The Yale Law Journal


CAFA and Federalized Ambiguity: The Case for Discretion in the Unpredictable Class Action

25 Jan 2011
Michael D. Y. Sukenik & Adam J. Levitt

Millions of dollars in legal fees, along with a great deal of litigants’ and judges’ time, have been spent trying to unravel CAFA’s statutory framework and its practical meaning. Now, as confirmed by the recent jurisprudence, it is becoming increasingly clear that some of CAFA’s purportedly central tenets and its structural approach may be defective. With the broadening gap between Congress’s professed intent and the practical realities of CAFA’s application (or lack thereof), CAFA’s effects are now more questionable than ever. Indeed, despite congressional efforts to fundamentally rework the class action instrument, CAFA’s application remains plagued with systemic inconsistencies and fundamental inequities, which foster even more technical gamesmanship and efforts to thread the procedural needle than those that arose prior to CAFA’s enactment.

This Essay explores the defects in CAFA’s underpinnings. With CAFA perpetuating confusion arising under its structural flaws, a more rational approach is needed—one that recognizes the merits of primary federal jurisdiction over many cases but that also respects the interests of state courts. Rather than resurrecting the deflated jurisdictional framework or calling for a return to the pre-CAFA regime, we suggest an approach that bridges the gap between parties’ desire to litigate in their chosen forum and the states’ abilityto adjudicate claims under their own laws. By engrafting the “nerve center” test recently established by the Supreme Court in Hertz Corp. v. Friend onto the CAFA analysis, Congress can better advance its desire to reform class action litigation in an evenhanded and practical manner. Specifically, by enabling plaintiffs who file class actions in a corporate defendant’s “nerve center” to proceed in state court, CAFA will give appropriate deference to the plaintiffs’ choice of forum. At the same time, the removal mechanism would remain predominant while granting federal judges discretion to consider the competing policy and federalism interests implicated by class action litigation. Such a revision will firmly recognize the vitality and importance of our state courts in the class action context, while preserving federal adjudication in instances of questionable state interest.

I. Statutory Dissonance

Originally, a nonfederal question class action could proceed in federal court only if the party filing that action or the party seeking removal from state court demonstrated complete diversity among the parties and satisfied the individual amount-in-controversy requirements. This resulted in the predominance of state court adjudication of various actions. The cloning of suits often resulted in inconsistent judgments across different state courts, duplicative and wasteful proceedings, and pressure on defendants to settle claims to avoid expensive, multi-front battles.

To address these perceived inefficiencies and tactical snares, Congress enacted CAFA. The legislation attempted to counteract state courts’ purported pro-plaintiff bias and unsophisticated approach toward class action cases. At its core, CAFA lowers the burden for removing a state class action to federal court, thus enabling the adjudication of multiple actions in a single proceeding through the federal multidistrict litigation mechanism or by the ex ante pleading of several separate state-based classes in a single action. While providing for federal consideration of state class actions, CAFA marginalized the usually considerable weight given to a plaintiff’s choice of forum and concurrently deprived state courts of their ability to adjudicate matters involving state law issues or pertaining to corporate defendants headquartered in a given state.

CAFA also forced federal courts to resolve questions of first impression under various state laws, increasing their reticence to adjudicate the disputes without prior state court guidance. Just as importantly, federal adjudication often undermined class certification due to federal courts’ concerns over managing and applying the varied laws of numerous states in a single action. Indeed, such concerns directly contributed to the rejection of numerous proposed class actions. Because the failure to certify is often a death knell in class action litigation, litigants are frequently left without a resolution on the merits. Even in meritorious actions, individual suits for the claims at issue in a case are usually cost-prohibitive; the enormous litigation costs generally outweigh even the largest potential recovery for any class member. As the Supreme Court has explained:

  The policy at the very core of the class action mechanism is to overcome the problem that small recoveries do not provide the incentive for any individual to bring a solo action prosecuting his or her rights. A class action solves this problem by aggregating the relatively paltry potential recoveries into something worth someone’s (usually an attorney’s) labor.  

After its passage, CAFA’s failings received considerable scholarly attention. Much of that literature, however, focused on the burdens of proof needed to show federal jurisdiction and on the appellate standard of review. Remarkably, one of the central tenets of CAFA’s revolutionary approach was largely ignored. Under the traditional requirements for federal jurisdiction based on diversity of citizenship, the amount in controversy must exceed $75,000, and the case must involve citizens of different states. CAFA drastically altered this clear standard in two ways. First, CAFA eliminated the requirement of total diversity among parties: under the new, simple diversity metric federal jurisdiction exists if any member of a proposed class holds different citizenship from any defendant. Second, CAFA mandated that federal jurisdiction would apply whenever the amount in controversy for the class action exceeded $5,000,000, without reference to the individual plaintiff $75,000 controversy threshold mentioned earlier in the statute. Although seemingly clear at first glance, this second prong has created divergence among and within the circuits—a divergence that further underscores CAFA’s fundamental structural weaknesses.

The statutory tension is straightforward. After CAFA’s enactment, most courts tasked with administering federal class actions reflexively concluded that where plaintiffs sought an amount exceeding the $5,000,000 threshold (or, in a removal situation, where defendants ostensibly established that monetary threshold’s satisfaction), no thought needed to be given to the traditional $75,000 amount in controversy required in individual diversity actions. In fact, this reading of CAFA was thought so entirely uncontroversial that district courts did not even bother to note any alternative potential interpretation of the amount-in-controversy requirement.

More recently, however, the Eleventh Circuit, relying heavily on an earlier Ninth Circuit decision, declined to follow the herd. In Cappuccitti, Georgia consumers brought a class action under Georgia law in federal court, seeking recovery of fees that DirecTV charged to subscribers who cancelled their subscriptions prematurely. Reviewing the district court’s denial of DirecTV’s motion to compel arbitration, the Eleventh Circuit determined that the district court lacked subject matter jurisdiction to even consider the motion. The court reasoned that although the class action alleged damages exceeding $5,000,000 in the aggregate—as provided in § 1332(d)—no individual plaintiff or putative class member satisfied the $75,000 amount-in-controversy requirement mandated by § 1332(a). Observing that “CAFA did not alter the general diversity statute’s requirement[s]” under § 1332(a), the court concluded that the district court lacked subject matter jurisdiction to consider the merits of the arbitration clause dispute. Notably, the court refused to “transform federal courts . . . into small claims courts, where plaintiffs could bring five-dollar claims by alleging gargantuan class sizes to meet the $5,000,000 aggregate amount requirement.”

The Eleventh Circuit’s interpretation was clearly erroneous from both statutory and policy perspectives. On October 15, 2010, in a per curiam opinion, the Eleventh Circuit reversed its prior ruling, asserting that “[s]ubsequent reflection has led us to conclude that our interpretation was incorrect” and holding that “CAFA’s text does not require at least one plaintiff in a class action to meet the amount-in-controversy requirement of 28 U.S.C. § 1332(a).”

The Eleventh Circuit’s recent gyrations on CAFA’s amount-in-controversy issue merely add to the cacophony of voices that previously expressed dismay at CAFA’s confusing precepts. In reaching the result, the Eleventh Circuit relied upon § 1332(d)(11), which imposes additional prerequisites upon “mass actions,” where, “unlike in a class action, [plaintiffs] do not seek to represent the interests of parties not before the court.” That section specifies that jurisdiction exists only where an individual plaintiff avers a claim of more than $75,000, regardless of the aggregate $5,000,000 amount. The Ninth Circuit opinion relied upon by Cappuccitti as well as by a subsequent Ninth Circuit panel, supporting the demonstration of an individual amount in controversy, dealt with mass action removals rather than class actions. No such individual requirement is facially evident for the class action mechanism, suggesting that the Eleventh Circuit improperly conflated a state mass action with a class action.

Had the Eleventh Circuit not reversed itself, the policy implications of its initial Cappuccitti decision were momentous—and a direct result of CAFA’s ambiguity. Consider that under that ruling, consumer class actions would have been almost entirely removed from the federal docket, as such cases rarely involve plaintiffs with individual claims exceeding $75,000. In the absence of federal jurisdiction, an increasing number of class actions would have again been adjudicated at the state court level. Such an effect would have directly contradicted Congress’s intention in enacting CAFA, which endeavored to ensure “[f]ederal court consideration of interstate cases of national importance under diversity jurisdiction.” This is evident through the elimination of the complete diversity requirement in favor of minimal diversity and through the explicit provision for aggregation of class members’ claims to reach the amount-in-controversy requirement. In essence, CAFA’s goal was to bar the state courtroom door for the vast majority of multistate or nationwide class actions. The Eleventh Circuit’s decision directly undermined these interests, instead blocking access to the federal courthouse.

At the same time, the initial decision created a genuine problem for litigants. Both plaintiffs and defendants could no longer presume that minimal diversity and aggregation of costs would suffice to proceed in federal court. Instead, plaintiffs would need to consider whether a class action should be brought within the Eleventh Circuit—and potentially the Ninth Circuit as well—as a means of avoiding federal jurisdiction. Concurrently, defendants would attempt to transfer any class actions out of courts within the Eleventh Circuit’s geographical jurisdiction to protect the tactical advantage of federal court removal. Such unpredictability—and the inevitable forum-shopping that would have ensued—would raise the costs of litigation for all involved, as litigants would engage in the type of protracted scheming and briefing that prompted Congress to streamline the class action process in the first place.

While the Eleventh Circuit’s subsequent reversal may have resolved this immediate problem in the context of that particular litigation, its initial misinterpretation and misapplication of CAFA remains a strong indicator of the defects in the statute.

II. Toward an Improved Mechanism

In its current form, CAFA requires a federal district court to accept almost any removed class action from state court so long as it meets the minimal diversity and damages thresholds; the same is true for actions originally filed in federal court. Congress found such a broad extension of federal jurisdiction appropriate because of perceived class action abuses that occurred mainly in state and local courts “in cases that really ought to be heard in Federal court.” Exceptions to federal jurisdiction are rare, because very few national class actions satisfy CAFA’s stringent exception requirements. The so-called “home state” exception requires the district court to decline jurisdiction where more than two-thirds of the putative class members are citizens of the state in which the action was originally filed. Additionally, the “local controversy” exception allows the district court to decline to exercise jurisdiction where more than one-third but fewer than two-thirds of putative class members are citizens of the state where the action was originally filed. Indeed, Congress created these exceptions as a hedge against criticism that the federal government usurped state power.

But their practical effect is far removed from their stated intent. The overwhelming majority of class actions—the type targeted by Congress to prevent purported state court abuses—consists of nationwide or multistate putative classes of plaintiffs without regard to their citizenship. In fact, an action brought on behalf of a nationwide class is exceedingly unlikely to consist predominantly of citizens of a single state. Indeed, because it is axiomatic that most class actions seek to represent citizens of multiple states, the highlighted exceptions are the anomaly and usually occur in the significantly rarer instances of purely intrastate actions.

While this rubric may advance the goal of federalizing any class action involving interstate commerce, such an approach marginalizes three distinct and important interests. First, negligible deference is given to the plaintiff’s traditionally respected choice of a particular forum. Second, this framework disregards the legitimate and often significant interests of states in conducting and controlling litigation over matters brought in their courtrooms, even where an individual state’s citizens do not compose one-third of the putative class. Finally, CAFA’s procedural and practical obstacles greatly limit litigants’ ability to proceed on the merits, thus disincentivizing the prosecution of even meritorious claims through procedural denials on class certification or damages allocation.

Even more significantly, CAFA’s current wording unreasonably restricts a federal court’s discretion to decline to exercise jurisdiction over a class action properly brought in state court. This is significant for two reasons. First, Congress was “emphatic” in insisting that it was “not impinging in any way on the independence of the Federal judiciary [or] their discretionary judgments” in passing CAFA. The expansion of federal jurisdiction to cover almost all class actions, however, results in precisely this type of interference. Mandatory federal judicial authority over any class action satisfying CAFA deprives federal judges of the ability to remand a class action originally brought in state court regardless of the factual underpinnings of the case.

More crucial, however, is another problem. If Congress’s genuine goal in enacting CAFA was to remedy the abuses inherent in the prior system, then CAFA goes too far in attempting to cure these ills. The result has been the reflexive federalization of the vast majority of class actions and disregard for the various legitimate private and public interests mentioned above. A much simpler approach would serve the interests of CAFA supporters while protecting plaintiffs and states from federal domination.

Our proposed framework is simple and effective: discretion is placed in the hands of federal judges to determine, in the context of a removal petition, whether a particular multistate class action should be adjudicated in state or federal court. This contrasts with the existing statute, which requires federal judges to exercise jurisdiction over a removed action regardless of the plaintiffs’ selection of the forum or the forum state’s interest in the matter.

Importantly, this proposal aligns with and integrates the Supreme Court’s latest edict on diversity jurisdiction and state citizenship. In Hertz, a case involving removal to federal court under CAFA and a subsequent remand, the Court adopted the “nerve center” test for determining diversity of citizenship for corporate litigants. Under this test, a corporation is considered a citizen of the place where its “officers direct, control, and coordinate the corporation’s activities,” or its “nerve center.” Consider that, under the current framework, if a corporate defendant had its nerve center in Delaware, and a Delaware plaintiff brought a class action on behalf of a nationwide class in Delaware state court pursuant to Delaware law, CAFA would automatically provide for federal jurisdiction. Despite the reputation of Delaware courts for judiciously and capably adjudicating the most complex matters, the Delaware state court immediately would be stripped of its authority to interpret and enforce its own laws with regard to its citizens; concurrently, the plaintiffs would be deprived of their chosen forum.

Under our proposal, after an action is removed from state court, the federal judge would inquire as to whether the action was originally filed in a defendant’s “nerve center.” If so, the court would then determine whether the action satisfies the requisite jurisdictional threshold under 28 U.S.C. § 1332(a). If not—even assuming that all of the other CAFA jurisdictional requirements are satisfied—remand will occur as a matter of course, recognizing the considerable state interests implicated. Where the initial action is not filed in the corporate defendant’s nerve center, the federal court would balance the jurisdictional monetary threshold against CAFA’s broader mandate, which favors retention of the action in federal court. The court, in its discretion, could consider factors such as the traditional deference accorded a plaintiff’s choice of forum, the composition of the proposed multistate class, the amount of damages claimed, and the prevalence (or lack thereof) of home-state-specific issues of law implicated by the action. The federal judge would also consider the other factors set forth by Congress in § 1332(d) for situations where one-third of the litigants reside in the forum state. Not only would such an approach comport with CAFA’s dictates and the current state of CAFA-related jurisprudence, but it would also vest federal judges with discretion and flexibility to consider the practical realities and ramifications of the removal/remand issues before them.

Because Congress primarily focused on preventing abuse in state adjudication of multistate class actions, assuming that the mandatory remand factors described above are not satisfied, the federal judge would then ascertain whether the removed class action would be prone to abuse if remanded to the state court. Like Congress, the judge could consider the relationship of the action to the forum, the forum’s experience and history in adjudicating class actions, the forum’s willingness to approve unreasonable settlements, the forum’s capacity to adjudicate the action in a timely and adequate way, and any other factors indicative of abuse. Through this analysis, the federal court would determine whether plaintiffs engaged in inappropriate forum-shopping by bringing the state action. At the same time, the court would ensure that defendants do not run afoul of the federalism precedents established in the Erie line of casesby forum-shopping to escape a forum they viewed as less favorable to their interests.

Finally, the court’s analysis may conclude by scrutinizing the intrinsic reason for amending CAFA: namely, whether a federal class action is likely to receive a hearing on the merits or whether procedural roadblocks will prevent the litigants from receiving justice. The main factor in this prong of the analysis is the likelihood of certification under Federal Rule of Civil Procedure 23, taking into account the manageability issues resulting from inconsistent state laws. In some instances, the judge might decide that “significant manageability concerns” would preclude certification of various statewide classes. This conclusion would weigh in favor of remand of the several statewide actions to their respective originating state courts. Alternatively, were the judge to conclude that manageability would not automatically disqualify an action for class certification, the judge might very well retain the action. This factor cannot be considered in a vacuum, however. The judge could consider the potential for duplicative discovery and wasted resources in granting, denying, or postponing remand until the completion of certain pretrial proceedings.

Under the “nerve center” proposal advanced here, federal courts would retain the discretion to decide whether Congress’s concerns about potential abuses arising from the class action mechanism applied to a specific scenario. At the same time, the federal judiciary would consider the federalism concerns marginalized by CAFA by determining the forum state’s connection to the multistate class action and, thus, the relative merits of removal and remand.

The proposed approach is superior to the current framework because it posits a middle ground between the purported state court biases enabled by the pre-CAFA regime, on the one hand, and Congress’s reflexive federalizing of the vast majority of class action cases, on the other. Indeed, Congress’s efforts skewed the equities in a single direction, with minimal consideration of the corporate and geographic realities implicated by the class action mechanism. Rather than simply imposing an unwieldy bright-line rule, as CAFA presently does, the suggested approach takes the latter factors into account, incorporating the Supreme Court’s “nerve center” considerations and permitting federal judges latitude to render reasoned decisions on the facts.


Five years into CAFA’s reign, the Act’s removal and remand provisions, which were murky to begin with, are now indecipherable. The Eleventh Circuit’s Cappuccitti decision should be viewed as a call for statutory revision before CAFA becomes a victim of its own systemic flaws. Our proposal is not a radical recasting of CAFA. Rather, it is a modest but effective effort to rectify some of CAFA’s fundamental structural failures in a cohesive and consistent manner. The “nerve center” approach harmonizes some of CAFA’s larger conceptual gaps with the jurisprudential exposition of those gaps in cases such as Cappuccitti. This approach forges a rational middle ground with the sole objective of restoring predictability and order to a crucial aspect of class action litigation that has been in upheaval since CAFA’s enactment.

Although implementing this proposal would almost certainly not stem the fertilizing tide for the vast majority of class action lawsuits, this proposal recognizes the substantial state interests implicated in many of these actions as well as the current state of CAFA-related jurisprudence. Our approach seeks to reduce the confusion and inefficiencies that plague CAFA (and litigants) by providing an analytical mechanism to enable courts legitimately and consistently to determine appropriate contexts for CAFA removal and remand.

Michael Yanovsky Sukenik is a law clerk to the Honorable Marjorie O. Rendell of the United States Court of Appeals for the Third Circuit; he graduated from the University of Chicago Law School and Northwestern University. Adam J. Levitt is a partner in the Chicago office of Wolf Haldenstein Adler Freeman & Herz LLC; he is a graduate of the Northwestern University School of Law and Columbia College, Columbia University. The authors are grateful to John E. Tangren and Richard J. Arsenault for providing generous feedback and insight. Finally, the authors would like to thank Brad Tennis and the editors of The Yale Law Journal for their invaluable editorial contributions.

Preferred citation: Michael D. Y. Sukenik & Adam J. Levitt, CAFA and Federalized Ambiguity: The Case for Discretion in the Unpredictable Class Action, 120 Yale L.J. Online 233 (2011),