Plaintiff lawyers must be mighty allergic to federal court. They perform all sorts of maneuvers to avoid CAFA removal of mass actions. For example, they will artificially subdivide their cases into groups of under 100. And/or they will disclaim any intent to try the cases together. Do these circumventions work? Perhaps most important, since so many of these CAFA avoidances occur in California, will such circumventions work in the Ninth Circuit?

Maybe.

At first, the Ninth Circuit permitted these evasions in a couple of decisions, creating a split with the Seventh and Eighth Circuits. But then the Ninth Circuit took the cases up en banc. The Ninth Circuit is so large that an en banc panel does not include all of the judges. But an entire en banc panel always includes the Chief Judge. That turned out to be important in the 2014 Corber en banc case because the dissenter in one of the earlier panel decisions was Chief Judge Gould. Guess who authored the Corber en banc opinion? Chief Judge Gould took a pragmatic approach to what counts as a “joint trial” for purposes of CAFA. That approach put the Ninth Circuit in alignment with the Seventh and Eighth Circuits and concluded that a proposal for a joint trial may be made implicitly as well as explicitly. Yes, it would be simpler to administer a bright line rule requiring plaintiffs to utter the magic words “joint trial,” but such a rule “would ignore the real substance” of plaintiffs’ proposals. The plaintiffs had sought coordination “for all purposes.” They had argued in the California state court that coordination was needed to avoid “the danger of inconsistent judgments and conflicting determinations of liability.” That smells like a request for something that would actually or functionally be a joint trial. The Ninth Circuit held that CAFA removal was proper under such circumstances.

Goodbye circuit split, hello sanity. We praised the Corber decision here.

But the Corber opinion possibly suggested a road map — or another set of magic words — that might work to make federal jurisdiction disappear. What if plaintiffs explicitly limited their request for coordination “solely for pretrial purposes”? We all know that such a statement would be disingenuous. But would it work? Would it keep the cases in the pro-plaintiff maw of California’s coordination process?

The other shoe has fallen (sort of), the magic words have been uttered (sort of), and plaintiffs followed the road map (sort of). In Dunson et al. v. Cordis Corp., 2017 U.S. App. LEXIS 6446 (9th Cir. April 14, 2017), the Ninth Circuit upheld a remand of a mass action because the plaintiffs had not proposed a joint trial. (We have written about the Dunson case before.) Instead, the plaintiffs had argued that consolidation “for purposes of pretrial discovery and proceedings, along with the formation of a bellwether-trial process, will avoid unnecessary duplication of evidence and procedures in all of the actions, avoid the risk of inconsistent adjudication, and avoid many of the same witnesses testifying on common issues in all actions, as well as promote judicial economy and convenience.” We think this should be enough for CAFA removal, but the Dunson court held otherwise.

As an initial matter, the court says the appeal would be easy to resolve if the plaintiffs had simply sought consideration for “all pretrial purposes, including discovery and other proceedings,” and stopped there. The Dunson court would easily have held that there was no request for a joint trial and thus, no basis for CAFA jurisdiction. But the plaintiffs did not stop there. They went on to wax poetically about the virtues of a bellwether trial process. Do we now have a request for a joint trial?

The Dunson court held that it all came down to what sort of bellwether trial was being sought. Sometimes, rarely, the result of a bellwether trial will be binding on the other cases. (For the moment, we are using deliberately vague language on this point. More to come.) That definitely would meet the definition of a joint trial. If that is what the plaintiffs want, they must go to federal court. But much more typical is a bellwether trial that would not be binding, but would be merely illustrative. Such a bellwether trial, according to the Dunson court, would not be a joint trial and would not support CAFA jurisdiction. The Dunson court assumes that when plaintiffs ask for a bellwether trial, they are asking for the non-binding member of the species. Putting the burden on the defendant to show that the plaintiffs were proposing a joint trial, the Dunson court held that such a showing had not been made, that the plaintiffs had not sought coordination “for all purposes” as in Corber, and that, thus, remand to state court was appropriate. A dismal day for the defense. (The Dunson court supported some of its reasoning by citing another less-than-delightful Ninth Circuit case, Briggs, which we dissected here.)

There are many problems with the Dunson decision, including its departure from the pragmatic approach of Corber. Experienced defense counsel know precisely what the plaintiffs want. They want a process that permits asymmetrical discovery where the defendants have to cough up millions of pages and scores of company witness depositions, while most of the plaintiffs’ individual cases hardly get tested. That is, plaintiffs want a settlement machine. The Dunson court pooh-poohed the preclusive effect of a bellwether trial because it would not have such effect on other plaintiffs. But the Dunson court was forced to acknowledge that, “True, a verdict favorable to the plaintiff in the bellwether trial might be binding on the defendant under ordinary principles of issue preclusion, but that is not enough” (emphasis in original). How fair is that? Moreover, the Dunson court ignores the plaintiffs’ own admissions of what they were up to in their consolidation request. The plaintiffs wanted to avoid the risk of “inconsistent adjudications” (we bet the plaintiffs are pretty selective when it comes to that aversion) and they defined that risk as “different results tried before different judge and jury, etc.” The Dunson court admitted that such language “does suggest that a joint trial would be needed to avoid the risk of inconsistent adjudication.” Yes. Yes, it does. But the plaintiffs parked that language in a portion of their briefs generally extolling the wonders of consolidation (and overlooking the massive prejudice to defendants that can arise from consolidation), and the plaintiffs did, after all, remember to insert a disclaimer that they were not seeking a joint trial.

Look, we clerked on the Ninth Circuit and will defend it against all the usual ideological attacks. But this time, the Ninth Circuit got CAFA removal wrong. It ignored the Supreme Court’s admonition in Standard Fire Ins. Co. v. Knowles — a case nowhere even cited in Dunson — not to “exalt form over substance” in assessing CAFA jurisdiction. Perhaps another en banc decision will ride to the rescue.