Casey Clarkson v. Alaska Airlines, Inc.

FOR PUBLICATION UNITED STATES COURT OF APPEALS FOR THE NINTH CIRCUIT CASEY CLARKSON, No. 21-35473 Plaintiff-Appellant, D.C. No. 2:19-cv- 00005-TOR v. ALASKA AIRLINES, INC.; OPINION HORIZON AIR INDUSTRIES, INC., Defendants-Appellees. Appeal from the United States District Court for the Eastern District of Washington Thomas O. Rice, District Judge, Presiding Argued and Submitted October 21, 2022 Portland, Oregon Before: Richard A. Paez and Bridget S. Bade, Circuit Judges, and Haywood S. Gilliam, Jr., * District Judge. Opinion by Judge Paez * The Honorable Haywood S. Gilliam, Jr., United States District Judge for the Northern District of California, sitting by designation. 2 CLARKSON V. ALASKA AIRLINES, INC. SUMMARY ** Labor Law The panel reversed the district court’s grant of summary judgment in favor of defendants Alaska Airlines, Inc., and Horizon Air Industries, Inc., and remanded, in a class action brought under the Uniformed Services Employment and Reemployment Rights Act (USERRA) by Casey Clarkson, a commercial airline pilot and military reservist. Clarkson alleged that because the airlines provided paid leave for non-military leaves, including jury duty, bereavement, and sick leave, the airlines were also required to pay pilots during short-term military leaves of thirty days or less. Under USERRA § 4316(b)(1), “a person who is absent from a position of employment by reason of service in the uniformed services” shall be “deemed to be on furlough or leave of absence” and shall be “entitled to such other rights and benefits not determined by seniority as are generally provided by the employer” to other employees on non- military furloughs or leaves of absence. Under 20 C.F.R. § 1002.150, the “non-seniority rights and benefits to which an employee is entitled during a period of service are those that the employer provides to similarly situated employees.” If the benefits vary according to the type of leave, the employee must be given “the most favorable treatment accorded to any comparable form of leave when he or she ** This summary constitutes no part of the opinion of the court. It has been prepared by court staff for the convenience of the reader.  CLARKSON V. ALASKA AIRLINES, INC. 3 performs service in the uniformed services.” To determine whether types of leave are comparable, the duration of the leave must be considered, as well as the purpose of the leave and the ability of the employee to choose when to take the leave. The panel held that the district court erred in concluding that no reasonable jury could find military leave comparable to non-military leave. In reaching this conclusion, the district court erred by comparing all military leaves, rather than just the short-term military leaves at issue here, with the comparator non-military leaves. The district court also erred by disregarding factual disputes about each of the three factors in the comparability analysis: duration, purpose, and control. The panel held that because factual disputes existed, comparability was an issue for the jury. The panel therefore reversed and remanded. It instructed that on remand, the district court should consider in the first instance the issue whether “pay during leave” was a standalone benefit that the airlines provided under their collective bargaining agreements to any employee on leave. 4 CLARKSON V. ALASKA AIRLINES, INC. COUNSEL Jonathan E. Taylor (argued), Deepak Gupta, Peter Romer- Friedman, and Robert Friedman, Gupta Wessler PLLC, Washington, D.C.; Adam T. Klein and Michael J. Scimone, Outten & Golden LLP, New York, New York; Vincent Chang, Block & Leviton LLP, San Francisco, California; R. Joseph Barton and Colin M. Downes, Block & Leviton LLP, Washington, D.C.; Matthew Z. Crotty, Crotty & Son Law Firm PLLC, Spokane, Washington; Thomas G. Jarrard, The Law Office of Thomas G. Jarrard PLLC, Spokane, Washington; for Plaintiff-Appellant. Anton Metlitsky (argued), Mark W. Robertson, and Charles Mahoney, O’Melveny & Myers LLP, New York, New York; Jason Zarrow, O’Melveny & Myers LLP, Los Angeles, California; Tristan Morales, O’Melveny & Myers LLP, Washington, D.C.; Steven W. Fogg, Corr Cronin LLP, Seattle, Washington; Kathryn S. Rosen, Davis Wright Tremaine LLP, Seattle, Washington; for Defendants- Appellees. David T. Raimer and Douglas W. Hall, Jones Day, Washington, D.C., for Amicus Curiae Airlines for America.  CLARKSON V. ALASKA AIRLINES, INC. 5 OPINION PAEZ, Circuit Judge: 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. In the Selective Training and Service Act of 1940, Congress ensured for the first time—but not the last—that veterans returning to civilian jobs would not face discrimination on account of their service. Over the succeeding decades, re- employment rights were extended to military reservists and National Guardsmen. These protections remain all the more important today, as our nation relies on an all-volunteer military force. Indeed, just as the draft came to an end, Congress expanded servicemembers’ protections in the Veterans’ Reemployment Rights Act of 1974. Congress continued its tradition of recognizing the sacrifice and dedication of servicemembers in 1994 by enacting the Uniformed Services Employment and Reemployment Rights Act (“USERRA”). Today, USERRA § 4316(b)(1) requires employers to provide employees who take military leave with the same non-seniority rights and benefits as their colleagues who take comparable non-military leaves. See 38 U.S.C. § 4316(b)(1); 20 C.F.R. § 1002.150(a). Casey Clarkson (“Clarkson”), a commercial airline pilot and military reservist, claims that his employers failed to abide by this rule. Clarkson alleges that because Alaska Airlines and Horizon Air Industries (collectively, the “Airlines”) provide paid leave for non-military leaves including jury duty, bereavement, and sick leave, the Airlines are also required to pay pilots during short-term military leaves. The district court disagreed, granting 6 CLARKSON V. ALASKA AIRLINES, INC. summary judgment to the Airlines and concluding as a matter of law that military leave is not comparable to any other form of leave offered by the Airlines. We reverse. I. STATUTORY BACKGROUND A. Enacted in 1994, USERRA 1 contains “the most expansive protection to servicemembers yet enacted . . . [and] entitle[s] reservists and other military personnel to certain employment benefits while on leave.” Travers v. Fed. Express Corp., 8 F.4th 198, 201 (3d Cir. 2021) (cleaned up). 2 USERRA serves three primary purposes: (1) “[T]o encourage noncareer 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 . . . disruption . . . by providing for the prompt reemployment of [persons performing military service] 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. In short, USERRA recognizes that those who serve in the military should be supported, rather than penalized, for their service. USERRA enables servicemembers to strike a balance between fulfilling their military duties and civilian obligations, including civilian jobs, without suffering discrimination. See Travers, 8 F.4th at 199. While USERRA is “the most recent in a series of laws protecting 1 Pub. L. No. 103-353, 108 Stat. 3149 (codified at 38 U.S.C. § 4301). 2 For a fuller recounting of the history of USERRA and its predecessor statutes, see Travers v. Fed. Express Corp., 8 F.4th 198 (3d Cir. 2021), and Rogers v. City of San Antonio, 392 F.3d 758 (5th Cir. 2004).  CLARKSON V. ALASKA AIRLINES, INC. 7 veterans’ employment and reemployment rights,” Rogers v. City of San Antonio, 392 F.3d 758, 762 (5th Cir. 2004), Congress intended USERRA to build on the rights established in its predecessor statutes, including the Selective Training and Service Act of 1940 (“STSA”), Pub. L. No. 783, 54 Stat. 885, 890, and the Veterans’ Reemployment Rights Act of 1974 (“VRRA”), Pub. L. No. 93-508, 88 Stat. 1578, 1594 (codified at 38 U.S.C. § 2021). See also Travers, 8 F.4th at 199-201. The “large body of case law that had developed under those statutes remain[s] in full force and effect.” Rogers, 392 F.3d at 762 (quoting 20 C.F.R. § 1002.2). An understanding of the history of these protective statutes and the corresponding case law is thus critical to our evaluation of this case. USERRA’s predecessor statutes guaranteed that servicemembers who took leave from a civilian job for military service could return to that job without losing “seniority, status, or pay.” STSA, Pub. L. No. 783, § 8(b), 54 Stat. 885, 890; VRRA, Pub. L. No. 93-508, § 2021(a), 88 Stat. 1578, 1595. These rights were extended to military reserve members and National Guardsmen throughout the 1950s and 1960s. See, e.g., Pub. L. No. 305, § 261(f), 69 Stat. 598, 602 (1955); Pub. L. No. 86-632, § 5, 74 Stat. 467, 468 (1960); accord Monroe v. Standard Oil Co., 452 U.S. 549, 555 (1981); Rogers, 392 F.3d at 764. Reservists, however, did not receive “protection against discharges, demotions, or other discriminatory conduct once reinstated” until 1968. Monroe, 452 U.S. at 556-60. At that time, Congress enacted what became § 2021(b)(3) of the VRRA, 3 3 Before 1974, veterans’ re-employment rights provisions were codified at 50 U.S.C. § 459. These provisions were recodified without 8 CLARKSON V. ALASKA AIRLINES, INC. which provides: “Any person [employed by a private employer] shall not be denied retention in employment or any promotion or other incident or advantage of employment because of any obligation as a member of a reserve component of the Armed Forces of the United States.” Pub. L. No. 90-491, 82 Stat. 790 (1968). Both the House and Senate Reports on the proposed legislation explained that § 2021(b)(3) would “assure[] that these reservists will be entitled to the same treatment afforded their coworkers without such military obligation.” H.R. Rep. No. 90-1303, at 3 (1968); accord S. Rep. No. 90- 1477, at 1-2 (1968). By protecting reservists from discrimination, the provision would encourage voluntary military service. See Monroe, 452 U.S. at 557-60. In 1981, the Supreme Court clarified in Monroe that § 2021(b)(3) entitled reservists to equal, but not preferential, treatment as compared to their non-military coworkers. Id. at 557-66. Five years after the Court decided Monroe, the Third Circuit considered the application of § 2021(b)(3) in Waltermyer v. Aluminum Co., 804 F.2d 821 (3d Cir. 1986). There, the Third Circuit concluded that a National Guardsman was entitled to holiday pay for a holiday that occurred while he was on military leave because employees on non-military leaves, including jury duty, bereavement leave, and sick leave, received holiday pay. Id. at 825. Although the collective bargaining agreement limited holiday pay to employees who worked during the holiday week or employees on certain enumerated leaves (which did not include military leave), the court concluded that the list substantive change in the VRRA. See Coffy v. Republic Steel Corp., 447 U.S. 191, 194 n.2 (1980).  CLARKSON V. ALASKA AIRLINES, INC. 9 of enumerated leaves focused on “involuntary” leaves. Id. Because the National Guardsman’s military leave was also involuntary, the court held that it would be equitable, and not preferential, to require the employer to provide holiday pay under § 2021(b)(3). Id. Congress later codified the Waltermyer decision in USERRA § 4316(b)(1), the provision at issue here. As the House Report on the bill explained: The Committee intends to affirm the decision in Waltermyer v. Aluminum Co. of America, 804 F.2d 821 (3d Cir. 1986) that, to the extent the employer policy or practice varies among various types of non-military leaves of absence, the most favorable treatment accorded any particular leave would also be accorded the military leave, regardless of whether the non-military leave is paid or unpaid. H.R. Rep. 103-65, pt. 1, at 33-34 (1993); accord S. Rep. 103- 158, at 58 (1993). B. Section 4316(b)(1) expanded non-seniority employment benefits for servicemembers, including reservists and Guardsmen, who must take leave from civilian jobs to perform their military duties. Under § 4316(b)(1), “a person who is absent from a position of employment by reason of service in the uniformed services” shall be “deemed to be on furlough or leave of absence,” and shall be “entitled to such other rights and benefits not determined by seniority as are generally provided by the employer” to other employees on 10 CLARKSON V. ALASKA AIRLINES, INC. non-military furloughs or leaves of absence. 38 U.S.C. § 4316(b)(1). 4 The Department of Labor’s (“DOL”) implementing regulation for § 4316(b)(1) explains that the “non-seniority rights and benefits to which an employee is entitled during a period of service are those that the employer provides to similarly situated employees by an employment contract, agreement, policy, practice, or plan in effect at the employee’s workplace.” 20 C.F.R. § 1002.150(a). The regulation then explains that if the benefits vary according to the type of leave, the employee must be given “the most favorable treatment accorded to any comparable form of leave when he or she performs service in the uniformed services.” Id. at § 1002.150(b) (emphasis added). The regulation also explains how to determine if types of leave are comparable: 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, 4 Under USERRA, the term “rights and benefits” is defined as “the terms, conditions, or privileges of employment, including any advantage, profit, privilege, gain, status, account, or interest (including wages or salary for work performed) that accrues by reason of an employment contract or agreement or an employer policy, plan, or practice and includes rights and benefits under a pension plan, a health plan, an employee stock ownership plan, insurance coverage and awards, bonuses, severance pay, supplemental unemployment benefits, vacations, and the opportunity to select work hours or location of employment.” 38 U.S.C. § 4303(2).  CLARKSON V. ALASKA AIRLINES, INC. 11 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. at § 1002.150(b). Clarkson challenges the Airlines’ compliance with this regulation. II. DISTRICT COURT PROCEEDINGS Clarkson filed this action against the Airlines in 2019, alleging that the Airlines violated § 4316(b)(1) of USERRA by failing to pay pilots who took short-term military leave while paying pilots who took comparable non-military leaves. He filed the action on behalf of himself and all others similarly situated. The district court certified a “Paid Leave Class” in August 2020, defined as: “All current or former Alaska or Horizon pilots who have taken short-term military leave from October 10, 2004 through the date of the judgment.” Although the certification order did not define short-term military leave, the parties and court understood it to mean military leaves of thirty days or less. Clarkson and the members of the class were or are employed as pilots by Alaska or Horizon. While employed, Clarkson and the other class members took at least one military leave of thirty days or less. These facts are undisputed. What the parties cannot agree on, however, is whether short-term military leave is comparable to the non-military leaves offered by the Airlines, namely jury duty, bereavement leave, or sick leave. 5 In seeking summary 5 Clarkson also argues that vacation leave may be comparable to short- term military leave. The Airlines do not address vacation leave, asserting that such argument is precluded because Clarkson did not specifically identify vacation leave as a comparator in his complaint. Clarkson’s 12 CLARKSON V. ALASKA AIRLINES, INC. judgment, the Airlines argued that military leave is not comparable to non-military leave as a matter of law. In the alternative, the Airlines argued that even if the leaves were comparable, “pay during leave” is not a benefit offered in their collective bargaining agreements (“CBAs”) to which the class members are entitled. Clarkson did not address this second argument, and instead argued that a reasonable jury could find that some or all of the non-military leaves offered by the Airlines are comparable to short-term military leave. As in the district court, the parties’ arguments focus on the three comparability factors outlined in the regulation: (1) duration of leave, (2) purpose of leave, and (3) ability of the employee to choose when to take the leave (also referred to as “control”). 20 C.F.R. § 1002.150(b). In the district court, each side submitted an expert report comparing the duration of military leave to the duration of other types of leave. The Airlines’ expert conducted his analysis using all military leaves rather than just short-term military leaves of thirty days or less, analyzed the frequency of leaves, and focused on the length of leaves at the highest percentiles rather than the average lengths. As a result, his comparator numbers look quite different from the numbers produced by Clarkson’s expert, who focused on the average, mode, and median days taken of short-term military leave. The parties complaint repeatedly references “other” types of leave besides jury duty, bereavement, and sick leave, and Clarkson included vacation as a comparator leave in his opposition to summary judgment. However, the district court did not address vacation leave as a comparator. Because it is clear that a reasonable jury could find jury duty, bereavement, or sick leave comparable to short-term military leave, we need not address vacation leave to reverse the grant of summary judgment. On remand, the district court may address vacation leave as a comparator in the first instance.  CLARKSON V. ALASKA AIRLINES, INC. 13 also presented conflicting evidence about the purpose of military leave and the ability of pilots to control when they take such leave. The Airlines contend that pilots take military leave to pursue a parallel career, while Clarkson maintains that military leave allows pilots to perform a civic duty and public service. The Airlines also contend that pilots have near total control over when to take military leave, while Clarkson argues that the military schedule is not so flexible. On each factor, the district court determined there were no genuine issues of material fact and concluded that military leave is not comparable to any other leave as a matter of law. The district court thus did not address the “pay during leave” issue and granted summary judgment to the Airlines. Clarkson timely appealed. 6 We reverse. III. STANDARD OF REVIEW We review de novo a district court’s summary judgment ruling to determine whether “there are any genuine disputes of material fact and whether the district court correctly applied the relevant substantive law.” Reynaga v. Roseburg Forest Products, 847 F.3d 678, 685 (9th Cir. 2017) (quoting Dominguez-Curry v. Nevada Trans. Dep’t, 424 F.3d 1027, 1033 (9th Cir. 2005)). “An issue of fact is genuine ‘if the evidence is such that a reasonable jury could return a verdict for the nonmoving party.’” Id. at 685-86 (quoting Villiarimo v. Aloha Island Air, Inc., 281 F.3d 1054, 1061 (9th Cir. 2002)). A grant of summary judgment is “proper only where there is no genuine issue of any material fact or where viewing the evidence and the inferences which may be drawn therefrom in the light most favorable to the adverse 6 We have jurisdiction under 28 U.S.C. § 1291. 14 CLARKSON V. ALASKA AIRLINES, INC. party, the movant is clearly entitled to prevail as a matter of law.” Sandvik v. Alaska Packers Ass’n, 609 F.2d 969, 974 (9th Cir. 1979) (quoting Smith v. Gross, 604 F.2d 639, 641 (9th Cir. 1979)). As we discuss below, the district court failed to heed these summary judgment principles in its comparability analysis. IV. DISCUSSION In entering judgment for the Airlines, the district court concluded that no reasonable jury could find military leave comparable to non-military leave. In reaching this conclusion, the district court erred by comparing all military leaves, rather than just the short-term military leaves at issue here, with the comparator non-military leaves. The court also erred by disregarding countless factual disputes about each of the three factors in the comparability analysis: duration, purpose, and control. The court seemingly considered only the evidence presented by the Airlines when it concluded no reasonable jury could find for Clarkson. Because factual disputes exist, comparability is an issue for the jury. A. Under USERRA, Courts Must Consider the Length of Leave at Issue In concluding that military leave is not comparable to other leaves as a matter of law, the district court first agreed with the Airlines that military leave must be considered as a “general category of leave,” Clarkson v. Alaska Airlines, Inc., 2021 WL 2080199, at *4 (E.D. Wash. May 24, 2021), rather than in an “individualized analysis.” Id. at *5. The district court thus decided that all military leaves taken by pilots at the Airlines—whether two days or two years—must be grouped together when assessing the most significant comparability factor: duration. 20 C.F.R. § 1002.150(b).  CLARKSON V. ALASKA AIRLINES, INC. 15 Although we have not yet addressed whether military leave should be considered based on its length or categorically when assessing USERRA violations, we hold that examining the length of leave at issue is the correct approach. Cf. Paige v. California, 291 F.3d 1141, 1147 (9th Cir. 2002) (determining the proper comparator groups is a question of law). The plaintiff, as master of the complaint, can limit the request for recovery of benefits to specific, shorter military leaves. To follow the district court’s approach and consider military leaves categorically would render USERRA’s protections meaningless. Military leaves vary greatly in length, and the longest leaves can last years. Were we to adopt a categorical approach to military leaves, no other type of leave would look similar, and servicemembers would not be protected under § 4316(b)(1). Indeed, the statute’s implementing regulation suggests that the military leave at issue should be compared with the alleged comparable leaves. Section 1002.150 states that duration is the most significant factor to compare when determining if “any two types of leave are comparable.” In context, “any two types of leave” must refer to (1) military leave and (2) another employer-offered leave. As the regulation explains, “a two-day funeral leave will not be ‘comparable’ to an extended leave for service.” 20 C.F.R. 1002.150(b). But while an “extended” military leave is not comparable to a “two-day funeral leave,” it is entirely possible that a two-day military leave is comparable to a two-day funeral leave. Although our sister circuits have not addressed this question directly, their opinions support the conclusion that a plaintiff may define his claim by the particular length of the military leave at issue. As the Seventh Circuit explained: “It is up to the employee to demonstrate that any given 16 CLARKSON V. ALASKA AIRLINES, INC. stretch of military leave is comparable to a form of nonmilitary leave that is accorded a benefit.” White v. United Airlines, Inc., 987 F.3d 616, 624 (7th Cir. 2021) (emphasis added). 7 We thus conclude that the district court erred when it compared non-military leaves offered by the Airlines to all military leaves taken by pilots at the Airlines. Clarkson limited his claim to military leaves of thirty days or less. Thus, the relevant question is whether such short-term leaves are comparable to the other leaves offered by the Airlines. As discussed next, this is a question for the jury. B. Whether Short-Term Military Leave Is Comparable to Other Types of Leaves Is a Jury Question “Comparability is a question of fact.” Syufy Enters. v. Am. Multicinema, Inc., 793 F.2d 990, 1003 (9th Cir. 1986). It is thus a question for the jury unless “the facts of a case suggest that no reasonable jury could see enough commonality for a meaningful comparison.” Howell v. 7 The district court acknowledged White but stated it “is not inclined to apply the Seventh Circuit’s White holding to this case, as the Ninth Circuit has not yet spoken on the issue.” Clarkson, 2021 WL 2080199 at *5. The Fifth Circuit has also denied summary judgment in a similar § 4316(b)(1) case, noting that there “are genuinely disputable issues as to the material facts of whether involuntary non-military leaves, not generally for extended durations . . . are comparable to each plaintiff’s military leaves taken for service in the uniformed services.” Rogers, 392 F.3d at 771-72 (emphasis added); see also Tully v. Dep’t of Just., 481 F.3d 1367, 1368 (Fed. Cir. 2007) (holding that employee was not entitled to holiday pay because a two-and-a-half-year military leave was not comparable to other leaves); Waltermyer, 804 F.2d at 825 (holding that employee was entitled to holiday pay because a two-week military leave was comparable to other leaves).  CLARKSON V. ALASKA AIRLINES, INC. 17 Wexford Health Sources, Inc., 987 F.3d 647, 657 (7th Cir. 2021) (quoting Rozumalski v. W.F. Baird & Assocs., Ltd., 937 F.3d 919, 927 (7th Cir. 2019)) (holding that comparability in employment discrimination cases is a jury question); see also Wheeler v. Georgetown Univ. Hosp., 812 F.3d 1109, 1116 (D.C. Cir. 2016) (same); Graham v. Long Island R.R., 230 F.3d 34, 39 (2d Cir. 2000) (same). The district court considered the three factors outlined in the regulation—duration, purpose, and control—and found that no jury could find in Clarkson’s favor given “significant differences” between military leave and non-military leave as to each factor. The district court thus granted the Airlines summary judgment because it found that there “are no genuine issues of material fact as to whether military leave is comparable to other forms of leave covered by the CBAs; they are not comparable.” Clarkson, 2021 WL 2080199 at *9. This conclusion was erroneous. At the summary judgment stage, “the judge’s function is not . . . to weigh the evidence and determine the truth of the matter[,] but to determine whether there is a genuine issue for trial.” Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 249 (1986). As explained below, the record evidence presents genuine issues of material fact as to each factor that, when viewed in the light most favorable to Clarkson, would allow a reasonable jury to find in his favor. 1. Duration Duration is the “most significant factor” in the comparability analysis. 20 C.F.R. § 1002.150(b). After considering the evidence, the district court determined that “[g]iven the significant differences in duration and frequency, military leave is not comparable to jury duty, 18 CLARKSON V. ALASKA AIRLINES, INC. bereavement leave, and sick leave.” Clarkson, 2021 WL 2080199 at *6. This holding ignored significant factual disputes over the leave data presented by the Airlines. The district court’s analysis was also flawed by including frequency as an integral factor in the duration analysis. a. Disputed statistical evidence Although the parties agree on the underlying dataset regarding the duration of military and non-military leaves, their agreement ends there. Each side’s expert used that dataset to present different opinions on the comparability of the duration of military and non-military leave. When parties “offer conflicting inferences drawn from the [statistical] evidence . . . there is a genuine dispute of material fact . . . and [the] action must proceed to trial.” Paige, 291 F.3d at 1147. It is for the jury, not the court, to decide how to weigh these conflicting inferences. Tyson Foods, Inc. v. Bouaphakeo, 577 U.S. 442, 459 (2016) (“Once a district court finds evidence to be admissible, its persuasiveness is, in general, a matter for the jury.”); City of Pomona v. SQM N. Am. Corp., 750 F.3d 1036, 1049 (9th Cir. 2014) (“Where two credible experts disagree, it is the job of the fact finder, not the trial court, to determine which source is more credible and reliable.”). Instead of recognizing this factual dispute, the district court adopted the Airlines’ evidence and granted their motion for summary judgment. In its analysis of the parties’ arguments, the court failed to consider the evidence in the light most favorable to Clarkson. Instead, the court relied only on the Airlines’ statistical evidence. This is evident from the district court’s repeated references to the Airlines’ Reply Statement of Material Facts, including to several facts that Clarkson disputes or materially clarifies. Further, as  CLARKSON V. ALASKA AIRLINES, INC. 19 discussed below, the district court’s review of the statistical evidence was tainted by its decision to consider military leave categorically. To conclude that the leaves are incomparable, the district court compared the longest military leave (185 days at Alaska, excluding the 99th percentile, and 70 days at Horizon, excluding the 95th percentile) with the longest jury duty leave (6 days at Alaska and 5 days at Horizon), the longest bereavement leave (6 days at Alaska and 3 days at Horizon), and the longest sick leave (5 days at Alaska and 4 days at Horizon). 8 But Clarkson presented a different analysis of the data in his opposition to summary judgment. His analysis showed that Alaska pilots took an average of 3.10 days of short-term military leave, 2.94 days of jury duty leave, 2.77 days of bereavement leave, and 2.52 days of sick leave. 9 Horizon pilots took an average of 4.23 days of short- term military leave, 2.66 days of jury duty leave, 2.48 days of bereavement leave, and 2.17 days of sick leave. 10 The 8 The jury duty and bereavement leave data presented by the Airlines covers the period between October 2004 and December 2020 for Alaska pilots and between October 2008 and December 2020 for Horizon pilots. The military leave data described above is from those time periods. The sick leave data, however, only covers the period from September 2013 to December 2020 for Alaska pilots and from February 2010 to December 2020 for Horizon pilots. During those periods, the comparable longest military leave is 122 days for Alaska pilots (excluding the 99th percentile) and 58 days for Horizon pilots (excluding the 95th percentile). 9 Again, because the sick leave data covered a different time period, the comparable average short-term military leave from September 2013 to December 2020 is 2.50 days for Alaska pilots. 10 The comparable average short-term military leave from February 2010 to December 2020 is 4.23 days for Horizon pilots. 20 CLARKSON V. ALASKA AIRLINES, INC. mode and median days of short-term military leave appear even more similar to the mode and median days of non- military leaves. The district court did not mention nor discuss Clarkson’s analysis. Nor did the district court mention the Airlines’ expert’s finding that the longest short- term military leave, which is the appropriate comparator, was 6 days for Alaska pilots and 9 days for Horizon pilots (excluding the 90th percentiles). Using these statistical measures, a reasonable jury could find in Clarkson’s favor. As noted above, summary judgment is only appropriate “where viewing the evidence and the inferences which may be drawn therefrom in the light most favorable to the adverse party, the movant is clearly entitled to prevail as a matter of law.” Sandvik, 609 F.2d at 974 (quoting Smith, 604 F.2d at 641). Here, the evidence put forth by Clarkson—and by the Airlines—could allow a jury to infer that the duration of military leave is comparable to the duration of jury duty, bereavement, or sick leave. The district court erred in concluding otherwise. b. Frequency The district court’s decision is also flawed because it treated frequency as an integral part of the duration analysis. As the court put it, “frequency is useful in the duration analysis.” Clarkson, 2021 WL 2080199 at *5. While it is true that the implementing regulation leaves room for courts to consider factors besides duration, purpose, and control, see 20 C.F.R. § 1002.150(b) (“[O]ther factors such as [purpose and control] should also be considered.” (emphasis added)), the factors enumerated in the regulation should be weighed most heavily when considering whether two leaves are comparable. Cf. Sanders v. City of Newport, 657 F.3d 772, 780-81 (9th Cir. 2011) (interpreting federal labor statute  CLARKSON V. ALASKA AIRLINES, INC. 21 “in light of the text of the pertinent DOL regulations”). Section 1002.150(b) instructs that duration is the “most significant” factor in the comparability analysis, and it does not mention frequency. See 20 C.F.R. § 1002.150(b). Frequency is not encompassed within duration. Rather, the two terms convey distinct concepts. Duration is “the time during which something exists or lasts.” Merriam-Webster Dictionary, https://www.merriam- webster.com/dictionary/duration (2022). Frequency is “the number of repetitions of a periodic process in a unit of time.” Merriam-Webster Dictionary, https://www.merriam- webster.com/dictionary/frequency (2022). Including frequency in the duration analysis effectively undermines the purpose of USERRA. 11 Congress intended USERRA and its predecessor statutes to protect reservists during their “frequent absences from work” with the full understanding that those frequent absences could “cause considerable inconvenience to an employer.” Monroe, 452 U.S. at 565. Nevertheless, “Congress has provided . . . that employers may not rid themselves of such inconveniences and productivity losses by discharging or otherwise disadvantaging employee-reservists solely because of their military obligations.” 12 Id. 11 This error is especially significant given the “canon that provisions for benefits to members of the Armed Services are to be construed in the beneficiaries’ favor.” King v. St. Vincent’s Hosp., 502 U.S. 215, 220 n.9 (1991) (citing Fishgold v. Sullivan Drydock & Repair Corp., 328 U.S. 275, 285 (1946)). 12 Regulations implementing other provisions of USERRA also explicitly state that “the timing, frequency, and duration of [a] person’s training or service . . . shall not be a basis for denying protection.” 38 U.S.C. § 4312(h) (emphasis added). 22 CLARKSON V. ALASKA AIRLINES, INC. In justifying the use of frequency, the Airlines rely on the very rationale that Congress sought to prohibit: “The fact that military duty leave happens much more frequently . . . than, say, jury duty leave, means that providing paid military leave is substantially more costly and burdensome to operations than providing paid jury duty leave.” But Congress enacted USERRA to “prohibit discrimination against persons because of their service in the uniformed services,” 38 U.S.C. § 4301, regardless of whether those anti-discrimination protections increase employers’ costs and burdens. See S. Rep. No. 90-1477, at 1 (explaining the statute was intended to protect reservists who were discriminated against because of frequent obligations). The Airlines also attempt to justify the use of frequency by arguing that “frequency is highly relevant in this case due to the peculiar military leave patterns unique to pilots,” who allegedly take more frequent leaves than other military reservists. But USERRA is no less protective of pilots than of any other military reservist. The Airlines’ (and amicus’s) policy arguments that pilots will use their frequent military duties to intentionally create conflicts to garner double pay, or that airlines will eliminate benefits such as paid bereavement leave or paid jury duty leave to avoid paying for military leave, are not compelling. The Airlines are only required to provide equal treatment, not preferential treatment, to employees taking military leave. Thus, if the employer only provides three days of paid bereavement leave per year or only offers the difference in pay between the employee’s salary and the compensation for jury duty (as Horizon does), that is all the employer would be required to provide to the servicemember. We are thus unpersuaded by the Airlines’ arguments that a ruling in  CLARKSON V. ALASKA AIRLINES, INC. 23 Clarkson’s favor would drastically increase costs or convince employers to cease offering other paid leaves. 2. Purpose In considering the purpose factor, the district court found that “the evidence supports [the Airlines’] position that a significant purpose of military leave under the CBAs is to allow employees to pursue parallel careers.” Clarkson, 2021 WL 2080199 at *7. By accepting this position, the district court rejected Clarkson’s evidence demonstrating that the primary purpose of military leave is to perform a civic duty and public service. Clarkson’s argument is both intuitive and well-supported by the record. In his deposition, Clarkson explained that he joined the military “to serve.” Alaska representatives also repeatedly acknowledged in their depositions and answers to interrogatories that public service is a primary purpose of military leave. Nonetheless, the Airlines attempt to undermine Clarkson’s testimony by pointing to a single question and answer from Clarkson’s deposition: Q: Would you consider it accurate to describe military service for Horizon pilots as a second career? A: I don’t know if we could consider it a second career. I would say parallel career. The district court, however, may not judge credibility, weigh the evidence, or resolve factual disputes at summary judgment. See Anderson, 477 U.S. at 255. The court thus erred when it held that there are no genuine disputes about 24 CLARKSON V. ALASKA AIRLINES, INC. the purpose of military leave. 13 It is for the jury to determine how much weight to give to the evidence presented about the purpose of military leave. 3. Control Finally, the district court found that “pilots have a greater degree of control over their ability to take military leave and schedule around such leave” than employees taking non- military leaves. Clarkson, 2021 WL 2080199 at *8. In so finding, the district court again accepted the Airlines’ factual contentions and ignored Clarkson’s factual presentation. Given the record evidence, however, a reasonable jury could find that pilots do not have significantly more control over military duty than they do over other types of leave. At both Airlines, pilots bid on their monthly schedules in advance and must provide notice of known absences, including scheduled military duty, at the time of bidding. The scheduling systems are designed to schedule pilots to avoid these absences, but conflicts can still occur because bidding takes place by seniority. To avoid conflicts after the schedules are released, pilots can trade trips with other pilots on a voluntary basis (with some limitations) and pick up or drop trips (with approval). The parties agree about this basic process. They also agree that when a pilot is in the military, his or her military service is involuntary. The parties disagree, however, about the level of control that pilots have to schedule their military duty so that it does 13 The district court’s conclusion that the purpose of military leave is to promote pilots’ own individual interests in a “parallel career” makes little sense when military leave is protected by USERRA, which Congress enacted explicitly to encourage public service in the military. See 38 U.S.C. § 4301.  CLARKSON V. ALASKA AIRLINES, INC. 25 not conflict with their work schedules at the Airlines. The Airlines contend that pilots have “tremendous” flexibility to schedule their military service, citing the testimony of Ronald Limes, Alaska’s Base Chief Pilot and former Air Force reservist. 14 In his deposition, Mr. Limes explained that as a commander in the reserves, he had “very little” flexibility, but as a rank-and-file servicemember, he had “tremendous” flexibility to schedule squadron trainings on weekends. Clarkson maintains that pilots’ control over military duty, especially duties beyond those squadron trainings, is more limited. Pilots can work with their military schedulers to resolve conflicts, but the ability to reschedule military duty depends on many factors including the pilot’s training requirements, military needs, and the availability of opportunities to engage in specific kinds of training. Clarkson also argues that many of the leaves at issue in this case were unknown to pilots more than a few weeks in advance, rendering pilots’ control over the scheduling process irrelevant. In his deposition, Clarkson explained that predetermined monthly military schedules would often change with little notice. Only the drill weekends, which were set annually by the governor, were unlikely to change. Given this factual record, the district court erred in finding that “[t]o the extent there is any scheduling conflict, pilots can . . . work with the bidding system and other pilots to ensure their flight schedules for [the Airlines] 14 The Airlines also argue that the “most pointed evidence” comes from Captain James Meldrum, an Alaska pilot who refused to sign a declaration stating: “Even if a pilot volunteers for military duty it is the military, not the pilot, who dictates the time and duration of that leave.” However, in his deposition, Captain Meldrum explained it is only his “belief” that military pilots have control; he has never been in the military himself. 26 CLARKSON V. ALASKA AIRLINES, INC. accommodate their military leave schedules.” Clarkson, 2021 WL 2080199 at *8. The evidence shows that pilots can try to trade shifts with other pilots and that pilots can try to work with their military scheduler to rearrange their duty periods. 15 A jury could reasonably conclude that because of last minute assignments, pilots do not have enough control over their schedules to prevent conflicts. A jury could also 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. These comparability determinations are for the jury to make, not the court. 16 15 USERRA’s legislative history addresses employees’ ability to reschedule work, as reflected in the House Report on the bill that became § 4316(b)(1): [T]o the extent the employer policy or practice varies among various types of non-military leaves of absence, the most favorable treatment accorded any particular leave would also be accorded the military leave . . . . Thus, for example, an employer cannot require servicemembers to reschedule their work week because of a conflict with reserve or National Guard duty, unless all other employees who miss work are required to reschedule their work. H.R. Rep. 103-65, pt. 1, at 33-34 (1993) (citations omitted); see also Rogers, 392 F.3d at 767-68 (citing the House Report). 16 Clarkson also argues that because military leave is involuntary, the “control” factor must come out in his favor. Clarkson relies on Waltermyer, which explained that military leave is comparable to jury duty leave, bereavement leave, and sick leave because all such leaves signaled “lack of choice by the employees.” 804 F.3d at 825. But, in  CLARKSON V. ALASKA AIRLINES, INC. 27 *** As explained above, the district court improperly resolved factual disputes as to each factor in the comparability analysis in order to grant summary judgment. Comparability is fundamentally an issue for the jury, and where reasonable jurors could return a verdict in favor of the nonmoving party, factual disputes must be resolved by a jury. See, e.g., Reynaga, 847 F.3d at 685 (citations omitted). Here, Clarkson has presented persuasive evidence in support of his claim. We therefore reverse the grant of summary judgment and remand for further proceedings consistent with this opinion. C. Pay During Leave Finally, the Airlines argue that “pay during leave” is not a standalone benefit that they provide under their CBAs to any employee on leave, rendering Clarkson’s claim moot even if short-term military leave is comparable to another type of leave. Because the district court found that military leave was not comparable to any other leaves, it did not address this issue and “decline[d] to adopt a specific interpretation of the ‘rights and benefits’ definition.” Clarkson, 2021 WL 2080199 at *3. Although our sister circuits have addressed this issue, see Travers, 8 F.4th at 199; White, 987 F.3d at 619, we remand for the district court Waltermyer, the court concluded that: “Particularly important is the fact that the guardsmen have no individual voice in selecting the weeks they will be on active duty.” Id. Here, the Airlines have presented evidence to demonstrate that pilots may have some “voice in selecting” their schedules. The governing regulation also clearly states that the factor to consider is “the ability of the employee to choose when to take the leave,” 20 C.F.R. § 1002.150(b), not whether the leave is voluntary. 28 CLARKSON V. ALASKA AIRLINES, INC. to consider the “pay during leave” issue in the first instance. CONCLUSION For the foregoing reasons, we reverse the district court’s grant of summary judgment to the Airlines. On remand, the district court should consider the “pay during leave” issue in the first instance. REVERSED AND REMANDED.