explaining that the 90-day deadline “implements the policy behind the already extant requirement of Fed. R. Civ. P. 23(c)(1) that class certification decisions be made as soon as practicable”
How later courts described this case
- explaining that the 90-day deadline “implements the policy behind the already extant requirement of Fed. R. Civ. P. 23(c)(1) that class certification decisions be made as soon as practicable”
- deferring to district court’s denial of class certification and noting that “Dellums consisted only of those persons arrested on the steps of the U.S. Capitol, a much more cohesive group than the approximately 7,000 putative class members here who were arrested throughout the city.”
- noting that Rule 23's requirement of adequate representation encom- passes "concerns about the competency of class counsel" (internal quotation marks and citation omitted)
- noting that “the stakes in the present case are much different [than those in a case in which in-junctive relief was granted to a class of plaintiffs with respect to the same arrests] given the claims for money damages”
Written by the judges who cited it.
The opinion
MIKVA, Circuit Judge,
concurring in
part and dissenting in part:
Whatever can be said of hard cases is equally pertinent to old cases. This dispute, vintage 1971, is before this court on questions that are yet at the threshold *1417 stage. Tempting as it might be to find a rug under which this case might be swept, I think it makes bad law. I dissent from that part of the majority’s opinion which denies the parties the right to proceed as a class. I believe the court is wrong in holding that the questions of law and fact common to the members of the putative class in this case do not sufficiently predominate over questions affecting only individual members as to warrant class certification. To explain my basis for dissenting, it is necessary to develop the facts of this litigation in somewhat more detail than found in the majority’s opinion.
The case involves another piece of the troublesome response taken by government officials to the “May Day” demonstrations against the Vietnam War in 1971. The “field arrest procedures” referred to by the majority were designed to deal with massive civil disorders in the District of Columbia. These procedures contemplated that after an arrest the arresting officer would complete a Field Arrest Form by filing in the officer’s name, unit and badge number, the identity of the arrestee, and the circumstances of the arrest. A photograph of the arrestee and the arresting officer together was also to be taken at the scene. Thereafter, the arrestee would be transferred to other police personnel for removal from the scene and booking, allowing the arresting officer to return to duty. The procedures also provided for expeditious post-arrest processing, in which the completed arrest forms and Polaroid photos would be accepted by the local courts as evidence of probable cause in lieu of direct testimony by the arresting officer.
With this expedited arrest and booking procedure in place, the Washington Police Department faced the antiwar demonstrations of late April and early May, 1971. In the initial stages of these demonstrations, prior to May 3,1971, the Police Department arrested several hundred demonstrators throughout the city. In doing so, they initially adhered to the field arrest procedures described above. At 6:23 a.m. on May 3,1971, hov/ever, defendant Jerry Wilson, then Chief of the Washington Metropolitan Police Force, issued an order suspending the field arrest procedures. This suspension remained in effect until May 4, 1971, at 5:40 a.m.
During this period, nearly 8,000 people were arrested. Having foregone the use of formal arrest procedures, however, the police lacked any records by which to establish probable cause. Accordingly, they initiated special booking procedures under which volunteer attorneys from the Department of Justice were instructed to:
record the name, address and physical description of each arrestee. In the blank marked ‘Original Charge’ they were instructed to enter the words ‘Disorderly Conduct’. At their initial briefing, they had been given a list containing the names, badge numbers and unit designations of seven police officers. Under the caption ‘Name of Arresting Officers,’ they were told to insert ‘one name taken seriatim from [this] list of seven’; and they were specifically told to leave blank that portion of the form in which the circumstances of arrest were to be recorded.
Sullivan v. Murphy, 478 F.2d 938, 951 (D.C.Cir.1973). Following their arrest, the arrestees were crowded, some for as long as seventy hours, into detention centers around the city.
In Sullivan v. Murphy, a class action brought on behalf of all those arrested without field arrest forms, we declared all of these nearly 8,000 arrests presumptively unlawful under the fourth amendment; without photographs and field arrest forms, the police simply could not as a general matter establish probable cause. We noted, however, that if probable cause could be proven in any particular case on remand, that particular arrest would be lawful. The government stated on remand that “an affirmative showing of probable cause [could] not be made in any of the pertinent cases” and therefore a final order was issued by the district court declaring all the arrests invalid. 380 F.Supp. 867, 868 (D.D.C.1974). Every member of the *1418 putative class in this case was a member of the class in Sullivan and hence the arrest of every individual in the class is presumptively unlawful. Against this background, I now turn to the legal issues presented by this case.
I.
I agree with the majority that, while procedural defects in intervention motions should generally be excused, in the context of this case the failure to serve the intervention motion on the federal defendants for two years justifies dismissing the case against these defendants.
Because the majority concludes that the case against the federal defendants should be dismissed on this ground, I assume that the majority’s discussion about the suitability of class treatment applies to only the claims against the District of Columbia and its local officials. Were this discussion intended to cover the federal defendants as well, it would be particularly troubling, for it seems indisputable that the dominant question with respect to the federal defendants would be the extent to which they were involved in and can be held responsible for the actions of local officials on May 3, 1971. That question, in my view, would make class treatment appropriate for the suit against the federal defendants. I do not read the majority opinion as expressing any view on the merits of a class suit against the federal defendants, however, and therefore address myself to only the class allegations against the District and its officials.
II.
The class action mechanism set forth in the Federal Rules of Civil Procedure is an intensely practical device, designed, as the majority notes, to foster efficient and economical multi-party litigation. Indeed, the very purpose of the major revisions made in 1966 to the class action rules was to “substitute functional tests for the concep-tualisms that characterized practice under the former rule.” 7 C. Wright & A. Miller, Federal Practice and Procedure § 1753, at 538 (1972). To determine whether class treatment is appropriate, therefore, it is necessary to look to the realities of the litigation and to assess whether, as a practical matter rather than as an abstract inquiry, class treatment is the most efficient means of conducting the action. “The proper standard [for determining whether a class action should be certified] under Rule 23(b)(3) is a pragmatic one.” 7A Id. § 1778, at 53.
The majority seems to recognize these principles in explaining its conclusion that class certification decisions are to be reviewed under an abuse of discretion standard. Yet the majority completely turns its back on these principles a mere two pages later when it states that a court is foreclosed from considering whether, as a practical matter, individual defenses to the action are likely to be real factors at trial. See Maj.Op. at n. 8. Instead, as the majority would seem to have it, as long as such defenses exist in theory and as an abstract possibility, they are enough to allow a trial judge to refuse to certify an action that, pragmatically viewed, would best be run as a class suit. There is no legal or logical basis for such an impractical result.
To determine whether common questions predominate a trial judge must inevitably make some effort to assess the realities of a lawsuit and to assay the issues that are likely actually to be in contention. For example, if several defenses are raised to a class complaint, some of which apply to the entire class and some of which are more individualized, the trial judge will of necessity have to make some judgment on which defenses the lawsuit is likely to turn. That judgment obviously will require a realistic appraisal and estimation of the number of individuals to whom the particularized defenses are likely to apply. In a products liability class action suit against a drug manufacturer, for instance, the manufacturer may assert that the drug does not cause the disease at issue and may also argue that many patients knew of the relevant dangers at the time they took the drug. If the individual defense would be a *1419 complete defense, the trial judge surely is not barred from estimating, before he decides the class certification question, how many individuals could be vulnerable to that defense. A procedural mechanism designed to be functional and pragmatic cannot be held hostage by a defendant who merely raises the theoretical possibility that individualized defenses will apply to all plaintiffs.
There is also no legal basis for such a result. Eisen v. Carlisle & Jacquelin, 417 U.S. 156 , 94 S.Ct. 2140 , 40 L.Ed.2d 732 (1974), relied on by the majority, does not stand for the proposition that a court cannot realistically appraise a class complaint to determine whether common questions predominate over individual ones. Eisen holds merely that a court cannot conduct a “preliminary hearing” into the merits and conclude that the plaintiff is likely to prevail on the litigation as a basis for shifting to the defendant the cost of providing notice to absent class members. Eisen thus prevents a district court from requiring defendants to a class action to pay the costs of the litigation before any definitive resolution of the defendants’ liability has been made. But surely Eisen does not prevent a court from determining how many individuals will be affected by certain defenses in an effort to gauge, not whether the class will prevail, but whether common questions transcend merely individual ones. See Strong v. Arkansas Blue Cross & Blue Shield, Inc., 87 F.R.D. 496 (E.D.Ark. 1980) (holding that, while it is not appropriate to consider the merits in determining certification, the admissibility of all evidence which has relevance to the merits is not barred if it is relevant to the purpose of refuting or supporting the existence of a class); see also Dura-Bilt Corp. v. Chase Manhattan Corp., 89 F.R.D. 87 (S.D.N.Y. 1981) (if there is no chance at all that a reasonable investor would have been able to discover an alleged fraud, plaintiff need not establish due diligence and due diligence is thus removed as an individual question of fact which could possibly predominate over common questions).
In fact, to the extent there is precedent on this question, it supports the view that inquiry into the actual facts of a case is proper to determine whether common questions predominate. In General Telephone Company of the Southwest v. Falcon, 457 U.S. 147 , 102 S.Ct. 2364 , 72 L.Ed.2d 740 (1982), the Court stated that a sufficient indication that an employer used similar means to discriminate in both hiring and promotion decisions would justify a single class action on behalf of both potential hirees and potential promotees. Id. at 158 , 102 S.Ct. at 2371 . As a corollary, when class plaintiffs claim that some of the same decisional bases were used for hiring and promotion decisions, a court will have to evaluate the extent of the overlap to determine whether the common factors are strong enough to justify a single class action on behalf of both hirees and promo-tees. Similarly, to determine whether plaintiffs have standing to seek injunctive relief for future injuries, courts must assess in real-world terms the likelihood that the plaintiffs will be subjected in the future to the complained-of conduct. See City of Los Angeles v. Lyons, 461 U.S. 95 , 103 S.Ct. 1660, 1667 , 75 L.Ed.2d 675 (1983). There is simply nothing forbidden, or unusual, in the need for a court to look into the actual facts of a case in an effort to resolve threshold questions such as whether common questions predominate over individual ones.
Moreover, this circuit has already confronted virtually the identical question regarding commonality presented in this case and resolved it in a way contrary to the majority’s disposition today. In Dellums v. Powell, 566 F.2d 167 (D.C.Cir.1977), a class of some 2,000 people who had been arrested during an anti-war protest at the Capitol sought damages from, inter alia, Chief James Powell of the United States Capitol Police. In opposing class certification, Chief Powell pointed to several record facts which showed that some of the demonstrators had been engaging in unlawful conduct at the time of the arrest — four to five had been climbing lampposts, one or two had been writing on the walls of the Capitol, and a group of around 100 had *1420 disobeyed orders to disperse. After a realistic appraisal of this evidence, we rejected the argument that the possible existence of these individualized defenses should preclude class certification. Chief Powell’s objections were “largely theoretical,” we said, because it was clear that at the time of the lawsuit he “had no means of identifying” any of the individuals who had been allegedly arrested lawfully. Id. at 191 . Del-lums thus clearly rejected the majority’s position that it is impermissible to pierce theoretical individualized defenses to evaluate the realistic likelihood that at trial they will in practice outweigh the issues common to the class.
Applying these principles to this case, I conclude that the class should have been certified. As to the false arrest claim, it simply blinks reality to think that probable cause could be demonstrated for a significant number of the 8,000 arrests at issue. Appellees deliberately failed at the time to maintain proper arrest records and, as in Dellums , have thus virtually foreclosed any opportunity to show that the arresting officer had probable cause to make an arrest in any particular case. Indeed, on remand in Sullivan v. Murphy, the companion suit to this case in which expungement of arrest records was sought, the District stated that “an affirmative showing of probable cause [could] not be made in any of the pertinent cases” and therefore a final order was issued by the district court declaring all the arrests invalid. 380 F.Supp. 867, 868 (D.D.C.1974). Similarly, in Dellums v. Powell, a damages action arising out of other mass arrests during the May Day period, defendants failed to take discovery from a single absent class member. 566 F.2d at 191 . In view of the facts of this case and these prior encounters with the District’s mass arrest procedures, it seems quite clear to me that, pragmatically viewed, the dominant if not sole issue at trial on the false arrest claim would be the amount of damages that ought to be awarded to the class or to various sub-classes.
The majority recognizes that bifurcation of trials into liability and damages phases is a well-accepted means for dealing with actions similar to the present one, but then asserts that a very different situation is presented when, as in the false arrest claims, the issue of liability vel non turns upon highly individualized facts such as whether probable cause for the arrest existed. I view this purported distinction as highly formalistic, for it makes the viability of a class action depend on whether the very same issue is treated as a defense to the action or as a damage mitigation measure. The emptiness of this distinction is demonstrated by the fact that the complaint also includes conspiracy counts, to which it is no defense to assert that the defendants also had proper motives for their conduct, see Halberstam v. Welch, 705 F.2d 472, 477 (D.C.Cir.1983); such an assertion is relevant to a conspiracy charge only at the stage of determining whether any actual damages have been suffered. Thus, under the liability/damages distinction offered by the majority, the false arrest claim could not proceed as a class action while the conspiracy counts could, even though the very same potential individualized defenses would be raised in both. That result makes little sense, especially in the context of a rule designed to be practical.
Without launching into an extended debate with the majority on the substantive law of civil conspiracy, I believe the conspiracy allegations can be viewed separately from the underlying tort claims in this action. First, it is not the law of this circuit that a conspiracy claim which arises under federal law is sustainable only after an underlying tort claim has been established; the passage to this effect in Hal-bertsam v. Welch, 705 F.2d at 479 , cited by the majority, is no more than a statement of the elements necessary to prove a conspiracy under District of Columbia law. Indeed, the court in Halberstam made clear that it was not expressing an opinion on whether “the element of combination alone may make certain acts unlawful— even though they would not be so if performed independently by individuals.” Id. *1421 at 477 n. 7; see also W. Prosser, Law of Torts § 46, at 293 (4th ed. 1971) (“[I]t now seems generally agreed, although there has been authority to the contrary, that there are certain types of conduct ... in which the element of combination adds such a power of coercion ... that it makes unlawful acts which one man alone might legitimately do.”). In a case such as this, in which federal and local officials are charged with plotting massive civil rights violations, common sense, sound policy and governing law all suggest that the agreement alone suffices to establish liability. See Hobson v. Wilson, 737 F.2d 1 at 52 (D.C.Cir.1984) (“To make the conspiracy actionable, there must also be an overt act in furtherance of the object of the conspiracy that injures plaintiff in his person or property, or, in a section 1985(3) action, which deprives him of having or exercising any right or privilege of a citizen of the United States.”) (emphasis added). Hobson does not suggest that the overt act must itself be tortious. Moreover, to the extent an underlying tort is required, there is nothing to suggest that each member of the plaintiff class need prove that he was the specific victim of tortious action — it may suffice that the conspirators undertook some tortious conduct with respect to the class as an entity. That requirement is clearly met in this case. It is unnecessary to resolve these questions about the substantive nature of federal civil conspiracy law, however, in order to convince me that these questions, and the ability of plaintiffs to establish the existence, scope, and nature of any conspiracy among the defendants, are much more central to this litigation than the question whether the defendants can establish probable cause for the otherwise unlawful arrest of some of the plaintiffs.
I would therefore hold that the false arrest, constitutional, and conspiracy claims, all of which have the same basic structure, are appropriate for class treatment.
As to the abuse of process claim, the majority’s position is in my mind even less convincing. The majority states that “the problem in the present case is that it is unclear how many putative class members were allegedly victimized by abuses of the judicial process.” As an empirical matter, this statement is flawed; the District admits that “almost all plaintiffs were subject to some sort of criminal or juvenile proceedings following their detention, the form of proceeding varypng] from forfeiture of collateral to no papering to prosecution.” District Brief at 25. More importantly, the majority’s position drastically alters the burden of proof for class certification questions.
While the majority correctly notes that some circuits have stated in passing that plaintiffs bear the burden of establishing that the requirements for class treatment have been met, it is equally correct that many courts and commentators have concluded that Rule 23 is directed to the trial court and that Rule 23 determinations are to be made in a nonadversarial context. See generally 1 H. Newberg, Class Actions § 2076 (“Burden of proof concepts ... are not appropriate in dealing with a determination respecting whether [an action brought under Rule 23] should be permitted to be maintained.”); City of Philadelphia v. Emhart Corp., 50 F.R.D. 232, 235 (E.D.Pa.1970) (stating that plaintiff has burden of showing he meets Rule 23 but then sua sponte dividing the claimed class into subclasses and allowing the class action to proceed). Moreover, even if as a formal matter the plaintiff bears this burden, the very cases cited by the majority make clear that the trial judge cannot, as in this case, passively rest on the pleadings and dismiss an action for failure of the complaint to suggest viable subclasses that would make the action manageable; as the court stated in Doctor v. Seaboard Coast Line R.R., 540 F.2d 699 , 707 (4th Cir.1976):
In determining whether the plaintiff has met his or her burden, the trial, court may look to the pleadings “but [t]he determination usually should be predicated on more information than the complaint itself affords.” The court may, and often does, permit discovery relating to the *1422 issues involved in maintainability, and a preliminary evidentiary hearing may be appropriate or essential as a part of the vital management role which the trial judge must exercise in class actions to assure that they are both meaningful and manageable, (quoting Huff v. N.D. Cass Co., 485 F.2d 710, 713 (5th Cir.1973) (emphasis added).
See also Dillon v. Bay City Construction Co., 512 F.2d 801, 804 (5th Cir.1975) (“Additionally, the plaintiffs were entitled to discovery which would bear on the always troublesome questions of whether this was or ought to be considered a class action ... and the terms and conditions, if any, on which it could proceed.”).
With respect to the abuse of process claim, the obvious, and traditional, means of dealing with the majority’s concern would be to sub-divide the class into those who were subject to prosecution, those who were subject to forfeiture of collateral, and those subject to no papering, and to award damages accordingly; to collect damages, a plaintiff would have to demonstrate, as always, the class to which he belonged. The solution is not to require the plaintiffs to make this demonstration in their complaint on pain of losing the opportunity to litigate as a class. Contrary to the majority’s intimation, see Maj.Op. at 1415 n. 10, I am not engaging “in the business of creating classes and subclasses never suggested by the parties;” my contention is simply that the district court’s failure to explore the possibility of subclass treatment through an evidentiary hearing and other avenues, such as creating a discovery schedule for both sides on the probable cause issue, constitutes reversible error. I would therefore remand so that the district court, after an evidentiary hearing, would enter findings of fact as to why the class action rule, which is to be given a liberal construction rather than be contracted into nothingness, should not favor class or subclass treatment of this action.
It is of course true, as the majority states, that district judges have significant discretion in determining whether a class action is the most appropriate means of conducting a particular lawsuit. But the extent of that discretion should vary with the nature of the determination which the district judge has made; the less a determination depends on the district judge’s familiarity with the facts and the particular case, the less the basis for appellate court deference. For example, in National Association for Mental Health, Inc. v. Califa-no, 717 F.2d 1451, 1459 (D.C.Cir.1983), cited by the majority, it was entirely appropriate for this court to accord substantial deference to a district judge’s determination that representation by the class representative had not been adequate; the district judge was uniquely situated to make this determination. In this case the situation is very different. Despite the vintage of this litigation, no proceedings had yet taken place before the district judge that made him any more familiar with the potential complexities of class treatment of this case than an appellate court would be; the district judge decided the certification issue essentially on the bare papers and without conducting any hearing. Under these circumstances, the rationales for deferring to the discretion of the district court are significantly attenuated.
More importantly, the fact that a district judge has some discretion in class certification decisions does not mean that those decisions can be made in a way that undercuts the purposes and policies of Rule 23. As other courts of appeals have recognized, “[although the district court has broad discretion in determining whether a particular action [is suitable for class treatment], the application of these requirements should not be so strictly applied that the policies underlying class action would be undermined.” Weathers v. Peters Realty Corp., 499 F.2d 1197 (6th Cir.1974). The failure to certify this class, which the majority approves only by turning its back on the actual circumstances spawning this lawsuit and on the issues that would in practice predominate at trial, sufficiently contravenes the purposes of the “functional” and “pragmatic” approach embodied in Rule 23 that I would overturn the district *1423 judge’s refusal to certify the class. I therefore dissent from that part of the majority’s opinion that holds otherwise.
III.
I am also somewhat troubled by the majority’s approval of the decision to deny class certification on the grounds that the motion to certify was untimely. I would require, as have most other courts and as has been advocated by eminent commentators, that a class certification motion not be denied on timeliness grounds unless the defendants can make some showing that they have actually been prejudiced by the delay. See Senter v. General Motors Corp., 532 F.2d 511, 521 (6th Cir.), cert, denied, 429 U.S. 870 , 97 S.Ct. 182 , 50 L.Ed.2d 150 (1976); see also Larionoff v. United States, 533 F.2d 1167, 1183 (D.C. Cir.1976) (“absent some showing of actual prejudice either to the Government or to the absent class members, we cannot conclude that the simultaneous entry of the judgment and the class action certification was reversible error.”); see generally 7A C. Wright & A. Miller, Federal Practice and Procedure § 1785 (1972 & 1982 Supp.). Whether such a showing could be made in this case is unclear; no discovery was taken of the original thirty-nine plaintiffs before the class certification motion was filed and hence it would be somewhat anamalous for the defendants to assert that the delay in moving for certification undermined their discovery efforts. I see no reason to elaborate my views on this question at length, however, for in light of the adoption of Local Rule l-13(b) in 1973, few if any cases will be affected by the majority’s timeliness discussion. I therefore merely note my reservation on this question.