Ruiz v. Citibank, N.A.

93 F. Supp. 3d 279, 2015 WL 1254820

• Court: District Court, S.D. New York
• Decided: March 19, 2015
• Docket: Nos. 10 Civ. 5950(KPF)(RLE), 10 Civ. 7304(KPF)(RLE)
• Status: Published

Opinion

OPINION AND ORDER

KATHERINE POLK FAILLA, District Judge.

Plaintiffs in this consolidated matter move to certify a class action against Defendant Citibank, N.A. on behalf of current and former personal bankers for unpaid overtime in violation of the New York Labor Law (the “NYLL”), the Illinois Minimum Wage Law, and the District of Columbia Minimum Wage Act Revision Act. Defendant Citibank moves to decertify a collective action previously certified under the Fair Labor Standards Act of 1938 (the “FLSA”). For the reasons set forth in this Opinion, Plaintiffs’ motion is denied and Defendant’s motion is granted.

BACKGROUND¹

A. Procedural Background

Plaintiff Digna Ruiz, a resident of New York, filed a complaint in this District on August 10, 2010, seeking to bring a nationwide collective action under the FLSA and a statewide class action under the NYLL. (Ruiz Compl. (Dkt. # 1) ¶¶ 3-4). . Ruiz alleged that Citibank had failed to compensate its personal bankers for overtime hours worked in violation of both laws. (Id,.). Fredrick Winfield, Zulma Muniz, James Steffensen, and Adoram Shen — residents of, respectively, Washington, D.C., Illinois, Virginia, and California — filed a complaint in this District on September 22, 2010, seeking to bring a nationwide collective action under the FLSA; statewide class actions under the labor laws of Washington, D.C., Illinois, and California;² and a nationwide class action under the Employee Retirement Income Security Act (“ERISA”), 29 U.S.C. § 1132(e)(1). (Win-field Compl. (Winfield Dkt. #1)). On October 26, 2010, the cases were designated as related and consolidated before the Honorable John G. Koeltl, United States District Judge. (Winfield Dkt. # 6).

Defendant Citibank moved in response to the Amended Winfield Complaint (Win-field Dkt. #30) to dismiss all claims and, further, to strike the requests for injunc-tive relief on standing grounds. (Winfield Dkt. # 33). Judge Koeltl granted the motion in part and denied it in part, dismissing the Winfield Plaintiffs’ ERISA claims. Winfield v. Citibank, N.A., 842 F.Supp.2d 560 (S.D.N.Y.2012). In addition, the Court, after limited discovery, granted conditional certification of the- FLSA collective action in both cases and authorized the issuance of notice to all personal bankers employed at Citibank within the relevant time period. Winfield v. Citibank, N.A., 843 F.Supp.2d 397 (S.D.N.Y.2012).³

During the class notice period, notice was sent to over 6,000 current and former Citibank personal bankers potentially eligible to join the FLSA collective action or the NYLL action, of whom 437 opted in to the FLSA collective action (including Plaintiff Ruiz). (Linthorst Decl. ¶¶2-3).⁴

Following the opt-in period, on October 18, 2013, the Honorable Ronald L. Ellis, United States Magistrate Judge, to whom the consolidated action has been assigned for general pretrial purposes, ordered that discovery proceed in a bifurcated fashion, focusing first on Plaintiffs’ anticipated motion for class certification of their state law claims and Defendant’s anticipated motion to decertify the FLSA collective action. (Dkt. # 123). In accordance with Plaintiffs’ request, and over Defendant’s objections, Magistrate Judge Ellis further ordered that discovery inquiries directed at Plaintiffs should be limited to a sample of the opt-in plaintiffs rather than every opt-in plaintiff. (Id.). The parties settled upon 30 representative opt-ins, with 15 chosen by Plaintiffs and 15 chosen by Defendants; for logistical reasons, only 23 such opt-ins were actually deposed (the “Sample Opt-ins”). (Linthorst Decl. ¶ 7). During the discovery period, the case was reassigned to the undersigned from Judge Koeltl. (Dkt. # 131).

Upon the conclusion of the certification-focused period of discovery, the parties filed the instant motions for certification of the state law classes under Rule 23 (Dkt. # 182) and decertification of the FLSA collective action (Dkt. # 178) on April 30, 2014. Oppositions to those motions were filed on May 30, 2014 (Dkt. # 188, 190), and replies were filed on June. 20, 2014 (Dkt. # 197, 200). Defendant’s sur-reply in further opposition to the motion for certification was filed on July 14, 2014 (Dkt. #204), and the briefing was complete with the filing of Plaintiffs’ sur-sur-reply in further support of the motion for certification on August 4, 2014. (Dkt. # 209). The Court now considers the motions.

B. Factual Background

1. The Parties

Defendant Citibank is a global bank with roughly 1,000 domestic branches in 13 states and the District of Columbia. (Dkt. # 46 Ex. 10 ¶ 7). Each branch is managed by a branch manager, and branch managers are supervised by roughly 65 area directors. (Id.). Each branch generally has a certain number of tellers, between one and ten personal bankers, and (depending on the size of the branch) other positions, such as assistant branch manager. (Id. ¶¶ 4-6).

Plaintiffs worked for Citibank as personal bankers during the relevant time period. The FLSA collective action consists of over 400 current or former personal bankers who have opted in- to the collective action; the putative New York class consists of over 2,000 current or former personal bankers who worked for Citibank in New York during the relevant period; the putative Illinois class consists of over 330 personal bankers; and the putative D.C. class consists of 16 personal bankers. (Tyner Decl. ¶ 3). '

2. Personal Banker Compensation

Citibank’s personal bankers are compensated on an hourly basis (Tyner Decl. Ex. 10), and are classified as non-exempt, overtime-eligible employees (id. Ex. 66). Personal bankers perform a number of client-related services, but their primary task is the sale of various banking services to current and potential clients. (Id. Ex. 10). Personal bankers have sales “hurdles,” requiring them to amass monthly sales credits equal to a percentage of their annual salary, and above which they may receive additional compensation of 15% of sales credits up to a certain threshold. (Lin-thorst Decl. Ex. 56-F (“2009 Individual Seller Plan Brochure”), G (“2010 Individual Plan Document”), H (“2011 Individual Plan Document”), I (“2012 Personal Banker Plan Document”), J (“2013 Individual Banker Plan Document”)).⁵

Failure to meet the sales goals is governed by the “Performance Management Progression” (or, in other years, the “Seller Corrective Action Process”), which sets escalating penalties for failure to meet goals (either the sales hurdle or a percentile rank within the seller class): an informal warning for missing the goals two out of three months; a performance improvement plan for missing the goals four out of six months; a final warning if performance does not improve on the performance improvement plan; and ultimately the possibility of termination. (Tyner Decl. Ex. 27, 28). Deposition testimony from Plaintiffs and Sample OptAIns suggests that penalties for failure to meet the sales targets were at the discretion of the branch manager, and unevenly imposed. (See Drago Dep. 167-68; Drews Dep. 114-15; Win-field Dep. 178). The evidence similarly suggests a striking lack of uniformity, both between personal bankers and among personal bankers from month to month, in how difficult it was to meet the sales targets while working 40 hours per week. (See Def. Decert. Br. 21-22 (collecting testimony)).

3. Citibank’s Timekeeping Policies and System

Citibank’s 2009 employee handbook contains several explicit instructions to employees that they are entitled to time-and-a-half overtime pay for hours worked beyond 40 hours; that although overtime must be approved in advance, overtime must be paid for time worked regardless of preapproval; that time must be recorded accurately; and that “ ‘off-the-clock’ work is strictly prohibited. Managers may not request or require ‘off-the-clock’ work.” (Linthorst Decl. 56-C, at 23-25). The handbook also informs employees that concerns about overtime pay may be raised with Citibank’s Human Resources Department, and that “[r]etaliation against any employee for raising a concern is strictly prohibited.” (Id. at 25). Citibank’s 2013 and 2011 employee handbooks contain similar language. (Id. Ex. 56-A, 56-B).

Personal bankers’ hours are tracked through Citibank’s North America Time & Attendance (“NATA”) System. (Linthorst Decl. Ex. 57 (Sumoela Decl.) ¶¶ 4-5). According to Citibank’s training materials, it was the primary responsibility of personal bankers to input their hours using the NATA system, but their branch managers would' review and approve the hours. (Id. Ex. A, B, C, D, E). The training materials nowhere indicate that branch managers should independently edit their employees’ hours, though they do 'indicate that it is a branch manager’s responsibility to “[Complete and submit the time record on behalf of their employee when they are unable to complete/submit their time record.” (Id. Ex. E). The deposition of Lori Sumoela, a Human Resources Risk Control Analyst at Citibank, indicates that branch managers could unilaterally edit the timesheets of personal bankers, though such edits would be logged and an audit report showing the changes could be accessed by both employees and managers. (Sumoela Dep. 54-56). In addition, a feature was added in 2010 that would automatically alert an employee if someone edited his or her timesheet. (Id. at 70).

Prior to 2011, either an employee or a manager could approve a timesheet to send it to Human Resources (Sumoela Dep. 54); in 2011, a feature was added requiring the employee to affirm that the timesheet was complete and accurate before it would be sent to Human Resources (Sumoela Decl. ¶¶ 11-13). If an employee declared that the timesheet was not accurate, a Human Resources representative would contact the employee (id. at ¶ 13; Tyner Decl. Ex. 132), though Plaintiffs suggest that such contact was primarily a device to pressure employees into selecting “I Agree” rather than to resolve concerns (PL Decert. Opp. 13-14).

In addition to its NATA system, Citibank tracked when employees logged onto and off of their work computers at the beginning and end of the workday. (See Tyner Decl. Ex. 124). An internal Citibank email suggests that managers were generally expected to review the audit logs to confirm the accuracy of timesheets, and that same email indicates that one branch manager’s comparison of the audit logs with timesheets suggested that personal bankers were working significant amounts of unreported overtime. (Id.). Citibank’s third-party security service also tracked the entry of alarm codes that were disarmed and armed at the beginning and end of every day, when the branch was opened and closed; some employees were given alarm codes, and each code was unique to an employee. (Wolter Dep. 42-45).

4. Citibank’s Efforts to Reduce Overtime Expenditures

During the relevant period, Citibank’s senior management sent out several directives to limit expenses, including among them overtime pay. (See, e.g., Tyner Decl. Ex. 70. (email identifying promotional items, stationery and supplies, travel and entertainment, and overtime pay as areas in need of expense management)). Even here, however, a common goal is difficult to discern: Different communiques set different targets, with several indicating a goal of zero overtime and others indicating nonzero targets. (Compare id. Ex. 67, 68, 72, 77 with id. Ex. 56 (25% reduction), 57 (50% reduction), 59 (75% reduction), 71 (goal of $11,000 per month for a branch)).

Some, but not all, of the directives to reduce overtime paid contained warnings that overtime incurred must be paid, and that the goal was to reduce overtime incurred. (See Tyner Decl. Ex. 67 (“That said, we absolutely MUST pay our PBs [Personal Bankers] for the time they were on the PB call last Thursday evening, even if they took the call from home. Be SURE that they record their hour (well actually minutes) that they were on that call.”), Ex. 79 (“Note — As always, we pay employees for the OT hours they work and will approve such hours via the standard process.”), Ex. 81 (“You may also have instances when OVERTIME is unavoidable. It is your responsibility to ensure your employees record their TIME WORKED accurately on their TIME-SHEETS.”)). Many of the emails also emphasize that overtime must be preap-proved, though these generally state that (in accordance with Citibank policy) the approval must come before the overtime is incurred, rather than suggesting that there be nonpayment of overtime that has already been incurred. (See, e.g., id. Ex. 80 (“Inform your Employees that ANY OVERTIME ... need[s] your approval before OVERTIME IS INCURRED.... Please check timesheets on NATA with your [Assistant Managers] EVERY WEEK and report any OVERTIME to Syed BEFORE it is incurred.”))- Other emails make reference to adhering to the “staffing model” and carefully scheduling shifts, indicating a focus on efficient scheduling as a mechanism for reducing unnecessary overtime. {See, e.g., id. Ex. 70 (“Adherence to the staffing model and limiting overtime are critical.”)).

The results of this push at the individual level were indeterminate. Citibank paid personal bankers a total of $2.39 million in overtime pay in 2009, $2.17 million in 2010, $2.07 million in 2011, $2.26 million in 2012, and $4.06 million in 2013. (Linthorst Decl. Ex. 48 (Hammond Decl.) ¶ 4). Perhaps more significantly, reviewing the payroll records of the Sample Opt-Ins, Citibank determined that every Sample Opt-In and named Plaintiff received overtime pay;⁶ as a group they received overtime pay in 45% of their pay periods during their employment at Citibank, ranging from 4% for Sample Opt-In Drews (2 out 57 pay periods) to 99% for Sample Opt-In Kosiba (75 out of 76 pay periods). (Hammond Decl. ¶¶ 7, 13, 20). And yet despite the documentary evidence of overtime pay for each of the Sample Opt-Ins, a number of them testified to being told by branch managers not to enter any timesheets reflecting more than 40 hours worked and to edit any such timesheets they had entered, and further testified that they understood these instructions to be the result of broader Citibank policy on overtime pay. {See Tyner Decl. Ex. 156 (collecting testimony)). Other Plaintiffs received overtime pay in nearly all periods, however, and could not recall any instances of entering inaccurate timesheets or having their time-sheets edited. {See Hammond Decl. ¶¶ 20 (Kosiba, overtime in 75 of 76 pay periods), 27 (Pikulin, overtime in 38 of 39 pay periods); Kosiba Dep. 151 (unable to remember ever failing to record overtime, or branch manager ever making changes to her knowledge); Pikulin Dep. 98, 127-28 (told to record all time, and unable to remember not recording all time or being paid for all time); Shen Dep. 228 (never threatened or disciplined for working overtime, and unable to remember being instructed to work off the clock or failing to enter overtime)).

DISCUSSION

A. Plaintiffs’ Motion to Certify the State Law Classes Under Rule 23 Is Denied

1. Applicable Law

a. Unpaid Overtime Claims Under New York, Illinois, and D.C. Law

By way of background, the FLSA requires all qualifying employers⁷ to pay employees no less than the minimum wage, and to compensate employees for hours worked in excess of 40 per work week at a rate not less than one-and-one-half times the regular rate of pay. 29 U.S.C. §§ 206(a)(1), 207(a)(1). Each of the relevant state statutes similarly specifies additional compensation for overtime work. The New York State Department of Labor has announced that an “employer shall pay an employee for overtime at a wage rate of one and one-half times the employee’s regular rate in the manner and methods provided in and subject to [the FLSA.]” N.Y. Comp.Codes R. & Regs. tit. 12, §§ 142-2.1, 142-2.2; see also N.Y. Lab. Law §§ 198(1— a), 663(1). The Illinois Minimum Wage Act provides that “no employer shall employ any of his employees for a workweek of more than 40 hours unless such employee receives compensation for his employment in excess of the hours above specified at a rate not less than 1 1/2 .times the regular rate at which he is employed.” 820 Ill. Comp. Stat. 105/4a(l). Finally, the District of Columbia Minimum Wage Act Revision Act provides that “[n]o employer shall employ any employee for a workweek that is longer than 40 hours, unless the employee receives compensation for employment in excess of 40 hours at a rate not less than 1 1/2 times the regular rate at which the employee is employed.” D.C.Code § 32-1003(c).

b. Class Certification Under Fed.R.Civ.P. 23

To meet the standard for certification of a class action under Federal Rule of Civil Procedure 23, Plaintiffs must establish that:

(1) the class is so numerous that joinder of all members is impracticable; (2) there are questions of law or fact common to the class; (3) the claims or defenses of the representative parties are typical of the claims or defenses of the class; and (4) the representative parties will fairly and adequately protect the interests of the class.

Fed.R.Civ.P. 23(a). In addition to these four requirements — often labeled numer-osity, commonality, typicality, and adequacy — a class must meet one of the three standards set forth by Rule 23(b). Here, Plaintiffs seek to certify three classes under Rule 23(b)(3), which requires that “the court finds that the questions of law or fact common to class members predominate over any questions affecting only individual members, and that a class action is superior to other available methods for fairly and efficiently adjudicating the controversy.”

“[A] district judge may not certify a class without making a ruling that each Rule 23 requirement is met,” In re Initial Pub. Offerings Sec. Litig., 471 F.3d 24, 27 (2d Cir.2006), and “[t]he party seeking certification must establish the Fed. R.Civ.P. 23 requirements by a preponderance of the 'evidence,” Pa. Pub. Sch. Emps.’ Ret. Sys. v. Morgan Stanley & Co., 772 F.3d 111, 119 (2d Cir.), as amended (Nov. 12, 2014). “Rule 23 does not set forth a mere pleading standard,” Wal-Mart Stores, Inc. v. Dukes, — U.S. -, 131 S.Ct. 2541, 2551, 180 L.Ed.2d 374 (2011), and courts must undertake a “rigorous analysis” to ensure that the requirements of Rule 23 have been satisfied, Gen. Tel. Co. of Sw. v. Falcon, 457 U.S. 147, 161, 102 S.Ct. 2364, 72 L.Ed.2d 740 (1982). “Frequently that ‘rigorous analysis’ will entail some overlap with the merits of the plaintiffs underlying claim.” Dukes, 131 S.Ct. at 2551. Here, Defendant vigorously challenges whether the D.C. class has met the numerosity requirement, and challenges, all three state classes on whether the named plaintiffs’ claims are typical of the class as a whole, whether the case presents sufficient commonality to justify proceeding ón a class basis, and whether such common inquiries, to the extent they exist, predominate over the individual questions.

With the 'exception of the adequacy of the proposed class counsel, which is not at issue, and the numerosity requirement with regard to the D.C. class, which presents an independent bar to certification of that class, the issues tend to merge into the determination of whether the state classes share common questions, susceptible to classwide proof, that advance the litigation to a sufficient extent to justify “an exception to the usual rule that litigation is conducted by and on behalf of the individual named parties only.” Califano v. Yamasaki, 442 U.S. 682, 700-01, 99 S.Ct. 2545, 61 L.Ed.2d 176 (1979); see Falcon, 457 U.S. at 157 n. 13, 102 S.Ct. 2364 (“The commonality and typicality requirements of Rule 23(a) tend to merge.”); In re Whirlpool Corp. Front-Loading Washer Products Liab. Litig., 722 F.3d 838, 853 (6th Cir.2013) (“[Cjommonality and typicality tend to merge with the requirement of adequate representation.... Due to the intertwined nature of commonality, typicality, and adequate representation, we consider them together.”); Saleem v. Corporate Transp. Grp., Ltd., No. 12 Civ. 8450(JMF), 2013 WL 6061340, at *7 (S.D.N.Y. Nov. 15, 2013) (“[Tjhe predominance inquiry is similar to, but more demanding than, the commonality inquiry.”).

The ramifications of Dukes are still being teased out by the courts, but a few observations can be made: Courts have not found that “Dukes bars certification in wage and hour cases,” Morris v. Affinity Health Plan, Inc., 859 F.Supp.2d 611, 616 (S.D.N.Y.2012) (collecting cases), and Defendant wisely does not attempt such an argument. Nor should it be suggested that Dukes bars certification of a class across multiple branches. That said, Dukes makes plain that the requirements of Rule 23 are not mere speedbumps on the road to certain certification, but rather that “[tjhe party seeking ‘class certification must affirmatively demonstrate ... compliance with the Rule,’ and a district court may only certify a class if it ‘is satisfied, after a rigorous analysis,’ that the requirements of Rule 23 are met.” In re Am. Int’l Grp., Inc. Sec. Litig., 689 F.3d 229, 237-38 (2d Cir.2012) (latter alteration in original) (quoting Dukes, 131 S.Ct. at 2551).

Another focus of the Dukes opinion was the issue of scale. As Judge Sand of this District has observed,

The Supreme Court suggested (when not explicitly stating) that the sheer size of the class and the vast number and diffusion of challenged employment decisions was key to the commonality decision. This makes a great deal of sense when the purpose of the commonality enquiry is to identify “some glue holding the alleged reasons for all of [the challenged] employment decisions together.”

Chen-Oster v. Goldman, Sachs & Co., 877 F.Supp.2d 113, 119 (S.D.N.Y.2012) (alteration in original) (citing Dukes, 131 S.Ct. at 2552). While Dukes certainly does not foreclose nationwide class actions, fairly read, it requires plaintiffs — like Ruiz⁸— who seek to challenge a nationwide policy to present qualitatively and quantitatively sufficient evidence demonstrating that policy. In other words, though anecdotal evidence need not follow a strictly proportional relationship to the size of the class, evidence that might suffice to demonstrate a local or regional policy might nonetheless fail to suffice at the national level. See Dukes, 131 S.Ct. at 2556 n. 9. Plaintiffs can, of course, frame their class allegations as they see fit; those plaintiffs who elect, however, to claim injuries in consequence of a nationwide policy are required under Dukes to present more, and broader, evidence of that policy.

Finally, Dukes suggested that in the absence of an express company policy that violated employee-plaintiffs’ rights, plaintiffs could nonetheless obtain class certification to challenge a practice that had sufficiently pervaded the company that it had become a de facto policy. Even here, however, plaintiffs are required to identify the “glue” holding together these disparate exercises of managerial discretion and rendering them suitable for classwide resolution. Specifically, the Court required plaintiffs to demonstrate “a common mode of exercising discretion that pervades the entire company,” since “it is quite unbelievable that all managers would exercise their discretion in a common way without some common direction.” 131 S.Ct. at 2554-55.

As a practical matter, employee-plaintiffs can rarely point to an explicit policy of their employer that is violative of their rights, including their rights under the relevant wage and hour laws. Courts have recognized this fact, and proof of de facto policies has therefore become the coin of the realm. To prove such policies, plaintiffs could present evidence of the implementation or recognition of these sub si-lentio policies at the senior management level, but such smoking guns are also quite rare. In the alternative, plaintiffs must present enough evidence to confidently suggest a uniform, or nearly uniform, practice occurring across branch offices. Perhaps even before Dukes, but certainly after that decision, this evidence entails either (i) comprehensive statistical analy-ses or (ii) anecdotal evidence that reaches a certain critical mass. Plaintiffs do not attempt the former, and as set forth in the remainder of this section, do not succeed in putting forth the latter.

2. Analysis

a. The State Law Classes Lack Questions of Law or Fact Common to the Classes Under Rule 23(a)(2)

It has been noted that, under Rule 23(a)(2)’s commonality requirement, “[e]ven a single common legal or factual question will suffice.” Freeland v. AT & T Corp., 238 F.R.D. 130, 140 (S.D.N.Y.2006) (citing In re Agent Orange Prod. Liab. Litig., 818 F.2d 145, 166-67 (2d Cir.1987)). Yet that common question must materially advance the litigation. As the Supreme Court has approvingly quoted, the language of Rule 23(a)(2) “is easy to misread, since '[a]ny competently crafted class complaint literally raises common questions.’ ” Dukes, 131 S.Ct. at 2551 (alteration in original) (quoting Richard Nagareda, Class Certification in the Age of Aggregate Proof, 84 N.Y.U. L.Rev. 97, 131-32 (2009)). Thus “[w]hat matters to class certification ... is not the raising of common questions — even in droves — but, rather the capacity of a classwide proceeding to generate common answers apt to drive the resolution of the litigation.” Id. (alteration and emphasis in original) (quoting Nagareda, supra, at 132).

Here, certain common questions are apparent and easily susceptible to classwide proof: Did Citibank have a national sales quota formula for personal bankers? Did Citibank attempt to reduce overtime hours paid to personal bankers from 2007 to 2012? Yet these types of questions, though generating common answers, are not apt to drive the resolution of the litigation. Setting production targets is a perfectly acceptable employment practice, as are “customary management admonitions to supervisors to watch and maintain control of work assignments in order to avoid unnecessary overtime.” Joza v. WW JFK LLC, No. 07 Civ. 4153(ENV)(JO), 2010 WL 8619661, at *2 (E.D.N.Y. Sept. 10, 2010) (denying FLSA and New York Labor Law claims after bench trial). And Citibank’s timekeeping system, which allowed branch managers to edit and approve timesheets, may have initially lacked certain desirable checks that were later added, but Plaintiffs have not demonstrated that it was systematically abused or designed to facilitate such abuse.

In this regard, Citibank’s -policies are comparable to Wal-Mart’s policy of delegating significant discretion to managers over pay and promotions, which, though vulnerable to abuse, is “a very common and presumptively reasonable way of doing business.” Dukes, 131 S.Ct. at 2554. And as with Dukes again, the only evidence of uniform corporate policy is Citibank’s entirely legal employment practices — here, repeated admonitions in training materials, handbooks, and communications that any overtime incurred must be paid. Cf. Hinterberger v. Catholic Health Sys., 299 F.R.D. 22, 51-52 (W.D.N.Y.2014) (“Because the question of CHS’s NYLL liability will come down to whether, on a department-specific basis, a particular supervisor did not follow CHS’s directive to devise and implement a method for reporting mealtime work, discouraged the reporting of mealtime work, or knew the work was performed but did not audit and correct time cards as required, Plaintiffs have not met the commonality prerequisite.”).

Plaintiffs identify four variations of the same common issue: “whether [personal bankers] were not properly compensated for overtime as a result [of] Citibank’s ‘de facto’ policy against the payment of overtime.” (Pl. Cert. Br. 2122). And indeed, this question lies at the very heart of the litigation, suggesting “that determination of its truth or falsity will resolve an issue that is central to the validity of each one of the claims in one stroke.” Dukes, 131 S.Ct. at 2551. Yet as noted above, this common question must be capable of “generating] common answers apt to drive the resolution of the litigation.” Id. (quoting Nagareda, supra, at 132). “Quite obviously, the mere claim by employees of the same company that they have suffered [a similar injury] gives no cause to believe that all their claims can productively be litigated at once.” Id. If the common question posed by Plaintiffs were, standing alone, sufficient to satisfy Rule 23’s commonality requirement, it would render commonality the sort of “mere pleading standard” rejected by Dukes. Accordingly, Plaintiffs seek to demonstrate by three routes a common plan or scheme to encourage and condone unpaid overtime: first, by demonstrating the common experiences of putative class members by anecdotal- testimonial evidence; second, by producing communications from and among Citibank managers demonstrating a common approach to overtime; and third, by arguing that knowledge of violations — an element of all three state legal standards for overtime violations- — -is common across the class. None of these avenues, either individually or in tandem, leads to a common answer.

Plaintiffs’ first argument is that the competing demands for high production and low hours worked created hydraulic pressure that predictably resulted in managers requiring off-the-clock work. Yet such a theory depends on demonstrating a relatively uniform response across branch managers; elsewise, Plaintiffs cannot identify “a common mode of exercising discretion that pervades the entire company.” Dukes, 131 S.Ct. at 2554-55. Managers exercised enormous discretion over the extent to which overtime was approved and overtime-reduction policies were enforced, and over the penalties given for failure to meet performance targets; as Citibank points out, the record contains several instances of managers reminding personal bankers to record more hours or overtime. (See Def. Decert. Reply 4 n. 6 (collecting record evidence)). As the Supreme Court pointed out, in a company with facially valid policies and wide discretion it is to be presumed that “most managers” operate in accordance with the law, rather than in violation of it. Dukes, 131 S.Ct. at 2554. “In such a company, demonstrating the invalidity of one manager’s use of discretion will do nothing to demonstrate the invalidity of another’s.” Id. In response to this hurdle, Plaintiffs have amassed not-insignificant anecdotal evidence to suggest that certain branch managers applied improper pressure, and that certain personal bankers felt pressured to work off the clock. Yet Plaintiffs’ anecdotal evidence is deficient in both scale and uniformity.

As noted supra, the Supreme Court has disclaimed the notion that anecdotal evidence must follow a strictly proportionate relationship to the size of the class. Dukes, 131 S.Ct. at 2556 n. 9. The Court distinguished, however, between cases in which 40 accounts of discrimination represented roughly one account for every eight class members, id. at 2556 (citing Teamsters v. United States, 431 U.S. 324, 338, 97 S.Ct. 1843, 52 L.Ed.2d 396 (1977)), and Dukes itself, in which 120 affidavits represented roughly one account for every 12,-500 class members, id. Here, the 31 named Plaintiffs, Declarants, and Sample Opt-Ins identified at Exhibit 130 of Plaintiffs’ Tyner Declaration represent roughly one account for every 76 of the roughly 2,346 putative class members (see Tyner Decl. ¶ 3). Yet even this ratio is unduly generous to Plaintiffs. If Plaintiffs seek to prove a nationwide policy that was disseminated throughout the three class states, the denominator (all personal bankers over the relevant period nationwide) would be significantly higher than 2,346, and the ratio of anecdotal accounts to personal bankers correspondingly lower.⁹ If Plaintiffs seek to prove uniformity of experience solely within the three state classes, only 15 of these 31 plaintiffs worked at branches in New York (11), Illinois (3), or the District of Columbia (1); thus, the 15 relevant anecdotal accounts represent only one in 156 class members (or, more precisely, one in 182 for New York, one in 110 for Illinois, and one in 16 for the District of Columbia). This sort of scattershot evidence does not meet the requirements of Rule 23 as envisioned in Dukes. See, e.g., In re Bank of Am. Wage & Hour Emp’t Litig., 286 F.R.D. 572, 580-81 (D.Kan.2012) (rejecting certification of statewide class where plaintiffs offered 100 declarations out of putative class of 35,000 (1 in 350)).¹⁰

Moreover, even if this number of anecdotal accounts were sufficiently large to confidently be extrapolated to the class as a whole, the contradictions and heterogeneity within the group suggest the lack of a common answer. (See, e.g., Drago Dep. 57 (testifying to different practices with regard to overtime at different branches), 167-68 (no written warning for repeatedly missing sales targets); Ho Dep. 82-83 (testifying that Citibank sent her on several all-expenses paid trips for her sales performances), 148 (“[I]n Jackson Heights branch it is always busy.... You won’t think about you won’t meet your goal. I never think about that because it is so busy. It is so easy to meet the goal.”), 160-62 (testifying that she made her performance goals in 79 of 87 months, at times met multiples of her goal, and received tens of thousands of dollars in performance bonuses over her seven years at Citibank); Kosiba Dep. 151 (unable to remember ever failing to record overtime, or branch manager ever making changes to her knowledge); Pikulin Dep. 98,127-28 (told to record all time, and unable to remember not recording all time or being paid for all time); Shen Dep. 228 (never threatened or disciplined for working overtime, and unable to remember being instructed to work off the clock or failing to enter overtime); Winfield Dep. 178 (never disciplined for failing to meet sales targets)).¹¹ Citibank’s anecdotal evidence thus suggests that although the competing pressures not to request overtime and to meet performance targets may have existed at a companywide — and therefore class-wide — level, they did not replicate themselves with sufficient uniformity across the branches to result in a common answer that goes to the ultimate question of liability-

In addition to anecdotal testimonial evidence from personal bankers, Plaintiffs point to the higher levels of Citibank’s management and argue that, despite the formal policies regarding overtime pay, there is sufficient evidence of a de facto plan or policy to encourage off-the-clock work. Such evidence has not been presented to this Court and, it would appear, does not exist. Plaintiffs’ evidence of such a de facto policy consists of inferences drawn from a comparatively miniscule quantity of emails among branch and area managers, and the testimony of certain personal bankers. As previously noted, many of the emails contain explicit admonitions that any overtime incurred must be paid.¹² Those emails that lack such admonitions overwhelmingly maintain their focus on minimizing the amount of overtime worked, rather than declining to pay for overtime already accrued.¹³ The testimony of personal bankers as to a de facto nationwide policy, meanwhile, is well outside the bounds of what such deponents might personally know; such testimony consists of personal bankers stating what their branch managers told them about what their branch managers’ superiors told their branch managers.¹⁴ The Court acknowledges that those engaging in a common scheme to violate the law will rarely coordinate such efforts by formal announcement. Yet the burden still rests upon Plaintiffs to show, by a preponderance of the evidence, that a common plan exists that is subject to testing at the classwide level. That burden has not been met.

Finally, Plaintiffs suggest that even if the wrongful behavior did not occur across the class, Citibank was generally put on notice as to the prevalence of denial of overtime wages, thus answering a significant common question. Even if such a common question suffices to meet the requirements of Rule 23(a)(2), the Court once again finds that the record betrays the absence of a common answer. Plaintiffs have brought forward only a handful of emails demonstrating that awareness of overtime violations percolated above the branch manager level, and these instances brought about immediate efforts to rectify the violations. In one instance, a new branch manager discovered that employees had been pressured by the previous branch manager into working off-the-clock. (Dkt. # 67 Ex. 1 (Henderson Decl.) ¶ 6). In response, Citibank’s Human Resources department interviewed each personal banker at the branch, and paid each of them the highest amount of uncompensated overtime — 300 hours — estimated by any one of them. (Id. at ¶¶ 8-9). In another instance, Sample Opt-In Schornstein was paid nearly $12,000 for unreported overtime when her branch was investigated by Citibank for underreporting time. (Schornstein Dep. 183). While it is true that Citibank is liable for every violation of which a branch manager became aware, “[i]n the absence of a company-wide policy or practice, plaintiffs will have to demonstrate that each individual manager had actual or constructive knowledge that plaintiffs were performing off-the-clock work without proper compensation. The record suggests that the knowledge of each individual manager varies widely.” Zivali v. AT & T Mobility, LLC, 784 F.Supp.2d 456, 468 (S.D.N.Y.2011).

The instant case is similar to that brought on behalf of personal bankers against Wells Fargo and Wachovia in Fernandez v. Wells Fargo Bank, N.A., Nos. 12 Civ. 7193(PKC), 12 Civ. 7194(PKC), 2013 WL 4540521 (S.D.N.Y. Aug. 28, 2013). The plaintiffs’ theory was that “the personal bankers of Wells Fargo ... were subject to a common, [statewide] policy to limit their recorded hours or require off-the-clock work.” Id. at *5. Yet as in this case, “[n]owhere in plaintiffs’ papers do they cite a specific, concrete, management directive concerning plaintiffs’ off-the-clock work or any purported requirement not to record hours worked.” Id. And similarly, the plaintiffs’ evidence consisted of email correspondence submitted with little or no context, statements made by several individuals without personal knowledge, and a significant amount of anecdotal evidence, much of which contained contradictions, and some of which lay either temporally or geographically outside the bounds of the proposed class. Id. at *5-9.¹⁵ The court there denied the motion for class certification, finding that “[l]imited, anecdotal evidence concerning a de facto compensation policy is insufficient to establish commonality.” Id. at *5.

The cases upon which Plaintiffs rely are largely distinguishable, though the Court will not pretend that the entirety of federal Rule 23 jurisprudence (even, if not especially, post-Dukes) can be reconciled. Some present common questions of whether the formal classification of employees or activities as not falling within the scope of the applicable overtime law was correct; a review of the case law confirms that these “misclassification” cases are far more susceptible of classwide resolution than “off-the-clock” cases such as the instant one. In Jackson v. Bloomberg, L.P., 298 F.R.D. 152 (S.D.N.Y.2014), for example, the common questions included whether a class of employees was properly classified as exempt and whether certain work activities were properly characterized as sufficiently minimal to obviate the need for recording. Id. at 162-64; see also Jacob v. Duane Reade, Inc., 602 Fed.Appx. 3, 6, 2015 WL 525697, at *2 (2d Cir. Feb. 10, 2015) (summary order) (“[T]he common contention to be proved is whether Duane Reade misclassified its employees as exempt from New York’s overtime requirements.”). In other cases, the evidence of a de facto policy that deviated from a formal policy is significantly stronger than in the instant case. See, e.g., Whitlock v. FSL Mgmt., LLC, No. 3:10 Civ. 00562(JHM), 2012 WL 3274973, at *6-7 (W.D.Ky. Aug. 10, 2012) (discussing meeting notes of enforcement of unpaid promotional duties by the manager of the entire class). And in still other cases, the anecdotal evidence was far more compelling due to the more compact nature of the class. See Jimenez v. Allstate Ins. Co., No. 4:10 Civ. 8486(JAK), 2012 WL 1366052 (C.D.Cal. Apr. 18, 2012) (certifying a class of claims adjusters across 13 California market claims offices within a “claims service area” under the supervision of a single manager); Mahoney v. Farmers Ins. Exch., No. 09 Civ. 2327, 2011 WL 4458513, at *8-9 (S.D.Tex. Sept. 23, 2011) (granting certification of an FLSA collective action across three office locations, but collecting cases denying certification of larger and more dispersed classes); cf. Dukes, 131 S.Ct. at 2556 & n. 9 (discussing the persuasiveness of anecdotal evidence based upon comparison to the size of the potential class).

The Court does not downplay the evidence amassed by Plaintiffs of individual violations, of systematic violations at the branch level, and even of violations induced by certain area directors across multiple branches. But the record in its totality shows that Citibank’s formal, com-panywide policies are entirely legal and appropriate, and that there is no evidence of a common plan or scheme to subvert these policies. A classwide determination of liability thus depends upon demonstrating that appropriate policies have reliably translated themselves into inappropriate managerial behavior across the width and breadth of the class. Such evidence is simply not to be had from this record. All of the Sample Opt-Ins were paid overtime in at least some pay periods, with wide variation between them. While paying some overtime does not relieve Citibank of liability for the overtime it failed to pay, the variation suggests that the effects of Citibank’s purported “no overtime” policy were far from uniform. To be clear, the Court does not suggest that every branch must have implemented unlawful policies for class certification to be appropriate. But at a minimum Plaintiffs must demonstrate “common direction” in the allegedly unlawful behavior or “a common mode of exercising discretion that pervades the entire company.” Dukes, 131 S.Ct. at 2554-55. The Court finds the evidence insufficient to demonstrate either common direction or a common mode of exercising discretion, and accordingly finds an absence of common questions susceptible to classwide proof.

b. To the Extent There Are Common Questions of Law or Fact, They Do Not Predominate Over Individual Questions, and a Class Action Is Not a Superior Method of Adjudication Under Rule 23(b)(3)

Rule 23(b)(3) requires, in part, that “the questions of law or fact common to class members predominate over any questions affecting only individual members.” Yet the Court has found, under the standards set forth in Dukes, that because the plaintiffs have provided “no convincing proof of a companywide [unlawful] policy ... they have not established the existence of any common question.” 131 S.Ct. at 2556-57. If there are no common questions, then common questions cannot predominate over individual questions. See Amchem Products, Inc. v. Windsor, 521 U.S. 591, 624, 117 S.Ct. 2231, 138 L.Ed.2d 689 (1997) (finding that .“the predominance criterion is far more demanding” than the commonality criterion). While the Court accordingly need not fully address the parties’ disputes over predominance, it is worth noting that Defendants have raised a significant number of necessarily individualized inquiries into liability that confirms the impracticability of proceeding as a class, largely concerning the questions of whether personal bankers worked unreported overtime, at whose direction, and with whose knowledge. Such questions speak to both predominance and superiority. See Johnson v. Nextel Commc’ns Inc., 780 F.3d 128, 148 (2d Cir.2015) (“Because liability for a significant bloc of the class members and damages for the entire class must be decided on an individual basis, common issues do not predominate over individual ones and a class action is not a superior method of litigating the case.”).

These questions of liability for unreported overtime are distinct from questions going to damages. The weight of circuit authority holds that the presence of some potential class members who have suffered no damages is not fatal to certification, so long as liability for any damages is susceptible to common proof. See In re Whirlpool, 722 F.3d at 855 (rejecting, in finding common liability for the sale of washing machines susceptible to mold, the argument “that the certified class is too broad because it includes Duet owners who allegedly have not experienced a mold problem and are pleased with the performance of their Duets”). Here, however, the issue is not merely that damages are highly particularized, but that Plaintiffs have failed to make the necessary showing that whatever damages exist are traceable to a common plan of scheme attributable to Citibank as a whole. Because the defect goes to liability and not damages, it cannot be cured by simply defining those personal bankers who worked no unpaid overtime as outside the class (see Pl. Cert. Reply 3 & n. 7) — a suggestion that, as Defendant points out, would raise serious concerns about the impermissible creation of a “fail-safe class” (see Def. Cert. Sur-Reply 2-3 (collecting cases)). It is not the existence of class members who suffered no harm that defeats certification, but rather the lack of common questions driving liability for even those class members who did suffer harm.

Furthermore, the necessity of additional individualized inquiries over damages, while not inherently fatal to class certification, “should be considered at the certification stage when weighing predominance issues.” Roach v. T.L. Cannon Corp., 778 F.3d 401, 408 (2d Cir.2015). Plaintiffs argue that Defendant’s database containing computer log-in and log-out times can effectively resolve the question of individualized damages in a single blow, by comparing the computer audit logs to the timesheets in Defendant’s NATA system. (See Pl. Cert. Br. 31-32). However, such information would not obviate the need for individualized inquiries. Several Plaintiffs testified that they regularly worked before logging in and after logging out (see, e.g., Lindsey Dep. 120-21; Nessler Dep. 174-78; Ruiz Dep. 338-39; Winfield Dep. 317-18); conversely, the audit logs would not reveal the length of lunch breaks, or whether a logged-in employee was researching sales opportunities or taking advantage of online shopping sales (see Bodt Dep. 6870). This Court, like other courts tp confront the same issue, finds that audit logs developed for a distinct purpose (monitoring computer activity) are not a reliable way of calculating hours worked. See Howard v. CVS Caremark Corp., No. 13 Civ. 4748(SJO), 2014 WL 7877404, at *9-11 (C.D.Cal. Dec. 9, 2014) (finding CVS’s prescription-tracking system, Rx Connect, “to be poor ‘glue’ for holding together Plaintiffs’ claims” of unpaid off-the-clock overtime work due to “several shortcomings with the Rx Connect records”); In re Bank of Am., 286 F.R.D. at 588 (“[I]t is not possible for anyone to accurately ascertain whether an employee in fact worked off-the-clock simply by comparing the transactional data with that employee’s timekeeping records.”). In short, the plethora of individualized questions going to liability and damages would defeat class certification under Rule 23(b)’s predominance requirement even if a common policy or practice could be found under Rule 23(a).

c. The Washington, D.C., Class Additionally Lacks Numerosity Under Rule 23(a)(1)

As an additional defect, the Washington, D.C. class, consisting of 16 personal bankers, lacks sufficient numerosity to make joinder of all members impracticable. See Siegel v. Bloomberg L.P., No. 13 Civ. 1351(DLC), 2013 WL 4407097, at *2 (S.D.N.Y. Aug. 6, 2013) (noting that “[n]u-merosity is presumed when a class consists of forty or more members,” and denying certification to a class consisting of 33 members). Plaintiffs’ mere assertion that “joinder would be unwieldy and burdensome” does not make it so. (Pl. Cert. Br. 20 n. 96). Though the lack of commonality, predominance, and superiority is sufficient to bar certification of the class, the lack of numerosity is additionally fatal.¹⁶

B. Defendant’s Motion to Decertify the FLSA Collective Action Is Granted

Effectively as a cross-motion, Defendants have moved to decertify the FLSA collective action that was previously certified by the Court. As set forth herein, the critical inquiry under FLSA is whether the opt-in plaintiffs are “similarly situated.” The Court’s does not, to be clear, undertake the “rigorous analysis” prescribed by Dukes, but it is still required to find sufficient evidence of common treatment. Despite years of discovery, Plaintiffs have failed to adduce enough evidence to meet even this lesser burden. Accordingly, the Court grants Defendants’ motion.

Section 216(b) of the FLSA authorizes collective actions to recover damages for unpaid wages where all employees are “similarly situated.” 29 U.S.C. § 216(b). “When deciding whether to certify a class under 29 U.S.C. § 216(b), district courts in the Second Circuit apply a two-step process.” Morano v. Intercontinental Capital Grp., Inc., No. 10 Civ. 2192(KBF), 2012 WL 2952893, at *4 (S.D.N.Y. July 17, 2012). The first step of this process, often termed conditional certification,¹⁷ requires plaintiffs to make “only a ‘modest factual showing’ that the plaintiff and potential opt-in plaintiffs ‘together were the victims of a common policy or plan that violated the law.’ ” Id. (quoting Myers v. Hertz Corp., 624 F.3d 537, 555 (2d Cir.2010)). At the second step, typically the defendant moves for decertification, and “the district court will, on a fuller record, determine whether a so-called ‘collective action’ may go forward by determining whether the plaintiffs who have opted in are in fact ‘similarly situated’ to the named plaintiffs.” Myers, 624 F.3d at 555. At this second stage, “[t]he court must apply a more ‘stringent standard’ of proof ... for determining whether plaintiffs are similarly situated for the purposes of the FLSA.” Moruno, 2012 WL 2952893, at *5.

It has been noted that “the great majority of cases involve certification at the initial stages of the litigation,” Torres v. Gristede’s Operating Corp., No. 04 Civ. 3316(PAC), 2006 WL 2819730, at *9 (S.D.N.Y. Sept. 29, 2006); perhaps for this reason, “[t]he Second Circuit has yet to prescribe a particular method for determining whether members of a class are similarly situated” at the second, more stringent stage, Pefanis v. Westway Diner, Inc., No. 08 Civ. 002(DLC), 2010 WL 3564426, at *4, (S.D.N.Y. Sept. 7, 2010). Some courts, respecting Congress’s admonition not to import Rule 23’s requirements into the FLSA, have adopted a mul-tifactor approach, see Torres, 2006 WL 2819730, at *9 (adopting the Tenth Circuit’s “ad hoc” approach (citing Thiessen v. Gen. Elec. Capital Corp., 267 F.3d 1095, 1102 (10th Cir.2001))), while others have framed the inquiry in terms similar to Rule 23’s commonality requirement, see Pefanis, 2010 WL 3564426, at *4 (requiring “a persuasive showing that the original and opt-in plaintiffs were common victims of a FLSA violation pursuant to a systematically-applied company policy or practice such that there exist common questions of law and fact that justify representational litigation”). Other courts have gone further, analogizing the FLSA certification inquiry to the overall standard of Rule 23. See Espenscheid v. DirectSat USA, LLC, 705 F.3d 770, 772 (7th Cir.2013) (“[Tjhere isn’t a good reason to have different standards for the certification of the two different types of actions, and the case law has largely merged the standards, though with some terminological differences.” (collecting cases)); accord Gardner v. W. Beef Props., Inc., No. 07 Civ. 2345(NGG) (JMA), 2013 WL 1629299, at *6 (E.D.N.Y. Mar. 25, 2013) (“The more opt-ins there are in the class, the more the analysis under § 216(b) will mirror the analysis under Rule 23.”), report and recommendation adopted sub nom. White v. W. Beef Props., Inc., No. 07 Civ. 2345(RJD)(JMA), 2013 WL 1632657 (E.D.N.Y. Apr. 16, 2013). But see Hernandez v. Fresh Diet, Inc., No. 12 Civ. 4339(ALC) (JLC), 2014 WL 5039431, at *8 (S.D.N.Y. Sept. 29, 2014) (“A number of district courts in this Circuit have observed that the second-stage ‘similarly situated’ analysis under FLSA § 216(b) is considerably less stringent than the requirement of Rule 23(b)(3) that common questions predominate.” (internal alteration and quotation marks omitted)).

Ultimately, this Court agrees with the Tenth Circuit that, functionally, “there is little difference in the various approaches” to motions to decertify a collective action under the FLSA. Thiessen, 267 F.3d at 1105. The Supreme Court, analyzing the same “similarly situated” standard of 29 U.S.C. § 216(b) that is incorporated into both the FLSA and the Age Discrimination in Employment Act (the “ADEA”), noted that collective actions allow the judicial system to “benefit[ ] by efficient resolution in one proceeding of common issues of law and fact arising from the same alleged discriminatory activity.” Hoffmann-La Roche Inc. v. Sperling, 493 U.S. 165, 170, 110 S.Ct. 482, 107 L.Ed.2d 480 (1989). The Seventh Circuit has noted the similarity between this analysis and the efficiency motivations underlying Rule 23. See Espenscheid, 705 F.3d at 772 (“[T]he provisions of Rule 23 are intended to promote efficiency as well, and in that regard are as relevant to collective actions as to class actions.” (internal citations omitted)). Given this harmony of animating principles, it is not mere coincidence that courts facing parallel motions to decertify an FLSA collective action under Section 216(b) and to certify a class action under Rule 23 have tended to allow either both actions or neither to proceed on a collective basis. See Gardner, 2013 WL 1629299, at *6 n. 3 (collecting cases).

At the first step in the FLSA certification process, Judge Koeltl granted Plaintiffs’ motion for conditional certification and authorized notice to be sent out to potential opt-in plaintiffs. The Court, noting Plaintiffs’ “minimal burden of showing that they are similarly situated to one another and to potential opt-in plaintiffs” at this stage, found that they had made the “modest factual showing” necessary for the Court to conditionally certify the class. Winfield, 843 F.Supp.2d at 401-10. In doing so, the Court noted several distinctions between the standard necessary for conditional certification and the fuller inquiries required both for Rule 23 certification and for upholding conditional certification in the face of a subsequent motion to decertify. In particular, it noted that at that earlier stage “the court does not resolve factual disputes, decide substantive issues going to the ultimate merits, or make credibility determinations,” id. at 402 (quoting Cunningham v. Elec. Data Sys. Corp., 754 F.Supp.2d 638, 644 (S.D.N.Y.2010)); that “courts in this Circuit regularly rely on [hearsay] evidence to determine the propriety of sending a collective action notice,” id. at 402-03 (alteration in original) (quoting Moore v. Eagle Sanitation, 276 F.R.D. 54, 59 (E.D.N.Y.2011)); and that the commonality requirement of Dukes is not applicable to a motion for conditional certification, id. at 409-10. And while accepting at that stage the testimony of several plaintiffs and declar-ants as to the misbehavior of several branch managers, the Court nonetheless cautioned that “[i]f further discovery reveals that such managers were indeed acting aberrantly, the defendant remains free to renew this argument by bringing a motion to decertify the class following the close of discovery.” Id. at 409. Defendant has indeed renewed the argument in its motion to decertify the collective action, and has done so successfully.

With discovery relating to certification and decertification now closed, the Court turns to the fuller consideration of the record contemplated by the FLSA. Without belaboring the point, Defendant’s motion to decertify the collective action succeeds for exactly the same reason Plaintiffs’ motion to certify the class actions fails — a lack of commonality, even under the less stringent FLSA analysis. Plaintiffs’ evidence at the point of collective certification consisted primarily of anecdotal testimony as to several managers’ misbehavior, and statements that those managers attributed the need to work unpaid overtime to company policy. Winfield, 843 F.Supp.2d at 405-08. After nearly two years of discovery, dozens of deposition, and hundreds of thousands of pages of document discovery (Linthorst Decl. ¶¶ 4-16), Plaintiffs have advanced the ball very little in demonstrating a common plan or scheme. As discussed above, Plaintiffs still rely largely on anecdotal allegations of violations, secondhand statements regarding companywide policy to force unpaid overtime attributed to branch managers, and a pair of entirely appropriate workplace policies that interacted — with highly uneven and uncertain effect — across Citibank’s many branches. Such evidence may suffice for conditional certification, but it does not provide a persuasive showing that the opt-in plaintiffs are “similarly situated,” and thus does not survive Citibank’s motion to decertify the collective action.

To be clear, the Supreme Court’s analysis of Rule 23’s commonality requirement in Dukes does not govern the “similarly situated” analysis. See 7B Charles Alan Wright & Arthur Miller, Federal Practice & Procedure § 1807 (3d ed. as modified 2014) (collecting cases and determining that district courts have “uniformly rejected” the argument that Dukes tightened the standard for FLSA collective actions). Yet the critical difference between Rule 23(b)(3) class actions and FLSA collective actions — the opt-out versus the opt-in model — carries less weight here due to the manner in which discovery proceeded. As this Court has expatiated in the preceding section, even when one examines only the sample Opt-in Plaintiffs, massive disparities are apparent in the policies of their branch managers, the difficulty in meeting their sales targets, and the frequency with which they received overtime. (See supra at n. 11). Because Plaintiffs have not shown a common policy that operated to common effect, or some other mode of evidencing shared experiences, they cannot proceed as a collective action. See Hernandez, 2014 WL 5039431, at *4 (decertify-ing collective action because “Plaintiffs are unable to carry their burden to demonstrate that the Defendants had a common policy or plan in violation of the FLSA that negatively impacted the original and opt-in Plaintiffs” (internal alterations and quotation marks omitted)). To find otherwise would reduce Section 216(b)’s requirement that plaintiffs be “similarly situated” to a mere requirement that Plaintiffs share an employer, a job title, and a professed entitlement to additional wages.

Where the collective action as envisioned cannot proceed, “the case may be divided, if appropriate, into subgroups. Alternatively, the claims of the opt-in plaintiffs may be dismissed without prejudice and the action proceed for the original plaintiffs alone, but not as a collective action.” Pefanis, 2010 WL 3564426, at *4 (internal citation omitted). Here, there is no apparent subdivision that would correct the fatal flaws in Plaintiffs’ attempt to maintain a collective action. As Citibank notes, looking at the 37 named Plaintiffs and Sample Opt-Ins who participated in discovery, there is only a single branch manager who oversaw more than one member of this group, having supervised three different Sample OpWIns at varying points in time. (Linthorst Decl. ¶ 14). Though consideration of the full class of 436 opt-in plaintiffs would no doubt reveal additional overlap, their employment would still have taken place at hundreds of branches scattered across the country under an even greater number of branch managers. And even branch location might not suffice as a class metric, given the potential individualized defenses to liability. Given that the number of necessary subclasses would likely approach the number of plaintiffs, there are no apparent efficiencies to be gained from proceeding collectively rather than by individual suit. Cf. Morano, 2012 WL 2952893, at *7 (de-certifying a previously conditionally certified FLSA collective action, and refusing to subdivide where there were individualized defenses and -multiple branch locations). Accordingly, the Court dismisses the claims of the opt-in plaintiffs without prejudice.

CONCLUSION

Because the proposed class actions lack common questions of law or fact, Plaintiffs’ motion to certify class actions under the New York Labor Law, the Illinois Minimum Wage Law, and the District of Columbia Minimum Wage Act Revision Act is DENIED. In addition, because the opt-in plaintiffs are not similarly situated, Defendant’s motion to decertify the collective action under the FLSA is GRANTED. The claims of the plaintiffs who have opted in to the collective action are DISMISSED without prejudice.¹⁸ Accordingly, the Clerk of Court is directed to terminate Docket Entries 178,182, and 187.

The Court further grants the parties’ letter requests to file redacted versions of their briefs and to file certain exhibits to the declarations in support of and opposition to their briefs under seal. Having weighed the factors set forth by the Second Circuit in Lugosch v. Pyramid Co. of Onondaga, 485 F.3d 110 (2d Cir.2006), the Court finds that the parties’ interest in preserving Plaintiffs’ financial privacy and protecting the confidentiality of certain of Defendants’ manuals and training materials outweigh the public’s interest in disclosure.

The remaining parties are directed to appear before the Court for a status conference on April 7, 2015, at 10:00 a.m., in Courtroom 618 of the Thurgood Marshall Courthouse, 40 Foley Square, New York, New York 10007.

SO ORDERED.

1 . The facts in this Opinion are drawn from the parties' declarations ("[Name] Decl.”) and exhibits thereto submitted with the parties’ briefs. Deposition testimony is cited as "[Name] Dep.” Except where indicated otherwise, docket entry numbers refer to the Ruiz Action, No. 10 Civ. 5950, rather than the Winfield Action, No. 10 Civ. 7304.

Defendant’s opening brief in support of the motion for decertification is referred to as "Def. Decert. Br.”; Plaintiffs’ opposition as "PL Decert. Opp.”; and Defendant’s reply brief as "Def. Decert. Reply.” Plaintiffs’ opening brief in support of the motion for certification is referred to as "PL Cert. Br.”; Defendant’s opposition as "Def. Cert. Opp.”; Plaintiffs’ reply as "PL Cert. Reply”; Defendant’s sur-reply as "Def. Cert. Sur-Reply”; and Plaintiffs’ sur-sur-reply as "PL Cert. Sur-Sur-Reply.”

2 . Plaintiffs have not moved to certify the California state law class due to the subsequent certification of such a class in California state court on January 23, 2013. (Pl. Cert. Br. 3 n. 2 (citing Davis v. Citibank, N.A., No. 30-2008-00060145 (Cal.Sup.Ct. Orange Cty.))). That case reached a settlement shortly thereafter. (Reiss Decl. Ex. 1).

3 . Judge Koeltl allowed notice to be sent to those potential class members who had been employed as personal bankers at any point after August 6, 2007 (three years before the filing of the Ruiz Complaint), and to those who had been employed as personal bankers in New York at any point after August 6, 2004, due to the NYLL’s six-year statute of limitations. Winfield, 843 F.Supp.2d at 410-11. Judge Koeltl noted that Defendant could subsequently challenge the timeliness of any class member’s claims. Id.

4 . A number of these opt-in Plaintiffs have been dismissed from the action for failure to comply with discovery obligations. (See, e.g., Dkt. #213). Citibank additionally maintains that 156 of the 437 opt-in plaintiffs’ claims are time-barred by the FLSA’s three-year statute of limitations. (Def. Decert. Br. 1). The precise number of FLSA opt-in plaintiffs is not relevant to the disposition of the motions at issue.

5 . In 2009 and 2010, the monthly quota varied from 28% to 35% of annual salary depending on the age of the branch; in 2011 it was set at 32%; in 2012 at 16%; and in 2013 at 20%. ees 2009-2013 Individual Seller Plan Brochures). Some adjustments to sales credits allocated per transaction type were made over the years. (Id..).

6 . Sample Opt-In Valencia was not paid overtime in any of his 18 pay periods during the May 2009 to December 2013 period, though he was paid overtime in 10 of his 50 pay periods from 2007 to May 2009. (Hammond Decl. ¶ 30).

7 . The FLSA applies to employers with an annual gross volume of sales of at least $500,000. 29 U.S.C. § 203(s)(1)(A)(ii).

8 . Plaintiffs seek Rule 23 certification only for three statewide classes. Yet rather than prove a common policy or practice within each relevant state, Plaintiffs seek to demonstrate a nationwide policy that can be imputed to each state. (See Pi. Cert. Br. 2 ("The directive against overtime expense came from top management and thus was common to [personal bankers] nationwide.”), 5 ("[Personal bankers] across the country testified that they were subject to this disciplinary procedure....”), 11 ("Branch managers delivered and strictly enforced the national 'no overtime’ directive from management.”); see also id. at 11 (citing evidence from Florida and California), 13 (Texas)). Such a decision was strategic — and understandable — inasmuch as it allowed Plaintiffs to pursue simultaneously a collective action under the FLSA.

9 . According to USBankLocations.com, 363 of Citibank's 955 U.S. branches (38%) are located in New York (273), Illinois (72), or the District of Columbia (18). See U.S. Bank Locations, www.usbanklocations.com (last visited March 19, 2015).

10 . As noted supra, the limited scope of this anecdotal evidence is due to Plaintiffs’ Pyrrhic victory in having discovery from the Opt-In Plaintiffs limited to a sample of 30 such Plaintiffs. (See Dkt. # 123).

11 . Defendant has submitted numerous "happy camper” affidavits of personal bankers who did not opt in to the FLSA collective action and were not, therefore, among the Sample Opt-Ins designated by Magistrate Judge Ellis for discovery. (See Reiss Decl. Ex. 7 to Ex. 19). The Court disregards these affidavits for the purposes of deciding the instant motions, finding sufficient variation even among the Sample Opt-Ins and named Plaintiffs to disprove the existence of common questions susceptible to classwide proof.

12 . In one particularly misleading citation, Plaintiffs point to an email in which a Citibank manager (presumably an area manager, though she is not identified) states, "As you can feel expense management is highly visible with tremendous pressure. [Overtime] is something you control ... if incurred we pay but I view as something you control and [it should] not be a smprise at the end of the month." (Tyner Decl. Ex. 101 (emphasis added)). Plaintiffs omit the rather important italicized portion. (See PL Cert. Br. 10).

13 .For example, one email to which Plaintiffs ascribe significance states: "Awareness of the cost of scheduling Personal Bankers to work more hours since their base is higher. We need to reduce [overtime pay] and our Sellers [are] our biggest opportunity.” (Tyner Decl. Ex. 48 (emphasis added)). Plaintiffs have, however, uncovered some emails suggesting a desire to deny overtime already incurred (though even these indicate that the solution is adjusting work hours later in the week rather than simply not paying for all hours worked). (See id. Ex. 110 (“If you have OT that is not approved, you will not be paid for it.... If you feel that you might have some OT hours, see Anthony or myself ASAP so that we can work out the rest of the week’s hours.”)).

14 . Such evidence, if offered at trial, would likely be admissible because both statements fall within Federal Rule of Evidence 801(d)(2)(D)’s hearsay exception for statements by a party-opponent’s employee within the scope of the relationship. Nevertheless, the Court is required to weigh competing evidence, even if overlapping with the merits of the claim, to determine whether Rule 23’s commonality requirement has been met by a preponderance of the evidence. Though the question of credibility inquiries has usually arisen in the context of competing expert testimony, the analysis in that context is instructive:

Rigorous analysis need not be hampered by a concern for avoiding credibility issues; as noted, findings with respect to class certification do not bind the ultimate fact-finder on the merits. A court’s determination that an expert’s opinion is persuasive or unpersuasive on a Rule 23 requirement does not preclude a different view at the merits stage of the case.

In re Hydrogen Peroxide Antitrust Litig., 552 F.3d 305, 324 (3d Cir.2008), as amended (Jan. 16, 2009). Similarly, the requirement of rigorous analysis must extend to evaluating the reliability of a handful of declarations of second- or thirdhand knowledge in establishing Citibank's nationwide corporate policy. Cf. Suvill v. Bogopa Serv. Corp., No. 11 Civ. 3372(SLT)(RER), 2014 WL 4966029, at *8 (E.D.N.Y. Sept. 30, 2014) ("As a preliminary matter, the Court hesitates to give weight to those assertions in the declarations submitted by plaintiffs’ counsel that have not been corroborated by deposition testimony because similar statements in the declarations have proven inaccurate and many of these assertions are not based on personal knowledge.”). Put simply, in light of Dukes, the Court is unwilling to certify classes of this magnitude based on imprecise recollections of information for which Plaintiffs have no firsthand knowledge, particularly where such recollections are (i) not uniform among class members, (ii) undercut by extensive contemporaneous evidence, and (iii) contradicted by the only company-wide evidence presented to the Court.

15 . It is worth reiterating the scattered and frequently acontextual nature of Plaintiffs’ email evidence in this case, and the concentration of much of this evidence among the-California branches that are no longer directly relevant to the state law classes (except by demonstration of a nationwide policy that can be extrapolated to the relevant states).

16 . The Court accordingly need not address whether the Rules Enabling Act prevents certification of an opt-out class for violation of Washington, D.C.’s minimum wage laws, which provide an opt-in mechanism. (See Def. Cert. Opp. 6; Pl. Cert. Reply 14-15).

17 . The Court agrees with Judge Cogan's observations in Morangelli v. Chemed Corp., 275 F.R.D. 99, 103 n. 1 (E.D.N.Y.2011):

Courts, including this one and the Second Circuit, see e.g., Kuebel v. Black & Decker Inc., 643 F.3d 352, 357 (2d Cir.2011), have called this stage of the FLSA proceeding, "conditional certification." This is a misnomer and may obfuscate the leniency of the standard employed to authorize plaintiffs' counsel to send notices of the action. When the Court allows notices to be sent out, it is only making a preliminary determination — often based solely on allegations — of whether plaintiffs are "similarly situated” under 8 U.S.C. § 216. See C. Wright, A. Miller & E. Cooper, 17A Federal Practice and Procedure § 1807 (3d ed.2011). The Court is not assuming that a class exists as Rule 23 used to permit courts to do, see 1 Joseph M. McLaughlin, McLaughlin on Class Actions § 3:6 (6th ed.2010) (explaining the history of "conditional certification” of class actions); there is no class in a collective action. Nor is the Court certifying anything — class or otherwise.

Cognizant of this imprecision, the Court nonetheless adopts the "certification-decertifi-cation” parlance used by other courts to describe the two-step process under the FLSA. ’

18 . This Order does not revive the claims of those sample opt-in plaintiffs who have previously been dismissed with prejudice for failure to comply with discovery orders in this case. Furthermore, the Court dismisses with prejudice the claims of opt-in plaintiffs • Michael Billy, Carol Bongiorno, Lucas Brandt, Azeem Karmally, Joanne Laurenzana, Jose Pena, Robert Pray, and Yu Tang for the same reasons set forth in the Court’s Order of August 19, 2014 (Dkt. #213).