chevron-down Created with Sketch Beta.
March 18, 2013 Articles

The Limited Scope of the Ascertainability Requirement

Courts should take care not to stretch this useful requirement beyond its intended purposes.

By Daniel Seltz and Jordan Elias

The law of ascertainability at class certification is straightforward. A class definition satisfies Rule 23’s implied ascertainability requirement if the class is defined by reference to objective factors and it is administratively feasible to determine whether a particular individual belongs to the class. As summarized in the Manual for Complex Litigation, a class definition must “avoid subjective standards (e.g., a class member’s state of mind) or terms that depend on resolution of the merits (e.g., ‘persons who were discriminated against’).” Manual for Complex Litigation (Fourth) § 21.222, at 270 (2004).

There are several sound reasons for the ascertainability requirement. First, courts require classes to be defined in a clearly ascertainable manner to identify who must receive notice of class certification under Rule 23(c)(2). Second, the definition of an ascertainable class will identify who will be bound by any judgment, thus avoiding subsequent litigation over a judgment’s preclusive effect. Third, the ascertainability requirement ensures it will be feasible to identify class members who may ultimately be entitled to damages or other relief. Defining class membership by reference solely to objective factors also avoids fail-safe classes, in which a class is defined in terms of the merits of a plaintiff’s claims. Members of a fail-safe class can be identified only after the defendant has been found liable, meaning that a favorable ruling will apply only to the plaintiff and not to the class, in addition to making class notice impossible.

The implied ascertainability requirement, however, should not be stretched beyond its limited purposes. By inquiring into ascertainability, courts safeguard and promote the efficiencies promised by Rule 23. But the requirement should not be a vehicle for introducing new or insurmountable burdens to class certification through overly broad or rigid interpretations. Courts should not demand precision as to the identities of individual class members at class certification, or showings that the class will ultimately prevail or that class members will ultimately be entitled to share in potential damages. The ascertainability analysis instead serves the more limited purposes of ensuring that notice may issue, that potential claims processes are administratively feasible, and that those who will be bound by any judgment can be identified.

Common Misuses and Case Studies

One common misuse of the ascertainability requirement is to insist that all of the class members’ identities be known when a plaintiff moves for class certification. This is unnecessary. Donovan v. Philip Morris USA, Inc., 2012 U.S. Dist. LEXIS 37974, at *91 (D. Mass. Mar. 21, 2012). One formulation is that, for class certification to be proper, the class must be “ascertainable” but not necessarily “ascertained.” Manual for Complex Litigation § 21.222, at 270; 1 Newberg on Class Actions § 3:3, at 163 (5th ed. 2011). So long as a court can determine “the general outlines of the membership of the class,” an ascertainable class may be found. In re Checking Account Overdraft Litig., 275 F.R.D. 666, 672 (S.D. Fla. 2011) (quoting 7A Wright, Miller & Kane, Federal Practice & Procedure: Civil 3d § 1760 (3d ed. 2005)).

The ascertainability requirement thus dovetails with Rule 23’s notice provisions, which do not require individual notice in all cases but only “the best notice that is practicable under the circumstances[.]” Fed. R. Civ. P. 23(c)(2)(B). As a result, many courts have analyzed ascertainability in conjunction with the feasibility of notice. In doing so, a court confirms that membership in the class is sufficiently ascertainable, not for its own sake, but “to permit a court to ‘decide and declare who will receive notice, who will share in any recovery, and who will be bound by the judgment.’” Donovan, 2012 U.S. Dist. LEXIS 37974, at *91–92 (quoting Kent v. SunAmerica Life Ins. Co., 190 F.R.D. 271, 278 (D. Mass. 2000)).

For example, in a case alleging that the defendant violated the Fair and Accurate Credit Transactions Act by printing more than the last five digits of customers’ credit or debit card numbers or expiration dates on receipts, the court found the class sufficiently definite because it was objectively defined: Class members “were either provided a receipt or they were not.” Tchoboian v. Parking Concepts, 2009 U.S. Dist. LEXIS 62122, at *14 (C.D. Cal. July 16, 2009). Despite the possibility “that there may be some difficulty in ascertaining the class,” class members could self-identify after notice. Id. at *13–14. Indeed, as the California Court of Appeal recognized, “[t]he necessity for class members to prove their own damages does not mean the class is not ascertainable.” Clothesrigger, Inc. v. GTE Corp., 191 Cal. App. 3d 605, 617 (1987); see, e.g., Parkinson v. Hyundai Motor Am., 258 F.R.D. 580, 594 (C.D. Cal. 2008).

Nor does the requirement that class members’ identities must ultimately be determined mean that there must be a simple or foolproof way of identifying class members post-judgment. Although defendants’ records often contain a complete list of class members—for example, a list of persons employed during the class period or a list of those who purchased a defendant’s product—this type of documentary proof is not required to conclude that it will be administratively feasible to identify class members. Class cases can pose unique administrative challenges; yet, “in identifying members, time and labor alone, even where significant, is not a reason to forego proceeding via a class.” Kendrick v. Standard Fire Ins. Co., 2010 U.S. Dist. LEXIS 135694, at *37 (E.D. Ky. Sept. 30, 2010) (citing Mitchell-Tracey v. United Gen. Title Ins. Co., 237 F.R.D. 551, 560 (D. Md. 2006)). The need to review files manually is not dispositive. If it were, defendants sued for wrongful conduct “could escape class-wide review due solely to the size of their businesses or the manner in which their business records were maintained.” Kendrick, 2010 U.S. Dist. LEXIS 135694, at *38–39.

Taken to an extreme, permitting defendants to evade class actions based on their own insufficient records would give companies an incentive to engage in shoddy record-keeping or even to destroy their files. However, in Marcus v. BMW of North America, LLC, 687 F.3d 583, 592–94 (3d Cir. 2012), the Third Circuit found that the proposed class—of owners and lessees of new BMWs purchased or leased in New Jersey with certain “run-flat tires” that had gone flat or been replaced—was not ascertainable, largely because BMW did not maintain records that could clearly identify class members. By contrast, in another case in which the defendant did not possess complete records (regarding sales of an allegedly defective product), the district court found an ascertainable class because a combination of “sales records, publication of notice, and verification by photograph” could identify class members; and the Seventh Circuit affirmed the class certification order. Saltzman v. Pella Corp., 257 F.R.D. 471, 476 (N.D. Ill. 2009), aff’d, 606 F.3d 391 (7th Cir. 2010).

In other cases in which defendants lacked records, courts reasoned that class members’ ability to identify themselves by reading the class definition and locating their own records was sufficient to defeat ascertainability challenges. See, e.g., Tchoboian, 2009 U.S. Dist. LEXIS 62122, at *13–14; Clothesrigger, 191 Cal. App. 3d at 617; Parkinson, 258 F.R.D. at 594 (“Because the proposed class definition allows prospective plaintiffs to determine whether they are class members with a potential right to recover, the defined class is sufficiently ascertainable.”); but see In re Teflon Prods. Liab. Litig., 254 F.R.D. 354, 361–63 (S.D. Iowa 2008) (finding that three proposed subclasses could not be ascertained; citing class representatives’ deposition testimony to deem as unreliable the class members’ memories and beliefs regarding their purchases and ownership of products that were the subject of alleged false representations).

As long as notice can issue and a claims process is administratively feasible (even if laborious), it is permissible to allow the exact contours of class membership to form after trial. A prime example is In re Methyl Tertiary Butyl Ether (MTBE) Products Liability Litigation, 241 F.R.D. 185 (S.D.N.Y. 2007), in which the proposed definition of a subclass consisted of owners of real property “in the vicinity” of a Maryland gas station whose underground tanks allegedly leaked a dangerous chemical. The court was not troubled by the lack of geographical precision in the class definition, explaining that the group of homeowners injured by the alleged leakage could not be determined until the jury “decide[d] how the MTBE spread after the underground storage tank leaked it into the ground.” Thus, although the term “vicinity” was imprecise, it was capable of objective determination “once a jury makes its findings of fact.” Id. at 195–96.

As MTBE shows, a class definition may be framed such that whether a given person belongs to a class or is entitled to relief cannot be determined until after fact-finding and judgment. It is the class, not all its constituent members, that must be ascertained at class certification. See id.

In this regard, courts should take care not to expand the definition of impermissible “fail-safe” classes from classes defined in terms of the defendant’s liability to classes defined in terms of those who sustained harm as a result of the defendant’s conduct. Determining the existence and parameters of causation is usually the province of the fact finder, and it is often impossible to define a class more precisely than those who sustained harm “as a result” of a challenged practice. Two recent cases illustrating this principle are Norwood v. Raytheon Co., 237 F.R.D. 581 (W.D. Tex. 2006), and In re Checking Account Overdraft Litigation, 275 F.R.D. 666 (S.D. Fla. 2011).

In Raytheon, the plaintiffs proposed a class consisting of all radar technicians, operators, and mechanics who worked with the defendants’ radar devices and who suffered illness or death as a result of exposure to ionizing radiation. The defendants claimed the class could not be ascertained because the plaintiffs failed to specify a method for determining causation. The court disagreed: “[I]t is not fatal to Plaintiffs’ proposed definition that it includes an element of causation without providing an objective method for determining causation.” 237 F.R.D. at 586. 

In Checking Account Overdraft, a defendant bank likewise contended that the proposed class could not be ascertained because it was defined as including account holders who “incurred an overdraft fee as a result of Union Bank’s practice of sequencing debit card transactions from highest to lowest.” In rejecting this argument and certifying the class, the court noted that the plaintiffs’ expert could identify which bank customers “incurred additional overdraft fees as a result of high-to-low re-sequencing of debit transactions by comparing that re-sequencing” to alternative debit posting scenarios, whereas “the finder of fact will ultimately have to decide which of these scenarios is most consistent with Union’s duty of good faith and fair dealing[.]” 275 F.R.D. at 672–73.

Unlike with a fail-safe class, determining class membership did not require any finding that the challenged conduct—posting instantaneous debit transactions from highest to lowest dollar amount—was unlawful. Moreover, the class as certified encompassed all customers who incurred overdraft fees under the bank’s high-to-low method that would not have been assessed under the bank’s prior method of posting debits from low to high; this was the broadest possible group of customers who could claim to have incurred additional fees because of the re-sequencing. Therefore, a final judgment in the bank’s favor, or for the class but relying on a chronological posting comparison, would preclude subsequent claims of all class members.

Just as a court or jury’s fact-finding may refine who is in the class and who is entitled to recovery, so may a claims administration process fulfill the same tasks. And there is no need for a plaintiff to show that all of the objectively defined class members will share in any eventual recovery, as illustrated by the decision in In re Chocolate Confectionary Antitrust Litigation, 2012 U.S. Dist. LEXIS 174681 (M.D. Pa. Dec. 7, 2012). In certifying a class of direct purchasers of chocolate candy bars, the court agreed with the plaintiffs that any class members unaffected by the alleged price-fixing conspiracy “may be removed at the claims phase of the litigation.” Id. at *99 n.30 (citing In re OSB Antitrust Litig., 2007 U.S. Dist. LEXIS 56584, at *26–27 (E.D. Pa. Aug. 3, 2007)).

It also goes without saying that the ascertainability requirement does not necessitate a showing at class certification that the challenged practice actually occurred, let alone that the defendant will be held liable. Courts rightly reject attempts by class action defendants to attack an otherwise objectively defined class on the grounds that they did not engage in the alleged conduct. For instance, the plaintiffs in Taylor v. Housing Authority of New Haven, 257 F.R.D. 23 (D. Conn. 2009), claimed the defendants discriminated against them by failing to provide accommodations for their disabilities as required by federal law. According to the defendants, the proposed class of handicapped public housing residents could not be ascertained because the Housing Authority of New Haven had “no policy discriminating against Disabled Section 8 Households” and encounters with Housing Authority staff accounted for the treatment of various residents. The court disposed of this argument as premature, noting that it “misses a critical distinction between considerations of ascertainability and the merits.” Questions about what the Housing Authority’s “policy or practice is and whether it is lawful are the core merits questions of this lawsuit, and they need not be resolved to determine the ascertainability of the class.” Id. at 28–29 (emphasis in original). After trial, however, the court decertified the class based on extensive findings in favor of the defendants. See Taylor v. Housing Auth. of New Haven, 267 F.R.D. 36, 62–66 (D. Conn. 2010).

Like the court in Taylor, the First Circuit in Matamoros v. Starbucks Corp., 699 F.3d 129 (1st Cir. 2012), wasted little time in dismissing a merits challenge disguised as an ascertainability argument. The claim there was that Starbucks violated a Massachusetts law by allowing shift supervisors to share pooled tips with the baristas they managed. Starbucks appealed the district court’s judgment for the plaintiff class, together with the incorporated order certifying a class of individuals who were employed in Massachusetts as Starbucks baristas. Starbucks asserted on appeal that the class was not ascertainable because certain baristas performed managerial tasks that rendered them ineligible to receive tips under the interpretation of Massachusetts law that supported the district court’s class-wide damages award. “This argument trenches on the frivolous,” the First Circuit declared, as the class was objectively defined by reference to job titles. Id. at 139.

Similarly, the settlement class in Union Asset Management Holding A.G. v. Dell, Inc., 669 F.3d 632 (5th Cir. 2012), was adequately and objectively defined. It consisted of “[a]ll persons who purchased or otherwise acquired the common stock of Dell Inc., directly or beneficially, between May 16, 2002 and September 8, 2006, inclusive, and who were damaged thereby.” A group of class members objected to the settlement and then appealed the district court’s order approving it; they contended that the “damaged thereby” phrase in the class definition would require “mini-trials” to ascertain whether Dell’s alleged securities fraud harmed each individual claimant. But the Fifth Circuit applied common sense to find the challenged phrase “superfluous.” Id. (citing In re Initial Pub. Offering Sec. Litig., 671 F. Supp. 2d 467, 492 (S.D.N.Y. 2009)). The court found the definition to be “a mechanical and objective standard, in no way an individualized causal determination on the merits,” as Dell shareholders incurred the alleged harm “just by holding stock, and a quick look at the trading records is all that [was] required to determine whether someone did so.”

Avoiding Overly Broad Applications

In conclusion, overbroad application of the ascertainability requirement may raise unwarranted barriers to certification of consumer class actions, a category of cases the Supreme Court has deemed vital to the purpose of Rule 23: In enacting Rule 23(b)(3), “the Advisory Committee had dominantly in mind vindication of ‘the rights of groups of people who individually would be without effective strength to bring their opponents into court at all.’” Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 617 (1997) (citation omitted); see also Phillips Petroleum Co. v. Shutts, 472 U.S. 797, 809 (1985); Deposit Guar. Nat’l Bank, Jackson, Miss. v. Roper, 445 U.S. 326, 339 (1980).

Given the strong policy favoring vindication of such rights, a court faced with a poorly drafted class definition, especially in a consumer case, should first consider whether a better definition can be crafted before analyzing whether class certification should be denied on ascertainability grounds. And the court should take care not to interpret the ascertainability requirement to require that all individual class members be identified at the time of class certification or that the class will ultimately recover in the event of a judgment for plaintiffs. Those holdings would transform a useful, limited requirement into a steep barrier with the potential to reduce the deterrent effect of the class action device.

Daniel Seltz and Jordan Elias – March 18, 2013