Printer Friendly

The way forward after Wal-Mart.

The Supreme Court's decision denying certification of a class action in Wal-Mart Stores, Inc. v. Dukes (1) elicited a strong dissent from Justice Ginsburg, (2) and widespread criticism in liberal circles, (3) but in several important respects, the decision was unanimous. All the Justices agreed that a class action could not be certified under Federal Rule of Civil Procedure 23(b)(2), (4) the rule that governs class actions in which injunctive relief is "appropriate respecting the class as a whole." (5) Instead, the class could be certified (if at all) only under the more stringent provisions of Rule 23(b)(3), (6) typically reserved for damage class actions. (7) The holding that divided the Justices concerned the failure of the plaintiffs' case to meet the requirement of Rule 23(a)(2) that there were "questions of law or fact common to the class." (8) On this issue, the five Justices conventionally identified as conservative were in the majority and the liberals were in dissent. Yet even on this issue there were points of apparent agreement, and one of them was the need to inquire into the merits to determine whether the prerequisites for certification were satisfied. (9) On this point, the majority and the dissenters disagreed only over how strong the plaintiffs' evidence on the merits really was.

As that disagreement makes clear, a consideration of the merits deeply affects almost all certification decisions. If Wal-Mart had actually discriminated against women in pay and promotions, as Justice Ginsburg plainly suspected it had, (10) then the argument for certification would have been strengthened by the need to prevent future discrimination and to compensate victims of past discrimination. If, on the contrary, the evidence was too weak to support this conclusion, as Justice Scalia argued for the majority, (11) then certification should have been denied. What is true in the particular case also is true for entire categories of litigation: the more meritorious the underlying claim of class-wide liability, the stronger the arguments for certification. Part I of this article situates this commonly accepted observation in the perennial disputes over substance and procedure within the specific context of class actions, both for the Title VII claims at issue in Wal-Mart and for class actions generally.

The opinion in Wal-Mart expands upon this observation in a different direction, by disapproving a broad interpretation of the holding in Eisen v. Carlisle & Jacquelin (12) that an inquiry into the merits cannot be used to shift the cost of notice to the defendant immediately upon certification of a class action. Wal-Mart makes the merits a component of the certification process, first, by requiring the party seeking certification to "affirmatively demonstrate" that the requirements of Rule 23 have been met, and second, by recognizing that such proof usually involves an examination of the merits. (13) Prior cases and previous commentary have recognized these points, but these sources fail to articulate exactly what an inquiry into the merits entails, or how it relates to other procedural devices that involve an examination of the merits before trial, such as motions to dismiss for failure to state a claim and motions for summary judgment. (14) Part II of this article examines this issue and seeks to put certification decisions in their proper place within the structure of civil litigation, consisting of pleading, discovery, and summary judgment, followed by settlement or trial.

That inquiry, in turn, leads to the larger question of how to reform class action procedure. If "one size does not fit all," as John Coffee has previously pointed out, then it is the merits that determine which size fits in different class actions, both in quantitative terms--in determining the optimal number of class members--and in qualitative terms--in defining the scope and procedure for different class actions. (15) As Richard Marcus has recently emphasized, an examination of the merits is crucial to the gatekeeping function of federal courts in controlling aggregate litigation. (16) An inquiry into the merits hardly resolves all the pressing questions raised by class actions as we have them now, but it offers a place to start in framing solutions tailored to the need for aggregate litigation in different areas of law. Part III argues that an inquiry into the merits provides a suitable vehicle for considering changes in class action practice. Even such seemingly "trans-substantive" requirements as adequacy of representation can be implemented only by reference to substantive law. Whether there are conflicts of interest within the class, or whether the class attorney can effectively represent the class, cannot be decided without considering substantive law. Neither can the choice of whether to certify a class action be decided under the different subdivisions of Rule 23(b).

In Wal-Mart itself, the Court invoked substantive law in requiring certification of class actions for back pay under subdivision (b)(3). (17) The Court rejected "Trial by Formula" as an impermissible infringement on the defendant's right to oppose individual relief to any particular class member. (18) The Rules Enabling Act (19) and the Due Process Clause, (20) according to the Court, required the preservation of fights once conferred by substantive law and prevented certification of claims for back pay under subdivision (b)(2). Yet even accepting this conclusion, a close look at the substantive law under Title VII supports the consideration of approximate remedies in certification decisions--not because the courts can require "Trial by Formula," but because the parties often engage in "Settlement by Formula." The tendency towards this form of settlement constitutes a proper consideration in deciding whether common issues predominate over individual issues as required by subdivision (b)(3).

Looking to the merits provides a way to differentiate among class actions, both in fine-grained analysis of particular claims and in broad terms defined by different areas of law. The cases and commentary on class actions presume that class actions must be divided into conventional categories, such as mass torts, consumer class actions, civil rights claims, and securities class actions, but they provide little more than a pragmatic justification for this division. (21) A look at the merits reveals how closely procedure and substance are fused together in class action practice, and paradoxically enough, how disaggregated the treatment of class actions must be. This was tacitly accepted as the premise of most analyses of class actions before Wal-Mart. It is now the only way forward after it.

I. SOME REALISM ABOUT RESULTS: WHY THE MERITS MATTER

Even a casual look at the opinions in Wal-Mart reveals the different attitudes of the majority and the dissent to the merits of the case. Where Justice Scalia expressed skepticism of the plaintiffs' evidence of sex discrimination, Justice Ginsburg regarded it with evident sympathy, as an example of pervasive preconceptions about gender roles. She effectively turned his skepticism on its head, transforming the entire complexion of the case and infusing the contents of Rule 23 with her view of the merits. The plaintiffs sought to represent a class of 1.5 million current and former female employees of Wal-Mart who had allegedly suffered sex discrimination in the subjective process by which Wal-Mart's managers and supervisors made decisions on pay and promotions. If the plaintiffs had presented evidence of a stark disparity--such as the complete absence of women from higher levels of management or higher levels of compensation--they would have magnified both the merits of their claim and their arguments for certification. The clearer and larger the wrong, the greater the need for aggregate litigation in order to remedy it. The "inexorable zero" of no representation of women in better paying jobs, as an early Title VII decision called it, (22) would have supported both a finding of a pattern or practice of discrimination and certification of a correspondingly broad class action.

As the record stood in Wal-Mart, the evidence fell far short of such clarity. The existence of a disparity in the pay and promotions of women was taken for granted, both within Wal-Mart as compared to men and outside Wal-Mart as compared to the promotion of women in other stores. The district court relied on
 largely uncontested descriptive statistics which show that women
 working in Wal-Mart stores are paid less than men in every region,
 ... that the salary gap widens over time even for men and women
 hired into the same jobs at the same time, that women take longer
 to enter into management positions, and that the higher one looks
 in the organization the lower the percentage of women. (23)


The district court also found that women constituted about sixty-five percent of hourly employees, but only thirty-three percent of management employees. (24) Connecting that disparity to the subjective decisions of Wal-Mart's many supervisors and managers presented the critical problem in the plaintiffs' case. The evidence of systemic discrimination came in three forms: anecdotal evidence of discrimination in the experience of the named plaintiffs and 120 other members of the class; a regression analysis finding national and regional disparities in the rates of pay and promotion of women at Wal-Mart; and the expert testimony of a sociologist that Wal-Mart's corporate culture made the discretionary decisions of managers and supervisors susceptible to sex discrimination. (25)

The majority rejected the adequacy of this evidence to generate common questions of law and fact, for both empirical and doctrinal reasons. The anecdotal evidence suffered from the small number of individual cases relative to the size of the class, over 100 but still less than .01% of the entire class, and because of the disjunction between individual instances of discrimination and issues common to the class. No aggregation of individual cases, no matter how large, could establish commonality in the absence of common features among them. The plaintiffs' regression analyses sought to control for a variety of factors that might have legitimately affected pay, such as job performance and length of time with the company, and found a residual disparity between men and women that could, in their expert's opinion, only be explained by sex. (26) But that evidence did not, in the majority's view, connect the pattern of lower pay with the decisions of particular managers. That task was left to the evidence of implicit bias. Apart from other problems with the testimony of the plaintiffs' expert on this issue, he admitted that he could not say "[w]hether 0.5 percent or 95 percent of the employment decisions at Wal-Mart might be determined by stereotyped thinking." (27) All in all, that left the majority with the conclusion that in a company of this size, "it is quite unbelievable that all managers would exercise their discretion in a common way without some common direction." (28)

The Justices who dissented on the existence of common issues, of course, took a different view of the evidence, as did the district court, upon whose opinion the dissenters relied for their view of the facts. (29) Justice Ginsburg reached this conclusion: "The plaintiffs' evidence, including class members' tales of their own experiences, suggests that gender bias suffused Wal-Mart's company culture." (30) Whether or not she was right, her view of the evidence on the merits undoubtedly affected her view on the issue of certification, just as it did for the majority. That view carried over to substantive legal doctrine under Title VII, which she interpreted to impose a greater obligation on employers to counteract the potential bias in subjective decision-making by their managers and supervisors. (31) She would have filled the gap between Wal-Mart's individualized decisions on pay and promotions with a more sympathetic view both of the plaintiffs' evidence and the substantive law that supported their claims. The majority took the opposite view on this question, insisting on the requirement that the plaintiffs point to a specific employment practice to prove a claim of disparate impact under Title VII. (32) In the end, the decision in Wal-Mart may come to stand as much for its insistence on the strict standards of proof for class claims under Title VII as for its interpretation of the requirements of Rule 23.

The connection between Title VII and Rule 23 goes back several decades. Just as the admonition against considering the merits dates from Eisen, the necessity of looking to the merits dates from the same era. Decisions on both sides were handed down not long after the amendments to Rule 23 in 1966; and specifically under Title VII, the Court authorized a limited inquiry into the merits in General Telephone Co. of Southwest v. Falcon. (33) The decision in Falcon figured prominently in the Court's analysis of commonality in Wal-Mart, first, in explicitly qualifying the admonition in Eisen against considering the merits, and second, in the need to bridge the gap between subjective decision-making and class-wide issues. (34) The Court's reliance upon Falcon also undermines a common criticism of Wal-Mart: that the decision constitutes just another example of the conservative judicial activism typical of the Roberts Court. Conservative though the decision may be, its doctrinal roots lie in structural features of class action practice. In Falcon, the Court was unanimous in reversing certification of the class, eliciting only a partial dissent on the need to remand the case at all. (35) For that matter, Eisen itself was unanimous in forbidding an inquiry into the merits solely in order to shift the costs of notice onto the defendant. (36)

Developments since these early decisions have only succeeded in emphasizing how fundamental they are to the modern version of Rule 23. A proposed amendment to the rule to specifically require an inquiry into "the probable success on the merits of the class claims, issues, or defenses," was thoroughly considered by the Advisory Committee on Civil Rules, but never approved. (37) In the meantime, more modest changes to the rule have been made, for instance, on the timing of certification decisions. (38) Thus the rule as it stands contains no independent requirement of probable success on the merits. Yet, by the same token, many elements of the rule implicate a consideration of the merits. Just to take the most obvious example, a settlement can be approved only if "it is fair, reasonable, and adequate," (39) a finding that requires an assessment of how likely class members were to obtain relief and in what amount and kind. Commonality, typicality, adequacy of representation, the need for class-wide injunctive or declaratory relief, and the predominance of common issues--indeed, the very definition of the class--all require an analysis of the nature of the named plaintiffs' claims. That analysis presupposes that only substantial claims count--those which have survived at least a motion to dismiss for failure to state a claim, which typically precedes a motion for certification. (40) In retrospect, Eisen inevitably had to be limited to accommodate inquiries into the merits that overlap with the requirements of Rule 23.

Falcon not only recognized why the merits mattered on certification, but how they mattered. In this respect, Falcon exhibits a striking similarity to Wal-Mart. It, too, involved the step from individualized, subjective decision-making to claims that the lower court characterized as "across-the-board" discrimination. The Supreme Court disagreed, partly because the plaintiff had lost on his individual claim of discrimination in promotions in the district court, but mainly because this individual claim differed fundamentally from the class claim: it concerned promotions, while the class claim concerned hiring; and it involved evidence of intentional discrimination, while the class claim relied upon statistical evidence of disparate impact. The connection between the plaintiff's individual claim and his class claim was entirely lacking, as was evidence that would have bridged this gap. Wal-Mart presented much the same question since it involved individual instances of discrimination that were scaled up into an attack on general practices of subjective decision-making.

The same problem of bridging the gap between the claims of individual class members and class claims comes up in class actions far afield from employment discrimination, in areas as different as mass torts and securities class actions. The Supreme Court reversed the certification of a "settlement only" class action in Amchem Products, Inc. v. Windsor (41) partly because of "the disparate questions undermining class cohesion" on the asbestos claims asserted on behalf of the class. (42) These concerned different products, different periods and duration of exposure, in different ways, resulting in different injuries (or for some class members, with "exposure only" claims, no injury at all). The substantive law provided no "glue," (43) to use the term from Wal-Mart, to hold the class together. Conversely, in Erica P. John Fund, Inc. v. Halliburton Co., (44) the Court approved certification of a class action based on the presumption of "fraud on the market" applicable to all purchases of securities in an efficient market. (45) The substantive law supplied the question of materiality common to the class that met the requirements of Rule 23. These examples could be multiplied across other areas of law. (46)

Upon examination, the influence of the merits on decisions whether or not to certify a class action is so obvious because the stakes in such decisions are so large. If certification is denied, class members with claims too small to support separate lawsuits have no prospect of recovery. As recognized long ago, denial of certification signals the "death knell" of the action. (47) Granting certification, by contrast, signals to everyone involved that the case deserves much greater attention and the investment of further resources, whether by way of settlement negotiations, discovery, or summary judgment. The gatekeeping function of certification identifies those cases worth the added expense of litigation on a larger scale over a longer period involving greater complexity. Although the formal requirements of Rule 23 depart from a straightforward predictive judgment whether the costs of such litigation might be worth the benefits, they cannot exclude it from the urgent practical considerations that must inform decisions on certification. The merits necessarily go into any such judgment. The question is not whether they do so, but how they should--a subject taken up in the next part of this Article.

II. SOME FORMALISM ABOUT PROCEDURE: CERTIFICATION IN THE STRUCTURE OF A CIVIL ACTION

The correct role of the merits in a decision on certification depends upon the proper place of such a decision in the sequence of steps that constitute a civil action. The earlier the certification decision is made, the more it resembles an issue of pleading and the more it depends upon the same resources and arguments. The later it occurs, the more it can draw on evidence developed for adjudication and trial. Placement ultimately is a matter of logic and function, but it begins as one of timing. The current version of Rule 23 requires that decisions on certification be made at "an early practicable time after a person sues or is sued as a class representative." (48) This provision replaced one in the 1966 version of the rule that was markedly more stringent, requiring certification decisions to be made "as soon as practicable after commencement of the action." (49) The Advisory Committee justified the change to the current language because the previous version of the rule "neither reflects prevailing practice nor captures the many valid reasons that may justify deferring the initial certification decision." (50) The current version of the rule stands suspended somewhere between "as early as practicable" and no later than necessary, or in procedural terms, roughly between pleading and summary judgment.

Justice Scalia gestured in that direction in Wal-Mart when he characterized Rule 23 as setting forth more than "a mere pleading standard. A party seeking class certification must affirmatively demonstrate his compliance with the Rule--that is, he must be prepared to prove that there are in fact sufficiently numerous parties, common questions of law or fact, etc." (51) This passage perhaps minimizes the standards for pleading, which recently have been made stricter, (52) but its emphasis upon proof brings certification of class actions closer to the merits. Plaintiffs must present evidence that the requirements of Rule 23 have been met. This showing nevertheless leaves proof of compliance with the rule at one remove from the merits. The plaintiff need not demonstrate the elements of a claim, but only that proof of those elements exhibits the qualities required by the rule. For certain claims, those qualities might be established with little or no evidence. For instance, suppose that Wal-Mart had a policy that limited the availability of pregnancy leave. (53) A plaintiff could establish common issues simply by proving the existence of this policy, without necessarily establishing whether it complied with Title VII. No further proof would be needed.

The problematic cases resemble Wal-Mart itself, in which a pattern of adverse effects on the basis of race or sex has to be connected to some specific employment practice. (54) These cases also arise more frequently as remaining forms of discrimination become more subtle and more difficult to prove. Employers have every incentive to eliminate general practices that are facially discriminatory, either as a precaution against litigation or in settling claims once litigation occurs. That leaves for litigation more complicated cases implicating practices that are not themselves obviously discriminatory but that might be connected to discriminatory effects. In those cases, the existence of common questions under Rule 23 quickly becomes entangled with the question of class-wide liability. As Justice Scalia framed the problem, relying upon the work of Richard Nagareda, it is "the capacity of a classwide proceeding to generate common answers apt to drive the resolution of the litigation." (55) The existence of a common answer still remains distinct from what that common answer is--either liability or not--although the evidence used to prove the ultimate issue of liability plainly overlaps with the derivative issue of commonality. If the pattern of discriminatory results has no connection to any discernible employment practice, a class cannot be certified for the same reason that the defendant cannot be held liable. Discriminatory effects alone do not violate Title VII. The defendant must be found to have engaged in some form of prohibited discrimination, and it must be connected to the harm; just as in tort law a defendant can be held liable only upon a finding of negligence and proximate cause.

It follows that the plaintiff must present some evidence on the merits in order to demonstrate compliance with Rule 23 when the issues overlap. The crucial question is: how much evidence? Scholars who have previously addressed this issue have proposed the showing necessary to obtain a preliminary injunction: likelihood of success on the merits. (56) That standard suffers from some ambiguity, which might make it either too strong or too weak to assess the merits as they are relevant to certification. If it is taken literally, it appears to be no different from proof of the merits at trial, which must be by a preponderance of the evidence on most issues in ordinary civil litigation. A likelihood of success appears to be the same as proof "more probably than not." In order to avoid this equivalence, courts have discounted the standard for issuance of a preliminary injunction by requiring only "some likelihood" of success or a "better than negligible" chance of success, or more generally, by applying a "sliding scale" that adjusts the necessary showing according to the relative harm to the parties. (57) The greater the harm to the plaintiff in denying the injunction, the lower the threshold of proof, while the greater the harm to the defendant in granting the injunction, the higher the threshold. (58) Conceivably, the standard of proof for certification could be adjusted in similar fashion to take account of the relative harm of denying or granting certification, but the baseline for making any such adjustment would be difficult to discern. It would take the already discretionary standards for preliminary injunctions and make them far more indeterminate by forcing the court to predict the consequences of granting or denying certification.

A further conceptual problem with this standard arises from the mismatch between directly inquiring into the merits and meeting the terms of Rule 23. The merits matter only insofar as they are reflected in the terms of the rule, in such provisions as those on typicality, commonality, and predominance. The ultimate inquiry into the defendant's liability makes a difference only insofar as it affects the derivative inquiry whether that determination can be made on a class-wide basis. "Likelihood of success on the merits" does not capture the derivative role of examining the merits at certification. As discussed earlier, the existence of class-wide issues might be clear even if the issues themselves are not because of ongoing disputes on the merits.

The standard for summary judgment has a much greater affinity to the two-level analysis needed for certification. To oversimplify only a little, the current standard for summary judgment under Rule 56--whether there is a "genuine dispute as to any material fact" (59)--could be modified to fit the elements of Rule 23, and in particular, the requirement of commonality, simply by adding a single phrase--"common to the class as a whole." At least in form, Rule 56 looks to the nature of the issues before the court, although, of course, it also looks to the merits. The rule requires an examination of the record to determine the existence of issues warranting trial, just as a motion for certification searches for issues warranting class-wide determination. When certification involves an examination of the evidence, the accepted standards for summary judgment can be easily adapted to that end.

The adaptation must proceed by recognizing the usual sequence of motions in pretrial practice, from motions to dismiss at the pleading stage, to certification after limited discovery, and then to full discovery and summary judgment. This order sometimes varies, most frequently if the defendant would prefer a quick dismissal of the complaint for failure to state a claim, even if it is not binding on the class. The statute of limitations might then operate to preclude further litigation on behalf of class members who have not filed pending claims. If the defendant (or the plaintiff also) wants to resolve the merits before certification, then the flexibility in timing recognized in the current rule enables them to do so. (60) A motion to dismiss for failure to state a claim provides them with another vehicle for adjudication insofar as questions of law can be resolved before certification.

The existing vehicles for addressing the merits before trial work well enough, in fact, that they raise the question whether a separate inquiry into the merits tailored to the requirements of Rule 23 has any point at all. If the relevant legal issues can be resolved on the pleadings and the triable issues of material fact can be resolved on summary judgment, why look at them again in ruling on a motion for certification? The trial court could just consolidate motions to dismiss and motions for summary judgment with motions to certify. The standard answer goes back to timing and the preclusive effect of a judgment in a class action. The 1966 amendments to Rule 23 sought to eliminate the practice of "one-way intervention" under the previous version of the rule. (61) Under the predecessor to Rule 23(b)(3), "spurious class actions" for damages could result in a judgment entered before membership in the class was conclusively determined. (62) In those cases, class members had the option to take advantage of a favorable judgment on liability by intervening to join the class action and obtain relief, or of staying on the sidelines and avoiding the preclusive effect of a judgment in favor of the defendant. (63) The 1966 version of the rule therefore insisted upon early decisions on certification so that the preclusive effect of any resulting judgment could be settled before a ruling on the merits.

Resolving a motion to dismiss or a motion for summary judgment before certification under the current rule has the same lopsided preclusive effect as one-way intervention under the old rule. If the case is dismissed before the class is certified, the class is not bound by it at all, (64) and in general, the longer certification is delayed, the greater the risk of effectively returning to the practice of one-way intervention. The drafters of the current version of the rule were willing to run this risk, but only within limits. Rule 23(c)(1)(A) still requires rulings on certification at "an early practicable time." (65) Delay, to the extent appropriate, can be accommodated within this language or the adjacent provision in Rule 23(c)(1)(C) allowing rulings on certification to be re-examined at any time in the proceedings. Wal-Mart insists that the plaintiff "affirmatively demonstrate" that the requirements of Rule 23 have been met. (66) Failure to meet that burden ordinarily should result in denial of certification, just as a plaintiffs failure to meet the burden of production in opposition to a motion for summary judgment should ordinarily result in dismissal. (67) By the same token, however, the reasons that support postponing ruling on one motion support postponing ruling on the other. The overriding concern should be to keep rulings on certification, insofar as they implicate the merits, consistent with rulings on summary judgment. If the two get out of sync, certification might be denied despite the existence of class-wide issues that should go to trial. A motion to certify by the plaintiff should not be denied when a motion for summary judgment by the defendant, raising the same issues on the merits, would also be denied.

The two motions, of course, do not entirely overlap. Because of the derivative nature of the inquiry into the merits on certification, it does not replicate all the elements of a ruling on summary judgment but might leave open crucial issues on the merits. Suppose, for instance, that in Wal-Mart the plaintiffs had submitted evidence of specific aspects of the company's "corporate culture" that inclined supervisors and managers to disfavor women, such as a policy against promoting single parents. Even if such a facially neutral policy had a disparate impact on women, it still could be justified as "job related for the position in question and consistent with business necessity." (68) The question whether such a policy applies to the entire class differs from the question whether it violates Title VII, and the former can be answered affirmatively without reaching the latter.

By contrast, if the disputed questions were whether the employer had such a policy or whether the policy had an adverse impact upon women, the merits would be much more deeply intertwined with certification. On one reading of Wal-Mart, the plaintiff would have to prove these facts in order to establish commonality. The policy's existence and effects would have to generate more than common questions. These questions would also have to generate common answers on the defendant's liability to individual class members, which in the absence of any policy or effect, they would not. Yet if the evidence supported reasonable inferences either way on these questions, the necessary common answers should be found to be present. The case should survive a motion for summary judgment and go to trial (or more realistically, the defendant could justifiably be forced into serious settlement negotiations) on class-wide issues of liability. Just as evidence of a general defense of business justification would support class-wide adjudication, the plaintiff's production of evidence supporting a finding that a specific practice caused disparate impact on the basis of sex should do the same. The standard for meeting the burden of production in opposition to a motion for summary judgment should be sufficient to resolve issues on the merits relating to certification.

That standard has other features, explored by Kevin Clermont, in the related context of proving jurisdictional facts that overlap with the merits. He would find prima facie evidence of such overlapping facts sufficient to sustain jurisdiction. (69) Taking "prima facie" to mean sufficient to satisfy the burden of production, his approach would also result in application of the standard for summary judgment. Although he supports a weaker showing in some contexts, he rightly rejects proof of the disputed facts themselves. (70) Such findings of fact, if made in a preliminary jurisdictional ruling, would risk prejudging the merits without adequate discovery, denying the plaintiff the right to jury trial, and re-introducing problems of limited preclusion (analogous to those of one-way intervention under previous class action practice). (71) Most of these problems can be avoided by following the model of a motion for summary judgment which, when it is denied, does not preclude further examination of the merits. Even when it is granted in an individual action, it does not preclude other plaintiffs from bringing similar claims if they can assemble better evidence in support of their claims than the previous plaintiff. Class members after denial of certification are left in no worse position than subsequent plaintiffs in a series of individual actions.

A further advantage of this standard can be discerned in its application to a case that closely resembles Wal-Mart: the Seventh Circuit's decision in McReynolds v. Merrill Lynch, Pierce, Fenner & Smith, Inc. (72) The plaintiff there represented a class consisting of 700 African-American stock brokers whose average compensation lagged behind that of white stock brokers. (73) In that respect, the case resembled Wal-Mart, as it did in the crucial issue whether this disparity in pay could be attributed to a specific employment practice of the defendant. The case departed from Wal-Mart, not just in the relatively small size of the class, but in the plaintiffs' identification of two specific practices that might have contributed to the disparity in pay. (74) Individual brokers could choose to work in teams whose performance in bringing in and servicing accounts was then pooled to determine earnings. (75) African-Americans were chosen to participate in teams at half the rate of whites. (76) Another practice that determined compensation involved "account distribution": the allocation of accounts of departing brokers, which was made based on past performance of the brokers competing for the accounts in the same office. (77) Both practices were established by central management and applied throughout the company although they had subjective components, particularly in the decision of individual brokers to form teams. (78) Both the plaintiffs and On this record, the district court denied certification of a class, but the court of appeals reversed in an opinion by Judge Posner. (80) He first distinguished Wal-Mart because of the presence of specific company policies and their potential effect on compensation, analogizing the formation of teams to the creation of "little fraternities":
 [A]s in fraternities the brokers choose as team members people who
 are like themselves. If they are white, they, or some of them
 anyway, are more comfortable teaming with other white brokers.
 Obviously they have their eyes on the bottom line; they will join a
 team only if they think it will result in their getting paid more,
 and they would doubtless ask a superstar broker to join their team
 regardless of his or her race. But there is bound to be uncertainty
 about who will be effective in bringing and keeping shared clients;
 and when there is uncertainty people tend to base decisions on
 emotions and preconceptions, for want of objective criteria.
 ....

 And likewise with regard to account distributions: if as a
 result of racial preference at the team level black brokers
 employed by Merrill Lynch find it hard to join teams, or at least
 good teams, and as a result don't generate as much revenue or
 attract and retain as many clients as white brokers do, then they
 will not do as well in the competition for account distributions;
 and a kind of vicious cycle will set in. (81)


This passage in the opinion offers no more than impressionistic reasoning about how discrimination by individual brokers could result in disparate impact against the entire class. It certainly could be disputed, as it was by Merrill Lynch, which offered evidence that traced the disparity in pay to the reduced contacts that African-American brokers had, apart from their employment, with potential customers who were wealthy. (82) Whether or not this evidence negates the inferences to be drawn from the plaintiffs statistics on the reduced presence of African-Americans on teams and the reduced value of account distributions to them, (83) as it does pose the right question: whether these direct consequences of the company's policies connected discrimination by individual white brokers with the pattern of lower pay for African-American brokers, and in particular, whether the plaintiffs could survive a motion for summary judgment on this issue.

A version of this last question was presented in Wal-Mart, but the plaintiffs there did less to fill in the missing pieces in a far larger puzzle, leaving even the dissent searching far afield for further evidence. (84) The Seventh Circuit did not have to go so far to certify the class in McReynolds. It might have reached the wrong result, but it did so on grounds distinguishable from those in Wal-Mart--at least on the issue of commonality. In other respects, as we shall see in the next part of this article, the decision could be criticized as inconsistent both with Wal-Mart and with the structure of Rule 23. The Seventh Circuit ordered the class to be certified under Rule 23(b)(2) and limited it to the issue of injunctive relief under Rule 23(c)(4), preserving the ability of class members to individually seek compensatory relief based on a finding of class-wide liability. (85) The need to resort to such expedients, which follow a long tradition of manipulating Rule 23 in order to reach sensible results, depends upon the absence of alternatives more consistent with the literal terms of the rule. Part III takes up this subject and examines the influence, if not the command, of substantive law in devising these alternatives.

III. THE EFFECT OF SUBSTANTIVE LAW: REMEDIES, SETTLEMENT, AND CERTIFICATION

Most of the scholarship on Rule 93 rests on the unspoken assumption that analysis of the rule must be segmented into different fields of substantive law in order to offer an analysis at an informative level of detail. Articles have focused on class actions in areas such as mass torts, securities, civil rights, employment discrimination, consumer fraud, and antitrust. (86) Some are even narrower, examining only asbestos litigation or only tobacco claims. (87) Others take a broader view. One influential study has examined class actions for damages certified under Rule 23(b)(3) and concluded with a chapter on "The Great Big Question About Class Actions." (88) That question, as the authors framed it, is whether damages class actions "on balance, serve the public well." (89) Yet the limitation of their study to damage class actions and their focus primarily upon consumer class actions makes it clear that this single large question quickly dissolves into many subsidiary questions. (90) One of those questions concerns the scope of subdivision (b) (3), as opposed to subdivision (b) (2), an issue that came up in both Wal-Mart and McReynolds.

Judicial decisions do not, in so many words, tailor application of Rule 23 to different claims, but at crucial points they invoke the substantive law to determine whether the rule's requirements have been met. (91) Just as it did on the issue of commonality, the majority in Wal-Mart relied on the substantive law to disapprove of certification under subdivision (b)(2) of class actions seeking back pay. Like damages, the Court reasoned, awards of back pay had to be determined on an individualized basis rather than through statistical approximation, requiring certification of the class under subdivision (b) (3), typically used for damage class actions, rather than subdivision (b) (2), used for injunctions. (92) The Court denounced the resort to statistics in these circumstances as "Trial by Formula" in a part of the opinion joined by all the Justices. (93) The Court's unanimity on this point was all the more surprising because it was unnecessary to the decision. The Court correctly held that class actions seeking substantial amounts of back pay--and in Wal-Mart back pay would have amounted to hundreds of millions of dollars--do not fit within the terms of subdivision (b) (2), which require "final injunctive relief ... appropriate respecting the class as a whole." (94) Individual awards of compensatory relief, whether in the form of damages or back pay, vary from one class member to another and might be denied for some class members altogether. They are not, like a class-wide injunction, "appropriate respecting the class as a whole."

On its face, the opinion in Wal-Mart rejects sampling as any basis for approximate relief, at both the remedy phase of a class action and at certification. (95) Yet the undoubted supremacy of substantive law in determining remedies does not make sampling irrelevant to certification. If it did, it would have the perverse effect of sacrificing the very improvement that class actions offer over individual litigation in enforcing substantive rights. Class actions make the greatest difference in cases when approximate relief provides the only effective means of deterrence and compensation. Nothing in the law on remedies, at least under Title VII, prevents "bifurcation" of a class action into class-wide and individual phases. The opinion in Wal-Mart itself noted this feature of Title VII law, which originated in the first wave of Title VII class actions and government enforcement actions. (96) These decisions, focused upon remedies, presupposed some form of aggregate litigation that can be divided into class-wide and individual components: usually class-wide determination of liability, including consideration of class-wide injunctive or declaratory relief, followed by individualized determinations of individual compensatory relief. In this respect, Title VII does not negate, but instead supports, certification of class actions for individual relief.

After Wal-Mart, class actions seeking back pay must be certified under subdivision (b) (3), with its enhanced requirements of predominance of class-wide issues over individual issues, and if a class is certified, the right of class members to notice and to opt-out. The possibility of "Trial by Formula" can no longer be used to dilute these requirements. (97) Nevertheless, since most class actions settle before trial, and virtually all result in settlement at the remedy stage, even after a finding of liability, "Settlement by Formula" remains a realistic prospect even if "Trial by Formula" is barred. Most class actions result in the award of some form of approximate relief granted by agreement between the parties. The question is whether the evolution of class actions towards "Settlement by Formula" after certification supports consideration of this factor in making the certification decision itself.

The early decisions on remedies under Title VII took the propriety of certification for granted. The decisions established a presumption in favor of back pay upon a finding of class-wide liability. (98) This presumption was then extended to other forms of individual relief, such as remedial seniority, and from private class actions to public pattern-or-practice actions. (99) The latter resemble class actions insofar as they seek relief for a class of victims of discrimination, but they are brought by public officials and they are not binding on class members who choose not to participate in them. (100) In the leading decision on remedies in such actions, International Brotherhood of Teamsters v. United States, the Supreme Court observed that allocating relief among identified victims of discrimination necessarily involves "a degree of approximation and imprecision." (101) Remedial questions that involve a combination of individualized decision-making and necessary approximation have not stood in the way of certification. In damage class actions certified under subdivision (b)(3), the First Circuit has noted that "individuation of damages in consumer class actions is rarely determinative under Rule 23(b)(3)." (102) That proposition might well be tested in Comcast Corp. v. Behrend, (103) which the Supreme Court has recently decided to hear. (104) Among the questions presented there is whether the plaintiff presented adequate evidence of class-wide issues to support certification of an antitrust class action.

Yet wholly apart from adjudication of questions based on approximate relief, settlement on such terms remains the norm. As the Principles of the Law of Aggregate Litigation of the American Law Institute recognize, class actions and other forms of aggregate litigation "may also be settled on terms that may include remedies not available in contested lawsuits." (105) This principle simply codifies existing practice, in which the parties' incentive to settle leads them to adopt approximate forms of relief. As a leading study commented, "most judges anticipate that parties to a mass tort class action will settle the individual damage claims without trial." (106) They forego the cost of continued litigation which, in theory, might result in remedies more precisely tailored to the characteristics of individual class members, but only at the expense of greater litigation. The parties avoid this expense typically by substituting formulas in which the compensation awarded to class members is determined according to characteristics that can be readily ascertained, such as the nature of the injury, the amount of medical expenses, other property loss, and lost income of class members. (107) The formulas then yield tables compiled with multiple variables that approximate the relief due to class members.

The reliance upon such approximations only increases after certification, which leads to settlement at over twice the rate of settlement in cases not certified. (108) This rate further increases after a finding of liability, especially under fee-shifting statutes, like those under Title VII and other civil fights laws. (109) Under those statutes, defendants become liable for the plaintiffs' attorney's fees from the beginning of the action until it is concluded, so long as the plaintiffs obtain significant relief and no "special circumstances" make the award of fees unjust. (110) At the remedy phase of a class action, defendants face a large and growing bill for all the expenses of litigation, both their own and the plaintiffs'. Although the amount of attorney's fees awarded is subject to overall limits based on the degree of the plaintiffs' success on the merits, the actual calculation is by the "lodestar method" of multiplying the hours reasonably spent in the representation by a reasonable hourly rate. (111) A plaintiff's fees in overcoming opposition to individual relief, more often than not, would just be added to the lodestar calculation of fees awarded, leaving the defendant to face the risk of indefinitely rising costs after a finding of liability. Those costs only add to the defendant's incentives to settle, which already have been multiplied by the effect of certification in exposing it to class-wide liability. (112) After Wal-Mart, defendants can insist upon the right to contest the award of compensatory relief to individual class members, but they can exercise that right only if they are willing to foot the entire bill for continued litigation.

To put this point another way, Title VII gives defendants the substantive right to limit the relief ordered against them, and Rule 23 cannot take it away, but other provisions of substantive law can. As Wal-Mart emphasized, the Rules Enabling Act requires that the rule "not abridge, enlarge, or modify any substantive right," (113) and even more fundamentally, the Due Process Clause guarantees defendants the opportunity to be heard on every disputed issue. (114) Yet substantive law can both expand and restrict substantive rights, as fee-shifting statutes do in imposing added costs on defendants after a finding of liability. Of course, defendants might have that finding reversed, but it is the ex ante exposure to liability that compels them to settle, not the possibility of ex post vindication. The dynamics of settlement, not the terms of Rule 23, compromise the value of defendants' procedural rights in the remedy phase of a class action.

Should the incentives for settlement later in the case be taken into account earlier, when the decision whether or not to certify is made? It would be odd if the likely course of litigation after certification had no influence on how the certification decision was made. As we saw in Part I, judges routinely take account of the plaintiffs' likelihood of success on the merits, whether they acknowledge it or not. Since most cases settle, that assessment realistically encompasses the prospect of settlement on terms favorable to the class. Yet it would be equally odd if the likelihood of settlement determined the decision to certify, since certification itself affects the probability of settlement. Cases that are not certified are likely to be dropped, resulting in no significant settlement for the class, while cases that are certified usually proceed immediately to serious settlement negotiations. (115) Certification and settlement could easily become joint components of a self-fulfilling prophecy.

The way out of this circularity is to recognize the way that settlement matters, like the merits matter in Part II: not in determining the ultimate question whether the class will get relief, but in determining the derivative question whether relief could be awarded predominantly by resolution of common issues rather than individual issues. If a class-wide finding of liability exercises only a weak influence over individual class members' entitlement to relief, then common issues are not likely to predominate, either in settlement or in adjudication. Conversely, if such a finding establishes a strong presumption that all class members have been victims of discrimination, then relief is susceptible to efficient class-wide determination and common issues predominate. All of this depends upon the nature of the plaintiffs' claims and the record developed at certification. A plaintiff cannot short-circuit this inquiry by pointing to the likelihood of settlement if the case is certified and reaches the remedy phase. Nor can the defendant do the same by adamantly insisting that it will never settle any aspect of the case.

The willingness of the Supreme Court to entertain "settlement only" class actions, left open in Amchem Products, Inc. v. Windsor, lends support to this conclusion. (116) There the parties had already reached a settlement before the class had been certified. In a contested certification decision, settlement remains only a probability and the court's assessment necessarily remains predictive. Past experience reveals both the frequency and the terms on which class actions are resolved by agreement of the parties. Few, if any, class actions proceed to an indefinite number of individual hearings stretching out, like the Rule Against Perpetuities, beyond lives in being. Amchem dealt with a situation in which the probability of settlement was set at one. It need not be set in all other cases to zero. Established practice demonstrates, in Title VII class actions and in those in other fields, that an intermediate value can be found and that, for the remedy phase of a case, it more closely approaches one than zero.

None of this is to say that a class action could--or could not--have been certified under subdivision (b)(3) in Wal-Mart. The Court's holding that there was no commonality precluded certification under any subdivision of the rule and the Court's opinion went no further than preventing certification under subdivision (b)(2) of class actions seeking significant awards of back pay. Courts have routinely certified cases involving damage class actions under subdivision (b) (3) despite the fact that the relief awarded to individual class members must, if the defendant insists, be done in individualized proceedings. The same might have been true in an alternative version of Wal-Mart in which the threshold showing of commonality had been met. To take an example mentioned earlier, if women had been the victims of discrimination based on a policy of not promoting single parents, the effect of that policy on individual women could be ascertained more easily than the effect of the alleged discrimination in subjective decision-making in the actual case. A narrower class of women would have been affected and the instances of actual discrimination more easily ascertained, enough so that approximate disposition of their cases could feasibly be resolved by settlement.

The same surmise could be made on the facts of McReynolds. Among the 700 African-American stock brokers in that class action, only some would have been victims of the company policies determining pay by team membership and past performance, and only some fraction of their pay would have been affected by those policies. Upon a finding of liability, the parties could have settled the individual claims for back pay on that basis. Instead of predicting a settlement on these grounds, and certifying a class action under subdivision (b) (3), the Seventh Circuit adopted the awkward expedient of certifying one for an injunction under subdivision (b)(2). (117) The court then left the preclusive effect of any judgment of liability to be determined by collateral estoppel, allowing class members to invoke non-mutual offensive issue preclusion in support of any claim they had for back pay. (118) That reasoning, while inventive, circumvents the restraints on this form of issue preclusion--designed to prevent prospective plaintiffs from adopting a wait-and-see strategy with respect to pending litigation (119)--and the similar restraints on class actions under subdivision (b)(3)--designed under the modern rule to prevent "one-way intervention." (120) The court allows class members to take advantage of a favorable judgment and avoid an unfavorable one. Reliance upon the likelihood of settlement, although necessarily predictive and uncertain, does not similarly flout existing restraints on preclusion.

This part of the article has sought to dispel doubts that sampling and statistical approximation have no role to play in certification decisions after Wal-Mart. The Court was right to insist upon the crucial role of substantive law in making such decisions, but wrong to imply that it reinforces anachronistic tendencies that anchor class action practice to traditional procedures appropriate only for individual actions. Rule 23 does have the limited role that the Court assigned to it: creating basic procedures that conform to the Rules Enabling Act and the Due Process Clause. The constitutional requirements of notice and adequacy of representation have certain constant features and the rule as a whole is framed at a sufficiently high level of abstraction to apply to virtually any claim. In these respects the rule aspires to be "trans-substantive," mainly by serving as a skeleton to be fleshed out by reference to substantive law. (121) But the Court was wrong to rely upon substantive law to reject modern methods of sampling. These are supported, not undermined, by Title VII, which from the beginning has relied upon class actions and other aggregate forms of litigation to give relief to individual employees. There is no need, either as a matter of procedure or substance, to abandon this goal and to sacrifice the very improvements that class actions offer over individual litigation.

CONCLUSION

Consistent with the emphasis on substantive law offered in this article, Wal-Mart might be read more for its effect on Title VII than for its interpretation of Rule 23. On that reading, the case stands for the increasing acceptance by the Supreme Court of neutral employment practices that incorporate substantial elements of subjective decision-making. Claims of disparate impact, although still theoretically available against such practices, (122) remain constrained by the need to identify the precise way in which they cause adverse effects upon the plaintiff class. The Court has imposed similar constraints on claims of disparate impact in another recent case, Ricci v. DeStefano, (123) and it might have done so implicitly in Wal-Mart.

A substantive interpretation of the decision is the subject for another article, but it intersects with the analysis offered here. As argued in Parts I and II of this Article, the dominance of substantive law forms the major premise of most applications of the rule, both realistically in how judges make decisions and formalistically in giving content to the rule. Accordingly, as argued in Part III, the continued prevalence of "Settlement by Formula" should serve as an antidote to Wal-Mart's rejection of "Trial by Formula." This example also provides a lesson for future application of the rule: notwithstanding recent decisions of the Supreme Court, substantive law does not always point in a conservative direction. Instead, it points the only way forward after Wal-Mart.

(1) 131 S. Ct. 2541 (2011).

(2) Id. at 2561-67 (Ginsburg, J., concurring in part and dissenting in part).

(3) E.g., Judith Resnik, Fairness in Numbers: A Comment on AT&T v. Concepcion, Wal-Mart v. Dukes, and Turner v. Rogers, 125 HARV. L. REV. 78, 134-54 (2011) (criticizing Wal-Mart for diminishing plaintiffs' access to court); Suzette M. Malveaux, How Goliath Won: The Future Implications of Dukes v. Wal-Mart, 106 NW. U. L. REV. COLLOQUY 34, 35 (2011), available at http://www.law.northwestern.edu/journals/lawreview/ Colloquy/2011/18/LRColl2011n18Malveaux.pdf ("[The decision is a] major blow to the plaintiffs' case because of the unique and powerful role of a class action.").

(4) FED. R. CIV. P. 23(b)(2) ("A class action may be maintained if Rule 23(a) is satisfied and if the party opposing the class has acted or refused to act on grounds that apply generally to the class, so that final injunctive relief or corresponding declaratory relief is appropriate respecting the class as a whole.").

(5) Id.

(6) FED. R. CIV. P. 23(b)(3). Subdivision (b)(3) states:

A class action may be maintained if Rule 23(a) is satisfied and if the court finds that the questions of law or fact common to class members predominate over any questions affecting only individual members, and that a class action is superior to other available methods for fairly and efficiently adjudicating the controversy. The matters pertinent to these findings include:

(A) the class members' interests in individually controlling the prosecution or defense of separate actions;

(B) the extent and nature of any litigation concerning the controversy already begun by or against class members;

(C) the desirability or undesirability of concentrating the litigation of the claims in the particular forum; and

(D) the likely difficulties in managing a class action. Id.

(7) Wal-Mart, 131 S. Ct. at 2557-61 (majority opinion).

(8) Id. at 2550-57.

(9) See id. at 2552 n.6.

(10) See id. at 2562-64 (Ginsburg, J., concurring in part and dissenting in part) (giving examples of sex discrimination resulting from subjective decision-making).

(11) Id. at 2552-57 (majority opinion).

(12) 417 U.S. 156, 177 (1974).

(13) Wal-Mart, 131 S. Ct. at 2551-52.

(14) See, e.g., Gen. Tel. Co. of SW. v. Falcon, 457 U.S. 147, 160 (1982) ("Sometimes the issues are plain enough from the pleadings to determine whether the interests of the absent parties are fairly encompassed within the named plaintiff's claim, and sometimes it may be necessary for the court to probe behind the pleadings before coming to rest on the certification question."); Robert G. Bone & David S. Evans, Class Certification and the Substantive Merits, 51 DUKE L.J. 1251, 1279, 1329 (2002) (recommending a standard of likelihood of success on the merits similar to the standard used for preliminary injunctions).

(15) John C. Coffee, Jr., Litigation Governance: Taking Accountability Seriously, 110 COLUM. L. REV. 288, 350 (2010). Less kindly, he found scholarship focused on the general terms of procedural rules to be "characterized by an incisive rigor--and a rule-bound tunnel vision." Id. at 289.

(16) Richard Marcus, Reviving Judicial Gatekeeping of Aggregation: Scrutinizing the Merits on Class Certification, 79 GEO. WASH. L. REV. 324, 349-54 (2011).

(17) Wal-Mart, 131 S. Ct. at 2560 (citing the remedial provisions of Title VII to classify backpay as neither an injunctive nor declaratory relief).

(18) Id. at 2560-61.

(19) 28 U.S.C. [section] 2072 (2006).

(20) U.S. CONST. amend. V.

(21) See infra text accompanying notes 86-90.

(22) Int'l Bhd. of Teamsters v. United States, 431 U.S. 324, 342 n.23 (1977) (quoting United States v. T.I.M.E.-D.C., Inc., 517 F.2d 299, 315 (5th Cir. 1975)).

(23) Dukes v. Wal-Mart Stores, Inc., 222 F.R.D. 137, 155 (N.D. Cal. 2004), aff'd in part and remanded in part, 603 F.3d 571 (9th Cir. 2010), rev'd, 131 S. Ct. 2541 (2011).

(24) Id. at 146.

(25) Id. at 145-66.

(26) Id. at 159.

(27) Wal-Mart, 131 S. Ct. at 2554 (alteration in original).

(28) Id. at 2555.

(29) Id. at 2562-64 (Ginsburg, J., concurring in part and dissenting in part).

(30) Id. at 2563 (footnote omitted).

(31) See id. at 2565 (interpreting a prior decision to show the Court's awareness of "the problem of subconscious stereotypes and prejudices" (quoting Watson v. Fort Worth Bank & Trust, 487 U.S. 977, 990 (1988))).

(32) 42 U.S.C. [section] 2000e-2(k) (2006).

(33) 457 U.S. 147 (1982). The Court in Falcon heavily relied on the reasoning in Coopers & Lybrand v. Livesay that "the class determination generally involves considerations that are 'enmeshed in the factual and legal issues comprising the plaintiff's cause of action.'" 437 U.S. 463, 469 (1978) (quoting Mercantile Nat'l Bank v. Langdeau, 371 U.S. 555, 558 (1963)).

(34) Wal-Mart, 131 S. Ct. at 2552-53 & n.6 (majority opinion).

(35) Falcon, 457 U.S. at 161 (Burger, C.J., concurring in part and dissenting in part). The precedent on which Falcon relied, East Texas Motor Freight System, Inc. v. Rodriguez, 431 U.S. 395 (1977), also was unanimous.

(36) See Eisen v. Carlisle & Jacquelin, 417 U.S. 156, 179 (1974) (Douglas, J., dissenting in part) (expressing "general agreement with the phases of this case touched on by the Court").

(37) FED. R. CIV. P. 23(b)(3)(E) (Tentative Draft, 1995), available at http://www.us courts.gov/uscourts/Rules AndPolicies/rules/Reports/CV12-1995.pdf (Memorandum from Patrick E. Higginbotham, Chair, Advisory Committee on Federal Civil Rules, to Members of the Standing Committee on Rules).

(38) FED. R. CIV. P. 23(c)(1)(A).

(39) FED. R. CIV. P. 23(e)(2).

(40) See THOMAS E. WILLGING ET AL., FED. JUDICIAL CTR., EMPIRICAL STUDY OF CLASS ACTIONS IN FOUR FEDERAL DISTRICT COURTS 31 (1996).

(41) 521 u.s. 591 (1997).

(42) Id. at 624 (citing reasoning in Georgine v. Amchem Prods., Inc., 83 F.3d 610, 626 (3d Cir. 1996)).

(43) Wal-Mart Stores, Inc. v. Dukes, 131 S. Ct. 2541, 2552 (2011).

(44) 131 S. Ct. 2179 (2011).

(45) Id. at 2185.

(46) See 7AA CHARLES ALAN WRIGHT ET AL., FEDERAL PRACTICE AND PROCEDURE: CIVIL [section][section] 1776, 1781-1783 (3d ed. 2005 & Supp. 2012) (citing cases on application of Rule 23 to class claims in constitutional law, civil rights, antitrust, securities fraud, consumer law, and environmental law).

(47) See Coopers & Lybrand v. Livesay, 437 U.S. 463, 469-77 (1978) (denying interlocutory appeal on that ground before amendment adding Rule 23(f) on appeals).

(48) FED. R. CIV. P. 23(c)(1)(A).

(49) UNITED STATES SUPREME COURT, FEDERAL RULES OF CIVIL PROCEDURE: As AMENDED THROUGH JULY 1, 1966 (The Foundation Press, Inc. 1967).

(50) FED. R. CIV. P. 23(c)(1)(A) advisory committee's notes (2003 amendments).

(51) 131 S. Ct. 2541, 2551 (2011).

(52) Ashcroft v. Iqbal, 556 U.S. 662, 678 (2009) ("[A] complaint must contain sufficient factual matter, accepted as true, to 'state a claim to relief that is plausible on its face.'" (quoting Bell Atlantic Corp. v. Twombly, 550 U.S. 544, 570 (2007))).

(53) Cf. Newport News Shipbuilding & Dry Dock Co. v. EEOC 462 U.S. 669 (1983) (holding that the Pregnancy Discrimination Act makes it clear that it is unlawfully discriminatory for an employer to exclude pregnancy coverage from an otherwise inclusive benefits plan).

(54) Wards Cove Packing Co. v. Atonio, 490 U.S. 642, 643 (1989).

(55) Wal-Mart, 131 S. Ct. at 2551 (quoting Richard Nagareda, Class Certification in the Age of Aggregate Proof 84 N.Y.U.L. REV. 97, 131-32 (2009)).

(56) Bone & Evans, supra note 14, at 1329-31; Marcus, supra note 16, at 349. The American Law Institute would go further and require resolution of relevant disputed issues of fact by a preponderance of the evidence, although it would not make such a finding binding at trial on the merits. AM. LAW INST., PRINCIPLES OF THE LAW: AGGREGATE LITIGATION [section] 2.06 (2010).

(57) For all of these strategies invoked in a single opinion, see Ty, Inc. v. Jones Grp., Inc., 237 F.3d 891, 895, 897 (7th Cir. 2001).

(58) Lawson Prods., Inc. v. Avnet, Inc., 782 F.2d 1429, 1433-34 (7th Cir. 1986). Some courts have applied a version of the Hand formula, awarding a preliminary injunction only if the harm to the plaintiff from denying the injunction, multiplied by the plaintiff's likelihood of success at trial, exceeds the harm to the defendant from granting the injunction, multiplied by the defendant's probability of success. See United States v. Carroll Towing Co., 159 F.2d 169, 173 (2d Cir. 1947) (formulating the Hand formula). But Professor Laycock points out that not only do the variables involved resist any kind of factual quantification, but that the variables cannot even be "conceptualized ... in theory as having discrete values.... Rather, these variables stand for ranges of possible developments, with the probabilities changing at every point of the range." DOUGLAS LAYCOCK, THE DEATH OF THE IRREPARABLE INJURY RULE 120 (1991). Problems of variable factual and conceptual precision are only multiplied in the class action context.

(59) FED. R. CRV. P. 56(a).

(60) See FED. R. CIV. P. 23(c)(1)(A) advisory committee's notes (2003 amendments).

(61) See FED. R. CIV. P. 23(c)(3) advisory committee's notes (1966 amendments) ("Under proposed subdivision (c)(3), one-way intervention is excluded; the action will have been early determined to be a class or nonclass action, and in the former case the judgment, whether or not favorable, will include the class, as above stated.").

(62) Id.

(63) See FED. R. CIV. P. 23(c)(1)(A) advisory committee's notes (1966 amendments).

(64) Smith v. Bayer Corp., 131 S. Ct. 2368, 2380 (2011) (holding that if a class is not certified, there is no preclusion of later action seeking class certification); Taylor v. Sturgell, 553 U.S. 880, 900-01 (2008) (holding that there is no preclusion by virtual representation without privity with party to prior action).

(65) FEB. R. CIV. P. 23(c)(1)(A).

(66) Wal-Mart Stores, Inc. v. Dukes, 131 S. Ct. 2541, 2551 (2011).

(67) Celotex Corp. v. Catrett, 477 U.S. 317, 322 (1986).

(68) Civil Rights Act of 1991, 42 U.S.C. [section] 2000e-2(k) (1) (A) (i) (2006).

(69) Kevin M. Clermont, Jurisdictional Fact, 91 CORNELL L. REV. 973, 973 (2006).

(70) Id. at 988-90, 998.

(71) Id. at 990-92. For similar arguments, see Marcus, supra note 16, at 354-68; George Rutherglen, Title VII Class Actions, 47 U. CHI. L. REV. 688, 727-30 (1980).

(72) 672 F.3d 482 (7th Cir. 2012), cert denied, 81 U.S.L.W. 3062 (Oct. 1, 2012). Another case with the same name subsequently was filed alleging discrimination against in the operation of Merrill Lynch's bonus and retention program. McReynolds v. Merrill Lynch, Pierce, Fenner & Smith, 694 F.3d 873, 878-79 (7th Cir. 2012). That case was dismissed for failure to state a claim before a class was certified, but not for reasons that cast doubt upon certification of the class in the earlier case. Id. at 888-89 (reasoning that the class in the later case could pursue their claims, to the extent they were meritorious, in the earlier case). the defendant submitted expert testimony and statistical evidence on these practices and their effect on compensation. (79)

(73) McReynolds, 672 F.3d at 483, 490.

(74) Id. at 488.

(75) Id. at 488-89.

(76) Expert Report of Janice Fanning Madden & Alexander Vekker for Plaintiffs at 68, McReynolds v. Merrill Lynch, Pierce, Fenner & Smith, 2010 WL 3184179 (N.D. Ill. Aug. 9, 2010) (No. 05 C 6583).

(77) Id.

(78) Id.

(79) McReynolds v. Merrill Lynch, Pierce, Fenner & Smith Inc., 2010 WL 3184179, at *4, *6 (N.D. Ill. Aug. 9, 2010) (No. 05 C 6583).

(80) McReynolds, 672 F.3d at 482.

(81) Id. at 489-90.

(82) McReynolds, 2010 WL 3184179, at *6.

(83) Id. at *5.

(84) Justice Ginsburg relied, for instance, upon evidence of sex discrimination in orchestral auditions. Wal-Mart Stores, Inc. v. Dukes, 131 S. Ct. 2541, 2564 n.6 (2011) (Ginsburg, J., concurring in part and dissenting in part).

(85) McReynolds, 672 F.3d at 491-92.

(86) See, e.g., Richard A. Booth, Class Conflict in Securities Fraud Litigation, 14 U. PA. J. BUS. L. 701 (2012) (arguing that securities fraud class actions should be treated as derivative actions because of the unique issues of adequate representation in the securities fraud legal context); Coffee, supra note 15, at 318-25 (analyzing developments under Private Securities Litigation Reform Act); Brandon Garrett, Aggregation and Constitutional Rights, 88 NOTRE DAME L. REV. 593 (2012) (arguing for a prominent role for class actions to enforce constitutional rights); David Marcus, Flawed But Noble: Desegregation Litigation and Its Implications for the Modern Class Action, 63 FLA. L. REV. 657, 678-94 (2011) (examining the origins of Rule 23(b)(2) and its assumptions about solidarity among civil rights plaintiffs); David Rosenberg, Mandatory-Litigation Class Action: The Only Option for Mass Tort Cases, 115 HARV. L. REV. 831, 838 (2002) (arguing for the relevance of tort law as opposed to a narrow "proceduralist" analysis of class actions).

(87) See, e.g., Samuel Issacharoff, "Shocked" Mass Torts and Aggregate Asbestos Litigation After Amchem and Ortiz, 80 TEX. L. REV. 1925, 1930-32 (2002) (grounding analysis of mass torts class actions in the empirical specifics of modern asbestos litigation) ; see also Elizabeth J. Cabraser, Unfinished Business: Reaching the Due Process Limits of Punitive Damages in Tobacco Litigation Through Unitary Classwide Adjudication, 36 WAKE FOREST L. REV. 979 (2001) (analyzing the impact of class actions on tobacco tort litigation); Robert L. Rabin, The Tobacco Litigation: A Tentative Assessment, 51 DEPAUL L. REV. 331, 332-47 (2001) (analyzing tobacco class actions and their effects).

(88) DEBORAH R. HENSLER ET AL., CLASS ACTION DILEMMAS 401-70 (2000).

(89) Id. at 401.

(90) They readily concede the significance of the merits and substantive law in assessing the overall utility of class actions, even if they recognize how elusive the answers to even narrower questions are. Id. at 416-24.

(91) See Ortiz v. Fibreboard Corp., 527 U.S. 815, 845 (1999) (emphasizing the risk of abridging substantive rights under state tort law of asbestos victims based on a liberal interpretation of "limited fund" class actions under Rule 23(b)(1)(B)); Amchem Prods., Inc. v. Windsor, 521 U.S. 591,594-95 (1997) (finding no adequate representation because of the diversity of present and potential medical injuries from exposure to asbestos); In re Rhone-Poulenc Rorer Inc., 51 F.3d 1293, 1300-01 (7th Cir. 1995) (rejecting certification of a class in part because of varied state tort law).

(92) Wal-Mart Stores, Inc. v. Dukes, 131 S. Ct. 2541, 2557 (2011).

(93) Id. at 2561.

(94) FED. R. CIV. P. 23(b)(2).

(95) Wal-Mart, 131. S. Ct. at 2561.

(96) Id.

(97) Id.

(98) Albemarle Paper Co. v. Moody, 422 U.S. 405, 421 (1975).

(99) Int'l Bhd. of Teamsters v. United States, 431 U.S. 324, 334-35 (1977); Franks v. Bowman Transp. Co., 424 U.S. 747, 767-68 (1976).

(100) Gen. Tel. Co. v. EEOC, 446 U.S. 318, 332-33 (1980).

(101) Teamsters, 431 U.S. at 372.

(102) Smilow v. Sw. Bell Mobile Sys., Inc., 323 F.3d 32, 40 (1st Cir. 2003). For decisions after Wal-Mart certifying class actions under Rule 23(b) (3) despite the presence of individual claims for damages, see In re Whirlpool Corp. Front-Loading Washer Prods. Liab. Litig., 678 F.3d 409, 421 (6th Cir. 2012); Ross v. RBS Citizens, N.A., 667 F.3d 900, 909 n.7 (7th Cir. 2012); In re Zurn Pex Plumbing Prods. Liab. Litig., 644 F.3d 604, 618-20 (8th Cir. 2011); Sullivan v. DB Invs., Inc., 667 F.3d 273, 296-301 (3d Cir. 2011) (en banc).

(103) Comcast Corp. v. Behrend, 11-864, 2012 WL 113090 (U.S. June 25, 2012) (granting certiorari).

(104) Id.

(105) AM. LAW INST., PRINCIPLES OF THE LAW OF AGGREGATE LITIGATION [section] 3.01(b) (2009).

(106) HENSLER ET AL., supra note 88, at 111.

(107) The case studies of damage class actions in HENSLER ET AL., supra note 88, all resulted in settlements approximating relief in this fashion. Id. at 454-59. For examples of such "settlement grids" see Issacharoff, supra note 87, at 1928-29, 1934-35 (noting agreement on value of asbestos claims and use of settlement grids); George Rutherglen, Distributing Justice: The September 11th Victim Compensation Fund and the Legacy of the Dalkon Shield Claimants Trust, 12 VA. J. Soc. POL'Y & L 673, 692-95 (2005) (noting use of presumptive awards in Dalkon Shield and September 11th settlement funds).

(108) WILLGING ET AL., supra note 40, at 60 (higher rate of settlement of certified cases, not counting those certified only for settlement).

(109) 42 U.S.C. [section][section] 1988, 2000e-5(k) (2006).

(110) Christiansburg Garment Co. v. EEOC, 434 U.S. 412, 416-17 (1978) (citing Newman v. Piggie Park Enters., 390 U.S. 400, 402 (1968)).

(111) Hensley v. Eckerhart, 461 U.S. 424, 433 (1983).

(112) In re Rhone-Poulenc Rorer, Inc., 51 F.3d 1293, 1298 (7th Cir. 1995). The "in terrorem" of effect of certification upon defendants has been widely accepted in the secondary literature and among the class action bar. HENSLER ET AL., supra note 88, at 106-08; Bone & Evans, supra note 14, at 1291-1305.

(113) 28 U.S.C. [section] 2072 (2006).

(114) Mullane v. Cent. Hanover Bank & Trust Co., 339 U.S. 306, 314 (1950) ("An elementary and fundamental requirement of due process in any proceeding which is to be accorded finality is notice reasonably calculated, under all the circumstances, to apprise interested parties of the pendency of the action and afford them an opportunity to present their objections.").

(115) WILLGING ET AL., supra note 40, at 60-61, 179-80.

(116) 521 U.S. 591, 620 (1997).

(117) McReynolds v. Merrill Lynch, Pierce, Fenner & Smith, Inc., 672 F.3d 482, 491-92 (7th Cir. 2012).

(118) Id. at 492.

(119) Parklane Hosiery Co. v. Shore, 439 U.S. 322, 331 (1979) (limiting this form of preclusion "where a plaintiff could easily have joined in the earlier action").

(120) See supra note 61.

(121) See Robert M. Cover, For James Win. Moore: Some Reflections on a Reading of the Rules, 84 YALE L.J. 718, 732-33 (1975) (expressing skepticism about procedural rules that are interpreted uniformly to cover a wide range of claims).

(122) See Wal-Mart Stores, Inc. v. Dukes, 131 S. Ct. 2541, 2554 (2011) (acknowledging the application of disparate impact to subjective decision-making).

(123) Ricci v. DeStefano, 557 U.S. 557, 583 (2009) (finding that a test with disparate impact could have been justified as job related).
COPYRIGHT 2012 University of Notre Dame Law School
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2012 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:class action procedure reform
Author:Rutherglen, George
Publication:Notre Dame Law Review
Date:Dec 1, 2012
Words:12447
Previous Article:Killing them with kindness: examining "consumer-friendly" arbitration clauses after AT&T Mobility v. Concepcion.
Next Article:Regulatory litigation in the European Union: does the U.S. class action have a new analogue?
Topics:

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters