HALEY v. DELTA AIR LINES INC (2026)

ORDER

This matter is before the Court on Plaintiffs' motion for class certification. (Doc. 183.) After careful consideration, the Court enters the following order.

I. Introduction

This is a putative class action brought by current and former Delta Airlines pilots who served in the U.S. Armed Forces during their employment at Delta. Plaintiffs contend that Delta violated The Uniformed Service Employment and Reemployment Rights Act (“USERRA”), 38 U.S.C. § 4301 et seq., by failing to provide paid leave to pilots for short-term military absences, while simultaneously compensating pilots for other types of short-term absences such as jury duty leave, bereavement leave, and sick leave. Plaintiffs seek to pursue their USERRA claim against Delta on behalf of themselves and other Delta pilots by certifying two classes: (1) an Injunction Class consisting of current Delta pilots who also serve in the military; and (2) a Damages Class consisting of pilots who previously took military leave of 30 days or less.

After carefully parsing the parties' arguments for and against class certification, the Court concludes that certification of the two classes proposed by Plaintiffs is appropriate under Federal Rule of Civil Procedure 23. For starters, Delta applies a uniform leave policy to its pilots through collective bargaining agreements. The core question posed by Plaintiffs' USERRA claim therefore is whether Delta's uniform leave policy violates USERRA. To answer that question, USERRA requires a jury to compare short-term military leave (which Delta does not compensate) to other forms of short-term leave (which Delta does compensate) across three factors: (1) the duration of the leave, (2) the purpose of the leave, and (3) the ability of employees to choose when to take the leave. Myrick v. City of Hoover, Alabama, 69 F.4th 1309, 1319 (11th Cir. 2023).

In the Court's view, a jury can determine whether short-term military leave is comparable to other forms of leave on a classwide basis, using evidence that is common to all Delta pilots who are (or were) members of the military and took short-term military leave. Because the comparability inquiry will resolve Delta's liability to all putative class members in one fell swoop, litigating Plaintiffs' USERRA claim through a class action offers substantial advantages over requiring every class member to bring an individual case. That is true even though, as Delta emphasizes, damages for each pilot who took short-term military leave will have to be calculated individually. At bottom, USERRA's comparability inquiry is the principal issue that will drive the resolution of this case, and it predominates over the individual issues that Delta has identified. Accordingly, Plaintiffs' motion for class certification is due to be granted.

II. Background

Plaintiffs are current or former Delta employees who serve (or previously served) in the U.S. Armed Forces while employed at Delta. (Second Amended Complaint, Doc. 182 ¶¶ 9-11.) Delta has a uniform employment policy of not compensating pilots for short-term military absences, while simultaneously providing paid leave for other types of short-term absences. Specifically, Delta has a “Pilot Working Agreement” (“PWA”)—essentially, a collective bargaining agreement—which governs the employment of its pilots. (First Downes Decl., Doc. 118-6 ¶ 4.) Since at least 2007, the PWAs have provided compensation for jury duty leave, bereavement leave, and sick leave, but not for short-term military leave (“STML”) of 30 consecutive days or fewer. (Id.) Plaintiffs contend that Delta's leave policy violates Section 4316 of USERRA by failing to provide paid leave to employees who take STML. (Doc. 182 ¶¶ 70-80.)

Before addressing the pending motion, the Court summarizes the relevant procedural history. On March 29, 2022, then-Chief Judge Timothy Batten denied Delta's motion to dismiss Plaintiffs' USERRA claim. (Doc. 42.) After a period of discovery, on June 12, 2023, Plaintiffs filed their first motion for class certification. (Doc. 118.) To the surprise of both the Court and Delta, Plaintiffs' motion sought the appointment of Benjamin Best, who was at that time not a named plaintiff in this matter, as the sole class representative. (Id.) The Court then held evidentiary hearings on Plaintiffs' motion for class certification on November 28, 2023, and January 17, 2024. (Doc. 135, 151.) But because Delta had not had the opportunity to depose or otherwise obtain discovery with respect to Mr. Best, after the evidentiary hearings, the Court reopened class discovery and directed the parties to submit supplemental briefing. (Doc. 149, 153.)

Following this additional discovery and briefing, on December 22, 2025, the Court denied without prejudice Plaintiffs' first motion for class certification on standing and procedural grounds. (Doc. 181.) Among other things, the Court found that neither of the named Plaintiffs had standing to pursue injunctive relief, and was concerned that, as an absent putative class member, Mr. Best could not cure the standing deficiency. (Id. at 3-12.) However, the Court provided Plaintiffs with leave to amend their complaint to add Mr. Best as a plaintiff and renew their motion for class certification. (Id. at 12-16.)

Plaintiffs have now amended their complaint and, on January 12, 2026, filed their renewed motion for class certification. Plaintiffs bring a single claim alleging that Delta violated USERRA by failing to provide compensation to pilots who take short-term military leave on the same basis as pilots who take comparable forms of non-military leave. (Doc. 182 ¶¶ 70-80.) In their renewed motion, Plaintiffs seek certification of the following two classes:

(1) All current and former pilots who work or worked for Delta Air Lines, Inc. (“Delta”) in the United States or its territories or possessions who took short-term military leave (i.e., a military leave of 30 consecutive days or fewer) from their employment from Delta as a pilot from April 30, 2007 to the date of judgment in this action, as reflected by Delta's military leave data (the “Damages Class”); and

(2) all pilots currently employed by Delta who are also currently members of the uniformed services (the “Injunction Class”).

(Doc. 183 ¶ 1.) Plaintiffs request that Mr. Best, who is now a named Plaintiff, serve as the sole class representative for both classes. (Id. ¶ 3.) Mr. Best is currently a pilot at Delta and simultaneously serves as a Major in the Michigan Air National Guard of the United States Air Force. (Transcript of Jan. 17, 2024, Hearing, Doc. 159 at 13-14 (pp. 12-13).)¹

Plaintiffs' renewed motion for class certification is ripe for review. In resolving Plaintiffs' renewed motion, the Court considers the parties' briefing and evidence submitted in connection with the first motion for class certification, as well as the briefing and evidence submitted with the instant motion. (See Doc. 118, 123, 126, 127, 128, 129, 138, 139, 143, 147, 148, 154, 159, 169, 170, 171, 177, 178, 179, 183, 185, 190.)

III. Legal Standard

“Before a district court may grant a motion for class certification, a plaintiff seeking to represent a proposed class must establish that the proposed class is adequately defined and clearly ascertainable.” Little v. T-Mobile USA, Inc., 691 F.3d 1302, 1304 (11th Cir. 2012) (cleaned up). If this threshold condition is satisfied, the plaintiff must then show that the requirements of Federal Rule of Civil Procedure 23(a) and 23(b) are met. See id.

Rule 23(a) requires a plaintiff to demonstrate 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). “These four requirements commonly are referred to as ‘the prerequisites of numerosity, commonality, typicality, and adequacy of representation,’ and they are designed to limit class claims to those ‘fairly encompassed’ by the named plaintiffs' individual claims.” Piazza v. Ebsco Indus., Inc., 273 F.3d 1341, 1346 (11th Cir. 2001) (quoting Gen. Tel. Co. of S.W. v. Falcon, 457 U.S. 147, 156 (1982)).

“In addition to establishing the Rule 23(a) requirements, ․ plaintiff[s] must also establish that the proposed class satisfies at least one of the three requirements listed in Rule 23(b).” Little, 691 F.3d at 1304. Rule 23(b) provides that a class action may be maintained only where one of the three following requirements is met:

(1) prosecuting separate actions by or against individual members of the class would create a risk of prejudice to the party opposing the class or to those members of the class not parties to the subject litigation, see Fed. R. Civ. P. 23(b)(1);

(2) the party opposing the class has acted or refused to act on grounds that apply generally to the class, so that final injunctive or declaratory relief is appropriate respecting the class as a whole, see Fed. R. Civ. P. 23(b)(2); or

(3) the court finds that the questions of law or fact common to the members of the class predominate over any questions affecting only individual members, and a class action is superior to other available methods for fair and efficient adjudication of the controversy, see Fed. R. Civ. P. 23(b)(3).

Here, Plaintiffs seek certification of one subclass under Rule 23(b)(2) and another under Rule 23(b)(3).

“Rule 23 does not set forth a mere pleading standard.” Wal-Mart Stores, Inc. v. Dukes, 564 U.S. 338, 350 (2011). Since class actions are “an exception to the usual rule that litigation is conducted by and on behalf of the individual named parties only,” the party seeking to certify a class action bears the burden of establishing compliance with Rule 23's requirements through “evidentiary proof.” See Comcast Corp. v. Behrend, 569 U.S. 27, 33 (2013) (cleaned up); Green-Cooper v. Brinker Int'l, Inc., 73 F.4th 883, 888 (11th Cir. 2023); see also Halliburton Co. v. Erica P. John Fund, Inc., 573 U.S. 258, 275 (2014) (“[P]laintiffs wishing to proceed through a class action must actually prove—not simply plead—that their proposed class satisfies each requirement of Rule 23[.]”).

As such, the district court must “conduct a ‘rigorous analysis’ to determine whether the movant carried his burden[.]” Brown v. Electrolux Home Prods., Inc., 817 F.3d 1225, 1234 (11th Cir. 2016) (citation omitted). Although district courts do not determine the merits of plaintiffs' claims at the class certification stage, they “can and should consider the merits of the case to the degree necessary to determine whether the requirements of Rule 23 will be satisfied.” Vega v. T-Mobile USA, Inc., 564 F.3d 1256, 1266 (11th Cir. 2009) (quoting Valley Drug Co. v. Geneva Pharms., Inc., 350 F.3d 1181, 1188 n.15 (11th Cir. 2003)). Ultimately, “[t]he presumption is against class certification,” and a district court that has doubts about whether Rule 23's requirements have been met should refuse certification. Brown, 817 F.3d at 1233-34.

IV. Statutory Framework

“Military reservists play a vital role in our nation's defense policy.” Myrick v. City of Hoover, Alabama, 69 F.4th 1309, 1312 (11th Cir. 2023). “[T]o mitigate the employment disadvantages that stem from non-career military service[,]” Congress enacted the Uniformed Services Employment and Reemployment Rights Act of 1994 (“USERRA”), 38 U.S.C. § 4301 et seq. Id. at 1314; see Clarkson v. Alaska Airlines, Inc., 59 F.4th 424, 428 (9th Cir. 2023) (“For over sixty years, our nation has encouraged military service by continually easing the burden on servicemembers who must juggle military duties with civilian jobs.”). USERRA serves three primary purposes: (1) “[T]o encourage service in the uniformed services by eliminating or minimizing the disadvantages to civilian careers and employment which can result from such service”; (2) “to minimize the disruption to the lives of persons performing service ․ as well as to their employers ․ by providing for the prompt reemployment of such persons upon their completion of such service”; and (3) “to prohibit discrimination against persons because of their service in the uniformed services.” 38 U.S.C. § 4301(a).

Among other things, USERRA “forces employers to give employees on military leave the same rights and benefits provided to similarly situated employees on non-military leave[.]” Myrick, 69 F.4th at 1314. Specifically, Section 4316(b) of USERRA provides that

a person who is absent from a position of employment by reason of service in the uniformed services shall be ․ entitled to such other rights and benefits not determined by seniority as are generally provided by the employer of the person to employees having similar seniority, status, and pay who are on furlough or leave of absence ․

38 U.S.C. § 4316(b). The Department of Labor's (“DOL”) implementing regulation for Section 4316(b) requires that “[i]f the non-seniority benefits to which employees on ․ leave of absence are entitled vary according to the type of leave, the employee must be given the most favorable treatment accorded to any comparable form of leave ․” 20 C.F.R. § 1002.150(b) (emphasis added). The regulation also explains how to determine whether certain types of leave are comparable to military leave:

To determine whether any two types of leave are comparable, the duration of the leave may be the most significant factor to compare. For instance, a two-day funeral leave will not be “comparable” to an extended leave for service in the uniformed service. In addition to comparing the duration of the absences, other factors such as the purpose of the leave and the ability of the employee to choose when to take the leave should also be considered.

Id.

The Eleventh Circuit has explained that these provisions “set forth a two-step process for determining which benefits are available to employees on military leave.” Myrick, 69 F.4th at 1314. “First, the military employee must identify a group of non-military employees ‘having similar seniority, status, and pay who are on ․ leave of absence[.]’ ” Id. (quoting 38 U.S.C. § 4316(b)(1)(B)). “Second, the military employee must prove that those employees took a form of non-military leave that is comparable to military leave.” Id. at 1315 (citing 20 C.F.R. § 1002.150(b)). “Three factors inform [the] comparison between leaves: (1) the duration of the leave, (2) the purpose of the leave, and (3) the ability of employees to choose when to take the leave.” Id. at 1319. A military employee who satisfies both requirements “is entitled to the same benefits that the similarly situated employees received while on the comparable form of leave.” Id. at 1315.

V. Motion to Partially Exclude Expert Opinions of Dr. Lee

Before turning to the merits of Plaintiffs' motion for class certification, the Court must address Plaintiffs' motion to partially exclude the expert opinions of Dr. Darin Lee, who serves as an expert witness for Delta. (Doc. 127.) Plaintiffs move to exclude two categories of Dr. Lee's opinion testimony. First, Plaintiffs seek to exclude Dr. Lee's opinion that Delta's leave data is insufficient to accurately determine the dates that its employees took military leave. (Doc. 127-1 at 8-15.) Second, Plaintiffs seek to exclude Dr. Lee's opinions that the Damages Class definition's limitation to leaves of 30 days or fewer is “arbitrary” and that there is “substantial heterogeneity” in the patterns of short-term military leave taken at Delta. (Id. at 15-20.)

In determining the reliability and relevance of proffered expert testimony, the Court performs a “gatekeeping function.” Daubert v. Merrell Dow Pharm., Inc., 509 U.S. 579, 589 n.7 (1993). Under Rule 702 of the Federal Rules of Evidence, “expert testimony is admissible if (1) the expert is qualified to testify regarding the subject of the testimony; (2) the expert's methodology is sufficiently reliable as determined by the sort of inquiry mandated in Daubert; and (3) the expert's testimony will assist the trier of fact in understanding the evidence or determining a fact at issue.” Chapman v. Procter & Gamble Distrib., LLC, 766 F.3d 1296, 1304-05 (11th Cir. 2014) (citing United States v. Frazier, 387 F.3d 1244, 1260 (11th Cir. 2004)) (internal quotation marks omitted); see Fed. R. Evid. 702. The proponent of the expert opinion bears the burden of showing that each of these requirements is met. Chapman, 766 F.3d at 1304-05.²

In Daubert, the Supreme Court identified four factors to guide district courts in their evaluations of the reliability of expert testimony:

(1) whether the expert's methodology has been tested or is capable of being tested;

(2) whether the theory or technique used by the expert has been subjected to peer review and publication;

(3) whether there is a known or potential error rate of the methodology; and

(4) whether the technique has been generally accepted in the relevant scientific community.

United Fire & Cas. Co. v. Whirlpool Corp., 704 F.3d 1338, 1341 (11th Cir. 2013) (per curiam) (citing Daubert, 509 U.S. at 593-94). But these factors are not “a definitive checklist or test,” Daubert, 509 U.S. at 593, and they are always to be “applied in case-specific evidentiary circumstances,” United States v. Brown, 415 F.3d 1257, 1266 (11th Cir. 2005).

District courts have “considerable leeway in deciding in a particular case how to go about determining whether particular expert testimony is reliable.” Kumho Tire Co., Ltd. v. Carmichael, 526 U.S. 137, 152 (1999). In assessing the reliability of expert testimony, the Court's focus “must be solely on principles and methodology, not on the conclusions that they generate.” Chapman, 766 F.3d at 1305 (quoting Daubert, 509 U.S. at 594-95) (emphasis omitted). “It is ‘proper’ and ‘necessary’ for the trial judge ‘to focus on the reliability’ of a proffered expert's ‘sources and methods.’ ” Id. at 1306 (quoting Kilpatrick v. Breg, Inc., 613 F.3d 1329, 1335 (11th Cir. 2010)).

At this stage of proceedings, the Court denies Plaintiffs' motion to exclude without prejudice. With respect to the first category of challenged opinion testimony, Plaintiffs contend that Dr. Lee's conclusion regarding the accuracy of Delta's leave data appears to rest, at least in part, on a faulty legal interpretation of USERRA. (Doc. 127-1 at 8-9.) Specifically, Dr. Lee notes that he was

instructed by counsel for Delta that the leave of absence from employment that is covered under and protected by USERRA is the entire period of time that the employee is performing service in the uniformed service because of a leave event and the additional period of time the employee is unavailable to work based on the military event, regardless of whether that entire period spans scheduled work days and/or scheduled days off.

(Dr. Lee Expert Report, Doc. 118-11 ¶ 8(a).) Plaintiffs argue that Dr. Lee's assumption contradicts Section 4316(b)(1) of USERRA.

However, the Court largely agrees with this aspect of Delta's interpretation of USERRA. Section 4316(b)(1) states that “a person who is absent from a position of employment by reason of service in the uniformed services shall be-- (A) deemed to be on furlough or leave of absence while performing such service; and (B) entitled to such other rights and benefits ․ as are generally provided by the employer” to similarly situated employees on a leave of absence. 38 U.S.C. § 4316(b)(1). Plaintiffs emphasize the phrase “absent from a position of employment,” arguing that employees are only “absent” if they are not present when they are scheduled or expected to be at their employment. (Doc. 127-1 at 10-11.) Plaintiffs provide the example of an employee who works a Monday through Friday schedule and is ordered to perform military drills on a Saturday. (Id. at 11.) Plaintiffs note that this employee would not be “absent from [her] position of employment by reason of [military] service[,]” because she was not expected to be at her position of employment on a Saturday. (Id.) For purposes of this discussion, the Court will accept this particular hypothetical posited by Plaintiffs. Because the employee's job did not require her to be at work on a Saturday, there is a strong argument that she was never “absent from [her] position of employment” within the meaning of USERRA. 38 U.S.C. § 4316(b)(1). Dr. Lee's testimony to the contrary therefore may conflict with USERRA. (See Doc. 127-3 at 6-7 (pp.62-63) (testifying that a pilot who works a Monday-to-Friday schedule but performed military service on Saturday took a military leave of absence under USERRA).).

That said, even if the hypothetical Monday-to-Friday employee was considered “absent” from their employment on a Saturday under USERRA, they would not be entitled to any additional pay or benefits as a result. An employee does not receive additional pay or benefits simply for conducting their own affairs on a day off for which they would not ordinarily receive pay. And Plaintiffs do not seek damages for military service that Delta pilots performed on scheduled days off or vacation days. As such, this hypothetical example of the Monday-Friday employee who performs military service on a Saturday has little relevance to the Court's class certification analysis.

Instead, the thrust of Dr. Lee's analysis and Delta's argument regarding the accuracy of military leave dates reflected in Delta's data focuses on a different factual scenario. Dr. Lee contends that Delta's self-reported military leave data is unreliable for purposes of defining the class (and calculating damages) because certain pilots may have reported multiple periods of STML shorter than 30 days, even though they actually performed military service for a continuous period longer than 30 days. (See Doc. 118-11 ¶¶ 13-35; Doc. 128 at 10-16.) This argument rests on the assumption that one continuous, uninterrupted period of military leave should be counted using the start date and end date of that military service, even if it was self-reported to Delta as a series of shorter STML periods. (Doc. 128 at 10-16.) Consider, for instance, a pilot who performed military service for a continuous period of 35 days but reported their military leave to Delta as several shorter periods of STML because they had scheduled days off in that 35-day period for which they did not need to report military leave. Delta argues that, under USERRA, the Court should consider that pilot's leave as a 35-day period of military leave, notwithstanding that it appears in Delta's leave data as multiple shorter periods of STML. (Id.) This issue is significant because a 35-day military leave would not satisfy the class definition proposed by Plaintiffs, but a 35-day leave broken up into shorter periods of 30 days or less might appear to fall within the class definition.

While it disagrees with Delta's ultimate arguments regarding the propriety of class certification, the Court finds this portion of Delta's interpretation of USERRA unobjectionable. Section 4316(b)(1) states that a person who is absent from their position of employment “shall be ․ deemed to be on furlough or leave of absence while performing such service[.]” 38 U.S.C. § 4316(b)(1) (emphasis added). The implication of this clause is that, once a person leaves their position of employment for military service, they are deemed to be on a leave of absence for the full length of time during which they are “performing such service[.]” Such a reading is consistent with the Eleventh Circuit's decision in Myrick, which explained that when a servicemember is “absent for active-duty service in the United States military for a continuous, uninterrupted period of service, ․ his leave is best viewed as a single period.” Myrick, 69 F.4th at 1320 n.7.³

Plaintiffs do not appear to dispute the principle that a continuous, uninterrupted period of military leave should be measured using its start and end date, notwithstanding how that leave is reflected in Delta's leave data. For instance, Plaintiffs state that

if a period of military service produces an absence from civilian employment, that period of service (and statutorily protected travel and reemployment time) is protected under USERRA 4316(b) (subject to the notice requirements and other provisions of the statute) and the employee is on a military leave of absence until they present for reemployment in accordance with 38 U.S.C. § 4312.

(Doc. 129 at 9 (emphasis added).) Moreover, with respect to Mr. Best, Plaintiffs appear to accept that a “one 39 day, long-term period of [military] leave” he took in 2017 should be treated as a single period of leave under USERRA, even though it was recorded as multiple periods of STML in Delta's leave data. (Doc. 171 at 17.)

Except for the Monday-to-Friday employee hypothetical discussed above, the Court finds that the legal assumption upon which the challenged portion of Dr. Lee's opinion testimony relies is consistent with USERRA. It is therefore not inconsistent with USERRA for Dr. Lee to argue that Delta's self-reported leave data is unreliable because it may record longer, continuous periods of military leave as multiple shorter periods of STML. The Court declines to exclude the subject portion of Dr. Lee's opinion testimony under Daubert and fully considers that testimony in the class certification analysis that follows.

With respect to the second category of Dr. Lee's opinion testimony, the Court does not find the parties' dispute material to the resolution of Plaintiffs' motion for class certification. For instance, Plaintiffs take issue with Dr. Lee's statement that the class definition is “arbitrary.” (Doc. 127-1 at 15-16; see Doc. 118-11 ¶ 8(e), 8(h).) As Plaintiffs point out, it is ultimately the Court's role to determine whether the class definition satisfies the requirements of Federal Rule of Civil Procedure 23. At the same time, Dr. Lee's testimony underlying his “arbitrariness” finding—namely, that many instances of military leave fall on both sides of the 30-day STML cutoff proposed by Plaintiffs—is sufficiently reliable to pass muster under Daubert, as Plaintiffs do not appear to dispute. (Doc. 127-1 at 15 n.4.) Even though the Court does not find persuasive Dr. Lee's statement that the class definition is “arbitrary” because some number of pilots who took 31-day periods of STML will be excluded from the class, it is unnecessary to exclude his testimony from consideration under Daubert, especially when there is no dispute as to the underlying statistical observation.

The Court agrees with Plaintiffs that Dr. Lee has not adequately defined what he means by “substantial” or “extreme” heterogeneity. (Id. at 19-20; see Doc. 118-11 ¶¶ 44-62.) At his deposition, Dr. Lee was asked, “What do you mean by ‘substantial heterogeneity’?” (Lee Dep., Doc. 127-3 at 12 (p.118).) He answered, “Well, when I look at [the data], I don't see a little heterogeneity, I see a lot of heterogeneity. So, you know, again, it's not a -- there is a lot of heterogeneity. That's all – that's what I can say.” (Id.) Dr. Lee also conceded that he did not calculate statistical measures for heterogeneity, for instance, by analyzing mathematical variance and standard deviations across the leave data. (Id. at 14 (p.121).) As such, it appears unlikely that Dr. Lee has employed a statistical methodology that is widely accepted in the scientific community to reach his conclusion that there is “substantial heterogeneity” or “extreme variance” in the leave data.

That said, at this stage of the proceedings, it is unnecessary for the Court to make a determinative ruling as to the admissibility of this portion of Dr. Lee's testimony. Dr. Lee's heterogeneity analysis has little, if any, bearing on whether class certification is appropriate here. Class certification in this case hinges on whether Plaintiffs' USERRA claim and the requisite comparability inquiry it calls for can be proven through classwide evidence. The statistical heterogeneity of Delta's leave data does not drive the resolution of that question. In fact, Dr. Lee has not even attempted to compare the statistical heterogeneity of STML to other forms of non-military leave. Delta appears to concede that Dr. Lee's opinions regarding heterogeneity are “largely irrelevant, as Delta has not opposed class certification based on the heterogeneity of the leaves[.]” (Doc. 128 at 7.)

The Court will save for another day the question of whether these portions of Dr. Lee's expert report and testimony should be excluded under Daubert. Plaintiffs' motion to partially exclude Dr. Lee's expert opinions (Doc. 127) is denied without prejudice. Plaintiffs, however, may renew their motion if any of Dr. Lee's challenged opinions become pertinent at a later stage of the case.

VI. Discussion of Class Certification Requirements

With those preliminary matters settled, the Court addresses the substantive requirements for class certification.

A. Ascertainability

Before turning to the enumerated Rule 23(a) requirements, the Court must decide whether the proposed classes are “adequately defined and clearly ascertainable.” Little, 691 F.3d at 1304. A class is “clearly ascertainable” if its membership is “capable of being determined” by objective criteria. Cherry v. Dometic Corp., 986 F.3d 1296, 1303 (11th Cir. 2021) (quoting Ascertain, Webster's New International Dictionary (3d ed. 1993); Ascertainable, Webster's New International Dictionary (3d ed. 1993)) (internal quotations marks omitted). Ascertainability does not require proof of administrative feasibility. Id. at 1303-04. Indeed, membership in a class “can be capable of determination without being capable of convenient determination.” Id. at 1303.

The Court finds that both classes proposed by Plaintiffs are adequately defined and clearly ascertainable. With respect to the proposed Damages Class, Delta's military leave data preliminarily identifies pilots who took STML—defined as military leave of 30 consecutive days or fewer—during the relevant period. Plaintiffs' economic expert, Dr. Robert J. Cavazos, identified 119,285 instances of STML taken by 3,361 pilots during the period from January 1, 2007, to September 30, 2022. (Dr. Cavazos Expert Report, Doc. 118-5 ¶ 16.) The proposed Injunction Class is even more straightforward. The class of pilots currently employed by Delta who are also serving in the military is both well-defined and “capable of determination.”

Delta contends that neither proposed class is ascertainable, but its argument appears to rest on administrative feasibility rather than whether the proposed class is capable of determination. (Doc. 123 at 18-19.) Specifically, Delta notes that to determine whether a pilot is in the Damages Class, “the Court must be able to determine when the period of military leave begins and ends.” (Id. at 19.) According to Delta, “pilots frequently do not report their entire period of military leave; as a result, it is not possible to determine who is in the class using objective criteria and available data.” (Id.) In other words, Delta, relying on Dr. Lee's opinion, suggests that because pilots self-report their military leave and may underreport the length of certain periods of military leave, Delta's military leave data is unreliable. See supra at 14-21 (discussing Dr. Lee's opinion testimony addressing this issue).

This argument focuses on administrative feasibility—the ease with which class members can be identified—rather than ascertainability—whether the class is capable of determination at all. In any event, the Court finds that it is administratively feasible to identify the members of the proposed Damages Class, notwithstanding Delta's concerns about the reliability of the self-reported leave data. As is undisputed, the military maintains “Point Credit Accounting and Reporting System” (“PCARS”) data that tracks when servicemembers perform military service. (Doc. 170 at 15; Doc. 171 at 17-18; see Doc. 118-11 at 19 n.14.) If Plaintiffs prevail on the merits of their USERRA claim, class members' PCARS data can be used to verify the periods of military leave reported to Delta. To the extent that Delta contends that the Damages Class is not ascertainable due to the potential need to cross-check potential class members' Delta leave records with their PCARS data, the Court disagrees. As several courts have held, “the need to review individual files to identify [class] members [is] not [a] reason[ ] to deny class certification.” Young v. Nationwide Mut. Ins. Co., 693 F.3d 532, 539 (6th Cir. 2012) (collecting cases); see Byrd v. Aaron's Inc., 784 F.3d 154, 170-71 (3d Cir. 2015) (“There will always be some level of inquiry required to verify that a person is a member of a class[.] ․ Such a process of identification does not require a ‘mini-trial,’ nor does it amount to ‘individualized fact-finding,’ and ․ indeed must be done in most successful class actions.” (quotation omitted)).

Moreover, the Court finds Delta's arguments regarding the alleged unreliability of PCARS data to be unpersuasive. Delta contends that Mr. Best's PCARS data is unreliable “because it conflicts with both Delta's leave data and Best's military flight logs.” (Doc. 170 at 15.) But Delta only identifies a single, one-day discrepancy between Mr. Best's PCARS data and his military flight logs—namely that his PCARS data recorded military service on December 6, 2023, but his flight logs showed that he flew on December 7, 2023. (Id.) The Court will require more than a single, one-day discrepancy before broadly holding official military service records unreliable.

While the Court recognizes that putative class members' PCARS records may conflict to some degree with Delta's self-reported leave data, that is precisely the verification role that the PCARS data is expected to play. As explained above, Delta has raised a concern that its leave data is unreliable and does not reflect putative class members' true dates of military service because it is self-reported by its pilots. That is where the PCARS data comes in. A factfinder can and likely would disregard self-reported leave data to the extent that it conflicts with official military service records like PCARS. It is also far from clear, based on the record before the Court, that the discrepancies between Delta's leave data and PCARS will be insurmountable. For instance, only 5 days of the STML that Mr. Best reported to Delta, representing approximately 3.7% of his total STML days, were not reflected in his PCARS data. (See Third Downes Decl., Doc. 171-1 ¶ 2 (“Cross referencing Best's short-term military leave entries in Delta's data and Best's PCARS, I identified five days of short-term military leave that were not reflected by a corresponding service record in his PCARS, representing approximately 3.7% of Best's short-term military leave days as reflected in Delta's data.”).); see also Huntsman v. Southwest Airlines Co., No. 19-CV-00083-PJH, 2021 WL 391300, at *13 (N.D. Cal. Feb. 3, 2021) (certifying a USERRA class notwithstanding a 23% error rate between the class representative's military records and self-reported military leave dates).

With respect to the Injunction Class, Delta argues that the “Injunction Class suffers from the same issues” as the Damages Class, but it does not explain how that is so. (Doc. 123 at 20.) Unlike the Damages Class, the proposed definition of the Injunction Class does not distinguish between short-term and long-term military leave. Rather, it simply defines the class as people meeting two objective criteria—those (1) currently employed by Delta as a pilot; and (2) currently serving in the military. Such a class is certainly capable of determination.

As such, the Court finds that both the Damages Class and Injunction Class proposed by Plaintiffs are adequately defined and clearly ascertainable. Plaintiffs have also shown that there is an administratively feasible way to identify class members using objective criteria.

B. Rule 23(a) Requirements

1. Numerosity

Rule 23(a)(1) requires Plaintiffs to show that “the class is so numerous that joinder of all members is impracticable.” See Vega, 564 F.3d at 1266. The general rule in the Eleventh Circuit for meeting the numerosity requirement is that “less than twenty-one is inadequate, more than forty adequate, with numbers between varying according to other factors.” Cox v. Am. Cast Iron Pipe Co., 784 F.2d 1546, 1553 (11th Cir. 1986).

Here, the numerosity requirement is easily satisfied for the Damages Class because Delta's military leave data identifies 3,361 pilots who took a total of 119,285 instances of STML from January 1, 2007, to September 30, 2022. (Dr. Cavazos Report, Doc. 118-5 ¶ 16.) With respect to the Injunction Class, Plaintiffs estimate that there are approximately 669 Delta pilots serving in the military based on the number of unique pilot employee IDs present in Delta's military leave data for September 2022, the last month for which data was provided. (Doc. 118-6 ¶ 2.) Delta does not appear to dispute numerosity. Accordingly, the Court finds that the numerosity requirement is met for both proposed classes.

2. Commonality

Rule 23(a)(2)'s commonality requirement imposes on the plaintiff a “relatively light burden,” Vega, 564 F.3d at 1268, to show that “there are questions of law or fact common to the class,” Wal-Mart, 564 U.S. at 349 (quoting Fed. R. Civ. P. 23(a)(2)). “This part of the rule ‘does not require that all the questions of law and fact raised by the dispute be common,’ or that the common questions of law or fact ‘predominate’ over individual issues.” Vega, 564 F.3d at 1268 (internal citation omitted). “[F]or purposes of Rule 23(a)(2) even a single common question will do.” Carriuolo v. Gen. Motors Co., 823 F.3d 977, 984 (11th Cir. 2016) (quoting Wal-Mart, 564 U.S. at 359). But “[t]hat common contention ․ must be of such a nature that it is capable of classwide resolution—which means 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.” Wal-Mart, 564 U.S. at 350. The crux of the commonality requirement “is not the raising of common ‘questions’—even in droves—but rather, the capacity of a class-wide proceeding to generate common answers apt to drive the resolution of the litigation.” Wal-Mart, 564 U.S. at 350 (quoting Nagareda, Class Certification in the Age of Aggregate Proof, 84 N.Y.U. L. Rev. 97, 132 (2009)).

Plaintiffs contend—and Delta does not appear to dispute—that Delta maintained a uniform policy of not providing paid leave to its pilots for STML, while providing compensation for other types of leave such as jury duty, bereavement leave, and sick leave. Because Delta applied a uniform leave policy to all pilots, resolution of their USERRA claims depend on several common issues of law and fact. See Wal-Mart, 564 U.S. at 355 (indicating that “a uniform employment practice” could “provide the commonality needed for a class action”). Specifically, resolution of the putative class members' claims depends on whether Section 4316(b) of USERRA requires Delta to provide paid leave to its pilots who take STML. The first common question raised by that inquiry is whether paid leave is a right or benefit under Section 4316(b) of USERRA. See, e.g., Huntsman v. Sw. Airlines Co., No. 19-CV-00083-PJH, 2021 WL 391300, at *4 (N.D. Cal. Feb. 3, 2021) (finding that “whether paid leave is a ‘right and benefit’ that must be provided equally under USERRA § 4316(b)” is a common question under Rule 23(a)); Clarkson v. Alaska Airlines Inc., No. 2:19-CV-0005-TOR, 2020 WL 4495278, at *4 (E.D. Wash. Aug. 4, 2020) (“Plaintiff identifies a common legal question for the Paid Leave Class as whether paid leave is one of the ‘rights and benefits’ that must be provided equally to employees on military leave under USERRA[.]”). That question was decided in the Court's order on Delta's motion to dismiss, (see Doc. 42 at 7-17), but it still plays a role in the Rule 23 commonality analysis.

More significantly, however, the factfinder will have to determine whether STML is “comparable” to a form of paid leave such as jury duty leave or bereavement leave that Delta provides to pilots. As explained in the Court's discussion of the predominance requirement, see infra at 55-89, the comparability of STML to other forms of leave—based on “(1) the duration of the leave, (2) the purpose of the leave, and (3) the ability of employees to choose when to take the leave”—will be an issue capable of common resolution across all putative class members. Myrick, 69 F.4th at 1319. The comparability inquiry will almost exclusively drive the resolution of Delta's liability to the putative class members' on their USERRA claim. See Huntsman, 2021 WL 391300, at *8 (finding that the comparability inquiry was one of two common questions “likely to drive resolution of the litigation[,]” given the context that “Southwest maintains uniform practices for each work group”); see also Scanlan v. Am. Airlines Grp., Inc., 102 F.4th 164, 170 (3d Cir. 2024) (noting that “[c]omparability is critical to the pilots' USERRA claims” in a class action case brought by pilots). The Court therefore finds that Rule 23(a)'s commonality requirement is satisfied.

3. Typicality

Rule 23(a)(3) requires that “the claims or defenses of the representative parties are typical of the claims or defenses of the class[.]” Fed. R. Civ. P. 23(a)(3). “Although typicality and commonality may be related,” the Eleventh Circuit distinguishes the two concepts “by noting that, ‘[t]raditionally, commonality refers to the group characteristics of the class as a whole, while typicality refers to the individual characteristics of the [class representative] in relation to the class.’ ” Vega, 564 F.3d at 1275 (quoting Piazza, 273 F.3d at 1346).

To establish typicality, plaintiffs must show “a sufficient nexus exists between the legal claims of the named class representatives and those of individual class members to warrant class certification.” Piazza, 273 F.3d at 1346; Ault v. Walt Disney World Co., 692 F.3d 1212, 1216 (11th Cir. 2012). “A sufficient nexus is established if the claims or defenses of the class and the class representative arise from the same event or pattern or practice and are based on the same legal theory.” Kornberg v. Carnival Cruise Lines, Inc., 741 F.2d 1332, 1337 (11th Cir. 1984). Yet, typicality “does not require identical claims or defenses.” Id. “A factual variation will not render a class representative's claim atypical unless the factual position of the representative markedly differs from that of other members of the class.” Id. For example, “a defense which is unique to the class representative and which has the potential to become ‘the focus of the litigation’ may suffice to destroy typicality.” In re Scientific-Atlanta, Inc. Sec. Litig., 571 F. Supp. 2d 1315, 1327 (N.D. Ga. 2007) (citations omitted). Ultimately, “[t]he premise of the typicality requirement is simply stated: as goes the claim of the [class representative], so go the claims of the class.” Hillis v. Equifax Consumer Servs., Inc., 237 F.R.D. 491, 499 (N.D. Ga. 2006) (quoting Sprague v. Gen. Motors Corp., 133 F.3d 388, 399 (6th Cir. 1998)). “Typicality cannot be satisfied when a [class representative] who proved his own claim would not necessarily have proved anybody else's claim.” Id. (cleaned up).

Plaintiffs' sole claim for damages is that Delta violated Section 4316(b)(1) of USERRA by failing to provide—as a matter of policy—paid leave to pilots who took STML. (Doc. 22 ¶¶ 61-69.) Mr. Best is a pilot at Delta and simultaneously serves as a Major in the Michigan Air National Guard. (Doc. 159 at 13-14 (pp. 12-13).) Mr. Best has shown that he and the proposed class members were subject to the same leave policy through Delta's PWAs. (Zawislak Dep., Doc. 123-3 at 22-23 (pp.21-22); Doc. 118-6 ¶ 4.) Moreover, Mr. Best has testified that he did, in fact, take STML on several occasions. (Doc. 159 at 16 (p.15).) On this score, Mr. Best's damages claim appears to be typical to those of the proposed class. See, e.g., Huntsman, 2021 WL 391300, at *8 (“It is evident that the nature of Huntsman's claim is the same as any other class member because each class member has been denied payment for short-term military leave.”). Further, with respect to the Injunctive Class, Mr. Best is currently employed at Delta while also serving the Michigan Air National Guard. (Doc. 159 at 13-14 (pp. 12-13).) Again, there is no dispute that he is subject to Delta's uniform policy of not providing compensation for STML.

Delta, however, argues that it has a unique defense against Mr. Best's USERRA claims, rendering him atypical from other members of the proposed class. (Doc. 170 at 5-9.) Specifically, Delta contends that Mr. Best routinely failed to provide sufficient notice of his military leave in violation of USERRA. (Id. at 7.) In support of this proposition, Delta begins with an implementing regulation of USERRA, 20 C.F.R. § 1002.85, which Delta claims “unambiguously require[s] employees to provide notice of their military duty ‘as far in advance as is reasonable under the circumstances.’ ” (Doc. 170 at 6 (quoting 20 CFR § 1002.85(d)).) Delta also points to its own flight operations manual which instructs that “[w]hen a pilot becomes aware of a military obligation, notification should be given as soon as the requirement for duty is known. This allows Delta to better serve the pilot's needs and optimizes Crew Scheduling/Crew Resources functions while complying with USERRA.” (Doc. 170 at 6 (quoting Nov. 28, 2023, Hr'g, Def. Ex. 1, Delta Flight Operations Manual, § 27.5.3.1.1.).) According to Delta, Mr. Best violated USERRA's notice requirement and Delta's leave policies, because he would only provide notice when he learned that his military duty would conflict with his Delta schedule, rather than when he first became aware of a military service obligation. (Id. at 6-7.)

The Court is unpersuaded by Delta's argument. As an initial matter, the Court does not read USERRA to require employees taking military leave to provide advance notice of a specific length. Indeed, the very regulation that Delta relies upon, states in full:

Although USERRA does not specify how far in advance notice must be given to the employer, an employee should provide notice as far in advance as is reasonable under the circumstances. In regulations promulgated by the Department of Defense under USERRA, 32 CFR 104.6(a)(2)(i)(B), the Defense Department “strongly recommends that advance notice to civilian employers be provided at least 30 days prior to departure for uniformed service when it is feasible to do so.”

20 C.F.R. § 1002.85(d) (emphasis added). In other words, while the implementing regulations indicate that employees must provide some advance notice, and recommend at least 30 days of notice, USERRA does not require notice of a particular length. Delta has failed to produce a single court decision in any jurisdiction that has recognized an employee's failure to provide notice of a sufficient length as a defense to a USERRA claim.⁴ Moreover, Plaintiffs have not brought a breach of contract claim but rather a USERRA claim. Thus, the violation of any internal policy requiring pilots to notify Delta as soon as they learn of a military obligation would not provide a defense under USERRA.

Even if the length of advance notice provided by an employee had some bearing on the strength of a USERRA claim, the Court would still find that Mr. Best's claim was sufficiently typical of those of the proposed class to satisfy Rule 23(a)(3). As the Eleventh Circuit has recognized, typicality “may be satisfied despite substantial factual differences ․ when there is a strong similarity of legal theories.” Loc. 703, I.B. of T. Grocery & Food Emps. Welfare Fund v. Regions Fin. Corp., 762 F.3d 1248, 1259 (11th Cir. 2014) (quoting Williams v. Mohawk Indus., Inc., 568 F.3d 1350, 1357 (11th Cir. 2009)). Different pilots may have provided varying amounts of advance notice for their STML, but the core legal theory underlying Mr. Best's claim and those of the proposed class is the same. That is, Plaintiffs argue that Delta violated USERRA by failing to provide any compensation for STML, while doing so for comparable types of leave.

Further, the Court observes that because of Delta's failure to provide compensation for STML, many pilots at Delta may have been incentivized to wait to inform Delta of military leave only after receiving their flight schedules and learning of potential conflicts. (See Doc. 118-11 ¶ 8(c) (noting that Delta pilots had a financial incentive to schedule STML around their flight schedules because they were not compensated for STML).) If that is so, two consequences follow. First, Mr. Best's practice of waiting to receive his flight schedule before notifying Delta of military leave may, in fact, be typical of many other pilots at Delta who serve in the military. Second, it might also signify that Mr. Best's practice was “reasonable under the circumstances,” in accordance with 20 C.F.R. § 1002.85(d)'s recommendation that “an employee should provide notice as far in advance as is reasonable under the circumstances.”

The Court therefore finds that Mr. Best has satisfied Rule 23(a)'s typicality requirement.

4. Adequacy of Representation

To prove that representation is adequate, Plaintiffs must demonstrate that “the representative parties will fairly and adequately protect the interests of the class.” Fed. R. Civ. P. 23(a)(4). This inquiry “serves to uncover conflicts of interest between named parties and the class they seek to represent.” Amchem Prod., Inc. v. Windsor, 521 U.S. 591, 625 (1997). “This ‘adequacy of representation’ analysis encompasses two separate inquiries: (1) whether any substantial conflicts of interest exist between the representatives and the class; and (2) whether the representatives will adequately prosecute the action.” Valley Drug Co., 350 F.3d at 1189 (cleaned up). “Because all members of the class are bound by the res judicata effect of the judgment, a principal factor in determining the appropriateness of class certification is the forthrightness and vigor with which the representative party can be expected to assert and defend the interests of the members of the class.” London v. Wal-Mart Stores, Inc., 340 F.3d 1246, 1253 (11th Cir. 2003) (quoting Lyons v. Ga.-Pac. Corp. Salaried Emps. Ret. Plan, 221 F.3d 1235, 1253 (11th Cir. 2000)).

The Court is satisfied that Mr. Best's interests are aligned with the proposed class and that he is willing to prosecute this action with forthrightness and vigor. (See Best Decl., Doc. 118-4.) Further, Plaintiffs' counsel are knowledgeable of the relevant law, experienced in complex litigation, and able to devote the time and resources that litigation of this nature requires. (See Anderson Decl., Doc. 118-2; Barton Decl., Doc. 118-3.)

Delta challenges the adequacy of Mr. Best as a class representative, arguing that his “positions generate conflicts among the class because his practice of not providing Delta with ‘pre-posting’ notice artificially narrows the scope of the putative class.” (Doc. 170 at 8.) In particular, Delta notes that Mr. Best only seeks to represent a class of pilots who took STML “as reflected by Delta's military leave data.” (Id.) This “narrowed” class definition, Delta claims, “does not adequately protect the interests of pilots who took STML but which does not appear in ․ Delta's data because, for example, they sufficiently notified Delta in advance of the bid period to avoid a conflict in scheduling and their schedules were built around military duty.” (Id.)

For a conflict to exist, the interests of the putative class members must be opposed, such that “some party members claim to have been harmed by the same conduct that benefitted other members of the class.” In re Blue Cross Blue Shield Antitrust Litig. MDL 2406, 85 F.4th 1070, 1091 (11th Cir. 2023) (quoting Valley Drug, 350 F.3d at 1189). There is no such conflict present here because Mr. Best does not have interests or objectives that are opposed to those of putative class members. Mr. Best and the putative class members are alleged to have been harmed in the same manner—that is, by Delta's failure to provide compensation for STML pursuant to its uniform leave policy.

That Plaintiffs have elected to define the class here more narrowly than they could have—namely, by using Delta's leave data to identify class members and instances of STML—does not generate conflicts between Mr. Best and other proposed members of the class. First, Plaintiffs' proposed class definition may exclude certain Delta pilots from the class altogether because they simply do not fall within the proposed class definition. Any such Delta pilot would not qualify as a class member and their interests are thus irrelevant to the adequacy analysis. See G.H. v. Tamayo, 339 F.R.D. 584, 590 (N.D. Fla. 2021) (“[T]he possibility that class members' interests may conflict with the interests of individuals who are not class members is not part of the adequacy analysis.”). Of course, because these pilots would be excluded from the class altogether, they would not be bound by any judgment and would be free to bring their own individual claims. See Original Brooklyn Water Bagel Co. v. Bersin Bagel Grp., LLC, 817 F.3d 719, 727 (11th Cir. 2016) (“[T]he settlements of class action suits bind only the members of that class—they do not affect individuals who opt out of or were otherwise not included in the class.”).

Second, there may be Delta pilots who fall within the class definition because they took some STML reflected by Delta's leave data, but the class definition may exclude some of their STML that is not documented by the leave data. (Doc. 170 at 7-8.) As Plaintiffs argue, however, it is appropriate for them to elect to pursue a narrower set of claims in a class action suit based on administrative feasibility or other strategic considerations. See, e.g., Todd v. Tempur-Sealy Int'l, Inc., No. 13-CV-04984-JST, 2016 WL 5746364, at *5 (N.D. Cal. Sept. 30, 2016) (“A strategic decision to pursue those claims a plaintiff believes to be most viable does not render her inadequate as a class representative.”); Manno v. Healthcare Revenue Recovery Grp., LLC, 289 F.R.D. 674, 696 (S.D. Fla. 2013) (“As numerous courts have held, the named plaintiff is not rendered inadequate just because he elects to pursue some remedies, but not others, so long as putative class members are given adequate notice and opportunity to opt out.”); see also Stanich v. Travelers Indem. Co., 259 F.R.D. 294, 307 (N.D. Ohio 2009) (“[C]ourts have found that concerns related to the potential res judicata effect of abandoning a certain claim in favor of another claim do not necessarily create a conflict between the class representative and the absent class members.”). Here, it would be more difficult for a putative class member to prove a USERRA violation, if any, where the STML was not recorded in Delta's leave data. Among other problems, a pilot whose STML was not recorded by Delta's leave data may have difficulty showing that Delta had any notice of the STML. (Doc. 171 at 12-13.) And it would almost certainly be less administratively feasible to resolve USERRA claims based on unrecorded STML on a classwide basis. Against that background, Plaintiffs' strategic choice to pursue only USERRA claims based on STML documented by Delta's leave data is reasonable.

Furthermore, Delta does not explain how Plaintiffs' choice to focus exclusively on recorded STML claims causes some class members “to have been harmed by the same conduct that benefitted other members of the class[,]” such that there is a “fundamental conflict” between Mr. Best and other class members. In re Blue Cross, 85 F.4th at 1091 (quoting Valley Drug, 350 F.3d at 1189). There is no such conflict present here because both Mr. Best and the putative class members are alleged to have been harmed by the same conduct in the same manner—that is, by Delta's failure to provide compensation for STML pursuant to its uniform leave policy. Mr. Best does not have interests or objectives that are opposed to those of putative class members. Instead, he has made a reasonable strategic choice to pursue the classwide resolution of only those instances of STML that are recorded in Delta's leave data.

As such, Mr. Best's decision to narrow the class definition to pilots who took STML as recorded by Delta's leave data does not render Mr. Best inadequate as a class representative. In addition, to the extent that Delta challenges Mr. Best's adequacy on the ground that he failed to provide sufficient notice of his STML to Delta, (Doc. 170 at 6-9), the Court finds that argument unpersuasive for the reasons discussed in the preceding section discussing typicality. See supra at 32-36.

C. Rule 23(b)(2) Injunctive Class

Having found that Mr. Best has met the requirements of Rule 23(a), the Court turns to Rule 23(b). Plaintiffs move for certification of both a Rule 23(b)(2) class for injunctive relief and a Rule 23(b)(3) class for monetary damages. (Doc. 183 ¶¶ 1-2.) The Court first addresses certification of the proposed Injunction Class pursuant to Rule 23(b)(2) and then considers certification of the Damages Class under Rule 23(b)(3).

A Rule 23(b)(2) class seeking injunctive relief may be certified if “the party opposing the class has acted or refused to act on grounds that apply generally to the class, so that final injunctive relief or corresponding declaratory relief is appropriate respecting the class as a whole[.]” Fed. R. Civ. P. 23(b)(2). The “key” to demonstrating a Rule 23(b)(2) class is “the indivisible nature of the injunctive or declaratory remedy warranted—the notion that the conduct is such that it can be enjoined or declared unlawful only as to all of the class members or as to none of them.” Wal-Mart, 564 U.S. at 360 (quoting Nagareda, Class Certification in the Age of Aggregate Proof, 84 N.Y.U. L. Rev. 97, 132 (2009)). That is, “Rule 23(b)(2) applies only when a single injunction or declaratory judgment would provide relief to each member of the class. It does not authorize class certification when each individual class member would be entitled to a different injunction or declaratory judgment against the defendant.” Id.

Plaintiffs seek to certify an Injunction Class of “all pilots currently employed by Delta who are also currently members of the uniformed services ․” (Doc. 183 ¶ 1.) With respect to the Injunction Class, Plaintiffs request that the Court “[d]eclare that Delta's policy or practice by which Delta fails to pay Plaintiff Best and the Injunction Class when they take short-term military leave, while paying employees who took other, comparable forms of non-military leave, violates the rights of Plaintiff and the Injunction Class under USERRA § 4316[.]” (Doc. 182 at 25.) Plaintiffs further request that the Court

[o]rder Delta to pay pilots who take short-term military leave on the same basis as pilots who take leave for the forms of comparable short-term, non military leave found to be comparable consistent with the Court's declaration and to the extent there is a difference among the comparable non-military leaves, pay based on the one that would be most favorable to servicemembers[.]

(Id.)

Plaintiffs argue that certification of the proposed Injunction Class is appropriate because Delta subjects all pilots who are members of the military to a uniform leave policy that denies any compensation for STML. (Doc. 183-1 at 25-26.) If the Court were to find that Delta's leave policy violates USERRA, Plaintiffs contend, the Injunction Class would be entitled to a declaration of the same and a corresponding injunction requiring Delta to provide compensation for STML on the same basis as it provides compensation for comparable short-term, non-military leave. (Id. at 26-27.)

Delta opposes certification of the proposed Injunction Class on two grounds. First, Delta asserts that Plaintiffs' request for injunctive relief is predominantly about monetary damages and would “requir[e] Delta to pay an indeterminate amount of damages to class members going forward[,]” which, per Delta, violates the U.S. Supreme Court's holding in Wal-Mart v. Dukes, 564 U.S. 338 (2011). (Doc. 185 at 10-13.) Second, Delta argues that Plaintiffs' requested injunction is an improper “follow-the-law” injunction that is insufficiently specific to “show[ ] that a single injunction would benefit the whole class.” (Id. at 13-14.) The Court addresses Delta's arguments out of order.

1. Whether Plaintiffs' Requested Injunction Would Provide Relief to Each Class Member

Contrary to Delta's contention,⁵ the Court finds that Plaintiffs have sufficiently demonstrated that their requested injunction would “provide relief to each member of the class.” Wal-Mart, 564 U.S. at 360. The proposed Injunction Class is made up of Delta pilots who are concurrently members of the military and therefore have obligations tied to their military service. If Plaintiffs were to prevail on the merits of their USERRA claim, the proposed injunction would make each member of the Injunction Class eligible for compensation for STML “on the same basis as pilots who take leave for the forms of comparable short-term, non[-]military leave found to be comparable” by the factfinder at trial. (Doc. 182 at 25.) For example, if the factfinder finds that a period of STML was comparable to an equivalent period of jury duty leave under USERRA, the proposed injunction would require Delta to provide every Injunction Class member taking STML with the corresponding compensation available to pilots taking jury duty leave under the subject PWA.

Moreover, complying with the injunction would not require an individualized inquiry into each class member's claims or situation; instead, Delta would have to change its uniform leave policy as it applies to every class member to comply with the Court's determination of what USERRA requires. See Melanie K. v. Horton, No. 1:14-CV-710-WSD, 2015 WL 1308368, at *5 (N.D. Ga. Mar. 23, 2015) (“These policies and practices are equally applicable to each class member, and injunctive or declaratory relief addressing the policy with respect to the class as a whole is appropriate.”); Hernandez v. Medows, 209 F.R.D. 665, 673 (S.D. Fla. 2002) (“This purported policy is equally applicable to each class member of the proposed class. Injunctive or declaratory relief settling the legality of the policy with respect to the class as a whole is appropriate.”). Put simply, the requested “injunction would benefit the class as a whole, in the same way.” Doe v. MG Freesites, LTD, 707 F. Supp. 3d 1157, 1186 (N.D. Ala. 2023).

Further, the Court is unpersuaded by Delta's arguments that the proposed injunction is an improper “follow-the-law” injunction or lacks sufficient detail to enable the Court to assess whether it would provide relief to each class member. (See Doc. 185 at 13-14.) Plaintiffs seek an injunction requiring Delta to provide compensation for STML on the same basis as comparable forms of short-term, non-military leave. Plaintiffs have alleged that Delta provides compensation for jury duty leave, sick leave, and bereavement leave and that these forms of leave are comparable to STML under USERRA. (Doc. 182 ¶¶ 54-69.) At this preliminary class-certification stage, of course, it remains to be decided which, if any, of these forms of non-military leave are comparable to STML. However, if Plaintiffs prevail on the merits of their USERRA claim, the Court could properly order Delta to comply with USERRA by providing compensation for STML at the same rate as it does for comparable non-military forms of leave (as determined by the factfinder). Such an injunction would not merely direct Delta to “follow the law.” Instead, the injunction would identify specific comparable forms of leave and instruct Delta to provide compensation for STML on the same basis as those forms of leave. Delta does not explain how Plaintiffs could ask for more specific injunctive relief as to their USERRA claim at this stage of the case, where the factfinder has not yet determined what forms of leave are comparable to STML.⁶

In short, the Court finds that Plaintiffs have carried their burden to show that their requested injunction would provide relief to each putative class member on a classwide basis.

2. Whether Plaintiffs' Requested Injunction Improperly Seeks Monetary Damages

Delta asserts that certifying Plaintiffs' proposed Injunction Class under Rule 23(b)(2) would violate the Supreme Court's decision in Wal-Mart v. Dukes, 564 U.S. 338 (2011), because Plaintiffs' request for injunctive relief is “predominantly about monetary damages” and would “requir[e] Delta to pay an indeterminate amount of damages to class members going forward.” (Doc. 185 at 10-13.) Plaintiffs contend that their request for injunctive relief “seeks to prevent a future injury, rather than remedy a past harm,” and therefore does not seek monetary damages at all. (Doc. 190 at 18-21 (emphasis removed).) Plaintiffs also highlight that several courts in the Eleventh Circuit have approved hybrid certification both before and following the Supreme Court's decision in Wal-Mart. (Id. at 21-22.)

The Court finds that Plaintiffs have the better argument. As a threshold matter, the Eleventh Circuit has explained that “hybrid class actions” with separate (b)(2) injunctive and (b)(3) damages classes may be maintained in appropriate cases:

It is possible to create hybrids in given cases. Since in theory there should be no hard requirement that (b)(2) be mutually exclusive, and since subpart (c)(4)(A) allows an action to be maintained “with respect to particular issues,” the fact that damages are sought as well as an injunction or declaratory relief should not be fatal to a request for a (b)(2) suit, as long as the resulting hybrid case can be fairly and effectively managed.

Williams, 568 F.3d at 1360 (quoting Holmes v. Cont'l Can Co., 706 F.2d 1144, 1158 n. 10 (11th Cir. 1983)); see also Ebert v. Gen. Mills, Inc., 823 F.3d 472, 480 (8th Cir. 2016) (“The use of ․ hybrid certification, insulating the (b)(2) class from the money-damage portion of the case, is an available approach that is gaining ground in class action suits.”); Jancik v. WebMD LLC, No. 1:22-CV-644-TWT, 2025 WL 560705, at *9 (N.D. Ga. Feb. 20, 2025) (“Although the Court has held that certification under Rule 23(b)(3) is appropriate, Jancik may seek certification under Rule [23(b)(2)] as well.”); Doe, 707 F. Supp. 3d at 1187 (“Defendants argue that Plaintiff cannot be granted certification of a Rule 23(b)(2) class because she is also seeking damages separately in a Rule 23(b)(3) class. To the contrary, it is permissible to certify both types of classes.”); see also In re Blue Cross, 85 F.4th at 1084, 1091-92, 1102 (approving a class action settlement agreement with both a Rule 23(b)(2) injunctive relief class and a Rule 23(b)(3) damages class that was represented by the same named plaintiffs and counsel); 2 Newberg and Rubenstein on Class Actions § 4:38 (6th ed.) (“Certification of a hybrid action is often thought to be the best of both worlds, achieving the judicial economies associated with group litigation while also respecting the due process rights of individuals with monetary claims should the defendant be found liable during the first phase of the trials.”). To the extent that Delta contends that Plaintiffs may not simultaneously seek certification of a Rule 23(b)(2) injunctive relief class and a Rule 23(b)(3) damages class, the Court disagrees.

Further, the Court disagrees with Delta's arguments concerning Wal-Mart's application to this case. Wal-Mart held that a group of female employees alleging sex discrimination in the workplace could not proceed as a Rule 23 class because their treatment was the result of “literally millions of [subjective] employment decisions” made by a decentralized network of managers rather than a general, company-wide policy of discrimination. Wal-Mart, 564 U.S. at 352. Because there was no “glue holding the alleged reasons for all those [employment] decisions together,” Wal-Mart concluded that the plaintiffs could not meet Rule 23(a)'s commonality requirement. Id.

Relevant here, the Wal-Mart Court also determined that the plaintiffs' “claims for backpay were improperly certified under [Rule] 23(b)(2).” Id. at 360 (emphasis added). The Court noted that it had previously “expressed serious doubt about whether claims for monetary relief may be certified under” Rule 23(b)(2) “at least where ․ the monetary relief is not incidental to the injunctive or declaratory relief.” Id. Although it did not conclusively resolve whether a class seeking monetary damages could ever be certified pursuant to Rule 23(b)(2), the Wal-Mart Court concluded that claims for individualized monetary damages, like backpay, are inconsistent with Rule 23(b)(2)'s requirement that plaintiffs seek “a single injunction or declaratory judgment [that] provide[s] relief to each member of the class.” Id.

This case, however, is unlike Wal-Mart because the proposed Rule 23(b)(2) class does not seek backpay, or any form of retrospective monetary damages. Rather, Plaintiffs seek an injunction that would change Delta's uniform leave policy to comply with USERRA. That the injunction might require Delta to provide compensation for STML in the future does not take the class outside of Rule 23(b)(2)'s ambit, because the requested injunction would change Delta's uniform leave policy in a manner that provides relief to each member of the class.

In fact, even if Plaintiffs were to prevail in obtaining injunctive relief, the requested injunction may not require the future payment of compensation to members of the Injunction Class, and it certainly would not require payment of any damages based on past conduct that occurred before the injunction was issued. Consider again the hypothetical where a jury finds that a period of STML is comparable to an equivalent period of jury service leave but no other form of leave. In that case, Delta could choose between two paths to comply with any ensuing injunction. Delta could provide members of the Injunction Class who take STML with the same compensation afforded pilots who take equivalent periods of jury leave. Alternatively, Delta could cease providing any compensation to pilots who take leave for jury service (so long as that is not otherwise prohibited by law). See White v. United Airlines, Inc., 987 F.3d 616, 624 (7th Cir. 2021) (“USERRA mandates only equality of treatment; it does not specify how generous or how parsimonious an employer's paid leave policies must be.”). If Delta selected the second option, it would not be required to provide any monetary benefits to pilots taking STML moving forward. Moreover, the injunction would not require Delta to provide damages for any previous instances of STML taken by class members that were uncompensated. If it were to require compensation at all, it would only do so moving forward from the date the injunction was issued. Indeed, Plaintiffs seek to recover individual damages for past uncompensated STML through a separate Rule 23(b)(3) class, not the Rule 23(b)(2) class. See MG Freesites, 707 F. Supp. 3d at 1187 (explaining that certification of separate (b)(2) and (b)(3) classes “insulates the (b)(2) class piece from the money damage portion of the case, hence complying with Wal-Mart's admonition against adjudicating individual damage claims in a (b)(2) class action” (quoting 2 Newberg and Rubenstein on Class Actions § 4:38)); see also 7 Newberg and Rubenstein on Class Actions § 23:30 (6th ed.) (Wal-Mart “held that a court cannot certify a (b)(2) class to adjudicate individualized monetary damages, but it did not foreclose certification of a (b)(2) class for injunctive purposes simply because money damages—that may be adjudicated in a (b)(3) format (or individually)—are also being sought in the case.”).

The Eleventh Circuit's decision in AA Suncoast Chiropractic Clinic, P.A. v. Progressive Am. Ins. Co., 938 F.3d 1170 (11th Cir. 2019) (“Suncoast”) illustrates the distinction between backward-looking monetary damages and prospective injunctive relief in the Rule 23(b)(2) context. In Suncoast, a group of healthcare providers brought a putative class action against an insurance company, Progressive, alleging that it employed a claims-handling process that violated Florida law. Suncoast, 938 F.3d at 1172. The district court declined to certify a (b)(3) damages class, but certified a (b)(2) injunction class of providers who had purportedly been harmed by the claims-handling process. Id. at 1173. The Eleventh Circuit found that certification of the (b)(2) class was improper because the providers' requested injunction was “not designed to address the treatment of future claims; it would instead, ․ restore claimants to the claims-handling process free of the improper cap on PIP benefits imposed by Progressive.” Id. at 1175. In reaching that conclusion, the Eleventh Circuit expressly distinguished between relief seeking to remedy “past” or “retrospective” harms, which cannot be sought through a Rule 23(b)(2) class, and “an injunction ․ geared toward preventing future harm[,]” which Rule 23(b)(2) permits. Id. at 1175-76.

The problem with the requested injunction in Suncoast was that it attempted to redress the past denial of insurance coverage by “mandat[ing] that Progressive reprocess all [past] claims” that were allegedly affected by the unlawful claims-handling policy. Id. That, the Eleventh Circuit said, was not a request to prevent future injury, but rather an improper attempt to obtain retrospective relief while sidestepping the requirements of Rule 23(b)(3). Id. at 1178-79 (“In short, the plaintiffs are not really interested in final prospective equitable relief at all; they are singularly focused on recovering a retrospective damages remedy, and Rule 23(b)(3), not (b)(2), governs certification of a damages class.” (quotation omitted)). However, the Eleventh Circuit recognized that “in the real world, there are likely people with a future interest in having Progressive manage its [claims] in a particular way” that could seek injunctive relief and certification under Rule 23(b)(2). Id. at 1175.

By contrast to the healthcare providers in Suncoast, Plaintiffs' (b)(2) class seeks prospective equitable relief, not retrospective money damages. Plaintiffs request an injunction that would restrain only future violations of USERRA by compelling Delta to change its uniform leave policy—a policy that, as previously explained, applies to every member of the proposed Injunction Class. Plaintiffs also do not pursue their claims for compensation of past STML through their proposed injunction. The injunctive relief requested here would only address class members' ability to access benefits when they take STML in the future. Plaintiffs seek certification of a separate 23(b)(3) class that would recover potential damages, if any, for past STML that Delta failed to compensate.

In sum, Plaintiffs' proposed Injunction Class does not seek retrospective damages, but rather a single injunction that would alter Delta's uniform leave policy to prevent future injury and provide relief to each member of the class. Because that is precisely what Rule 23(b)(2) permits, the Court will certify Plaintiffs' Injunction Class.

D. Rule 23(b)(3) Damages Class

Plaintiffs seek to certify the following Damages Class under Rule 23(b)(3):

All current and former pilots who work or worked for Delta Air Lines, Inc. (“Delta”) in the United States or its territories or possessions who took short-term military leave (i.e., a military leave of 30 consecutive days or fewer) from their employment from Delta as a pilot from April 30, 2007 to the date of judgment in this action, as reflected by Delta's military leave data[.]

(Doc. 183 ¶ 1.) For certification of a damages class under Rule 23(b)(3), Plaintiffs must demonstrate “(1) that common questions of law or fact predominate over questions affecting only individual class members (“predominance”); and (2) that a class action is superior to other available methods for adjudicating the controversy (“superiority”).” Vega, 564 F.3d at 1265. The Court begins with the predominance inquiry and then turns to the issue of superiority.

1. Predominance

With respect to the predominance requirement, certification under Rule 23(b)(3) requires a district court to find that “questions of law or fact common to class members predominate over any questions affecting only individual members[.]” Fed. R. Civ. P. 23(b)(3). The “predominance inquiry tests whether proposed classes are sufficiently cohesive to warrant adjudication by representation.” Tyson Foods, Inc. v. Bouaphakeo, 577 U.S. 442, 453 (2016) (quoting Amchem Products, Inc. v. Windsor, 521 U.S. 591, 623 (1997)). To determine whether common issues predominate, a district court “must first identify the parties' claims and defenses and their elements” and “then classify these issues as common questions or individual questions by predicting how the parties will prove them at trial.” Brown, 817 F.3d at 1234. Moreover, a district court must decide all questions of fact and law that bear on the predominance inquiry, even if those questions also overlap with the merits. Id. at 1237.

“An individual question is one where members of a proposed class will need to present evidence that varies from member to member, while a common question is one where the same evidence will suffice for each member to make a prima facie showing or the issue is susceptible to generalized, class-wide proof.” Tyson Foods, 577 U.S. at 453 (cleaned up). When “one or more of the central issues in the action are common to the class and can be said to predominate, the action may be considered proper under Rule 23(b)(3) even though other important matters will have to be tried separately, such as damages or some affirmative defenses peculiar to some individual class members.” Id. (cleaned up)

The predominance inquiry calls for a “qualitative assessment” of the “relative importance” of common versus individual questions. Brown, 817 F.3d at 1235 (quoting Butler v. Sears, Roebuck & Co., 727 F.3d 796, 801 (7th Cir. 2013)); see also Tyson Foods, 577 U.S. at 453 (noting that the predominance requirement asks “courts to give careful scrutiny to the relation between common and individual questions in a case”). As a “rule of thumb[,]” if common issues predominate over individualized questions, “the addition or subtraction of any of the plaintiffs to or from the class [should not] have a substantial effect on the substance or quantity of evidence offered.” Brown, 817 F.3d at 1235 (quoting Vega, 564 F.3d at 1270). Put another way, if “the addition of more plaintiffs leaves the quantum of evidence introduced by the plaintiffs as a whole relatively undisturbed, then common issues are likely to predominate.” Id. (quoting Vega, 564 F.3d at 1270). “District courts should assess predominance with its overarching purpose in mind—namely, ensuring that ‘a class action would achieve economies of time, effort, and expense, and promote ․ uniformity of decision as to persons similarly situated, without sacrificing procedural fairness or bringing about other undesirable results.’ ” Id. (quoting Amchem Prods., 521 U.S. at 615).

a. USERRA's Comparability Inquiry

Plaintiffs bring a single claim alleging that Delta violated USERRA by uniformly failing to provide compensation to pilots who take short-term military leave on the same basis as pilots who take comparable forms of non-military leave. Under USERRA, a person who takes military leave from their employment is “entitled to such other rights and benefits not determined by seniority as are generally provided by the employer of the person to employees having similar seniority, status, and pay who are on furlough or leave of absence ․” 38 U.S.C. § 4316(b). The Eleventh Circuit has explained that this provision “set[s] forth a two-step process for determining which benefits are available to employees on military leave.” Myrick, 69 F.4th at 1314. “First, the military employee must identify a group of non-military employees ‘having similar seniority, status, and pay who are on ․ leave of absence[.]’ ” Id. (quoting 38 U.S.C. § 4316(b)(1)(B)). “Second, the military employee must prove that those employees took a form of non-military leave that is comparable to military leave.” Id. at 315 (citing 20 C.F.R. § 1002.150(b)). A military employee who satisfies both requirements “is entitled to the same benefits that the similarly situated employees received while on the comparable form of leave.” Id. In accordance with implementing regulations issued by the DOL, the Eleventh Circuit has held that “[t]hree factors inform [the] comparison between leaves: (1) the duration of the leave, (2) the purpose of the leave, and (3) the ability of employees to choose when to take the leave.” Myrick, 69 F.4th at 1319.

The Court must now “classify” each element of the USERRA leave analysis as a “common” or “individual” question “by predicting how the parties will prove them at trial.” Brown, 817 F.3d at 1234. The first step of the USERRA leave analysis—as set forth in Myrick—is a common question capable of determination through classwide proof. Plaintiffs have already identified several different forms of non-military paid leave offered to Delta pilots through the terms of the current (and past) PWAs that govern (or governed) the employment of Delta pilots. The PWAs apply classwide and their terms do not vary from pilot to pilot. Thus, Plaintiffs can use common proof to “identify a group of non-military employees ‘having similar seniority, status, and pay who are on ․ leave of absence[.]’ ” Id. (quoting 38 U.S.C. § 4316(b)(1)(B)).

Duration. Turning to the second step of the USERRA leave analysis, the Court begins with the first factor regarding the comparability of leave—the duration of the leave. “In order to determine whether any two types of leave are comparable, the duration of the leave may be the most significant factor to compare.” 20 C.F.R. § 1002.150.

The duration element of comparability strikes the Court as straightforward. In the Court's view, with respect to duration, a 5-day period of STML is comparable to a 5-day period of bereavement leave. Likewise, in terms of duration, a 14-day period of jury duty leave is comparable to a 14-day period STML. See Myrick, 69 F.4th at 1320 (finding that the defendant city “was inclined to provide paid administrative leave for up to around sixteen months[,]” which was comparable in duration to the plaintiffs' “longest instances of military leave” of a similar length); 20 C.F.R. § 1002.150 (“For instance, a two-day funeral leave will not be ‘comparable’ to an extended leave for service in the uniformed service.”). Delta has produced data reflecting the dates of STML and other forms of leave, including non-military leave. (Doc. 118-5 ¶ 9; Doc. 118-11 ¶ 10a.) Because the factfinder can use Delta's leave data to identify whether the duration of STML matches the duration of other non-military forms of leave, the Court believes that the duration aspect of comparability can be determined using classwide proof.

Delta, however, contends that Plaintiffs cannot use classwide proof to compare the duration of STML to other forms of leave because Delta's leave data does not reliably reflect the true duration of each STML period. (Doc. 170 at 12-16.) For example, Delta highlights that its leave data reflects that, in April and May 2017, Mr. Best took four STMLs of two, seven, fourteen, and one day(s), respectively. (Id. at 12.) By contrast, Mr. Best's military records show that these STMLs are part of one continuous period of military service of 39 days. (Id.) Plaintiffs do not appear to dispute that, according to his military records, Mr. Best's military leave in April and May 2017 lasted 39 days, which conflicts with Delta's leave data. (See Doc. 171 at 17 (“Delta's data reflects multiple short-term periods of leave by Best in Spring of 2017, but his PCARS reflects one 39 day, long-term period of leave․ USERRA requires the entire period of service be treated as a leave.”).)

Despite this conflict between Mr. Best's military records and Delta's self-reported leave data, the Court is unpersuaded by Delta's argument that potential reliability issues with its leave data precludes comparison of the duration of STML with the duration of other forms of leave. First, even if there are some flaws in Delta's leave data, a jury could still rely on it to conclude that the duration of STML is comparable to the duration of other forms of leave. That is, the reliability problems go to the weight of the data, not whether it can serve as common proof. Further, Plaintiffs have submitted evidence suggesting that the PCARS data will provide a meaningful cross-check on Delta's leave data. (See Third Downes Decl., Doc. 171-1 ¶ 2 (“Cross referencing Best's short-term military leave entries in Delta's data and Best's PCARS, I identified five days of short-term military leave that were not reflected by a corresponding service record in his PCARS, representing approximately 3.7% of Best's short-term military leave days as reflected in Delta's data.”).)

Second, even if a proportion of the datapoints of STML in the aggregate dataset are not truly instances of STML (because they were part of longer, continuous periods of military leave), a factfinder could still find, on the basis of classwide evidence, that STML of a certain length is comparable in duration to other non-military forms of leave. A concrete example helps illustrate this point. Consider, hypothetically, that Delta's leave data shows that the putative class members took a total of 1,000 5-day periods of STML for which they did not receive any compensation. Consider further that Delta's leave data shows that its pilots took a total of 250 5-day periods of jury leave, which were compensated. Surely, a factfinder could (and probably would) conclude that the 5-day periods of STMLs were comparable to the 5-day periods of jury leave in terms of duration. Indeed, both periods are exactly the same length. Now assume that half of the 5-day STML datapoints are inaccurate because they were part of longer periods of military leave, so there are actually only 500 accurate 5-day periods of uncompensated STML in the dataset. Would such a finding preclude the factfinder from reaching the same conclusion regarding comparability of duration? The Court thinks not. The data would still show that Delta routinely provides compensation for 5-day periods of jury duty leave, but not for 5-day periods of STML. On that basis, a jury could find that a 5-day period of STML is comparable in duration to a 5-day period of non-military leave under USERRA. The mere fact there are fewer instances of 5-day STMLs in the dataset does not render a 5-day STML any less comparable, in terms of duration, to a 5-day jury duty leave.

The Eleventh Circuit's decision in Myrick is instructive. In Myrick,

[t]he district court analyzed [the duration] factor by grouping military leave and paid administrative leave into two categories based on length—short-term leave and long-term leave. In the short-term-leave category, the court placed military leave for training and administrative leave for brief events like jury duty. It concluded that short-term military leave lasted three times longer (thirty-seven days) than short-term administrative leave (thirteen days). In the long-term-leave category, the district court placed military leave for deployment and investigative administrative leave. It found that both lasted, on average, sixteen months.

The district court ultimately concluded that military leave was not comparable in duration to paid administrative leave because the three-to-one difference between short-term military leave and short-term administrative leave was significant. In reaching this decision, the district court disregarded the similarity between military leave for deployment and investigative administrative leave, labeling the long-term leaves “outliers.”

Myrick, 69 F.4th at 1319-20. The Eleventh Circuit, however, “s[aw] things differently.” Id. at 1320. Rather than viewing the longest instances of administrative leave as “outliers[,]” the Myrick Court determined that “they set the upper strata of paid administrative leave that Hoover was willing to provide its employees.” Id. That is, the longest instances of administrative leave “demonstrate[d] that Hoover was inclined to provide paid administrative leave for up to around sixteen months—the same average length as the longest instances of military leave.” Id. As such, the Eleventh Circuit determined that “[h]ad the Officers been placed on paid administrative leave instead of military leave, they would have received holiday pay and accrued benefits for each period of service, including those shorter than sixteen months.” Id. (emphasis added). Thus, the two forms of leave—military leave and administrative leave—were comparable in duration. Id.

The same sort of duration analysis is possible on a classwide basis here. For instance, if Delta's leave data shows that Delta regularly provides compensation for jury duty leave up to 15 days or less, a jury could find that this was the “upper strata” of paid jury duty leave that Delta was willing to pay. The jury could therefore also find that STMLs of 15 days or less are comparable in duration to jury duty leave.⁷ This finding would apply classwide based on proof—namely, Delta's leave policies and leave data—that is common to the class. Moreover, the fact that Delta's leave data may include some instances of STML that were longer than they appear in the data would not undermine the factfinder's comparability analysis.

To be sure, Defendant's argument might have consequences for the individual calculation of damages for each putative class member at the claims administration stage, if any, of the case. Should Plaintiffs prevail on their USERRA claim, the Court will have to ensure when calculating damages that a period of STML that was recorded as a certain number of days in Delta's leave data, but was in reality longer (e.g., as shown through a cross-check of PCARS data), is properly recalculated. But that issue does not undermine the notion that the comparability inquiry as to the duration of different forms of leave can be conducted on a classwide basis. Accordingly, the Court finds that comparability with respect to duration presents a common question capable of classwide determination.

Purpose. Plaintiff contends that the purpose of different forms of leave will be established using common proof, for example, through the testimony of Delta employees and experts. (Doc. 183-1 at 17-19, 29.) Delta does not appear to dispute that comparability with respect to the purpose of leave is capable of determination using classwide proof. Indeed, Myrick found that military leave and the paid administrative leave at issue there “serve[d] similar ends” by enabling the defendant city “to comply with the law” and “to shield employees from unnecessary hardship.” Myrick, 69 F.4th at 1319. More recently, the Eleventh Circuit determined that Delta's provision of “known leaves of absence” was not comparable to long-term military leave with respect to purpose:

[A]s to purpose, Delta offers senior pilots a known leave of absence when the company is overstaffed and needs to furlough employees. Pilots are considered active while on a known leave of absence—and can be recalled—but do not fly any flights. Long-term military leave, on the other hand, allows a pilot to perform military service and Delta cannot call the pilot back to work.

Sorenson v. Delta Air Lines, Inc., 174 F.4th 54, 63 (11th Cir. 2026). The Third Circuit has determined, in another USERRA case brought by airline pilots, that “[a] reasonable jury reviewing the evidence in this case could conclude that jury-duty leave and military leave have a common purpose: civic duty.” Scanlan, 102 F.4th at 171 (“Indeed, the parties agree that those two types of leave both entail performing a public service.”); see also Clarkson, 59 F.4th at 437 (concluding that a jury could find “that the primary purpose of military leave is to perform a civic duty and public service”).

These cases assess the purpose of different forms of leave at a level of generality or abstraction that does not require individualized proof that varies from pilot to pilot. That approach is also appropriate here. The Court finds that Plaintiff can use classwide proof to establish that STML and other forms of leave serve similar purposes.

Control. The parties contest whether the record evidence demonstrates that military pilots have control over when to take military leave and if that determination is susceptible to common proof. Plaintiffs argue that Mr. Best's testimony is sufficient to permit a “jury [to] conclude that—as with pilots who take jury-duty or bereavement leave—most pilots who take military leave lack the ability to control their schedules.” (Doc. 190 at 14 (quoting Scanlan, 102 F.4th at 171).) Mr. Best testified that he was required to participate in a bidding system for scheduling his military duties where he would have to submit his availability to his unit. (Transcript of Jan. 17, 2024, Hearing, Doc. 159 at 48-49 (pp. 47-48); Best Dep., Doc. 170-2 at 78-80 (pp.77-79).) He indicated he could “[n]ot really” swap flights with other pilots in the Air National Guard and that he had no flexibility in determining when “drill weekends” were set. (Doc. 159 at 49-50 (pp. 48-49).) However, Mr. Best also testified that after flights were scheduled, he was able to notify his unit that he was no longer available for a flight to which he had been assigned. (Doc. 170-2 at 82-83 (pp.81-82).)

Delta contends that “[a] reservist's control over when to take leave and the frequency and duration of such leave are individual questions ‘not susceptible to a common answer for all pilots.’ ” (Doc. 185 at 6.) Delta points to Plaintiff Reep's testimony that he had flexibility in scheduling some of his military duties. (See Doc. 123 at 37 n.15; Reep Dep., Doc. 123-2 at 91-110 (pp. 90-109).) For instance, Plaintiff Reep testified that he was able to submit preferences for certain flights and could exert additional influence on military flight schedules when he was a flight commander. (Doc. 123-2 at 91-100 (pp. 90-99).) He also explained that he could sometimes swap flights with other pilots in his unit and would regularly volunteer for certain assignments. (Id. at 105-06, 108-10 (pp. 104-05, 107-09).)

The Court finds, based on the record before it, that the scheduling of military duties for pilots may vary in nuanced ways from pilot to pilot and assignment to assignment. For example, Plaintiff Best testified that he could not regularly swap flights with other pilots in his military unit, while Plaintiff Reep testified that he could. Further, Plaintiff Reep indicated that as a flight commander, he could influence scheduling in ways that pilots in other positions could not. Plaintiff Best noted that certain assignments like drill weekends had no flexibility, but he could change his availability with respect to other assignments. Perhaps unsurprisingly, given the nature of military operations in general and flying in particular, pilot scheduling in the military appears to be a complicated endeavor.

That said, the Court does not find that these individual variations are fatal to the notion that the comparability of forms of leave is capable of classwide determination with respect to the control factor. Although Delta attempts to compare the “control” factor among different forms of leave at an extremely fine-grained level of particularity, the Court is unpersuaded that USERRA's statutory text and implementing regulations require that degree of analysis. Section 4316(b) of USERRA provides that a person on military leave is “entitled to such other rights and benefits ․ as are generally provided by the employer ․ to employees having similar seniority, status, and pay who are on ․ leave of absence ․” 38 U.S.C. § 4316(b) (emphasis added). While the DOL's implementing regulations require a military employee to show that he or she “took a form of non-military leave that is comparable to military leave[,]” Myrick, 69 F.4th at 1315 (emphasis added), the regulations expressly use the language of a comparable “form” or “type” of leave. 20 C.F.R. § 1002.150(b) (emphasis added). The use of such language suggests that the comparison between two different forms or types of leave should occur at some higher level of generality or abstraction than the micro-level of how every employee schedules every particular instance of leave. See Form, Black's Law Dictionary (12th ed. 2024) (“The outer shape, structure, or configuration of something, as distinguished from its substance or matter[.]”); Form, Merriam Webster Dictionary, https://www.merriam-webster.com/dictionary/form (last accessed July 21, 2026) (“[T]he shape and structure of something as distinguished from its material[.]”); Type, Merriam Webster Dictionary, https://www.merriam-webster.com/dictionary/type (last accessed July 21, 2026) (“[A] particular kind, class, or group[.]”).

Indeed, if forms of leave were to be compared at such a micro-level, that would appear to undermine the ability to ever compare two different forms of leave as to control under USERRA. Every form of leave has some variation between individual instances of a person taking that form of leave, including with respect to a person's ability to choose when to take that form of leave. Consider the example of bereavement leave. At an intermediate level of abstraction, one might say that an employee who takes bereavement leave to grieve for a family member's death has no ability to choose when to take the leave. After all, an employee does not have any control over when his or her family member passes. If one squints hard enough, however, one can see variation between the ability to control different instances of bereavement leave. Some employees may have no control over when to schedule bereavement leave; the funeral is scheduled by another family member and the employee attends. But what of the employee who is responsible for organizing the funeral? That person may be able to schedule the funeral between a range of dates and therefore has some “ability ․ to choose when to take the leave[.]” 20 C.F.R. § 1002.150(b).

Take another example: jury duty leave. Again, at a certain level of generality, one might find that employees on jury duty do not have the ability to control when to take leave. The prototypical example is that a court mails the employee a jury summons, and the employee then dutifully shows up at the courthouse on the date and time specified on that summons. In the Court's experience, reality is far more complicated. Many potential jurors seek to postpone their jury service when it conflicts with work, education, or family obligations. And many jurisdictions permit potential jurors to reschedule their jury duty dates one or more times. Some jurisdictions may permit more flexibility in jury service postponement than others. See, e.g., Clarkson, 59 F.4th at 439 (“[P]ilots receive advance notice of jury duty and have some flexibility to reschedule it.”).

This same sort of intra-form variation, as the limited record before the Court reflects, appears to be true of military leave taken by pilots. Some military assignments may be scheduled among a set of limited choices; other duties may present no flexibility whatsoever. The Court, however, highly doubts that USERRA should be interpreted to preclude comparison between different forms of leave merely because there is some variation among individual instances of taking each form of leave. That would appear to make the comparability inquiry set forth in the DOL regulations a dead letter. Instead of comparing different forms of leave, employers and employees would have to compare every instance of leave across the three factors enumerated in USERRA's implementing regulations. Not only would such a rule be unworkable,⁸ but it would also severely undermine USERRA's purpose of minimizing the disadvantages and disruptions associated with military service.

The better reading is that USERRA's implementing regulations, by using the terms “form” and “type” of leave, call for a factfinder to make qualitative comparisons between forms of leave at a higher level of generality than at the granular level of every individual instance of leave. Those comparisons, of course, can and should account for the range of control that each form of leave presents. To be sure, the factfinder in this case may credit Delta's argument that military pilots often have some flexibility to schedule their STML and therefore STML taken by pilots may not be comparable to certain forms of leave that have less flexibility (such as bereavement leave). Equally, the factfinder might determine that pilots exercise similar levels of control over STML and other forms of leave, such as jury duty leave, which often can be rescheduled. However, the mere fact that there is some variation in the scheduling procedures for different military assignments (and other non-military forms of leave) does not defeat the ability to compare STML taken by pilots with other non-military forms of leave on a classwide basis.

Although the case law applying USERRA in the Eleventh Circuit and elsewhere has not explicitly addressed the level of generality or abstraction at which USERRA's comparability inquiry operates with respect to the control factor, the reasoning set forth in those cases points towards more generalized comparisons. For instance, the Myrick Court found that paid administrative leave and military leave were “similar in terms of control” because “[m]ilitary employees do not control when they will be summoned for active-duty service, just as non-military employees do not control when [the City of] Hoover will launch an investigation and place them on administrative leave.” Myrick, 69 F.4th at 1319. Although Myrick suggested that evidence that certain military assignments were voluntary could have altered its determination of control as a factor, it did not suggest that the comparative inquiry had to proceed on an individual leave-by-individual leave basis. See id. at 1319 n.6. Likewise, Sorenson determined that “as to the ability to choose when to take leave, Delta pilots can decide for themselves if they want to take a known leave of absence when it is offered. Military pilots have no choice. They must take leave when they are ordered to report to military service.” Sorenson, 174 F.4th at 64. That is, the Eleventh Circuit discussed the ability of “Delta pilots” or “[m]ilitary pilots” to choose when to schedule the relevant forms of leave in general, not on an individualized assignment-by-assignment basis. Id.

In a USERRA class action brought by pilots against American Airlines, the Third Circuit determined that “a jury could conclude that—as with pilots who take jury-duty or bereavement leave—most pilots who take military leave lack the ability to control their schedules.” Scanlan, 102 F.4th at 171 (emphasis added). A Ninth Circuit decision in another pilot class action noted that “[a] jury could ․ find that pilots' level of control over their military duty is comparable to their level of control over other types of leaves. For instance, pilots receive advance notice of jury duty and have some flexibility to reschedule it. Pilots can also use sick leave for pre-scheduled appointments or procedures.” Clarkson, 59 F.4th at 439. These sister circuits thus endorsed a generalized comparison between different forms of leave, at least with respect to the control factor, rather than a specific instance-by-instance inquiry.

Here too, the comparison between pilots' ability to choose when to take STML and other forms of leave can be determined on a classwide basis through common proof. Plaintiffs may rely on their own and other putative class members' testimony to establish the scheduling procedures for the military assignments that pilots complete while on STML. Likewise, Plaintiffs may also testify regarding the level of control they have to choose when to take other forms of leave such as jury duty or bereavement leave. The jury can then make comparability determinations based on that common proof. The Court therefore finds that comparability with respect to control also presents a common question capable of classwide determination.

b. USERRA's Notice Requirement

Delta also contends that whether pilots provided “sufficient advance notice” of each instance of STML is an individualized issue that predominates over common questions in the case. (Doc. 123 at 21-24; Doc. 170 at 10-12.) As explained above with respect to the typicality requirement, the Court finds that while USERRA requires service members to provide advance notice of military leave to their employers, it does not require notice of a particular length. See 20 C.F.R. § 1002.85 (explaining that USERRA requires service members to provide advance notice to their employers but “does not specify how far in advance notice must be given to the employer”); supra at 32-36 (finding that an employee need only provide some advance notice of military leave to satisfy USERRA, rather than notice of any particular length). Accordingly, Plaintiffs and other class members will have to show that they provided advance notice for each instance of STML they seek damages for, though they need not demonstrate that they provided notice of a particular or “sufficient” length. Id.

Nevertheless, the Court agrees with Delta that whether the notice requirement was satisfied for each period of STML presents an individualized issue. Plaintiffs will have to prove advance notice was provided for each individual instance of STML. Contrary to Delta's position, however, the Court finds that this issue is not one that predominates over common questions in the case. Plaintiffs have presented evidence that Delta has data that should resolve the notice question for many of the instances of STML taken by class members. From 2018, pilots reported their STML through a Delta system called “iCrew” that records the date and time that the STML was reported to Delta. (See Zawislak Dep., Doc. 138 at 56 (p.55) (testifying that iCrew records “the date [the military leave] was reported to the company”).) In addition, from 2017, Delta maintained a military leave “notification log” that also recorded the date on which Delta was notified of STML. (Id. at 56-57 (pp.55-56) (noting that military leave notification log reflects “when Delta was aware” of the leave).) As such, to the extent a particular instance of STML was recorded in the iCrew system or military leave notification log, the question of whether advance notice was provided to Delta can be resolved with reference to that data.

Delta, however, argues that the iCrew and notification log data does not cover a large portion of the STMLs included in the class definition because Plaintiffs seek to certify a class of Delta pilots who took STML from April 30, 2007, to the date of judgment in this action. (Doc. 170 at 11 n.4; see Doc. 183 ¶ 1). Delta also contends that the iCrew and military notification log data is incomplete because pilots can give notice of STML to Delta verbally. (Doc. 170 at 11 n.4.) Those points are well taken, but the Court believes that they can be addressed though an administratively feasible process. Should Plaintiffs prevail on their claims, there will be a damages and/or claims administration stage where individualized damages are calculated for each class member. At that stage, the Court can require that class members provide evidence that they gave Delta advance notice as to each instance of STML for which they seek damages. This process need not be administratively burdensome. For instance, in the absence of Delta records showing when it received notice, class members could sign a written form declaration, under penalty of perjury, attesting to the fact that they provided advance notice to Delta of the relevant periods of STML for which they seek compensation.

In short, although the question of whether advance notice was provided to Delta for each period of STML is an individual one, the Court finds that it does not predominate over the common issues presented by Plaintiffs' USERRA claim. Moreover, even if the Court were to later determine that its prediction about the feasibility of individually demonstrating notice at the claims administration stage was incorrect, the Court could amend the class definition to include only the date ranges for which Delta's iCrew system and military notification log provide evidence of advance notice.

c. Damages

The Court now turns to Delta's argument that the individualized calculation of damages will predominate over the issues common to the class. As a threshold matter, “the presence of individualized damages issues does not prevent a finding that the common issues in the case predominate.” Allapattah Servs., Inc. v. Exxon Corp., 333 F.3d 1248, 1261 (11th Cir. 2003); see Carriuolo v. Gen. Motors Co., 823 F.3d 977, 988 (11th Cir. 2016) (“[I]ndividualized damages calculations are insufficient to foreclose the possibility of class certification, especially when, as here, the central liability question is so clearly common to each class member.”); Brown, 817 F.3d at 1239 (“The ‘black letter rule’ recognized in every circuit is that ‘individual damage calculations generally do not defeat a finding that common issues predominate.’ ” (quoting William B. Rubenstein, Newberg on Class Actions § 4:54 (5th ed.))); Klay v. Humana, Inc., 382 F.3d 1241, 1259 (11th Cir. 2004). Indeed, “[d]istrict courts have many tools to decide individual damages: ‘(1) bifurcating liability and damage trials with the same or different juries; (2) appointing a magistrate judge or special master to preside over individual damages proceedings; (3) decertifying the class after the liability trial and providing notice to class members concerning how they may proceed to prove damages; (4) creating subclasses; or (5) altering or amending the class.’ ” Brown, 817 F.3d at 1239 (quoting In re Visa Check/MasterMoney Antitrust Litig., 280 F.3d 124, 141 (2d Cir. 2001), abrogated in part on other grounds by In re Initial Pub. Offerings Sec. Litig., 471 F.3d 24 (2d Cir. 2006)). Therefore, individualized damages issues only predominate over common questions “if computing [damages] will be so complex, fact-specific, and difficult that the burden on the court system would be simply intolerable or if significant individualized questions go to liability.” Green-Cooper, 73 F.4th at 893 (cleaned up).

At the class certification stage, Plaintiffs must prove “that a reliable damages methodology exist[s],” but need not demonstrate “the actual damages [that] [P]laintiffs sustained.” Id. Specifically, Plaintiffs must show that their proposed damages model is consistent with their theory of liability. See Comcast, 569 U.S. at 35 (“[A] model purporting to serve as evidence of damages in this class action must measure only those damages attributable to that theory.”); Carriuolo, 823 F.3d at 988 (“[A]ny model supporting a plaintiff's damages case must be consistent with its liability case.” (quoting Comcast, 569 U.S. at 35)). Further, any proposed model cannot be “arbitrary” or “speculative, and must at least measure damages as a matter of “just and reasonable inference[.]” Comcast, 569 U.S. at 35 (quotation omitted). That said, “there is no categorical prohibition on a court relying on an unexecuted damages model to certify a class[.]” Schultz v. Emory Univ., No. 23-12929, 2024 WL 4534428, at *6 (11th Cir. Oct. 21, 2024) (quoting Lytle v. Nutramax Labs., Inc., 99 F.4th 557, 575 (9th Cir. 2024), amended and superseded on other grounds, 114 F.4th 1011 (9th Cir. 2024), cert. denied, 145 S. Ct. 1308 (2025)) (noting that an expert's “skeletal [damages] models may well be sufficient evidence that, at trial, the models can be fleshed out into actual class-wide damages calculations”).

Plaintiffs' economic expert, Dr. Cavazos, sets forth a proposed damages model for calculating the compensation that a class member would be due for each period of STML. (See Doc. 118-5 ¶¶ 21-30.) In his expert report, Dr. Cavazos starts with the premise that pilots “receive[ ] pay either for flights flown or for days of reserve duty[.]” (Id. ¶ 26.) Dr. Cavazos states that “the amounts that Class Members would have received had they continued working during periods of short-term military leave can be computed by determining the amounts that would have been paid [to] the employee for dropped flights or for missed periods of reserve availability resulting from [STML], in accordance with the terms of the applicable collective bargaining agreement.” (Id.) Dr. Cavazos provides the following example for calculating damages with respect to one instance of STML taken by a pilot:

[C]onsider the example of the Delta pilot with ID#723560, who took a period of short-term military leave on October 28, 2019 through November 1, 2019, at which time he or she had a pay rate of $262.84 per hour. Under the terms of the applicable collective bargaining agreement, that pilot was entitled to five hours and fifteen minutes of pay for each day of unscheduled legal duty leave or the value of rotations dropped or reserve availability missed for scheduled legal duty leave (See Exhibit “A”, no. 29 p. 13-3.). They were also entitled to be paid as shown in their line for up to four consecutive days for a death in the immediate family (See Exhibit “A”, no. 29 p. 13-6 to 13-7.). And they were entitled to between 50 and 270 sick leave credit hours per year, depending on seniority, with pay and credit based on a formula taking into account the scheduled value of rotations lost for regular line pilots and the reserve guarantee and on-call days missed for reserve pilots (See Exhibit “A”, no 29 p. 14-3 and p. 14-5.). If I were provided with information regarding whether this pilot was assigned a regular line or flying reserve as of the dates of their leave, and then either the duration in hours of trips dropped by reason of the short-term military leave or the reserve availability days missed by reason of the short-term military leave (as applicable), ․ I would be able to calculate the additional wages to which they would be entitled under scenarios where it was determined that each of these forms of leave were comparable to short-term military leave. As to a scenario where it was determined that leave to perform jury duty was comparable to short-term military leave, the Delta employee with ID#723650 would be entitled to either five hours and fifteen minutes of pay per day of short-term military leave (i.e., $1,450.87) or the value of trips dropped or reserve availability missed by reason of the short-term military leave. As to a scenario where it was determined that leave due to death of an immediate family member was comparable to short-term military leave, the pilot would be entitled to be paid as shown in their line for the first four days of short term military leave (i.e., the value of trips dropped or reserve availability missed in the first four days of the short-term military leave). And as to a scenario where it was determined that that sick leave was comparable to short-term military leave, the pilot would be entitled to the scheduled value of the rotations lost because of the short-term military leave, a pro rata share of their average line value, a reserve pro rata share for on-call days lost due to the short-term military leave, or a pro rata portion of their reserve guarantee, as applicable (See Exhibit “A”, no. 29 p.14-5.). To the extent that data showing the specific trips or on-call days dropped by pilots is unavailable, damages could still be calculated by using the compensation procedures for known leaves, which are based on paying and crediting a pro rata portion of a pilot's average line value or reserve guarantee.

(Id. ¶ 28.)

Essentially, Dr. Cavazos sets forth a damages model that mirrors the methods that Delta uses to calculate compensation for the allegedly comparable forms of leave—jury (legal) duty leave, bereavement leave, and sick leave. The formula for compensation of these forms of leave is set forth in the PWAs that were applicable to pilots during the relevant time periods encompassed by the class definition. (See, e.g., Doc. 118-7 at 170-71 (legal duty), 173-74 (bereavement leave), 181-84 (sick leave).) Because Plaintiffs' proposed damages model follows the compensation methodology for each of the forms of leave that Plaintiffs allege are comparable to STML, the damages model matches the theory of liability underlying Plaintiffs' USERRA claim.

Nevertheless, Delta argues that Plaintiffs' damages model is insufficient and that individualized damages issues will predominate over common questions. First, Delta asserts that Plaintiffs' damages model is not “attributable to” Plaintiffs' theory of liability and would overcompensate some class members and undercompensate others. (Doc. 123 at 30.) But Delta does not fully explain why that is so when Plaintiffs' damages model expressly follows the same methods that Delta uses to calculate compensation for allegedly comparable forms of leave. For instance, Delta takes issue with Dr. Cavazos' statement that, if STML is determined to be comparable to bereavement leave, a pilot would be “entitled to be paid as shown in their line for the first four days of short term military leave (i.e., the value of trips dropped or reserve availability missed in the first four days of the short-term military leave).” (Doc. 118-5 ¶ 28; see Doc. 123 at 30.) Delta argues that this assumption “ignores the possibility that the pilot notified Delta of her STML before the schedule was posted, that Delta then accommodated her schedule around the STML, and/or the pilot then achieved a full schedule by adding or swapping flights that might have interfered with her STML.” (Doc. 123 at 30.) Thus, Delta says, the damages model would overcompensate the pilot for damages she did not suffer.

The Court does not interpret Plaintiffs' damages model that way. The model attempts to calculate the amount of pay that a class member would have received had they taken the comparative form of leave—in this example, bereavement leave—instead of STML. Delta's hypothetical raises the question, if a pilot had rescheduled their flights around bereavement leave, would they be entitled to compensation for “the value of trips dropped or reserve availability missed in the first four days of the short-term military leave”? (Doc. 118-5 ¶ 28.) If the answer to that question is no, as Delta's argument appears to presuppose, then Plaintiffs' damages model also would not provide any compensation for STML in that same scenario (because no “trips [were] dropped” and/or no “reserve availability [was] missed”). (Id.)

Delta also uses the bereavement leave example to argue that “if the pilot's STML was longer than four days, and she was forced to drop flights on the fifth and sixth days for purposes of STML, the example would result in underpayment.” (Doc. 123 at 31.) That appears to misunderstand the theory of liability underlying Plaintiffs' USERRA claim. Under USERRA, Plaintiffs are only entitled, if they prevail, to the same amount of compensation they would have received had they taken the comparative form of leave instead of STML. See 38 U.S.C. § 4316(b) (providing that servicemembers are “entitled to such other rights and benefits ․ as are generally provided by the employer” to similarly situated employees on a leave of absence (emphasis added)); White, 987 F.3d at 624 (“USERRA mandates only equality of treatment; it does not specify how generous or how parsimonious an employer's paid leave policies must be.”). As such, if Delta only provides 4 days of compensation for bereavement leave, even when the leave lasts longer, the same rule applies to STML. Just as with bereavement leave, Plaintiffs would only receive compensation for 4 days of STML even if they took a longer period of STML. Thus, Delta's argument on this score demonstrates that Plaintiffs' damages model does match their theory of liability under USERRA.

Second, Delta argues that “the inherent variability in the scheduling and pay of Delta pilots precludes a model that would calculate damages on a classwide basis.” (Doc. 123 at 31.) Delta further contends that Plaintiffs cannot satisfy the predominance requirement because calculating individual damages “will be so complex, fact-specific, and difficult that the burden on the court system would be simply intolerable.” (Id. (quoting Brown, 817 F.3d at 1239-40).) Specifically, Delta asserts that there is no “easy way” for Plaintiffs to calculate the “dropped flights” or “missed periods of reserve availability” resulting from each period of STML. (Id. at 32.) Delta explains that it schedules pilots for flights using a “Preferential Bidding System (‘PBS’)” by which pilots bid on certain flight schedules “depending on seniority, base location, and other factors[.]” (Id.) Delta emphasizes that its “pilots are not paid a fixed salary nor are they paid a fixed number of hours[,]” but instead are paid by the flight hour or for on-call reserve periods depending on the situation. (Id. at 33.) According to Delta, “[b]ecause of the flexibility in scheduling and the variability in reporting of military leaves, it is not possible to determine with any degree of accuracy what a pilot would have been paid had she not taken a military leave.” (Id.)

Plaintiffs respond that Delta's argument is essentially that it “is impossible to determine for any individual pilot what damages they are entitled to for any of Delta's violations of USERRA § 4316(b).” (Doc. 126 at 25.) Plaintiffs contend that “the damages analysis need not determine the amount that a pilot would have earned in a world where they never performed military service.” (Id. at 26-27.) Rather, class members are “only entitled to the same non-seniority rights and benefits that Delta provides to employees who take leaves comparable to short-term military leave[.]” (Id. at 27.) Plaintiffs argue that the damages model set forth by Dr. Cavazos is reasonable because it “examines the comparator leave and affords a pilot with the same rights and benefits as a pilot who took that comparator leave instead of short-term military leave” in accordance with the applicable PWA. (Id.) As the relevant PWAs detail, the formulas for calculating compensation vary between each of the allegedly comparable form of leave. To the extent that the PWAs call for compensation for flights dropped or periods of reserve availability missed, Plaintiffs note that “Delta maintains records of pilot schedules as originally issued to pilots, before any modifications due to military leave.” (Doc. 126 at 28; see Zawislak Dep., Doc. 123-3 at 178-79 (pp.177-78) (stating that Delta archives the original pilot schedules issued to pilots).) Plaintiffs argue that they can rely on the archived pilot schedules to identify dropped flights and missed reserve availability by comparing final pilot schedules to original pilot schedules. (Doc. 126 at 28.)

In the Court's view, Plaintiffs have the better argument at this stage of proceedings. Although individual calculation of damages will be required, the Court does not find that the calculations will be “so complex, fact-specific, and difficult that the burden on the court system would be simply intolerable.” Brown, 817 F.3d at 1240 (quotation omitted). Dr. Cavazos has outlined a reasonable approach to damages that closely follows the applicable PWAs' policies for providing compensation for allegedly comparable forms of leave. While Delta highlights several complexities with respect to pilot scheduling, Delta does not persuasively explain how these issues would preclude the calculation of damages in the same manner that compensation for comparable forms of leave is calculated under the applicable PWAs. After all, Delta does not contest that it regularly calculates and provides compensation for allegedly comparable forms of leave such as jury duty leave, bereavement leave, and sick leave. It is therefore somewhat implausible, based on the present record, that Dr. Cavazos could not estimate the compensation a pilot would have received had that pilot taken a comparable form of leave rather than STML.

That said, the Court recognizes that, at present, Dr. Cavazos' damages model remains to be executed. For instance, Plaintiffs indicate that they will use Delta's pilot scheduling archives to estimate the compensation that a pilot would have received for STML had the pilot taken a comparable form of leave. But they have not yet performed such a calculation with actual scheduling data. Should Plaintiffs prevail on demonstrating Delta's liability under USERRA, the Court can and will reassess whether individual damages can be calculated without the need for proving damages in an individual action. If Plaintiffs cannot later show that their damages model will produce reliable estimates of the compensation each class member is due, the Court will consider decertifying the class following a trial on liability and requiring class members to prove damages in individual proceedings. However, at this preliminary stage, Plaintiffs have satisfied their burden to demonstrate “that a reliable damages methodology exist[s][.]” Green-Cooper, 73 F.4th at 893. Dr. Cavazos' proposed damages model constitutes “sufficient evidence that, at trial, the models can be fleshed out into actual ․ damages calculations.” Schultz, 2024 WL 4534428, at *6. Accordingly, the Court finds that the individualized calculation of damages will not predominate over the issues common to the class.

d. Weighing the Relative Importance of Common and Individualized Questions

After analyzing the common and individualized questions posed by Plaintiffs' USERRA claim, the Court concludes that the common issues in this case outweigh the individualized issues of notice and calculation of damages. The core question in this case is whether, under USERRA, STML is comparable to other forms of leave that Delta provides compensation for—namely, jury duty leave, sick leave, and bereavement leave. As explained above, that comparability inquiry will be resolved exclusively through classwide proof. With respect to liability, the comparability inquiry will drive resolution of Plaintiffs' USERRA claim.

To be sure, should Plaintiffs prevail on the merits of their USERRA claim, the calculation of damages will require individualized proof. Class members will also have to show that they provided notice to Delta as to each period of STML for which they seek compensation. On this record, however, the Court does not view those issues as sufficiently burdensome or complex to defeat predominance. The Court has several options to manage individualized damages inquiries. See Brown, 817 F.3d at 1239. If necessary, the Court can modify the class definition or decertify the class following a determination on liability and require each class member to prove damages in an individual proceeding. But given that Plaintiffs have identified over 3,000 Delta pilots who took STML within the proposed class period, it would be unreasonable (and wasteful of judicial resources) to require each pilot to make a separate showing of liability under USERRA when liability can be decided using classwide proof. See, e.g., Klay, 382 F.3d at 1260 (“It is ridiculous to expect 600,000 doctors across the nation to repeatedly prove these complicated and overwhelming facts [related to liability].”). Under the circumstances of this case, the Court finds that the relative importance of the common issues concerning liability on Plaintiffs' USERRA claim predominates over the individualized damages and notice issues. See, e.g., Huntsman, 2021 WL 391300, at *14 (“[T]he common issues identified by plaintiff, whether paid leave is a ‘right and benefit’ under 38 U.S.C. § 4316(b) and whether short-term military leave is comparable to other forms of paid leave, can be resolved on a classwide basis. Because plaintiff's USERRA claim turns on Southwest's generally applicable policies, predominance is met.”); Clarkson, 2020 WL 4495278, at *6 (concluding that “central USERRA questions about liability predominate over individual considerations such as damages amounts” in a putative class action brought by pilots against their airline).

2. Superiority

To certify a class under Rule 23(b)(3), the Court must find that “a class action is superior to other available methods for fairly and efficiently adjudicating the controversy.” Fed. R. Civ. P. 23(b)(3). The following factors are relevant to the Court's inquiry:

(A) the class members' interests in individually controlling the prosecution or defense of separate actions;

(B) the extent and nature of any litigation concerning the controversy already begun by or against class members;

(C) the desirability or undesirability of concentrating the litigation of the claims in the particular forum; and

(D) the likely difficulties in managing a class action.

Fed. R. Civ. P. 23(b)(3). “Certification should not be denied for lack of superiority solely because a class action would make a district court's task complex or difficult.” In re Conagra Peanut Butter Prods. Liab. Litg., 251 F.R.D. 689, 699 (N.D. Ga. 2008). The “focus” of the superiority inquiry “is not on the convenience or burden of a class action suit per se, but [instead] on the relative advantages of a class action suit over whatever other forms of litigation might be realistically available to the plaintiffs.” Klay, 382 F.3d at 1269.

“[W]here a court has already made a finding that common issues predominate over individualized issues, [it] would be hard pressed to conclude that a class action is less manageable than individual actions.” Williams, 568 F.3d at 1358 (quoting Klay, 382 F.3d at 1273). That is, “the predominance analysis ․ has a tremendous impact on the superiority analysis ․ for the simple reason that, the more common issues predominate over individual issues, the more desirable a class action lawsuit will be as a vehicle for adjudicating the plaintiffs' claims.” Klay, 382 F.3d at 1269.

With respect to the first superiority factor, “[t]here is no reason to believe that the putative class members in this case have any particular interest in controlling their own litigation[.]” Id. The putative class members share the same interest in showing that STML is comparable to forms of leave that Delta compensates and that they are therefore entitled to compensation. As to the second factor, there is no indication, nor has Delta suggested, that a critical mass of putative class members have separately pursued USERRA claims against Delta for unpaid STML. Id.

The Court also finds, with respect to the third factor, that “it is desirable to concentrate this litigation in a single forum.” Id. at 1270. As set forth above, common issues predominate over individualized issues in this case. Because Delta's liability to putative class members under USERRA will be proven or disproven using common evidence, litigating Plaintiffs' USERRA claim as a class action will “offer substantial economies of time, effort, and expense for the litigants as well as for the court.” Id. (cleaned up). By contrast, “[h]olding separate trials ․ would be costly, inefficient, and would burden the court system by forcing individual plaintiffs to repeatedly prove the same facts and make the same legal arguments before different courts.” Id. (quotation omitted); see also Carriuolo v. Gen. Motors Co., 823 F.3d 977, 989 (11th Cir. 2016) (“Because common questions of law and fact predominate, class-wide adjudication appropriately conserves judicial resources and advances society's interests in judicial efficiency.”).

As to the fourth and final factor, “the likely difficulties in managing a class action” do not defeat superiority here. Fed. R. Civ. P. 23(b)(3)(D). “The manageability inquiry focuses on whether a class action ‘will create relatively more management problems than any of the alternatives,’ not whether it will create manageability problems in an absolute sense.” Cherry, 986 F.3d at 1304 (quoting Klay, 382 F.3d at 1273). “[M]anageability problems will rarely, if ever, be in themselves sufficient to prevent certification.” Id. (cleaned up). “Courts are generally reluctant to deny class certification based on speculative problems with case management․ Even potentially severe management issues have been held insufficient to defeat class certification.” Klay, 382 F.3d at 1272-73 (quotation omitted). “[T]he factor of manageability is ordinarily satisfied so long as common issues predominate over individual issues[.]” Williams, 568 F.3d at 1358.

The manageability of this matter, as compared to the alternatives, does not pose a barrier to class certification. As the Court has explained, USERRA's comparability inquiry is the principal issue that will drive the resolution of this case, and it predominates over the individual issues that Delta has identified. Thus, resolving Delta's liability under USERRA on a classwide basis will be more manageable than considering over 3,000 separate lawsuits brought by each individual class member. Klay at 1273 (comparing the manageability of a class action to “600,000 separate lawsuits [brought] by [putative] class members”).

Moreover, as the Court has explained throughout this order, there are solutions to the issues that Delta has raised that might bear on manageability. For instance, Delta has raised concerns about the reliability of its pilots' self-reported military leave data and the administrative feasibility of identifying periods of STML. But class members' PCARS data can be used to verify periods of STML that pilots reported to Delta. Delta has not persuasively argued that such cross-verification would be burdensome, especially given that such a process would involve a straightforward comparison of leave dates.

While there may be some complexity in calculating individual damages, the Court has several tools to manage the damages stage of litigation, including requiring class members to prove damages in individual proceedings before a special master or Magistrate Judge, or decertifying the class after a trial on liability. See Brown, 817 F.3d at 1239. Plaintiffs, moreover, have proposed a damages model that adheres to the same formulas that Delta itself uses to calculate and provide compensation for paid forms of leave such as such as jury duty leave, bereavement leave, and sick leave. Although Plaintiffs' damages model has not yet been executed, the present record before the Court indicates that, should Plaintiffs succeed on the merits of their USERRA claim, Plaintiffs will be able to estimate the compensation a pilot would have received had that pilot taken a comparable form of leave rather than STML.

The Court therefore does not predict “any specific management problems—aside from the obvious ones that are intrinsic in large class actions—that would render a class action impracticable ․” Id. at 1273; see Jancik v. WebMD LLC, No. 1:22-CV-644-TWT, 2025 WL 560705, at *9 (N.D. Ga. Feb. 20, 2025) (“[N]o significant manageability concerns appear to outweigh the manageability concerns that would exist if this case was litigated through separate actions.”). Considering the four factors enumerated by Rule 23(b)(3) and the other circumstances presented by this case, the Court concludes that a class action is superior to individual actions for adjudicating the putative class members' USERRA claim.

VII. Conclusion

For the foregoing reasons, Plaintiffs' motion to partially exclude Dr. Lee's expert opinions (Doc. 127) is DENIED without prejudice. Plaintiffs' motion for class certification (Doc. 183) is GRANTED. The Court certifies the following Injunction Class under Rule 23(b)(2):

All pilots currently employed by Delta who are also currently members of the U.S. Armed Forces.

In addition, the Court certifies the following Damages Class under Rule 23(b)(3):

All current and former pilots who work or worked for Delta in the United States or its territories or possessions who took a military leave of 30 consecutive days or fewer from their employment with Delta as a pilot from April 30, 2007, to the date of judgment in this action, as reflected by Delta's military leave data.

Plaintiff Benjamin A. Best is named as the Class Representative of both classes. Pursuant to Rule 23(g), Plaintiffs' counsel, R. Joseph Barton and Michael Scimone, are appointed as Co-Lead Class Counsel. Further, Stephen Anderson will serve as Liaison Class Counsel.

The parties are DIRECTED to meet and confer regarding next steps in this case and to file a joint status report within 30 days of the entry of this order. The joint status report should address: (1) whether the parties believe mediation would be productive; (2) a proposed schedule for conducting additional discovery and filing motions for summary judgment; and (3) any other topics the parties deem pertinent.

SO ORDERED, this 7th day of August, 2026.

FOOTNOTES

1.   The Court refers to the page numbers generated by CM-ECF, even when the internal pagination of a document differs. At times, for the sake of clarity, the Court will also include the internal pagination in brackets, denoted by “p.”

2.   Rule 702 states: “A witness who is qualified as an expert by knowledge, skill, experience, training, or education may testify in the form of an opinion or otherwise if the proponent demonstrates to the court that it is more likely than not that: (a) the expert's scientific, technical, or other specialized knowledge will help the trier of fact to understand the evidence or to determine a fact in issue; (b) the testimony is based on sufficient facts or data; (c) the testimony is the product of reliable principles and methods; and (d) the expert's opinion reflects a reliable application of the principles and methods to the facts of the case.” Rule 702 was amended effective December 1, 2023, to “clarify and emphasize” that expert testimony “may not be admitted unless the proponent demonstrates to the court that it is more likely than not that the proffered testimony meets the admissibility requirements set forth in the rule.” See Fed. R. Evid. 702 advisory committee note to 2023 amendment.

3.   Indeed, as Delta points out, any contrary interpretation would defy common sense. If scheduled days off were not counted in calculating the length of a military leave, every military leave, no matter how long, would be measured as a series of STMLs merely because the leave period included certain days off. For instance, the military leave of a Monday-to-Friday employee who was away from their employment for a year would be measured as 52 separate STML periods from Monday to Friday every week, instead of as a continuous one-year period, because weekends would not be included in the leave period. This would appear to conflict with USERRA's implementing regulations, which provide that “a two-day funeral leave will not be ‘comparable’ to an extended leave for service in the uniformed service.” 20 C.F.R. § 1002.150(b).

4.   The only case that Delta cites is a district court decision on a motion for renewed judgment as a matter of law, or alternatively, for a new trial. (Doc. 123 at 22 (citing Reese v. Army Fleet Support, LLC, No. 16-CV-227-TFM, 2018 WL 4214143, at *3 (M.D. Ala. Sept. 4, 2018).) That decision, however, only noted in passing that USERRA requires a plaintiff to prove that he “gave advance notice of [his] service[.]” Reese, 2018 WL 4214143, at *3. It does not discuss the notice requirement in any detail and does not suggest that a plaintiff must demonstrate notice of any particular length.

5.   Although Delta states that Plaintiffs “have not shown that a single injunction would benefit the whole class[,]” that is all it says. (Doc. 185 at 13.) Delta does not explain that statement or connect it to the specific injunctive relief requested by Plaintiffs in this case.

6.   Delta points to a case where the Eleventh Circuit found that the plaintiff “never identified exactly what injunctive or declaratory relief it was seeking” because it merely requested “such declaratory and injunctive relief as appropriate in order to compel and ensure [the defendant's] future compliance with law.” Lakeland Reg'l Med. Ctr., Inc. v. Astellas US, LLC, 763 F.3d 1280, 1291 (11th Cir. 2014). By contrast, Plaintiffs' requested injunction here is considerably more specific. Plaintiffs request that the Court “[o]rder Delta to pay pilots who take short-term military leave on the same basis as pilots who take leave for the forms of comparable short-term, non[-]military leave” as determined by the factfinder. (Doc. 182 at 25.) The Court has no difficulty discerning the nature of the injunctive relief that Plaintiffs are seeking under USERRA.

7.   Of course, the data may also show that certain periods of STML are not comparable in duration to certain periods of jury service. For instance, if Delta's leave data shows that the maximum length of jury leave that any pilot has taken is 10 days, the factfinder might determine that periods of STML which exceed 10 days are not comparable to jury duty leave in terms of duration.

8.   Consider, for instance, an employer based in downtown Atlanta who provides compensation for up to 10 days of jury duty leave. Consider further that the employer was faced with a military employee's claim that his STML of 10 days or less should be compensated on the same terms as jury duty leave under USERRA. When comparing the two forms of leaves in terms of control, which specific instance of jury duty leave should the employer choose? Should the employer assess control over scheduling jury duty leave based on the scheduling rules for jury duty in state court in Fulton County, Georgia, or the scheduling rules for jury duty in the U.S. District Court for the Northern District of Georgia? What about employees who do not live in Fulton County and commute to work from surrounding counties? Should the comparability inquiry regarding control vary based on where each employee lives and their jurisdiction's specific rules? As these questions illustrate, if the comparability inquiry were to take place at such a granular level—rather than at a level of abstraction permitting a generalized comparison between two different forms of leave—it would quickly become unworkable.

SARAH E. GERAGHTY, United States District Judge