March 20 and 21, 1997

The Civil Rules Advisory Committee met on March 20 and 21, 1997, at the University of Alabama School of Law. Committee members also attended the CJRA Implementation Conference held at the School of Law by the American Bar Association from March 20 through March 22. The meeting was attended by Judge Paul V. Niemeyer, Chair, and Judge John L. Carroll, Judge David S. Doty, Francis H. Fox, Esq., Assistant Attorney General Frank W. Hunger, Mark O. Kasanin, Esq., Judge David F. Levi, Carol J. Hansen Posegate, Esq., Judge Lee H. Rosenthal, Professor Thomas D. Rowe, Jr., Judge Anthony J. Scirica, Judge C. Roger Vinson, and Phillip A. Wittmann, Esq. Edward H. Cooper was present as reporter. Richard L. Marcus attended as Special Reporter for the Discovery Subcommittee. Former Committee chair Judge Patrick E. Higginbotham also was present. Sol Schreiber, Esq., attended as liaison member from the Committee on Rules of Practice and Procedure. Judge Eduardo C. Robreno represented the Bankruptcy Rules Committee, and Judge Jerome B. Simandle represented the Committee on Court Administration and Case Management. Peter G. McCabe and John K. Rabiej represented the Administrative Office of the United States Courts, and Karen Kremer of that office also attended. Donna Stienstra represented the Federal Judicial Center. Observers included Deborah Hensler and James Kakalik of the RAND Institute for Civil Justice; Judith Resnik; and Jonathan W. Cuneo and Alfred W. Cortese, Jr.

Judge Niemeyer opened the meeting with a report on the Federal Judicial Center program to expand the coverage of jury voir dire practices in its judicial education programs. He also noted several bills pending in Congress on subjects that may be of interest to the Committee. One proposes a study of judicial activism. Another is the reintroduction of the Sunshine in Litigation Act that has engaged the Committee's attention in connection with the continuing study of discovery protective orders and Civil Rule 26(c). A third would add a new 28 U.S.C. § 1292(b)(2) to provide for interlocutory appeal from class action certification orders, in terms that parallel the interlocutory appeal proposal published for comment in August, 1996, as new Civil Rule 23(f). The fourth carries forward the "Contract with America" proposals with respect to offers of judgment and Civil Rule 68.

Judge Niemeyer also reported on the January meeting of the Committee on Rules of Practice and Procedure. He discussed with the Standing Committee the discovery project being launched by this Committee, the status of the public hearings on the proposed class-action rule amendments, and this Committee's work with the Committee on Court Administration and Case Management in framing a Civil Justice Reform Act report for the Judicial Conference. He noted that the Standing Committee had approved the recommendation to revise the Civil Rules and forms to reflect the statutory abolition of the option to appeal from the judgment of a magistrate judge to a district court. The Judicial Conference has approved this recommendation and submitted it to the Supreme Court for transmittal to Congress in time to take effect on December 1, 1997.

Discussion of these reports noted that Civil Rule 68 has long engaged the Committee's attention. The issues have proved difficult, and many of the suggestions for revision test the limits of the Rules Enabling Act process. It may prove wise to defer further consideration pending developments in Congress. Rule 68 may yet provide the occasion for exploring means of cooperating with Congress in matters that involve the Civil Rules but that may best be addressed through the exercise of Congressional power to make substantive law.


The Civil Justice Reform Act requires that the Judicial Conference report to Congress on experience under the Act. The Committee on Court Administration and Case Management has primary responsibility for drafting a report to be considered by the Judicial Conference. This process was discussed briefly in open session, and later -- again briefly -- in executive session. It was pointed out in the open session that the Civil Rules Advisory Committee has a deep interest in the results of the local district experiments under the CJRA. Many of the principles and techniques fostered by the CJRA have been embodied in the Civil Rules. Experience under the act will call for study of other possible changes in the rules. This Committee worked with the Committee on Court Administration and Case Management in furtherance of this interest, and the cooperative endeavor proved highly successful.

Civil Rule 23

Civil Rule 23 will be the focal point of the May 1 and 2 meeting of the Committee. It was noted that the public comments and testimony on the proposals published in August, 1996, provided broad, deep, and varied reactions. The comments went not only to the proposals that were made but also to matters that had been considered by the Committee but deferred and to matters that had not been considered by the Committee. The preliminary discussion at this meeting is designed to help form the agenda for the May meeting.

A first quick summary of the comments and testimony observed that there was much controversy surrounding the Rule 23(b)(3)(F) proposal that would allow consideration of the balance between probable individual relief and the costs and burdens of class litigation. Much controversy also surrounded the Rule 23(b)(4) settlement class proposal. Although there were spirited comments addressed to many of the other proposals, most advanced issues that the Committee had already explored in depth.

The comments and testimony also addressed more fundamental challenges to the nature of Rule 23. Many witnesses stated that the number of class actions has expanded dramatically in the last two or three years, often in state courts. The insurance "rounding up" case from Texas became a symbol of a deep dispute about the purpose of class litigation. To many, the case represents the best of class actions, providing very small individual awards but forcing disgorgement of a large total sum wrongfully taken from very many people. To others, the case represents the worst of class action excesses. The question thus framed is whether Rule 23 is -- or should be -- a private attorney-general device that enables self-appointed representatives and counsel to enforce public claims. This use of Rule 23 has many steadfast supporters. It is challenged, however, by others who believe that class actions should serve only the procedural purpose of achieving efficiency through aggregation.

With this introduction, it was suggested that there are more than a thousand class action settlements every year. Perhaps 50 of them might fairly be characterized as "bad" dispositions. The balance between good and bad dispositions demonstrates the success achieved by Rule 23.

The dilemma posed by the current debate includes Enabling Act concerns. Any change in Rule 23 will, in some sense, have substantive consequences. Rule 23(b)(3) has had enormous substantive consequences. Substantive effects will follow from changes that expand it, narrow it, or expand it in some directions while narrowing it in others. The central recurring question is whether the class action is an appropriate regulatory device without regard to the benefits reaped by individual class members.

Many of the comments suggested that the proposals simply do not go far enough to restrain the unfortunate excesses of contemporary class litigation. Should the Committee undertake a more fundamental review of Rule 23?

A more specific suggestion, taken up by many of the comments, is that the opt-out class should be replaced -- in some settings or in all settings -- by an opt-in class. One approach would be to adopt the opt-in class for cases that seem to offer only de minimis individual relief. The question whether there is sufficient private interest to justify private adversary litigation could be tested by limiting the class to those who opt in. If the aggregate benefits actually sought by willing class members who choose to opt in justify the costs and burdens of class litigation, well and good. It even would be possible to leave the decision whether to pursue the litigation to the class representatives and counsel: if they are willing to pursue the action on behalf of those who have opted in, the action can proceed without any requirement that a judge attempt to balance benefits against costs.

The opt-in class proposal led to renewed discussion of the (b)(3)(F) small-claims proposal. It was stated that the purpose of the proposal was to separate out the "coupon" class, the class that seeks only the substantive goals of deterrence and disgorgement. The published proposal presents the difficult problem of striking a balance between costs and benefits. An opt-in class alternative would alleviate this problem. Opt-in classes need not lead to a proliferation of opt-in class actions growing out of the same underlying events. One answer would be to apply claim preclusion against any potential member of an opt-in class who had actual notice of the class but chose not to opt in. Even without preclusion, however, there often would not be a series of class actions. The risk of returning to the pre-1966 "one-way intervention" practice through nonmutual issue preclusion could be met by providing that potential members of the opt-in class could not use any judgment to support issue preclusion.

Turning to settlement classes, it was agreed by consensus that action on the (b)(4) proposal should be deferred until the Supreme Court has decided Amchem Products, Inc. v. Windsor, No. 96-270, argued on February 19, 1997.

Another theme sounded during the public comment period was that many of the proposals seem driven by the growing use of class actions to dispose of mass tort litigation, particularly dispersed mass torts. It was noted that recent appellate decisions seem to have exerted a substantial restraining effect on certification of mass-tort classes (and suggested that this result shows the importance of the interlocutory appeal proposal). The suggestion was made that the Committee should reexamine the possibility, abandoned some years ago, of a new and separate rule for mass tort classes. This approach would avoid the danger that changes made in Rule 23 to address mass tort problems may cause unnecessary difficulties in many other fields characterized by mature and useful class-action practice. The difficulties, however, are manifold. Perhaps the most direct difficulty is in defining the boundaries of a "mass tort" rule. It might be limited to personal injury cases, perhaps covering such matters as thresholds for numbers of victims, dispersion of injuries in time and place, and severity of injuries; a different emphasis on the value of "issues" classes; special rules for choice of law; and particular answers for the subsequent stages of assessing such necessarily individual issues as injury, specific causation, contributory fault, and damages. Provision might be made for the problem of "future" claimants who have been exposed to a harm-causing agent but have not yet suffered injury. Efforts might even be made to integrate such a rule with supporting legislation. The potential artificiality of excluding property damage claims might be met by allowing property damage claims to be resolved under the rule so long as the personal injury threshold were met. If these boundary problems can be surmounted, a more fundamental challenge will remain. The content of the rule must be defined. The definition must account for the predictable fact that most of the claims in most of these classes will arise under state law, not federal. Great care must be taken to avoid untoward substantive impact on these state-law claims.

Another proposal advanced by several witnesses was that the "common evidence" element of predominance should be made an explicit factor in (b)(3) classes. The proposal would require that common proof resolve all, or substantially all, elements of class members' claims. The result will be to avoid the need for thousands of individual "minitrials" after a reduced set of common issues is resolved on a class basis. It was rejoined that in fact there are not thousands of minitrials. Defendants do not insist on this approach. And a commonality of proof requirement, taken very far, would make class litigation almost impossible.

Some of the comments urged that any Rule 23 revision should address problems of notice. Notices in (b)(3) classes now are commonly unintelligible -- even sophisticated lawyers must spend hours attempting to decipher them. A "plain English" requirement should be added. Something also might be done to authorize lower cost notice in small-claims classes, and perhaps to require notice in (b)(1) and (b)(2) classes.

Attention turned from these new proposals to specific issues addressed to the published proposals. It was urged that the "maturity" element added to factor (b)(3)(C) was addressed to mass tort cases, and should be removed if mass torts are not to remain a focus of any Rule 23 revisions. Demanding maturity in other settings could upset well-established practices.

If the balancing approach to small-claims classes in proposed factor (b)(3)(F) goes forward, drafting changes may be required. Many have suggested that it should be made clear whether aggregate class benefits may be considered, and whether the projection of probable relief requires or permits a preliminary evaluation of the merits.

The hearing requirement added to subdivision (e) in conjunction with the (b)(4) settlement class proposal was the target of several comments. It has been urged that courts now routinely hold hearings, but do not hold hearings when an action brought as a class action is dismissed before class certification and in circumstances that do not involve any risk of collusion or injury to putative class members. There is little need for this revision unless it is tied to settlement class provisions or other changes in subdivision (e). And there may be risks. It has been urged that many pro se actions are brought as class actions on claims that warrant immediate dismissal; the risk that a hearing must be had in every such case is inappropriate. It also has been suggested that many cases involve first a hearing and approval of a settlement, then administration of the settlement over a protracted period, and finally dismissal; the proposal might be read to require a second hearing before the final dismissal.

The first task for the May meeting will be to decide what steps to take next with the Rule 23 proposals. For some time it has been supposed that it would be best to address Rule 23 once, in a single package. Circumstances, however, may have changed in ways that support separation of the initial package. The challenges raised by the comments and testimony make it appropriate to consider the prospect of proposing more fundamental changes in Rule 23. The Supreme Court's consideration of an important settlement-class case counsels delay at least on settlement-class proposals. In this setting, it may be appropriate to consider separating the package. Two or even three tracks might be followed. Some proposals could be carried through the regular steps that follow publication: they can be reconsidered in light of the comments and testimony, revised if appropriate, and -- if the revisions to not require a second publication -- sent ahead for submission to the Judicial Conference, the Supreme Court, and Congress. Others might be held for further study, recognizing that additional proposals are likely to require some time for deliberation, likely further publication, and further consideration. It is always possible that some proposals might be found in a posture that warrants immediate publication; if that should happen, it would be necessary to decide whether to delay immediate action even on proposals that otherwise would be ready to go ahead now.

A set of materials dealing with these matters will be prepared for the May agenda.

At the invitation of the Committee, Deborah Hensler of the RAND Institute for Civil Justice described the design and ongoing progress of a RAND study of class actions. The study is designed to follow a different methodology than the study done for the Committee by the Federal Judicial Center, and to supplement it by looking at different sources and kinds of information. One element is to attempt to develop a sense of the number of class actions by gathering information from electronic data bases, to be supplemented by interviews with counsel. Attention also is being paid to trends. It seems clear even now that class actions remain a relatively rare phenomenon in the total universe of litigation. Corporations and others facing large numbers are talking of dozens or scores of class actions, not hundreds. The "Fortune 50" are the most frequent targets, and they face far larger numbers of individual actions than class actions. Everyone reports significant growth, often speaking of a doubling or tripling in the last three years. Most of the growth, particularly in the last year and a half, has been in state court class actions. The subjects of class litigation also are more diverse than in the past. Mass tort claims are not frequently certified for class treatment, and many of the initial certifications have failed. There also seem to be mass tort property damage cases that are much like personal injury cases; drawing a boundary between "mass tort" classes and other classes may prove difficult. It is clear that a single event many indeed generate multiple class actions -- they are viewed as "competing" classes, whether by independent filings in the same jurisdiction or by filings in different jurisdictions. A second element is to measure the dynamics of class litigation. This stage will rely exclusively on interviews with lawyers. Lengthy interviews have been completed with some 50 different people, including leading practitioners in the plaintiffs' bar, corporations in most sectors of the economy, and so on. Much attention has focused on small damages cases. Generally they do not involve classes all of whose members have suffered only small injuries. And generally they do not involve mere "technical" violations of the law. Some do seem to involve individual claims too small to support the cost of individual notice. This is a rapidly changing area of practice. The lawyers involved in this practice are excited by the challenge, and among the finest lawyers in the country. And these kinds of cases would persist even if Rule 23 were repealed; they would remain as families of related cases, managed together.

Discovery Committee

The Discovery Committee led discussion of the project to review the discovery rules and practice. Complaints of discovery abuse continue to be pressed. It remains difficult, however, to define abuse. Equal difficulties arise with attempts to diagnose, measure, or cure abuse. The other broad issues are no easier. Is discovery too costly by some measure, either generally or in more specific ways that can be profitably addressed by rules changes? Are there proposals that will reduce cost or delay and prove acceptable to all sides?

The project now is in the phase of developing a "smorgasbord of ideas" for the September meeting. The October meeting will pick out the ideas to be developed by the Discovery Committee for the March, 1998 meeting.

It was recognized that there is a powerful view that no changes should be made in the discovery rules. The "no changes" view is particularly popular among judges and academics.

The Discovery Committee began work by holding a January conference in San Francisco in conjunction with the Rule 23 hearings. The lawyers invited to the conference were not a random group. They were selected because of their rich experiences and demonstrated interest in continuing procedural reform. Collectively, they have many years of experience from practice in all sections of the country, representing parties who span the full spectrum of federal-court litigants. The meeting generated substantial levels of agreement on some topics, and disagreement on others. The result was not a crisis report, nor any demand for radical relief. There was a clear consensus that something much like modern discovery is essential. The question is how much discovery, not whether there should be discovery.

There also seemed to be agreement that there are no general problems arising from practice under Rules 33 (interrogatories), 35 (physical and mental examinations), or 36 (requests for admissions). Nor was there much concern with civility.

Compared to the "no changes" view, the January conference showed a different view. They believe that discovery has gotten worse over the last five years. The problem is not so much abuses as the demands in the "documents" case. There seems to be a substantial practice of over-discovery, but it does not seem to involve calculated abuse for tactical advantage. The document-discovery input is enormous. The actual output of materials useful for pretrial or trial purposes is minuscule. The problems are aggravated in the "one-way" case in which one party holds almost all the documents; when both parties hold substantial volumes of documents, it is much more likely that they will cooperate to find means to manage reasonable discovery.

Apart from the cases with massive document discovery, there may be some problems with concealment. Rules that effect unintentional waiver of privilege also may deserve attention; quite apart from the intrinsic merit of the waiver rules, the fear of waiver exacts a high cost in reviewing documents for discovery responses.

There also was a consensus at the San Francisco meeting that national uniformity is desirable. The Department of Justice seems particularly interested in achieving more uniformity. One obvious need for attention is the variety of disclosure practices that have emerged from the Civil Justice Reform Act programs and the authorization given by Civil Rule 26(a)(1).

There is a strong sense that in most cases discovery is not a problem. The problems seem to be associated with "complex" cases. Defining complex cases may not be easy, however, and there are good reasons to fear an attempt to cure whatever problems may arise in complex cases by rules changes that will apply to cases that generally to not generate problems.

The protective order question remains part of this more general discovery project. The years of study and the two published proposals to amend Rule 26(c) brought this topic close to completion, but the conclusion was deferred with the thought that the discovery terrain might be changed by the broader project. One of the problems addressed by the published proposal continues to demand attention -- discovery materials produced in one action may be returned, destroyed, or sheltered by a continuing protective order that forces the parties to parallel litigation to unnecessary work in duplicating the same discovery efforts.

General discussion suggested that one approach to document discovery is to make the demanding party sort it out. If a demand seems excessive, rather than produce the responding party can simply force a motion to compel. The process that leads to a motion can lead to a reasonable result.

It was observed that problems often arise from delegation of discovery to the youngest lawyer involved with a case. Inexperienced lawyers do not know what they will need for trial, and fear criticism if they do not ask for enough.

It also was suggested that the key to successful discovery is active judicial oversight. There are some reasonable grounds for disagreement among the parties. Ready access to a judge can help the process immeasurably. Although the RAND report on CJRA experience suggests that magistrate judges can play a useful role, involvement of a district judge can be important. One task may be to attempt to sort out the frequency and success of different patterns of judicial behavior: how many judges hold themselves available for telephone discovery conferences? How many delegate problems to magistrate judges? How many take the view that the parties should resolve all discovery disputes for themselves?

Some lawyers have urged that it should be possible to work out a standard protocol identifying the types of documents that are reasonably discovered in various types of litigation. The protocols could be generated by bringing together plaintiff and defendant lawyers from each area of practice. Securities lawyers could work out a protocol for securities cases, antitrust lawyers for antitrust cases, employment discrimination lawyers for employment discrimination cases, and so on. Other lawyers have expressed doubt about this approach, and observe that often it will be difficult to determine which category matches a particular case.

The view was expressed that much good is done during the premotion conferences that most districts require. The lawyers work out most problems without judicial intervention. And magistrate judges accomplish a lot in resolving the problems that the lawyers cannot work out.

The Discovery Committee is working with the Federal Center to develop a questionnaire to be sent to all lawyers involved in a sample of 1,000 recently concluded cases. There is a limit, unfortunately, on how much can be asked. The more complex the questionnaire, the lower the level of response will be.

Disclosure practices under Rule 26(a)(1) and under local district variations also were discussed. It was urged that there is a substantial and unfortunate delay at the start of trial by the combined effect of the disclosure rule, the suspension of discovery and disclosure until the Rule 26(f) conference, and the mandatory scheduling order provisions of Rule 16(b). One response has been to issue an "initial" scheduling order at the start of the litigation, subject to revision when a regular scheduling order can be entered under Rule 16(b). It has been suggested that there is a tension between Rule 16 and Rule 26, but at the same time the theory of the 1993 amendments was that the Rule 26(f) conference is necessary to make the Rule 16(b) scheduling order effective.

The possible sources of information on the working of various local disclosure practices, including the "national" rule, were discussed. The RAND CJRA data base is available for study by the Administrative Office and Federal Judicial Center, and RAND itself will be asking some further questions. The data base includes information on lawyer hours and "judge minutes" devoted to discovery, and on discovery motions. The data will be searched to see what kinds of cases generate high levels of discovery, or frequent discovery disputes. These findings can be related to the policies used -- early mandatory disclosure, voluntary disclosure, and so on.

There are a growing number of local CJRA reports. The Eastern District of Pennsylvania has done an elaborate study of disclosure. The Eastern District of New York also has an elaborate study; it may prove instructive to compare their experience with disclosure to experience in the adjacent Southern District of New York, which has rejected disclosure. The Federal Judicial Center has full data on district-level practices, but the options commonly made available to individual judges within each district make it difficult to achieve district-wide comparisons. This phenomenon accounts for the choice to base the FJC survey on a case-level comparison, not a district-by-district approach.

It was suggested that the FJC study might usefully ask for lawyer responses to half a dozen policy questions, seeking "positions, not data."

Many lawyer associations have been asked to contribute ideas to the discovery project. Among them are ATLA, the ABA Litigation Section, the Defense Research Institute, Lawyers for Public Justice, and the American College of Trial Lawyers. The American College has been involved in the launching of the discovery project, and it is hoped that all of these groups -- as well as any others than can be brought into the process -- will provide much help.

Other Rules

Other pending Rules topics were addressed briefly.

The Copyright Rules will be on the agenda for discussion at the September meeting if time allows; otherwise they will be addressed at the October meeting.

The proposals to revise the Admiralty Rules may be ready in time for presentation at the May meeting. If not, they will be on the agenda for one of the fall meetings.

A Department of Justice proposal to amend Rules 4 and 12 to extend the time to answer in "Bivens" actions was presented in draft form. This proposal will be on the agenda for discussion at the September or October meetings.

Judge Higginbotham

Judge Higginbotham was presented a Resolution of the Judicial Conference of the United States recognizing his service as Chair of the Civil Rules Advisory Committee and as a member, from 1987 to 1993, of the Federal-State Jurisdiction Committee. In accepting the resolution, he observed that it is very important that the Advisory Committee continue the openness policy that it has been following, and that it continue to be willing to engage the hard issues. Congress will be deferential to the process as long as the Committee continues to engage the important issues openly, thoughtfully, and rationally. He also noted that the Committee has been fortunate to have the very strong and thoughtful support of the Administrative Office staff.

Respectfully submitted,

Edward H. Cooper, Reporter