DEEP DIVE: New TCPA Certification Decision Highlights Unpredictability in Rule 23 Analysis Used By Courts in TCPA Cases
As I have said many times, TCPA class action defense is some of the most complex litigation out there. And with billions on the line, this line of work is not for the faint of heart or the green of horn. And when it comes to opposing certification a Defendant must be prepared for anything.
And I do mean anything.
The order granting certification in Abboud v. Agentra, Civil Action No. 3:19-CV-00120-X, 2020 U.S. Dist. LEXIS 168142 (N.D. Tx. Sept. 14, 2020) is about the most troubling and unlikely Rule 23 analysis I’ve ever encountered. And while I don’t want to be unnecessarily harsh on the analysis afforded by a sitting federal judge-that always seems to come back to haunt you—the analysis here really needs to be discussed with some frankness for educational purposes.
This is a bit of a longer post–a deep dive, as the title implies– so make sure you have 10 or so good minutes to spend with me.
For the uninitiated, to certify a case Plaintiff must prove—using evidence—that: i) a proposed class is sufficiently numerous to make joinder impracticable; ii) he has suffered a harm akin to that suffered by the class; ii) that common issues exist between class members that are likely to lead to common answers pertinent to liability on common proof; and iv) that the attorneys representing the class are qualified to do so. For a damages class action—which all good TCPA cases are—a Plaintiff must prove, again using evidence, that common issues actually predominate over individualized issues impacting the claims of class members. In most jurisdictions, the class members must also be ascertainable—which means there has to actually be a way to find folks in the class without individualized effort. Finally, there’s a squishy “superiority” requirement that imbues a level of discretion into the analysis.
TCPA cases are sometimes certifiable—especially where all of the numbers called come from a single source—but most often cannot be certified because there will be individualized issues surrounding consent, as the Fifth Circuit Court of Appeals first recognized way back in 2008. See Gene & Gene, LLC v. BioPay, LLC, 541 F.3d 318, 326 (5th Cir. 2008). So, for instance, if a Plaintiff claims he or she received a call after revoking consent, a class cannot ordinarily be certified because the only way to determine whether a person is similarly situated to the Plaintiff would be to review all of the facts and circumstances surrounding each purported revocation of consent by a class member to assess their claim. That is an individual review, not a common review. So the case cannot be certified. On the other hand, if the Plaintiff can show that everyone in the class provided consent in the same (purportedly deficient) manner—or did not provide consent at all—using common proof, then the case is certifiable because there is no individual inquiry required—the claims of everyone in the class can be tried at once.
Abboud breaks that mold rather entirely. Here are the classes that were certified:
Text Class: The text class is defined as any person in the United States who, from January 15, 2015 to the date notice is sent to the class, received one or more text messages from Agentra in the same manner as Abboud.
Agent Class: any person in the United States who, from January 15, 2015 to the date notice is sent to the class, was called on their cellular telephone by Agentra to promote insurance products using the same equipment as was used to call Abboud.
Even a quick review of the definitions demonstrate that these classes could never be properly certified unless the Defendant had absolutely no evidence of consent. The reason is simple: the definitions do not take into account that calls to class members might be perfectly legal depending on how the caller obtained the numbers at issue. So not everyone in the class is going to be in the same position vis their TCPA claim; some might have valid claims. Some might not. So there is no core commonality holding the classes together.
Yet the Abboud court certified these classes, which should never have happened.
So what went wrong? Let’s take a look.
The Court first addresses ascertainability. The analysis here isn’t great but courts often fumble with the amorphous rules on ascertainability. The Court finds that class members are ascertainable using the following process:
potential class members could present phone bills, or other forms of proof, showing that phone numbers used by Agentra called their phone. If the potential class members then signed affidavits saying they had never given Agentra permission to solicit them, and Agentra could not offer proof to the contrary, it would objectively give rise to the likelihood that the potential class members did not consent to the call.
So what’s wrong with that? Well, pretty much everything.
In the first place, the analysis presented would identify class members with valid claims. But the ascertainability analysis is supposed to look at how to identify members of the classes. Period.
Here the classes are defined as everyone that received a call or text in the same manner or from the same equipment as the Plaintiff. Finding class members that did not consent is, therefore, not necessary to establish ascertainability. So while the failure to take consent into account in the class definition is fatal to certification—more on that below—it actually makes ascertainability quite simple: all Plaintiff had to do was introduce evidence of the calls placed by Defendant using the same equipment that was used to call the Plaintiff.
But notice that the ascertainability analysis here doesn’t actually get to that key requirement. Just because someone received a call without consent does not actually address the key commonality requirement class membership—that the manner and equipment is the same. Yet the Court’s ascertainability analysis focuses solely on the fact that the message came from Agentra; it does not address or mention how the call or text was sent at all. So the ascertainability methodology focuses on the wrong thing—consent—and fails to address the key requirement that the text or call was made in the proper (common) manner. Eesh.
And, of course, even if the methodology approved by the Court were actually designed to ascertain class members—again, it isn’t—the process would not suffice under rule 23 anyway because it requires tremendous individualized inquiry. A declaration by declaration review followed by an assessment of evidence followed by… what? A trial? To determine whether a class member has been properly ascertained? Not the way this is supposed to work folks. Indeed, a proper showing of ascertainability requires proof that such individualized review will not be necessary—that’s the entire point of this implied requirement.
But, again, ascertainability can be tricky and I’ve seen lots of courts struggle with it and even approve similar errant approaches. If this was the only problem with the analysis this blog wouldn’t be so long. Unfortunately, however, the analysis in the Abboud decision only becomes more problematic and, thus, gentle reader, we still have a long road ahead together.
Turning to numerosity, a Plaintiff must prove–again using evidence—that there are more than 40 class members to certify a case. Here that showing should have been easily met—again all Plaintiff had to do was introduce evidence that Agentra sent more than 40 messages using the same equipment it used to send messages to Abboud. Unfortunately we do not know whether or not Plaintiff made that showing because the Court focuses solely on the allegations of the complaint:
Abboud alleges that there are thousands of class members, making joinder of all the parties impossible. This satisfies the Fifth Circuit’s numerosity requirement, which it suggested is a class of at least over forty individuals.
Oh no. This is a cardinal sin. A plaintiff cannot simply rely on allegations to prove numerosity—evidence is required. The mere allegation that the class is too numerous to make joinder practicable, by itself, is never sufficient to certify a case. Fleming v. Travenol Lab., Inc., 707 F.2d 829, 833 (5th Cir. 1983); Zeidman v. J. Ray McDermott & Co., Inc., 651 F.2d 1030, 1038 (5th Cir. 1981) (a plaintiff cannot rely on conclusory allegations that joinder is impracticable; he must show some evidence or reasonable estimate of the size of the class).
Yet the Abboud court does what it cannot do; it accepted allegations as evidence sufficient to certify the class.
All right. So that was a swing and a miss. But—as I alluded to above—the classes are broadly defined so there is a high likelihood that the classes are numerous so… no harm no foul?
But now things get very odd. Turning to commonality– again a requirement showing that all class member claims are subject some form of common proof on an issue critical to liability—the Court offers this brief analysis:
The members falling into the text and agent classes will all benefit if the Court finds that Agentra did use an automatic telephone dialing system or willfully acted to contact individuals without their consent or against their wishes.
Again the classes, as defined, are simply everyone that received a call using the same equipment or in the same manner—whatever that means—that was used to communicate with Plaintiff. So while a common finding of ATDS might be possible, there is simply zero opportunity for the court to assess consent across a class wide basis, much less on common proof. And while leaping to conclude—with no evidence identified or analysis presented—that willfulness can be determined on a classwide basis is startling (willfulness is virtually never subject to common proof) the final clause is absolutely astounding. Determining whether or not a text was received “against [the] wishes” of class members is quintessentially impossible in a properly certified class action—whenever subjective determinations are required (i.e. any review of a class member’s state of mind) a class is NOT certifiable because of the individualized inquiries needed to read the class member’s minds. Yet the Court here not only ignores this impregnable hurdle to certification—it affirmatively adopts the assessment of class member’s desire to receive the calls or texts as a common issue among class members.
This is truly baffling. The Court has taken something that should absolutely destroy certification and used it as a basis to certify the case.
But we move forward.
Next the Court gives very short shrift to the typicality and adequacy requirements, determining, in essence, that Plaintiff says she’ll participate in the case and she received the text messages at issue so she’s fine. As to her lawyers, the Court notes that their bios suggest they’ve done this before. Ok. Not great analysis but this can be forgiven. Again, courts are busy. A cut-to-the-chase approach on these elements is to be expected at times, even if the evidentiary support seems flimsy.
But now we get to the really really problematic part of the analysis. Predominance.
As I explained above, to certify a case common issues among class members must predominante over individualized issues to be tried between class members. This is discretionary and a bit of an “eye of the beholder” situation sometimes, but the necessary analysis is clear: the court must evaluate the common issues to be tried, on the one hand, and the individual issues to be tired, on the other.
Here is the entirety of the Abboud court’s predominance analysis:
The classes appear narrowly tailored enough so that if Agentra or its agents violated the Telephone Consumer Protection Act by using an automatic telephone dialing system, or by some other means, then that finding would trigger a verdict to eligible class members predominating any of Abboud’s individual claims.
What in the world?
Again, cutting through the fact that the class is not narrowly defined (and not limited to individuals that received calls without consent, the court weighs the wrong issue entirely—predominance does not look at whether class issues predominate over the Plaintiff’s individual case; predominance looks at whether common issues of proof predominate over individualized issues of proof as to the claims of class members.
So the Abboud court badly confuses the appropriate analysis and reaches a conclusion that simply does not address the requirements of Rule 23(b)(3). What the court needed to focus on was how the Plaintiff was going to actually prove TCPA violations using common evidence. But that critical issue was addressed solely in the ascertainability analysis and in a way that demonstrated, conclusively, that individualized issues of proof must predominate—there was simply not articulable basis for common proof on the issue of consent presented anywhere in the opinion. So the case should not have been certified. See Gene & Gene at 328.( “plaintiffs must advance a viable theory employing generalized proof to establish liability with respect to the class involved, and… district courts must only certify class actions filed under the TCPA when such a theory has been advanced.”)
Now, in fairness to the Court, I have not read the parties’ briefing here and it is perfectly plausible that the Court’s analysis ended up askew because the parties simply did not properly brief issues properly or otherwise confused matters needlessly. I’m sure matters were not laid out as succinctly as they are on TCPAworld. That seems particularly likely on the ascertainability analysis. And I believe deeply it is the Defense counsel’s obligation to break complicated issues down and make them simple and quickly digestible for a court.
But still, Abboud is a tough ruling and analysis like that applied in Abboud is difficult to predict or prepare for. At bottom, therefore, rulings like Abboud do make it more difficult to predict the outcome of straightforward cases, imbuing an additional level of uncertainty into a TCPAWorld already riddled with uncertainty.
Let’s hope a reconsideration motion or interlocutory appeal is on the horizon.
We’ll keep an eye on this.