Walter Dellinger

September 25, 2002

I. THE CURRENT ASBESTOS-LITIGATION CRISIS
A. An Overview of the Crisis.
When asbestos product liability lawsuits emerged almost thirty years ago,2 no one could have predicted that courts today would be facing what the United States Supreme Court has aptly termed an "asbestos-litigation crisis." Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 597 (1997). The Occupational Safety and Health Administration ("OSHA") promulgated its first asbestos regulation in 1971, and followed up with increasingly stringent regulations in the years to follow.3 By the early 1970s, "use of new asbestos essentially ceased in the United States." In re Joint E. & S. Dists. Asbestos Litig., 129 B.R. 710, 737 (Bankr. E. & S.D.N.Y. 1991) (Weinstein, J.), vacated, 982 F.2d 721 (2d Cir. 1992), opinion modified on reh'g, 993 F.2d 7 (2d Cir. 1993) (reviewing history of asbestos use). It seemed to many that, after the 1970s and 1980s, asbestos litigation would be a serious but diminishing problem. See Victor E. Schwartz & Leah Lorber, A Letter to the Nation's Trial Judges: How the Focus on Efficiency Is Hurting You and Innocent Victims in Asbestos Liability Cases, 24 Am. J. Trial Advoc. 247, 248 (2000) ("Schwartz & Lorber").
The opposite is true. Instead of declining, asbestos filings are multiplying exponentially. In 1991, there were approximately 100,000 asbestos cases pending in courts around the country. By 1999, that number had doubled to roughly 200,000. See The Fairness in Asbestos Compensation Act of 1999: Legislative Hearing on J.R. 1283, Before the House Comm. on the Judiciary, 106th Cong. at 4 (July 1, 1999) (statement of Prof. Christopher Edley, Jr., Harvard Law School) ("Edley House Testimony"). New cases are now filed at a rate greater than ever before. See id. In 2001 alone, plaintiffs filed at least 90,000 new claims, see Alex Berenson, A Surge in Asbestos Suits, Many by Healthy Plaintiffs, N.Y. Times, Apr. 10, 2002, at A-1, and up to 700,000 more cases are expected by the year 2050, Mass Tort Litigation Report Discusses Resolving Asbestos Cases Over Next 20 Years, 14 Mealey's Litig. Rep.: Asbestos 22 (June 18, 1999). All told, the number of future claimants could reach as high as 3.5 million. See Report of the Judicial Conference Ad Hoc Committee on Asbestos Litigation 5 (Mar. 1991) ("Judicial Conference Report").
In short, "the asbestos litigation crisis not only remains with us, but has in important respects grown worse." The Fairness in Asbestos Compensation Act of 1999: Legislative Hearing on S. 758, Before the Senate Judiciary Comm., 106th Cong. at [1] (Oct. 5, 1999) (statement of Prof. Christopher Edley, Jr., Harvard Law School) ("Edley Senate Testimony"). In 1991, the Judicial Conference described a looming "disaster of major proportions." Judicial Conference Report at 2. Since that time, the rate of new filings and the mounting number of pending cases have only exacerbated the crisis. Long delays in resolving claims remain routine. Christopher Edley, Jr. & Paul C. Weiler, Asbestos: A Multi-Billion-Dollar Crisis, 30 Harv. J. on Legis. 383, 394 (1993) ("Edley & Weiler"). Bankruptcies increasingly threaten the ability of asbestos defendants to compensate seriously ill plaintiffs, now and in the future. To date, more than 56 companies have been driven into bankruptcy. See Mark A. Behrens, Editorial, When the Walking Well Sue, Nat'l L.J., Apr. 29, 2002, at A12; Mark D. Plevin & Paul W. Kalish, Where Are They Now? A History of the Companies That Have Sought Bankruptcy Protection Due to Asbestos Claims, Vol. 1, No. 1 Mealey's Asbestos Bankr. Rep., Aug. 2001. In the last two years, this process has accelerated dramatically, forcing at least 18 companies with more than 100,000 employees into bankruptcy. See Alex Berenson, supra, at A1.4 More companies will follow, probably by the end of this year. See RAND Rep. at 50 (predicting that "[a]ll of the major asbestos defendants are likely to be in bankruptcy within 24 months"). And each new bankruptcy puts "mounting and cumulative" financial pressure on the remaining defendants, including defendants who are increasingly removed from any wrongdoing, and whose resources are limited, Edley & Weiler, supra, at 392. The total cost to the economy is already staggering, and may reach as high as $200 billion in the future.5

B. The Courts' Contribution to the Crisis.
The origins of the wave of asbestos litigation that began in the 1970s are well known. In the 1940s and 1950s, millions of American workers were exposed to asbestos, usually with few or no precautions. Resulting illnesses began to appear by the 1960s, and, because some asbestos-related diseases have latency periods of up to 40 years, injuries continued to emerge in later decades. Recent estimates suggest that hundreds of thousands of Americans were injured by exposure to asbestos and that thousands have died or will die as a result. See Edley & Weiler at 388-89; Judicial Conference Report at 2. Absent congressional action - action for which the United States Supreme Court has consistently pled6 - it was inevitable that asbestos litigation would present a problem for the courts.
What is harder to understand is why a problem that should have begun to resolve itself by the 1990s has instead worsened dramatically. It is here that the courts themselves share some of the blame. With the best of intentions, many courts have adopted both procedural and substantive rules intended to facilitate resolution of asbestos claims. Those efforts have been massively counterproductive. Lowering the legal barriers to recovery may seem attractive in individual cases, but in the aggregate, it only fuels the fire, inviting more and more claims with little regard for merit. See Schwartz & Lorber at 248-251; see also House Hearing on H.R. 1283 (statement of Prof. William N. Eskridge, Yale Law School) ("Eskridge Testimony") (describing judicial contribution to asbestos-litigation crisis).

1. Procedural Shortcuts.
Faced with hundreds or even thousands of asbestos claims on their dockets, courts have struggled to find ways of speeding final decision or settlement. One "near-heroic effort[] . . . to make the best of a bad situation," Ortiz, 527 U.S. at 865 (Rehnquist, C.J., concurring), involved mass settlements of hundreds of thousands or even millions of claims aggregated under Rule 23 of the Federal Rules of Civil Procedure. But that route was invalidated by the Supreme Court in Amchem and Ortiz: even the most pressing efficiency interests, the Supreme Court held, cannot justify distortions of the civil justice system that are fundamentally unfair to the parties involved. Amchem, 521 U.S. at 620-29; Ortiz, 527 U.S. at 841-61.
Other courts have turned to mass joinders, "jumbo" consolidations and "conjoinders," aggregating under any number of labels thousands of claims against dozens or hundreds of defendants in an effort to produce quick settlements with low transaction costs. See Eskridge Testimony at 13 (describing pressure on defendants to settle on terms favorable to plaintiffs). Typically, the claims are so disparate - injured plaintiffs joined with the unimpaired, plaintiffs exposed to asbestos in different settings and even in different decades - that they would not remotely qualify for aggregation under normal circumstances. See id.; Schwartz & Lorber at 256-57 ("In other cases that do not involve asbestos, judges would not consolidate or join cases when plaintiffs suffer completely different types of injuries."). In the asbestos context, however, courts see no choice but to forgo standard procedural protections in an effort to streamline resolution.
Even if this trade-off were acceptable - and the Supreme Court, in cases like Amchem and Ortiz, has suggested strongly that it is not - it has proven entirely counterproductive. As it turns out, bending procedural rules to put pressure on defendants to settle, see Eskridge Testimony at 39-40, brings no lasting efficiency gains. Rather than making cases go away, it invites new meritless ones. As Professor Eskridge explains, "[J]udicial experimentation has sacrificed both [procedural protections] and efficiency, by helping create a juggernaut whereby jumbo settlements generate more lawsuits." Id.; see also Schwartz & Lorber at 249. This effect should not be surprising:
Judges who move large numbers of highly elastic mass torts through their litigation process at low transaction costs create the opportunity for new filings. They increase demand for new cases by their high resolution rates and low transaction costs. If you build a superhighway, there will be a traffic jam.
Francis E. McGovern, The Defensive Use of Federal Class Actions in Mass Torts, 39 Ariz. L. Rev. 595, 606 (1997).7

2. Unimpaired Plaintiffs.
The courts' substantive rulings in asbestos cases also have contributed to the litigation crisis. Of special concern are substantive rules that make it easier for unimpaired or only mildly impaired plaintiffs to recover. For it is by now widely acknowledged that claims by the relatively unimpaired are at the heart of the continuing asbestos-litigation crisis. "No serious analyst believes that the increased number of filings is due to an increased prevalence of asbestos-related disease. . . . Rather, the new filings represent claims of people who have been exposed to asbestos . . . but are not impaired by an asbestos-related disease and likely never will be." Edley Senate Testimony.
Some unimpaired plaintiffs, though they have been exposed to asbestos, show no physical symptoms at all. Others show "pleural plaques" or "pleural thickening," physical changes in the lungs that do not affect lung functions and do not necessarily lead to or increase the risk of asbestos-related disease. Mild forms of asbestosis, a set of lung disorders, also may be present without significant impairment or any medical link to more severe illnesses. What all of these unimpaired or less-impaired plaintiffs have in common is that they do not suffer from the kinds of asbestos-related cancers - most often, mesothelioma - or severe asbestosis prevalent in asbestos plaintiffs of earlier decades. See Edley & Weiler at 393.
In Amchem, 521 U.S. at 629, Justice Breyer served that "up to one half of asbestos claims are now filed by people who have little or no physical impairment." That number is perhaps too conservative. For instance, Professor Edley estimated in 1992 that claims by unimpaired plaintiffs then accounted for 60 to 70 percent of new claims, with the trend toward unimpaired claimants steadily increasing, Edley House Testimony at 5, and some current estimates are as high as 90 percent, see Jennifer Biggs et al., Overview of Asbestos Issues and Trends 3 (Dec. 2001) (http://www.actuary.org/mono.htm). Whatever the precise percentage, mass filings by unimpaired or mildly impaired claimants are the "wild card" that caused earlier predictions of a decline in litigation to be so far off the mark.
The problem presented by these claims is self-evident: they divert scarce resources from the truly ill claimants who need them most. Backlogs of claims by the unimpaired or mildly impaired slow the judicial process, delaying resolution for those with fatal diseases and elderly claimants. And payments to the unimpaired or mildly impaired are rapidly exhausting limited assets that should go to "the sick and the dying, their widows and survivors." In re Collins, 233 F.3d 809, 812 (3rd Cir. 2000), cert. denied, 532 U.S. 1066 (2001) (internal quotation omitted); see also Edley & Weiler at 393. Indeed, lawyers who represent asbestos plaintiffs with cancer share this concern, recognizing that recoveries by the unimpaired may so deplete available resources that their clients will be left without compensation. See "Medical Monitoring and Asbestos Litigation" - A Discussion with Richard Scruggs and Victor Schwartz, Vol. 17, No. 3 Mealey's Litig. Rep.: Asbestos, Mar. 1, 2002, at 39 (quoting plaintiffs' attorney Richard Scruggs).
A number of factors help to explain this phenomenon. Some plaintiffs exposed to asbestos may feel compelled to file suit despite the absence of symptoms for "fear that their claims might be barred by the statute of limitations if they wait until such time, if ever, that their asbestos-related condition progresses to disability." In re Asbestos Cases, 586 N.E.2d 521, 523 (Ill. App. 1991); see also Mark Behrens and Monica Parham, Stewardship for the Sick: Preserving Assets for Asbestos Victims Through Inactive Docket Programs, 33 Tex. Tech. L. Rev. 1 (2001) (proposing inactive dockets as solution to problem); Judicial Conference Report at 25-26 (discussing similar proposals). Other plaintiffs, aware that many asbestos defendants are filing for bankruptcy, may seek compensation now because they worry that it will not be available later.
Again, however, the courts' own rulings in asbestos cases are a major contributor to the problem. Rulings loosening procedural rules have on a systemic level opened the floodgates to claims by unimpaired plaintiffs. Some courts have done this simply by recognizing as a compensable injury pleural thickening, visible only on an x-ray and entirely harmless. See Edley House Testimony at 5. Others have allowed unimpaired claimants to sue for medical monitoring, or for the fear of future injury. In Metro-North Commuter Railroad Company v. Buckley, 521 U.S. 424 (1997), the Supreme Court refused to authorize an asbestos-related medical monitoring claim under the Federal Employers' Liability Act, 45 U.S.C. §§ 51, 53 & 56, recognizing that such a claim would extend to "tens of millions of individuals," expose defendants to unlimited liability, and thus drain the pool of resources available for meritorious claims by plaintiffs with serious present harm. Id. at 442. Nevertheless, several states permit medical monitoring claims under state law. See Bower v. Westinghouse Elec. Corp., 522 S.E.2d 424 (W. Va. 1999).
Thus, substantive rulings regarding unimpaired plaintiffs have, like procedural shortcuts, become a part of the very problem they are designed to address. However well-intentioned, they inevitably encourage plaintiffs to sue even in the absence of any injury, and encourage aggressive lawyers to seek out unimpaired clients. See In re Joint E. & S. Dist. Asbestos Litigation, 129 B.R. 710, 748 (E. &. S.D.N.Y. 1991) (describing lawyers who have "arranged through the use of medical trailers . . . to have x-rays taken of thousands of workers without manifestations of disease and then filed complaints for those that had any hint of pleural plaques"); Pamela Sherrid, Looking for Some Million Dollar Lungs, U.S. News & World Rep., Dec. 17, 2001, at 36 (lawyers advertise with solicitations reading: 'Find out if YOU have MILLION DOLLAR LUNGS!'"). The upshot, of course, is that judicial resources and defendant assets are diverted from the truly sick claimants who need them most.

II. A CASE IN POINT: WEST VIRGINIA'S RECENT HANDLING OF ASBESTOS LITIGATION THROUGH MASS "CONJOINDERS."
West Virginia is perhaps the case in point for describing the asbestos-litigation crisis. The West Virginia courts believe that they have no choice but to "adopt diverse, innovative, and often non-traditional judicial management techniques to reduce the burden of asbestos litigation." State ex rel. Appalachian Power Co. v. MacQueen, 479 S.E.2d 300, 304 (W. Va. 1996). In practice, this has meant two mass asbestos trials, with a third mass trial--originally filed by over 8,000 plaintiffs against over 250 defendants--beginning this very week. But this judicial "innovation" has not solved West Virginia's asbestos problem. Instead, it has aggravated it. As one West Virginia trial judge involved in asbestos litigation has ruefully acknowledged:
I will admit that we thought that [an early mass trial] was probably going to put an end to asbestos, or at least knock a big hole in it. What I didn't consider was that that was a form of advertising. That when we could whack that batch of cases down that well, it drew more cases.
In re Asbestos Litigation, Civ. Action No. 00-Misc.-222 (Nov. 8, 2000) (transcript of hearing before Judge John A. Hutchinson).
The mass trial beginning this very week is the most aggressive example of West Virginia's "innovative" approach to asbestos litigation. In November of 2000, all "asbestos-based personal injury cases in West Virginia" were referred to the MLP by order of the West Virginia Chief Justice. State ex rel. Allman v. MacQueen, 551 S.E.2d 369, 372 (W. Va. 2001). The next month, those pending asbestos cases were transferred to a trial judge, who entered a "master plan" anticipating a series of group trials containing between 20 and 100 plaintiffs each. MacQueen, 551 S.E.2d at 372. In July 2001, the West Virginia Supreme Court rejected the trial judge's plan, holding that a more expeditious approach was required. Id. at 375. Specifically, the court held that all of the thousands of pending asbestos trials were to commence in just one year, regardless of the circumstances. The court designated an additional judge to supervise the administration of the asbestos litigation, and ordered a report to the Chief Justice on the status of the case with a view toward a July 1, 2002 commencement of trial. Id.
On September 6, 2001, the new trial judge ruled that he would conduct a single "mass trial" of all asbestos claims. Over the due-process objections of many defendants, the judge entered the official "Trial Scheduling Order" ("TSO") on February 26, 2002. Under that TSO, the mass trial was set to resolve approximately 8,000 cases against three "groups" of defendants: manufacturers, premises owners, and employers. The first three phases of the mass trial - keyed to the three groups of defendants - will determine the fault of each defendant. The Group I trial, which includes over a hundred defendants, will resolve "the common issues of fault of all manufacturers and distributors of asbestos-containing products, as well as any defendants whose purported fault is based on an allegation of conspiracy, tortious joint venture, or other tortious combination with a manufacturer or distributor."
At the end of each of these three "fault" trials - but before any determination of causation or injury - the jury will consider punitive damages. For any defendant whose conduct warrants an award of punitive damages, the jury will select a "punitive damages multiplier" - that is, the number by which any subsequent award of compensatory damages should be multiplied to arrive at a punitive damages award. Causation and compensatory damages will be determined only after the fault phase is completed, either through mini-trials or through mini-trials in combination with a statistical matrix by which early verdicts are extrapolated to the remaining claims.
On April 25, 2002, the West Virginia Supreme Court denied several defendants' request for relief and approved this mass trial plan. The decision of the West Virginia Supreme Court runs roughshod over the Due Process Clause of the Fourteenth Amendment. The hundreds of defendants include premises owners, manufacturers, employers, and insurance carriers. The thousands of plaintiffs worked at hundreds of locations across the country, in different types of jobs, at different time periods spanning six decades, with different degrees of exposure and different individual health backgrounds. They were exposed to hundreds of different asbestos-containing products with different applications, instructions, and warning labels, and are asserting different theories to recover for different injuries. In short, apart from the fact that their claims involve alleged exposure to asbestos, these approximately 8,000 plaintiffs have nothing in common.8 But under West Virginia's special mass tort rule, the liability of hundreds of defendants to these thousands of plaintiffs will be resolved at once. There will be no inquiry into whether that mass adjudication affords defendants a fair opportunity to defend themselves. A defendant will not, for example, have any opportunity to show that the claims against it have no logical relationship to the claims made against other defendants, or to demonstrate that the tremendous differences among defendants will be lost during a mass trial.
To be sure, States possess considerable flexibility in creating rules of civil procedure. That flexibility, however, is ultimately constrained by the Due Process Clause. Hansberry v. Lee,. 311 U.S. 32, 40-42 (1940); Logan v. Zimmerman Brush Co., 455 U.S. 422, 432 (1982) ("because minimum [procedural] requirements [are] a matter of federal law, they are not diminished by the fact that the State may have specified its own procedures that it may deem adequate") (citation omitted). Of special relevance here, the Court has on several occasions invalidated state rules of civil procedure when they afford individuals affected by mass litigation inadequate protections. See Hansberry, 311 U.S. at 40-42; Phillips Petroleum Co. v. Shutts, 472 U.S. 797 (1985); Eisen v. Carlisle & Jacquelin, 417 U.S. 156 (1974).
Due process limits on state authority to aggregate cases for trial protect important interests of defendants. Specifically, many courts have recognized that cases joined, or conjoined, for trial must have enough in common so that a defendant forced to deal with several plaintiffs at once is in fact defending only against a single and narrowly presented legal situation. See, e.g., Garber v. Randell, 477 F.2d 711, 716 (2d Cir. 1973). In addition, even if cases have something in "common," and whatever broad level of abstraction, cases should not be tried together if they unfairly prejudice the parties to that joint trial. See, e.g., Glussi v. Fortune Brands, Inc., 714 N.Y.S.2d 516, 518 (N.Y. App. 2000).
These twin precautions--what I call "commonality" and "prejudice"--are doubly important in cases involving not only multiple plaintiffs but also multiple defendants. Without a careful inquiry into commonality and prejudice, the proceeding may easily become so large and confusing that it is impossible for each defendant to present a meaningful defense: evidence inadmissible as to one defendant may be admitted as to others, see Cain v. Armstrong World Industries, 785 F. Supp. 1448, 1457 (S.D. Ala. 1992), and the complexity of the proceedings may make it impossible for the jury to sort through the evidence and various defenses to tailor a verdict to each party's culpability, see Logan, 455 U.S. at 433 (due process guarantees "the aggrieved party the opportunity to present his case and have its merits fairly judged").
The West Virginia mass trial amply illustrates the wisdom of those constitutional limits. The single common element in the 8,000 claims massed for trial is that the word "asbestos" appears in each complaint. The thousands of plaintiffs have been "exposed to different asbestos-containing products, for different amounts of time, in different ways, and over different periods. Some [plaintiffs] suffer no physical injury or have only asymptomatic pleural changes, while others suffer from lung cancer, disabling asbestosis or from mesothelioma." Amchem, 521 U.S. at 624 (quotation omitted).
The mass trial will also make it impossible for a defendant to assert its unique defenses in a meaningful way. The Group I mass trial originally included over a hundred defendants, each of which manufactured or distributed a different asbestos-containing product or products. Each defendant's liability will be assessed in conjunction with that of over a hundred other manufacturers, distributors, and alleged conspirators. Evidence concerning the alleged knowledge and conduct of all these other defendants will be admitted into the mass proceeding, where it will almost inevitably tar any other defendant, as well. Gwathmey, 215 F.2d at 154 (cumulative effect of evidence against some defendants prejudices jury against all defendants in consolidated case). And the sheer quantity and complexity of the information that will be presented is virtually certain to overwhelm the jury, making it impossible to distinguish one defendant or defense from another and to render a fair verdict on any one defendant's unique defenses. Malcolm, 995 F.2d at 352 (finding that "sheer breadth of the evidence" when 48 asbestos cases are consolidated makes it impossible to prevent jury confusion).
Departures from prevailing practice like the West Virginia handling of asbestos claims should be particularly suspect in this context. The Supreme Court has recognized the critical importance of the traditional protections that attend class-action aggregations. Those protections will be rendered meaningless if state procedural innovations, like the West Virginia "conjoinder," operate unchecked by traditional aggregation rules and standards. Whatever the label affixed by a State, a mass aggregation that implicates the rights of plaintiffs to adequate representation and the rights of defendants to a fair opportunity to defend should be accompanied by the traditional protections - including the standard judicial inquiry into commonality and risk of prejudice and jury confusion. See, e.g., Joan Steinnman, Reverse Removal, 78 Iowa L. Rev. 1029, 1042 (1993) (noting concern that mass consolidations lack the "procedural safeguards that due process and codified rules demand in class actions of similar magnitude").
The due process concerns at issue in the West Virginia case are especially troubling because post-trial review of mass aggregations is effectively unavailable, as the "paucity of appeals challenging trial settings of multiple [consolidated] claims" attests. In re Ethyl Corp., 975 S.W.2d at 610-11. Given the enormous potential liability that mass aggregations pose for defendants, combined with scrutiny from financial markets, aggregated proceedings exert powerful pressure on defendants to settle even meritless cases. Aggregation may raise the stakes of litigation to the point where a defendant simply cannot risk trial, regardless of the merits, thus opening the door to what are effectively "blackmail" settlements. "The risk of facing an all-or-nothing verdict presents too high a risk, even when the probability of an adverse judgment is low." Castano v. Am. Tobacco Co., 84 F.3d 734, 746 (5th Cir. 1996); see also In re Rhone-Poulenc Rorer, Inc., 51 F.3d 1293, 1298-99 (7th Cir. 1995) (same); Fed. R. Civ. P. 23(f), advisory committee notes (providing for interlocutory review of class certification decisions because certification "may force a defendant to settle rather than incur the costs of defending a class action and run the risk of potentially ruinous liability"). In In re Chevron, which involved an aggregation of 3,000 personal-injury claims against a single defendant, the Fifth Circuit concluded that the decision to aggregate would "probably not [be] effectively reviewable after trial. The pressure on the parties to settle in fear of the result of a perhaps all-or-nothing . . . trial is enormous." 109 F.3d at 1022. Outside observers increasingly agree. See, e.g., The Fairness in Asbestos Compensation Act of 1999: Legislative Hearing on H.R. 1283 Before the House Committee on the Judiciary, 106th Cong. (July 1, 1999) (Prof. William N. Eskridge, Yale Law School) ("Especially in state courts, defendants in the typical [asbestos] jumbo consolidation now face an Armageddon scenario if they do not settle on terms favorable to plaintiffs.").
Indeed, all of the parties to the West Virginia litigation fully understand the coercive pressures at work. Lawyers for unimpaired plaintiffs argued against Judge MacQueen's original small-group trial plan on the express basis that it would not impose "enough leverage on [defendants] to cause them to settle a thousand cases." In re Asbestos Litig., Civ. Action No. 00-Misc.-222 (Cir. Ct. Kanawha Cty., W. Va. March 16, 2001), Transcript of Hearing at 72-73 (statement of James F. Humphrey). The West Virginia courts themselves have relied on the certainty that the planned mass trial will provoke mass settlements - in order to explain why the mass trial will not be as hopelessly sprawling and confusing as it now appears. It is one thing to take account of the possibility of settlements driven by external factors in planning an aggregated trial. But it is something else entirely when the anticipated settlements are driven by the mass trial proceeding itself, so that the coercive nature of a mass trial becomes its own justification.9