The D.C. Circuit recently deepened a circuit split over whether district courts may certify a “fail-safe” class. In In re White, 64 F.4th 302 (D.C. Cir. 2023),the D.C. Circuit agreed that fail-safe classes are generally improper, but rejected the views of other circuits that categorically forbid such classes . Instead of what it described as an “extra-textual” limitation on class certification, the D.C. Circuit held that the existing requirements of Rule 23 (and a district court’s discretion to alter proposed class definitions) should be used to prevent certification of fail-safe classes.

Fail-safe classes

A fail-safe class is “one that is defined so that whether a person qualifies as a member depends on whether the person has a valid claim.” Messner v. Northshore Univ. Healthsystem, 669 F.3d 802, 825 (7th Cir. 2012). Clear-cut examples of a fail-safe class are classes of “all individuals who the defendant defrauded” or “all applicants against whom the defendant illegally discriminated.”

Defining a class in terms of the underlying merits poses at least two major problems. First, although identifying whether someone ultimately is a class member is impossible until the merits have been decided, class notice must be sent out long before that point. Second, if the only members of a fail-safe class are people with winning claims (i.e., those who have been defrauded or discriminated against), then the class has no members if the defendant wins at trial—meaning that no one is bound by a final judgment in the defendant’s favor. Subjecting a defendant to a class-wide trial that it can only lose but never win is profoundly unfair.

Unsurprisingly, most courts of appeals have held that it’s improper to certify fail-safe classes.  See, e.g., Olean Wholesale Grocery Coop., Inc. v. Bumble Bee Foods LLC, 31 F.4th 651, 669 n.14 (9th Cir. 2022) (en banc); Cordoba v. DIRECTV, LLC, 942 F.3d 1259, 1276-77 (11th Cir. 2019); Orduno v. Pietrzak, 932 F.3d 710, 716 (8th Cir. 2019); McCaster v. Darden Rests., Inc., 845 F.3d 794, 799 (7th Cir. 2017); Byrd v. Aaron’s Inc., 784 F.3d 154, 167 (3d Cir. 2015); In re Nexium Antitrust Litig., 777 F.3d 9, 22 (1st Cir. 2015); EQT Prod. Co. v. Adair, 764 F.3d 347, 360 n.9 (4th Cir. 2014); Young v. Nationwide Mut. Ins. Co., 693 F.3d 532, 538 (6th Cir. 2012).

But the Fifth Circuit has refused to hold that fail-safe classes are categorically impermissible, In re Rodriguez, 695 F.3d 360, 370 (5th Cir. 2012), and in White, the D.C. Circuit has joined the minority view. White, 64 F.3d at 313-15.

The D.C. Circuit’s analysis

In White, the D.C. Circuit was faced with an ERISA class action, in which the plaintiff sought to certify a proposed class of current or former employees of the defendant (or their beneficiaries) who had “been denied vested rights” to retirement benefits. 64 F.4th at 306. The district court denied certification, deeming the class definition “impermissibly fail-safe” because “whether retirement rights had vested was an issue to be resolved in the case.” Id. at 305-06. The D.C. Circuit then granted Rule 23(f) review and reversed.

The D.C. Circuit acknowledged that the “concerns” that other courts have expressed about fail-safe classes are “understandable.”  Id. at 313.  But the court explained that, in its view, “Rule 23 is a carefully structured rule that, properly applied, already addresses relevant defects in the class definition.” Id. at 313. Rather than “deploy[] a textually untethered and potentially disuniform criterion,” the D.C. Circuit reasoned, “courts should stick to Rule 23’s specific requirements when making class certification decisions.” Id. at 313-14.

The D.C. Circuit explained that properly applying Rule 23’s express requirements likely would “assuage” any “‘fail-safe’ concerns”:

  •  “[N]umerosity must exist throughout the litigation,” “[y]et a class that could be defined to have zero members if the plaintiffs lose is not numerous at all.” Id. at 314.
  •  “[A] circular class definition could reveal the lack of a genuinely common issue of law or fact.” Id.
  •  “Typicality too should be a hard hill to climb if the named plaintiffs might not be members of the class come final judgment,” and “the same applies to the adequacy prerequisite.” Id.
  • “[A] class action would fail to be a superior device for resolving a dispute if the class would collapse should the plaintiffs lose on the merits.” Id.

The court added that Rule 23(c)’s provisions for what must be included in a class notice—including a plain-language explanation of “the binding effect of a class judgment on members”—are also “fatal to an indeterminate class definition.” Id.

In other words, the court concluded, “apply[ing] the terms of Rule 23 as written” should “eliminate most, if not all, genuinely fail-safe class definitions.” Id. And in the “rare cases” in which a fail-safe class “hurdles all of Rule 23’s requirements,” district courts can likely change the wording of the class definition, either sua sponte or by “work[ing] with counsel,” to “eliminate the problem.” Id. at 314-15.

Because the district court’s rejection of the proposed fail-safe class “bypassed” consideration of “Rule 23’s requirements,” the D.C. Circuit reversed and remanded.

The upshot

Although deepening the circuit split regarding fail-safe classes, the D.C. Circuit took pains to emphasize that, under its approach, fail-safe classes should routinely flunk Rule 23’s requirements for class certification. In practice—to the extent they are not doing this already—companies that are opposing class certification should explain why putative fail-safe classes fail to satisfy Rule 23’s express requirements.

That said, we think the better view of Rule 23 is that the majority rule against fail-safe classes itself follows from faithful application of Rule 23’s requirements and is not an additional requirement. After all, as the White court pointed out, when a putative class is a fail-safe class, it almost never should satisfy Rule 23(a)’s prerequisites for class certification or Rule 23(b)(3)’s superiority requirement.

Moreover, the majority rule against fail-safe classes has several virtues.

First, the rule is simple and easy to administer. Given the relative complexity of class certification doctrine, a clear and easy-to-apply rule benefits courts and parties alike.

Second, by framing the issue as compliance with a welter of Rule 23 requirements, the D.C. Circuit’s approach may cause some district courts to defer consideration of the issue until the class certification stage, after discovery is complete. Yet an early motion to strike a fail-safe class would allow for efficient resolution of the class certification question earlier in the litigation—something that Rule 23 itself encourages.

Third, encouraging district courts to attempt to resolve concerns with fail-safe classes by sua sponte redefining the proposed class can lead to unfairness and improper class certifications. In opposing class certification, defendants generally confine their briefing to the issues posed by the plaintiffs’ proposed class definition—not whatever alternative the court might later formulate after briefing and a hearing. The district court’s alternative version may suffer from yet other fatal defects. To avoid these problems, district courts should ensure that the parties are given an opportunity to brief any proposed class definition, even if the district court is the one proposing it.