ELEVENTH CIRCUIT REVERSES COURSE ON ASCERTAINABILITY: Here’s Why the Court’s Surprising Rejection of “Administrative Feasibility” Matters So Much in TCPAWorld
We all have one of those friends or siblings that goes through a phase where they just really need a lot of attention.
Usually, they were a mousy quiet type before but something happens–usually someone “less deserving” starts getting noticed– and then, all of a sudden, they’ve simply had enough and start acting out for attention.
I kind of feel like that’s happening in the Eleventh Circuit right now. The Court has come of age and is tired of being overshadowed by its sister circuit– it wants some attention and it wants it right now.
In its latest “notice me” move, an Eleventh Circuit panel just held that the “administrative feasibility” test on ascertainability, which richly permeates the certification case law in the jurisdiction, is no more. This is a pretty astounding move since the Eleventh Circuit Court of Appeals has repeatedly–albeit in unpublished decisions– articulated a robust ascertainability standard that has lead dozens of district courts to deny certification for want of a clear way to identify class members, including many TCPA cases.
Nonetheless, in Garret v. Domestic, Case No. 19-13242 (11th Cir. 2021) the Eleventh Circuit held yesterday that administrative feasibility is not actually an implied requirement of Rule 23 after all. Instead, all a plaintiff must do to assure a class is “ascertainable” is to define a class that can be theoretically (rather than actually) identified: a proposed class is ascertainable if it is adequately defined such that its membership is capable of determination.
Talk about a reversal. Arguably the Eleventh Circuit just went from the circuit with the strongest ascertainability requirement to the circuit with the weakest. And in a published decision to boot.
But let’s back up. What is “ascertainability” anyway and why does the loss of the “adminsitrative feasibility” test matter in TCPA class actions?
Glad you asked.
To certify a class you have to be able to identify who is in the class. In broadest terms, this is what is meant by ascertainability: you have to be able to ascertain who is in a class as defined.
That seems like a simple proposition but there’s actually a bit of nuance to it. In a TCPA class action, for instance, there are generally three big pieces to ascertainability:
Is the class defined using objective criteria or is it defined using the merits of the claim?
Can class member identifiers actually be located in the defendant’s records?
Can class member names and addresses actually be found using those identifiers?
In the old days (like Monday) a Plaintiff would have to show all three things in order to certify a TCPA class action in the Eleventh Circuit. As of today, however, a Plaintiff may only have to demonstrate the first one to certify a class (the formulation is still a bit fuzzy and consideration of administrative feasibility is now backloaded into the “superiority” element of Rule 23(b)(3) but we’ll get to that in a second.)
But let’s walk through this.
First, even under Garrett the class still must be clearly and precisely defined. “Capable of identification” is a pretty weak standard but the word “capable” in this context is generally applied to mean “based upon objective criteria” as opposed to based upon merits inquiries. So, for instance, a class definition based upon whether class members “consented to receive calls” would not be ascertainable even under Garrett–consent is a merits issue that must be determined by a judge or a jury-whereas a class defined as “individuals that purportedly provided consent by visiting x website” would pass the first (and perhaps only) prong of ascertainability– whether someone visited a website is (at least theoretically) capable of identification.
But now let’s step from theory into reality.
How does one actually find the individuals that visited the website? Let’s imagine first a webform submission that requires only a phone number to be provided for more information, along with a check box. It is possible that those phone numbers reside in the Defendant’s records in a neat and tidy little box somewhere. In which case– prong 2 is easy to meet. But in many instances, webform submissions might be stored without specific reference to source, or the webform accepted by the consumer might change over time. So it becomes literally impossible (in reality) to identify class membership based upon data storage limitations– you simply cannot find identifiers in the available data elements that match the class description.
Prong 2 becomes even more compelling when the class is (improperly) defined based upon the presence or absence of consent. For instance, we commonly see classes defined based upon individuals who did not actually visit a website and provide consent even though a Defendant has a record of a lead from that website. In that instance, it is quite impossible for a Defendant to figure out which of its seemingly valid leads aren’t really valid.
A similar circumstance often arises with recycled or wrong number phone calls. If a caller buys 1,000 leads from a website it is almost certain that some unknown number of those lead forms involves a wrong number. Why? Because humans are limited clumsy beings that insert wrong numbers on webforms by accident. They are also petty obnoxious beings that insert wrong numbers on webforms on purpose. They are also fickle erratic beings that change phone numbers frequently. You get the point.
In all three of these circumstances, there is likely a definable class, but there is no way to reliably find data to identify the class members in the defendant’s records. Its a big guessing game. And certifying such a case is a fool’s errand.
That is what is meant by “administrative feasibility”– it's just not feasible from a practical standpoint to certify one of these cases because there’s just no way to find class members without abusive horrifying “turn a company inside out” tactics.
But under the Eleventh Circuit’s new ascertainability formulation, it simply doesn’t matter–well, sort of. Still getting to that.
There was one last prong to ascertainability which was–in my mind–always kind of silly to begin with. And that’s the “finding actual class members” prong. As described above, in many many instances finding class members in TCPA class actions is literally impossible. But in some cases, the data matching the class is easy enough to find and it is just a question of getting names and addresses for those class members.
For instance, in a single source TCPA fax class action –i.e. all numbers faxed came from a lead list supplied by a third party with a monolithic consent basis– the universe of phone numbers receiving faxes might be readily identifiable. (Sure, there’s always the question of which faxes were actually read and which actually went through and (now) which were sent to physical vs. virtual fax machines, but ignore all that for this example.) But that just means you have a list of numbers. You still need to take that list of numbers and use it to find class members.
There is a cottage industry of experts (both plaintiff and defense) that make a living supplying reports on whether phone numbers can be used to identify class members reliably. I, frankly, never like diving into prong 3 of ascertainability because prong 1 and prong 2 are usually far more fruitful. But theoretically, you could defeat certification based solely on the inability of transforming phone numbers into a usable class list. And, in the Eleventh Circuit, some district courts have denied certification on just that basis.
While prong 2 ascertaianability is still likely alive and well in a separate context–still getting to that– prong 3 ascertainability feels pretty dead. When Garrett talks about administrative feasibility “rarely if ever” being dispositive it seems to be pointing primarily at the inability to identify actual class members from available data. Word to the wise: don’t rely on prong 3 ascertainability anymore.
Ok so what about prong 2?
Garrett opens an ascertainability window, even as it closes a door. It notes–as it must– that Rule 23(b)(3) contains a “superiority” requirement that looks, in part, at whether a class action would be “manageable.” And management of a class action looks at the ability to find class members in a TCPA defendant’s records as much as anything else. As the court puts it:
A difficulty in identifying class members is a difficulty in managing a class action… If there appear to be unusually difficult manageability problems at this step, a district court ha[s] discretion to insist on details of the plaintiff’s plan for notifying the class and managing the action. And it has discretion to decertify a certified class that turns out to be unmanageable.
This is the part of Garrett that TCPA defendants must hold onto. When class member identification is impossible–as it often is in the TCPA context–that issue must be presented as part of the manageability requirement.
And don’t be shy about demanding a plan from the Plaintiff’s counsel on how they intend to identify class members and try the case. The Eleventh Circuit has a good body of district court authority requiring Plaintiffs to provide such plans, and Garrett seems to approve of, rather than roll back, this case law.
Remember also, of course, that the Eleventh Circuit has given TCPA defendants a bunch of huge and important positive rulings. So this one hurts a little less as a result. Nonetheless, Garrett is a huge ruling for TCPAWorld–and outside of it–and all class action litigators should take note.