From Casetext: Smarter Legal Research


Judicial Panel on Multidistrict Litigation
Jan 1, 2005
228 F.R.D. 673 (J.P.M.L. 2005)


January 1, 2005


On February 18, 2005, President Bush signed into law the Class Action Fairness Act of 2005 ("Act"). The new law became effective at that moment, applying to all actions filed thereafter. The law will have immediate and salutary impact on abuse of the class action device, although limitations in the Act's scope may reduce its overall impact and certain shortcomings in its drafting have left a number of significant issues to be resolved by the courts.

Courts have rejected efforts to allow removal of actions pending on the effective date. See Knudsen v. Liberty Mut. Ins. Co., 411 F.3d 805, 2005 WL 1389059 (7th Cir. June 7, 2005); Pritchett v. Office Depot, Inc., 404 F.3d 1232 (10th Cir. 2005). Knudsen held that amendment of the class definition would not. It may be possible, however, that amendment to add new claims might give rise to federal jurisdiction under the Act. See Knudsen, 2005 WL 1389059, at *2.

I. Overview of the Class Action Fairness Act of 2005

The Act is made up of a number of discrete components, most of which will function essentially in isolation from one another. Apart from the expansion of federal court jurisdiction, which will result in a greater proportion of class actions being adjudicated in federal rather than state court, the Act's provisions apply only to class actions filed in, or removed to, federal court.

The major sections and a brief description of the change they will bring are:

A. Section 3, a self-described "Consumer Class Action Bill of Rights," establishes four specific reforms applicable in all class actions, each directed to specific shortcomings of current practice. This section regulates coupon settlements, so-called "zero-value" settlements, and discrimination among class members based on where they reside, and it creates a new notice mechanism requiring notice to state and federal officials before approval of a settlement.
B. Section 4 of the Act creates jurisdiction in federal court over some interstate class actions involving at least $5,000,000 in claims. In many ways, this is the most far-reaching change in the new law.
C. Section 5 creates a mechanism for removing class actions filed in state court to federal court, including extending the time during which removal can be sought.
D. Section 6 requires that, by February 18, 2006, the federal courts' administrative body (the Judicial Conference) report to Congress on class action settlements and recommend "best practices" for handling class actions.

See Class Action Fairness Act of 2005 § 3, codified at 28 U.S.C. § 1712.

See Class Action Fairness Act of 2005 § 3, codified at 28 U.S.C. § 1713.

See Class Action Fairness Act of 2005 § 3, codified at 28 U.S.C. § 1714.

See Class Action Fairness Act of 2005 § 3, codified at 28 U.S.C. § 1715.

See Class Action Fairness Act of 2005 § 4, codified at 28 U.S.C. § 1332(d).

See Class Action Fairness Act of 2005 § 5, codified at 28 U.S.C. § 1453.

See Class Action Fairness Act of 2005 § 6, not codified in U.S.C.

Additional provisions of the law are a recital that it does not in any way circumscribe the rulemaking authority of the Supreme Court and Judicial Conference and that it would take effect upon signing, February 18, 2005. By its terms, it applies only to civil actions commenced after the effective date.

II. Overview of Class Action Issues

A. History of the Act

The Act, although passed with remarkable speed in 2005, had numerous precursors. The Senate Judiciary Committee had conducted hearings on similar legislation since at least the late 1990s. The House of Representatives had previously passed the Interstate Class Action Jurisdiction Act of 1999 and earlier Class Action Fairness Acts of 2001 and 2003. Indeed, the speed with which the Act passed Congress in 2005 apparently left no time to address a provision that had become entirely anachronistic: Section 7 of the Act directs that certain amendments to Federal Rule of Civil Procedure 23 — which had been approved by the Supreme Court of the United States on March 27, 2003, and which took effect December 1, 2003 — nonetheless, "shall take effect on the date of enactment of this Act or on December 1, 2003 . . ., whichever occurs first." This provision, which presumably originated in the Class Action Fairness Act of 2003, was rendered meaningless by the passage of time and plainly should have been removed before the legislation reached the floor of the Senate or House.

Perhaps because the Act had existed in some form for several years, as ultimately enacted, it was pushed through Congress without amendment. Among the amendments that were unsuccessfully proposed were a proposal to exclude mass actions from the Act's scope, as well as provisions limiting the secrecy of settlement agreements, addressing choice-of-law issues in multistate actions, and carving out from the Act's scope actions brought by state Attorneys General or actions based on state civil rights or wage discrimination laws.

B. "Reform" of Class Actions

Class actions have been the subject of widespread criticism from many quarters. Many of these concerns were addressed in the 2003 amendments to the federal rules of civil procedure. Statutes have also addressed some of the criticisms, though often only in the context of certain types of class actions. For example, although concerns about the appointment of lead plaintiffs and hiring of lead counsel were addressed in the Private Securities Litigation Reform Act of 1995, this reform applies only in securities class actions. The problem remained in other classes.

Similarly, the Multiparty, Multiforum Trial Jurisdiction Act of 2002 ("MMTJA") expanded federal court jurisdiction to cover certain largescale accidents. Specifically, the requirement of complete diversity between plaintiffs and defendants was eliminated for cases involving "a single accident, where at least 75 natural persons have died in the accident at a discrete location. . . ." Commentators properly viewed this as a possible model for dealing with class action litigation, and in many ways the Act was cast in the mold of the MMTJA. Federalization of class actions as a means of dealing with overlapping class actions has also been discussed for years.

28 U.S.C. § 1369(a); see generally Passa v. Derdarian, 308 F. Supp. 2d 43 (D.R.I. 2004) (applying statute to cases arising from a nightclub fire). The statute is analyzed in Angela J. Rafoth, Congress and the Multiparty, Multiforum Trial Jurisdiction Act of 2002: Meaningful Reform or a Comedy of Errors?, 54 DUKE L.J. 255 (2004).

See, e.g., Laura Offenbacher, Note, The Multiparty, Multiforum Trial Jurisdiction Act: Opening the Door to Class Action Reform, 23 REV. LITIG. 177, 192-93 (2004).

See, e.g., Alan B. Morrison, Removing Class Actions to Federal Court: A Better Way to Handle the Problem of Overlapping Class Actions, 20 STAN. L. REV. 101 (2005) [forthcoming] (discussing of merits of 2004 predecessor of the 2005 Reform Act).

Finally, in a change that applied to all class actions in federal court, the Supreme Court amended the federal rules of civil procedure and of appellate procedure in 1998 to expand the availability of interlocutory appellate review of class certification decisions. While that change in the rules has led to a significant increase in the number of appellate decisions that address class certification issues, the circuit courts have varied widely in their willingness to grant petitions for interlocutory review, with the result that a number of circuits continue to have few, if any, precedential decisions that elaborate on the requirements of Rule 23. If the Act leads to a significant increase in the number of class actions being heard in federal court, all of the circuits will presumably have little choice but to begin to develop case law interpreting Rule 23, and the Supreme Court will then likely see a greater number of circuit splits that should be resolved, all of which may have the salutary effect of creating something closer to a nationwide set of standards for certification of class actions.

III. The Major Impact of the Act: Federalization of Class Actions

Although the Act has not federalized all class actions, it has eliminated the most significant barrier faced by defendants who sought to remove such actions to federal court. In the context of class actions or mass actions, the long-standing requirement of complete diversity between plaintiffs and defendants is a thing of the past. Unfortunately, in an attempt to balance the competing interests between actions that largely involve a single state (and should therefore be litigated in that state's courts) and actions that involve multiple states, the Act introduces significant gray areas that may only be made black and white after the various courts of appeals have imposed their own interpretive glosses on the Act's somewhat opaque commands. If the courts follow their traditional approach of declining jurisdiction whenever possible (an approach that some say is the only permissible approach for courts of limited jurisdiction), then it is likely that the Act will not have quite as expansive an effect as Congress may have envisioned.

A. Federal Court Jurisdiction

As a threshold matter, the Act permits class actions to be brought in or removed to federal court if two requirements are met: (1) at least $5 million must be in controversy, exclusive of interest and costs; and (2) at least one plaintiff and one defendant must be citizens of different states or of a state and a foreign country.

1. Amount in controversy

In one respect, the issues surrounding the amount-in-controversy requirement should pose no difficulties in application, for it will be presumably treated consistently with the courts' existing treatment of the $75,000 amount-in-controversy requirement found in § 1332(a). The Act also makes clear that for purposes of satisfying the $5 million threshold, all of the individual class members' claims are to be aggregated.

See id. § 1332(d)(6). The aggregation issue has been frequently litigated prior to the passage of the Act, with a resulting circuit split on whether 28 U.S.C. § 1367 (the supplemental jurisdiction statute) overruled the holding in Zahn v. International Paper Co., 414 U.S. 291, 94 S.Ct. 505, 38 L.Ed.2d 511 (1973), that each class member must independently satisfy the amount-in-controversy requirement. The Supreme Court recently held that § 1367 effectively overruled Zahn. See Exxon Mobil Corp. v. Allapattah Svcs., Inc., 125 S.Ct. 2611, 2005 WL 1469477 (2005).

The difficulty that courts will likely face in determining whether the jurisdictional amount is satisfied relates to the size of the plaintiff class. In the abstract, aggregation is simple enough to apply, but in practice, it only works if there is some means by which to determine both the size of the claims and the number of claimants. Although there certainly are cases in which the parties may be able to determine with some ease (very likely from records in the defendant's possession) how many members are likely in a putative plaintiff class, there are many other cases in which the parties have no way of making any sound estimate of the number of potential class members. The Act provides no guidance on how these determinations are to be made, and this uncertainty will affect not only how courts rule on motions to remand or to dismiss for lack of subject matter jurisdiction but also the timing of removal decisions by defendants, which are discussed in more detail below.

2. Minimal diversity

Historically, the federal courts' diversity jurisdiction has only been available to litigants when the citizenship of all plaintiffs is diverse from that of all defendants. As noted above, the MMTJA expressly expanded federal court jurisdiction to include cases in which only minimal diversity exists — that is, when at least one plaintiff's citizenship is diverse from that of at least one defendant. While that aspect of the statute has not been tested, the interpleader statute ( 28 U.S.C. § 1335) also grants federal courts jurisdiction in cases of minimal diversity, and that statute has been upheld by the Supreme Court.

See 28 U.S.C. § 1332(a); see also Strawbridge v. Curtiss, 7 U.S. (3 Cranch) 267, 2 L.Ed. 435 (1806).

For purposes of diversity of citizenship under the Act, parties' citizenship either of a state or of a foreign country is considered. So a party from a foreign country is deemed diverse from an adverse party from any state, just as are parties who are citizens of different states. See 28 U.S.C. § 1332(d)(2). It appears that, consistent with how § 1332(a) has been applied, parties from foreign states are not considered diverse for this purpose (in other words, practically speaking, all foreign countries are treated as if they are the same, single foreign country). See generally Allendale Mutual Ins. Co. v. Bull Data Sys., Inc., 10 F.3d 425, 428 (7th Cir. 1993) (acknowledging that many courts have "state[d] that the presence of foreign parties on both sides of a litigation . . . destroys complete diversity," but noting qualification to rule); see also Grupo Dataflux v. Atlas Global Group, L.P., 541 U.S. 567, 120 S. Ct. 1920, 1923, 158 L.Ed.2d 866 (2004) ("aliens were on both sides of the case, and the requisite diversity was therefore absent").

See State Farm Fire Cas. Co. v. Tashire, 386 U.S. 523, 87 S.Ct. 1199, 18 L.Ed.2d 270 (1967).

For purposes of determining diversity of citizenship under the Act, the court is to consider whether the citizenship of "any member of a class of plaintiffs" is "different from" that of "any defendant." This issue presents problems similar to those presented by the amount-in-controversy, for the identity (and therefore the citizenship) of putative class members is frequently not apparent at the time a lawsuit is filed. Plaintiffs' counsel may often avoid this problem by ensuring that the citizenship of at least one named class representative is diverse from that of at least one defendant, but, so long as the citizenship of a putative (but unnamed) class member is diverse from that of a defendant, it appears that the Act intends the minimal diversity requirement to be satisfied.

The Act also changes how unincorporated associations' citizenship is determined, but only for purposes of § 1332(d) and not, apparently, for purposes of diversity jurisdiction under § 1332(a). Many courts have treated unincorporated associations — whether in the form of partnerships or limited liability corporations or other non-corporate forms — as possessing the citizenship of all of the entity's constituent members, but that is no longer the rule in cases governed by § 1332(d).

See 28 U.S.C. § 1332(d)(10) ("For purposes of this subsection and section 1453, an unincorporated association shall be deemed to be a citizen of the State where it has its principal place of business and the State under whose laws it is organized.").

See, e.g., Grupo Dataflux v. Atlas Global Group, L.P., 541 U.S. 567, 124 S.Ct. 1920, 1932 n. 1, 158 L.Ed.2d 866 (2004) (Ginsburg, J., dissenting) ("Although the Court has never ruled on the issue, Courts of Appeals have held the citizenship of each member of an LLC counts for diversity purposes."); Carden v. Arkoma Assocs., 494 U.S. 185, 196, 110 S.Ct. 1015, 108 L.Ed.2d 157 (1990) (in determining "the citizenship of an artificial entity," including "an unincorporated association," court must consider the citizenship of all its members).

3. Effect of certification ruling

Unlike the supplemental jurisdiction statute, which expressly authorizes a district court to dismiss or remand an action once it has "dismissed all claims over which it has original jurisdiction," the Act says nothing about what a district court is to do if it denies class certification and the remaining individual claims would not otherwise be subject to federal jurisdiction. In diversity cases, of which actions subject to the Act would appear to be a species, the traditional rule is that if the necessary diversity and amount-incontroversy requirements are met at the time of filing (or of removal), then subsequent changes in those facts do not affect the court's jurisdiction. Following that rule, after denial of a class certification motion, the district court would have no discretion to decline to adjudicate the remaining named plaintiffs' claims, even if there were no independent jurisdictional basis for them to remain in federal court.

See, e.g., Rosado v. Wyman, 397 U.S. 397, 405 n. 6, 90 S.Ct. 1207, 25 L.Ed.2d 442 (1970); St. Paul Mercury Indem. Co. v. Red Cab Co., 303 U.S. 283, 289-90, 58 S.Ct. 586, 82 L.Ed. 845 (1938); see also Grupo Dataflux, 541 U.S. at 574, 124 S. Ct. at 1926 (change in party's citizenship post-filing does not cure jurisdictional defect at time of filing).

If one were to view the individual claims as merely pendent to the putative class claims, there is a similar tradition — now embodied in the supplemental jurisdiction statute but with its origins in United Mine Workers of America v. Gibbs — that when "the federal claims are dismissed before trial, even though not insubstantial in a jurisdictional sense, the state claims should be dismissed as well."

Id. at 726, 86 S.Ct. at 1139.

As courts of limited jurisdiction, the federal courts certainly may be tempted to dismiss the remaining named plaintiffs' claims after class certification has been denied, but that could lead to practical problems if the named plaintiffs are free to seek again to represent a plaintiff class when their claims are remanded to, or filed for the first time in, state court. Once such an intent is made clear, any defendant would seemingly be permitted again to remove the case to federal court, so long as the newly proposed class action satisfies the requirements of section 1332(d), after which the district court would likely again deny class certification. If the federal courts conclude that they are not free under the Act to dismiss the remaining named plaintiffs' claims, they would avoid the prospect of such yo-yo-like proceedings.

It is difficult to predict how this issue will play out in the courts, but one can be reasonably confident that the courts will, in time, develop a rule (or competing rules) to address this gap in the Act's provisions.

B. Abstention

In cases in which the amount-in-controversy and minimal diversity requirements are satisfied, federal district courts in some circumstances must and in other circumstances may nonetheless decline to hear class actions depending on the composition of the putative plaintiff class, the citizenship of the "primary" defendant(s), and the relationship of the claims made to a particular state.

See 28 U.S.C. §§ 1332(d)(3) (discretionary abstention), 1332(d)(4) (mandatory abstention).

1. Mandatory abstention

Like the MMTJA, the Act seeks to exclude from its scope cases that are predominantly connected with a single state and its citizens. The MMTJA provides that "a district court shall abstain from hearing any civil action . . . in which (1) the substantial majority of all plaintiffs are citizens of a single state of which the primary defendants are also citizens; and (2) the claims asserted will be governed primarily by the laws of that State." The Act's mandatory abstention provision is a bit more complicated, requiring that a district court decline to exercise jurisdiction whenever (1) two-thirds or more of the plaintiffs in the proposed class are citizens of the state in which the action was filed, and (2) either (a) the "primary defendants" are also citizens of the state in which the action was filed, or (b) principal injuries from the defendants' alleged conduct, or related conduct, occurred in the state in which the action was filed; no similar class action has been filed in the preceding three years; and at least one of the defendants from whom "significant relief" is sought and whose alleged conduct forms a "significant basis" for the plaintiffs' claims is a citizen of the state in which the action was filed.

This provision — which, like the MMTJA provision, appears to operate as a mandatory abstention provision rather than as a limitation on the court's jurisdiction — attempts to provide greater precision than does the MMTJA provision. Where the MMTJA refers to "a substantial majority" of the plaintiffs, the Act's provision is triggered by "two-thirds or more" of the plaintiffs. Both the MMTJA and the Act rely on an otherwise unfamiliar, undefined concept of the "primary" defendant, but the Act appears to contain an implicit definition of what constitutes a "primary" defendant: one "from whom significant relief is sought," "whose alleged conduct forms a significant basis for the claims asserted by the proposed plaintiff class." While the Act's use of the word "significant" is itself undefined, its focus on the relief sought and the relationship between the alleged conduct and the claims asserted provides greater focus than does the MMTJA's completely undefined use of the "primary defendant" label.

For a discussion of the potential ramifications of this distinction, see Passa v. Derdarian, 308 F.Supp.2d 43, 55-57 (D.R.I. 2004).

These standards will likely pose many of the same problems for courts and litigants as do the Act's basic jurisdictional requirements. How, for example, are the parties to determine the citizenship of all potential class members and then arithmetically ascertain whether two-thirds of them reside in a particular state? What if the issue is a close one, but after an opt-out class has been created, it becomes clear that two-thirds of the plaintiff class members are from the primary defendants' home state, in which the action was originally filed? Since the statute contains no limit on when the mandatory abstention provisions may be applied, it would appear that under such a hypothetical set of facts, the court would be obligated to remand the case to state court. If, after such a remand, the plaintiffs sought to expand the scope of the plaintiff class with the result that fewer than two-thirds of the class members were citizens of the forum state, then the Act would appear to authorize removal to federal court. Such yo-yo-like proceedings can be avoided, however, if the courts develop a rule that abstention decisions under the Act should only be made at the outset of a case and not be revisited thereafter.

2. Discretionary abstention

In addition to the mandatory abstention provision described above, the Act (unlike the MMTJA) contains a discretionary abstention provision as well. If more than one-third but less than two-thirds of the plaintiff class members are citizens of the state in which the action was filed, then the district court "may, in the interests of justice and looking at the totality of the circumstances, decline to exercise jurisdiction . . . based on consideration of" six articulated factors: (1) whether the claims involve "matters of national or interstate interest"; (2) which state's (or states') laws will apply; (3) whether the complaint has been pleaded in a manner that seeks to avoid federal jurisdiction; (4) whether the state in which the action was filed has "a distinct nexus with the class members, the alleged harm, or the defendants"; (5) whether the number of plaintiffs who are citizens of the state in which the action was filed "is substantially larger than the number of citizens from any other State, and the citizenship of the other members of the proposed class is dispersed among a substantial number of states"; and (6) whether a similar class action has been filed within the preceding three years.

The Act provides no basis for weighting any of these six elements any differently than the remaining elements. To the extent that this provision depends on the court's and the parties' ability to know the number of class members and their citizenship, it presents the same difficulty as do the jurisdictional and mandatory abstention provisions. Apart from that shortcoming, however, the factors provide some guidance for district courts that are grappling with whether a class action actually belongs in federal court. This provision, as with a number of other Act provisions, all but requires that the individual circuit courts of appeal add their own interpretive glosses to these factors in order to further constrain the district courts' exercise of their discretion whether to abstain or not.

C. Subject Matter Carve Outs

The Act contains additional jurisdictional limitations, or carve outs. It does not apply to any action in which (1) the potential number of plaintiff class members is less than 100 or (2) the primary defendants are states, state officials, or other governmental agencies "against whom the district court may be foreclosed from ordering relief" (presumably because of sovereign immunity and the Eleventh Amendment).

The Act also does not apply to actions (1) concerning a "covered security" under the Securities Act of 1933 and the Securities Exchange Act of 1934; (2) that relate to the "right, duties (including fiduciary duties), and obligations relating to or created by or pursuant to any security [as defined by the `33 Act]"; or (3) that relate to the "internal affairs or governance of a corporation or other form of business enterprise" and are brought under the laws of the state in which the enterprise is organized.

28 U.S.C. § 1332(d)(9). Curiously, the carveout in (d)(5) states that both the jurisdictional provision in (d)(2) and the abstention provisions in (d)(3) and (d)(4) do not apply, while the carveout in (d)(9) refers only to the jurisdictional provision. Nonetheless, it is difficult to imagine how the abstention provisions can apply to actions that fall within the scope of (d)(9), for if there is no jurisdiction, there is no opportunity to abstain.

These exclusions from the scope of the Act's coverage are plainly limitations on jurisdiction and not abstention provisions as are found in sections 1332(d)(3) and (d)(4). The class-size limitation suffers from the same shortcoming as all of the class-size requirements in the statute, but certainly there will be cases in which it is apparent at the outset that there are no more than 100 potential class members, and regardless of their citizenship and the amount in controversy, those cases must remain in state court unless there is an independent basis for federal jurisdiction. The sovereign immunity limitation's use of "may" suggests that if a sovereign immunity defense could be advanced by a defendant in good faith ( i.e., consistent with Fed.R.Civ.P. 11), then the action falls outside the Act's grant of jurisdiction. Many of the cases involving securities under the `33 and `34 Acts may have independent bases for federal jurisdiction, but if they do not, those actions are to remain in state court. Finally, the corporate governance exclusion appears to have been made in recognition of the interest that a state's courts have in applying its own corporate governance law and the skill that some courts ( e.g., the Delaware Court of Chancery) have developed in resolving such disputes.

D. Removal

The new removal provisions are found in 28 U.S.C. § 1453, and there are three significant ways in which the procedures governing removal as it is available in ordinary diversity cases have been made more liberal. Removal of class actions is governed, as is removal of diversity actions generally, by § 1446, but § 1453 makes certain exceptions. First, where § 1446 requires that all defendants join in or consent to a notice of removal, § 1453 states that a class action "may be removed by any defendant without the consent of all defendants." Second, where § 1441(b) precludes removal on diversity jurisdiction grounds of actions in which any defendant is a citizen of the state in which the action is pending, the Act states that class actions "may be removed . . . without regard to whether any defendant is a citizen of the State in which the action is brought." Third, where § 1446(b) precludes removal on diversity jurisdiction grounds of an action more than one year after the action was commenced, the Act states that "the 1-year limitation . . . shall not apply" to class actions.

See, e.g., 13B CHARLES ALAN WRIGHT, ARTHUR R. MILLER EDWARD H. COOPER, FEDERAL PRACTICE AND PROCEDURE: JURISDICTION § 3723 (3d ed. 1998) ("a cardinal rule is that only defendants may remove and all defendants must join in the notice of removal" (footnotes omitted)).


Id. The Act also provides when citizenship is to be determined, both for purposes of jurisdiction and abstention: either "as of the date of the filing of the complaint or amended complaint, or, if the case stated by the initial pleading is not subject to Federal jurisdiction, as of the date of service by plaintiffs of an amended pleading, motion, or other paper, indicating the existence of Federal jurisdiction." 28 U.S.C. § 1332(d)(7). This rule appears to be generally consistent with the traditional requirement under § 1441 that citizenship, and the existence of diversity, be determined at the time of removal. See, e.g., Grupo Dataflux v. Atlas Global Group, L.P., 541 U.S. 567, 124 S. Ct. 1920, 1925-26, 158 L.Ed.2d 866 (2004) (discussing Caterpillar Inc v. Lewis, 519 U.S. 61, 117 S.Ct. 467, 136 L.Ed.2d 437 (1996)).

The principal difficulty with removal of class actions will involve deciding at what point, in cases in which there is doubt about the existence of diversity or satisfaction of the jurisdictional amount, a defendant is on notice that jurisdiction under § 1332(d) exists. At that point — no matter how far the litigation has proceeded in state court — any defendant may remove the case to federal court. By removing the one-year limitation found in § 1446(b), and by expressly stating that the jurisdictional provision applies "to any class action before or after the entry of a class certification order by the court," the Act invites defendants to consider delaying removal of an action as long as possible. While there would be risks in delaying removal when it is apparent that federal jurisdiction exists, in cases in which the existence of jurisdiction is not certain, defendants may choose to have state courts handle the case to a certain point and then, if dissatisfied with the rulings to date (including, possibly, a ruling certifying a class), and if it is at that point apparent that federal jurisdiction exists (because either the citizenship of class members or amount in controversy has been clarified), then the Act seems to contemplate permitting defendants to remove actions to federal court, at which point the question of class certification (and perhaps other antecedent rulings) could be revisited. This sort of forum shopping may happen only in rare cases in which the citizenship of the class members or, more likely, the amount in controversy is difficult to determine at the outset, but when the facts permit it, nothing in the Act would appear to preclude it.

E. Appeal of Remand Decisions

Delving into an already murky area — the ability of litigants to appeal district court's decisions on motions to remand — the Act both expands the availability of such appeals and requires that the Courts of Appeals expedite consideration of them. This provision appears to be effective, as two courts of appeals have heard appeals under the Act and issued decisions during the first four months after enactment of the Act.

See Knudsen v. Liberty Mut. Ins. Co., 411 F.3d 805, 2005 WL 1389059 (7th Cir. June 7, 2005); Pritchett v. Office Depot, Inc., 404 F.3d 1232 (10th Cir. 2005) (decided April 11, 52 days after Act's effective date).

In the ordinary case, section 1447(d) provides that an order remanding a case to state court "is not reviewable on appeal or otherwise," unless the case was removed under section 1443, which governs certain civil rights claims. The Supreme Court has confined the category of nonreviewable remand orders to those that are "based on grounds specified in § 1447(c)." Section 1447(c) governs motions to remand that are based on any ground other than lack of subject matter jurisdiction.

Things Remembered, Inc. v. Petrarca, 516 U.S. 124, 127, 116 S.Ct. 494, 133 L.Ed.2d 461 (1995) (discussing Thermtron Prods., Inc. v. Hermansdorfer, 423 U.S. 336, 96 S.Ct. 584, 46 L.Ed.2d 542 (1976)).

The Act expressly permits discretionary appeals from orders granting or denying remand motions in class actions, "notwithstanding section 1447(d)." Section 1453 contains a carveout that is identical to section 1332(d)(9) — relating, generally, to securities and corporate governance class actions — but does not include a provision similar to section 1332(d)(5), which carves out the small class (fewer than 100) or sovereign immunity cases.

In addition to expanding the availability of review of remand orders, the Act requires extremely speedy resolution of any appeals that a Court of Appeals does accept. A petition for permission to appeal must be filed within seven days. Such a petition must presumably comply with Fed.R.App.P. 5. There is no limitation on how long the Court of Appeals may take to rule on the petition, but if it accepts review, it must then rule on the merits ("render[] judgment") "not later than 60 days after the date on which such appeal was filed." The court may extend that time for up to 10 days or, with the agreement of all parties, for a larger period of time. If the Court of Appeals has not ruled within the permitted 60 days (or as extended), the appeal is deemed denied.

The text of the Act actually provides that application must be made to the Court of Appeals "not less than 7 days after entry of the order." 28 U.S.C. § 1453(c)(1). Within the context of the Act, it seems fairly clear that the word "less" was intended to be "more." Precisely how the Courts of Appeals will resolve this problem remains to be seen (though it seems unlikely that the courts will permit appeals to be filed years after a remand decision is made). Until the appellate courts have addressed this issue, the safest course may be to petition for permission to appeal an order remanding or refusing to remand a class action on the seventh day following the entry of such order.

Sixty days is a short time period for a Court of Appeals to have briefing, hold oral argument, and rule, which will presumably have two consequences: (1) the Courts of Appeals will grant such petitions sparingly, and (2) when they are granted, parties should anticipate extremely expedited proceedings.

The 60-day period begins to run on "the date on which such appeal was filed." One might read this to mean when the petition for permission to appeal was filed, but that is likely not a sound reading of the statute. Fed.R.App.P. 5(d)(2) provides that when a petition to appeal is granted, the date on which the petition is granted is considered the date of the filing of the notice of appeal. Until permission is granted, there is no appeal, so reading Rule 5(d)(2) in conjunction with section 1453, it appears that the 60-day time period begins to run on the date on which an order is entered granting a party's petition to appeal.

F. The Benefit of Federalization

There appear to be two primary benefits in bringing interstate class actions to federal court. First, there are significant concerns about some state court judges' independence and the wisdom of having them decide cases involving national classes when they run for reelection and are possibly beholden to local attorneys. Some state court systems also have appellate systems that do not offer any appellate review as a matter of right, making these fora less acceptable for cases involving distant litigants. In this regard, the concern is essentially the same for all diversity cases, but the class action context raises the stakes, and justifies more ready access to federal courts. The second, and probably stronger, justification for allowing class cases to be brought in federal court relates to the efficient handling of the cases. By allowing the cases to be brought in or removed to federal court, the Act essentially puts them within the purview of the Judicial Panel on Multidistrict Litigation, and allows them to be transferred for centralized handling in a single district.

The Panel ability to manage complex litigation by centralizing cases before a single judge is well known and well accepted. The Panel and the practice of assigning multiple cases to a single judge have been widely emulated in state court systems as well, though this does not impact multistate class actions. Transfer of class actions — particularly multiple class actions with potentially conflicting class definitions — is a likely result once they are before the Panel. The Panel itself has consistently recognized that conflicting or multiple class actions constitute a strong reason justifying transfer by the Panel. It stated early in its existence:

See, e.g., Earle F. Kyle, IV, The Mechanics of Motion Practice Before the Judicial Panel on Multidistrict Litigation, 175 F.R.D. 589, 590 (1998) ("Consolidated or coordinated pretrial proceedings resulting from Panel transfer orders can provide tremendous benefits. Parties can save time and expense by avoiding duplicative discovery and minimizing the burden on non-parties who may be compelled to testify or produce documents in similar actions scattered around the country.").



Our experience since [an earlier case] has reinforced our belief that resolution of potentially conflicting class actions are best made by the transferee court. . . . Such a potential for conflicting or overlapping class actions presents one of the strongest reasons for transferring such related actions to a single district for coordinated or consolidated pretrial proceedings which will include an early resolution of such potential conflicts.

Arntl Bros. Constr. Co. v. American Radiator Standard Sanitary Corp. (In re Multidistrict Private Civil Treble Damage Litigation Involving Plumbing Fixtures), 308 F. Supp. 242, 243-44 (J.P.M.L. 1970). That wisdom continues, or is even more pronounced, today. See Bishop v. Maytag Corp. (In re Maytag Corp. Neptune Washer Prods. Liability Litigation), 333 F. Supp. 2d 1382 (J.P.M.L. 2004) (3 putative class actions centralized); In re Urethane Antitrust Litigation, 333 F. Supp. 2d 1379 (J.P.M.L. 2004) (Panel notes danger of conflicting classes).

Because the Panel has no jurisdiction over actions pending in state court, it has not been able to centralize these cases by assign to a single judge.

Congress's most recent resort to federalization of state-court claims as a means of dealing with the burgeoning case management challenges of complex civil litigation was the Multiparty, Multiforum Trial Jurisdiction Act of 2002, which was adopted as part of appropriations for the Justice Department. That law addressed, in a quite limited way, one of the two most important limitations on the transfer power for complex cases contained in 28 U.S.C. § 1407. Under that statute, the Judicial Panel for Multidistrict Ligation is authorized to transfer cases pending in multiple federal district courts for consolidated pretrial proceedings. Commentators have recognized for years that the limitation of transfer being available only to federal court actions is a significant limitation on the Panel's power. This legislation does not remedy that general problem in any way, but it moots it by allowing class actions to be filed in or removed to federal court.

28 U.S.C. § 1369.

See 21st Century Department of Justice Appropriations Authorization Act, Pub.L. 107-273, 2002 H.R. 2215, 116 Stat. 1758 (Nov. 2, 2002).

See generally HERR supra note 43.

See generally HERR, supra note 43, at § 3:13.

IV. The Curious Treatment of Mass Actions

The Act includes a curious provision limiting the power of the courts to deal with "mass actions." Despite treating these cases as class actions for most purposes, it expressly limits the power of the Judicial Panel on Multidistrict Litigation to transfer them for coordinated pretrial proceedings.

A. Mass Actions Defined

"Mass action" is not a term of art in the world of civil procedure, but it is defined in the Act as "any civil action . . . in which monetary relief claims of 100 or more persons are proposed to be tried jointly on the ground that the plaintiffs' claims involve common questions of law or fact," except for five categories of cases:

1. class actions in federal court under the diversity statute, as amended;
2. cases where "all of the claims in the action arise from an event or occurrence in the State in which the action was filed, and that allegedly resulted in injuries in that State or in States contiguous to that State";
3. cases where the claims are joined upon motion of a defendant;
4. cases "all of the claims in the action are asserted on behalf of the general public (and not on behalf of individual claimants or members of a purported class) pursuant to a State statute specifically authorizing such action"; and
5. cases where "the claims have been consolidated or coordinated solely for pretrial proceedings."

Mass actions are presumably intended to include a variety of types of actions assembled to avoid class action treatment but still accomplish some economy of numbers. The most common of these, "mass-joinder" actions, are cases in which numerous plaintiffs attempt to join their cases together. Mississippi used to allow virtually unlimited joinder of parties in this matter, but has more recently conformed to a practice more like that under Fed.R.Civ.P. 20, allowing joinder only of related claims, not merely similar claims. The definition of "mass action" is broad, however, and it would also extend to other situations. A state-court attempt to order a mass consolidation, so long as it was not requested by a defendant, would potentially invoke the rule and removal of the case to federal court would be possible. In this situation, the defendant might welcome the special treatment mass actions receive.

Compare Am. Bankers Ins. Co. of Fla. v. Alexander, 818 So. 2d 1073, 1078 (Miss. 2001) (allowing joinder of 1,371 claimants in single action) with Harold's Auto Parts, Inc., v. Mangialardi, 889 So.2d 493, 495 (Miss. 2004) (terming joinder of 264 tort cases against numerous different defendants as "a perversion of the judicial system unknown prior to the filing of mass-tort cases") and Janssen Pharmaceutica, Inc., v. Armond, 866 So.2d 1092 (Miss. 2004) (severance of 56 product liability claims ordered).

B. Treatment Like Class Actions

For purposes of jurisdiction, including original diversity jurisdiction and removal jurisdiction, the Act deems mass actions to be class actions. Additionally, the statute extends to these actions the tolling of the statute of limitations during pendency of the action in federal court after removal.

Once removed to federal court, mass actions are subject to management orders by the federal judge. Where the action is the result of mass-joinder foreign to practice under Fed.R.Civ.P. 20, severance may be possible, although this may also have the effect of destroying federal jurisdiction, depending on citizenship and the amounts claimed by individual plaintiffs.

C. Limitations on Transfer under 28 U.S.C. § 1407

Mass actions that are removed to federal court are not subject to transfer by the Judicial Panel on Multidistrict Panel unless a majority of the plaintiffs request transfer or the action is either certified as a class action or the plaintiffs "propose that the action proceed as a class action." This provision is a confounding one, for it denies the federal courts one of the most effective tools available to foster the efficient and fair handling of complex cases. In the case of mass tort litigation involving a single product, the courts may be faced with a master, centralized docket in one court of cases transferred by the Panel, but pockets of removed "mass action" cases that will continue to pend in other federal courts, unable to be transferred.

D. Potential Solutions

The broad definition of "mass actions" coupled with the requirement that they be treated differently from other, similar cases creates an impractical situation. It may be possible in some cases to transfer these cases under the venue statutes, although this option would be available only for transfer to another jurisidiction where venue would have been proper initially.

See 28 U.S.C. § 1404(a). MDL transfer under § 1407 is available to any district, without regard to venue. See HERR, supra note 43, at § 3:4.

V. Approval of Settlements and Attorney Fees

The Act's provisions relating to approval of settlement are, at best, curious. They address only some of the major issues surrounding class actions, and include an elaborate but toothless notice mechanism that doesn't seem to address any known problem.

A. Coupon Settlements

Coupon (also called "scrip" or "in-kind") settlements have been the subject of much criticism from scholars and consumer advocates. "Coupon" settlements have the potential to be attractive, creative settlement vehicles where the settling defendant can provide something of greater value to the recipients than it may cost to provide. In that situation, they may present "win-win" devices. It is beyond dispute, however, that they have been powerful tools to misstate or inflate the true "value" of a settlement to a class, essentially by providing a settlement coupon that is not readily capable of evaluation to "fair market value," where no market exists for the coupon results. Additionally, the $100 coupon may be theoretically available to 12 million class members, but if only 180,000 are redeemed, it is hardly fair to value the settlement overall on the coupons that are never redeemed (even aside from whatever implications non-redemption would have on the true value of the coupon).

See generally Note, In-Kind Class Action Settlements, 109 HARV. L. REV. 810 (1996). For detailed analysis of mechanisms whereby coupon settlements are rendered less valauble, if not valueless, see Christopher R. Leslie, A Market-Based Approach to Coupon Settlements in Antitrust and Consumer Class Action Litigation, 49 UCLA L. REV. 991 (2002).

The Act's mechanism for dealing with coupon settlement arrangements is to mandate that the amount of any attorney fee recovery by class counsel be based on the value of the coupons actually redeemed by class members. As a practical matter both basing fees on value and limiting that value to the coupons redeemed, rather than their theoretical value, will make coupon settlements significantly less attractive to class counsel. The statute requires the court to hold a hearing on approval of a coupon settlement, and authorizes the court to hear expert evidence on the valuation question.

28 U.S.C.A. § 1712(e). The Act has been cited as providing guidance on the propriety of limiting fee awards to a percentage of the actual value of amounts claimed by class members. See Fears v. Wilhelmina Model Agency, Inc., 2005 WL 1041134, at *4-5 (S.D.N.Y. 2005); cf. In re Conpact Disc Minimum Advertised Price Antitrust Litigation, 370 F.Supp.2d 320 (D.Me. 2005) (court calcualtes fee based on 2% of coupons actually redeemed as Act would require; but case in not subject to Act).

The likely effect of this provision will be to discourage the use of coupons. The process of determining the value of the coupons will be arduous. The value will not be determinable until the time for their redemption has passed, and will often require retention of additional experts. Class counsel will not have an artificial reason to favor a coupon over cash. The remedy is not as Draconian as some courts have suggested — requiring class counsel to accept their fees in coupons and giving cash to the class members — but it will level the playing field between cash and non-cash deals.

B. Geographic Favoritism

The Act restricts settlements that favor some class members over others based solely on geography. It specifically prohibits approval of settlements that provide "for the payment of greater sums to some class members than to others solely on the basis that the class members to whom the greater sums are paid are located in closer geographic proximity to the court." This provision is particularly intended to stop the practice in some state court settlements, now removable to federal court under the Act, whereby in-state class members were treated more favorably than out-of-state claimants.

C. Negative Value Settlements

The Act deals with zero-value settlements (also called "negative-value" settlements) — those where a class member may owe more for fees than is recovered — by permitting court approval of a settlement only if the court makes a finding that the non-monetary benefits of the settlement outweigh what would otherwise be an economic loss. This is not a trivial concern. In one notorious class action settlement, one of the class members received a $2.19 credit to his account as the benefit from the litigation, and a $91.33 charge to his account for his share of the attorney fees awarded by a distant state court judge. The statutory language is both mandatory and clear in requiring unanimity: the court must not approve a settlement in which "any class member" has to pay more in fees to class counsel than received in settlement. The only exception to the mandatory language requires a judge to make findings to support a determination that injunctive or other noneconomic relief justifies a recovery that would otherwise be prohibited as having "negative value."

See 28 U.S.C.A. § 1713.

See Kamilewicz v. Bank of Boston Corp., 100 F.3d 1348, 1349 (7th Cir. 1996) (Easterbrook, J., dissenting from denial of rehearing en banc). Negative value suits generally are discussed in Linda S. Mullenix, Negative Value Suits, NAT'L L.J. 11 (March 22, 2004).

D. Notice

The Act contains new notice provisions that are complex and problematic. They require notice to public officials who may be difficult to identify, but make the proper identification of them (and delivery of notice) crucial to obtaining final adjudication of the class claims by means of the "final" settlement.

The Act imposes new requirements for notice to state and federal officials. These notices must be given within 10 days after the class action is filed. Notice must be given to an "appropriate State official of each state in which a class member resides," and "the appropriate Federal official." The appropriate official is defined by the statute in § 1715(a). For the notice to state officials, the primary regulatory official or licensing agency for defendant is appropriate, or in the absence of such an official, the state attorney general. For the Federal government, the appropriate official is generally the Attorney General, but in the case of "a Federal depository institution, a State depository institution, a depository institution holding company, a foreign bank" or subsidiary, as those terms are defined in 12 U.S.C.A. § 1813, the primary regulator of that institution.

The notice to these officials must contain specific information. Section 1715(b) requires that the notice consist of:

1. A copy of the complaint (or access to it if filed electronically);
2. Notice of any pending hearing;
3. Any notice to class members relating to exclusion from the class or a statement that no such right exists and notice of any proposed settlement;
4. Any settlement or other agreement between class counsel and the defendant;
5. Any judgment or notice of dismissal;
6. Information on the number of class members in the state for the State official's notice; and
7. Any written opinion about the foregoing.

This form of notice has no clear precursor in class action practice, and it is not clear what vice it will remedy. The Act nowhere gives any substantive or procedural rights to the recipients of the notice — it does not suggest that they have a remedy. In many cases, they would not have standing to object or intervene in the action. There has been little complaint that public officials have not had notice of class actions, so this elaborate mechanism seems something of a solution seeking a problem. The law does not do anything to change or improve the form of notice given to class members.

Cf. MMTJA, 28 U.S.C. § 1369(d) (injured persons granted right to intervene in removed action, without regard to venue).

One effect of this notice is to start a 90-day period during which the court cannot give final approval to a settlement. This provision is not particularly troublesome, although it will add some delay to many class action settlements. The bigger problem is the risk that failure to give the proper notice to the correct public official may render the class settlement non-binding on a class member, at the member's option. This provision is likely to be especially problematic because it will potentially cast many settlements into doubt, or will permit collateral attack on court-approved settlements and judgements. It does not limit the time during which such an attack might be made, and for settlements involving future claimants, the settlement process could become rife with doubt. One doubts that Congress intended to burden the settlement process in this way, but it it seems a certain consequence of this provision.

28 U.S.C.A. § 1751(e)(1).

If the purpose of the new notice requirement is the prevention of collusive settlements, it fails completely. It simply does not reach the worst of the cases: the truly collusive suit where the defendant is rewarded for settling on terms unfair to the class but favorable to class counsel. If the class counsel sues in a jurisdiction where a settlement is likely to be approved, the defendant will have little or no incentive to remove the class action to federal court, and none of the Act's reforms will be invoked. Indeed, in this respect, by creating a disincentive to remove a case to federal court, the Act may stand as a barrier to reform rather than the bearer of it.

VI. Shortcomings of the Act

A. Technical Flaws

The effective-date provision relating to the federal rules is an obvious technical flaw in the statute, but not one that probably requires a cure. The drafting of the timing provision relating to appeals is much more likely to create confusion, and should be fixed. Numerous other problematic provisions are identified above. They may well require extensive litigation to provide meaningful interpretation, but they probably will not prompt serious problems.

B. Missed Opportunities

The bigger problem with this legislation is that it reformed some of problems with class actions, but left others untouched. As discussed above, it did not meaningfully address concerns about truly collusive settlements. This legislation presented an opportunity deals with the so-called " Lexecon" problem, but it not only did not address Lexecon, it potentially limited the ability of the Judicial Panel on Multidistrict Litigation to deal with mass torts effectively by carving out "mass actions" from the transfer power.

In 1998 the Supreme Court decided Lexecon Inc. v. Milberg Weiss Bershad Hynes Lerach, 523 U.S. 26, 118 S. Ct. 956, 140 L.Ed.2d 62 (1998), and invalidated the practice of transferee courts to entertain motions to transfer cases to themselves for trial pursuant to 28 U.S.C. § 1404(a). Before being struck down by the Court, the self-transfer power had earned general acceptance because it worked well in practice and permitted the federal courts to be better managers of their burgeoning dockets; its end was marked by courts needing to look carefully at the jurisdictional underpinnings and the determination that Congress had never authorized nor intended this practice. The obvious solution to this problem is to pass legislation effectively overruling Lexecon. Congress has had numerous proposed bills to accomplish this, but has not yet acted.

The problem of collusive settlements is potentially serious regardless of how widespread it may be, and the Act did nothing to address the problem. The Act does nothing to address "clear sailing" agreements, which could have been easily prohibited in a class member's "Bill of Rights."

See generally William Henderson, Clear Sailing Agreements: A Special Form of Collusion in Class Action Settlements, 77 TUL. L. REV. 813 (2003).


Although hardly a model of legislative drafting, the Class Action Fairness Act of 2005 is generally a fair and effective attempt to deal with some of the issues confronting the court in connection with class actions. The majority of class actions will be heard in federal court, and because of the unlikelihood of remand, will be resolved there. Its provisions, coupled with the 2003 amendments to Fed.R.Civ.P. 23, should permit class actions to be handled fairly and more efficiently around the country, and should foster greater fairness to plaintiffs and defendants in these vexing cases. It could have accomplished more, and has a few glaring problems that should be fixed by Congress, but on balance, it is probably legislation that will accomplish a little less than both its proponents and detractors claimed.

Summaries of


Judicial Panel on Multidistrict Litigation
Jan 1, 2005
228 F.R.D. 673 (J.P.M.L. 2005)
Case details for


Case Details


Court:Judicial Panel on Multidistrict Litigation

Date published: Jan 1, 2005


228 F.R.D. 673 (J.P.M.L. 2005)