They Don’t Call It the Wild West for Nothing: The Ninth Circuit Reverses Denial of Class Certification because Trial Court Kept out Inadmissible EvidenceThe Ninth Circuit’s decision not to grant en banc rehearing in Sali v. Corona Regional Medical Center should all but guarantee that the issue of expert testimony at the class certification stage is heading to the Supreme Court.

Sali involved a declaration that was created by a paralegal at one of the firms representing the plaintiffs and then offered to establish typicality under Rule 23(a)(3). The paralegal purported to analyze various time entries, and offered opinion testimony that the named plaintiffs’ claims were typical of the class members’ claims. Everyone agreed that the declaration was not admissible. The district court kept the spreadsheet out and, because the plaintiffs had no other evidence of typicality, declined to certify the class. The plaintiffs appealed.

The Ninth Circuit reversed. Instead of engaging directly on the issue of whether the declaration was inadmissible, it decided that the district court should not have considered the question of admissibility at all, stating “[i]nadmissibility alone is not a proper basis to reject evidence in support of class certification.” In place of the bright-line rule of admissibility, the Ninth Circuit proposed a sliding scale where the district court should consider whether the evidence could conceivably be offered in admissible form at trial—but that at the class certification stage, admissibility problems merely go to the weight of the evidence, not whether it should be considered.

The Ninth Circuit frequently referred to the admissibility standard as “formalistic” or overly restrictive. For example, in noting that the paralegal declaration was the only evidence of typicality, the Ninth Circuit called the district court’s evidentiary standard “narrow” and said the remaining evidence “tells us nothing about the satisfaction of the typicality requirement”—which is why the district court denied class certification for lack of evidence of typicality.

This decision has far-reaching consequences.

  • First, the practice-point takeaway is that defendants in the Ninth Circuit cannot rely on evidentiary objections alone to avoid class certification. Instead, they must engage on the merits of the plaintiff’s evidence while simultaneously pointing out any admissibility problems that exist. Defendants must also consider making their own record affirmatively showing that the class cannot be certified.

 

  • Second, allowing inadmissible evidence tilts what are supposed to be the neutral scales of justice decidedly in favor of class certification. By expanding the relevant record to include matters that could never be admitted as evidence or as expert testimony, the Ninth Circuit’s rule all but invites plaintiffs to rely on unfounded theories, unsupportable conclusions, or incompetent witnesses. While it is hoped that district courts will continue to use their discretion to certify only such classes as are supported by sufficient evidence, the Ninth Circuit’s decision will of course lead to more classes being certified in that circuit, particularly when that court seems to believe that the Rules of Evidence and Rules of Civil Procedure are excessively “formalistic” and “narrow” and therefore to be applied sparingly.

 

  • Third, this issue is now likely to head to the United States Supreme Court because the circuits have irreconcilably split on this issue. The Ninth Circuit has adopted the loosest rule of admissibility, while the Eighth Circuit has also held evidence need not be admissible to be considered at class certification (In re Zurn Pex Plumbing Prods. Liab. Litig., 644 F.3d 604, 611 (8th Cir. 2011)). On the other side, the Third, Fifth, and Seventh Circuits have required district courts to decide admissibility questions at class certification—at least when it is critical to certification issues (In re Blood Reagents Antitrust Litig., 783 F.3d 183, 187 (3d Cir. 2015) — “We join certain of our sister courts to hold that a plaintiff cannot rely on challenged expert testimony, when critical to class certification, to demonstrate conformity with Rule 23 unless the plaintiff also demonstrates, and the trial court finds, that the expert testimony satisfies the standard set out in Daubert”; Unger v. Amedisys Inc., 401 F.3d 316, 325 (5th Cir. 2005) — “When a court considers class certification based on the fraud on the market theory, it must… base its ruling on admissible evidence”; Messner v. Northshore Univ. Health Sys., 669 F.3d 802, 812 (7th Cir. 2012) — requiring a district court to make “a conclusive ruling on any challenge to that expert’s qualifications or submissions before it may rule on a motion for class certification” when the expert’s testimony is “critical”). Unpublished opinions from the Sixth and Eleventh Circuits also require admissibility as a threshold (see In re Carpenter Co., 2014 WL 12809636, at *3 (6th Cir. 2014); Sher v. Raytheon Co., 419 F. App’x 887, 890 (11th Cir. 2011)). Given that the Ninth Circuit’s opinion candidly acknowledges that “[o]ther circuits have reach varying conclusions” on this issue and that the dissent from the denial of en banc rehearing notes that Sali “puts our court on the wrong side of a lopsided circuit split,” this circuit split calls for resolution. Indeed, the Ninth Circuit appears to have split with itself in Sali, given that a different panel of the court had previously held that evidence must be admissible to be considered at class certification in Ellis v. Costco Wholesale Corp., 657 F.3d 970 (9th Cir. 2011).

 

  • Fourth, when the time comes, we anticipate that the Supreme Court will reject the Ninth Circuit rule. Analagous dicta from Wal-Mart Stores, Inc v. Dukes already signaled which way the court is likely leaning: “The District Court concluded that Daubert did not apply to expert testimony at the certification stage of class-action proceedings. We doubt that is so….” Moreover, the Comcast decision turned on the evidentiary value of expert testimony admitted by the District Court. If the Supreme Court shared the Ninth Circuit’s view, it likely would not have attached such significance to the scope of the expert’s opinion in that case. Indeed, the Supreme Court has repeatedly stated that Rule 23 is not a mere pleading standard, yet the Ninth Circuit has effectively converted it to such.

 

  • Lastly, hiding beneath the larger issues in this case is a lesson about the importance of local rules. It is possible that the Ninth Circuit’s analysis in this case is driven by the Central District of California’s local rules that require a plaintiff to move for class certification within 90 days of commencing a putative class action. That compressed timeline does not allow for much discovery on the evidence supporting the class certification prerequisites, and virtually requires a truncated evidentiary record. While Rule 23 requires deciding class certification as soon as practicable, 90 days is too fast for most cases, particularly those of any complexity. Class certification can be dipositive of a case in practice and so should not be rushed. Had the local rules allowed adequate time for discovery, the District Court could have relied on a more mature record and applied evidentiary standards more akin to summary judgment.

First Circuit Restricts Class Certification of Classes Containing Uninjured PersonsIn recent years, courts have reached divergent conclusions about the circumstances in which a damages class containing uninjured persons can be certified. Although there is some room to debate what constitutes injury, it is well established that individual litigants who have not suffered any injury at all should not recover; after all, injury in fact is a bedrock element of standing, and it is well-settled that the purely procedural class action device cannot be used to expand existing substantive rights or create new substantive rights that did not otherwise exist. However, in class cases, there is considerable dispute among circuit courts as to when courts should deal with the presence of uninjured persons in the putative class, and how uninjured persons should factor into the class certification decision.

On the one hand, the Seventh and Ninth Circuits have held that the presence of some—but not “a great many”—uninjured persons in the putative class is not a bar to class certification. Instead, these courts have sanctioned the use of the post-judgment claims administration process to identify and remove uninjured class members. Neither court has yet attempted to explain exactly how many uninjured members constitute “a great many,” nor have they provided much clarity on exactly how to sort it all out at the end.

On the other hand, the majority of courts to consider the issue, including the Second, Fifth, Eighth, and D.C. Circuits, have broadly held that “no class may be certified that contains members lacking Article III standing,” and required that classes “be defined in such a way that anyone within it would have standing.” These courts have placed the burden squarely on plaintiffs to show that “they can prove, through common evidence, that all class members were in fact injured” by the alleged misconduct.

Previously, the First Circuit had attempted to carve out somewhat of a middle ground. In In re Nexium Antitrust Litigation, the court acknowledged that each absent class member must have suffered an injury in fact to recover, but nevertheless affirmed certification of a class that “probably” contained “a de minimis number of uninjured parties.” The court reasoned that defendants’ mere speculation that a few members of the class may have been uninjured was insufficient to rebut the plaintiffs’ showing that for the vast majority of class members, injury in fact could be proven by class-wide proof. The defendants also failed to show that it would be difficult to identify the handful of hypothetically uninjured persons.

Last month, however, the First Circuit came more in line with the majority view. In In re Asacol Antitrust Litigation, the court reversed certification of a consumer class that included a significant number (approximately 10 percent) of uninjured persons. In Asacol, the defendant manufacturer of a drug used to treat ulcerative colitis was accused of violating the Sherman Act by pulling Asacol from the market shortly before the drug’s patent was set to expire and replacing it with two similar, patent-protected drugs. The theory of liability was that by prematurely pulling Asacol from the market, the defendant had precluded entry of lower-cost, generic alternatives in the treatment of ulcerative colitis. The plaintiff class in Asacol comprised consumers who had purchased Asacol before it was taken off the market and who then had purchased at least one of the two replacement drugs after Asacol was taken off the market.

At the class certification stage, the evidence showed that approximately 10 percent of the putative class members would not have switched from Asacol to a generic even if the defendant had not adopted its allegedly anticompetitive strategy and generic versions of Asacol had been available. The district court concluded that persons who would not have switched to a generic did not suffer any injury from the alleged anticompetitive conduct. Yet, the district court certified a class containing those uninjured persons anyway, reasoning that the uninjured class members could be removed post-judgment during the claims administration process and that the total amount of the damages award could be reduced to account for the removal of uninjured class members.

In reversing certification, the First Circuit rejected the district court’s position that it would be appropriate to remove uninjured class members post-judgment during the claims administration process. As the court explained, to move the adjudication of injury-in-fact to the claims administration process would deprive the defendant of its fundamental Seventh Amendment right to a trial by jury on all of the elements of the plaintiffs’ causes of action. Moreover, the defendant had introduced substantial evidence showing that identifying the uninjured class members would require precisely the sort of individualized inquiries antithetical to both the predominance and superiority requirements of Rule 23(b)(3). As a result, the court reasoned that class certification would be inappropriate. The court distinguished its Nexium decision for two reasons: (1) unlike in Nexium, in Asacol a more than “de minimis” number of class members were uninjured; and (2) unlike in Nexium, in Asacol the defendant came forward with substantial evidence both of the number of uninjured persons and the difficulty of proving injury on a class-wide basis – and made clear that it intended to present that evidence at trial.

There are several important takeaways for class action defendants in courts that follow the Asacol rationale:

1. Asacol’s impact will likely be broader than antitrust cases. Although there is some language in the Asacol opinion that purports to limit its holding to antitrust or similar claims, in which “injury in fact” is an affirmative element of the cause of action, Asacol will likely have a much broader application. That is because Article III itself requires that every federal plaintiff have suffered an “injury in fact” to have standing to recover in federal court, and it is unlikely that the current Supreme Court will allow Rule 23 to serve as an end-run on that constitutional requirement.

2. Defendants should introduce evidence at class certification concerning the number of uninjured class members and the difficulty of identifying them. Asacol reinforces the importance to class action defendants to introduce substantial evidence at the class certification stage of the number of uninjured absent class members and the difficulties in proving injury on a class-wide basis. Although plaintiffs bear the burden of proving that the requirements of Rule 23 have been met, as a practical matter, defendants will likely have to level a genuine challenge to allegations of injury in fact and to the plaintiffs’ ability to prove injury on a class-wide basis to defeat class certification because of the presence of uninjured class members. Speculation or unsubstantiated theories that the class may contain uninjured persons probably won’t cut it. Expert testimony will often be required to meet this practical burden.

3. Consider challenging certification even where the number of uninjured class members seems de minimis. One of the stated bases for the court’s decision in Asacol was the fact that the number of uninjured class members was not de minimis. As the court explained, it would have had far fewer concerns if the class had contained just a few uninjured members who could have easily been “picked off” by the defendant “in a manageable, individualized process at or before trial.” This seems to us to be a point of internal inconsistency in the court’s rationale. The Seventh Amendment does not have a “de minimis number of litigants” exception, yet the First Circuit’s approach arguably still seems to.

Assuming the de minimis exception continues to survive to some degree, this does not necessarily mean that defendants should not challenge certification where it appears that only a de minimis number of class members were uninjured. Most importantly, to the extent a court adopts a procedure that contemplates identifying and removing the de minimis number of uninjured class members post-judgment, we believe, as the First Circuit seemed to also conclude, that such a procedure would violate the Seventh Amendment, the Rules Enabling Act, and Article III, none of which have a de minimis exception. We think it likely that the Supreme Court will soon make that clear, perhaps even in a petition for certiorari in this very case should one be filed.  Therefore, we recommend persistently raising and preserving the argument that certification of a damages class is inappropriate unless each and every class member has suffered injury. Unless a class defendant makes that argument, it may be deemed to have forfeited its Seventh Amendment rights, given that those rights can always be waived.

In addition, even if the court attempts to mitigate Seventh Amendment concerns by adopting a process that would allow the jury to identify the uninjured persons at trial, that still does not necessarily mean that class certification would be appropriate. On the contrary, if the process by which uninjured persons are identified would require highly individualized—and likely unmanageable—factual inquiries, then Rule 23(b)(3)’s predominance and superiority requirements might be defeated. That arguably was the case in Asacol, where injury turned on an individualized, subjective inquiry of each putative class member’s purchasing decisions.