My theme for discussing Professor Cooper's paper(1) is prompted by the impending retirement of Senator Daniel Patrick Moynihan. Many may recall how he first came to broad public attention. In .January 1969, as an advisor to President-elect Richard Nixon, Moynihan wrote a memorandum concerning the

course the new President should adopt toward race relations in the wake of the civil rights activism of the 1960s. Borrowing the Earl of Durham's 1839 suggestion for the British attitude toward Canada,(2) Moynihan recommended that the new administration pursue a policy of "benign neglect."

After the New York Times obtained a copy of the memorandum and published it,(3) Time reported that "the document caused a sensation."(4) Noting that Moynihan was a long-standing Democrat who had served in the Johnson Administration and was "generally regarded as the most liberal of close Presidential advisers," the Wall Street Journal opined on "Mr. Moynihan's Apostasy."(5) It concluded that the memorandum was leaked to enrage civil rights leaders "because of the incendiary phrase."(6) Anthony Lewis opined that African Americans would rightly interpret neglect as hostile,(7) and the Times editorially concluded that the phrase "may have been felicitous when applied to Canada by the Earl of Durham in 1839 but it suggests no program for black America in 1970."(8) Congress also discussed the implications of the memorandum.(9)

So anyone who borrows the Earl of Durham's phrase another time must do so diffidently. Yet the procedures that undergird the debate about mass tort litigation also owe much to a different form of 1960s activism--active employment of the federal courts to foster multiparty litigation. The 1966 amendments to Rule 23 were partly designed to facilitate civil rights litigation,(10) and the framers were certainly aware of the possible use of the new provisions of Rule 23(b)(3) in mass accident litigation.(11) The 1968 adoption of the Multidistrict Litigation Act.(12) embraced another theme of the era--achieving efficiency by combining federal lawsuits from across the land. The combination of these two creations of the Sixties has proved critical to much of current mass tort litigation.

Since the Sixties there have been proposals to change Rule 23, but these have not been adopted except for one relatively small provision.(13) Regarding legislation and rulemaking, therefore, it seems that benign neglect has been the order of the day,(14) even in the face of increasing enthusiasm in some quarters for action to respond to the challenges of mass torts. In 1991, for instance, the Judicial Conference's Ad Hoc Committee on Asbestos Litigation concluded that there was a "litigation impasse [that] cannot be broken except by aggregate or class proceedings,"(15) and urged Congress to act to resolve the problem.(16) That did not happen, and the courts were left to improvise using existing procedural tools. In 1999, Chief Justice Rehnquist joined the Court's decision in Ortiz v. Fibreboard Corp.(17) decrying the "elephantine mass of asbestos cases,"(18) a but added a separate opinion concluding that the Court's rejection of a class action settlement designed to solve those problems was correct because "we are not free to devise an ideal system for adjudicating these claims."(19) Accordingly, "[u]nless and until the Federal Rules of Civil Procedure are revised, the Court's opinion correctly states the existing law."(20) Perhaps this was an invitation from on high to change the rules.

Much study has been done on how an "ideal system" might be designed, and how Rule 23 might be amended to further that goal. Throughout the 1990s, the Advisory Committee on Civil Rules has been considering possible changes to Rule 23.(21) Shortly after Chief Justice Rehnquist's call for action in Ortiz, there were hearings in Congress about a possible solution to the asbestos litigation problem.(22)

Professor Cooper's paper moves beyond asbestos litigation to propose a statutory "all-encompassing" solution to the problems of mass tort litigation. In addition, it proposes a rule revision that would not require statutory facilitation. For those in a position to act on these suggestions, the question is whether the time for benign neglect has passed. To shed light on that question, this Commentary will play devil's advocate and present the arguments for adhering to a policy of benign neglect.

I. BENIGN NEGLECT IN THE PAST

Before addressing the current proposals, it is important to recall our experience with benign neglect during the last quarter century. This is not the first time there have been urgent calls for class action reform. To the contrary, twenty-five years ago there was what Arthur Miller called a "holy war" over Rule 23.(23) The criticism took discordant themes. One decried the "enabling" features of the 1966 amendments. It was typified by Milton Handler's denunciation of antitrust class actions as "legalized blackmail,"(24) and the condemnation of the Eisen class action as a "Frankenstein Monster posing as a class action."(25) Others found that there was too much restraint on the availability of class actions. Writing in the Columbia Law Review, a representative of the Department of Justice urged that class action reform was necessary to end "[s]ubstance's [i]ndenture to [p]rocedure."(26) He decried the ability of defendants to frustrate class certification by challenging the existence or predominance of common questions, or by attacking the adequacy of representation afforded by the person who brought the suit.(27)

Faced with this turmoil, the Advisory Committee concluded that any modification of class action practice should come from Congress, a position approved by the Judicial Conference.(28) The Department of Justice proposed legislation to rectify the problems that it believed affected class action practice,(29) but that bill was not adopted. In effect, benign neglect became the order of the day.

Yet the roof did not fall in. In part, this was due to restrictive Supreme Court class action decisions on subject matter jurisdiction,(30) notice to class members in Rule 23(b)(3) class actions,(31) and appellate jurisdiction.(32) But more generally, as predicted by Professor Miller in 1979, it seems that the courts mastered the difficulties presented by Rule 23. Miller noted that there had first been a period of "euphoria over the rule's potential" among the legal community' in the second half of the 1960s during which not enough attention was paid to the rule's requirements.(33) This initial period of excitement was followed by a period of reaction from about 1969 to 1974 when "antipathy to the class action became palpable" among courts.(34) Miller believed that the pendulum then began to swing again, and that we then entered a phase during which "[i]nstead of wielding a meat axe, courts increasingly ... operat[ed] with a scalpel."(35) In his view, benign neglect would work. When the RAND Corporation announced its class action study in 1999, it seemed to confirm that Miller's prediction had proved correct: "[A] s the years passed, the legal system gradually acclimated itself to the 1966 rule. Courts pulled back from their initial enthusiastic support, litigation patterns became more predictable and therefore easier for corporations to adjust to, and the clamor for rule revision died down."(36) Indeed, things became so tranquil that in 1988 the New York Times reported that class actions were "dying."(37)

Shortly after Professor Miller wrote, however, seeds of new turmoil about class actions were sown. The opening shot was fired in 1981 by Judge Spencer Williams in San Francisco, who was presiding over personal injury suits brought by users of the Dalkon Shield contraceptive device. Appreciating the prospect of burgeoning mass tort ligation,(38) Judge Williams saw Rule 23 as the key to a cure. He therefore certified a nationwide mandatory class action for punitive damages and a statewide Rule 23(b)(3) opt-out class action for compensatory damages for California users of the Dalkon Shield. He saw class treatment as essential to avoid the "unconscionable possibility that large numbers of plaintiffs who are not first in line at the courthouse door will be deprived of a practical means of redress."(39) When the Ninth Circuit held that Rule 23 did not permit what he had done, Judge Williams wrote an article lamenting the possible demise of mass tort class actions and forecasting that "until Congress addresses these questions by enacting comprehensive federal products liability law ... the inequities and shortcomings of the present system require that we judges work in an innovative fashion, adapting aspects of the current system to address these challenging problems."(40)

Judge Williams was right about judicial innovation; as Professor Cooper says, the courts have been "remarkably inventive" in addressing the problems of mass tort litigation with existing procedural tools.(41) As I have written elsewhere, federal judges engaged in this effort often seemed animated by a substantive preference to subordinate punitive damages to compensatory relief and to ensure that priority in obtaining relief would be given to those suffering the most serious injuries.(42) These efforts have relied not only on Rule 23, but also the power of the Panel on Multidistrict Litigation to combine federal cases under [sections] 1407, most prominently evidenced in the 1991 order transferring all asbestos personal injury actions to Philadelphia.(43)

There seem again to have been some elements of overenthusiasm in this process of creativity. It is chilling, for example, to learn of the insouciance exhibited by the district judge who certified a mandatory class action in Bendectin litigation in 1984.(44) This early experience seems to parallel the "euphoria" Professor Miller detected in the wake of the 1966 amendments to Rule 23.(45) There has surely been a retrenchment in the 1990s. The Supreme Court has twice struck down class action settlements in asbestos litigation.(46) The courts of appeals have proved inhospitable to mass tort class certification as well.(47) In some instances, these courts appeared to distend prevailing doctrine on appealability,(48) but the 1998 amendment to add Rule 23(f)(49) should provide a ready avenue for such review in the future. So if there was too much enthusiasm for certifying mass tort class actions initially, it does seem that the over-enthusiasm has been reined in, and that a period of balanced treatment might result from more frequent appellate review under Rule 23(f).

Somewhat similarly, the Supreme Court's decision that transfer for trial is not permitted under [sections] 1407 may put the brakes on overly aggressive use of the transfer device to combine cases.(50)

Altogether, these developments show that the attitude toward mass tort combined litigation in federal court has changed significantly since the Advisory Committee began studying the problem in 1991. The period of enthusiasm that may have existed then has been supplanted by a more skeptical view, and it is possible that more nuanced use of existing procedural devices will characterize the future if benign neglect carries the day so far as legislation and rule amendment are concerned. Against that background, we can turn to Professor Cooper's specific proposals.

II. THE "ALL-ENCOMPASSING" STATUTORY SOLUTION

Like others, proceduralists dream of all-encompassing solutions to pressing problems, and Professor Cooper has shared his dream with us. It weaves together a variety of existing devices. Because the objective is to resolve all claims, the proposal relies on the class action. To manage the process, it drafts the Judicial Panel on Multidistrict Litigation. But the closest analogy seems to be interpleader. In 1967, the Supreme Court noted that "federal interpleader was not intended to serve the function of a `bill of peace' in the context of multiparty litigation arising out of a mass tort."(51) That is exactly what Professor Cooper's proposal is designed to do, however, and it is therefore to be embodied in a new statute to be codified as 28 U.S.C. [sections] 1335A, right after the provision setting forth the main ingredients of the current interpleader.

A. What the Proposal Does Not Do

Although Professor Cooper's proposal is "all-encompassing," it fails to satisfy some of the concerns voiced by critics of mass tort litigation. It is therefore worthwhile to begin by identifying some prominent omissions.

1. Individual litigation will remain necessary, and may even include class actions.

Critics of mass tort litigation have objected to dispersed repetitive lawsuits about the same product. For those who emphasize maturity as a necessary predicate for a reliable combined disposition, repetition is essential until a clear pattern emerges. For example, with regard to asbestos litigation, the Third Circuit recognized that "[t]he procedures of the traditional tort system proved effective in unearthing the hazards of asbestos."(52) In addition, the U.S. Supreme Court observed that plaintiffs' lawyers "honed the litigation of asbestos claims to the point of almost mechanical regularity."(53)

As noted below, the role of maturity under Professor Cooper's new proposal is not entirely clear. It is clear, however, that some litigation must occur before the proposed procedure can be initiated. Either 100 or more suits have to be filed or 100 claimants must assert claims against the defendant invoking the new procedure.(54) As Professor Cooper recognizes, there is no limitation on plaintiff class actions or other aggregated actions.(55) If one wants to put an end to dispersed or repetitive litigation, this solution is partial at best.

2. The proposal places no limitation on consumer class actions.

Some of the most vigorous criticisms of class actions have been directed at consumer class actions. Such actions bundle together thousands of small-claims plaintiffs, supposedly providing too much leverage to plaintiffs' lawyers. Recently, for example, there was much press coverage about a settlement by Toshiba of a class action charging that defendant's laptop computers had a flaw that could cause errors in recording data onto floppy disks. Although it asserted that nobody had even claimed that there was such a flaw until the suit was filed, and that it could not duplicate the alleged fault in its laboratory under normal operating conditions, Toshiba settled for an amount estimated to be $1 billion because a bad loss at trial could threaten the company's existence.(56) Soon thereafter, it was reported that this settlement had inspired "copycat" class actions against other manufacturers who vowed to fight them.(57)

The possibility that the class action would force such a settlement without regard to the merits is indeed troubling,(58) but it is not addressed by this proposal, which is limited to claims for personal injury or death, or for physical property damage. Cases like Toshiba's would not be included. Of course, that follows from the fact that this proposal addresses mass torts and not other problems, but it is worth recalling that this "all-encompassing" solution does not deal with some important areas of concern.

3. There is no provision for controlling punitive damages or ensuring that the worst go first.

Starting with Judge Williams in 1981,(59) federal judges have regularly expressed uneasiness about the possibility that repeated punitive damages awards could absorb funds that would better be spent on compensating injured victims. Similarly, many have expressed dismay that, at least in asbestos litigation, it often seems that claimants suffering no present impairment are receiving money while those facing death are forced to wait.(60)

The all-encompassing proposal contains no provisions for solving these problems. Until the new procedure is invoked, there is no impediment to litigation as it now occurs. Once the procedure is invoked, it does provide for a stay of pending actions,(61) but it has no provision for subordinating punitive damages claims to compensation claims or for granting priority to those worst injured. It may be that one could expect a mass tort court presiding over the cases to look with favor on these techniques, but there is nothing in the statute or the commentary provided thus far that endorses that result.

4. It does not preclude a sweetheart settlement of a class action.

Another abiding concern is the "reverse auction" possibility that a defendant could locate a pliant plaintiffs' lawyer and arrange a global class action settlement on favorable terms.(62) In large measure, this problem results from the fact that such a lawyer cannot threaten to go to trial unless the case is certified as a class action for litigation purposes, but certification for purposes of settlement may be available. Because the new procedure :authorized by [sections] 1335A comes into play only if the defendant initiates a mass tort action, there is nothing to stop a defendant who prefers the collusive settlement from pursuing that route if a pliant judge cart be found.

B. What the Proposal Does Do

Though there are a variety of criticisms not addressed by Professor Cooper's all-encompassing proposal, it is certainly not timid. To the contrary, [sections] 1335A includes several radical features and raises a number of questions about potentially unfortunate results should this solution to the problem of mass torts be adopted. We turn to those now.

1. It leads to a binding, no-opt-out decree as to all present and future claimants and defendants.

The defendant invoking the new mass tort action procedure must join all present and future claimants, all present and future defendants, and any insurer of any of those defendants possibly obliged to indemnify.(63) In furtherance of this broad rule of joinder, the broadest possible personal jurisdiction is afforded.(64) All pending actions are to be stayed,(65) and even a bankruptcy filing by one of the defendants does not interrupt the mass tort proceedings, for it must be filed in the district in which the mass tort action is pending.(66)

This is truly a nuclear weapon for mass tort litigation problems. It could result in a proceeding incomparably broader than any yet attempted. To put that observation in context, recall that the Supreme Court called the class action in Amchem "sprawling,"(67) and that the Fifth Circuit in Castano said that tobacco suit "may be the largest class action ever attempted in federal court."(68) Because both could be dwarfed by what [sections] 1335A contemplates, there is at least a reason to pause.

Beyond that, consider the difficulties facing the party who attempts to file such a mass tort action. Section 1335A(d) declares that the party filing the action must name all those who should be joined. How much effort is required to identify those people? Current claimants may not be difficult to identify,(69) but future claimants present much greater difficulties for mass-marketed products. Naming all parties possibly liable presents different sorts of difficulties. How well can a current defendant foresee all possible grounds for liability that may be asserted? In asbestos litigation, for example, there was a period during which claims against tobacco producers were suggested, either for contribution to asbestos producers or for direct liability to claimants.(70) Must a party inaugurating the new mass tort action be imaginative enough to foresee all such developments?

2. It leaves the decision whether to authorize the mass tort action to proceed to the Judicial Panel on Multidistrict Litigation, which must determine whether transfer under the new procedure is "desirable."

The Panel must dismiss the proceeding if it is not properly brought as a mass tort action. But the threshold for invoking the new procedure has intentionally been set low--100 claimants seeking an aggregate of $10,000,000, which could be satisfied in the case of a plane crash.(71) Professor Cooper correctly notes that this low threshold presumes that parties eligible to invoke the new procedure will exercise a good deal of self restraint because they will not want to shoulder the onerous burdens of initiation or maintenance of the mass tort action unless essential.(72) Nonetheless, the possibility of bluffing(73) suggests that proceedings may sometimes be initiated when they should not.

That is presumably why the statute endows the Panel with discretion to dismiss if transfer is not "desirable." Professor Cooper is confident that the Panel "will enter this new regime with a wealth of experience with large-scale pretrial consolidations, and will develop comparable experience that should support far more discerning decisions than any statute could dictate."(74) The statute therefore provides the Panel with virtually no guidance as to how it should make this decision. This paucity of authority stands in contrast to the extended list of factors provided for deciding whether to approve class settlements.(75) Given the magnitude of the task of adjudicating a mass tort action, one could well wish for more guidance. If there are criteria, it would be good to set them out, and if there are no criteria that can be set out, this proposal asks Congress to confer an astonishing amount of almost unreviewable authority on the Panel.

3. It directs the mass tort court to grasp the choice-of-law nettle.

This proposal builds on other choice-of-law proposals for mass tort litigation, such as the one made by the American Law Institute.(76) Unless the Panel dismisses, it transfers the mass tort proceeding to a mass tort court, which may in some instances be a state court. That court, in turn, is required to choose the law of a single state or country to govern all tort issues not governed by federal law.(77) This choice-of-law authority enables the mass tort court to overcome the obstacles to class action treatment that often result from applying divergent state law. Doing so, however, raises substantial federalism issues that are emphasized by the fact that proposed [sections] 1335A includes its own Mass Torts Rules Enabling Act. Mass tort rulemaking would be liberated from the constraints that usually prohibit rules that modify substantive rights.(78) The choice-of-law provision in [sections] 1335A exempts federal law from alteration by the mass tort court, but all substantive rights--including federal ones--would be up for grabs in the rulemaking process. The Amchem Court's concern with the differences in outcome in different states(79) is thus solved, but the broad consequences of the change are similarly made clear.

Against this background, it is somewhat odd that no mention is made of the federal courts' substantive preferences about punitive damages and ensuring that those suffering greater damages be allowed to go first.(80) One of the reasons for the federal courts' impatience about these issues is concern that the states may not have taken them seriously. As the Third Circuit remarked in 1986, states presented with these issues would probably act in "a parochial and nearsighted manner" because imposing a more equitable regime on their own residents would simply disadvantage them compared to the residents of other less enlightened states.(81) So it is unlikely that the law of any state would contain the ingredients that the federal courts have sought. Still, the mass tort court could be influenced in its choice-of-law decision by such substantive preferences. At least the mass tort court is not allowed to apply the law of more than one state or country. If it were, it would have even greater latitude for substantive preferences, for it could cobble together a collection of legal doctrines that reflect the judge's substantive preferences even though they do not exist together in the law of any jurisdiction. Nevertheless, the overall consequence is to broaden the powers of federal judges and rulemakers considerably.

The choice-of-law provision also has enormous strategic importance. Recall that the statute provides no guidance on whether the Panel should dismiss. One possible criterion for that decision would be whether the litigation is mature. It is easy to understand why that is not included--if massive litigation is bound to follow, there is no reason to think the mass tort court cannot properly manage the sequence of decision so that the claims are mature before class action treatment is employed. At the same time, the choice-of-law provision means that the filing of the mass tort action can change the ground rules in a very significant way. For some litigants, the chosen law will be less desirable than that which would otherwise apply. Those eligible to file mass tort actions therefore may be influenced by their expectations about whether they will reap a favorable change in law by doing so.

The existence of the choice-of-law question therefore may be inconsistent with the desire for "maturity." Should decisions already made under that other, more attractive, law be given any weight by the mass tort court? If a state-court class action is about to reach judgment, should that case be stayed pending the decision by the mass tort court whether the law of that state should be applied? Indeed, if the choice-of-law question remains an enigma until resolved by the mass tort court, is there any reason for the Panel to give a thought to the posture of litigation at the time the mass tort action is filed before it? This sort of wild card might well cause considerable confusion.

4. It creates substantial opportunities for bluffing.

The new mass tort act would give every defendant the equivalent of a litigation nuclear bomb. The Panel might dismiss a case that satisfies the statutory prerequisites on the ground that mass tort transfer is not "desirable," but one could hardly be sure that would happen. The statutory threshold for invoking the mass tort procedure is very low, and one confronting unfavorable law or juries might well be inclined to use the weapon to her advantage.

In our New World Order, we worry about terrorists obtaining nuclear weapons. In a similar vein, one could consider whether we might be arming some "terrorist" litigants with this nuclear weapon. Professor Cooper sensibly suggests that "it does not seem likely that many defendants will want to bring massive aggregation down on themselves without a compelling need."(82) Although this argument makes sense, it does not dispel concern about bluffing.

One possibility is bluffing directed at plaintiffs. A defendant might use the threat of a mass tort filing to extract concessions from plaintiffs who had obtained, or could reasonably expect to obtain, success in conventional litigation if they feared that the mass tort action (and its choice-of-law consequences) posed a threat to those successes. Since a mass tort action could be initiated even after a finding of liability, it would hang over all other proceedings, subject to the possibility that the Panel would find that a mass tort proceeding is not "desirable" in circumstances in which a completed proceeding in another court would be disrupted.

Another possible bluff is that the "weak sister" defendant might try to extract concessions from other defendants by threatening to throw the entire proceeding into a mass tort action. Recall that all defendants and all of their insurers would be sucked into such a proceeding. While a defendant might hesitate to bring this down on its own head, it might see an advantage in threatening to bring a mass tort action down on the heads of a variety of well-heeled co-defendants and insurers. At least a bluff to take such an action might benefit a party, and those considering creating this new bluffing opportunity should reflect on the potential costs that could result.

5. It requires elaborate and unprecedented revision of lines of appellate review.

Arranging for appellate review always becomes more complicated in proceedings that are moved from one place in the country to another. Proposed [sections] 1335A takes that sort of complexity to a new level because transfer by the Panel can be to a state court,(83) and issues that relate only to individual claims may be transferred to any court competent to entertain the issues.(84) Thus, dispersed appeals could be a possibility, as might review by a state appellate court (if a state court is the mass tort court) of rulings by an Article III judge (if certain issues are transferred to a federal court for resolution in connection with a mass tort action entertained by a state court judge). The proposed statute resolves these issues in a provision on appeals.(85) In this way, Professor Cooper tries to avoid results that would be "outside most concepts of our federal structure,"(86) although he notes that some of the results would be "startling."(87) For present purposes, the problem of appeal points out another? way in which adopting this approach would result in radical revision of our current system.

6. It contemplates requiring the court to identify and appoint as class representatives "the most capable representatives available."

Seemingly taking the lead of the Private Securities Litigation Reform Act,(88) the new Rule 23.3 proposed to accompany new [sections] 1335A directs the court to make a similar effort.(89) This is an understandable objective given the stakes, but it is difficult to see how the court is to do this job. In securities fraud actions, the starting point is to give a preference to the party with the largest financial stake in the suit.(90) How could the court develop a similar rule of thumb in mass tort actions? Among those injured, claimants likely to have the largest claims might be sensible choices, although one would expect them to be the sickest. For a class of "future" plaintiffs who may not be represented by those currently injured,(91) it is even more difficult to determine which claimant should be selected. Yet including this requirement presumably makes failure to select the best person an argument for opposing a settlement. Again, the requirements for this extraordinary new proceeding emphasize the dramatic nature of its changes.

In sum, not only would significant objectives sought by many concerned with recent class action practice not be accomplished by this proposal,(92) but its adoption would raise difficult and potentially dangerous new issues.(93) Accordingly, benign neglect might be the better course and a more modest program of revising Rule 23 adopted instead.

III. THE RULE 23 ADJUSTMENT

Without any action by Congress, the Advisory Committee could propose to modify Rule 23 further. It circulated proposed changes to the rule in 1996.(94) Armed with the insights gained from that experience, the Advisory Committee could make different proposals, and Professor Cooper has suggested several.(95) As with the statutory proposal, one can look to what these proposals do and do not accomplish to determine whether they provide a reason for action.

A. What the Proposed Rule Amendments Do Not Do

1. The amendments place no constraint on class certification by state courts.

At most, federal rulemakers can alter practices in federal court. But to the extent federal courts become less hospitable to activities deemed abusive by some, those who wish to pursue similar objectives may simply shift to state court. There is some reason to think that they have been doing so.(96) There is also some reason to believe that certain states--perhaps even certain specific counties(97)--are magnets for such activity.

Whether this side effect of federal rule revision should be considered important is debatable. Federal rulemakers hardly can constrain their improvements in federal practice due to perceived deficiencies in state court practice. Moreover, there is no reason to think that states themselves are incapable of changing their ways. Alabama, for example, recently enacted legislation to ensure that "drive-by certification" would no longer occur within its borders.(98) So even if this factor has weight, it may be that benign neglect is as effective as a more active stance.

2. The amendments offer no direction about whether to certify a settlement class.

The proposed amendment to Rule 23(b) in effect "promulgates" the result in Amchem without providing any guidance about how to apply it. At least some believe that Amchem could be improved upon,(99) and one could certainly want more direction about how it should be applied. The proposed amendment offers nothing of the sort, however, in marked contrast to the detailed treatment of settlement approval proposed to be added to Rule 23(e).(100) Since the amendments effectively do nothing that Amchem has not already done, one could easily conclude that there is no need to make them.

3. The amendments offer no guidance on "limited fund" certification under Rule 23(b) (1) (B) in the mass tort setting.

Some of us have long questioned whether the limited fund concept could ever be adapted to mass torts.(101) Clearly the Court viewed limited fund certification as central to its decision in Ortiz.(102) Clearly Chief Justice Rehnquist thought that revisions to the rule could have made a difference in the answer reached by the Court--that Rule 23(b) (1) (B) does not in current form permit the sort of certification done by me judge.(103) So one purpose of amendment might be to revise the rule to facilitate limited fund certification in circumstances like Ortiz, but that would entail confronting a variety of thorny policy issues.

An alternative objective would be to clarify the criteria discussed in Ortiz. On that score, it seems to this reader that the Court's opinion raises more than the usual number of questions that might be addressed by amendments to the rule. Undoubtedly others have their own favorites, but here is a partial list:

   (1) Should limited fund certification be forbidden altogether in mass tort
   litigation?(104)

   (2) By what standard should the court determine whether' there is a limited
   fund?(105)

   (3) If the "net assets plus insurance" approach may be used to determine
   the available :assets, how is the court to make the decision what those
   assets total?(106)

   (4) How should the court determine the total of probable claims?(107)

   (5) Should claims for punitive damages be handled differently in the
   limited fund scenario?(108)

   (6) If limited fund treatment is proper in a mass tort situation, how
   should the fund be divided?(109)

   (7) What claimants have to be included?(110)

   (8) Can the debtor keep part of its assets because of the transaction costs
   saved by the combined resolution?(111)

   (9) How does the right to jury trial bear on limited fund class action
   settlements?(112)

Obviously not all of these questions are susceptible to resolution by rule amendment. The Court's opinion, however, is couched in terms of the intentions of the rulemakers in 1966, and new intentions could lead to different results. At least Chief Justice Rehnquist seems confident that rule amendments could change the status quo regarding the issues addressed by the Court.(113) Thus that possibility seems worthy of attention.

4. The amendments do not eliminate the problem of the reverse auction.

By writing settlement classes into the rule without any specification about how the court should approach them differently (if at all), the proposal does not address the problem of a reverse auction.(114)

5. The amendments only implicitly disavow the 1966 Advisory Committee Note disapproving mass accident class actions.

In 1966, the rulemakers foresaw the use of the class action in mass accident cases and tried to fend off that possibility with disapproving language in the Committee Note.(115) Under the pressures of mass tort litigation, the courts began in the 1980s to relax their opposition to mass tort class actions. By then, at least, some of those connected to the 1966 Advisory Committee Note had disavowed their prior attitudes.(116) One does not amend rules to bury Notes, but given the prominence of this particular Note some official disavowal might be urged. Amchem itself seems to have done the job,(117) however, so no further action is required, and in any event a retreat from the 1966 Note seems implicit in the amendments proposed.

B. What the Proposed Rule Amendments Do

1. The amendments provide enhanced settlement evaluation guidance in Rule 23(e).

Although the framers of the 1966 amendments to Rule 23 clearly intended to permit settlement of class actions, and to make such settlements binding on class members, they provided courts with almost no guidance for deciding whether to approve settlements. Fortifying this provision therefore seems a relatively obvious improvement on the rule.(118)

The proposed amendments ensure that there will be a hearing on the proposed settlement.(119) They provide that any class member may object, and then obtain "discovery reasonably calculated to appraise the apparent merits of the class claims, issues, or defenses, and to disclose the course of settlement negotiations and the terms of any incidental agreements or understandings."(120) Whether this discovery should reach all factors that the rule says bear on settlement could present some interesting issues, for those factors include "the total resources available to the parties agreeing to pay money under the settlement"(121) and whether the attorneys fees proposed to be paid are reasonable.(122) The amendments also guarantee an award of attorneys fees for a successful objection to the settlement and permit such an award in connection with an unsuccessful objection.(123) Whether this provision would provoke too many objections is debatable. There were certainly vigorous objectors to the settlements in Amchem and Ortiz, but it is less clear whether objectors usually have been sufficiently willing to get involved. In addition, the proposed amendment permits the court to refer the initial appraisal of the settlement to a magistrate judge or a person specially appointed for an independent investigation.(124) Some courts have appointed a guardian ad litem for a class in the settlement context, and it is good to see this idea included in the proposed rule.

Besides dealing with the procedure for settlement approval, the proposed amendments list factors that the court should consider "among other factors" in deciding whether to approve a proposed settlement. Here maturity is the first consideration on the list, as it bears on "the ability to assess the probable outcome of a trial on the merits."(125) This should go a considerable distance toward resolving the concerns raised about the 1996 proposed amendments,(126) which included maturity as a factor that might work mischief in antitrust, employment discrimination, or securities fraud cases. Like a number of cases,(127) the factors call for a comparison of the likely outcome at trial with the risks :and costs of trial.(128)

As noted below, it may be that such provisions can be adequately spelled out by caselaw.(129) If a rule change is needed, however, these specifics seem a very good starting place for developing a new Rule 23(e).

2. The amendments create an absolute right to opt out after settlement notice is given in all class actions.

Presently, of course, the right to opt out exists only in Rule 23(b) (3) class actions, and it is not connected to settlement, :although it may be made available at the time of settlement if the class has not been certified theretofore. The proposed amendment radically enlarges the role of the opt-out, providing that a settlement of a class action binds a class member only after notice of the terms of the settlement is provided to the class member and a chance to opt out is afforded the class member.(130)

Some may believe recent developments to indicate that there must always be an opt-out right because due process permits binding effect only where one exists.(131) But none of the cases on which this view is based involved the sorts of claims asserted in traditional Rule 23(b) (1) and (b)(2) classes, and it is hard to believe that these class actions are necessarily burdened with ensuring a right to opt out to make them constitutional.

Such a right would seem impossible to implement in some (b) (2) situations.(132) In a school desegregation class action, for example, what exactly does it mean to say that some members of the plaintiff class have opted out? Are they then exempt from the remedy developed by the court? In cases involving structural injunctions, this sort of solution will not work. True, there may be serious disagreements within the class about what remedy to seek--such as whether plaintiffs should prefer busing or other measures to deal with the problems caused by past discrimination.(133) Some may feel that these suits are artifacts of the past, but they are still being filed.(134) There are serious drawbacks to allowing opting out of (b) (2) classes that are settled.

Similar difficulties would result from allowing opting out in (b) (1) class actions. The central reality is that Rule 23(b) (1) provides an adjunct to Rule 19 for cases in which the necessary parties are too numerous to be joined. But the thrust of Rule 19 is to require that the absent parties be joined if it is feasible to do so, not just if they want to be.(135) Allowing them a right to opt out is inconsistent with the motivation for (b)(1) treatment. Recall the motivation for using this provision in mass tort cases--intercepting early punitive damage claims that would siphon off funds otherwise available for paying compensatory damages.(136) Although there have been extraordinary cases in which opting out in (b) (1) class actions has been allowed,(137) the idea is incompatible with the reason for having a mandatory class in the first place.

Moreover, allowing a right to opt out in the proposed rule amendment is jarringly different from the treatment of similar issues under the [sections] 1335A proposal discussed above. The statute would direct that all present and future claimants, defendants, and insurers be joined and that none can opt out. If due process requires a right to opt out, it is surely curious to suggest a statute that would ensnare the world.

Yet another curious feature of this right to opt out is that it exists only in the event of settlement. What exactly does that mean? In (b)(2) actions, there may be considerable litigation before such a settlement emerges. Consider Martin v. Wilks,(138) in which there was very vigorous litigation of employment discrimination claims against the City of Birmingham, Alabama, leading ultimately to a settlement about the decree that should be entered to undo the discriminatory practices the court had found to exist. There is some reason to think that this sort of negotiated decree is common in structural injunction cases, where the court lacks expertise to design the decree and the law condemns certain practices but does not prescribe a specific solution to the problems they create.(139) Martin v. Wilks dealt with the question whether the resulting decree could bind whites who were not included in the class, but did not suggest that the class members themselves were entitled to a right to opt out after the settlement was announced. Some of those class members also opposed the settlement;(140) would class members under the new rule have an automatic right to opt out in such a case? How then would the city handle them in its hiring and promotion decisions? Imposing the opt-out right on settled (b)(2) actions seems likely to work considerable mischief.

The settlement opt out raises interesting questions in connection with (b) (3) damage actions as well. If the critical objective of the opt-out right is to make the binding effect of the decree consensual, one could also ask why the opt-out right does not apply if the case is fully litigated. Under current Rule 23(c)(2), of course, there is a right to opt out even though the case was certified for litigation, albeit not one required to be made available after a settlement is announced.(141) Perhaps the notion is that if a damages class action is suitable for litigation we have no reason to worry about the remedy afforded after trial because that is our best approximation of what the parties should receive. If the case is certified for full litigation, however, one would think that the reverse-auction problem(142) has largely been solved since the plaintiffs' lawyers have the threat of trial to back up their settlement negotiations. What reason is there to guarantee a second right to opt out if there is a settlement?

Finally, it is worth noting that the opt-out right dooms any settled class action involving future class members unless all of them in effect are to occupy an "opt in" status. There is obviously no way to notify them about the terms of the decree. Consider again the school desegregation case. Children who have not yet entered school (perhaps not even been born) would normally be included in a class consisting of present and future students in the school system. How will the decree work if each of them can opt out on reaching school age?

In sum, there are many reasons to approach the proposed opt-out provision with great caution.

3. The amendments authorize class certification solely for settlement purposes in (b)(1) and (b)(2) class actions.

The Advisory Committee's 1996 proposal for settlement class actions was limited to cases brought under (b)(3).(143) But mass tort class actions often are packaged under other portions of Rule 23. We have already seen the variety of questions that persist about using Rule 23(b)(1)(B) in such cases.(144) Such cases may also be packaged as (b)(2) class actions when they assert some sort of medical monitoring claim.(145) Broadening the settlement class authority in this way would capture these situations.

The problem is that it would capture a great deal more and disregard the fundamental idea of a mandatory class, a point analogous to the one made about allowing opting out in such cases.(146) The basic idea of a mandatory class is that everyone has to be joined in the same proceeding or the objectives of Rule 19 will be frustrated.(147) If that predicate is satisfied, class certification is mandatory in the sense that the court must do it for all purposes, not just settlement. So the entire idea of a settlement class seems at odds with the precepts of (b)(1), and probably (b)(2).

In sum, the rule amendment proposal offers promise of clarifying and improving practice under Rule 23(e), but other features of the proposal cause great uneasiness. Given that important rule-interpretation questions are omitted and presumably left to caselaw development, it may be that benign neglect is again the wisest course.

IV. SHOULD BENIGN NEGLECT AGAIN BE EMBRACED?

Certainly the state of race relations in this country does not indicate that benign neglect was the best policy for them. This paper raises devil's advocate arguments for persisting in the 1978 decision of the Advisory Committee not to propose substantial changes for Rule 23, and leaving the question to Congress.(148) It attempts to identify reasons for considering Professor Cooper's proposals for legislation or rule amendment insufficient--discussing the features they lack--as well as reasons for thinking them overkill--exploring the possible disadvantages of the features they have.

Beyond those concerns, there are other considerations that one trying to decide whether to persist in a policy of benign neglect might think important. Accordingly, the paper will end with a list of some of those:

1. Can the desirable changes be effected without passing legislation or amending Rule 23?

Obviously some cannot, or at least cannot easily. For the federal courts to curtail inappropriate state court class certification, for example, requires legislation absent a constitutional ground for invalidating such class certification.

Other positive developments might be achievable by caselaw. The proposed amendment to Rule 23 to provide directives about how to handle the question of class settlement approval(149) would replace the delphic current rule language with much helpful detail. But there is no longer a clean slate in interpreting Rule 23(e). The courts have been working out the ways to handle it since 1967, when a district court in Philadelphia first confronted the question whether court approval was required in a settled case in which a class had not yet been certified.(150) This effort has yielded may opinions since; Professor Cooper thus correctly cites Judge Scirica's thoughtful discussion of these problems in a 1998 case.(151) Amending the rule would not throw all this judicial experience out the window; the proposed amendment lists factors that may be pertinent "among other factors."(152) But one could at least ask whether there is a need to revise a rule when there is already a body of interpretive (and creative) caselaw, as well as considerable treatment of the problems in the treatises.(153)

2. Is it desirable to undermine the transsubstantive orientation of the rules in this way?

The rules' transsubstantive orientation has come under much attack recently, but it is still a fundamental feature of the overall scheme. As Professor Carrington has reminded us in the pages of this journal,(154) there are profound risks associated with making rules for only one kind of case. But that is clearly what the all-encompassing solution proposed by Professor Cooper would do. Indeed, his proposed new Rule 23.3 would override "inconsistent" provisions of other Civil Rules as well. That sort of thing is not unprecedented,(155) but might wisely be avoided. On the other hand, including all types of actions under the proposed amendment to Rule 23 creates problems of its own by threatening undesirable results in cases that are not mass torts.(156)

3. Should the Judicial Branch abandon the field?

It is too late in the day to believe that the Judicial Branch can stop procedural development by deciding that it is unwarranted. If the Civil Justice Reform Act(157) did not prove that, the Private Securities Litigation Reform Act(158) should have. So the pleasant notion that the transsubstantive ideal can be pursued without regard to political realities should perhaps be banished.

Leadership and sage counsel are extremely important in the difficult activity of fashioning appropriate methods of addressing these problems. One could easily say that Congress has evinced some willingness to move quickly with ,dubious solutions. For the Advisory Committee (or other pertinent parts of the Judicial Branch) to foreclose participation could be a very large mistake.

4. Is it too soon to act?

It is hard to identify a congeries of procedural problems that has received more attention over the last decade than mass torts. One might even suggest that this mass torts symposium be the last one for a while. Yet we must also keep in mind what Judge Becker said about these problems in 1998: "In my 27 years on the bench, I have never seen an area in as much ferment as this class action area is, this mass tort and consumer class action is now."(159) Trying to fashion final solutions to these problems before the ferment settles down further may be unwise.

5. Are we fighting the last war?

Writing about the class action tumult in 1979, Professor Miller enumerated the types of cases in which class actions were thought to matter--antitrust, civil rights, consumer, environmental, and the like--but there was no mention of personal injury tort suits.(160) That was not surprising given the antagonistic Advisory Committee Note,(161) but it should give us pause. The tocsin for using class actions in mass torts was sounded by Judge Williams(162) only two years later, yet Miller (then Reporter to the Advisory Committee on the Civil Rules) did not see this problem coming. The problems that weighed so heavy in 1979, meanwhile, had largely disappeared a decade later. How confident can we be that today's problems will persist and that there is not something new just around the corner that will preoccupy us in ten years?

6. Unless we act on what we have learned, will that mean that all this effort was in vain?

The Advisory Committee has been working on the issue of class actions and mass torts for almost a decade. It has sponsored a number of conferences, issued proposed amendments that generated four volumes of analysis,(163) and prompted the formation of the Mass Torts Working Group, which issued its own extensive report with a large appendix.(164) The process has involved innumerable lawyers, judges, and law professors.

Even if no statute or rule change results from this activity, it has provided immense value to the handling of these cases. For one thing, the work product of this activity provides an invaluable resource for those who now and later address these questions. For another, the process of amassing this database has disseminated and improved the knowledge available. The spread and improvement of knowledge may be the most important way of solving the present problems, rather than legislation or rulemaking. So if benign neglect prevails, and neither new rules nor new statutes emerge, there is nevertheless reason for the Advisory Committee to think that this has been a job well done.

(1) See Edward H. Cooper, Aggregation and Settlement of Mass Torts, 148 U. PA. L. REV. 1943 (2000).

(2) See Peter Kihss, "Benign Neglect" on Race Is Proposed by Moynihan, N.Y. TIMES, Mar. 1, 1970, at 1 (reporting that Moynihan told an interviewer that he drew the phrase from a report by the Earl of Durham that concluded that Canada had become more competent in governing itself "through many years of benign neglect").

(3) See Text of the Moynihan Memorandum on the Status of Negroes, N.Y. TIMES, Mar. 1, 1970, at 69.

(4) Moynihan's Memo Fever, TIME, Mar. 23, 1970, at 15.

(5) Mr. Moynihan's Apostasy, WALL ST. J., Mar. 13, 1970, at 8.

(6) Id.

(7) See Anthony Lewis, Neglect--Benign or Hostile ?, N.Y. TIMES, Mar. 2, 1970, at 36.

(8) Editorial, Neglect--But Not "Benign, "N.Y. TIMES, Mar. 3, 1970, at 40.

(9) It appears that the memorandum was discussed in Congress, and it is reproduced at least twice in the Congressional Record. See 116 CONG. REC. 6070-71 (1970) (reproducing the text printed in the N.Y. TIMES, supra note 3, at 69); 116 CONG. REC. 6124-25 (1970) (reproducing the same text); see also 116 CONG. REC. 7339 (1970) (reproducing Mr. Moynihan's Apostasy, supra note 5).

(10) See 7A CHARLES ALAN WRIGHT, ARTHUR R. MILLER & MARY K. KANE, FEDERAL PRACTICE AND PROCEDURE [sections] 1775, at 470 (2d. ed, 1996) [hereinafter WRIGHT, MILLER & KANE] ("[S]ubdivision (b)(2) was added to Rule 23 in 1966 primarily to facilitate the bringing of class actions in the civil rights area.").

(11) Of course, they expressed their opposition to that sort of class action. See infra notes 115-17 and accompanying text for a more complete discussion.

(12) Pub. L. No. 90-296, 82 Stat. 109 (1968) (codified as amended at 28 U.S.C. [sections] 1407 (1994)).

(13) See infra note 49 and accompanying text (discussing FED. R. CIV. P. 23(f)).

(14) Of course, Congress has acted in the specialized area of securities fraud litigation. See Private Securities Litigation Reform Act of 1995, Pub. L. No. 104-67, 109 Stat. 737 (codified as amended in scattered sections of 15 U.S.C.) (modifying some aspects of securities-fraud class actions).

(15) JUDICIAL CONFERENCE OF THE U.S., REPORT OF THE AD HOC COMMITTEE ON ASBESTOS LITIGATION 19 (1991).

(16) See id. at 32-34.

(17) 527 U.S. 815, 119 S. Ct. 2295 (1999).

(18) Id. at 2324 (Rehnquist, C.J., concurring) (quoting Justice Souter's majority opinion, 119 S. Ct. at 2302).

(19) Id.

(20) Id.

(21) The commencement of this review in connection with mass torts was in 1991. See Robert G. Bone, Rule 23 Redux: Empowering the Federal Class Action, 14 REV. LITIG. 79 (1994) (describing 1991-1994 considerations of the Advisory Committee about amending the rule). In 1996, the Committee produced proposed amendments to Rule 23 that were published for public commentary. See Preliminary Draft of Proposed Amendments to the Federal Rules of Civil Procedure, 167 F.R.D. 523, 559-66 (1996). After the period of public comment on those proposals was completed, Chief Justice Rehnquist appointed a Working Group on Mass Torts consisting of representatives of the Civil Rules Advisory Committee and a number of other Judicial Conference committees, and that group submitted a report to the Chief Justice on February 15, 1999, recommending the appointment of an Ad Hoc Committee on Mass Torts to continue this work. See ADVISORY COMM. ON CIVIL RULES & WORKING GROUP ON MASS TORTS, REPORT ON MASS TORT LITIGATION 69-70 (Feb. 15, 1999) [hereinafter REPORT ON MASS TORT LITIGATION]. As of this writing, the Chief Justice has not appointed such a committee.

(22) See Stephen Labaton, Asbestos Cases in for Overhaul by Lawmakers, N.Y. TIMES, June 28, 1999, at A1 (reporting that Congress would hold hearings on how to deal with the problems presented by asbestos litigation).

(23) Arthur R. Miller, Of Frankenstein Monsters and Shining Knights: Myth, Reality, and the "Class Action Problem, '92 HARV. L. REV. 664, 664 (1979).

(24) Milton Handler, The Shift from Substantive to Procedural Innovations in Antitrust Suits--The Twenty-Third Annual Antitrust Review, 71 COLUM. L. REV. 1, 9 (1971). It is worth noting that this phrase continues to be used by class action opponents. See L. Stuart Ditzen, Carmaker Sues Over Liability Lawsuit, PHILA. INQUIRER, Nov. 12, 1999, at B1 (reporting on a suit filed by an automobile manufacturer against lawyers who had earlier sued the carmaker, and quoting a representative of the manufacturer who said that "[f]or too long, trial lawyers have been exploiting class actions, turning these lawsuits into a form of legalized blackmail").

(25) Eisen v. Carlisle & Jacquelin, 391 F.2d 555, 572 (2d Cir. 1968) (Lumbard, C.J., dissenting).

(26) See Stephen Berry, Ending Substance's Indenture to Procedure: The Imperative for Comprehensive Revision of the Class Damage Action, 80 COLUM. L. REV. 299, 299 (1980) (critiquing class damage action procedure).

(27) See id. at 302-12.

(28) See Miller, supra note 23, at 684 & n.86.

(29) For a discussion of this proposed legislation, see Berry, supra note 26, ar 321-43.

(30) See Zahn v. International Paper Co., 414 U.S. 291, 301 (1973) (holding that in state-law class action, the claim of each class member must satisfy the amount-in-controversy requirements of 28 U.S.C. [sections] 1332 (a)). The Supreme Court granted certiorari to decide whether the adoption of 28 U.S.C. [sections] 1367 nullified Zahn, but affirmed when Justice O'Connor recused herself and neither side could muster a majority to resolve the question. See Free v. Abbott Lab., 51 F.3d 524, 529 (5th Cir. 1995) (holding that [sections] 1367 overturned Zahn and allowed a district court to exercise supplemental jurisdiction over members of a class although they did not meet the amount-incontroversy requirement, as did the class representatives), aff'd by an equally divided court, 120 S. Ct. 1578 (2000). Other courts have held that [sections] 1367 did not nullify Zahn. See, e.g., Meritcare Inc. v. St. Paul Mercury Ins. Co., 166 F.3d 214 (3d Cir. 1999). For discussion, see James Pfander, Supplemental Jurisdiction and Section 1367: The Case for a Sympathetic Textualism, 148 U. PA. L. REV. 109, 148-49 (1999).

(31) See Eisen v. Carlisle & Jacquelin, 417 U.S. 156, 175-76 (1974) (holding that all identifiable class members had to be notified of the suit individually by first class mail rather than by publication).

(32) See Coopers & Lybrand v. Livesay, 437 U.S. 463, 476-77 (1978) (rejecting the "death knell" doctrine that some courts of appeals had used to justify immediate appellate review of district court denial of class certification, and holding that review would be available under 28 U.S.C. [sections] 1291 only after entry of final judgment in the individual action remaining after denial of class certification, even if that would never happen because denial of class status was indeed the death knell of the suit). (33) Miller, supra note 23, at 678-79.

(34) Id. at 679.

(35) Id. at 680.

(36) DEBORAH R. HENSLER ET AL., CLASS ACTION DILEMMAS: EXECUTIVE SUMMARY 1 (1999).

(37) Douglas Martin, The Rise and Fall of the Class Action Lawsuit, N.Y. TIMES, Jan. 8, 1988, at B7.

(38) Judge Williams's forecast anticipated the views that would become widespread by the end of the 1980s:

      The latter half of the twentieth century has witnessed a virtual
   explosion in the frequency and number of lawsuits filed to redress injuries
   caused by a single product manufactured for use on a national level.
   Indeed, certain products have achieved such national notoriety due to their
   tremendous impact on the consuming public, that the mere mention of their
   names--Agent Orange, Asbestos, DES, MER/29, Dalkon Shield--conjure images
   of massive litigation, corporate stonewalling, and infrequent yet
   prevalent, "big money" punitive damage awards.

      In a complex society such as ours, the phenomenon of numerous persons
   suffering the same or similar injuries as a result of a single pattern of
   misconduct on the part of a defendant is becoming increasingly frequent.

In re Northern Dist. of Cal. "Dalkon Shield" IUD Prod. Liab. Litig., 526 F. Supp. 887, 892 (N.D. Cal. 1981), rev'd, 693 F.2d 847 (9th Cir. 1982).

(39) Id. at 893.

(40) Spencer Williams, Mass Tort Class Actions: Going, Going, Gone?, 98 F.R.D. 323, 325 (1983).

(41) Cooper, supra note 1, at 1944.

(42) See Richard L. Marcus, They Can't Do That, Can They? Tort Reform via Rule 23, 80 CORNELL L. REV. 858, 859-66 (1995) (discussing the substantive undercurrent in federal mass tort class actions).

(43) In re Asbestos Prod. Liab. Litig. (No. VI), 771 F. Supp. 415 (J.P.M.L. 1991).

(44) See MICHAEL GREEN, BENDECTIN AND BIRTH DEFECTS: THE CHALLENGES OF MASS TOXIC SUBSTANCES LITIGATION 211-12 (1996) (describing certification of a mandatory class in Bendectin litigation as a "judicial accommodation" to the settlement that was embodied in a "hastily-drawn order" even though "there was little legal basis for the certification of a rule 23(b)(1) class").

(45) See supra note 33 and accompanying text.

(46) See Ortiz v Fibreboard Corp., 527 U.S. 815, 119 S. Ct. 2295 (1999); Amchem Prods., Inc. v. Windsor, 521 U.S. 591 (1997).

(47) See, e.g., Castano v. American Tobacco Co., 84 F.3d 734 (5th Cir. 1996) (overturning certification of nationwide class of allegedly addicted smokers); In re American Med. Sys., Inc., 75 F.3d 1069 (6th Cir. 1996) (overturning certification of nationwide class of recipients of defendant's allegedly defective penile implants); In re Rhone-Poulenc Rorer Inc., 51 F.3d 1293 (7th Cir. 1995) (granting writ of mandate to overturn certification of hemophiliacs infected with HIV virus in suit against manufacturers of blood solids used by hemophiliacs that supposedly were source of infection); cf. Valentino v. Carter-Wallace, Inc., 97 F.3d 1227 (9th Cir. 1996) (overturning class certification on facts before it in nationwide suit alleging that pharmaceutical was defective, but refusing to say that certification could never be proper for multistate personal injury class).

(48) See e.g., In re Rhone-Poulenc Rorer Inc., 51 F.3d at 1294 (granting writ of mandate over objections of dissenter that the extraordinary circumstances warranting use of the writ were not present).

(49) FED. R. CIV. P. 23(f) (authorizing immediate appeal of orders denying or granting class certification at discretion of court of appeals).

(50) See Lexecon Inc. v. Milberg Weiss Bershad Hynes & Lerach, 523 U.S. 26, 39 (1998) (holding that transferee courts under 28 U.S.C. [sections] 1407 may not engage in self-transfer pursuant to 28 U.S.C. [sections] 1404(a)).

(51) State Farm Fire & Cas. Co. v. Tashire, 386 U.S. 523, 537 (1967).

(52) In re School Asbestos Litig., 789 F.2 [sections] 96, 1000 (3d Cir. 1986).

(53) Ortiz v. Fibreboard Corp., 527 U.S. 815, 119 S. Ct. 2295, 2303 (1999).

(54) The standard is described in two different ways. The statutory proposal says that the new procedure can be invoked by a defendant "against whom more than 100 persons have asserted claims." Cooper. supra note 1, at 1953 (proposing 28 U.S.C. [sections] 1335A(a)). But Professor Cooper's description says that the new procedure would not be appropriate unless 100 actions were filed. See id. at 1960 (arguing that the 100-action minimum is insufficient to merit use of the "heavy artillery" available in the statute).

Either approach has difficulties. Invoking the procedure as soon as 100 claimants have asserted claims against the defendant could mean that the proposed mechanism is available as soon as a class action involving a class of 100 or more class members is filed, or at least when the class is certified. Requiring that 100 separate suits be filed would prevent use of the procedure where ten overlapping class actions we. re filed on behalf of thousands, of even millions, of claimants.

(55) See id. (explaining that "there is nothing that prevents class actions filed by plain tiffs or other forms of aggregation from becoming mass tort[s].... [because i]t is the defendant's choice whether to invoke this new one-court, mandatory aggregation system").

(56) See Andrew Pollack, Toshiba Faces $1 Billion Bill Over a Lawsuit, N.Y. TIMES, Oct. 30, 1999, at A1 (noting that according to Toshiba executives, potential liability could have exceeded $9 billion).

(57) See Evan Ramstad & David P. Hamilton, Toshiba's $2.1B Settlement Inspires "Copy cat" Suits, S.F. DAILY J., Nov. 3, 1999, at 7.

(58) Cf. Janet Cooper Alexander, Do the Merits Matter? A Study of Settlements in Securities Class Actions, 43 STAN. L. REV. 497, 514-19 (1991) (finding that settlements in certified securities class actions almost always are for essentially the same proportion of possible damages, indicating that they are not motivated by the merits of the claims, which cannot all be of the same strength). But cf. Joel Seligman, The Merits Do Matter, 108 HARV. L. REV. 438, 453 (1994) (questioning Alexander's methods and conclusions).

(59) See supra notes 3840 and accompanying text.

(60) See Peter M. Schuck, The Worst Should Go First, 15 HARV. J.L. & PUB. POL'Y 541, 561 (1992) (urging that plaintiffs suffering the worst injuries be given precedence in the court system).

(61) See Cooper, supra note 1, at 1954 (delineating the provisions of 28 U.S.C. [sections] 1335A(f) as proposed).

(62) See John C. Coffee, Jr., Class Wars: The Dilemma of the Mass Tort Class Action, 95 COLUM. L. REV. 1343, 1367-84 (1995) (discussing risk of collusive settlements). For further elaboration on this and related topics, see John C. Coffee, Jr., Class Action Accountability: Reconciling Exit, Voice, and Loyalty in Representative Legislation, 100 COLUM. L. REV. 370 (2000).

(63) See Cooper, supra note 1, at 1953-54 ([sections] 1335A(d)).

(64) See id. at 1955 ([sections] 1335A(i)).

(65) See id. at 1954 ([sections] 1335A(f)).

(66) See id. at 1955 ([sections] 1335A(l)).

(67) Amchem Prods., Inc. v. Windsor, 521 U.S. 591,624 (1997).

(68) Castano v. American Tobacco Co., 84 F. 3d 734, 737 (5th Cir. 1996).

(69) Note that if the defendant invoking the procedure is a small player in a large industry, it may not be a party to a substantial proportion of the suits, thus making the process of identifying claimants more difficult even as to current plaintiffs.

(70) See, e.g., In re American Tobacco Co., 880 F.2d 1520 (2d Cir. 1989) (addressing discovery dispute resulting from filing of actions in state courts in many states against tobacco companies claiming injury due to a combination of tobacco smoking and exposure to asbestos).

(71) But see supra note 54 (discussing the difference between requiring that there be 100 claimants or 100 suits).

(72) See Cooper, supra note 1, at 1960.

(73) See comment 4 below in this section for a discussion of the potential for bluffing by defendants about use of the new procedure.

(74) Cooper, supra note 1, at 1962 (footnote omitted).

(75) See infra notes 118-29 and accompanying text. Compare also the list of factors that bear on superiority for common question class actions under the current rule. See FED. R. CIV. P. 23(b)(3).

(76) See AMERICAN LAW INST., COMPLEX LITIGATION: STATUTORY RECOMMENDATIONS AND ANALYSIS [sections] 6.01 (1994) (proposing choice-of-law rules for mass tort cases).

(77) See Cooper, supra note 1, at 1955 ([sections] 1335A(j)). This directive does not apply to interpretation of insurance policies.

(78) Section 1335A(e), Cooper, supra note 1, at 1954, allows rulemaking subject to 28 U.S.C. [subsections] 2073-2074. Notably absent are the limits in 28 U.S.C. [sections] 2072(b), which forbid rules that modify substantive rights.

(79) But compare Amchem, 521 U.S. at 610 n.14, in which the Court noted that the statewide average for mesothelioma recoveries in California is more than twice the maximum allowed under the payment schedule in the settlement agreement. This observation introduces even more uncertainty about the differences between states by suggesting that one should consider not only the variations in legal doctrine but also the habits of juries in different states.

(80) See supra notes 3843 and accompanying text (describing preference to subordinate punitive damages to compensatory damages and to ensure that the worst go first).

(81) In re School Asbestos Litig., 789 F. 2d 996, 1001 (3d Cir. 1986) ("A forum wishing to take the long-range view might find that its efforts were not only ineffective but unfair to its citizenry because claimants in the other states could drain off all the assets available for satisfaction of claims.").

(82) Cooper, supra note 1, at 1960.

(83) See id. at 1953 ([sections] 1335A(c)(4)).

(84) See id. at 1954 ([sections] 1335A(g)).

(85) See id. at 1955-56 ([sections] 1335A(m)).

(86) Id. at 1971.

(87) Id.

(88) See 15 U.S.C. [sections] 78u-4(a) (3) (B) (i) (Supp. II 1996) (directing the court to appoint as lead plaintiff in securities fraud class action the person "most capable of adequately representing the interests of class members").

(89) See Cooper, supra note 1, at 1958 (Proposed Rule 23.3(d)(2)) (requiring the court to "assure that claimants designated as representatives are the most capable representatives available from the represented class or subclass").

(90) See 15 U.S.C. [sections] 78u-4(a) (3) (B) (iii) (I) (bb) (creating a rebuttable presumption in favor of the person with the largest financial stake).

(91) See Cooper, supra note 1, at 1957 (Proposed Rule 23.3(d) (1)) ("Claimants whose claims have not yet accrued may not be included in a class with claimants who have present claims.").

(92) See supra notes 52-62 and accompanying text.

(93) See supra notes 63-91 and accompanying text.

(94) See Preliminary Draft of Proposed Amendments to the Federal Rules of Appellate, Civil, and Criminal Procedure, 167 F.R.D. 523, 559-66 (1996) [hereinafter Prelimary Draft].

(95) See Cooper, supra note 1, at 1981-89.

(96) Thus, the RAND Corporation reported as follows in 1999:

   We found no quantitative data to permit us to calculate growth trends. But
   we are persuaded by our interviews with plaintiff and defense counsel that
   there has been a surge in damage class actions in the past several years,
   particularly in state courts and in the consumer area. Many practitioners
   trace this growth to the curbs on securities litigation enacted by Congress
   [in the Private Securities Litigation Reform Act] in 1995. Faced with these
   curbs, they say, plaintiff attorneys looked for new types of suits to bring
   and found their opportunities in consumer complaints against business
   practices and products.

HENSLER ET AL., supra note 36, at 5.

(97) During the Mass Torts Symposium at the University of Pennsylvania on November 11-12, 1999, it was suggested that there are around a dozen identifiable counties in the country that are magnets for this sort of litigation.

(98) 1999 Ala. Acts 99-250 provides that a class action may not be certified without an evidentiary hearing if any party requests one, and affords sixty days' notice of the hearing. It also directs Alabama courts to require the proponent of certification to satisfy all the requirements of Alabama Rule 23, which is very similar to the Federal Rule 23.

It is clear, however, that previously some were very anxious to remain in Alabama state court. See, e.g., Davis v. Carl Cannon Chevrolet-Olds, Inc., 182 F.3d 792, 793 (11th Cir. 1999) (involving class action in Alabama state court containing what the court calls a "do not remove me disclaimer" in capital letters).

(99) See, e.g., Mark C. Weber, A Consent-Based Approach to Class Action Settlement: Improving Amchem Products, Inc. v. Windsor, 59 OHIO ST. L.J. 1155 (1998). This is an illustration, not a recommendation for adopting some specific proposal.

(100) See infra notes 118-29 and accompanying text.

(101) See Marcus, supra note 42, at 877-81 (detailing difficulties with applying Rule 23(b) (1) (B) in mass tort situations).

(102) Thus, the opinion begins by explaining that "[t] his case turns on the conditions for certifying a mandatory settlement class on a limited fund theory under Federal Rule of Civil Procedure 23(b)(1)(B)." Ortiz v. Fibreboard Corp., 527 U.S. 815, 119 S. Ct. 2295, 2302 (1999). Later the Court explained that "[t]he nub of this case is the certification of the class under Rule 23(b) (1) (B)." Id. at 2307.

(103) See supra note 20 and accompanying text.

(104) The Court went out of its way not to answer this question. It emphasized at one point that "[w]e do not, it is true, decide the ultimate question whether Rule 23(b) (1)(B) may ever be used to aggregate individual tort claims." 119 S. Ct. at 2314. It also said that "we cannot settle all the details of a subdivision (b) (1) (B) limited fund here (and so cannot decide the ultimate question whether settlements of multitudes of related tort actions are amenable to mandatory class treatment)." Id. at 2312. On the other hand, discussing the record before it, the Court said that "if Fibreboard's own assets would not have been enough to pay the insurance shortfall plus any claims in excess of policy limits, the projected insolvency of the insurers and Fibreboard would have indicated a truly limited fund." Id. at 2318. Therefore, the question seems open even under the current rule.

(105) The Court noted that the lower courts had developed two standards for demonstrating a limited fund--that the adjudication of individual claims "will inescapably compromise" the claims of others, or that there is a "substantial probability" such a compromise will occur. Id. at 2316 n.26. It declined to decide the appropriate standard because the case before it did not satisfy either. See id.

(106) The Court clearly held that the technique used in Ortiz was not proper, and that there must be a proceeding in the district court in which the proponents of limited fund treatment proffer evidence demonstrating the extent of assets available to pay claims. See id. at 2316. It observed that the insurance funds would be limited if there were aggregate limits in the policies, or if the insurers would obviously be rendered insolvent by success of many claimants. See id. at 2317. Since there was no such policy aggregate limit, nor evidence of looming insolvency due to open-ended liability for the insurers, these techniques did not work. The proponents of the settlement (and .Justice Breyer in dissent, see id. at 2326-29) urged that the settlement reached with the insurers established such a limit since it represented a discounted valuation of' the potential exposure that might be held unavailable altogether. The Court, however, held that the negotiation did not provide a basis for accepting this theory since the lawyers negotiating for the plaintiffs had a conflict of interest due to the fact that they would receive the proceeds of their separate settlement of pending claims only if the class settlement survived fairness scrutiny. See id. at 2317. The Court did not, however, hold that this discount method would always be improper in cases of disputes about insurance coverage. See id. ("This sense of limit as a value discounted by risk is of course a step removed from the historical model, but even on the assumption that it would suffice for limited fund treatment, there was no adequate finding of fact to support its application here.").

The additional question of noninsurance assets was not addressed by the Court, as the amount of those assets seemed not to be disputed. The certainty that they would not suffice should the insurance coverage disappear due to rulings in the California state courts in the insurance coverage litigation was also undisputed. It is worth noting, however, that in other cases setting a proper valuation for a going concern can be a challenge. In the A.H. Robins Dalkon Shield litigation, for example, Judge Williams treated as sufficient a showing in 1981 that the noninsurance net worth of the company was $280 million. Six years later, however, bankruptcy proceedings led to sale of the company for $2.475 billion, so the figure accepted by Judge Williams at least looks dubious. For discussion, see Marcus, supra note 42, at 878-79.

(107) In commercial cases, it may be appropriate to add up the claims in all the cases, but the Court recognized that this method would not work in mass tort cases. See Ortiz, 119 S. Ct. at 2316. As to both amount and likelihood of success on liability, there is no reason to believe that because claimant number one recovers $X claimant number n will recover some specific amount. See Marcus, supra note 42, at 878-79.

(108) Certainly punitive damages sometimes seem the wild card of this sort of litigation. However much difficulty there may be in forecasting probable compensatory recoveries, there is likely to be even more in forecasting punitive awards. Ortiz sheds no light on how to answer this question. It might be noted as well that the idea of subordinating punitive damages claims to compensatory damage claims in genuine limited fund situations has some appeal, but would almost certainly be beyond the power of the rules process.

Punitive damages also raise a separate problem of whether there is a legal limitation on total recoveries that could produce a different sort of limited fund situation. Courts have speculated that there may be some limit on the recurrent imposition of these damages, either as a matter of due process or as a matter of state law. Thus, some courts have used a "limited generosity" theory to justify treating the prospect of limits on punitive damage recoveries as warranting limited fund treatment. See, e.g., In re Agent Orange Prod. Liab. Litig., 100 F.R.D. 718, 728 (E.D.N.Y. 1983) (asserting that "[t]here must ... be some limit ... to the amount of times defendants may be punished for a single transaction"). Ortiz did not present this issue.

(109) I have previously observed that "Rule 23 (b) (1) (B) simply does not prescribe any standards for resolving competing claims to the limited fund, if one is found to exist." Marcus, supra note 42, at 880. In Ortiz, the Court seemed to find an implicit standard in the old cases it examined--"equity among members of the class." 119 S. Ct. at 2318. This appears to mean pro rata distribution. See id. at 2311-12, 2314 (justifying mandatory class treatment by referring to a limited fund, distributed on a pro rata basis, to satisfy claims based on a common theory of liability). So Ortiz does provide some light, but that raises the question of method of fixing the liabilities to be applied ratably against the limited fund, possibly implicating the right to jury trial issue raised as item (9) below.

(110) One of the problems with the limited fund class action, as compared to a bankruptcy, is that in effect it creates two classes of unsecured creditors. In bankruptcy, one would normally expect all unsecured creditors to share ratably in the debtor's assets. Based on its reading of cases using a limited fund technique before 1938, the Court said in Ortiz that it was permissible to limit the class to those who "might state a claim on a single or repeated set of facts, invoking a common theory of recovery, to be satisfied from the limited fund as the source of payment." Id. at 2311. In effect, then, there are two categories of unsecured creditors--those with claims based on a common theory or set of facts, and those whose claims are founded on other facts or theories. This approach leads to problems the Court did not address.

One problem is determining how far this circle of "common theory" claimants should be extended. In In re School Asbestos Litigation, 789 F. 2d 996 (3d Cir. 1986), the court rejected a limited fund class action for punitive damages brought on behalf of owners of properties who incurred costs removing asbestos. The claim was rejected in part because personal injury plaintiffs were not included, but they might seek punitive damages as well. Would this situation be treated as one in which the personal injury claimants and property damages claimants are invoking a "common theory of recovery?" Another problem is explaining how the limited funds of the debtor can be subdivided so that some are available for one class of "common theory" claimants while others are available for the rest. Unless the debtor can keep assets for the unsecured creditors excluded from the "common theory" group, it is difficult to see how those creditors can be kept out of the class action, or how a class action improves on handling the problem through a bankruptcy proceeding.

(111) Fibreboard contributed only about five percent of its net worth to the settlement pot. Yet the Court noted that in most previous cases on which it relied "the whole of the inadequate fund was to be devoted to the overwhelming claims." 119 S. Ct. at 2311. If that were done in a case like Ortiz there would have to be provision for all creditors, since otherwise the class would get all the assets and the other unsecured creditors would have none left for them. If the "limited fund" is all of the debtor's assets, bankruptcy would seem preferable to limited fund certification. The district court in Ortiz, however, found that the net worth of Fibreboard was less than the likely savings in defense costs resulting from rile settlement. See id. The Court did not entirely reject the idea that this argument could sometimes justify leaving equity in the debtor, although the Court found the concept inapplicable to the case before it. See id. at 2321 & n.35. If there is no way to leave something on the table, all the other unsecured creditors are deprived and there may be little incentive for companies to settle. Yet it is odd that companies should achieve discounted resolution of the class members' claims due to having limited assets while keeping significant portions of their assets. Perhaps only Chapter 11 bankruptcy treatment should permit this sort of situation to occur.

(112) The Court noted that "certification of a mandatory class followed by settlement of its action for money damages obviously implicates the Seventh Amendment jury trial rights of absent class members." Id. at 2314. It is difficult to understand how the Seventh Amendment bears on settlement, or why it should only affect settlement of mandatory class actions. Is there no similar limitation on the settlement of suits under Federal Rule of Civil Procedure 23(b)(2) for injunctive relief?. Is there a reason why the settlement of the scope of defendant's violation of plaintiffs' rights or the proper remedy is different when plaintiff has a right to jury trial? Unless one takes the view that, if the Seventh Amendment does not apply, there is no constitutional right to a "trial" at all, it is hard to understand why cases subject to the Seventh Amendment are different from those that are not (as would be true in many 23(b) (2) class actions).

(113) See supra note 20 and accompanying text.

(114) See supra note 62 and accompanying text.

(115)

   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 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.

FED. R. CIV. P. 23 advisory committee note to 1966 Amendments.

(116) Professor Charles Alan Wright summarized his views in court:

      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....

Transcript of Oral Argument, In re Asbestos Sch. Litig., 594 F. Supp. 178 (E.D. Pa. 1984), quoted in HERBERT B. NEWBERG & ALBA CONTE, NEWBERG ON CLASS ACTIONS [sections] 17.06, at 20 (3d ed. 1992).

Somewhat similarly, Judge Jack Weinstein offered the following observations about the Committee Note:

   As authority for this warning against attempts to use class actions in
   torts, the note cites an article II] wrote as a law professor. As a judge
   II have] been forced to ignore this indiscretion when faced with the
   practicalities of mass tort litigation. In the earlier 1960s we did not
   fully understand the implications of mass tort demands on our legal system.

Jack B. Weinstein & Eileen B. Hershenov, The Effect of Equity on Mass Tort Law, 1991 U. ILL. L. REV. 269, 288.

(117) The Court put the Note in its place:

   The Advisory Committee for the 1966 revision of Rule 23, it is true, noted
   that "mass accident" cases are likely to present "significant questions,
   not only of damages but of liability and defenses of liability, ...
   affecting the individuals in different ways." And the Committee advised
   that such cases are "ordinarily not appropriate" for class treatment. But
   the text of the Rule does not categorically exclude mass tort cases from
   class certification, and district courts, since the late 1970's, have been
   certifying such cases in increasing number.

Amchem, 521 U.S. at 625 (citations omitted).

(118) See William W Schwarzer, Settlement of Mass Tort Class Actions: Order Out of Chaos, 80 CORNELL L. REV. 837, 841 (1995) (noting that "Rule 23 contains no standards at all governing judicial approval of class action settlements" and arguing that it should).

(119) See Cooper, supra note 1, at 1981 (Proposed Rule 23(e)(2)) ("The court may not approve settlement ... without a hearing.").

(120) Id. (Proposed Rule 23(e) (4)).

(121) Id. at 1982 (Proposed Rule 23(e)(5)(D)).

(122) See id. (Proposed Rule 23(e) (5) (H)).

(123) See id. at 1981 (Proposed Rule 23(e) (4)).

(124) See id. at 1982 (Proposed Rule 23(e)(6)).

(125) Id. (Proposed Rule 23(e) (5) (A)).

(126) See Preliminary Draft, supra note 94.

(127) See, e.g., In re Corrugated Container Antitrust Litig., 643 F. 2d 195, 207 (5th Cir. 1981) (detailing standards for review of proposed settlement).

(128) See, e.g., Cooper, supra note 1, at 1982 (Proposed Rule 23(e)(5)(C), (E), and (F)).

(129) See infra notes 150-51 and accompanying text.

(130) See Cooper, supra note 1, at 1981 (Proposed Rule 23(e)(3)). This right does not exist if the class member was notified of a "less favorable" settlement and did not then opt out. See id.

(131) Professor Hazard's paper for this symposium suggests such a view, at least as to future claimants. See Geoffrey C. Hazard, Jr., The Futures Problem, 148 U. PA. L. REV. 1901 (2000).

(132) See Thomas v. Albright, 139 F 3d 227 (D.C. Cir. 1998) (holding that the district court abused its discretion in allowing opt-outs in a (b)(2) class action); see also Eubanks v. Billington, 110 F.3d 87, 94-95 (D.C. Cir. 1997) ("[A]s a general matter, courts should not permit opt-outs when doing so would undermine the policies behind (b) (1) or (b)(2) certification...."); In re Cincinnati Radiation Litig., 187 F.R.D. 549 (S.D. Ohio 1999) (providing that the class action is binding with regard to equitable relief, but allowing class members to opt out of monetary relief provisions).

(133) For a discussion of these problems, see Derrick A. Bell, Jr., Serving Two Masters: Integration Ideals and Client Interests in School Desegregation Litigation, 86 YALE L.J. 470 (1976); cf. William B. Rubenstein, Divided We Litigate: Addressing Disputes Among Group Members and Lawyers in Civil Rights Campaigns, 106 YALE L.J. 1623 (1997) (discussing the tension between individualism and group interests in such litigation).

This sort of tension is not limited to school desegregation. Consider, for example, Pratt v. Chicago Housing Authority, 155 F.R.D. 177 (N.D. Ill. 1994), in which residents of facilities operated by the Chicago Housing Authority ("CHA"), represented by civil liberties lawyers, sued the CHA challenging the constitutionality of its tough new policy of warrantless searches of apartments in response to incidents of violence or gunfire. The problem was that the raids never occurred sooner than two days after the event in question, and the court initially certified a class of residents of the CHA projects to challenge the policy. Thereafter the Local Advisory Council presidents for 18 of the 19 CHA developments moved to intervene on the defense side and presented a petition signed by 5000 other CHA tenants applauding the CHA's efforts to respond to violence in the projects. The court decertified the class. Would it have been sufficient to allow opting out? How much difference would that make if the court entered an injunction against the new policy?

(134) See, e.g., NAACP v. City of Thomasville Sch. Dist., 187 F.R.D. 690 (M.D. Ga. 1999) (certifying (b) (2) class in suit alleging that school district engages in racial segregation).

(135) See FED. R. CIV. P. 19(a) (directing that such people "shall be joined" if their joinder will not deprive the court of subject matter jurisdiction).

(136) See supra notes 38-39 and accompanying text.

(137) See County of Suffolk v. Long Island Lighting Co., 907 F. 2d 1295 (2d Cir. 1990) (upholding the district court's decision to allow a class member that had already litigated an individual claim to judgment at great expense to opt out of a Rule 23(b) (1) class action).

(138) 490 U.S. 755 (1989).

(139) See Abram Chayes, The Role of the Judge in Public Law Litigation, 89 HARV. L. REV. 1281, 1298-1300 (1976) (suggesting that such negotiated decrees are probably the norm in public law litigation).

(140) Martin, 490 U.S. at 775 (Stevens, J., dissenting) (describing objections of a group of black firefighters that the settlement's provisions were inadequate).

(141) Indeed, the provision in Rule 23(c)(1) that the certification decision be made "[a]s soon as practicable" would seem to mean that there would often be a delay between notice and the consummation of a settlement.

(142) See supra note 62 and accompanying text.

(143) See Preliminary Draft, supra note 94, at 559 (proposing new Rule 23(b)(4)).

(144) See supra notes 104-12 and accompanying text.

(145) See, e.g., Craft v. Vanderbilt Univ., 174 F.R.D. 396, 406-08 (M.D. Tenn. 1996) (certifying medical monitoring class under (b) (2) for women subjected to experiment using radioactive isotopes).

(146) See supra Part III.B.2.

(147) Consider the views of Judge Heaney, dissenting from a ruling that the Anti Injunction Act, 28 U.S.C. [sections] 2283, precluded an injunction against state-court litigation by class members in connection with an alleged Rule 23(b)(1)(B) federal-court class action:

      A mandatory class action, of course, has a restrictive effect on related
   proceedings in any other court--state or federal. This is because, by
   definition, members of such a class cannot pursue independent litigation of
   class claims.... [T]he implication of the majority's view is that mandatory
   classes are not truly mandatory--any member who has previously commenced
   independent litigation is somehow not subject to the ordinary rules of such
   class actions.

In re Federal Skywalk Cases, 680 F. 2d 1175, 1191 (8th Cir. 1982) (Heaney, J., dissenting).

(148) See supra note 28 and accompanying text.

(149) See supra text accompanying notes 118-28.

(150) See Philadelphia Elec. Co. v. Anaconda Am. Brass Co., 42 F.R.D. 324 (E.D. Pa. 1967) (holding that if precertification settlement purports to resolve class claims it is covered by FED. R. CIV. P. 23(e)).

(151) See In re Prudential Ins. Co. of Am. Sales Practice Litig., 148 F. 3d 283, 316-24 (3d Cir. 1998) (discussing a variety of factors pertinent to assessing adequacy of proposed settlement), cited in Cooper, supra note 1, at 1986.

(152) Cooper, supra note 1, at 1981-82 (Proposed Rule 23(e)(5)).

(153) See, e.g., 7B WRIGHT, MILLER & KANE, supra note 10, [sections] 1797.1 (discussing the factors considered for approval of Rule 23 settlements).

(154) See Paul D. Carrington, Making Rules To Dispose of Manifestly Unfounded Assertions: An Exorcism of the Bogy of Non-Trans-Substantive Rules of Civil Procedure, 137 U. PA. L. REV. 2067, 2069-87 (1989) (examining the risks of shifting to substance-specific rules).

(155) See SUPP. RULES FOR CERTAIN ADMIRALTY 8c MAR. CLAIMS, Supp. Rule A, ("The general Rules of Civil Procedure ... are also applicable to the foregoing proceedings except to the extent that they are inconsistent with these Supplemental Rules.").

(156) See supra notes 130-46 and accompanying text.

(157) Pub. L. No. 101-650, 104 Stat. 5089 (1990) (requiring each district to develop a plan to reduce litigation cost and delay).

(158) Pub. L. No. 104-67, 27, 109 Stat. 737, 738 (1995) (altering handling of securities fraud suits, including class action practices).

(159) The Future of Class Actions in Mass Tort Cases: A Roundtable Discussion, 66 FORDHAM L. REV. 1657, 1667 (1998) (comments of Judge Becker).

(160) See Miller, supra note 23, at 670-75 (enumerating types of cases that raised class action problems but not mentioning torts).

(161) See supra note 115 and accompanying text.

(162) See supra notes 38-40 and accompanying text.

(163) See ADVISORY COMM. ON CIVIL RULES, 1-4 WORKING PAPERS OF THE ADVISORY COMMITTEE ON CIVIL RULES ON PROPOSED AMENDMENTS TO CIVIL RULE 23 (1997).

(164) See REPORT ON MASS TORT LITIGATION, supra note 21.

RICHARD L. MARCUS, Horace O. Coil ('57) Chair in Litigation, University of California, Hastings College of Law. Since 1996, I have served as Special Reporter to the Advisory Committee on Civil Rules in connection with its review of the discovery rules, but not its study of Rule 23. The views expressed in this Commentary are entirely my own and do not reflect the views of the Advisory Committee, or any of its members. I am indebted to Mary Kay Kane for her comments on a section of this paper. This Commentary is based on a presentation at the Mass Torts Symposium held at the University of Pennsylvania on November 11-12, 1999.