TCPA cases often involve brightly-lit esoteric legal issues that attract law nerd moths like me from all corners.

Last week I explored one way intervention, and I had so much fun that this week I’d like to ruminate on the contours of a proper class definition.

Two recent cases frame this discussion. First, in Bais Yaakov v. ACT, Inc., Civ No. 12-40088, 2018 WL 5281746 (D. Mass Oct. 24, 2018) a court refused to certify a failsafe class, finding it uncertifiable and unamendable owing to predominance concerns. Second, in Eisenband v. Schumacher Auto., Inc., Case No. 18-cv-80911-BLOOM/Reinhart, 2018 U.S. Dist. LEXIS 181272 (S.D. Fl. Oct. 23, 2018) a court refused–with virtually no analysis–to strike an impermissible failsafe class at the pleadings stage. Both of these cases demonstrate that although failsafe classes are never certifiable–for the reasons described below–courts have routinely failed to exercise their gate keeper function to strike uncertrifiable classes  “at the point of minimum expenditure of time and money by the parties and the court.” See Bell Atlantic Corp., 550 U.S. at 558.

As I described last week, one way intervention is a class action doctrine built on assuring procedural fairness— specifically assuring that class members can’t opt in to good classes and avoid being bound by bad classes.  The rule against failsafe classes is built on an almost identical principles, although it manifests itself quite differently.

First, what is a failsafe class?

In simple language– it is as a class definition that is defined based–in whole or in part– upon the elements of a claim.

TCPA cases tend to generate classic failsafe classes by the barrel full. This is so because they are so easy to plead and assure an element of facial commonality between class members. Plus courts have proven reluctant–for reasons that defy explanation–to strike these facially inappropriate definitions at the pleadings stage, leaving TCPA plaintiffs with free reign to run amok with expensive class discovery.

Example, you demand?

Example you shall have:

All persons who received a call from Defendant using an ATDS or a pre-recorded voice message.

STOP

Whether or not a device is an ATDS or a pre-recorded voice are merits issues. This is so because the TCPA specifically uses those terms to determine an element of a claim. Anytime a class action complaint defines a class based upon the use of an “ATDS”, therefore, the class is instantly a failsafe.

Similarly:

All persons called using Defendant’s Super Duper dialer without express consent.

STOP

Here the ATDS component is no longer failsafe–the class looks to the Defendant’s use of a specific dialer as the criteria for class membership–but the phrase “without express consent” converts the definition into a failsafe. Again, whether or not a call was made “without express consent” is plainly a merits issue as a class member’s claim will likely rise or fall on whether or not consent was granted.

Ok. So now that we know what a failsafe class is, why is it so bad?

Two reasons. First, class definitions needs to be precisely defined using objective criteria. That’s the law in every single circuit. But district courts seem to forget that when approaching faisafe class definitions. Probably because they seem to nifty and useful from a commonality perspective. After all–everyone that called without consent will plainly be in a common position vis consent, and that’s a crucial merits issue! But its exactly because the definition is based upon the merits that the class is not precisely defined based upon facts. So a failsafe class can never meet the requirement of definite pleading of a class definition required everywhere.

So the first reason is killer and dead on, but its also a little technical. The second reason is all about fairness. Just like with one-way intervention, failsafe classes create an unfair imbalance between defendants and class members respecting the determination of substantive issues in the litigation. Specifically, everyone in the class will have already won on a crucial merits issue. And if the class loses on that merits issue the class members are not bound by the result–because they were never class members to begin with.

So let’s take the first example above. If that class is certified and the defendant uses a dialer that was determined not to be an ATDS in that case, the unnamed class member would be free to turn around and sue that same defendant regarding those very same calls. Why? Because the class member isn’t a class member. The determination that the dialer was not an ATDS converted the would-be class member into a mere bystander–the class only includes people called using an ATDS. The would be class member was determined not to have been called by an ATDS. So he or she is not bound by the result of the case by virtue of res juidicata–they were never technically parties to the case at all. Instead the class members would be impacted by the judgment only by virtue of the “slender reed of stare decisis.”

Pretty unfair right?

How about example number two? This one is even worse. After going through the massive and expensive catastrophe of analyzing class data to try to find people that Plaintiff’s expert contends did not consent based upon the defendant’s calling practices and campaign data–this is what always happens here folks, trust me–the case is certified for some reason and the parties are put to their proof. The defendant is entitled by the constitution to cross-examine class members about their purported lack of consent. Defendant attempts to call 1,000 would be class members to the stand to cross examine them regarding the circumstances leading up to the call and to introduce 1,000 call records to discredit their testimony that they did not consent. The court–refusing to recognize that these individualized issues should surely prevent certification or justify de-certification– limits the Defendant to but 40 mini-trials. Defendant conclusively proves that 35 of those individuals actually did consent. The court finds that those individuals consented–and instantly they are no longer parties to the case bound by the judgment because they do not fit within the class definition.

Insane right?

The second example demonstrates both why failsafe classes are so unfair but also why defining classes using objective criteria is absolutely essential. In order to prevent minitrials there must be a common set of facts that in and of themselves are dispositive of a merits issue “in one stroke.”  (That’s what the U.S. Supreme Court said in Dukes after all.) Otherwise the Defendant is hamstrung presenting evidence of a nuanced and varying defense across an entire certified class that may include millions of individualized stories.

Obviously such a class should never be certified as much because of a lack of administrative feasability as much as due to fairness concerns–but courts seem to get confused between “feasability” for ascertainability purposes and “feasability” for predominance purposes. The former is growing increasingly irrelevant to certification analysis, but the latter is absolutely crucial in Rule 23(b)(3) classes and failsafe classes will never afford predominance–they will always mask individualized issues. Literally, always. So they should never be allowed past the pleadings stage.

And no, this isn’t some sort of defense “gotcha” strategy where consumers can never define a certifiable class. True, TCPA cases are extremely difficult to certify, but defining them is not that hard for any consumer that was actually harmed. Indeed, its a two step process in TCPA cases (uh oh–I’m about to give away some serious secret sauce here) and no consumer lawyer should ever file a class action until they have both pieces of information:

1. What system was used by Defendant?; and
2. What actually happened that resulted in the call to the consumer allegedly without consent?

The class should then be defined based upon the facts that form that underlying claim.

For instance, a class such as “All individuals that Defendant called using its super duper dialer where the number was found on a bathroom wall.”

This class is defined using objective criteria and not the elements of the TCPA. Yet these facts might (I said might) form a common result on a merits issue resulting in disposition of some crticial issue in one stroke as Dukes required. So this is a properly defined and potentially certifiable class. And this is how TCPA classes should be defined–based on facts and not on the merits.

Why would I give away this advice?

Because leveraging uncertifiable failsafe classes allows plaintiff’s attorneys to make ridiculous discovery demands and drive up the costs of defense on callers to force a classwide settlement–and that has got to stop.

For instance, the “number off the bathroom wall Plaintiff” might be completely unique in his circumstance–a victim of some prankster or ne’er-do-well employee. Yet by defining a failsafe class definition he masks those inherent flaws in his ability to certify his class– probably no one else fits this scenario. Yet by ignoring the facts and focusing on the predicate right of recovery, the plaintiff makes it look like the class might be huge. Then he can seek discovery on every record of consent to call any customer no matter where the phone number came from. And courts will sometimes grant discovery demands like this! See Medina v. Enhanced Recovery Co., Ltd. Liab. Co., No. 15-14342-CIV, 2017 U.S. Dist. LEXIS 186651 (S.D. Fla. Nov. 9, 2017)(Overbroad TCPA class leaked past the pleadings stage resulting in an order compelling Defendant to produce every record of consent for every dialer call it made during the class period, to go along with every complaint made by any consumer.) This is so because the scope of discovery tracks the pleadings–the broader the complaint, the broader the scope of class discovery (at least in theory).

That’s a tragedy and not at all how class actions are supposed to work. Courts have been far too tolerant with consumer lawyers filing baseless class actions with no theory of generalized proof of commonality apparent at the pleadings stage, just hoping to back into some certification theory using abusive discovery as a tactic. I have seen this time and time again. Defendants foot the cost of defense–typically hundreds of thousands of dollars–and the consumer lawyers never have to pay for the huge cost and expense bills they rack up with no sound basis.

This is an awful racket and the rule against failsafe classes–coupled with the U.S. Supreme Court’s recognition that class defendants should not be subjected to extremely broad discovery until plausible facts establishing certifiability have been alleged–really ought to be sufficient to put a stop to it. But as Bais Yaakov and Eisenband demonstrate, courts are a long way from taking that edict seriously.