CitiGroup, Inc. (“Citi”) just defeated certification in a massive class action in the consumer-friendly Northern District of Illinois in the case of Tomeo v. CitiGroup, Inc., No. 13-C-4046, 2018 U.S. Dist. LEXIS 166117 (N.D. Ill. Sept. 27, 2018).  The lawyers and experts for Citi did a clinical job of dismantling Plaintiff’s bid to certify potentially massive revocation and wrong number classes, both of which were ultimately felled based on the predominance of individualized issues of consent.  Below is our play-by-play of Citi’s huge win in this case.

The Putative Classes

Plaintiff sought to represent two classes in this case.  The first – a revocation class – defined as all persons who were called by Citi using its Aspect UIP or Genesys dialers between October 27, 2010 and November 30, 2014 where Citi’s business records indicate the person requested not to be called.  The second – a wrong number class – defined as all persons who were called by Citi using its Aspect UIP or Genesys dialers between October 27, 2010 and November 30, 2014 where Citi’s business records indicate that Citi was told that it had called the wrong number.

The size of these potential classes were massive.  Just based on a limited universe of dialer data spanning calls made from October 2013 to August 2015, there were 138.7 million outbound calls made by Citi with just its Aspect UIP dialer.  And even if a small fraction of these calls were made without consent, in the aggregate that spells hundreds of millions, if not over a billion in potential exposure to statutory damages.

The Records

As is the case with most (if not all) TCPA class actions, certification came down to whether consent could be determined by common proof based on Citi’s own business records.  According to the Court’s ruling, three different types of computerized records contained information relevant to whether Citi had the necessary consent to contact a phone number.

The first, the CitiLink loan system, which contains basic accountholder information and Individual Note Screens where Citi representatives record the details of their interactions with the accountholder.

The second, Citi’s default-related loan servicing system which contains similar information as CitiLink, but only for accounts in default.

The third, FileNet – the imaging system used by Citi to store digital versions of correspondence, contracts, and other paper documents relevant to accounts.

These documents contained not only records and documentation concerning consent, but also revocation of consent.  Specifically, Citi had three types of flags relating to TCPA compliance: a “Do Not Call” flag, which is placed when an accountholder asks Citi to stop contacting a specific number, a “Cease and Desist” and “Case 998” flag relating to requests to stop all communications, and subsequent investigations relating to those requests, and a “WRNG” flag for when a Citi representative is informed that Citi called the wrong number.

Within this universe of documents, records, and data, the experts fought out whether issues of consent could be determined by common proof.  And Citi’s experts ultimately won the day.

The Experts

In Plaintiff’s corner – Jeff Hansen.  And if you’re a regular reader of this site, or involved in TCPA class litigation, then you’ve heard of the guy.

In Citi’s corner – Ken Sponslor and John Taylor.  As you’ll see below, these experts packed a potent 1-2 punch.  Taylor broadly analyzed and opined on class data, while Sponslor drilled deep into the files for the four class representatives to show that there was no way to determine consent on a class-wide basis without a file-by-file review.

The Opinions

The Court started by addressing Citi’s challenge to the Hansen’s opinions.  Long story short, the Court threw out the portion of Hansen’s opinion relating to ATDS use.  It found that his opinions were not reliable because he never inspected or tested the system, and that his report contained impermissible legal conclusions that the dialers used by Citi were an ATDS.  Hmm…  That sounds familiar.

However, it found that the other, more critical aspects of Hansen’s report concerning class data passed the Daubert relevance and reliability tests, as they were “reasoned and founded on data”, and would assist the trier of fact because the report “identifies which calls and texts Citi made to cell phones using the Aspect and Genesys dialers, and which of those calls and texts occurred after the individual associated with that phone number requested that Citi stop contacting them or told Citi that it had the wrong number.”  And in this part of the report, Hansen opined that consent could be determined on a class-wide basis by reference to the codes contained in Citi’s records described above.

But not so fast.  In rebutting Hansen’s opinions, Citi presented “specific evidence establishing that a significant percentage of the putative class consented to receiving calls.”  Through Taylor’s opinion, Citi established that the existence of each of the three flags/codes in its records did not necessarily constitute individuals who did not consent to be called.

Taylor started with a sample of over 1,000 Individual Note Screens, and established that consumers frequently went back and forth between consenting, and revoking consent, so the only way to determine consent for each accountholder was to individually review Citi’s files.  Taylor pointed to data which showed that in 17% of the files reviewed by Hansen, the accountholder had later re-consented to contact.  And with respect to the “WRNG” (wrong number) codes, Taylor similarly found that 15% of the time the number actually belonged to a borrower (whether indicated by a later incoming call by the borrower from the same number, or where multiple phone numbers for an account were flagged as wrong number).

Then to back up Taylor’s opinions, Citi introduced the report of Sponslor, who had done a more exhaustive review of the four original named plaintiffs and found that Citi had consent to call three of four Plaintiffs.  And with respect to the fourth Plaintiff, Citi had obtained the phone number when it was provided by the Plaintiff’s mother.  Thus, Sponslor opined, it would be necessary to consider information in the files beyond the Individual Notes Screens to determine whether there was consent.

The Court agreed, finding that “[t]he level of potential error within both classes is significant.”

Plaintiff’s Response

In response to the Taylor/Sponslor opinions, Plaintiff argued that Hansen had provided a common way of addressing the issue of consent because “he could automatically search the Individual Note Screens for evidence that the accountholder re-consented to a phone call,” by devising a keyword search for “re-consent phrases”.  But the Court rejected the idea, finding that “Hansen was unable to execute this mass-search to show that it could be done,” and that in any event “there is no one word that Hansen could search to determine the consent status for each potential class member at the time of each potential violation.”

The Court found this approach to be “equally problematic” with respect to the wrong number class, and was not convinced that Hansen had developed a common way to isolate those who had already consented to calls.  This is due to a circumstance present in many wrong number debt-collection classes: many customers would falsely claim that Citi had called the wrong number to avoid calls, and if the customer provided a number belonging to another person – say a family member – an inquiry into that customer’s authority to provide consent to call that number would be required.

But above it all, the Court was left unpersuaded by Hansen’s proposals because he “ha[d] not actually executed any of the methods that he proposes for determining consent or actual wrong numbers,” and Plaintiff could not “establish predominance based on Hansen’s proposed, but unexecuted and unproven analysis.”

Key Findings by the Court

After reviewing, analyzing, and digesting the parties’ respective evidence and expert reports, the Court denied class certification because “individualized issues of consent predominate, and thus [Plaintiff] has not carried his burden under Rule 23(b)(3)”.  And in reaching its holding, the Court made five key findings:

1. It didn’t matter that consent was an affirmative defense versus an element of Plaintiff’s claim: “The need for individualized inquires with respect to an affirmative defense may still defeat the predominance requirement.”
2. The issue of consent eclipsed any other common issues because “most or even all class members will require this inquiry,” and the issue directly affected “whether [Citi] was liable for some violation of the law.”
3. Consent is inextricably intertwined with the primary issue of liability to the point where it predominates over the other common issues in the case, and other common issues would not “sufficiently drive the resolution of the litigation to convince the Court to look past the issue of consent.”
4. Although the Taylor report was based on a statistical sampling which found consent issues with 15-17% of the class, the purpose of those reports were to identify some flaws in the potential classes, not all the flaws.  The point being, is these samplings were enough to establish that that individualized inquiries into consent would be required.
5. As explained above, in the face of Citi’s evidence that it had consent for a significant percentage of the potential class members, Plaintiff failed to establish a way to determine consent on a classwide basis.

As we’ve seen before, weak showings of individualized issues of consent are bound to result in class certification.  Defeating certification in revocation or wrong number TCPA class actions instead requires a robust showing – supported by “specific evidence” – not only that members of the class consented, but that those issues of consent cannot be resolved on a class-wide basis without individualized inquiries.  Citi and its lawyers put on quite the clinic in this case on how that can – and should – be done in these sorts of cases.