” that the total assets available to satisfy the claims of the class members would fall well below the likely total value of those claims, for in such circumstances the money would go to those claimants who brought their actions first, thereby “substantially impair[ing]” the “ability” of later claimants “to protect their interests.” And the District Court found there was indeed such a “risk.” 162 F. R. D., at 526.

    Conceptually speaking, that “risk” was no different from the risk inherent in a classic pre-Rules “limited fund” case. Suppose a broker agrees to invest the funds of 10 individuals who each give the broker $100. The broker misuses the money, and the customers sue. (1) Suppose their claims total $1,000, but the broker’s total assets amount to $100. (2) Suppose the same broker has no assets left, but he does have an insurance policy worth $100. (3) Suppose the broker has both $100 in assets and a $100 insurance policy.

    The first two cases are classic limited fund cases. See ante, at 16—17 (citing, e.g., Dickinson v. Burnham, 197 F.2d 973 (CA2 1952), cert. denied, 344 U.S. 875 (1952), an investors’ suit for the return of misused funds); ante, at 18 (citing, e.g., Morrison v. Warren, 174 Misc. 233, 234, 20 N. Y. S. 2d 26, 27 (Sup. Ct. 1940), a suit to distribute insurance proceeds to third party beneficiaries). The third case simply combines the first two, and that third case is the case before us.

    Of course the value of the insurance policies in our case is not as precise as the $100 in my example, nor was it certain at the time of settlement. But that uncertainty makes no difference. It was certain that the insurance policies’ value was limited. And that limitation was created by the likelihood of an independent judicial determination of the meaning of words in the policy, in respect to which the merits or value of the underlying tort claims against Fibreboard were beside the point.

    Nor does it matter that the value of the insurance policies in our case might have fluctuated over time. Long before the Federal Rules of Civil Procedure, courts permitted actions by one group of insurance policy holders to bind all policy holders, even where the group proceeded against an insurance-company-administered fund that fluctuated over time. See Hartford Life Ins. Co. v. IBS, 237 U.S. 662, 672 (1915) (life insurance fund which, like the fund before us, was administered through court-ordered rules that bound all policy holders).

    Neither does it matter that the insurance policies might be worth much more money if the California court decided the coverage dispute in Fibreboard’s favor. A trust worth, say, $1 million (faced with $2 million in claims) is a limited fund, despite the possibility that a company whose stock it holds might strike oil and send the value of the trust skyrocketing. Limitation is a matter of present value, which takes appropriate account of such future possibilities.

    I need not pursue the conceptual matter further, however, for the majority apparently concedes the conceptual point that a fund’s limit may equal its “value discounted by risk.” Ante, at 33. But the majority sets forth three additional conditions, which it says are “sufficient … to justify binding absent members of a class under Rule 23(b)(1)(B), from which no one has the right to secede.” Ante, at 20. Those three conditions are:

Condition One: That “the totals of the aggregated liquidated claims and the fund available for satisfying them, set definitely at their maximum, demonstrate the inadequacy of the fund to pay all the claims.” Ibid.; Part IV—A, ante.

Condition Two: That “the claimants identified by a common theory of recovery were treated equitably among themselves.” Ante, at 21; Part IV—B, ante.

Condition Three: That “the whole of the inadequate fund was to be devoted to the overwhelming claims.” Ante, at 20; Part IV—C, ante.

I shall discuss each condition in turn.


    In my view, the first condition is substantially satisfied. No one doubts that the “totals of the aggregated” claims well exceed the value of the assets in the “fund available for satisfying them,” at least if the fund totaled about what the District Court said it did, namely, $1.77 billion at most. The District Court said that the limited fund equaled in value “the sum of the value of Fibreboard plus the value of its insurance coverage,” or $235 million plus $1.535 billion. App. to Pet. for Cert. 492a. The Court of Appeals upheld the finding. 90 F.3d, at 982. And the finding is adequately supported.

    The District Court found that the insurance policies were not worth substantially more than $1.535 billion in part because there was a “significant risk” that the insurance policies would soon turn out to be worth nothing at all. 162 F. R. D., at 526. The court wrote that “Fibreboard might lose” its coverage, i.e., that it might lose “on one or more issues in the [California] Coverage Case, or that Fibreboard might lose its insurance coverage as a result of its assignment settlement program.” Ibid.

    Two California insurance law experts, a Yale professor and a former state court of appeals judge, testified that there was a good chance that Fibreboard would lose all or a significant part of its insurance coverage once the California appellate courts decided the matter. 90 F.3d, at 974. And that conclusion is not surprising. The Continental policy (for which Fibreboard had paid $10,000 per year) carried limits of $500,000 “per-person” and $1 million “per-occurrence,” had been in effect only between May 1957 and March 1959, and arguably denied Fibreboard the right to settle tort cases as it had been doing. See App. to Pet. for Cert. 267a. The Pacific policy was said (no one could find a copy) to carry a $500,000 per-claim limit, and had been in effect only for one year, from 1956—57. See ibid. To win significantly in respect to either of the two policies, Fibreboard had to show that the policies fully covered a person exposed to asbestos long before the policy year (say, in 1948) even if the disease did not appear until much later (say, in 2002). It also had to explain away the $1 million per occurrence limit in the Continental policy, despite policy language defining “one occurrence” as “ ‘[a]ll … exposure to substantially the same general conditions existing at or emanating from each premises location.’ ” Brief for Respondents Continental Casualty et al. 5. And Fibreboard had to show that its tort-suit settlement practice was consistent with the policy.

    The settlement value of previous cases also indicated that the insurance policies were of limited value. Fibreboard’s “no-cash” settlements (which required a settling plaintiff to obtain recovery from the insurance companies) were twice as high on average as were its comparable 40% cash settlements. App. to Pet. for Cert. 231a. That difference, suggesting a 50% discount for 40% cash, in turn suggests that settling parties estimated the odds of recovering on the insurance policies as worse than 2 to 1 against.

    The District Court arrived at the present value of the policies ($1.535 billion) by looking to a different settlement, the settlement arrived at in the insurance coverage case itself as a result of bargaining between Fibreboard and the insurance companies. See id., at 492a. That settlement, embodied in the Trilateral Agreement, created a backup fund by taking from the insurance companies $1.535 billion (plus other money used to satisfy claims not here at issue) and simply setting it aside to use for the payment of claims brought against Fibreboard in the ordinary course by members of this class (in the event that the federal courts ultimately failed to approve the Global Settlement Agreement).

    The Fifth Circuit approved this method of determining the value of the insurance policies. See 90 F.3d, at 982 (discussing value of Trilateral Agreement plus value of Fibreboard). And the majority itself sees nothing wrong with that method in principle. The majority concedes that one

“may take a settlement amount as good evidence of the maximum available if one can assume that parties of equal knowledge and negotiating skill agreed upon the figure through arms-length bargaining, unhindered by any considerations tugging against the interests of the parties ostensibly represented in the negotiation.” Ante, at 34.

    The majority rejects the District Court’s valuation for a different reason. It says that the settlement negotiation that led to the valuation was not necessarily a fair one. The majority says it cannot make the necessary “arms-length bargaining” assumption because “[c]lass counsel” had a “great incentive to reach any agreement” in light of the fact that “some of the same lawyers … had also negotiated the separate settlement of 45,000” pending cases, which was partially contingent upon a global settlement or other favorable resolution of the insurance dispute. Id., at 34—35 (emphasis added).

    The District Court and Court of Appeals, however, did accept the relevant “arms-length” assumption, with good reason. The relevant bargaining (i.e., the bargaining that led to the Trilateral Agreement that set the policies’ value) was not between the plaintiffs’ class counsel and the insurance companies; it was between Fibreboard and the insurance companies. And there is no reason to believe that that bargaining, engaged in to settle the California coverage dispute, was not “arms-length.” That bargaining did not lead to a settlement that would release Fibreboard from potential tort liability. Rather, it led to a potential backup settlement that did not release Fibreboard from anything. It created a fund of insurance money, which, once exhausted, would have left Fibreboard totally exposed to tort claims. Consequently, Fibreboard had every incentive to squeeze as much money as possible out of the insurance companies, thereby creating as large a fund as possible in order to diminish the likelihood that it would eventually have to rely upon its own net worth to satisfy future asbestos plaintiffs.

    Nor are petitioners correct when they argue that the insurance companies’ participation in setting the value of the insurance policies created a fund that is limited “only in the sense that … every settlement is limited.” Brief for Petitioners 28. As the District Court found, the fund was limited by the value of the insurance policies (along with Fibreboard’s own limited net worth), and that limitation arose out of the independent likelihood that the California courts would find the policies valueless. App. to Pet. for Cert. 492a. That is why the District Court said that certification in this case does not determine whether

“mandatory class certification is appropriate in the typical case where a class action is settled with a defendant’s own funds, or with insurance funds that are not the subject of genuine and vigorous dispute.” 162 F. R. D., at 527.

The court added that, in the ordinary case: “If the settlement failed, … the defendant would retain the settlement funds (or the insurance coverage), and there might not be the ‘impair[ment]’ to class members’ ‘ability to protect their interests’ required for mandatory class certification.” Ibid. In this case, however, if settlement failed, coverage “may well disappear … with the result that Class members could not then secure their due through litigation.” Ibid.

    I recognize that one could reasonably argue about whether the total value of the insurance policies (plus the value of Fibreboard) is $1.535 billion, $1.77 billion, $2.2 billion, or some other roughly similar number. But that kind of argument, in this case, is like arguing about whether a trust fund, facing $30,000 in claims, is worth $15,000 or $20,000 (e.g., do we count Aunt Agatha’s share as part of the fund?), or whether a ship, subject to claims that, by any count, exceed its value, is worth a little more or a little less (e.g., does the coal in the hold count as fuel, which is part of the ship’s value, or as cargo, which is not?). A perfect valuation, requiring lengthy study by independent experts, is not feasible in the context of such an unusual limited fund, one that comes accompanied with its own witching hour. Within weeks after the parties’ settlement agreement, the insurance policies might well have disappeared, leaving most potential plaintiffs with little more than empty claims. The ship was about to sink, the trust fund to evaporate; time was important. Under these circumstances, I would accept the valuation findings made by the District Court and affirmed by the Court of Appeals as legally sufficient. See supra, at 4.


    I similarly believe that the second condition is satisfied. The “claimants … were treated equitably among themselves.” Ante, at 21. The District Court found equitable treatment, and the Court of Appeals affirmed. But a majority of this Court now finds significant inequities arising out of class counsel’s “egregious” conflict of interest, the settlement’s substantive terms, and the District Court’s failure to create subclasses. See ante, at 35—39. But nothing I can find in the Court’s opinion, nor in the objectors’ briefs, convinces me that the District Court’s findings on these matters were clearly erroneous, or that the Court of Appeals went seriously astray in affirming them.

    The District Court made 76 separate findings of fact, for example, in respect to potential conflicts of interest. App. to Pet. for Cert. 392a—430a. Of course, class counsel consisted of individual attorneys who represented other asbestos claimants, including many other Fibreboard claimants outside the certified class. Since Fibreboard had been settling cases contingent upon resolution of the insurance dispute for several years, any attorney who had been involved in previous litigation against Fibreboard was likely to suffer from a similar “conflict.” So whom should the District Court have appointed to negotiate a settlement that had to be reached soon, if ever? Should it have appointed attorneys unfamiliar with Fibreboard and the history of its asbestos litigation? Where was the District Court to find those competent, knowledgeable, conflict-free attorneys? The District Court said they did not exist. Finding of Fact ¶372 says there is “no credible evidence of the existence of other ‘conflict-free’ counsel who were qualified to negotiate” a settlement within the necessary time. Id., at 428a. Finding of Fact ¶317 adds that the District Court viewed it as “crucial … to appoint asbestos attorneys who were experienced, knowledgeable, skilled and credible in view of the extremely short window of opportunity to negotiate a global settlement, and the very high risk to future claimants presented by the Coverage Case appeal.” Id., at 401a. Where is the clear error?

    The majority emphasizes the fact that, by settling the claims of a class that consisted, for the most part, of persons who had not yet asserted claims against Fibreboard, counsel assured the availability of funds to pay other clients who had already asserted those claims. Ante, at 35. The decision to split the latter “inventory” claims from the former “class” claims, however, reflected the suggestion, not of class counsel, but of a judge, Circuit Judge Patrick Higginbotham, who had become involved in efforts to produce a timely settlement. Judge Higginbotham thought that negotiations had broken down because the combined class was “too complex.” App. to Pet. for Cert. 316a—317a; see also id., at 397a. He thought “inventory” claim settlements could be used as benchmarks to determine future class claim values, id., at 316a—317a, and that is just what happened. Although the majority is concerned that “inventory” plaintiffs “appeared to have obtained better terms than the class members,” ante, at 38, Finding of Fact ¶329 says that class counsel

“used the higher-than-average [inventory plaintiff settlement values] … to achieve a global settlement for future claimants at similarly high values, effectively arguing they could not possibly accept less for a class of future claimants than they had just negotiated for their present clients.” App. to Pet. for Cert. 407a.

In addition, more than 150 findings of fact, made after an 8-day hearing, support the District Court’s finding that overall the settlement is “fair, adequate, and reasonable.” See id., at 500a—501a. And, of course, Finding of Fact ¶318 says that appointing other attorneys–i.e., those who had no inventory clients–would have “ ‘jeopardiz[ed] any effort at serious negotiations’ ” and “resulted in a less favorable settlement” for the class, or perhaps no settlement followed by no insurance policy either. Id., at 402a.

    The Fifth Circuit found that “[t]he record amply supports” these District Court findings. 90 F.3d, at 978. Does the majority mean to set them aside? If not, does it mean to set forth a rigid principle of law, such as the principle that asbestos lawyers with clients outside a class, who will potentially benefit from a class settlement, can never represent a class in settlement negotiations? And does that principle apply no matter how unusual the circumstances, or no matter how necessary that representation might be? Why should there be such a rule of law? If there is not an absolute rule, however, I do not see how this Court can hold that the case before us is not that unusual situation.

    Consider next the claim that “equity” required more subclasses. Ante, at 38—40. To determine the “right” number of subclasses, a district court must weigh the advantages and disadvantages of bringing more lawyers into the case. The majority concedes as much when it says “at some point there must be an end to reclassification with separate counsel.” Ante, at 39. The District Court said that if there had “been as many separate attorneys” as the objectors wanted, “there is a significant possibility that a global settlement would not have been reached before the Coverage Case was resolved by the California Court of Appeal.” App. to Pet. for Cert. 428a. Finding of Fact ¶346 lists the shared common interests among subclasses that argue for single representation, including “avoiding the potentially disastrous results of a loss . . . in the Coverage Case,” “maximizing the total settlement contribution,” “reducing transactions costs and delays,” “minimizing … attorney’s fees,” and “adopting” equitable claims payment “procedures.” Id., at 415a. Surely the District Court was within its discretion to conclude that “the point” to which the majority alludes was reached in this case.

    I need not go into further detail here. Findings of Fact ¶¶347—354 explain why the alleged conflict between pre- and post-1959 claimants is not significant. Id., at 415a—418a (noting that “the decision as to how to divide the settlement among class members” did not take place until after the Trilateral Agreement was agreed to, at which point money was available equally to both pre- and post-1959 claimants). Findings of Fact ¶¶355—363 explain why the alleged conflict between claimants with, and those without, current illnesses is not significant. Id., at 419a—422a (explaining why “the interest of the two subgroups at issue here coincide to a far greater extent than they diverge”). The Fifth Circuit found that the District Court “did not abuse its discretion in finding that the class was adequately represented and that subclasses were not required.” 90 F.3d, at 982. This Court should not overturn these highly circumstance-specific judgments.


    The majority’s third condition raises a more difficult question. It says that the “whole of the inadequate fund” must be “devoted to the overwhelming claims.” Ante, at 20 (emphasis added). Fibreboard’s own assets, in theory, were available to pay tort claims, yet they were not included in the global settlement fund. Is that fact fatal?

    I find the answer to this question in the majority’s own explanation. It says that the third condition helps to guarantee that those who held the

“inadequate assets had no opportunity to benefit [themselves] or claimants of lower priority by holding back on the amount distributed to the class. The limited fund cases thus ensured that the class as a whole was given the best deal; they did not give a defendant a better deal than seriatim litigation would have produced.” Ante, at 20—21.

That explanation suggests to me that Rule 23(b)(1)(B) permits a slight relaxation of this absolute requirement, where its basic purpose is met, i.e., where there is no doubt that “the class as a whole was given the best deal,” and where there is good reason for allowing the third condition’s substantial, rather than its literal, satisfaction.

    Rule 23 itself does not require modern courts to trace every contour of ancient case law with literal exactness. Benjamin Kaplan, reporter to the Advisory Committee on Civil Rules that drafted the 1966 revisions, upon whom the majority properly relies for explanation, see, e.g., ante, at 14, 15, 24, wrote of Rule 23:

“The reform of Rule 23 was intended to shake the law of class actions free of abstract categories … and to rebuild the law on functional lines responsive to those recurrent life patterns which call for mass litigation through representative parties. … And whereas the old Rule had paid virtually no attention to the practical administration of class actions, the revised Rule dwelt long on this matter–not, to be sure, by prescribing detailed procedures, but by confirming the courts’ broad powers and inviting judicial initiative.” A Prefatory Note, 10 B. C. Ind. & Com. L. Rev. 497 (1969).

The majority itself recognizes the possibility of providing incentives to enter into settlements that reduce costs by granting a “credit” for cost savings by relaxing the whole-of-the-assets requirement, at least where most of the savings would go to the claimants. Ante, at 44.

    There is no doubt in this case that the settlement made far more money available to satisfy asbestos claims than was likely to occur in its absence. And the District Court found that administering the fund would involve transaction costs of only 15%. App. to Pet. for Cert. 362a. A comparison of that 15% figure with the 61% transaction costs figure applicable to asbestos cases in general suggests hundreds of millions of dollars in savings–an amount greater than Fibreboard’s net worth. And, of course, not only is it better for the injured plaintiffs, it is far better for Fibreboard, its employees, its creditors, and the communities where it is located for Fibreboard to remain a working enterprise, rather than slowly forcing it into bankruptcy while most of its money is spent on asbestos lawyers and expert witnesses. I would consequently find substantial compliance with the majority’s third condition.

    Because I believe that all three of the majority’s conditions are satisfied, and because I see no fatal conceptual difficulty, I would uphold the determination, made by the District Court and affirmed by the Court of Appeals, that the insurance policies (along with Fibreboard’s net value) amount to a classic limited fund, within the scope of Rule 23(b)(1)(B).


    Petitioners raise additional issues, which the majority does not reach. I believe that respondents would likely prevail were the Court to reach those issues. That is why I dissent. But, as the Court does not reach those issues, I need not decide the questions definitively.

    In some instances, my belief that respondents would likely prevail reflects my reluctance to second-guess a court of appeals that has affirmed a district court’s fact- and circumstance-specific findings. See supra, at 4; cf. Amchem Products, Inc. v. Windsor, 521 U.S. 591, 629—630 (1997) (Breyer, J., concurring in part and dissenting in part). That reluctance applies to those of petitioners’ further claims that, in effect, attack the District Court’s conclusions related to: (1) the finding under Rule 23(a)(2) that there are “questions of law and fact common to the class,” see App. to Pet. for Cert. 480a; see generally Amchem, supra, at 634—636 (Breyer, J., concurring in part and dissenting in part); (2) the finding under Rule 23(a)(3) that claims of the representative parties are “typical” of the claims of the class, see App. to Pet. for Cert. 480a—481a; (3) the adequacy of “notice” to class members pursuant to Rule 23(e) and the Due Process Clause, see id., at 511a; see generally Amchem, supra, at 640—641 (Breyer, J., concurring in part and dissenting in part); and (4) the standing-related requirement that each class member have a good-faith basis under state law for claiming damages for some form of injury-in-fact (even if only for fear of cancer or medical monitoring), see App. to Pet. for Cert. 252a; cf., e.g., Coover v. Painless Parker, Dentist, 105 Cal. App. 110, 286 P. 1048 (1930).

    In other instances, my belief reflects my conclusion that class certification here rests upon the presence of what is close to a traditional limited fund. And I doubt that petitioners’ additional arguments that certification violates, for example, the Rules Enabling Act, the Bankruptcy Act, the Seventh Amendment, and the Due Process Clause, are aimed at or would prevail against a traditional limited fund (e.g., “trust assets, a bank account, insurance proceeds, company assets in a liquidation sale, proceeds of a ship sale in a maritime accident suit,” ante, at 15—16 (internal quotation marks and citations omitted)). Cf. In re Asbestos Litigation, 90 F.3d, at 986 (noting that Phillips Petroleum Co. v. Shutts, 472 U.S. 797 (1985), involved a class certified under the equivalent of Rule 23(b)(3), not a limited fund case under Rule 23(b)(1)(B)). Regardless, I need not decide these latter issues definitively now, and I leave them for another day. With that caveat, I respectfully dissent.