From Casetext: Smarter Legal Research

Morrow v. Navy Fed. Credit Union

United States District Court, Eastern District of Virginia
Apr 6, 2023
1:21-cv-0722 (MSN/LRV) (E.D. Va. Apr. 6, 2023)

Opinion

1:21-cv-0722 (MSN/LRV) 1:22-cv-0844 (MSN/LRV)

04-06-2023

SIOBHAN MORROW and TRACEE LE FLORE, individually and on behalf of all others similarly situated,, v. NAVY FEDERAL CREDIT UNION, Defendant. MARIA HART and TRACEE LE FLORE, individually and on behalf of all others similarly situated, Plaintiffs, v. NAVY FEDERAL CREDIT UNION, Defendant.


MEMORANDUM OPINION

Hon. Michael S. Nachmanoff United States District Judge

This matter comes before the Court on both Defendant's Motion to Exclude the Testimony of Arthur Olsen (Dkt. No. 79) and Plaintiffs' Motion for Class Certification (Dkt. No. 59). For both the reasons stated in open court and those that follow, the Court has qualified Arthur Olsen as an expert and certified Plaintiffs' class. See Dkt. No. 128.

I. BACKGROUND

The facts underlying this dispute are relatively straightforward. Plaintiffs are a putative class of accountholders suing Navy Federal Credit Union. In their Complaint, Plaintiffs allege that-because the credit union service agreement specifies that accountholders will only be charged International Service Assessment Fees (“ISA Fees”) for transactions made outside of the United States-Defendant Navy Federal breached its contractual obligations by charging those fees for transactions made while the accountholder was still in the country. Dkt. No. 1. However, on Defendant's motion, Judge O'Grady dismissed the case in October 2021, finding that the contract's language clearly and unambiguously precluded Plaintiffs' sought-after relief. See Dkt. No. 33. On appeal, the Fourth Circuit issued an opinion vacating Judge O'Grady's decision, holding that the terms of the contract were ambiguous and remanding the case back to this District for further proceedings (Dkt. No. 37 at 1, 13-14). The action was then reassigned to this Court.

Following remand, Defendant filed its Answer (Dkt. No. 49), and the matter proceeded through discovery. Currently before the Court, however, are Plaintiffs' Motion for Class Certification (Dkt. No. 59) and Defendant's Motion to Exclude the Testimony of Arthur Olsen (Dkt. No. 79). After reviewing the parties' briefs and holding oral argument, the Court issued oral rulings on both motions at its March 24 hearing. See Dkt. No. 127 (minute entry); Dkt. No. 129 (transcript of argument and oral pronouncement). The Court then issued a written order, indicating that this opinion would follow. Dkt. No. 128. On April 5, 2023, Defendant filed a Motion to Reconsider Class Certification. See Dkt. No. 141.

Defendant's reconsideration motion raises three purported deficiencies in the Court's March 24 Order: (1) that the Court failed to address Rule 23's requirements; (2) that the Court failed to define the class; and (3) that the Court failed to appoint class counsel. Dkt. No. 141. As noted from the bench and in the March 24 Order itself, however, the Court never intended the March 24 Order to stand alone. After rendering its oral ruling, the Court issued the written order only to make pellucidly clear that the 21-day clock for Defendant to produce the “National Data” had begun to run-notwithstanding the absence of the (then-forthcoming) written opinion. See Dkt. No. 97 (ordering Defendant to produce the National Data within 21 days of class certification). Because the Court addresses the parties' legal contentions today, the concerns raised in Defendant's motion are now moot.

II. DISCUSSION

At the heart of the issues is a disagreement about whether it is possible to identify a class of accountholders who were improperly charged ISA fees for transactions made while in the United States. Arguing that it is, Plaintiffs have presented the testimony of expert Arthur Olsen who asserts that he can (and has) developed an algorithm that would analyze Navy Federal's customer transaction data to identify the improper fee assessments through a systematic process of elimination. Challenging Olsen's “Proposed Method,” Defendant argues that Olsen's expertise lies only in analyzing existing data (as opposed to using existing data to make inferences about missing information), he is not qualified to vouch for the reliability of the Proposed Method. And because his testimony would have been the only evidentiary support for the Proposed Method, Defendant argues that Plaintiffs' class should not be certified because they failed to provide a reliable method of ascertaining members.

With that background, the Court recognizes that analyzing both the Motion to Strike and the Motion to Certify requires an understanding of Plaintiffs' “Proposed Method” of identifying which Navy Federal accountholders were incorrectly charge ISA Fees while in the United States. Accordingly, the Court will start there.

In their own words, Plaintiffs describe the Proposed Method as follows:

• Plaintiffs allege ISA Fees could not properly be charged on debit card transactions that occurred while the cardholder was physically in the U.S. That means any transactions that occurred while the debit card was physically present cannot possibly be an improper ISA Fee. Helpfully, the data has a flag indicating whether the card was physically present or not. Mr. Olsen proposes to remove “card present” transactions from his analysis altogether.
• This leaves all ISA Fees assessed where a card was not present (that is, the purchase took place via internet or phone or mobile application). Aware that certain accountholders could be present in a foreign country while they made such a purchase, Mr. Olsen proposes to exclude those transactions from his analysis. He does this with a simple “proximity” analysis: if the cardholder made an in-person, card present transaction in a foreign country within fourteen days of a card not present transaction, that person is reasonably understood to have been overseas when the card not present transactions occurred. In short, if there is evidence the person was outside of the U.S., all card not present transactions around that time will be presumed to have also occurred overseas-and therefore excluded from the analysis.
• This then leaves all ISA Fees where a card was not present and where there is no evidence the cardholder ever left the U.S. This is a reasonable way to identify damages and the analysis could end there.
• Mr. Olsen also proposes that-should the Court or the factfinder have any doubts about the validity of the foregoing analysis, an additional step or steps can take place to further validate the analysis. For instance, for further assurance, the body of ISA Fees can be cross-referenced against a list of merchants to eliminate from the analysis transactions with any merchant that requires physical presence. Additionally, for further assurance, the body of ISA Fees can be cross-referenced against the merchant location of other transactions that occurred close in time to the transaction at issue. These methods provide substantial protection to ensure the class is neither overinclusive or under-inclusive.
Dkt. No. 93 at 2-3.

“Card Not Present” data is not available through individual account statements for the months between January and August 2016. Plaintiffs propose two workarounds: (1) identify and include transactions that incurred ISA Fees within 14 days of other transactions in the United States or (2) use the VISA data to identify a “pool of merchants that have accepted payments in person in the foreign location of the merchant” and then “exclude ISA Fees that were assessed on transactions from those merchants for the January 7, 2016 to July 2016 time frame.” Dkt. No. 6-8 ¶¶ 31-32 (“Olsen Decl.”).

With that understanding, the Court will now turn to a discussion of the motions at issue.

A. Defendant's Motion to Strike

1. FRE 702 and Daubert

Under the Federal Rules of Evidence, a “witness who is qualified as an expert by knowledge, skill, experience, training, or education may testify in the form of an opinion or otherwise” if: (1) the expert's scientific, technical, or other specialized knowledge will help the trier of fact to understand the evidence or to determine a fact in issue; (2) the testimony is based on sufficient facts or data; (3) the testimony is the product of reliable principles and methods; and (4) the expert has reliably applied the principles and methods to the facts of the case. Fed.R.Evid. 702. A court's determination of whether expert testimony is admissible “is always a flexible one, and the court's conclusions necessarily amount to an exercise of broad discretion guided by the overarching criteria of relevance and reliability.” Oglesby v. GMC, 190 F.3d 244, 250 (4th Cir. 1999) (citing Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 589 (1993)); Westberry v. Gislaved Gummi AB, 178 F.3d 257, 261 (4th Cir. 1999) (“Rule 702 was intended to liberalize the introduction of relevant expert evidence.”). Further, “the court need not determine that the expert testimony a litigant seeks to offer into evidence is irrefutable or certainly correct.” Westberry, 178 F.3d at 261. Criticisms as to the inadequacies of an expert study or report, such as the failure to include certain variables, for example, “are more appropriately considered an objection to the weight of the evidence rather than its admissibility.” Rosenfeld v. Oceania Cruises, Inc., 654 F.3d 1190, 1193 (11th Cir. 2011) (citation omitted). Accordingly, “[o]nly if the expert's opinion is so fundamentally unsupported that it can offer no assistance to the jury must such testimony be excluded.Hartley v. Dillard's, Inc., 310 F.3d 1054, 1061 (8th Cir. 2002). Moreover, doubts regarding whether an expert's testimony will be useful should be resolved in favor of admissibility. Maryland Shall Issue, Inc. v. Hogan, No. CV ELH-16-3311, 2021 WL 3172273, at *3 (D. Md. July 27, 2021) (“In effect, a tie goes to the proponent of the expert evidence.”). Under these standards, courts rarely find expert testimony inadmissible: “[A] review of the caselaw after Daubert shows that the rejection of expert testimony is the exception rather than the rule.” Cook v. Rockwell Int'l Corp., 580 F.Supp.2d 1071, 1082-83 (D. Colo. 2006).

2. Analysis

At times, both parties have suggested that the primary purpose of Olsen's testimony is to vouch for the reliability of the Proposed Method. However, in the Court's view, such characterizations paint things with too broad a brush. It is true that Olsen is a “data and database expert” with no expertise or experience in predicting cardholder locations using proxies and that he lacks special knowledge on how often a cardholder chooses to draw from their account by presenting their debit card, or how long a typical international trip would last, or even the reliability of the ”Card Present” and “Card Not Present” designations. Olsen has, however, created an algorithm designed to generate a proxy for the missing location data-specifically, whether an accountholder was in the United States when they made the ISA-fee-incurring transaction. And it is that latter point that the Court finds particularly salient.

Olsen intends only to testify as to his ability to carry out the steps required by the proposed methodology. And while Defendant suggests that narrowing the scope of Olsen's testimony to just that renders him no more useful than any lay witness that can perform “basic calculations,” Dkt. No. 101 at 7, the Court concludes otherwise. As Plaintiffs point out, “Olsen has analyzed historical bank data from over 100 of the largest financial institutions in the United States”-including Defendant Navy Federal-“in order to determine whether damaged class members could be ascertained and to calculate individual damages from those class members.” Dkt. No. 93 at 7. In other words, Olsen may not have expertise in the subject matter of the underlying dispute, but that fact alone does not foreclose the relevance (and, thus, the admissibility) of his testimony because Olsen's expertise as a data analyst cannot be reasonably questioned. See Mott Dep. at 40:19-41:19 (“A: [C]ould Mr. Olsen add numbers or subtract or divide from a database? All day long. He's made a good living off of that. I admire that.”). Indeed, in Gutierrez, a court found that, although he lacked subject matter expertise, Olsen could “convincingly” devise algorithms and make programmatic assumptions that filled the unavoidable gaps in the plaintiffs' theory with “professional and careful” work. Gutierrez v. Wells Fargo, 730 F.Supp.2d 1080, 1138 (N.D. Cal. 2010) And, in TD Bank, a different court accepted Olsen's testimony, noting that he “adequately demonstrated” his ability to “mine the Bank's customer transactional data, determine [the relevant datapoints], and adjust these calculations based upon data parameters established by the Court.” In re TD Bank Overdraft Litigation, 325 F.R.D 136, 173 (D.S.C. 2018). This Court will rely on those assurances. Cf. Dkt. No. 93 at 8-10 n.3-6 (collecting cases where Olsen was qualified to provide similar testimony). Accordingly, Olsen is qualified to testify on a narrow topic regarding the feasibility of devising an algorithm that can perform the steps laid out in the Proposed Method to generate a list of accountholders that were likely to have been improperly assessed ISA Fees for transactions completed while in the United States. See Dkt. No. 128.

Admittedly, Olsen does not have a perfect track record, and his testimony has been rejected by some courts. See Kilpatrick v. HomeAway, 2020 WL 7680558, at *2 (W.D. Tex. April 21, 2020). However, those courts were not controlled by the Fourth Circuit's acknowledgment that “[t]he main purpose of Daubert exclusion is to prevent juries from being swayed by dubious scientific testimony.” See Nease v. Ford Motor Co., 848 F.3d 219, 231 (4th Cir. 2017) (citing Zurn Pex Plumbing Prod. Liab. Litig., 644 F.3d 604, 613 (8th Cir. 2011) (holding that a defendant's “desire for an exhaustive and conclusive Daubert inquiry before the completion of merits discovery cannot be reconciled with the inherently preliminary nature of pretrial evidentiary and class certification rulings”)).

B. Plaintiff's Motion to Certify

Plaintiffs requested this Court to certify a class of all current and former Navy Federal Credit Union accountholders who opened their account prior to March 1, 2022, that were charged an ISA Fee on a debit card transaction made while the accountholder was located in the United States between January 7, 2016, and March 24, 2023. See Dkt. No. 80-2. However, although the parties' briefing suggests otherwise, the certification of Plaintiffs' class does not necessarily rise and fall with the qualification of their expert. To grant the certification motion, the Court must carefully apply the requirements set forth in the Federal Rules of Civil Procedure. See Fed.R.Civ.P. 23.

1. Rule 23

Rule 23 generally governs class certification. Under that rule, one or more members of a class may sue as representative parties on behalf of all if they satisfy the following threshold requirements: (1) numerosity; (2) commonality; (3) typicality; (4) adequacy of representation, and (5) ascertainability. See Peters v. Aetna Inc., 2 F.4th 199, 241-42 (4th Cir. 2021) (citing Fed.R.Civ.P. 23(a)). Once those threshold requirements are met, putative class members must also satisfy Rule 23(b) by showing that either: (a) individual actions would risk inconsistent or non-dispositive judgments; (b) that they are seeking class-wide injunctive or declaratory relief; or (c) that there are common legal or factual questions that predominate over any other concerns affecting individual members. Gunnells v. Healthplan Servs., Inc., 348 F.3d 417, 423 (4th Cir. 2003) (citing Fed.R.Civ.P. 23(b)).

It is also important to note that, while the certification inquiry and merits inquiries may sometimes overlap, “Rule 23 grants courts no license to engage in free-ranging merits inquiries at the certification stage.” EQT Prod Co. v. Adair, 764 F.3d 347, 358 (4th Cir. 2014) (citing Amgen Inc. v. Conn. Ret. Plans & Trust Funds, 133 S.Ct. 1184, 1194-95 (2013)). The Fourth Circuit has also instructed district courts to “give Rule 23 a liberal rather than a restrictive construction, adopting a standard of flexibility in application which will in the particular case best serve the ends of justice for the affected parties and promote judicial efficiency.” Gunnells, 348 F.3d at 424 (quoting In re A.H Robins & Co., Inc., 880 F.2d 709, 740 (4th Cir. 1989)).

Finally, when deciding whether or not to certify a putative class, “[d]istrict courts have wide discretion,” and the Fourth Circuit has committed to reversing certification decisions only when that discretion is abused. Central Wesleyan Coll. v. W.R. Grace & Co., 6 F.3d 177, 185 (4th Cir. 1993) (“Central Wesleyan”) (affirming a district court's decision to conditionally certify a class, notwithstanding underlying uncertainty about class membership, individualized liability issues, and other manageability concerns).

2. Analysis

In their motion to certify this matter as a class action, Plaintiffs assert that they meet each of Rule 23's requirements, and although, many of those discussions do not impact the Court's ultimate ruling, it is worth briefly discussing them here.

Plaintiffs deal with the first four requirements in short order, asserting that they can show: (1) numerosity (because Defendant's actions affected an untold number of its accountholders spread across the globe); (2) commonality (because the legal question of whether Defendant's policy of indiscriminately assessing the ISA Fees anytime an accountholder transacted with a foreign merchant was a contractual breach is common to each of the putative class members); (3) typicality (because the representative Plaintiffs, like all other class members, were improperly charged ISA Fees under Defendant's policy); and (4) adequacy (because Plaintiffs' counsel is experienced in complex class action litigation). Dkt. No. 73 at 16-20. Plaintiff's also assert that they can show that 23(b)'s predominance requirement is met (because the breach-of-contract question is the central inquiry and would not be superseded by any individualized inquiries into the amount of damages). Dkt. No. 73 at 20-22. Notwithstanding Defendant's various objections, the Court will accept those assertions.

The Court, however, will more closely scrutinize Plaintiffs' arguments regarding Rule 23's final requirement: ascertainability. The Fourth Circuit has held that for class members to be “ascertainable,” they must be “readily identifiable” through “objective criteria.” EQT Prod. Co. v. Adair, 764 F.3d 347, 358 (4th Cir. 2014). And while they must demonstrate that they will be able to do so without extensive and individualized fact-finding, plaintiffs “need not be able to identify every class member at the time of certification.” Id. Instead, courts must only require that-“at some point”-there will be some “administratively feasible” method of determining whether a particular individual is a member. Krakauer v. Dish Network, LLC, 925 F.3d 643, 658 (4th Cir. 2019).

Plaintiffs argue that, because Defendant does not collect data on an accountholder's physical location at the time of a given transaction (nor could it), the Proposed Method would serve as an adequate means of determining class membership once Navy Federal's databases are turned over. And to support that assertion, Plaintiff points to Olsen's testimony as evidence of the legitimacy of their claims. Dkt. No. 73 at 17 (suggesting that the “logic behind Mr. Olsen's analysis” can “systematically exclude ISA Fees made by accountholders located abroad.”). But, as Defendant argues (and as the Court found above), there is nothing in Olsen's testimony that speaks to the methodology's ability to produce a proxy for an accountholder's location. Still, the Court is not convinced that the Proposed Method is doomed as a result. The relevant caselaw requires only that Plaintiffs provide the Court with “objective criteria” capable of answering the who's-in-and-who's-out question later in the litigation. Plaintiffs have met this burden.

Defendant's arguments to the contrary are made primarily through Stephen Mott, an expert in the debit-card-payment industry. For example, in his testimony, Mott highlights that the Proposed Method (in its then-existing form) would not account for “Card Not Present” charges from ride-sharing apps for trips in foreign cities or from tolls on foreign roads, which should be treated the same as “Card Present” transactions and serve as the anchor point for a new 14-day window. Dkt. No. 80 at 21-23 (citing Mott Rep. ¶¶ 81-82, 94, 99 (identifying transactions that could be miscategorized if additional cross-checks were not incorporated into the Proposed Method)). While such a critique cannot be overlooked, it can also be overstated. As Plaintiffs point out in their response to Defendant's Daubert motion (Dkt. No. 93), the transactions that would have been incorrectly sorted under the previous model of the Proposed Method accounted for an almost negligible percentage the total transactions. Dkt. No. 93. Moreover (and more importantly), Plaintiffs also maintain that the Proposed Method is designed to be flexible and leaves room for Defendant, the Court, or Plaintiffs themselves to incorporate new cross-checks to account for things like ride-sharing and toll roads. Dkt. No. 93 at 2.

The Court finds that there may be other holes in the logic undergirding the Proposed Method. But the Court is unwilling to deny certification based only on the possibility of imprecision. See TD Bank, 325 F.R.D at 173 (certifying class even though Olsen's precise calculations “[would] need to be refined” as the litigation went on). It is also worth pointing out that it does not appear there is an alternative way to determine which account holders may have been improperly charged ISA Fees. Thus, if there is to be a class action, Plaintiffs have no choice but to rely on proxies. Cf. Gunnells, 348 F.3d at 426 (noting that “the adjudication of [the] matter through a class action” was “superior to no adjudication of the matter at all.”) (citing 5 Moore's Federal Practice § 23.48[1] (1997)). And because the Court finds that it would be patently unfair to require Plaintiffs to present a perfect mechanism for identifying that proxy before receiving access to Defendant's full records, the Court will exercise its broad discretion to certify Plaintiffs' class. See Central Wesleyan, 6 F.3d 177 (recognizing the district court's “considerable discretion” in this area and upholding a conditional certification order in the face of a “daunting number of individual issues” necessary to establish liability); Gunnells, 348 F.3d 417 (4th Cir. 2003) (relying heavily on Central Wesleyan and upholding a district court's conditional certification order in the face of uncertainty about the plaintiffs' likelihood to maintain class status).

In its March 24 Order, the Court “conditionally granted” Plaintiffs' certification motion. Dkt. No. 128. To be clear, the Court's use of conditional language was intended only to underscore that defendant retains the right to seek decertification if the implementation of the Proposed Method ultimately fails to adequately identify class members. The Court finds that Plaintiff's class is fully certified, notwithstanding the possibility of later decertification.

In sum, the Court finds that Plaintiffs have met their burden to certify the class, but-if, at any point in the litigation, it becomes apparent that the now-adopted methodology is unable to accurately identify accountholders who should be members of the class and separate them from accountholders that should not or if the burden of managing that process begins to outweigh the benefits of allowing Plaintiffs to proceed as a class-this Court will not hesitate to decertify, either sua sponte or on Defendant's motion. For now, however, the Court finds that the Proposed Method provides “objective indicia” that identifies potential members of Plaintiffs' class. The class members are thus ascertainable, and unless the Court later decides otherwise, this matter will proceed as a class action.

* * *

Accordingly, pursuant to the Court's March 24 Order, Defendant is required to produce the National Data on or before Friday, April 14, 2023. At which point, Plaintiffs will promptly identify class members in accordance with the procedures laid out in the Proposed Method. In addition, it is also

ORDERED that the Court's March 24 Order is AMENDED to clarify that Plaintiffs' class is fully certified; it is further

ORDERED that Plaintiffs' class shall be defined as: “All current and former Navy Federal Credit Union accountholders who opened their account prior to March 1, 2022, that were charged an ISA Fee on a debit card transaction made while the accountholder was located in the United States between January 7, 2016, and March 24, 2023”; it is further

ORDERED that-because the Court finds that the requirements of Fed. R. Civ. P 23(c) are satisfied, see Dkt. No. 61 ¶¶ 22-31 (detailing the class-action experience of counsel)- KalielGold PLLC, Kopelowitz Ostrow Ferguson Weiselberg Gilbert, Van Winkle Law Firm, and Carlson Lynch LLP are APPOINTED as counsel for the class of Plaintiffs; and it is further

ORDERED that, in light of the issuance of today's order, Defendant's Motion to Reconsider Class Certification (Dkt. No. 141) is DENIED AS MOOT.


Summaries of

Morrow v. Navy Fed. Credit Union

United States District Court, Eastern District of Virginia
Apr 6, 2023
1:21-cv-0722 (MSN/LRV) (E.D. Va. Apr. 6, 2023)
Case details for

Morrow v. Navy Fed. Credit Union

Case Details

Full title:SIOBHAN MORROW and TRACEE LE FLORE, individually and on behalf of all…

Court:United States District Court, Eastern District of Virginia

Date published: Apr 6, 2023

Citations

1:21-cv-0722 (MSN/LRV) (E.D. Va. Apr. 6, 2023)