DuVaney v. Delta Airlines, Inc.

• Court: District Court, D. Nevada
• Decided: April 1, 2025
• Docket: 2:21-cv-02186
• Status: Unknown

Opinion

1 UNITED STATES DISTRICT COURT

2 DISTRICT OF NEVADA

3 * * *

4 MARSHA R. DUVANEY, on behalf of Case No. 2:21-cv-02186-RFB-EJY herself and all others similarly situated, 5 Plaintiff, ORDER 6 v. 7 DELTA AIR LINES, INC., and THE 8 ADMINISTRATIVE COMMITTEE OF DELTA AIR LINES, INC., 9 Defendants. 10 11 Pending before the Court is Plaintiff’s Motion to Compel Defendants to Respond Fully to 12 Interrogatory 10 (ECF No. 78). The Court has considered the Motion, Defendant’s Opposition (ECF 13 No. 81), and Plaintiff’s Reply (ECF No. 85). 14 I. Background 15 The instant Motion arises out of a putative class action alleging violations of the Employee 16 Retirement Income Security Act of 1974 (“ERISA”). Plaintiff DuVaney, a former employee of 17 Northwest Airlines of thirty-three years, is a participant in Northwest Airlines’ Pension Plan for 18 Contract Employees (the “Contract Plan”), which is currently sponsored by Defendants. ECF No. 1 19 ¶¶ 1–2, 13. Plaintiff receives a Joint-and-Survivor Annuity (“JSA”) that provides a fixed monthly 20 benefit amount; Plaintiff’s husband, as named beneficiary, is entitled to receive a specified portion 21 of that benefit amount in the event Plaintiff predeceases him. Id. ¶¶ 3, 13. JSAs are provided as an 22 alternative to a Single Life Annuity (“SLA”) in which no benefit is provided to a surviving spouse 23 of the participant, and which therefore result in higher monthly benefits than JSAs. Id. ¶ 4. 24 Although the monthly benefits paid during the life of a participant will be less under a JSA, 25 § 205(d) of ERISA requires that any qualified JSA provide benefits that are the “actuarial equivalent” 26 of benefits that a participant would be entitled to under an SLA. 29 U.S.C. § 1055(d)(1)(B). Plaintiff 27 alleges that Defendants violated § 1055(d) by providing her JSA benefits that are not actuarially 1 a $12,605 reduction in benefits. ECF No. 1 ¶ 77. Plaintiff asserts this reduction is the result of the 2 “unreasonably low conversion factors” used by Defendants in calculating JSA benefits. Id. ¶ 75. 3 Plaintiff further alleges that the use of unreasonably low conversion factors affect not only 4 participants of the Contract Plan, such as herself, but also participants of the Northwest Airlines 5 Pension Plan for Salaried Employees (the “Salaried Plan”) and the Northwest Airlines Pension Plan 6 for Pilot Employees (the “Pilots Plan”).1 Id. ¶¶ 1, 7. 7 Based on the above allegations Plaintiff identifies a putative class that includes all retirees 8 who began receiving a JSA benefit or Qualified Pre-Retirement Survivor Annuity (which is similar 9 to a JSA but pays a benefit to a surviving spouse in the event the participant dies before retirement) 10 under any of the Plans on or after December 10, 2015. Id. ¶ 87. Based on data produced by 11 Defendants, this putative class contains approximately 4,272 members. ECF No. 81-1 ¶ 7. 12 Plaintiff served Defendants with an initial set of interrogatories on June 7, 2024, including 13 one that requested Defendants disclose the monthly amount of the SLA each retiree in the putative 14 class could have received at their Benefit Commencement Date (“SLA at BCD”). ECF No. 78 at 4. 15 This interrogatory was in addition to a prior request for production containing identical language. 16 ECF No. 81-2 at 15. Plaintiff asserts that the SLA at BCD for each putative class member is essential 17 for determining whether they are receiving an actuarially equivalent JSA amount, as well as 18 determining the harm each putative class member suffered. ECF No. 78 at 3. On July 15, 2024, 19 Defendants’ response directed Plaintiff to their previously produced spreadsheet, which Defendants 20 assert contains over forty seven fields of data for each participant and beneficiary (the “Data Set”). 21 Id. at 4. On July 12, 2024, Plaintiff asked for confirmation that the Data Set included each retiree’s 22 SLA at BCD,2 to which Defendants responded confirming the Data Set did not include this 23 information for all participants. ECF No. 81 at 8. 24

25 1 Collectively, the Contract Plan, the Salaried Plan, and Pilot Plan are referred to as the “Plans.” 2 There appears to be disagreement regarding the communication that occurred on July 12, 2024. Plaintiff 26 represents she told Defendants that the information was missing, while Defendants represent that Plaintiff merely asked whether this information was included in the Data Set. ECF Nos. 78 at 5; 81 at 8. This dispute is of no moment as both 27 parties acknowledge this communication took place over a month after Plaintiff’s initial interrogatories were served on 1 After additional discussion between the parties, Defendants provided a supplementary 2 response to the relevant interrogatory explaining that SLA at BCD is not maintained on their system 3 “in a … way … that could be retrieved and exported in a programmatic, automated way using their 4 system’s current functionality.” ECF No. 78-4 at 5. Following a further meet-and-confer, 5 Defendants represent that they engaged their recordkeeper, Conduent, to manually retrieve SLA at 6 BCD for a sample of fifty putative class members and provided this data to Plaintiff, but Plaintiff 7 rejected it. ECF No. 81-1 ¶ 8. Defendants further represent that, based on the time and expense 8 involved in producing this sample, Conduent estimated that producing the SLA at BCD for the 4,272 9 class members would take approximately 355 hours and cost $41,535.3 ECF No. 81-5 ¶ 8. Plaintiff 10 states that Conduent gave the same estimate to her in response to a subpoena. ECF No. 78 at 10. 11 After further failed attempts to resolve this issue, Plaintiff filed the instant Motion to Compel. 12 II. The Parties’ Arguments 13 As a threshold argument, Plaintiff asserts that Defendants have waived any objection to 14 responding to the interrogatory at issue by not raising a specific objection in their initial response. 15 ECF No. 78 at 11. Plaintiff alleges that Defendants offered only “boilerplate objections” without 16 specifically addressing the requested SLA at BCD amounts. Id. at 15-16. Separately, Plaintiff 17 addresses proportionality under Federal Rule of Civil Procedure (“FRCP”) 26(a) asserting disclosure 18 of the SLA at BCD for each purported class member is appropriate because of the public interest in 19 retirement plans, the amount in controversy is likely in the tens of millions, the SLA at BCD exists 20 within Defendants’ systems, Defendants’ significant resources relative to Plaintiff, the importance 21 of the SLA at BCD amounts in proving claims on behalf of the putative class, and the relatively low 22 cost of production compared to the alleged amount in controversy. ECF No. 78 at 12-14. Plaintiff 23 further argues that any burden or expense on Defendants is a result of their own decision to outsource 24 recordkeeping to a third-party, and that poor recordkeeping is no excuse to avoid discovery. Id. at 25 16-19. 26

3 In her Reply Plaintiff asserts Defendants are only quoting the amount Conduent would charge her directly to 27 produce the information. ECF No. 85 at 9. As Defendants’ Response and attached declarations make clear, this is the 1 Defendants argue that the manual retrieval of SLA at BCD for all 4,272 putative class 2 members is unduly burdensome and disproportionate to the needs of the case. ECF No. 81 at 12. 3 Defendants assert they did not waive this objection because, inter alia, Plaintiff’s initial 4 interrogatories did not raise the specific issue of whether SLA at BCD values were available for the 5 identified retirees, and a supplemental response with specific objections was provided once that issue 6 was identified. Id. at 22. In the alternative, Defendants argue that any failure to object should be 7 excused because of their good-faith efforts to promptly respond once Plaintiff raised the issue in her 8 July 12, 2024, communication subsequent to the service of her initial interrogatories on Defendants. 9 Id. at 23. 10 Turning to the merits of their objection, Defendants assert that Plaintiff’s expert could use 11 the Data Set to derive the SLA at BCD for each retiree, and that this is a far less burdensome means 12 than manual retrieval. Id. at 13-15. Defendants further argue that Plaintiff’s asserted damages 13 cannot, without more, outweigh the burden of manual production, and that maintaining SLA at BCD 14 data in useful form is not required by any law and would serve no business purpose. Id. at 17-21. 15 III. Discussion 16 District Courts have broad discretion to control the scope of discovery. Little v. City of 17 Seattle, 863 F.2d 681, 685 (9th Cir. 1988). Parties are permitted to seek discovery of any 18 nonprivileged matter that is relevant and proportional to the needs of the case, with proportionality 19 determined by reference to several factors including “whether the burden or expense of the proposed 20 discovery outweighs its likely benefit.” Fed. R. Civ. P. 26(b)(1). The party seeking to avoid 21 discovery bears the burden of explaining why discovery should be denied. U.S. E.E.O.C. v. Caesars 22 Entertainment, 237 F.R.D. 428, 432 (D. Nev. 2006). 23 Because Defendants do not challenge the relevance of the information being sought, the 24 Court limits its discussion to proportionality. As an initial matter, the Court agrees with Defendants 25 in that, to the extent they neglected to raise a specific objection to the relevant interrogatory, there is 26 good cause to excuse such waiver. See Fed. R. Civ. P. 33(b)(4) (“Any ground not stated in a timely 27 objection is waived unless the court, for good cause, excuses the failure.”) (emphasis added). The 1 had already responded by producing the Data Set and raising specific objections. Compare ECF 2 No. 81-2 at 15-16 with ECF No. 78-2 at 4-5. More importantly, the issue of whether SLA at BCD 3 values for every retiree was included in the Data Set was not raised by Plaintiff until over a month 4 after her initial interrogatories had been served. ECF No. 81-1 ¶¶ 9-10; ECF No. 78-1 ¶ 9. 5 While Plaintiff is of course correct that opposing parties may not “refuse discovery simply 6 by making a boilerplate objection that it is not proportional,” Fed. R. Civ. P. 26 advisory committee’s 7 note to the 2015 amendment, the fact that Plaintiff appears not to have indicated that Interrogatory 8 10 requested anything beyond what Defendants had already disclosed until well after the 9 interrogatory was served means Defendants had little reason to offer a specific objection. The Court 10 therefore finds sufficient good cause exists to excuse Defendants’ failure to offer a specific objection 11 in their initial response to Interrogatory 10 and that their supplemental response qualifies as a timely 12 objection. 13 Turning to proportionality, Defendants’ primary objection is that the time and cost involved 14 in manually retrieving SLA at BCD values for each putative class member would be unduly 15 burdensome. ECF No. 81 at 12-16. Though Plaintiff argues that any burden involved is not undue 16 by citing Holland-Hewitt v. Allstate Life Ins. Co., 343 F.R.D. 154 (E.D. Cal. 2022), the Court finds 17 this contention unpersuasive. It is true that the plaintiff in Holland-Hewitt argued that Allstate could 18 not establish undue burden because the contact information being sought was information Allstate 19 was responsible for tracking on a routine basis. Id. at 164. However, the court did not rule based 20 on this argument; rather, the court based its decision on the fact that Allstate “d[id] not suggest any 21 specific alternative timeline at all; d[id] not describe the actual process or what is involved in 22 preparing this information to produce, generally nor directly in response to the fact this information 23 is already available in anonymized form.” Id. at 174. In contrast, here, Defendants provided specific 24 details of the process required to manually retrieve SLA at BCD, as well as the time and cost 25 involved. ECF No. 81 at 10-11. Even if the court in Holland-Hewitt had based its decision on the 26 reasoning Plaintiff suggests, it would not be applicable to the facts of this case. Unlike an auto 27 insurer’s tracking of contact information as it relates to covered accidents, Defendants have no 1 benefit and who no longer have the option to select an SLA. 4 Plaintiff has not established that 2 retaining such data is routine practice for Defendants and, indeed, Defendants have averred the 3 opposite.5 Id. at 6. 4 For similar reasons, the Court finds Plaintiff’s arguments regarding supposedly poor 5 recordkeeping on Defendants’ part unavailing. The cases cited by Plaintiff all involve parties who 6 failed to maintain in an easily accessible manner information or documents they were either required 7 to maintain or the maintenance of which would be cautioned by competent business practices. See, 8 e.g., Pham v. Wal-Mart Stores, Inc., Case No. 2:11-cv-01148-KJD-GWF, 2011 WL 5508832 (D. 9 Nev. Nov. 9, 2011) (seeking production of injury reports); Union Commercial Services Ltd. v. FCA 10 Int’l Operations LLC, Case No. 16-cv-10925, 2018 WL 558760 (E.D. Mich. Jan 25, 2018) (invoices 11 from sales); Edwards v. Ford Motor Co., Case No. 11-cv-1058-MMA (BLM), 2012 WL 553383 12 (S.D. Cal. Feb. 17, 2012) (standardized reports relating to vehicle component issues). In this case, 13 Plaintiff is asking Defendants to produce a hypothetical benefit amount for an annuity that the 14 putative class members never chose and no longer have the option of choosing. Defendants cannot 15 be faulted for not maintaining this information in readily accessible form for each individual retiree. 16 For these reasons, while the Court agrees some SLA at BCD data may be necessary for Plaintiff to 17 establish her claims on behalf of the class, the Court is not convinced that requiring Defendants to 18 produce these values for the putative class is proportionate at this stage of litigation. 19 Nor has Plaintiff explained why the SLA at BCD value for each of the 4,272 purported class 20 members is necessary at this stage. Plaintiff’s Complaint alleges that the conversion factors used to 21 calculate JSA benefits, as specified in plan documents, are “unreasonably low,” thus resulting in the 22 putative class members receiving monthly benefits that are not actuarially equivalent to that which 23 they would have received under an SLA. ECF No. 1 ¶¶ 74–86. Establishing such a claim does not 24 require calculating the individual SLA at BCD value for each of the 4,272 retirees identified in the 25 Data Set because if Plaintiff’s theory is correct, and the factors utilized by Defendants are inherently

26 4 The Contract Plan states that “[n]o election made after the Annuity Starting Date shall be effective to change the form of pension.” ECF No. 36-2 at 32. 27 5 While Plaintiff cites to several provisions of ERISA for the proposition that Defendants have an obligation to 1 flawed, it is reasonable to infer that such an error will affect most if not every participant who 2 enrolled in a JSA within the specified time period. Plaintiff acknowledges as much in her Complaint. 3 Id. ¶ 76 (“While the amount of the loss suffered will vary depending on the ages of the participant 4 and beneficiary at the time of retirement, all participants receiving JSAs under the Plans are not 5 receiving actuarially equivalent benefits[.]”). 6 However, while production of SLA at BCD data for 4,227 individuals may be unduly 7 burdensome for Defendants at this stage, the Court does not find Defendants’ proposed alternative 8 appropriate either. Simply leaving Plaintiff’s experts to derive SLA at BCD values using the Data 9 Set would be to present Defendants with another avenue for attacking Plaintiff’s eventual motion 10 for class certification.6 Defendants characterize Plaintiff as admitting she could “back into” the SLA 11 at BCD using information provided in the Data Set. ECF No. 81 at 16 (quoting ECF No. 78 at 17). 12 However, Plaintiff in the cited portion of her brief only offers that she could “estimate” these values. 13 ECF No. 78 at 17 (emphasis added). In contrast to this estimate, Defendants acknowledge that 14 Conduent could produce the actual value by retrieving each retiree’s benefit commencement 15 package. ECF No. 81 at 10. Therefore, Defendants’ relative access to this data compared to 16 Plaintiff’s must be weighed against the burden in producing it. 17 In contexts such as this it is not unusual for courts to require Defendants to produce data for 18 a representative sample of the putative class so as to provide enough information to support class 19 certification. Romo v. GMRI, Inc., Case No. EDCV-12-0715-JLQ, 2013 WL 11310656, at *5 (C.D. 20 Cal. Jan. 25, 2013) (“Sampling is in fact a commonly used and reasonable compromise at the class 21 certification stage of a case.”). In Tinnin v. Sutter Valley Med. Found., Case No. 1:20-cv-00482- 22 NONE-EPG, 2021 WL 1315435 (E.D. Cal. Apr. 7, 2021), for example, the plaintiff sought 23 production of electronic payroll and timekeeping data for 7,400 employees in native format, each of 24 which would have to be manually retrieved by the employer. Id. at *3. The Eastern District of

25 6 Defendants’ reliance on James v. UMG Recordings, Inc., Case No. 11-cv-01613-SI (MEJ), 2013 WL 5978322 (N.D. Cal. Nov. 8, 2013), on this point is unavailing. Defendant in that case argued that producing the information 26 sought would require reprogramming its database and creating new software, id. at *2, whereas Defendants’ declarations specify a method of manual retrieval that, while costly and time consuming, does not require such drastic measures. 27 Further, the plaintiffs in James admitted they could extrapolate the requested information from materials currently 1 California found that such a request was unduly burdensome at the class certification stage, and 2 instead ordered the employer to produce a representative sample, leaving plaintiff the option to 3 request additional discovery after the class had been certified. Id. at *5. 4 While this is not a Fair Labor Standards Act collective action, the Court nonetheless finds a 5 representative sample is appropriate and that a sample size of approximately five percent of the 6 retirees in the Data Set, or approximately 215 individuals, is sufficient to give Plaintiff the data 7 necessary to argue class certification is appropriate. The Court is cognizant of the fact that 8 Defendants previously offered to provide such a sample to Plaintiff and, in fact, did direct Conduent 9 to produce the SLA at BCD for 50 individuals in the Data Set, which Plaintiff rejected.7 However, 10 Plaintiff at no point explains why a representative sample is not sufficient for purposes of certifying 11 the class. In the absence of a specific bases to reject a sample, the Court finds there is no reason a 12 sample will prevent Plaintiff from proceeding to certification. If Plaintiff finds the sample is for 13 some reason insufficient, Plaintiff is invited to raise this issue with the undersigned. 14 Because Defendants have already provided SLA at BCD values for 50 of the identified 15 retirees, the Court only requires Defendants to produce the difference between 215 individuals and 16 what Defendants have already produced. To ensure the individuals selected are representative of 17 the putative class as a whole, and to avoid any potential gamesmanship in the individuals selected, 18 Defendants must provide the SLA at BCD for every 25th retiree in the Data Set, continuing until 19 165 individuals are selected. With the 50 individuals for whom this data has already been produced, 20 Plaintiff will have a total sample of 215 retirees. Defendants must identify the individuals selected 21 in such a manner that allows Plaintiff, with reference to the Data Set previously produced, to confirm 22 that the sample consists of every 25th retiree. The Court finds that the information provided by this 23 sample will likely be sufficient to support a motion for class certification. If Plaintiff is successful 24 in certifying the class, she will retain the right to the SLA at BCD values for each individual class 25 member for purposes of calculating damages. 26 27 1 IV. Order 2 Accordingly, IT IS HEREBY ORDERED that Plaintiff’s Motion to Compel (ECF No. 78) 3 is GRANTED IN PART to the extent that Defendants must disclose SLA at BCD values for the 4 representative sample of putative class members described above. This sample shall be in addition 5 to the sample of 50 putative class members Defendants previously provided Plaintiff, and consist of 6 data for an additional 165 putative class members. Of the 4,272 retirees included in the Data Set, 7 Defendants must provide the SLA at BCD value for every twenty-fifth retiree to Plaintiff no later 8 than May 1, 2025. 9 IT IS FURTHER ORDERED the Defendants must identify to Plaintiff the members included 10 in the representative sample so as to allow Plaintiff to confirm that every twenty-fifth retiree in the 11 Data Set was selected and that Defendants did not intentionally select the members in a manner that 12 benefited their position. 13 IT IS FURTHER ORDERED that the parties must meet and confer after the sample data has 14 been disclosed to discuss any remaining issues arising from the sample data. This meet-and-confer 15 must take place no later than May 22, 2025. 16 IT IS FURTHER ORDERED that should the disputes regarding the sufficiency of the sample 17 data for class certification purposes remain following the meet and confer, the parties must file a 18 joint request for a hearing at which time the Court will address the remaining disputes. If the sample 19 provided is sufficient to allow Plaintiff to proceed to a motion for class certification, the parties must 20 file a status report providing a briefing schedule for class certification and proposed discovery plan 21 that will follow the decision on certification. 22 IT IS FURTHER ORDERED that discovery deadlines in the Court’s most recent Scheduling 23 Order (ECF No. 92) are vacated pending the outcome of the production ordered and anticipated 24 motion for certification that will follow. 25 Dated this 1st day of April, 2025. 26