IN RE: Jeh Charles JOHNSON, Secretary of the U.S. Department of Homeland Security, Petitioner.
Eight AfricanAmerican Secret Service agents who were denied promotions to the GS–14 or GS–15 level, allegedly because of their race, were certified by the district court to sue the Secretary of the Department of Homeland Security on behalf of a class comprising all similarly situated agents, of whom there are approximately 120. The Government argues the plaintiffs are not eligible to proceed as a class under Federal Rule of Civil Procedure 23 because they do not meet the requirements of Rule 23(a) that there be “questions of law or fact common to the class” and that “the representative parties will fairly and adequately protect the interests of the class,” nor the requirements of Rule 23(b)(3) that “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 Government asks this court to exercise its discretion under Rule 23(f) to grant interlocutory review of the class certification order on the ground either that the decision is “manifestly erroneous” or that it presents “an unsettled and fundamental issue of law relating to class actions ․ that is likely to evade end-of-the-case review.” In re Lorazepam & Clorazepate Antitrust Litig., 289 F.3d 98, 105 (D.C.Cir.2002). Because we conclude neither of these reasons for Rule 23(f) review is applicable, we shall not at this time review the order of the district court certifying the class.
I. Background *
The Secret Service employs about 3,000 Special Agents in seven major offices, each headquartered in Washington, D.C. Each office is run by an Assistant Director who reports to the Deputy Director and the Director of the Secret Service.
The Secret Service used a multistage process known as the Merit Promotion Plan (MPP) to select Special Agents for promotion to the GS–14 and GS–15 positions at issue in this case. In the first stage of the process, a Special Agent received numerical scores from (1) his or her immediate supervisor (First Level Evaluation), (2) a panel comprising a representative from each Assistant Director's office (Second Level Evaluation), and (3) for promotion to a GS–14 position, a panel comprising six peers and two alternates who are selected and convened annually by the Deputy Director (Peer Panel Evaluation). The MPP prescribed the criteria each of the three groups had to use to assign scores and the training that scorers received, provided common evaluation forms to scorers and applicants, and assigned weights to each of the component scores in order to come up with a final MPP score for each candidate. A Special Agent seeking a promotion used his or her final MPP score to “bid” for one or more available positions. Bidders for each position were ranked by their scores and the top 30 or the top 25%, whichever was greater, were placed on a “best qualified list” (BQL). Then the Advisory Board, comprising the Deputy Director, the seven Assistant Directors, and the Chief Counsel, reviewed the BQL for each position and made a recommendation to the Director of the Secret Service, who made the final promotion decision.
The named plaintiffs in this suit are current and former African–American Special Agents who bid for but did not receive GS–14 or GS–15 promotions under the MPP in the period from 1995 to 2005. They allege both that the Secret Service engaged in a pattern or practice of racial discrimination in making promotions and that the MPP had a disparate impact upon African–American Special Agents seeking promotions, in violation of Title VII of the Civil Rights Act of 1964, 42 U.S.C. § 2000e et seq., and the Civil Rights Act of 1991, 42 U.S.C. § 1981a. The plaintiffs request relief in the form of back pay and compensatory damages to the extent the discrimination cost them promotions, and also ask for a declaratory judgment and an injunction requiring the Secret Service to end its allegedly discriminatory promotion practices. In order to support their two theories of discrimination, the plaintiffs proffered evidence that African–American Special Agents were disadvantaged both in the scoring process and in the final selection process.
BQLs were used to narrow the pool of applicants for approximately half the vacancies that arose during the class period; for the other half, there were few enough applicants that all were placed on the relevant BQL. The plaintiffs have proffered statistical evidence that the use of scores pursuant to the MPP disproportionately disqualified African–American Special Agents from reaching the BQLs throughout the class period. Additionally, their statistical evidence shows that when African–American Special Agents were included in the BQLs, they had lower mean ranks than would be expected in the absence of discrimination and more frequently failed to score as high as did the lowest-scoring agent who was promoted that year. The plaintiffs have also proffered statistical evidence that even when African–American Special Agents made it onto the BQLs, for certain years in the class period (1998 to 2000 for GS–14 promotions and 2002 to 2005 for GS–15 promotions), fewer of them were selected for promotion than would be expected in the absence of discrimination. Additionally, the plaintiffs have pointed to a “substantial continuity” throughout the class period among the personnel serving on the rating panels, on the Advisory Board, and as the Director, suggesting class members experienced in common whatever racial bias may have affected the subjective elements of the promotions process.
The named plaintiffs seek to represent the class of all similarly situated African–American Special Agents. The order here challenged by the Government was issued after the plaintiffs' fourth attempt to define the class so as to meet the requirements of Rule 23 for certification. Rule 23(a) requires that (1) the class be so numerous that joinder is impracticable, (2) there be at least one question essential to the resolution of the suit that can be answered in common for all members of the class, (3) the claims of the named plaintiffs be typical of the claims of all class members, and (4) the representative parties, i.e. the named plaintiffs and their counsel, will fairly and adequately protect the interests of the class. Rule 23(b)(3) requires a class seeking damages to show 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.” FED. R. CIV. P. 23(b)(3).
The class, as certified by the district court, comprises the following individuals, numbering approximately 120:
[A]ll current and former African–American Special Agents who bid for promotion to a GS–14 position from 1995 to 2004 and were not promoted to GS–14 on the first bid list on which they bid; and all current and former African–American Special Agents who bid for promotion to a GS–15 position from 1995 to 2005 and were not promoted to GS–15 on the first bid list on which they bid; but excluding Special Agents who served as an Assistant Director, a Deputy Director, or the Director of the Secret Service during the class period.
Moore v. Napolitano, 926 F.Supp.2d 8, 35 (2013). The plaintiffs arrived at this definition after having been denied class certification for including in the putative class Special Agents against whom class members had made “direct accusations of discrimination.” Id. at 31–32 (quoting Moore v. Napolitano, 269 F.R.D. 21, 34 (2010)). Finding no such conflict of interest in a class excluding those agents, the district court determined the plaintiffs met the adequacy of representation requirement. Id. at 32. The court also analyzed the admissibility of the plaintiffs' statistical evidence and determined it was reliable and relevant to the question of discrimination vel non under Title VII. Id. at 27. Having determined that each of the four requirements of Rule 23(a) was met, that common issues predominated over individual ones, and that the class device was superior to other methods, the court certified the class under Rule 23(b)(3). Id. at 35. The Government then petitioned for interlocutory review per Rule 23(f), challenging the court's determinations with respect to commonality, adequacy of representation, predominance, and superiority.
This court has identified three reasons for which interlocutory review of a class certification order is appropriate under Rule 23(f):
(1) when a “questionable” class certification decision creates a “death-knell situation” for either party; (2) when the certification decision presents “an unsettled and fundamental issue of law relating to class actions ․ that is likely to evade end-of-the-case review”; and (3) when the certification decision is manifestly erroneous.
In re Veneman, 309 F.3d 789, 794 (2002) (quoting Lorazepam, 289 F.3d at 105). Other circumstances may also justify review per Rule 23(f), but we have cautioned that such review should be “granted rarely.” Lorazepam, 289 F.3d at 105.
The Government here invokes both the second and third reasons and additionally argues this case presents “special circumstances” of the sort present In re Rail Freight Fuel Surcharge Antitrust Litig ., 725 F.3d 244 (D.C.Cir.2013), where a “confluence of multiple rationales,” including a subsequent decision of the Supreme Court, justified immediate review of the class certification order. Id. at 250. None of these three justifications obtains in the present case.
A. Unsettled and Fundamental Issue Likely to Evade End–of–the–Case Review
The Government fails to demonstrate this case presents any unsettled and fundamental issue of law relating to class actions that is likely to evade end-of-the case review. To be sure, the Government points out areas of ambiguity in the law of class actions that are relevant to its case, but it fails to provide any reason we should resolve those issues now: The Government's briefs simply ignore the requirement that the issues be “likely to evade end-of-the-case review.” Lorazepam, 289 F.3d at 105. When pressed at oral argument the Government suggested settlement pressure would prevent end-of-the-case review, but that is no more than a belated argument that it faces a “death-knell situation.” Even if it had been timely raised, the argument would be misplaced, for it is appropriate only when “the grant of class status raises the cost and stakes of the litigation so substantially that a rational defendant would feel irresistible pressure to settle.” Rail Freight, 725 F.3d at 251. The costs and stakes here are insignificant in relation to the litigating resources of the defendant, upon which it has drawn for more than 14 years in tenaciously defending this case. We are confident, therefore, that any novel legal question raised by this case can be addressed after judgment.
B. Manifest Error
The Government next contends the class certification decision is manifestly erroneous. This is a difficult standard to meet; we have never before granted Rule 23(f) review on the basis of a manifest error and other circuits, too, have indicated there is a high bar for doing so. See, e.g., Chamberlan v. Ford Motor Co., 402 F.3d 952, 962 (9th Cir.2005) (“It is difficult to show that a class certification order is manifestly erroneous unless the district court applies an incorrect Rule 23 standard or ignores a directly controlling case. Class certification decisions rarely will involve legal errors, however, simply because class actions typically involve complex facts that are unlikely to be on all fours with existing precedent.” (citations omitted)).
Despite its need to clear this high bar, the Government has made almost no effort to explain which of the district court's alleged errors it thinks manifest. Out of an abundance of caution, therefore, we consider whether any of its challenges to the class certification requirements—commonality, adequacy of representation, predominance, and superiority—is based upon an error that can be deemed manifest. We conclude that none of these rulings on the requirements for certification is manifestly erroneous because the court applied the correct standards and the cases relied upon by the Government do not squarely foreclose the class certification here. We therefore decline to exercise our discretion under Rule 23(f) to conduct a more searching inquiry now into the application of those standards to the facts here.
The district court found the commonality requirement was satisfied because every class member's claim of employment discrimination could be moved toward resolution by answering the common contention, supported by some evidence, that there was a policy or practice of racial discrimination infecting the promotion process at the Secret Service in the period 1995–2005. Moore, 926 F.Supp.2d at 28–30. The Government argues this conclusion is at odds with the Supreme Court's decision in Wal–Mart Stores, Inc. v. Dukes, ––– U.S. ––––, 131 S.Ct. 2541, 2550–57 (2011), in which a putative class alleging employment discrimination failed the commonality requirement. The Court there stated that “[w]ithout some glue holding the alleged reasons for all [the promotion] decisions together, it will be impossible to say that examination of all the class members' claims for relief will produce a common answer to the crucial question why was I disfavored.” Id. at 2552. According to the Government, there is no such glue in this case because class members allege discrimination at different stages in the MPP process, where they faced different decision makers. Id. at 2551 (class “claims must depend upon a common contention for example, the assertion of discriminatory bias on the part of the same supervisor”). The Government also relies upon this court's decision in DL v. District of Columbia, 713 F.3d 120 (2013), in which, as in Wal–Mart, class members' experiences varied to the point that their complaints could not be resolved together. The facts in the present case, however, are sufficiently different from the facts in those cases, and the claims of the class members here are sufficiently alike, so as not squarely to foreclose the district court's determination, in effect, that it is indeed possible to find a common answer to the crucial question why was I disfavored. See Wal–Mart, 131 S.Ct. at 2551, quoting Richard A. Nagareda, Class Certification in the Age of Aggregate Proof, 84 N.Y.U. L.Rev. 97, 132 (2009) (“What matters to class certification ․ is not the raising of common ‘questions'—even in droves—but, rather, the capacity of a class-wide proceeding to generate common answers apt to drive the resolution of the litigation”).
In Wal–Mart, the only feature common to all promotion decisions was the policy of delegating those discretionary decisions to individual store managers. See id. at 2554. In essence, all they had in common was that each one (or, more precisely, each manager's decision) was different. In contrast, under the MPP here at issue, every class member was evaluated upon the same criteria and scored using the same numerical system. Furthermore, every promotion decision was ultimately made by the Director of the Secret Service. Although different decision makers no doubt injected some subjectivity into the evaluations of different class members, which might be enough to cast doubt upon the commonality of their claims were we engaged in a more searching inquiry, the policy here was far from the complete delegation involved in Wal–Mart; hence, if the district court erred in coming to a different conclusion, then the error is certainly not manifest and so need not be reviewed at this stage in the litigation.
In DL, all the class members alleged they had been deprived of a “free appropriate public education” as a result of various failings at different stages in the District of Columbia's process to “identify, locate, evaluate, and offer special education and related services to disabled preschool-age children.” See 713 F .3d at 122. Because the class members were suing for injunctive relief to obtain improvements at various different stages in the delivery of disability services, the court held there was no question capable of common resolution for all class members. See id. at 125, 127–28. Here, all class members are seeking damages to compensate them for the allegedly discriminatory decisions not to promote them. The Government seeks to differentiate class members by the stage of the promotion process in which they may have encountered discrimination; for instance, it suggests class members who applied for promotions in years for which there was no statistically significant disparity in the BQL-to-promotion rate are in actuality complaining only about discrimination from the scoring and ranking process, not about discrimination in management's subjective decision-making process. These distinctions, however, are not indicated by DL because all the class members here are seeking the same relief and, in so doing, must all show that promotion decisions made pursuant to the MPP were affected by discrimination generally. Because there is a good deal of commonality in the way all those promotion decisions were made, the district court did not manifestly err in finding the class claims could be moved together toward resolution.
2. Adequacy of representation
The Government next contends the district court erred in holding the named plaintiffs could adequately represent the entire class because some class members had served as scorers in the evaluation process at the first stage of the MPP. The Government argues this decision “cannot be squared with” Wagner v. Taylor, 836 F.2d 578 (D.C.Cir.1987), in which we held the named plaintiff could not adequately represent the class in part because he had “accused his own supervisor, who is a potential class member, of racial discrimination.” 836 F.2d at 595. In order to eliminate just such a conflict, the present plaintiffs amended the definition of the putative class in their motion for certification specifically to exclude Special Agents who were part of upper management and were therefore responsible for selecting candidates for promotion from the BQLs.** The district court then distinguished Wagner and found the named plaintiffs could adequately represent the class on the basis that “the record does not show specific allegations of discrimination by class members against other class members.” Moore, 926 F.Supp.2d at 32. We see no manifest error in this ruling. If class members neutrally applied a flawed rating system and thereby reached a discriminatory result, then they were not themselves discriminating and therefore have no apparent interest that is in conflict with the attempt to prove other agents were denied promotions because of their race.
3. Predominance of the common issues
The district court held common issues predominated over individual issues in this suit because there are no individual issues involved in determining whether the MPP was discriminatory and because the court, when and if it ever reaches individualized questions, will send those questions to separate hearings outside the class proceedings. Moore, 926 F.Supp.2d at 33–34 & n.7. In the end, that is, only common issues will be handled on a classwide basis, so common issues necessarily predominate. Insofar as the Government argues those issues cannot be resolved on a classwide basis and so should weigh against predominance, that is nothing more than a restatement of the Government's commonality argument. Therefore, the district court could have erred only if it was incorrect in thinking it lawfully could bifurcate the suit and certify the class for the purpose of answering only the common questions concerning discrimination vel non.
Because we see no facial defect in the district court's reasoning and there is no binding precedent on point, we conclude the district court did not make a manifest error in certifying the class solely to resolve the questions concerning discrimination. The Supreme Court has instructed that in pattern-or-practice class actions, the plaintiffs can first “demonstrate that unlawful discrimination has been a regular procedure or policy followed by an employer,” after which members of the class will have the benefit of the presumption at his or her individualized hearing (now known as a Teamsters hearing) that “individual [employment] decisions were made in pursuit of the discriminatory policy.” Int'l Bhd. of Teamsters v.. United States, 431 U.S. 324, 359–60 (1977) (extending Franks v. Bowman Transp. Co., 424 U.S. 747 (1976)). In Wal–Mart, the Supreme Court held that a defendant must have the opportunity in a Teamsters hearing to rebut this presumption of discrimination on an individualized basis; “a class cannot be certified on the premise that [the defendant] will not be entitled to litigate its statutory defenses to individual claims.” 131 S.Ct. at 2561. Here, the district court anticipated individualized Teamsters hearings and so, in effect complied in advance with that aspect of the Wal–Mart decision. Unlike the lower courts in Wal–Mart, the district court here did not certify the class on the ground that it could resolve individualized issues on a classwide basis. The practical effect of the district court's decision is that it certified a “class action with respect to particular issues,” which it is expressly authorized to do by Rule 23(c)(4).
We recognize there is a controversy over the proper use of issue classes, especially when the result is to isolate a particular issue that would otherwise derogate from the predominance of common issues in a 23(b)(3) class action. See generally Robert H. Klonoff, The Decline of Class Actions, 90 Wash. U.L.Rev.. 729, 807–15 (2013) (surveying courts' and scholars' different approaches to Rule 23(c)(4) and observing “[t]his rule has created significant conflict and confusion among the courts”); Jenna G. Farleigh, Note, Splitting the Baby: Standardizing Issue Class Certification, 64 Vand. L.Rev.. 1585 (2011) (discussing the controversy and observing that the First, Second, Third, Fourth, Seventh, and Ninth Circuits have held an issue class can be used to avoid a predominance problem whereas the Fifth Circuit has held the cause of action as a whole must pass the predominance test). For example, in a prior case before the district court, the judge took a different approach to separating the common and individualized issues, despite being faced with facts similar to those here. Taylor v. D.C. Water & Sewer Auth., 241 F.R.D. 33 (D.D.C.2007). The court there certified the pattern-or-practice class under Rule 23(b)(2) for the purposes of determining discrimination and granting injunctive relief but with respect to damages indicated it would later decide between certifying the class under 23(b)(3) and conducting Teamsters hearings. Id. at 47.
Because the appropriate use of an issue class was not raised or briefed in the present case, we shall not review at this time the district court's determination of how best to manage the issues before it. Cf. Veneman, 309 F.3d at 796 (declining to exercise 23(f) discretion because the important, unresolved issues were entirely unbriefed). We may note with confidence, however, that in general, a district court has a good deal of discretion in the management of the litigation before it, see, e.g., Cloverleaf Standardbred Owners Ass'n, Inc. v. Nat'l Bank of Washington, 699 F .2d 1274, 1277 (D.C.Cir.1983) (“A district judge, closer to the arena, is often better situated than is an appellate panel to survey the practicalities involved in the litigation” (quotation marks omitted)), making it difficult to conclude the court has erred, let alone manifestly erred, by in effect certifying an issue class action.
4. Superiority of the class device
In the district court, the Government suggested no way of handling class members' claims that would be superior to the class device, other than holding a separate trial for each individual. Its argument before us is that the district court's analysis was insufficiently “rigorous” to determine the class device was superior to that alternative.
Because the district court did identify factors that could make the class device superior, and because it dealt with the Government's arguments to the contrary in its analyses of the commonality and predominance requirements, the court's relatively brief discussion of superiority was sufficient unto the task. The court explained that the interests of efficiency and uniformity supported resolving the question of discrimination in one stroke rather than requiring the same question to be answered separately for each individual. Moore, 926 F.Supp.2d at 34. On the other hand, it acknowledged the Government's argument that class members might have enough at stake in this litigation to make individual trials viable. Id. It was not manifestly erroneous for the district court to hold that argument was outweighed by the court's and the plaintiffs' legitimate interests in efficiency and uniformity. Furthermore, given the district court's identification of common questions susceptible to classwide proof and its stated intention, if the plaintiffs prevailed on those questions, to hold Teamsters hearings for each class member, we cannot credit the Government's argument that the district court failed adequately to analyze how the class device would operate to conserve judicial resources.
C. Special Circumstances
Perhaps sensing it is not on solid ground with any of the reasons for interlocutory review we identified in Lorazepam, the Government contends this case presents “special circumstances” similar to those that justified our granting review under Rule 23(f) in Rail Freight, 725 F.3d 244. There the district court had certified the class prior to the Supreme Court issuing an opinion “with significant bearing” upon the class certification question at issue. Id. at 254. That development “tip[ped] the scales in favor of review” because the case nearly qualified for interlocutory review for the first reason set out in Lorazepam: There was a “questionable” class certification decision that created a near “death-knell situation.” Id. In this case, the Government points out that two of the decisions upon which it relies, Comcast Corp. v. Behrend, ––– U.S. ––––, 133 S.Ct. 1426 (Mar. 27, 2013), and DL, 713 F.3d 120 (Apr. 12, 2013), were issued after the class certification order (Feb. 25, 2013). As we have explained in the foregoing sections, however, there is no similarly close call in this case with respect to any of the three reasons given for interlocutory review in Lorazepam. Accordingly, there is no “confluence of multiple rationales” as there was in Rail Freight, 725 F.3d at 250; therefore, even assuming they are relevant, the two subsequent decisions do not tip the scales in favor of immediate review.
Although we recognize there are unsettled questions of law relating to class actions at issue in this case, now is not the appropriate time to resolve them. None of the district court's rulings in support of its order certifying the plaintiff class is foreclosed by controlling precedent and the unsettled questions are not likely to evade end-of-the-case review. As we have observed before, “[t]he sheer number of class actions, the district court's authority to modify its class certification decision, and the ease with which litigants can characterize legal issues as novel, all militate in favor of narrowing the scope of Rule 23(f) review.” Lorazepam, 289 F.3d at 105–06 (citations omitted). Accordingly, the Government's petition is
FOOTNOTE. Part I describes the facts as they were relevant to the certification order. They are taken from the opinion of the district court and the joint appendix provided by the parties and they are undisputed in this proceeding.
FOOTNOTE. In their third certification motion, the plaintiffs submitted and the court rejected the following class definition:[A]ll current and former African–American Agents who were employed as Criminal Investigators (GS/GM–1811) and who had the required time-in-grade to seek promotion to competitive positions at the GS–14 level at any time during the years 1995 to 2004, and/or who had the required time-in-grade to seek promotion to competitive positions at the GS–15 level at any time during the years 1995 to 2005.In their fourth certification motion, they proposed the following class definition, the last clause of which excluded upper management and which the court certified:[A]ll current and former African–American Special Agents who bid for promotion to a GS–14 position from 1995 to 2004 and were not promoted to GS–14 on the first bid list on which they bid; and all current and former African–American Special Agents who bid for promotion to a GS–15 position from 1995 to 2005 and were not promoted to GS15 on the first bid list on which they bid; but excluding Special Agents who served as an Assistant Director, a Deputy Director, or the Director of the Secret Service during the class period.Moore, 926 F.Supp.2d at 15, 35.
GINSBURG, Senior Circuit Judge: