We’ve blogged before about plaintiffs’ attempts to circumvent the “mass action” provisions in the Class Action Fairness Act of 2005 (“CAFA”), which  allow defendants to remove to federal court certain cases raising “claims of 100 or more persons that are proposed to be tried jointly.” 28 U.S.C. § 1332(d)(11)(B)(i). To evade removal, creative plaintiffs’ lawyers have subdivided their mass actions into parallel cases of fewer than 100 persons each. Some courts have gone along with the charade. See, e.g., Scimone v. Carnival Corp., No. 13-12291 (11th Cir. July 1, 2013); Abrahamsen v. ConocoPhillips, Co., 503 F. App’x 157, 160 (3d Cir. 2012); Anderson v. Bayer Corp., 610 F.3d 390, 392 (7th Cir. 2010); Tanoh v. Dow Chem. Co., 561 F.3d 945, 950-51 (9th Cir. 2009).

The fight over removal in these gerrymandered mass actions often boils down to one key question:  whether the parallel cases are “proposed to be tried jointly.”  If so, CAFA permits removal.

Recognizing this point, the plaintiffs in these cases frequently remain coy about—or outright deny—an intent to try the parallel mass actions jointly.  But they often go right up to the edge, urging the same state trial court to resolve threshold issues in the cases together—or even simply to consolidate the state-court actions outright. Then, these plaintiffs say, CAFA’s mass-action removal provision doesn’t apply because they say that they have had the claims “consolidated or coordinated solely for pretrial proceedings.” 28 U.S.C. § 1332(d)(11)(B)(ii)(IV) (emphasis added).

But not all courts are falling for this effort to elevate form over substance.

An important example is the Eighth Circuit’s recent decision in Atwell v. Boston Scientific, 2013 WL 6056782 (8th Cir. Nov. 18, 2013), in which the plaintiffs had subdivided their state-court mass action into three cases with fewer than 100 plaintiffs each. The defendants removed the cases under CAFA, explaining that the plaintiffs’ arguments in seeking to have the cases all assigned to a single judge demonstrated that the plaintiffs in fact proposed to try the three cases jointly.  Specifically, the plaintiffs had disclosed that they intended to “select [a] bellwether case to try,” and mentioned that assigning the cases to a single judge was needed to “‘avoid[] conflicting pretrial rulings,’ ‘provid[e] consistency in the supervision of pretrial matters,’ and [achieve] ‘judicial economy.’”

The district court remanded the cases back to state court, but the Eighth Circuit reversed.  The court of appeals concluded that these references revealed that the plaintiffs contemplated a joint trial on the merits—or at least would be affirmatively requesting such a degree of coordination among the cases as to amount for all practical purposes to a single joint trial.

But Atwell is in direct conflict with two decisions by a panel of the Ninth Circuit—underscoring the need for greater certainty and clarity in this area.

First, in Romo v. Teva Pharmaceuticals USA, Inc., 731 F.3d 918, 921 (9th Cir. 2013), the plaintiffs had sought joint treatment of cases under a California state-law procedure that allows coordination of certain civil actions “for all purposes.” Although the plaintiffs’ arguments in state court supporting coordination were not limited to pretrial proceedings—and explicitly referred to a need to avoid “duplicative and inconsistent rulings, orders or judgments”—the Ninth Circuit panel held that there wasn’t sufficient basis to conclude that plaintiffs were seeking a joint trial. Placing a thumb on the scale by asserting that there is a strong presumption favoring remand, the panel’s decision effectively amounts to a rule that plaintiffs must expressly request a single joint trial to trigger removal under CAFA

Second, in Corber v. Xanodyne Pharmaceuticals, No. 13-56306 (9th Cir. Sept. 24, 2013), a companion case to Romo, the same panel applied Romo and affirmed an order remanding the mass actions in a memorandum disposition.

The panel’s holding in Romo and Corber was not unanimous; Judge Gould dissented.  His opinion in Romo focused on “the import of the [plaintiffs’] coordination motion [in state court] and the reasons given for it.” In contrast with the panel majority, he recognized that “proposal[s] for joint trials may be made implicitly,” and that only “matters consolidated exclusively for pretrial purposes are not properly removed to federal court.” Because the plaintiffs’ request for coordination “for all purposes” wasn’t limited to pretrial proceedings—and indeed would amount to a joint trial in the long run—Judge Gould explained that he would have held that the removal under CAFA was proper. (Judge Gould also dissented in Corber for the same reasons he did in Romo.)

The defendants have petitioned the Ninth Circuit in Romo and Corber for rehearing en banc, noting the existence of a circuit split with several decisions of the Seventh Circuit, including In re Abbott Labs, Inc., 698 F.3d 568 (7th Cir. 2012). The Eighth Circuit’s decision in Atwell deepens that split even further.

The issue is an important one.  As we’ve mentioned before, the issue of jurisdictional gerrymandering under the mass-action provision reminds us of attempts to get around CAFA’s amount-in-controversy requirement. Last term, in Standard Fire Insurance Co. v. Knowles, 133 S.Ct. 1345 (2013), the Supreme Court put an end to plaintiffs lawyers’ practice of stipulating that the class would seek less than $5 million in damages in order to avoid removal. Among other reasons for the Court’s decision, allowing a putative class representative to bind absent class members before certification would “exalt form over substance,” and would allow plaintiffs to split up “a $100 million action into 21 just-below-$5-million state-court actions”—an outcome that the Court said “would squarely conflict with the statute’s objective.”

The panel’s approach in Romo seems to suffer from the same types of problems that the Supreme Court found in Knowles. In Romo, the majority took a formalistic look at the reasons the plaintiffs had given for seeking coordination, and concluded (among other things) that removal couldn’t be based on a few “isolate[ed]” references in the record to the need for consistent judgments. But other courts—such as the Eighth Circuit in Atwell—have rejected such a stance, instead taking a more practical view of how a case is likely to be litigated. In particular, these courts have focused on whether plaintiffs had invoked an open-ended procedural rule, or offered reasons that contemplated that the cases might be tried jointly—rather than on whether plaintiffs had used the “tried jointly” magic words. Atwell’s approach makes sense to us; a contrary rule requiring remand unless plaintiffs expressly seek to have claims “tried jointly” would frustrate the statutory purpose of CAFA, which exempts from removal only those qualifying mass actions in which the plaintiffs have requested consolidation or coordination “solely for pretrial proceedings.”

We’ll be watching for developments on this issue, which—if the Ninth Circuit does not rehear Romo and Corber en banc—ultimately may be destined for the Supreme Court.