Raymond et al v. Spirit AeroSystems Holdings, Inc. et al, No. 6:2016cv01282 - Document 1115 (D. Kan. 2023)

Court Description: MEMORANDUM AND ORDER granting 1035 Motion for Final Certification of a Collective Action; granting 1036 Motion to Exclude; granting in part and denying in part 1037 Motion to Exclude; granting 1040 Motion for Partial Summary Judgment; den ying as moot 1042 Motion to Exclude; granting 1043 Motion to Exclude; denying 1044 Motion to Strike; denying as moot 1046 Motion to Exclude; granting 1090 Motion to Strike. See Order for further details. Signed by District Judge John W. Broomes on 5/16/2023. (sz)

Download PDF

6:16-cv-01282-JWB Raymond et al v. SpiritCase AeroSystems Holdings, Inc. et alDocument 1115 Filed 05/16/23 Page 1 of 83 Doc. 1115 IN THE UNITED STATES DISTRICT COURT FOR THE DISTRICT OF KANSAS DONETTA RAYMOND, et al., Plaintiffs, v. Case No. 16-1282-JWB SPIRIT AEROSYSTEMS HOLDINGS, INC., and SPIRIT AEROSYSTEMS, INC., Defendant. MEMORANDUM AND ORDER This matter is before the court on Plaintiffs’ motion (Doc. 1035) for final certification of a collective action under the Age Discrimination in Employment Act (ADEA), and on numerous motions from Defendants, including a motion to strike or decertify the collective action claims (Doc. 1044), a motion for partial summary judgment on those claims (Doc. 1040), five Daubert motions to exclude testimony of witnesses (Docs. 1036, 1037, 1042, 1043, 1046), and a motion to strike testimony (Doc. 1090). The motions are fully briefed and are ripe for decision. For the reasons stated herein, the court grants Plaintiffs’ motion for certification of a collective action (Doc. 1035), denies Defendants’ motion to decertify (Doc. 1045), rules on the Daubert motions as indicated herein, and grants Defendants’ motion for partial summary judgment (Doc. 1040) on Plaintiffs’ collective ADEA claims. Dockets.Justia.com Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 2 of 83 I. Background The case arises from a 2013 reduction-in-force (“RIF”) at Spirit Aerosystems, Inc. (“Spirit”). As part of a company-wide effort to reduce costs, Spirit laid off 271 employees from its manufacturing facility in Wichita, Kansas, where it employed more than 4,000 workers. The Society of Professional Engineering Employees in Aerospace (“SPEEA”) represents two separate bargaining units – the Wichita Engineering Unit (WEU) and Wichita Technical and Professional Unit (WTPU) – both with their own collective bargaining agreement (“CBA”) with Spirit. Plaintiffs were Spirit employees in Wichita represented by SPEEA who lost their jobs in the layoffs. After the layoffs, some Plaintiffs tried to get rehired into positions at Spirit but were not rehired. Plaintiffs allege that in 2013, new CEO Larry Lawson began focusing on cutting labor costs and that Spirit “falsely claimed … that past tolerance of poor performance required tightening of performance standards.” (Doc. 1034 at 7.) Lawson decided to implement Spirit’s first-ever mass layoff and Spirit allegedly sought to eliminate “more expensive (older, sicker, and those who took leave) workers.” (Id. at 7.) To achieve this, Spirit allegedly changed its annual performance evaluation [“PM”] process to “focus on older and sicker workers to protect the younger workforce from the layoff” by mandating a 15-70-15 distribution in performance evaluations and by using a “calibration” process to “tighten” the standards. (Id. at 7-8.) Then, prior to the layoffs, Spirit conducted a retention ranking exercise as allowed under the CBAs, under which Spirit ranked SPEEA represented employees in one of three categories: A (the highest, most likely to be retained), B, or C (the lowest, least likely to be retained), pursuant to a 70-20-10 distribution mandated by the CBAs. Plaintiffs contend Spirit “manipulated” the CBAs by depriving longserving employees of a “bump” in retention ranking and by not requiring new employees to be 2 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 3 of 83 ranked. Plaintiffs also note that Spirit changed to self-funded health insurance in 2013 and contends the company targeted older workers with higher health care claim costs in the layoff. (Id. at 8-9.) Plaintiffs also claim Spirit discriminated against them in rehiring based on age. After the layoff, Spirit held various job fairs to hire for a number of positions. Plaintiffs claim Spirit “excluded virtually all Plaintiffs from being rehired” by placing previously laid-off worker applications in a pool that precluded the applicants from being considered. Plaintiffs assert that their statistical expert found, among other things, that “applicants 40 and over had only 76% the odds of a successful screening as under 40 applicants.” (Id. at 12.) Various subgroups of Plaintiffs assert claims under three statutes: the ADEA, 29 U.S.C. § 621; the Americans with Disabilities Act (ADA), 42 U.S.C. § 12112; and/or the Family Medical Leave Act (FMLA), 29 U.S.C. § 2615. The three main statutory claims are further divided into “termination” and “failure-to-rehire” claims, with some Plaintiffs asserting only terminationrelated claims, some asserting only failure to rehire claims, and some asserting both. Additionally, the claims are divided into disparate treatment and/or disparate impact theories. Plaintiffs assert the ADEA claims both collectively and individually. With respect to the collective ADEA claims, Plaintiffs allege a “pattern or practice” of age discrimination on Defendants’ part. A few Plaintiffs also assert individual retaliation claims. In the pretrial order, Plaintiffs “assume that this case will be tried initially as a collective action, with initial consideration of Plaintiffs’ pattern or practice and disparate impact collective action claims, pursuant to the requirements set forth in Int’l Bhd. Of Teamsters v. United States, 431 U.S. 324, 334-35 (1977).” (Doc. 1034 at 1.) In the “event of decertification or dismissal of any of the collective action claims, Plaintiffs request the Court’s permission to amend the pretrial 3 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 4 of 83 order.” (Id. at 2.) For their part, Defendants “dispute that this case may be tried as a pattern-orpractice case” and say any request to amend the pretrial order can be addressed after the court rules on the current motions.1 (Id.) The result of all this is a somewhat dizzying array of claims and parties, set forth by Plaintiffs in a table in the pretrial order, and summarized by them as follows: 1. Collective Termination claim under the ADEA and Teamsters alleging disparate impact.2 (15 plaintiffs) 2. Collective Termination claim under the ADEA alleging disparate treatment pattern or practice (15 plaintiffs). 3. Collective Hiring claim under the ADEA alleging disparate treatment pattern or practice (51 plaintiffs). 4. Individual termination claims under the ADEA alleging disparate impact (15 plaintiffs). 5. Individual termination claims under the ADEA alleging disparate treatment (15 plaintiffs). 6. Individual hiring claims under the ADEA alleging disparate treatment (51 plaintiffs). 7. Individual termination claims under the ADA alleging discriminatory disparate treatment (3 plaintiffs). 8. Individual termination claims under the FMLA alleging retaliatory termination (2 plaintiffs). 9. Individual hiring claims under the ADA alleging discriminatory disparate treatment (11 plaintiffs). 1 If Plaintiffs do not prevail on the first stage of their pattern-or-practice claims, they are nevertheless entitled to proceed on their individual claims. Thiessen v. Gen. Elec. Cap. Corp., 267 F.3d 1095, 1106 n.8 (10th Cir. 2001) (citing Cooper v. Fed. Reserve Bank of Richmond, 467 U.S.867, 878 (1984) (“It could not be more plain that the rejection of a claim of classwide discrimination does not warrant the conclusion that no member of the class could have a valid individual claim.”). The individual claims proceed under the normal McDonnell Douglas framework, rather than benefitting from a presumption of discrimination. Id. 2 Defendants object to the inclusion of this claim in the pretrial order, asserting it has no counterpart in the First Amended Complaint (Doc. 522), and arguing Teamsters does not address disparate impact claims. (Doc. at 23 n.3.) The court will address the disparate impact issue on summary judgment. 4 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 5 of 83 10. Individual hiring claims under the FMLA alleging retaliatory failure to hire (5 plaintiffs) 11. Individual termination claims under the ADA alleging discriminatory disparate impact (3 plaintiffs). 12. Individual hiring claims under the ADA alleging discriminatory disparate impact (11 plaintiffs). (Doc. 1034 at 22.) Defendants respond with a table of their own listing thirty-one defenses and showing the Plaintiffs and claims as to which each defense is asserted. (Id. at 23-32.) There are a number of pending motions, including motions relating to certification of collective claims under the ADEA (Docs. 1035, 1044), Daubert motions to strike or limit expert testimony (Docs. 1036, 1037, 1042, 1043, 1046, 1090), and Spirit’s motion for partial summary judgment, which seeks judgment on the collective claims (Doc. 1040). The court sought input from the parties on the sequence in which the motions should be decided. That request produced no consensus. (Doc. 1034 at 42-43.) The court concludes the most efficient manner of addressing the motions is to first address the question of certification, and then to proceed to the summary judgment motion, addressing the Daubert motions within the context of summary judgment as necessary. The Supreme Court recognized a format for adjudicating “pattern or practice” discrimination claims in Int’l Bhd. of Teamsters v. United States, 431 U.S. 324 (1977). Under such a claim, the plaintiff has the “initial burden to demonstrate that unlawful discrimination has been a regular procedure or policy followed by an employer….” Id. at 360. “At the initial, ‘liability’ stage of a pattern-or-practice suit the [plaintiff] is not required to offer evidence that each person for whom it will ultimately seek relief was a victim of the employer’s discriminatory policy.” Id. Rather, its burden is to establish a prima facie case that such a policy existed, after which the burden shifts to the employer to demonstrate that the plaintiff’s “proof is either inaccurate or 5 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 6 of 83 insignificant.” Id. If the employer fails to do so, the court may conclude that a violation has occurred and determine an appropriate remedy. If individual relief for victims is sought, “a district court must usually conduct additional proceedings after the liability phase of the trial to determine the scope of individual relief.” Id. Proof of a pattern or practice of discrimination supports an inference during this “remedial stage” that any particular employment decision was made in furtherance of the discriminatory policy, with the plaintiff only needing to show that an individual suffered an adverse employment action, and with the burden then on the employer to show there were lawful reasons for the action. See id. at 362. In other words, if the plaintiffs prevail in the first stage of trial, they “reap a significant advantage for purposes of the second stage: they are entitled to a presumption that the employer … discriminated against them.” Apsley v. Boeing Co., 691 F.3d 1184, 1194 (10th Cir. 2012) (quoting Teamsters, 431 U.S. at 361). The Tenth Circuit provided a roadmap for addressing certification and summary judgment motions relating to ADEA pattern-or-practice claims in Thiessen v. Gen. Elex. Capital Corp., 267 F.3d 1095, 1108 (10th Cir. 2001). The court first approved an “ad hoc” approach to certification of collective claims under 29 U.S.C. § 216(b),3 with an initial, more lenient “notice stage” where certification is granted as long as there are “substantial allegations” that putative class members were victims of a single policy or plan, followed by a second, stricter stage where the court reviews several factors to determine if the plaintiffs are in fact similarly situated.4 In making that determination, a court must take into account the nature of pattern-or-practice claims and the twostage Teamsters procedure, where the initial focus is not on individual employment decisions “but 3 The ADEA enforcement provision, 29 U.S.C. § 626(b), incorporates remedies available under the Fair Labor Standards Act (FLSA), including a provision allowing an action to be maintained by employees on behalf of themselves and other employees similarly situated. See id; 29 U.S.C. 216(b). 4 At this stage, the court examines: “(1) disparate factual and employment settings of the individual plaintiffs; (2) the various defenses available to defendant which appear to be individual to each plaintiff; (3) fairness and procedural considerations; and (4) whether plaintiffs made the filings required by the ADEA before instituting suit.” Thiessen, 267 F.3d at 1103 (citation omitted.) 6 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 7 of 83 on a pattern of discriminatory decision[-]making.” Thiessen, 267 F.3d at 1106. In Thiessen the district court erred by failing to consider this structure. Although “a district court considering a motion to certify is entitled to look past the pleadings and examine the evidence produced during discovery, … when an ADEA plaintiff relies upon a ‘pattern or practice’ theory and comes forward with legitimate evidence to support that theory, the district court must take into account” the Teamsters framework. Id. at 1108. Thiessen found “little case authority discussing summary judgment motions in pattern-orpractice cases,” but saw “no reason why summary judgment cannot be aimed at both the first and second stage issues,” provided such motions are analyzed in light of the order of proof unique to pattern-or-practice cases. Id. at 1108-09. Thus, at the first stage a summary judgment motion “must focus solely on whether there is sufficient evidence demonstrating that defendants had in place a pattern or practice of discrimination during the relevant limitations period.” Id. Prior to resolution of that issue, it might be improper for a court to even consider summary judgment pertaining to second stage (individual relief) issues, but if it did so the Teamsters framework would require the court to employ a presumption that the employment decisions were discriminatory. Id. at 1109. After reviewing the arguments and relevant case law, the court concludes it is appropriate to first address whether the ADEA claims are properly asserted collectively under 29 U.S.C. § 216. Accordingly, the court will first address Plaintiffs’ motion to certify (Doc. 1035) and Spirit’s motion to strike and decertify (Doc. 1044.) After determining whether collective claims are proper, the court will address Spirit’s motion for partial summary judgment, as well as any Daubert or other motions relating to expert testimony that may be relevant to summary judgment issues. 7 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 8 of 83 II. Certification of Collective Action. Plaintiffs move for final certification of their ADEA collective action. Specifically, they seek certification of a collective of 15 Plaintiffs asserting age discrimination in terminations, under both disparate treatment and disparate impact theories, and a collective of 51 Plaintiffs asserting age discrimination in hiring under a disparate treatment theory. (Doc. 1035.) Defendants move to strike the collective termination claims and to decertify the previously certified (conditionally) hiring collective. (Doc. 1044.) A. Standards. “Class actions under the ADEA are authorized by 29 U.S.C. § 626(b), which expressly borrows the opt-in class action mechanism of the Fair Labor Standards Act of 1938, 29 U.S.C. § 216(b) (1994). Section 216(b) provides for a class action where the complaining employees are ‘similarly situated.’” Thiessen v. Gen. Elec. Cap. Corp., 267 F.3d 1095, 1102 (10th Cir. 2001). Thiessen endorsed an ad hoc approach to certification, under which the court typically makes an initial “notice stage” determination of whether Plaintiffs are similarly situated, requiring only substantial allegations that the class members were victims of a “single decision, policy or plan.” Id. at 1102.5 After discovery, the court makes a second determination, using a stricter standard, and reviewing factors including the following: “(1) disparate factual and employment settings of the individual plaintiffs; (2) the various defenses available to defendant which appear to be individual to each plaintiff; (3) fairness and procedural considerations; and (4) whether plaintiffs made the filings required by the ADEA before instituting suit.” Id. at 1103 (citation omitted.) 5 Under the foregoing standards, the court previously approved the parties’ stipulation to conditionally certify a collective action on Plaintiffs’ failure to hire claims. (Doc. 404.) The conditional certification allowed notice to be sent to potential class members and provided for an opt-in period. (Id.) Spirit reserved its right to later seek decertification of the collective failure to hire claim. (Doc. 397 at 2.) Plaintiffs did not previously seek conditional certification of their termination claims, 8 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 9 of 83 B. Analysis Plaintiffs claim Spirit was concerned about its aging workforce and health care costs associated with older workers in Wichita, prompting the company to use a financial loss at its Tulsa facility as “a pretext for the age-based July RIF.” (Doc. 1038 at 3.) They claim Spirit “rigged its performance management and retention processes to target older workers” in the RIF. (Id. at 4.) According to Plaintiffs’ economist, “age disparity in termination beg[an] with disparately low performance ratings, [was] magnified by disparities in retention rankings, then flow[ed] through to termination outcomes,” such that Spirit “terminated a disproportionately large share of Older Employees in the July RIF.” (Id. at 8.) Plaintiffs also claim Spirit “refused to rehire the mostly older workers terminated in July 2013” for hundreds of jobs available after the RIF due to Spirit’s treatment of their applications. Because of a policy applied to the applications of all former employees, including Plaintiffs, their applications were routed to an “AP-Hold” status where they were routinely denied review. Plaintiffs’ statistician concluded that Plaintiffs were “disadvantaged by being stranded in the Applicant Hold Pool” resulting in significantly reduced chances of success in the hiring process. (Id. at 9.) Plaintiffs argue that although they held different jobs, “they all: (1) share the same critical characteristics: older (40+) salaried SPEEA workers; and (2) suffered from the same centrallyplanned and -executed policies and practices: effort to reduce workforce average age and payroll and healthcare costs through (a) changing Human Resources (HR) processes (PM and retention exercise); (b) eliminating protections for long-tenured workers, such as ‘bumping’ and recall rights; … (c) July RIF, and [d] [Spirit’s] years-long refusal to rehire for hundreds of open positions.” (Doc. 1038 at 12.) 9 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 10 of 83 The court agrees with Plaintiffs that they have sufficiently identified and cited evidence that they are similarly situated with respect to common company policies and processes at Spirit, processes which they claim were part of a Spirit plan to discriminate on account of age. Plaintiffs were all members of SPEEA and their employment was governed by one of two CBAs; they all underwent the same PM rating process; they were all ranked under the same retention rank process called for by the CBAs; and they were all selected for termination as part of the same RIF. The 51 Plaintiffs who allege Spirit unlawfully failed to hire them (collectively ”Hiring Plaintiffs”) were affected by a common company procedure of having the applications of former employees placed on AP-Hold status where they were unlikely to be considered. Plaintiffs claim all of the foregoing processes were tainted by age discrimination. They have cited some evidence in support of their allegations. In deciding these motions regarding certification and decertification, the court makes no determination as to whether Plaintiffs’ evidence is sufficient to support a claim of age discrimination under a pattern and practice theory; that is the subject of Defendants’ partial summary judgment motion and will be addressed in that motion. The point here is that Plaintiffs have alleged claims based on company-wide policies and cite evidence that they were subjected to those policies as part of a common group. Cf. Thiessen, 267 F.3d at 1107 (citing Adamson v. Bowen, 855 F.2d 668, 676 (10th Cir.1988) (holding that district court, in making class certification decision, should avoid focusing on merits of underlying class action); Anderson v. City of Albuquerque, 690 F.2d 796, 799 (10th Cir.1982) (reversing and remanding denial of class certification where district court indicated its belief that plaintiff could not prevail on individual claims). Such allegations, including the assertions that these policies were adopted with a discriminatory purpose or effect, present common factual questions shared by these Plaintiffs, such 10 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 11 of 83 that they are fairly considered to be “similarly situated” for purposes of asserting age discrimination claims under a pattern and practice theory. Defendants challenge the collective claims, contending Plaintiffs’ employment situations are “vastly different,” that Spirit’s defenses require individualized evidence, that procedural fairness weighs against conducting “15 mini-trials” twice (collectively and individually), and that some Plaintiffs failed to exhaust their claims. (Doc. 1045.) The court ultimately finds these assertions are outweighed by evidence that Plaintiffs were subjected to common policies and assert claims, essentially as a group, and by the common claim that these policies had a discriminatory purpose or effect. Cf. Thiessen, 267 F.3d at 1105 (“In particular, plaintiffs alleged that defendants, by way of the MDP and the blocker policy, had a company-wide policy of age discrimination.”) Defendants are undoubtedly correct that there is some individualized evidence relating to Spirit’s defenses, but as noted in Thiessen, the pattern or practice trial format contemplates treatment of such matters in a second stage of trial after a pattern or practice of discrimination has been established. See id. at 1107 (“Although it is true that defendants asserted ‘highly individualized’ defenses to each of the instances of individual discrimination asserted by plaintiffs, those defenses would not become the focal point until the second stage of trial and could be dealt with in a series of individual trials, if necessary.”) As for Defendants’ arguments that some Plaintiffs did not file or exhaust EEOC charges, the court has previously addressed the “piggyback” rule that allows workers asserting common claims to rely upon the timely filed charges of others. Under the circumstances, Defendants have not shown that the failure of any individual Plaintiff to file an EEOC charge merits denial of certification. Cf. Thiessen, 267 F.3d at 1110 (“The policy behind the single filing rule is that ‘[i]t would be wasteful, if not vain, for numerous employees, all with the same grievance, to have to process many identical complaints with the EEOC.’. . . As long as 11 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 12 of 83 the EEOC and the company are aware of the nature and scope of the allegations, the purposes behind the filing requirement are satisfied and no injustice or contravention of congressional intent occurs by allowing piggybacking.”) Plaintiffs’ motion for final certification (Doc. 1035) is GRANTED; Defendants’ motion to strike and decertify (Doc. 1044) is DENIED. The court finds that Plaintiffs are similarly situated within the meaning of § 216(b) on their collective claims. III. Spirit’s Motion for Partial Summary Judgment (Doc. 1040). Spirit’s motion for partial summary judgment seeks judgment in its favor on the three collective age discrimination claims asserted by Plaintiffs under the ADEA: (1) the pattern-orpractice disparate treatment termination claim; (2) the pattern-or-practice disparate impact termination claim; and (3) the pattern-or-practice disparate treatment failure-to-hire claim. (Doc. 1040.) A. Uncontroverted Facts. The court finds the following facts to be uncontroverted for purposes of summary judgment. The following statement excludes any facts asserted by the parties that are not properly supported by the materials cited in support of them. 6 Spirit is an aerospace manufacturer. Most of its domestic employees work at its Wichita, Kansas, headquarters, including (as of July 2013) 4,122 salaried and management employees. SPEEA represents two bargaining units of salaried Spirit employees in Wichita (“SPEEA employees”), each with its own CBA. The first unit is the WTPU and the second is the WEU. SPEEA job codes are divided into levels ranging from Level 1 or A (entry level), to Level 4 or D 6 Plaintiffs’ memorandum in opposition (Doc. 1070) purports to dispute a number of Defendants’ asserted facts without actually addressing the substance of the asserted facts. In those instances, the court has considered the asserted fact as undisputed for purposes of the motion provided Defendants have properly supported it with a citation to the record. See Fed. R. Civ. P. 56(e). 12 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 13 of 83 (expert/specialist) and Level 5 or 6 (consultant/senior consultant). Higher level employees have higher pay scales and are expected to have more and broader experience and skills than lowerlevel employees in the same job code. (Doc. 1041 at 6-7.) On July 25, 2013, Spirit announced separate layoffs of overhead (management and salaried) workers at three domestic sites, including Wichita. The Wichita layoff impacted 271 employees, including 50 non-SPEEA employees and 221 SPEEA employees, effective August 8, 2013. (Id.) Of the 221 laid-off SPEEA employees, 15 now allege unlawful termination (the “Termination Plaintiffs”) and 51 allege unlawful failure to hire (the “Hiring Plaintiffs”) in this case. Seven Termination Plaintiffs filed EEOC charges for discriminatory termination, and 16 Hiring Plaintiffs filed EEOC charges of hiring discrimination. (Id. at 8.) Spirit offered a severance package and paid nearly $6 million to 260 laid-off Wichita employees who acted on its offer in exchange for a release of all existing claims against Spirit, except Termination Plaintiffs (minus Bucchin, Doyon, Hatcher, Jackson, and Raymond) and most Hiring Plaintiffs. Spirit Executive and Human Resources (HR) Communications Manager Deborah Miller testified that in 2012, Spirit Chief Operating Officer (COO) David Walker encouraged managers to promote and guide “our younger people that were coming on, especially new managers” into leadership positions, stating he thought Spirit had a tradition of old Boeing DNA and mediocrity, and “the younger people were the future of the company” and were the key to rising above mediocrity. (Doc. 1070-23 at 5.) Spirit’s data as of June 2012 showed that the average age of its workforce company-wide was 45.9 in 2007 and that it had increased in 2011 to 46.4. (Doc. 1080-4 at 1.) As of 2012, the 13 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 14 of 83 average age of workers for the year 2017 was projected to be 45.6. While gathering age data in June 2013, HR Vice President (VP) Cassie Caster expressed surprise about the proportion of older workers in the Spirit workforce, writing in an email: “OMG – 76% of our domestic employees are age 40 and over? Am I reading this right? … 76%??” Carter testified she was surprised “at what that would do to our skill base if in the next few years people would decide … to leave. How would we get the work done?” She said, “I didn’t know what the number was, so I didn’t know what to expect, but two thirds being [near] the window when they could retire is a troublesome thing in workforce planning.” Carter denied, however, that Spirit ever had a goal, strategy, or initiative to lower the average age of the workforce. (Doc. 1041-29 at 45-46.) Spirit HR prepared periodic workforce planning spreadsheets for director-level or higher managers. These tools included only aggregate data, such as current and projected overall headcount, average age, or retirement risk within an organization. They did not identify individual employees and were not used to make individual employment decisions. First-level managers generally did not have access to workforce planning documents with demographic data such as age. They instead focused on budgeted headcounts (by number of heads, not labor dollars.) ADP, a vendor hired by Spirit to aid in hiring, likewise had no access to workforce planning documents, average age, or retirement risks. Spirit took a one-time October 2012 $590 million write-down, which was attributable in significant part to its Tulsa operations. Based on the financial difficulties the company was experiencing at the time, Spirit’s leadership committed to its Board of Directors that it would cut a minimum of $150 million from its operations. It focused its cost-cutting primarily on its Wichita operations, where it had by far the most employees and the oldest average workforce. (Doc. 10805 at 3.) The average age of the Wichita and Oklahoma workforce, respectively, was 47.5 and 45.3. 14 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 15 of 83 The Wichita workforce had a much longer average length of service and larger average salary than the Oklahoma workforce. (Id.) In October 2012, Spirit’s HR VP unveiled a “Cost Down Initiative” that discussed aggressively managing overhead and salaried workforce, including by “replac[ing] long service/high costs employees with new hires by offering a Voluntary Separation Package.” (Doc. 1080-7 at 3.) It noted that the current budget for personnel based on projected work equated to a headcount (of total employees) of 15,665, but the company’s current headcount was actually 16,301. (Id. at 4.) The initiative did not propose layoffs. Healthcare Costs In 2006, Spirit had launched “Healthy Spirit,” a program that encouraged workers to assess their health risks. In 2012, using data from that program, Spirit’s benefit consultant, Mercer, analyzed health costs and identified rising costs, including in the Wichita workforce. Mercer found that every increase in current age of employees adds approximately 4% in medical costs; predicted a number of high-cost claimants among SPEEA members; specified that high-cost claimants and age demographics were some of the factors that increase healthcare spending; and flagged “high utilization” and “large claims,” among other things, as self-funding risks. It noted that Spirit’s “population continues to age.” (Doc. 1070 at 5-6; Doc. 1081-4 at 7.) It identified five ways to keep healthcare costs down: reduce members; reduce services; manage vendors; better decisions by employees on plan usage; and improve/maintain health. (Doc. 1080-18 at 18.) After exploring cost-saving measures relating to employee benefits, Spirit elected in January 2013 to transition to self-funded medical plans, effective July 1, 2013. The change was projected to result in savings in 2013 of up to $14.8 million. Spirit had the option of purchasing stop-loss insurance, which “[p]rovides risk protection against catastrophic claims for self-funded 15 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 16 of 83 plans.” (Doc. 1059-25 at 2.) Spirit was informed that 60% of organizations its size purchased stoploss insurance while 40% did not. (According to one of Plaintiffs’ experts, as of 2013 87% of employers of a similar size had self-funded insurance and only about half (52%) purchased stoploss insurance.) A Mercer analysis indicated to Spirit that over the five-year preceding period, Spirit’s stop-loss premiums would have exceeded its claim costs, although in the period 2011-12 there was an increased number of high claimants, and stop-loss insurance would have resulted in reimbursements in excess of premium estimates. Spirit elected not to purchase stop-loss insurance, meaning it would be responsible for all healthcare claim costs of its covered workers. Plaintiffs contend they or their family members posed a high risk of incurring large medical costs. Their managers, however, would have had little or no knowledge of their particular medical conditions or costs, although they may have been aware that employees took leave. In June and July 2013, many managers did not understand that Spirit was changing to self-funded health insurance or the implications of that change. There is no evidence that managers who rated Plaintiffs in the 2012 PM ratings or 2013 retention exercise had any knowledge of Plaintiffs’ medical costs or that such knowledge played any part in the ratings given to Plaintiffs. Of the 221 laid-off SPEEA employees, at least 21 (including some Plaintiffs) were not directly enrolled in a Spirit medical plan, so their layoff could not have saved Spirit any medical costs. Another 50 of the 221 had employee-only coverage. Of Plaintiffs, some had varying types of coverage (employee only, employee and spouse, family), and two continued on Spirit insurance after their layoff as spouse-dependents. 16 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 17 of 83 2012 Performance Management (PM) Ratings In October 2012, Spirit announced a $134 million net income loss for the quarter, and it ended 2012 with net income down 82 percent. In response, Spirit worked to reduce labor (as well as non-labor) costs. Spirit had never conducted a management and salaried layoff in Wichita during then-CEO Jeff Turner’s tenure. Turner viewed layoffs as a last resort. To avoid layoffs, Spirit initially planned to save over $31 million in payroll by eliminating the lowest-performing ten percent of its workforce. Spirit anticipated it would use its annual Performance Management (PM) reviews, the process it used to evaluate salaried employees, to identify the lowest performers. Under the PM process, managers assign employees an overall PM rating on the following scale: Exceptional; Exceeds Expectations; Meets Expectations; Meets Some Expectations; and Unacceptable. (The three middle levels are sometimes shortened to Exceeds, Meets, and Meets Some.) In the past, Spirit had experienced “grade inflation” in PM ratings, with managers assigning consistently high ratings. In 2010, for example, 58% of employees were deemed Exceptional or Exceeds. For the 2012 PM reviews (which occurred in the first quarter of 2013), Spirit told managers their ratings needed to follow a “bell curve” distribution, with 15% of their employees rated high, 70% in the middle, and 15% at the low end. New employee hires (those who had been hired within 90 days) were exempted from the 2012 PM review. Spirit HR exercised greater control over the process than it had in the past in order to implement the new bell curve policy. Spirit HR instructed managers to give employees realistic, honest, and consistent ratings, and to rate on performance, and not on unlawful factors, including age, or on immaterial factors, including seniority. HR pressured managers to rate based on performance and indicated a need to identify poor performers. (Doc. 1080-8 at 4-8.) For example, one exercise asked managers to identify examples of employee 17 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 18 of 83 performance, including whether managers ever noticed opportunities for improvement. (Doc. 1070-23 at 7.) One manager, who said she was told the company needed to “get lean and mean” and had “dead weight” employees “just filling out their time,” testified she felt like the people she was being asked to rate lower “tended to be older, tended to have health problems.” (Id. at 24.) Spirit essentially required managers to adjust the rating scale so that a group of low performers would be identified. The process included “calibration” sessions organized by HR in which managers from across the organization met to review initial ratings, provide input, and foster consistency in ratings. During calibration sessions, HR kept a spreadsheet of employees to ensure that each employee was discussed. Each manager discussed his or her employees, declared how they would rate them, and others provided input. Managers from across the organization weighed in. If upperlevel managers disagreed, they hashed it out during the session, but no one could override a direct manager’s final rating. The ratings were ultimately made by individual managers, who had discretion within the foregoing bounds to rate their employees. Some managers did not like the new requirements or thought they were unfair; some were disciplined or fired when they resisted or refused to follow them. Managers’ own ratings were determined in part by how well they executed the PM process. The HR training sessions for managers included materials that identified four case studies of hypothetical employees. One involved an employee (“Sue”) who does what the manager wants, shows enthusiasm, and whose work is great; a second (“Gus”) has worked for 15 years with five different managers, whose work is not great, who had a negative attitude and no desire to learn anything new; a third (“Bill”) is a good employee but who didn’t complete an objective on time at a time when the employee was overwhelmed on other projects; and a fourth (“John”) who was a 18 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 19 of 83 new member of the team who was eager to learn, has taken initiative, and whose work is always high quality. The materials asked managers how they would rate each of these employees. The materials identified common mistakes in appraisal, including grade inflation, “nice guy effect” (rating on employee personality rather than performance), and “central tendency” (rating people in the middle because it is easier than identifying the top and bottom performers). (Doc. 1080-8 at 14-18.) The materials were part of a 29-page training packet for managers on the 2012 PM rating process that stressed the need to identify poor performers to improve Spirit’s overall operations. (Id. at 3-4.) HR permitted managers to communicate PM ratings to the employees only after final approval was received. Union employees had appeal rights concerning their ratings pursuant to the CBAs. For the 15 Termination Plaintiffs, their managers: (a) rated six of them “Meets Expectations” in 2012 (Doyon, Ensor, Hatcher, Jackson, Marks, Richardson) and the rest “Meets Some Expectations”; (b) rated four the same in 2012 as in 2010 (Ensor, Hatcher, Jackson, Richardson); and (c) did not rate one (Longan). Of the ten who had lower ratings in 2012 than in 2010, six (Bucchin, Marks, Meadows, Schmidt, Sha, West) had a new manager. In comparing 2012 PM ratings to 2010 PM ratings, Plaintiffs’ expert statistician reported that 62% of older workers (1006/1620) had their ratings reduced, compared to 49% of younger workers (168/342). (Doc. 1046-2 at 14.) No Plaintiff filed an EEOC charge related to this lawsuit until March 2014, more than 300 days after managers reported the 2012 PM ratings in April 2013. After the 2012 PM process, HR reviewed all low-rated employee files to determine whether to terminate the employee, to issue a performance coaching plan (PCP), or to issue an individual 19 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 20 of 83 development plan (IDP). Spirit identified only 50 underperformers for immediate termination, a fraction of what it was seeking, as most low-rated performers lacked prior documentation justifying dismissal. HR pushed managers to pursue “more exits and [PCP] plans.” (Doc. 1070 at 10.) Spirit expected to turn “99% of [improvement plans]” into “exits,” but 89% of employees were actually on track to meet their plans, which was “not in line with BOD [Board of Directors] expectations.” Thereafter, new CEO Larry Lawson, who favored cost reduction via layoffs, approved a RIF of 10% of the Wichita workforce. 2013 Pre-Layoff Retention Exercise CEO Jeff Turner retired in April 2013 just before Spirit hired Lawson as CEO. In his first month, Lawson learned that Spirit still faced $4.3 billion in financial losses and potential risks to “mitigate if [it] wanted to stay viable as a company.” (Doc. 1041 at 7.) The SPEEA CBAs govern the process for laying off SPEEA employees. They call for Spirit to periodically rank SPEEA employees through a retention exercise and they allow Spirit to conduct a new retention exercise prior to conducting a layoff. Per the CBAs, employees are ranked within retention groups (by job classification and exempt/non-exempt status [the WTPU] and by job classification and skill management code [the WEU]). All active employees within a retention group, regardless of level, are ranked together. Managers over employees within one retention group, even if they work in different organizations, meet together to rank the employees in that group. Per the CBAs, the top 70% are rated A, the next 20% rated B, and the lowest 10% rated C, with Cs being most vulnerable to layoff. (Doc. 1041 at 9.) In Article 7 of the WTPU CBA, which addresses workforce issues, Section 7.5(c) entitled “Retention Rating Group Makeup” provides as follows: “Management will assign the retention rating by Job Classification for each employee to who[m] this Article applies, with the basic 20 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 21 of 83 objective of identifying those employees who, in the opinion of Management, are best able to maintain or improve the efficiency of the Company, further its progress and success and contribute to the successful accomplishment of the Company’s current and future business.” (Doc. 1041-101 at 6.) Section 7.5 on Reductions-in-Force provides: “Should reductions-in-force become necessary, the Company will retain employees with the best performance or as warranted by business need in each classification.” (Id. at 4.) The WEU CBA contains similar provisions. (Doc 1041-102 at 8.) Retention is different than performance (PM ratings), and retention ratings (or rankings) do not directly correspond to PM ratings. In retention, (1) the decisionmakers and comparators (managers, by retention group) are different than in PM (managers, by organizational unit; across job codes); (2) employees are forcibly ranked, meaning 10% must be C-rated even if an entire group is exceptional; and (3) 2012 performance is only one of three factors considered in retention ratings. The three factors considered in the pre-RIF retention ratings were: performance (both 2012 PM ratings and 2013 performance), versatility, and criticality. (Doc. 1041 at 9.) Versatility measures an employee’s breadth of skills and ability to adapt to and perform different types of tasks, i.e., “how many skills you had” and proficiency in those skills, a partially objective/partially subjective standard. Criticality measures the importance of an employee’s knowledge, skills, or role. This covers factors such as the skills Spirit needs now and into the future and how hard it would be to absorb an employee’s work if they were no longer with Spirit. Different managers consider different skills critical. (Id. at 10.) When first-level direct managers cannot agree who should receive a lower rating to meet the 70/20/10 retention distribution, they sometimes solicit input from second-level managers and, 21 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 22 of 83 ultimately, use seniority as a tie-breaker, with the least-senior employee receiving the lower rating. (Id.) A SPEEA employee hired after the most recent retention exercise – that is, one “who enters the bargaining unit between retention rating reviews” – automatically receives a C rating until the next exercise. In other words, employees hired since the 2010 exercise were C-rated by default until the 2013 exercise. As of July 2013, only SPEEA employees hired after the 2013 exercise cut-off date were C-rated by default. (Id.) Under the CBAs, employees assigned an “unadjusted” C rating by their manager and who have at least 20 years of service receive a contractual seniority preference that bumps their rating to an “adjusted” B and gives them recall rights for their particular job. But the CBAs allow Spirit, in its sole discretion, to “designate” C-rated employees who will not receive that preference. (Doc. 1041 at 10-11.) Spirit conducted a retention exercise in June 2013, based on the SPEEA workforce as of May 20, 2013. Unlike prior exercises, Spirit “designated” nearly all C-rated employees so that the layoffs would in fact impact the lowest-rated 10% of each group and Spirit could retain its higherrated workers without regard to seniority. Employees could appeal C ratings and designations, although the time until the RIF was conducted made prevailing on appeal before the RIF was conducted impractical. Spirit cites evidence that a successful appellant after the RIF could have been reinstated with backpay. (Doc. 1084-2 at 3.) The designations had no impact on the ratings of most Cs, as 59% of the Cs (176/298) (including 126/221 layoffs) had less than 20 years of service. (Doc. 1041 at 11.) The retention exercise included procedural “guardrails,” including manager training, HR level monitoring of rating discussion and decisions, an appeals process guaranteed by the CBAs, 22 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 23 of 83 feedback from prior managers (for employees in a new area), and discussions with other managers to gather additional feedback and foster consistency. The direct managers had a list of employees who needed to be rated in each retention group, including their seniority, but without age, date of birth, or other demographic data. Spirit HR “remind[ed] managers that they cannot use [“age, … disability, … FMLA, personal, medical, or family issues”] in making their decisions.” (Doc. 1041 at 11.) Managers generally did not know the reasons an employee had taken leave. Spirit HR trained managers to make rating decisions on the merits and not on improper factors (including age) or immaterial factors (including seniority). Training materials stated that employees “at the bottom tier of performance,” with “skill deficiencies,” or “with behavior/Value deficiencies” were to be designated as C employees. (Doc. 1080-38 at 6.) Managers were told it was not acceptable to “designate or fill Category C with new employees,” as the company had invested time and money to attract new talent. (Id.) Accordingly, seniority should be used as a tiebreaker rather than as the primary basis for retention, and new employees “should be evaluated based on how they are progressing, their contribution to date, and their potential going forward.” (Id.) Neither the CBAs nor Spirit dictated what rating managers should assign to recent hires during the exercise. The retention exercise was conducted in June 2013; employees who were hired after May 20 were exempted from the exercise, in part because they had no significant performance record. Within the bounds set by the CBAs and HR, more than 300 Spirit managers rated SPEEA employees. Within each group, direct managers assigned a rating for their employees, which HR aggregated to see if the 70/20/10 distribution was met. If it was not, those managers met to discuss each employee. HR facilitated these meetings but did not rate employees or recommend ratings. 23 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 24 of 83 No one could change a direct manager’s ratings without that manager’s approval. (Doc. 1041 at 12.) The 2012 PM ratings, alone, did not determine retention ratings: (a) not all Cs were from the two lowest PM categories (Meets Some; Unacceptable); 96/298 (32%) of Cs had a Meets or Exceeds Expectations rating; (b) not all employees in the two lowest PM categories were Cs; 146/331 (44%) were A/B rated; and (c) most employees with “lowered” 2012 PM ratings (vs. 2010) were not in the two lowest PM categories or C-rated; 1,025/1,204 (88%) were not Cs. (Doc. 1041 at 12-13.) Different managers had different methods and reasons for ranking employees, weighting the retention factors, and assigning retention ratings. Various managers said they ranked Plaintiffs for the following reasons: Slusser considered versatility the biggest factor because his group did not have critical skills. After discussions with other managers, Slusser rated Doyon C because, relative to her group, she was either less versatile (less experience and narrower skill set) or a lower performer and deserved a lower rating. Connolly ranked his reports by skills and performance and assigned C ratings to the lowest 10%. He ranked Denny at the bottom because, although his 2013 performance had improved, he was less experienced and versatile than others. Walton wanted to rate Hatcher, who had no performance issues but lacked initiative and was not proactive, B, but after multiple meetings and heated arguments, he relied on the seniority tie-breaker to rate her – the least senior B-rated employee – C. Holter wanted to rate Marks B but agreed Marks was the right person to rate C to meet the mandatory distribution because while skilled, he did not apply himself and was “was the least of the group that we had.” Kivett (who became West’s manager in 2013) rated West C because her output and skill set were below average and below her level and experience. West’s former manager, who claims West deserved a higher 2012 PM rating than her new manager gave her, said she also likely would have rated West a C. 24 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 25 of 83 Alley rated Bucchin C because, while knowledgeable, he was confrontational, which inhibited his performance. Springer likewise rated Ensor, a knowledgeable employee, C because of his attitude and treatment of other employees. Leeks rated Sha C because he struggled to convert his good ideas into actionable uses and was low performing and not particularly versatile. Wilson rated Raymond C because of ongoing performance issues and negative responses when asked to take on new tasks. Koppelman rated Richardson a C because of lower technical skills. Walker rated Koch7 C because, while an effective performer, his skill set was limited to simple, non-critical tasks. Sanchez considered Bradley, in isolation, a B, but relative to her group she was a C because she had less experience and ability to perform multiple functions. Fuller rated Kilgroe C based on his understanding that anyone with a Meets Some rating had to be a C, even though Fuller (who had assigned the PM rating based on a prior manager’s input) believed Kilgroe’s PM rating was inaccurate. Bell rated Fondaw C based on guidance that employees who had received performance coaching in 2013 should be C-rated absent an executive exception. Bell’s manager requested (but did not receive) an exception, but understood that this “was a borderline call” given Fondaw’s prior performance deficiencies. (Doc. 1041 at 13-14.) Spirit also ranked its non-union (i.e. non-SPEEA) workers on an A-C retention scale, with some similar factors considered, although non-union workers were separately rated based on their performance and “KSA” - knowledge, skills, and abilities. They were given numerical scores corresponding to A, B, or C bands, under a process that had been used by Spirit in Oklahoma, although the distribution of the bands was not specified. (Doc. 1080-43 at 2.) 7 William Koch is listed as a Plaintiff but no claim for him is included in the Pretrial Order. 25 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 26 of 83 HR reviewed retention ratings to ensure exclusion of overt bias in the ratings. Of the 15 Termination Plaintiffs: five were Cs in 2010; six with lower 2013 rankings had a new manager. Nine had less than 20 years of service, so being designated did not impact their retention rating, including two who, after appeal, were not designated. At least four were retention rated in 2013 based on roles they had moved into within the last 1.5 years or so. As illustrated in the following table, the Spirit managers rated employees older than the Termination Plaintiffs as an A or B in all ten of their retention groups and rated an employee under age 40 as a C in seven of ten. Plaintiff (Age) Union Retention Group Bucchin (56) WTPU Denny (53) WTPU Doyon (51) WTPU Hatcher (53) WTPU Longan (65) Heston (55) Marks (53) Jackson (50) WTPU MW Exempt KH Non-Exempt CD, Exempt LP, Exempt KJ Non-Exempt Meadows (58) West (50) WTPU SE, Non-Exempt Raymond (59) WTPU NQ Exempt KS Non-Exempt Richardson (58) WTPU Ensor (55) Schmidt (53) WTPU Sha (62) WEU JE, Non-Exempt JA, 63D # of Older Workers Rated A/B (age of oldest) (oldest employees A/B-rated) 26 older than Bucchin (70) (27 of 30 oldest) 47 older than Denny (68) (69 of 75 oldest) 6 older than Doyon (71) (38 of 40 oldest) 20 older than Hatcher (65) (24 of 25 oldest) 4 older than Longan (69) 109 older than Heston 149 older than Marks 169 older than Jackson (21 of 22 oldest) 19 older than Meadows (70) 45 older than West (17 of 19 oldest) 9 older than Raymond (72) (21 of 23 oldest) 11 older than Richardson (74) 14 older than Ensor (15 of 17 oldest) 7 older than Schmidt (64) (28 of 30 oldest) 1 older than Sha (67) (20 of 23 oldest) (Doc. 1041 at 15.) 26 Under-40 Workers Rated C? Yes Yes Yes Yes Yes No No Yes Yes No Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 27 of 83 Some managers were not happy with the retention rating process or results, but they agreed that the retention decisions were made for non-age business reasons. Fuller in particular did not like the ranking or retention process and thought the concept that the company should reduce a certain percentage of its employees every year “didn’t make sense.” (Doc. 1070-8 at 2.) At an HR training session on the 2013 retention exercise, Fuller was upset because he felt he was being told that “[e]verybody who’s a ‘met some’ or lower [PM rating] is going to be considered a Retention 4” (low retention level) and that managers were not supposed to tell employees why they were getting that level. (Doc. 1041-36 at 9.) Fuller admitted he was “challenging” at the training session and said, “So you want us to lie to the employees.” (Id.) Fuller was suspended for five days after Spirit indicated his behavior was “not representative of a manager performing his responsibilities to support company initiatives.”8 (Id. at 12.) Despite the designations (which eliminated a bump in retention rating from C to B for 20 plus year employees), lower-tenure (less than three years) and lower-level (Levels 1 and 2/B) employees were statistically overrepresented in the C-rated employees, and longer-tenure (20 or more years) and higher level (Levels 4/D and 5) employees were underrepresented: Population < 2 Years of Service < 3 Years of Service >20 Years of Service Level 1 and 2/B Level 4/D and 5 % of Rated Population 9.5% (274/2,889) 12.3% (356/2,889) 45.5% (1,314/2,889) % of C-Rated Pool 13.1% (39/298) 16.4% (49/298) 40.9% (122/298) % Over / Under 38% 33% -10% 25.4% (735/2,889) 34.3% (991/2,889) 39.3% (117/298) 17.8% (53/298) 55% -48% (Doc. 1041 at 16.) 8 Fuller testified at one point that Spirit “didn’t really want to lay off new hires,” indicating he may have attempted (apparently unsuccessfully) to give a new hire a C rating, although he could not recall details of the incident. (Doc. 170-8 at 3-4) (“I – I think I started to, yes, and I was probably told not to. But, honestly, I don’t remember the details of that.”) 27 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 28 of 83 The July 2013 Layoff In July 2013, Spirit leadership called for a layoff of 10-15% of overhead positions in Wichita, except in ten excluded job or skill management codes. Among the non-excluded SPEEA jobs Spirit managers considered for layoff, the managers did not lay off anyone in 49 retention groups. Managers of SPEEA employees had already identified a bottom 10 percent of each retention group in the retention exercise. They selected for layoff all C-rated employees in every retention group in which layoffs occurred, except: (a) Spirit decided that managers could not select New Hires for layoff (the stated purpose of which was to protect Spirit’s ties with recruitment sources and not wasting its investments in recruiting), which impacted only 7 SPEEA New Hires, 4 of whom were over age 50 (51, 56, 58, 63), and none in the same group as any Termination Plaintiff; and (b) Spirit made 5 individual exceptions to retain C-rated employees in groups that experienced layoffs, 4 of them over age 40 (57, 52, 49, 43) and 1 age 39. Although Plaintiffs initially alleged, based on rumor and speculation, that Spirit leadership created a list of workers for managers to target for lowered ratings and layoff because of their age or medical costs (a position now abandoned), neither HR nor upper-level managers gave individual managers a list of employees to assign a low rating or to layoff or otherwise instructed managers how to rate specific employees. Non-SPEEA employees were selected for layoff by different decisionmakers (their managers) through a different process than the SPEEA employees, based mostly (“90 percent or better”) on performance. (Doc. 1041 at 18.) The July 2013 RIF produced savings to Spirit of $5.7 million in 2013 and $37.8 million in 2014 in salary and health care costs. 28 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 29 of 83 Layoff Demographics By not laying off anyone in 49 retention groups, Spirit retained 4 C-rated older employees and only 1 younger C-rated employee. In ten other retention groups, Spirit selected only younger workers or the same number or fewer older workers as younger workers. Only 33% (6/18) of the layoffs in those groups were over age 40, even though 70% (116/166) of the employees were over age 40. Spirit laid off the less expensive Level 1 and 2/B employees (94/572; 16.4%) at about four times the rate it laid off Level 4/D and 5 employees (43/1016; 4.2%). Spirit’s Hiring Policies and Practices A written Spirit policy lists terminable offenses, such as theft or fraud, that make employees “Not Eligible for Rehire.” Other employees who are involuntarily terminated are generally ineligible for rehire for 12 months. (Doc. 1041-108) (“Employees involuntarily terminated for reasons other than those outlined [above such as theft or fraud] will not be considered for rehire for 12 months following the date of termination. Depending on the circumstances, employees might be reinstated through a union grievance settlement earlier than 12 months.”) In 2009, Spirit had adopted a practice of vetting former employees who reapply. There was a concern at the time that people who had been terminated for cause were “slipping through the employment process,” leading the company to put in place a procedure to vet all former employees. If the person responded on pre-application questions that they had worked at Spirit before, someone in HR would vet their background to see why they were terminated. (Doc. 1041-78 at 23.) Laid-off SPEEA employees were not categorically ineligible for rehire. Spirit elected not to include a no-rehire clause in severance packages offered to Wichita employees. Hiring 29 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 30 of 83 managers had discretion and differing views on whether to hire eligible former employees for positions they were trying to fill. Spirit had largely frozen hiring for salaried SPEEA positions as of June 2013. Spirit’s HR worried this would put it behind in recruiting. Spirit’s policy as of August 2013 was that the company would hire in key roles but would do no hiring in job codes that were impacted by the RIF. The company would continue to hire hourly employees for backfill as needed and would continue to process offers already extended, including 6 SPEEA hires (1 rehire) who started between July 26 and August 23, 2013. There were no other SPEEA hires until 2014. (Doc. 10819; Doc. 1084 at 9.) Spirit engaged vendor ADP (and its predecessor) to manage its recruiting. ADP recruiters did not have knowledge of, or access to, candidates’ (or their dependents’) ages, medical conditions or costs, disabilities, Spirit leave usage, or Spirit performance or retention ratings. ADP recruiters post job openings in an applicant tracking system called RightThing Recruit (“RTR”), which job seekers could access through Spirit’s website. Spirit often receives hundreds of applications for a single requisition, which can include dozens of applications by former employees. In some cases, multiple Plaintiffs applied to fill the same job opening. To be considered for hire, a candidate ordinarily has to apply through the RTR portal. Whenever a candidate starts to apply, a new application is created in RTR. RTR automatically generates emails to candidates to confirm submission of an application and with status updates throughout the process. If there is no entry for a candidate in RTR, Spirit has no record of the purported application and would not have considered the candidate for the position. The application process has multiple steps, and candidates are assigned an RTR status that tracks their progress and records a final outcome. At any stage, a candidate may withdraw (or “deselect”) 30 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 31 of 83 an application. A candidate applies for a specific requisition by completing the applicant prescreen (“AP”) for the requisition. It includes “knockout” questions, where a disqualifying answer automatically removes the application from consideration, such as when candidates respond that they are not 18, are not authorized to work in the U.S., or are a foreign national who requires sponsorship. Applicants to some hourly jobs can be knocked out when they acknowledge they lack the requisite experience or lack the ability to work any shift. Knocked-out applications are coded “AP-Not Qualified.” Other prescreen answers can cause applications to be put on hold. Candidates who identify as a former Spirit employee were automatically assigned an “AP-Hold” status, regardless of the candidate’s age or the date of, or reason for, their exit from Spirit. Such applications do not automatically advance to a recruiter for consideration. Former employees assigned an AP-Hold status remained in that status pending a manual review by Spirit HR to determine eligibility for rehire. Given the volume of applicants, the review process is labor-intensive, as Spirit HR had to separately determine eligibility for each candidate. HR did not review prior performance records or instruct hiring managers whether to hire an eligible candidate. Spirit manually reviewed APHold candidates on an ad hoc basis, such as when a hiring manager asked to see the application of a candidate the manager knew had applied or when Spirit HR requested a bulk review of the APHold pool on a hard-to-fill requisition. If, as often happened, the requisition closed with candidates still in the AP-Hold pool, those applications would be assigned an “AP-Inactive-Not Selected” status, which meant no ADP recruiter or Spirit employee ever reviewed their application. The APHold process has been in place since 2009 and the ad hoc review practice has been used since 2011. (Doc. 1041-58 at 19-21.) 31 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 32 of 83 The uncontroverted facts show that Spirit at times engaged in a manual review vetting process when inquiries were made about former employees. In August 2013, a manager requested to hire a worker in Wichita who had been laid off by Spirit in Oklahoma (but who sometimes helped in Wichita). One Spirit HR employee emailed another indicating opposition, noting there was no requisition for the job and that “we should not be bringing people back that just got [laid] off and [who] got a severance package.” (Doc. 1081-11 at 1.) Another representative agreed, emailing Vice President of HR Justin Welner that it “seems to defeat the purpose of the layoff, hitting headcount/financial numbers,” adding that it “diminishes our defense with SPEEA when we say they were all C’s but some we want back,” and noting that in the Oklahoma layoff everyone was made ineligible for rehire. The HR representative suggested that we “at least incorporate into our Spirit hiring procedures a 12 month break in service before we will consider anyone for rehire….” (Id.) A few days later, on another manager inquiry about bringing back a laid off worker, HR discouraged it, saying “we shouldn’t be bringing people back who are recently laid off and received a severance package,” and that the expectation was “if there are other qualified candidates they should be provided to the manager first before we would explore hiring a recently laid off employee.” (Doc. 1081-19 at 1.) At some point after the 2013 RIF, Welner determined that the 12-month ineligibility policy would not apply to those who were laid off in the RIF. But according to Welner they “were also our bottom 10 percent for performance and capability,” and the company had spent a lot on severance packages, so although they were eligible, [t]hey weren’t likely to be hired back.” (Doc. 1041-78 at 12.) Spirit held its first post-RIF job fair in October 2013 for open hourly positions. (Doc. 108415.) At the time, Spirit had 487 open requisitions, but it only had three open, non-management salaried positions in Wichita. (Doc. 1062-3.) 32 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 33 of 83 An HR representative cautioned in a July 16, 2014, email that although RIF’d workers were eligible for rehire after one year, the company should perform “due diligence” and “vet the reason they were a designated C and [ensure] there are not past performance concerns.” (Doc. 1041-111.) The representative said Spirit had “paid a lot of money to get them out the door and we should not be passing up an opportunity to ‘upskill’ and get the right talent in the door.” (Id.) An ADP recruiter screens the resumes of candidates who pass the AP stage against the job qualifications on a first-in-first-out basis. For salaried positions, the recruiter identifies the top candidates for a phone interview (“PI”) stage to further assess the candidate. Once the recruiter identifies a full slate of candidates for the hiring manager, typically the top 3-5 (hourly) or top 8 or so (salaried) candidates per opening, the recruiter stops reviewing applications unless the hiring manager asks to see more. If a recruiter notices a candidate has an active application on another requisition, the new application is coded “Active on Other Req,” which puts it on hold without further consideration to avoid issues caused by multiple managers considering the same candidate for different jobs at the same time. Recruiters exercise judgment based on resumes, phone interviews, and position requirements to determine which candidates to forward to a hiring manager (“HM”) stage. Based on resumes and recruiter PI notes, the hiring manager (rather than Spirit HR) decides who to interview and who should be offered the position. Hired candidates are coded as “HI-Hired” in RTR. Spirit requires candidates for some jobs to complete a preemployment assessment. Candidates who fail it are coded “Not Qualified.” ADP also proactively searches for candidates who might be interest in open Spirit positions. It “resume mines” (“RM”) databases such as Indeed and invites prospective candidates to apply for a Spirit job. Job seekers can also join a “talent community” to receive notification of new job postings. ADP collects information about these prospects and assigns them an “RM” 33 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 34 of 83 status, but they still must go to Spirit’s recruiting portal and submit an application before they can be considered for a particular opening. Candidates who show a final “RM” status did not accept ADP’s invitation to apply for a job opening. When ADP recruiters search existing candidate profiles in RTR for a prospective candidate, a new entry for that candidate is created in RTR but is not assigned any status. Entries without any status are not applications for an open Spirit position. SPEEA employees had recall rights under a CBA unless they were designated by Spirit, but the recall process is separate from the process for applying to an open requisition. Post-Layoff Hiring Outcomes In the nine years since the 2013 layoff, many laid-off SPEEA employees have sought employment with Spirit. Spirit has reemployed 34 of them, including 5 Plaintiffs (Carter, Denny, Doyon, Jackson, Moran), and engaged 5 others (including Russell) as contractors. At least 2 declined a job offer (Williams, Hatcher). RTR data for requisitions on which Hiring Plaintiffs applied after the layoff (through March 31, 2017) show Spirit hired 799 older workers on those requisitions, in line with the age demographics of applicants. Extracting from the RTR data all entries for Hiring Plaintiffs created on or after July 25, 2013, through March 31, 2017 (the discovery period) captures 790 entries, with a total of 33 different disposition statuses in 8 different stages, including 37 RM or no-status entries. After excluding the (37) RM and no-status entries, the remaining 753 entries represent an application submitted through RTR by 48 Hiring Plaintiffs. Those 48 Plaintiffs submitted as few as 1 application (Moehring, Troilo, Byram, Gardner, Rahbar) and as many as 74 (Jackson), from as early as July 25, 2013 (while still employed by Spirit) to February 11, 2017, and to 467 different requisitions managed by 16 different ADP recruiters or Spirit HR staffers, covering 392 different 34 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 35 of 83 job titles and from college internship and entry-level positions to expert, management, director, and senior director level positions, located at 5 different Spirit locations in 3 different states. Of Hiring Plaintiffs’ 753 applications, at least 36 were submitted to a requisition after Spirit had already made its last job offer to any candidate on that requisition, such that the Hiring Plaintiff could not have been considered for that opening. Additionally, 3 applications are duplicates by the same Hiring Plaintiff to the same requisition. Of the remaining 715 applications, 329 were submitted to a requisition on which Spirit never hired anyone. Of the remaining 386 applications, 92 were submitted (a) for Hiring Plaintiffs who filed an EEOC Charge, but more than 300 days before filing a failure-to-hire Charge or anytime thereafter; or (b) for Hiring Plaintiffs who did not file such a Charge, more than 300 days before the first-filed or anytime after the last-filed failureto-hire Charge by any Hiring Plaintiff. Of the remaining 294 applications, 7 were automatically disqualified. Five were knocked out by their responses to prescreening questions (AP-Not Qualified) (Williams (2), West, Jackson, Martin); one failed a preemployment test for availability to work any shift (Jackson); and one was withdrawn (Lawson). Of the remaining 287 applications, 276 were “held” in RTR and not submitted for further review: 272 remained in AP-Hold (AP-Inactive-Not Selected); 4 were assigned “Active on Other Req” status. Of the remaining 11 applications: 1 led to a rejected job offer (Williams, age 52) on a requisition for which two other similarly-aged (51) Plaintiffs applied; and 7 were to requisitions on which Spirit hired someone over age 40, including 4 on which the hire was older or within one year of the age of the Hiring Plaintiff. Of the 7 applications on which Spirit did not hire someone older or within one year of the Hiring Plaintiff, an ADP recruiter rejected 4 at the resume-screening 35 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 36 of 83 stage; and as to 3 others a hiring manager who did not know the Hiring Plaintiff’s age or medical information hired another qualified candidate. Two of the Hiring Plaintiffs (Carter, Jackson) in the latter group were later hired by Spirit. Of the 790 entries discussed above, 36 (4.6%) reached a hiring manager. Of those, 27 were submitted for requisitions on which Spirit never hired anyone. B. Summary of Expert Opinions Relating to RIF Statistics 1. Plaintiff Expert Lance Kaufman Initial Report. Lance Kaufman is a consultant with a Ph.D. in Economics. He was “retained by Plaintiffs to evaluate disparities in terminations related to age and Spirit-covered medical expenses.” (Doc. 1042-2 at 5.) He analyzed Spirit data related to: 1. The impact of age and medical costs on changes in performance management and retention rankings from 2010 to 2013; 2. The impact of age on the relationship between performance ratings and retention rankings in 2013; 3. The impact of age on the July 2013 reduction in force; and 4. The impact of medical leave for Spirit-insured employees and family members on the July 2013 reduction in force. (Id.) His principal findings are: 1. Changes in Spirit performance management ratings and retention rankings between 2010 and 2013 show significant disparities for Older Employees. These disparities are statistically significant after controlling for Job Code, medical costs, and discipline. [footnote omitted.] Thus, these disparities are consistent with the Plaintiffs’ contention that in 2013, Spirit’s changes to the performance management and retention ranking processes disproportionately disadvantaged older employees. 2. Changes in Spirit performance management ratings and retention rankings between 2010 and 2013 show significant disparities for employees taking medical leave for themselves or their family members that was covered by Spirit health insurance (“Covered Leave”). These disparities are statistically significant after controlling for Job Code, age, and Discipline. Thus, these disparities are consistent with the Plaintiffs’ contention that Spirit’s July RIF was motivated in part by medical costs of employees. 36 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 37 of 83 3. Spirit terminated a disproportionately large share of Older Employees in the July RIF. This difference is statistically significant and remains statistically significant after controlling for Job Code, performance rating, medical costs, and discipline. 4. Spirit terminated a disproportionately large share of employees with Covered Leave in the July RIF. This difference is statistically significant and remains statistically significant after controlling for Job Code, performance rating, age, and discipline. (Id. at 5-6.) Kaufman first looked at employees whose PM rating was lower in 2012 than in 2010. Using that population of workers, Kaufman determined that 62.1% of older workers (1006/1620) experienced a rating reduction, while 49.1% of workers under 40 (168/342) experienced a rating reduction from 2010 to 2012. To account for potential differences in ratings across jobs, Kaufman performed a logistic regression of rating reduction using both age group and job code as dependent variables. Kaufman found that after controlling for job code, older workers had greater odds (a ratio of 1.63) of receiving a PM rating reduction, and the disparity was highly statistically significant. (Id. at 14-15.) Kaufman next observed that Spirit’s decision to “designate” all C retention-ranked employees ahead of the RIF meant that only older workers were deprived of a seniority bump under the CBAs, as no workers under 40 had 20 years of service, while 62% of older workers did. (Id. at 15-16.) Next, Kaufman examined retention rankings of what he termed “Low-Rated employees,” meaning those rated as “Meets Some” or “Unacceptable.” He examined 2010 retention ratings and found a 14.3% chance of older, Low-Rated workers (2/14) being ranked C, while 45.5% of younger Low-Rated workers (5/11) were ranked C, although this disparity was not statistically significant. By contrast, he found in the 2013 retention exercise that 67.5% of older Low-Rated workers were ranked C, while 33.3% of younger Low-Rated workers were ranked C. Kaufman found older Low-Rated workers in 2013 were twice as likely as younger Low-Rated 37 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 38 of 83 workers to be designated C, and that the disparity was highly statistically significant. (Id. at 1617.) Because Kaufman was retained in part to evaluate disparities in terminations related to Spirit-covered medical expenses, he attempted to evaluate disparity pertaining to “employees taking medical leave for themselves or their family members that was covered by Spirit health insurance,” a status termed by Kaufman as “Covered Leave.” To do so, Kaufman “constructed a proxy measure for medical expense by combining medical leave data with insurance data” using a somewhat complicated process, because Kaufman did not have historical medical expense information by employee. (See Doc. 1042-2 at 9-10.) Using the proxy of Covered Leave, Kaufman looked at C ranked employees in 2013 and compared those employees with and without Covered Leave, finding that 17.1% (26/152) of those with Covered Leave were ranked C, while 9.3% (191/2052) of those without Covered Leave were ranked C, a risk ratio of 1.84 and a highly statistically significant (4 out of 1,000) disparity. Kaufman next examined the impact of changes in retention rankings in 2013, looking at employees for whom retention rank was reduced from A or B in 2010 to C in 2013. He found 8.1% (132/1620) of older workers had a reduced ranking compared to 4.1% (14/342) of younger workers, a risk ratio of 1.99 that was statistically significant (1 out of 100). Kaufman performed a logistic regression said to control for performance rating, age, Covered Leave, discipline, year and job code level, and found older workers had more than double (2.41) the odds of being reduced to a C ranking, a statistically significant disparity representing 2.74 standard deviations. (Id. at 21.) Kaufman then examined the RIF itself, finding that among “Low-Rated” employees – that is, those with Meets Some or Unacceptable 2012 PM ratings -, 56.4% of older workers (163/289) were terminated compared to 28.9% of younger workers (24/83). He found the disparity was 38 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 39 of 83 highly statistically significant (2 out of 50,000 chance). With respect to employees with Covered Leave, Kaufman found a more than double risk that older workers would be terminated in the RIF compared to younger workers, 16.4% (36/220) to 7.2% (231/3202), with the disparity highly statistically significant (2 out of 1,000,000). Kaufman performed a logistic regression to account for the following factors: Covered Leave, Non-Covered Leave, Discipline, [2012] performance rating, and job code. Under that analysis, he found older workers and workers with Covered Leave each had more than double the odds (2.05 and 2.35) of being terminated, statistically significant disparities of just over 3 standard deviations. (Id. at 23.) Kaufman concluded his analysis showed “statistically significant age and health disparities in … Spirit’s 2013 performance management exercise, retention ranking process, and July RIF selections,” which “are consistent with Plaintiff’s assertions that Spirit employment practices harmed older workers.” (Id. at 24.) 2. Defendant Expert Paul White Initial Report Paul F. White is a labor economist with a Ph.D. in the field of economics. He was retained by Defendants to evaluate the claims in the First Amended Complaint and respond to Plaintiff’s economic experts, including Dr. Kaufman. (Doc. 1041-119.) White states that his statistical analysis “indicates that Spirit’s July 2013 RIF … did not disadvantage older workers.” (Id. at 5.) He faults Kaufman’s analysis, among other reasons, for “analyzing an incorrect and overly inclusive population, focusing on an arbitrary subset of the relevant population, making ill-founded assumptions,” and modelling errors. (Id. at 5-6.) White states that when he corrects for these errors, he finds no statistically significant difference in termination rates in the RIF. (Id. at 6.) White notes that instead of analyzing only the SPEEA employees on the Older Worker Benefit Protection Act (“OWBPA”) disclosure list, Kaufman 39 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 40 of 83 examined the entire list, even though over one-third of the workers on the list were not SPEEA members and were not subject to the same selection process (i.e. the retention rating exercise) as SPEEA members. White says Kaufman’s comparisons cannot and should not be used to describe the impact of the RIF on SPEEA members, observing that Plaintiffs’ claims in this suit “are limited to … SPEEA laid off employees….” (Id. at 25.) He says Kaufman’s analysis “cannot address whether SPEEA employees age 40 or over were more likely to be terminated in the July 2013 RIF.” (Id.) White also faults Kaufman for arbitrarily selecting a subset of “Low-Rated” workers to examine for age-related disparity, because “plaintiffs did not claim age discrimination among the ‘low-rated’ employees,” and the alleged age disparity disappears when the relevant population of all SPEEA workers is considered. According to White, “Dr. Kaufman’s own methodology and his own data demonstrates age neutrality in the RIF once applied to all the SPEEA employees on the OWBPA disclosure list.” (Id. at 7.) White’s analysis concluded there was no significant change in average age among the relevant population before and after the July 2013 RIF. He compared the average age of Wichita SPEEA represented employees before and after the July 2013 RIF and found it was 48.7 and 48.5, respectively, a statistically insignificant difference. He noted the average age of Wichita SPEEA employees from 2012 to 2017 was close to 48 throughout that period, all of which he concluded was inconsistent with Plaintiffs’ assertion that Spirit undertook a campaign to reduce older workers and replace them with younger workers. (Id. at 9-10.) White said that the relevant population on Plaintiffs’ claims are the 2,296 SPEEA members on the OWBPA disclosure list, of whom 221 were selected for layoff. Within that relevant population, he compared the number of older workers who were actually terminated against the 40 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 41 of 83 number of older workers that would be expected to be terminated based on their representation among all SPEEA employees on the list. He used multiple pools to reflect the fact that Spirit’s retention ranking was broken down by job code for WTPU employees, and by skill management code for WEU employees, and to reflect “the decentralized decision-making process Spirit used to make layoff selection decisions.” (Id. at 13.) He said not every employee on the OWBPA list should be compared to the entire pool of employees considered for layoff, but rather to employees within the same job code or skill management code. Based on a relevant population of 1,832 workers aged 40 and over, and 464 employees under age 40, and controlling for the foregoing factors, White said it would be expected that just over 177 workers age 40 or over would be selected for layoff, whereas there were actually 185 older workers selected and 36 younger workers. White said this difference (8 more older workers terminated than expected) represented 1.31 standard deviations and was not statistically significant. (Id.) White said a similar result is reached even using Dr. Kaufman’s “overly inclusive” entire OWBPA disclosure list. (Id. at 1516.) White addressed Dr. Kaufman’s findings and said his selection and comparison of “LowRated” employees was arbitrary, and that there is no evidence of age discrimination in the RIF when an appropriate pool of the relevant population is examined. White said the focus on LowRated workers was arbitrary because Plaintiffs’ claim is for age discrimination against all SPEEA members on the OWBPA list, not just the Low-Rated ones. Moreover, he notes the 2013 retention ranking was based not only on workers’ PM ratings, but also on their versatility and criticality, factors he said were ignored by Kaufman’s analysis. White said that because retention ranking included the latter factors, the 2013 retention rankings provided a more comprehensive measure of the factors considered by Spirit in the RIF than the 2012 PM ratings, and that when the former 41 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 42 of 83 are examined under a logistic regression, older workers were not statistically significantly more likely to be terminated in the 2013 RIF than younger workers. (Id. at 36.) White also noted that in 2013, 73 people received a C ranking who were nevertheless rated as Meets Expectations (i.e., not Low-Rated). White said when the bottom three PM categories are examined (Meets, Meets Some, Unacceptable), Kaufman’s “statistically significant adverse findings completely disappear,” and there is no statistically significant difference in assignment of C rankings between older and younger workers in the 2013 exercise. (Id. at 28.) White repeated the analysis with employees in one of the top three PM categories and with all SPEEA employees who were ranked in 2010 and 2012, and found older workers were not more likely than younger workers to receive a C ranking. White said the results showed that even if older employees were more likely to see a decline in PM rating from 2010 to 2012, they were not any more likely to receive a C ranking in 2013 that would lead to their termination. White found that when the assignment of C rankings among all SPEEA employees was examined, the disparity between older workers receiving Cs (10.3%) and younger workers receiving Cs (7.8%) was not statistically significant. (Id. at 29, Table 11.) White said Kaufman’s analysis of employees taking medical leave did not consider the relevant population, as it also included non-SPEEA workers on the OWBPA list. White further said the analysis was not tied to Plaintiffs’ claims of age discrimination, as whether or not workers with higher medical costs were more likely to be terminated “is ultimately unconnected to the issue of discrimination against older workers in the July 2013 RIF.” (Id. at 31.) White said Kaufman attempted to connect the two by relying on a general study finding that insurance claim costs were higher for older persons and assuming that older employees at Spirit generate higher medical costs to Spirit than younger employees. But if “Covered Leave” was a reasonable proxy for such costs, 42 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 43 of 83 White stated, then the proportion of older workers taking covered leave should be higher for older than younger workers, but White concluded that was not the case. (Id. at 32.) 3. Rebuttal and Supplemental Reports Kaufman and White each offered follow-up opinions, including Kaufman’s rebuttal report (Doc. 1042-3), White’s supplemental report (Doc. 1042-9), Kaufman’s supplemental report (Doc. 1042-4), White declaration (Doc. 1042-7), and Kaufman deposition (Doc. 1042-8). Kaufman’s rebuttal findings included his opinion that White’s analysis of average age before and after the RIF was not statistically valid, and that it does not reliably identify age disparities in terminations. (Doc. 1042-3 at 4.) He said the “pools” analysis used by White overly fragmented the data on terminations and made it too weak to identify significant disparities. Even so, Kaufman said a pools analysis shows statistically significant age disparities. Kaufman maintained that the OWBPA list and the Low-Rated employees were appropriate groups to analyze, but he said in any event his results are not materially impacted if limited only to SPEEA employees. Kaufman also criticized White’s rejection of the broader OWBPA list, saying White “provides no explanation for why potential discriminatory behavior would be limited to SPEEA employees.” (Id. at 10.) He said use of the more limited SPEEA subset would still show statistically significant disparities disadvantageous to older workers in PM rating reductions, C ranking reductions, and terminations of Low-Rated employees. (Id. at 11.) Kaufman emphasized that his analysis of PM rating and C retention reductions was intended to demonstrate that “age disparities exist in both the performance rating process and the retention ranking process,” not to connect those two matters to terminations. (Id. at 13.) Kaufman said controlling analysis of terminations using retention rankings, as White suggested, was not statistically valid “because the retention ranking process is subject to age 43 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 44 of 83 disparities.” (Id.) With respect to analysis of Low-Rated workers, Kaufman rejected White’s accusation that the analysis was arbitrary, stating that Low-Rated workers were targeted by Spirit and “[t]hese employees were terminated at a much greater rate,” which “leads to a reasonable conclusion that Spirit’s termination process for low rated employees differs from Spirit’s termination practices for employees assigned a ‘Meets’ performance rating.” (Id. at 14.) Kaufman believes that White’s inclusion of employees with “Meets” ratings in the analysis is inappropriate because it “fails to appropriately control for the relationship between performance rating and termination.” (Id.) Kaufman also defended his examination of Covered Leave, saying it was reasonable to assume employees with a need for medical leave will incur medical costs and some of those costs would be passed on to Spirit. In any event, Kaufman said his analysis showed a ranking and termination disparity with respect to employees with Covered Leave, not health care costs. (Id. at 16.) In rejecting Dr. White’s suggestion that his regression models should control for retention ranking rather than performance ranking, Dr. Kaufman said his analyses “suggest that the presence of age disparity in termination begins with disparately low performance ratings, is magnified by disparities in retention rankings, then flows through to termination outcomes,” and that controlling for retention rankings “risks explaining away the age effect and risks obscuring the true role of age in termination outcomes….” (Id. at 18.) Kaufman further asserted that retention ranking was not a good control because it was highly correlated with termination and would leave little explained variation in terminations. Kaufman also rejected White’s suggestion that he should have included both 2010 and 2012 PM ratings changes as controls in his regression models, saying there was too much correlation between the two. In sum, Kaufman said nothing in White’s report caused him to change his opinions. 44 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 45 of 83 White’s supplemental report noted a correction relating to errors in 13 Spirit identification (ID) numbers, which had some minor effect on his calculations and his review of Kaufman’s analyses. White again identified 3,473 total Wichita employees on the OWBPA disclosure list who were considered for layoff, of whom 2,296 were Wichita SPEEA employees, and of whom 221 SPEEA employees were selected for the RIF. White said Kaufman and another expert (Dr. Robert A. Bardwell) incorrectly identified 2,388 SPEEA employees on the list in their rebuttal reports, caused in part by a coding error that eliminated entries for duplicate Spirit ID numbers over a period of years, but which failed to account for whether the employee was a SPEEA member at the time of the RIF. (Doc. 1041-120 at 7-8.) White said this error meant Kaufman’s analysis could not provide useful evidence relating to SPEEA employees. White updated his own calculations and again found no statistically significant disparity in termination of older and younger SPEEA employees in the RIF. (Id. at 9.) White further found no statistically significant disparity using the entire OWBPA list, including non-union employees, as suggested by Kaufman’s analysis. (Id. at 11.) White corrected his calculations relating to the number of SPEEA employees within the various PM rating categories – including 70 rated “Meets” who were selected for the RIF – and again found no statistically significant disparity in older versus younger workers selected for the RIF from the three lowest PM categories. (Id. at 13.) White also said if Kaufman’s analysis of the effect of age on termination is limited to the 2,296 SPEEA employees on the OWBPA list, there is no statistically significant disparity between older and younger workers who were terminated in the RIF. (Id.) White again maintained that both the 2010 and 2012 PM ratings should be added as control factors in Kaufman’s regression model for the RIF, and that doing so showed no statistically significant disparity in ages of workers selected for the RIF. (Id. at 14.) 45 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 46 of 83 Kaufman’s responsive supplemental report (Doc. 1042-4) said the data corrections identified in White’s report did not alter the findings in Kaufman’s prior reports. (Id. at 4.) Kaufman concluded the new termination analyses in White’s supplemental report are erroneous; that White inappropriately removed relevant and statistically important control variables from his own analyses and from Kaufman’s; that the change in 13 Spirit IDs and White’s method of identifying SPEEA employees makes no difference in the results in the initial Kaufman report and no material difference in the results in Kaufman’s rebuttal report; that the “new analysis of SPEEA employee identification methodology in the White Supplemental Report is immaterial”; and that Kaufman’s findings “support an inference that Spirit’s RIF had a disparate impact on older employees.” (Id. at 7.) C. Applicable Standards 1. Summary Judgment. Summary judgment is appropriate if the moving party demonstrates that there is no genuine dispute as to any material fact, and the movant is entitled to judgment as a matter of law. Fed. R. Civ. P. 56(a). “[T]he mere existence of some alleged factual dispute between the parties will not defeat an otherwise properly supported motion for summary judgment; the requirement is that there be no genuine issue of material fact.” Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 247– 48 (1986) (emphases in original). “A fact is material if, under the governing law, it could have an effect on the outcome of the lawsuit. A dispute over a material fact is genuine if a rational jury could find in favor of the nonmoving party on the evidence presented.” Doe v. Univ. of Denver, 952 F.3d 1182, 1189 (10th Cir. 2020) (quoting Jones v. Norton, 809 F.3d 564, 573 (10th Cir. 2015)). Conclusory allegations are not sufficient to create a dispute as to an issue of material fact. See Hall v. Bellmon, 935 F.2d 1106, 1110 (10th Cir. 1991). The court views 46 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 47 of 83 the evidence and all reasonable inferences therefrom in the light most favorable to the nonmoving party. LifeWise Master Funding v. Telebank, 374 F.3d 917, 927 (10th Cir. 2004). 2. Pattern and Practice Claims. In Teamsters, the Supreme Court said the plaintiff could prove a discrimination claim by showing that an employer's conduct was part of a pattern or practice of discrimination against minority employees. Teamsters, 431 U.S. at 336. To do so, the plaintiff had to show “more than the mere occurrence of isolated or ‘accidental’ or sporadic discriminatory acts.” Id. It had to “establish by a preponderance of the evidence that racial discrimination was the company's standard operating procedure [--] the regular rather than the unusual practice.” Id. In Apsley v. Boeing Co., 691 F.3d 1184 (10th Cir. 2012), the Tenth Circuit outlined the order of proof on a pattern and practice claim: Under the ADEA, a class of plaintiffs proceeding under a pattern or practice theory must first make a prima facie showing that unlawful discrimination has been a regular procedure or policy followed by an employer or group of employers. If they succeed, the burden then shifts to the employer to defeat the prima facie showing of a pattern or practice by demonstrating that the plaintiffs' proof is either inaccurate or insignificant. If an employer fails to rebut the inference that arises from the [plaintiffs'] prima facie case, the finder of fact can conclude that a violation has occurred and the trial court can award prospective equitable relief. If the plaintiffs also seek individual relief for the victims of the discriminatory practice, the case moves into the second or subsequent stages. In these additional proceedings, it must be determined whether each individual plaintiff was a victim of the discriminatory practice. Importantly, by having prevailed in the first stage of trial, the individual plaintiffs reap a significant advantage for purposes of the second stage: they are entitled to a presumption that the employer had discriminated against them. Id. at 1194 (citations and quotation marks omitted.) At the first stage of a pattern or practice claim, a summary judgment motion “must focus solely on whether there is sufficient evidence demonstrating that defendants had in place a pattern or practice of discrimination during the relevant limitations period.” Thiessen, 267 F.3d at 1109. 47 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 48 of 83 3. Daubert. Federal Rule of Evidence 702, which controls the admission of expert witness testimony, provides: 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: (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 has reliably applied the principles and methods to the facts of the case. Fed. R. Evid. 702. Under this rule, the district court must satisfy itself that the testimony is both reliable and relevant, in that it will assist the trier of fact, before permitting a jury to assess such testimony. Schulenberg v. BNSF Ry. Co., 911 F.3d 1276, 1282 (10th Cir. 2018) (citing United States v. Nacchio, 555 F.3d 1234, 1241 (10th Cir. 2009) (en banc)). The district court must first determine whether the witness is qualified by knowledge, skill, training, experience, or education to render an opinion. Id. If so, the district court must determine whether the witness's opinion is reliable by assessing the underlying reasoning and methodology. Id. at 1283. The court is not required to admit opinion evidence that is “connected to existing data only by the ipse dixit of the expert,” and may exclude the opinion if “there is simply too great an analytical gap between the data and the opinion offered.” Id. (quoting Gen. Elec. Co. v. Joiner, 522 U.S. 136, 146 (1997)). But the rejection of expert testimony is the exception rather than the rule, and “[v]igorous crossexamination, presentation of contrary evidence, and careful instruction on the burden of proof are the traditional and appropriate means of attacking shaky but admissible evidence.” Daubert v. Merrell Dow Pharm., Inc., 509 U.S. 579, 596 (1993). 48 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 49 of 83 “The court has discretion to determine how to perform its gatekeeping function under Daubert.” In re EpiPen (Epinephrine Injection, USP) Mktg., Sales Practices & Antitrust Litig., No. 17-MD-2785-DDC-TJJ, 2020 WL 1164869, at *3 (D. Kan. Mar. 10, 2020) (citing Bill Barrett Corp. v. YMC Royalty Co., LP, 918 F.3d 760, 770 (10th Cir. 2019)). D. Analysis 1. Pattern or Practice - Termination Claim – Disparate Treatment. Plaintiffs claim that they were terminated on account of their age pursuant to a pattern or practice of age discrimination by Spirit. They assert both disparate treatment and disparate impact theories. As indicated by Apsley, the initial question is whether Plaintiffs have cited evidence from which a jury could reasonably find that “unlawful [age] discrimination has been a regular procedure or policy followed by” Spirit. Plaintiffs cite both anecdotal and statistical evidence which they believe supports this claim. Noticeably absent from Plaintiffs’ evidence is anything suggesting that prior to or aside from the 2013 RIF, Spirit had any history or common practice of engaging in age discrimination generally such that it could be fairly said to be the company’s “standard operating procedure.” Cf. Teamsters, 431 U.S. at 336. To the contrary, the uncontroverted facts show that Spirit employed a predominantly older workforce, including at its Wichita site. Defendants argue that against this background, the challenged RIF cannot support Plaintiffs’ claim because “a ‘one-shot’ layoff … cannot be a ‘pattern or practice.’” (Doc. 1041 at 40.) Plaintiffs’ allegation that Spirit altered its traditional rating employee practices to specifically target older workers in a one-time RIF in 2013 hardly seems consistent with a showing that discrimination was the company’s “regular rather than unusual” practice. Teamster, 431 at 336. Cf. Equal Emp. Opportunity Comm'n v. TriCore Reference Lab'ys, 849 F.3d 929, 937 (10th Cir. 2017) (“The claim requires proof that the employer 49 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 50 of 83 had a regular practice of discrimination—'more than the mere occurrence of isolated or ‘accidental’ or sporadic discriminatory acts.’”) (quoting Teamsters, 431 at 336). Still, the court finds no authority stating that a company RIF, in and of itself, cannot support a claim alleging a pattern or practice of discrimination. Cf. Thompson v. Weyerhaeuser Co., 582 F.3d 1125, 1127 (10th Cir. 2009) (affirming district court’s conclusion that pattern or practice framework could be invoked by plaintiffs in an ADEA case; plaintiffs had been terminated in a RIF). But assuming the rather dubious proposition that a one-time RIF can support a pattern or practice claim, Plaintiffs nevertheless fail to cite evidence to support such a claim here. The uncontroverted facts show that hundreds of different Spirit managers – not Spirit HR or executives – made the PM and retention rankings that led to the selections in the RIF on which Plaintiffs base their pattern or practice termination claims, and the facts further show that the managers made these rankings applying facially neutral standards. Plaintiffs nevertheless argue that Spirit’s HR “controlled performance management and retention processes” in order “to accomplish [Spirit’s] stated goal to cut high-cost, long-tenured employees….” (Doc. 1070 at 44.) But the uncontroverted facts show only that Spirit HR required managers to conform to bell-curve distributions in PM ratings and emphasized ratings based on performance rather than on other factors, including seniority, for the purpose of identifying lesser performing employees. In other words, Spirit HR set age-neutral guideposts within which managers exercised discretion to evaluate the employees under their supervision. Similarly, HR trained Spirit managers how to make ranking decisions on the 2013 retention exercise under a CBA-required distribution scale, but again directed them to rank based on performance, not upon improper factors, and directed managers to identify employees “at the bottom tier of performance” with “skill deficiencies” or “with behavior/Value deficiencies.” (Doc. 1080-38 at 6.) HR also discouraged managers from 50 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 51 of 83 “taking the easy way out” and simply filling up the C category with new employees, saying new employees “should be evaluated based on how they are progressing, their contribution to date, and their potential going forward.” Welner testified this was done because he wanted managers “to go through the hard work of truly identifying the bottom performers so that that’s the group we exited.” (Doc. 1070-33 at 9.) Cf. Pippin v. Burlington Res. Oil and Gas Co., 440 F.3d 1186, 1197 (10th Cir. 2006) (“Indeed, our cases have previously held that leaving out new employees from RIF decisions does not establish pretext.”) (citation omitted.) Some managers were obviously uncomfortable with or resistant to the change this represented from prior years (marked by inflated performance ratings by some managers and a habit of other managers to rank most employees in the middle of the performance scale) and to the concept of singling out and identifying employees whose performance, compared to others, was lacking. But nothing in the evidence cited by Plaintiffs reasonably supports an inference that HR’s company-wide training dictated or even suggested to managers who made the individual rating decisions that they should select employees for lower ratings based on their age. The ranking criteria imposed by HR were age-neutral and do not support an inference that Spirit endorsed or required managers to apply age-discriminatory criteria. Plaintiffs cite evidence of what they claim was a purposeful Spirit plan to eliminate older workers, beginning with the fact that Spirit tracked its workforce age by location. But tracking average ages in the manner cited is not necessarily nefarious and has entirely legitimate purposes, including managing business risks from the loss of employees to retirement. Cf. Apsley, 691 F.3d at 1205 (age-tracking documents citing aggregate age of workforce was “innocuous”). Plaintiffs also invoke what they characterize as Spirit’s “preoccupation,” “hyper focus,” or even “obsessive[] focus” on workforce age. (Doc. 1070 at 3, 37.) In support, Plaintiffs cite little more than random 51 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 52 of 83 comments by Spirit leaders engaged in long-term business planning who indicated managers should encourage and mentor younger workers because they were the “future of the company,” which is essentially a truism. Plaintiffs question whether it was justified to cut labor costs in Wichita at all, asserting that the $590 million write-down that prompted the cutbacks at Spirit were “mainly” due to problems in Tulsa and “the Wichita plant was financially sound.” (Doc. 1070 at 4.) They similarly contend that the RIF did not really make economic sense, but they cite no evidence that Spirit did not face some economic difficulty or that Spirit leaders did not genuinely believe the RIF was in furtherance of the overall financial interest of the company. In the same vein, Plaintiffs seek to support their theory that Spirit targeted older workers through a RIF by citing opinions of Kevin E. Cahill, an economist with a Ph.D., who was asked “to provide an applied microeconomic assessment of Defendant’s actions related to its workforce reductions, increases, and planning from 2012 through 2017.” (Doc. 1043-2.) After viewing various Spirit documents and SEC filings, comparing Spirit’s actions to the actions of other companies, and applying various formulas related to profit maximization, Cahill opines that “[t]he actual reason behind Spirit’s RIF … can be assessed through a review of Spirit’s actions,” and Cahill’s review of those actions “suggests that a change in business strategy was not the primary reason for the RIF.” (Id. at 7.) He believes Spirit’s “filings with the SEC do not depict a company that is addressing a systemic issue regarding the productivity of its workforce,” while Spirit’s internal documents “stress the importance of labor costs prior to the RIF, and noted improvements in labor costs after the RIF,” such that “[c]ollectively, the empirical evidence suggests that the reason behind Spirit’s RIF was profit maximization through workforce reductions that targeted Spirit’s most expensive workers – those who were older and those who had high healthcare costs.” (Id. at 8.) In sum, he believes the RIF “exhibits several characteristics that indicate the RIF targeted 52 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 53 of 83 its most expensive workers,” including: Spirit not mentioning in public statements a change in business strategy and “block[ing] the rehiring of the workers it laid off;” public statements that did not emphasize systemic performance issues within its workforce; and Spirit internal documents that “highlight a focus on Spirit’s most expensive workers,” behavior that is “consistent with” layoffs targeted at older workers and those who had high healthcare costs. (Id. at 24.) Defendants move to exclude Cahill’s testimony under Daubert. (Doc. 1043.) They contend he is not qualified to offer opinions on interpretations of language in SEC filings, that he seeks to impermissibly opine on Spirit’s motivations for the RIF and subsequent hiring, that his opinions would be unhelpful to the jury, and that they are unreliable because he failed to consider sufficient facts and reliably apply his principles and methods. In response, Plaintiffs assert that as an economist, Cahill “is qualified to examine and extrapolate from SEC filings,” that he relied on sufficient facts to offer a reliable and helpful opinion, that he has opined on Spirit’s actions and not its state of mind, after applying a “well-founded microeconomic profit maximization framework to empirical evidence,” and that he reliably applied his methodology to the facts. Plaintiffs contend Defendants arguments go to the weight, not the admissibility, of the testimony. After review of these arguments, the briefs, and the cited materials, the court concludes the motion to exclude Cahill’s testimony should be granted. As an initial matter, the court agrees with Defendants that Cahill offers opinions outside the area of his economic expertise when he interprets or opines on the significance of statements and omissions by Spirit in its SEC disclosures. (See Doc. 1042-3 at 8) (“Spirit’s filings with the SEC do not depict a company that is addressing a systemic issue regarding the productivity of its workforce.”) Among other things, Cahill purported to “evaluate how common it is for a publicly-traded company to implement a change in business strategy that entails large-scale layoffs, and to not disclose the change in 53 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 54 of 83 strategy in its public filings.” (Id. at 16-17.) He obtained information about Worker Adjustment and Retraining Notifications (WARN) from a Department of Labor database (that included information for 24 states (not including Kansas)) and searched WARN notices for the period 2012 to 2019, for public companies laying off between 200 and 500 workers. Among 64 companies found matching the criteria, two-thirds mentioned the RIF in their 10-K or 10-Q filings with the SEC, more than half quantified the costs associated with the RIF, and 73% mentioned a business restructuring. Cahill found “only six of the 64 companies … had a pattern that resembled Spirit’s actions (i.e., no mention of a strategic change and no mention of a willingness to rehire laid-off workers.)” (Id. at 18.) While this sort of statistical analysis is clearly within Cahill’s expertise, no showing is made that he has any expertise on the legal requirements of disclosure pertaining to layoffs in such filings, or, for that matter, the business significance of non-disclosure. Yet he considered Spirit’s asserted non-disclosure as a factor contributing to his conclusion that “the empirical evidence is inconsistent with a change in business strategy as the reason for Spirit’s RIF.” (Id. at 20.) That strays beyond any demonstrated economic expertise. Noting Spirit’s assertion that it targeted underperforming workers in the RIF to control costs, Cahill said that “[t]o evaluate the plausibility of this explanation … I look to what Spirit disclosed to the market about any systemic inefficiencies within Spirit’s workforce at the time.” (Id. at 20.) Cahill found significance in Spirit’s description in various 10-Ks of its workforce as a “Competitive and Predictable Cost Structure” or “stable and predictable.” In contrast, Cahill said, certain other businesses (he cited four) commented in 10-K filings on their workforce inefficiencies, which Cahill again found to be “empirical evidence … inconsistent with systemic labor force efficiencies being the reason for Spirit’s July 2013 RIF.” (Id. at 22-23.) Like his other conclusions, this one attributes significance to the particular language used in SEC filings in a manner that strays beyond 54 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 55 of 83 economic expertise, and it rejects Spirit’s asserted explanation as implausible without an adequate factual or logical foundation, simply on the grounds that Spirit’s manner of disclosure was less common among the filings he examined. In a manner closer to his area of expertise, Dr. Cahill proceeded to discuss certain economic incentives and consequences relating to a RIF, but his ultimate conclusions similarly are not shown to rest upon an adequate factual foundation or represent a reliable method of analysis. After noting that Spirit shifted to a self-funded health plan, which Cahill said would “increase the financial incentive [for] laying off such [high cost] workers,” and noting the costs associated with retaining experienced workers, and the tradeoffs involved from reduced costs and lost gains in productivity, Cahill asserted: Unlike the change-in-business strategy explanation or the productivity-based targeted-workforce-reduction explanation, the targeting of high-cost workers has a clear financial incentive tied to it and the empirical evidence, taken as a whole, suggests that Spirit’s decisionmakers responded to this incentive. (Doc. 1043-2 at 24.) It is not at all clear what “empirical evidence” led to this conclusion. See Roe v. FCA US LLC, 42 F.4th 1175, 1180 (10th Cir. 2022) (“nothing in either Daubert or the Federal Rules of Evidence requires a district court to admit opinion evidence that is connected to existing data only by the ipse dixit of the expert. A court may conclude there is simply too great an analytical gap between the data and the opinions offered.”) (citations omitted.) The imputation of company motives based on a review of economic incentives in the manner described has not been shown to be the product of a verifiable and reliable method. Moreover, it is clear to the court that this sort of testimony, aside from the problems noted above, would not be helpful to the jury in deciding the issues in this case. A jury can plainly understand without expert testimony the concept 55 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 56 of 83 that Spirit had some incentive to reduce its labor costs, including the costs it incurred for health care. Spirit’s motion to exclude the testimony is accordingly granted.9 Plaintiffs ultimately fail to cite any competent evidence that Spirit did not genuinely believe its cost-cutting measures through the RIF were in the financial interests of the company, even if Plaintiffs or selected economic experts question Spirit’s judgment. Nor have they cited evidence that Spirit intentionally targeted older workers in the RIF, or that Spirit’s decision to significantly cut its salaried workforce in Wichita was otherwise a pretext to fire older workers. Plaintiffs attempt to tie this case to Thiessen by asserting that on two occasions a Spirit executive referred to older workers as “blockers” of new talent. (Doc. 1070 at 4.) But Plaintiffs fail to support this allegation,10 and at any rate they cite no evidence of a Spirit “blocker policy” [i.e., a plan to remove older workers for the purpose of allowing advancement or hiring of younger workers] of a type that would support a pattern or practice claim. Cf. Thiessen, 267 F.3d at 1100 (company executives discussed plans to “remove blockers,” gave presentations on getting rid of “blockers,” and distributed a list of “possible blockers” to HR). The layoffs in Wichita constituted 9 Defendants’ motion to strike (Doc. 1090) additional opinions offered by Cahill in a declaration (Doc. 1043-2) is GRANTED. The declaration attempts to insert new expert analyses (specifically in paragraphs 10-14 of the declaration) on matters not previously addressed and was offered far beyond the deadline for disclosure of expert opinions. Nor is Plaintiffs’ argument persuasive that the new opinions are merely a permissible supplement to Dr. Cahill’s initial report or a proper response to the Daubert motion. The opinions offered include an entirely new analysis of data not previously addressed in Cahill’s initial report. Plaintiffs do not assert that the data was previously unavailable. Moreover, Defendants have demonstrated they will suffer prejudice given that they relied on the scope of the initial report in determining to file a Daubert motion rather than depose Cahill or obtain an expert for rebuttal. Given the impracticality of conducting a deposition at this point, the proper remedy is to exclude the opinions in paragraphs 10-14 of the declaration. 10 In support of this allegation at Doc. 1070 at 4, Plaintiffs cite “Ex. 47, 88557” and “Ex. 59, Lawson Dep. 129:5-14.” Exhibit 47 is an email that does not discuss “blockers.” (Doc. 1080-6.) The reference is apparently to another exhibit in which Spirit’s Sam Marnick, in discussing an incentive program for voluntary retirements as an alternative to a RIF, noted “there are quite a few more tenured ops managers (lots of experience but also blockers of development of less tenured folks.)” (See Doc. 1039-2.) Aside from the absence of any showing that a policy of removing older “blockers” was employed in implementing the RIF, the email itself refers only to employee “tenure,” not age. As for the cited excerpt from Exhibit 59, it shows that CEO Lawson testified, “[i]f we didn’t have the talent, we’d go get it. If we had the talent, we didn’t want to lose it,” and that his focus was to make sure that “[w]e retain the necessary talent we needed to run our business.” (Doc. 1070-43 at 2.) He said nothing about “blockers.” 56 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 57 of 83 a cost-cutting measure that resulted in a significant headcount reduction for both older and younger workers. The only context in which Plaintiffs cite evidence of Spirit discussing replacement of “long service/high cost employees with new hires” was in the context of earlier proposals to save money through voluntary retirement initiatives, not as part of the RIF that was designed to save money by reducing the number of overall workers. The uncontroverted facts show that even if Spirit HR “tightly controlled” the employee rating process by using manager “calibration” to enforce consistent rating standards, by emphasizing the criteria to be used in ratings, and by requiring that ratings fall into a mandatory distribution, the company nevertheless identified individuals selected for the RIF through a decentralized process of individual manager judgment. Plaintiffs make no showing that this process was somehow a product of intentional age discrimination by Spirit. At most, they cite a few random instances of manager disagreement with the philosophy of being required to identify the lowest performers under their supervision. And, as discussed in more detail below, the statistical impact of this decentralized decision-making process on older SPEEA workers fails to rise to the level necessary to show that Spirit’s standard operating procedure involved age discrimination. Among other things, Plaintiffs single out Spirit’s decision to “designate” all C rated employees in the retention exercise – which meant that SPEEA workers with 20-plus years did not receive a bump in retention rating – as “powerful evidence of age-based intent,” arguing it was “intentional, consequential, adverse treatment of older workers in a direct and obvious way.” (Doc. 1070 at 39.) It is true that Spirit had never before designated all C workers, but it is also true that Spirit had never before engaged in a RIF. The CBAs gave Spirit discretion to designate C workers in this manner, and presumably did so for the very reason Spirit said it used the option: because it 57 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 58 of 83 wanted the primary factor for retention to be work performance rather than seniority. Moreover, one of Spirit’s stated goals was to achieve a ten percent reduction in workforce. Tailoring the retention rating to result in ten percent of workers receiving a C retention rating allowed Spirit to focus the RIF on the C workers. If Spirit had allowed tenured workers to receive the retention “bump” to a B rating, then eliminating the C workers would not have achieved the desired reduction in headcount and would have required Spirit to develop additional criteria to determine which subset of B workers would be added to the RIF to meet the workforce reduction goals. Thus, designating all C workers prevented the tenure “bump” from frustrating and unnecessarily complicating efforts to achieve the goal of a ten percent RIF. The prospect of receiving a retention “bump” under the CBA no doubt raised the expectations of C-rated SPEEA workers with 20 or more years’ experience that they would be spared from the RIF. Spirit’s designation dashed those expectations, but Spirit was within its rights under the CBA to do so and, as a practical matter, the effect of the designation was that all SPEEA workers – young and old alike – were subjected to exactly the same standards for retention ranking and selection in the RIF. The fact that Spirit designated all C-rated employees does not lend support to the allegation that Spirit engaged in a pattern or practice of age discrimination. Cf. Pippin, 440 F.3d at 1197 (“Indeed, an employer ‘may choose to conduct its RIF according to its preferred criteria of performance ... and we will not disturb that exercise of defendant's business judgment.’” (citing Beaird v. Seagate Tech., Inc., 145 F.3d 1159, 1169 (10th Cir. 1998)). Plaintiffs also complain that training materials provided by Spirit HR “contrasted a longtenured, lazy, and difficult ‘Gus’ with a newly hired, enthusiastic ‘John,’ implying John should be rated more highly than Gus.” (Doc. 1070 at 8.) Aside from the remote likelihood that this particular hypothetical had any material impact on Spirit managers ranking employees, the most that can be 58 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 59 of 83 said of it is that it reiterated Spirit’s stated instructions that performance rather than tenure should be the primary consideration for managers when rating employees.11 Apparently acknowledging that the example itself is age-neutral (it mentions only an employee who is “a new member of your team” and another who has been “in his job for 15 years with five different managers,” Plaintiffs resort to complaining that it employed “classic stereotypes” about older workers. Plaintiffs cite the testimony of Dr. Toni S. Locklear, an Industrial/Organizational Psychologist, who mentions such stereotypes in offering opinions on “the validity, reliability, and fairness of the 2013 retention exercise, including the 2012 performance exercise, used to identify employees for layoff.” (Doc. 1036-2 at 2.) Dr. Locklear concludes: [T]he decision-making processes associated with Spirit’s July 2013 layoff were inconsistent with characteristics of successful downsizing effort. The flaws detailed [in the report] argue against the validity of the retention exercise used to identify SPEEA employees for termination, as well as the performance ratings that informed the retention exercise. Moreover, the lack of reliability in the criteria used to make retention ratings, and the Company’s focus on age and tenure, could have opened the door to biased judgments such as the consideration of legally protected characteristics. (Id. at 33.) The court grants Defendants’ motion (Doc. 1036) to exclude Dr. Locklear’s testimony under Daubert. As helpful as such analysis might be to organizations seeking guidance, the opinions offered here are not based on a sufficiently reliable method to warrant their admission at 11 Plaintiffs and their experts grossly exaggerate the implications and potential effect of this hypothetical. Several of their experts, including their economist Dr. Kaufman, single it out because “[t]hese case studies appear to indicate that tenure should be considered in developing performance ratings,” and “[t]enure is typically correlated with age.” (Doc. 1042-2 at 12.) The suggestion that this single example in training materials somehow green-lighted hundreds of Spirit managers into downgrading the ratings of older workers whose performance was equal to or better than that of younger workers is based on rank speculation. The example makes clear that the one worker with more tenure not only “hates change” and has a “negative attitude,” but also “his work product is not high quality,” he misses deadlines, doesn’t work well with his peers, and he won’t help other workers. The example clearly describes a poor worker who should, in contrast to “a new member of your team” whose “work product is always high quality and ahead of schedule” and who otherwise is an ideal employee, be rated as a lower-tier employee. Again, at most the example conveyed a message that tenure alone should not boost ratings of employees if their performance is poor. This was consistent with Spirit’s entirely legitimate emphasis on the primacy of job performance in employee ratings. 59 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 60 of 83 trial, nor would they be helpful to a jury’s understanding of the issues or in deciding the case. “Fairness” is of course a highly subjective standard, and the court is not entirely sure how Locklear’s opinions about the “validity” and “reliability” of the RIF process would assist the jury, but the analysis offered is too subjective to warrant admission in this matter. Whether or not Spirit conducted the RIF in accordance with Locklear’s idea of “best practices” is so tangential to the issues as to be excludable under Fed. R. Evid. 403, as it would likely mislead or confuse the jury in attempting to decide whether a pattern or practice of age discrimination has been established. Plaintiffs also cite expert testimony from Professor Elizabeth Pendo, J.D., who thinks Spirit “did not provide its management personnel with effective training or otherwise take sufficient steps to address the risk of reliance of management on common stereotypes and assumptions related to age, disability, or both,” and that this contributed to “a significant risk of discriminatory decision-making, especially in the absence of effective training .….” (Doc. 1037-2 at 5.) Professor Pendo’s report concedes that Spirit trained all employees including managers against prohibited discrimination, including age discrimination, but believes they could have done more training about avoiding common stereotypes and assumptions about older workers, and she invokes the ubiquitous John and Gus as having fostered negative stereotypes about older workers. (Id. at 12.) She also faults Spirit for using subjective evaluation criteria such as “leadership,” “dignity and respect for all,” and “teamwork,” because subjective criteria “should be coupled with training to reduce the risk that managers rely on common but impermissible stereotypes and assumptions about older workers when making subjective judgments,” and “I did not see any evidence of such training….” (Id. at 12-13.) It is undoubtedly true to say that Spirit could have provided more and better training in almost any area, including on age discrimination. But this sort of negligence analysis, aside from employing speculation about the impact of not giving more instruction on 60 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 61 of 83 stereotypes, would be unlikely to assist the jury in its evaluation of Spirit’s actions. It would inject possible confusion of the issues by asking the jury to “grade” Spirit’s thoroughness in training instead of focusing on whether a policy of engaging in age discrimination actually existed. For these reasons, Daubert would likely require exclusion of Pendo’s opinions, but even taking them at face value, the assertion that Spirit’s failure to provide more stereotype training created a “risk” of discriminatory results provides no reasonable support for a conclusion that Spirit in fact had a “standard operating procedure” of engaging in age discrimination. Cf. Apsley, 691 F.3d at 1205– 06 (“Our role is not to determine whether the Companies’ hiring process could have been better, but only whether a jury could discern from the evidence a pattern or practice of intentional age discrimination.”) (citing EEOC v. Picture People, Inc., 684 F.3d 981, 989 (10th Cir.2012) (“[A] court should not ‘act as a super personnel department that second guesses employers’ business judgments.’”)). Thus, even if Pendo’s opinions are considered, they provide no material support for a finding that Spirit engaged in a regular pattern or practice of age discrimination. Plaintiffs also seek to introduce opinions from Pendo that Spirit’s decision not to buy stoploss insurance “is consistent with Plaintiffs’ contention that in July 2013 Spirit attempted to mitigate and reduce its exposure to high health claims by terminating disproportionate numbers of older workers,” and with Plaintiffs’ contention that Spirit’s failure to rehire Plaintiffs “reflected a company priority not to employ older workers the company assumed to have high health costs.” (Doc. 1037-2 at 5.) The court grants Defendants’ motion to exclude these opinions under Daubert. Plaintiffs essentially want to have a law professor second-guess the business judgment of Spirit that it was more cost-effective to forego stoploss insurance than it was to purchase it. Stoploss insurance may mitigate the risk of incurring large health expenses, as Pendo emphasizes, but it obviously comes with a cost tradeoff in the form of a significant premium, and the failure to 61 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 62 of 83 purchase it is hardly evidence of a plan to fire older workers. According to Pendo’s own report, industry data showed that for companies of Spirit’s size with self-funded plans, only slightly more than half elected to purchase stoploss insurance. (Doc. 1037-2 at 22.) Pendo indicates, however, that Spirit’s choice is indicative of a plan to discriminate against older workers because the company extensively examined health care costs and had already undertaken some steps to reduce them, so a remaining way to reduce the risk of high costs would be to “reduce the number of covered individuals who have (or are believed to have) high health care costs,” which “is consistent with” Plaintiffs’ allegation that Spirit targeted older workers. This analysis is highly speculative and does not rest on an objective or demonstrably reliable basis. It basically infers a discriminatory purpose from the fact that the level of health care costs could be reduced by unlawfully terminating older workers, while ignoring economic disincentives that weigh heavily against such a path, including potentially enormous legal costs and the loss in quality of products from dismissing highly qualified workers because of their age. In fact, Plaintiffs’ elaborate theory that Spirit’s “obsessive” focus on health care costs drove its decision to conduct a RIF in the first place, that it drove Spirit’s choice of which workers would be let go, and led to the subsequent failure to rehire Plaintiffs, is built largely on speculation about this supposed economic incentive pertaining to older workers. But the evidence about Spirit’s collection and examination of aggregate worker data, including health care costs, shows no demonstrable connection to the rating and selection of workers for the RIF, or to the failure to rehire former employees after the RIF. “[G]ross statistical disparities ... alone may in a proper case constitute prima facie proof of a pattern or practice of discrimination.” Apsley, 691 F.3d at 1195 (quoting Hazelwood Sch. Dist. v. United States, 433 U.S. 299, 307–08 (1977)). See also Hazelwood, 433 U.S. at 308 n. 14 (citing disparities of 5 and 6 standard deviations.) On the statistical front, Plaintiffs cite the opinions of 62 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 63 of 83 Dr. Kaufman that Spirit’s 2012 PM ratings and 2013 retention rankings both resulted in statistically significant age disparities disfavoring older workers. (Doc. 1070 at 39.) Kaufman said his analysis “suggest[s] that the presence of age disparity in termination begins with disparately low performance ratings, is magnified by disparities in retention rankings, then flows through to termination outcomes.” (Doc. 1042-3 at 18.) The opinions offered by Kaufman, however, are not sufficient, either alone or in combination with the other evidence cited, to show gross statistical disparities that could establish a pattern or practice of age discrimination. An initial concern with Dr. Kaufman’s analysis was his use of all Wichita Spirit employees on the OWBPA disclosure list, both SPEEA and non-SPEEA workers, in measuring age disparities in the RIF. Using a total population of 3,273 workers on Spirit’s OWBPA Disclosure List, of whom 271 were selected for termination in the RIF, Kaufman performed various analyses and, using a regression analysis, ultimately concluded the data showed a statistically significant disparity of 3.03 standard deviations (SD) in terminations of older workers on the list. (Doc. 10422 at 23.) Although it is true that all of the employees on that disclosure list were in fact considered for the layoff, which would ordinarily make the entire pool the relevant comparison, an integral part of the justification for allowing Plaintiffs’ claims to proceed collectively in this case was their alleged common status and treatment as SPEEA members,12 including their assertion that Spirit rigged the 2013 retention exercise (applicable only to SPEEA workers) to disadvantage older SPEEA workers. The relevant population for making inferences about the collective termination claims as actually asserted, then, appears to be only SPEEA members on the OWBPA list, exclusive of the more than 1,000 non-union workers who were not subject to the same retention 12 See Doc. 1034 at 46 n. 6 (citing Doc. 522) (“[T]he Named Plaintiffs assert a termination claim for themselves and other former Wichita-based Spirit salaried workers, age 40 and older, who were represented by SPEEA and who were terminated in July 2013.”) 63 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 64 of 83 exercise and/or the same rating standards as SPEEA members. When Dr. Kaufman was challenged by Dr. White on this point, Kaufman asserted that White took an “overly narrow view” of relevance. (Doc. 1042-3 at 5.) Kaufman stated it had “been represented to me that the OWBPA List is legally relevant” and it was “statistically relevant” because “my findings do not materially change when I limit my analysis from the OWBPA List to the SPEEA employees identified as relevant by Dr. White.” (Id.) Using only the population of SPEEA workers as suggested by White, Kaufman performed a regression analysis with controls for Covered Leave and Discipline, that concluded there was a disparity of 2.05 standard deviations in the proportion of older SPEEA workers terminated in the RIF versus younger SPEEA workers. (Doc. 1042-4 at 10.) The Supreme Court has said, “in the context of grand juror selection, that ‘[a]s a general rule …, if the difference between the expected value and the observed number is greater than two or three standard deviations, then the hypothesis that the jury drawing was random would be suspect to a social scientist.’” Apsley, 691 F.3d at 1198 (quoting Castaneda v. Partida, 430 U.S. 482, 496 n.17 (1977). But “[s]tatistics must always be evaluated in the context of all of the surrounding facts and circumstances.’” Id. at 1195 (quoting Teamsters, 431 U.S. at 340). In connection with their termination claims, Plaintiffs rely on the statistical analysis provided by Dr. Kaufman. Defendants have moved under Daubert to exclude Kaufman’s opinions. (Doc. 1042.) The Daubert motion raises substantial questions about the methodology used by Dr. Kaufman and whether it is closely tied to the facts of the case. The court notes Dr. Kaufman acknowledged in his report that “Spirit instructed managers to rank employees utilizing the following assessment criteria: performance rating, versatility, and criticality of skills.” (Doc. 10422 at 13.) Kaufman nowhere addressed versatility or criticality of skills in his reports – apparently due to their subjectivity or the inability to statistically control for them – or how they may have 64 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 65 of 83 impacted retention rankings or the selection process for the RIF. Whether subjective or not, such facially neutral factors related to selection for the RIF could, absent explanation, account for a level of age disparity in RIF selection. The uncontroverted facts also show that in the 2013 retention rating, employees were rated against other workers within their specific job codes (and exempt or non-exempt status) or skill management codes, not as part of an overall pool of employees. Although Dr. White examined those particular pools and found no significant age disparities, Dr. Kaufman did not do so because “pooling on job code results in 151 pools,” an average of less than 2 terminations per pool, and this “fragments the data” and results in a “very weak test” that has a low chance of identifying disparities. (Doc. 1042-3 at 9.) While that may be true as a matter of statistical analysis, the result is that the test used by Dr. Kaufman was not closely tied to the actual manner in which the retention exercise was conducted. Dr. Kaufman also extensively injected the concept of “Covered Leave” into his analysis, a term he constructed as a proxy for employee medical costs covered by Spirit. This variable was itself based on a rather complex set of assumptions and inferences and its purpose and place in the statistical analysis of age disparities is unclear.13 It is apparently intended to suggest that a correlation between taking covered medical leave and selection for the RIF somehow reinforces an inference that older workers were targeted in the RIF. What is clear, however, is that the concept of Covered Leave, or alternatively the actual medical costs incurred by SPEEA workers, played 13 Responding to White’s assertion that Kaufman improperly assumed that Spirit incurred medical costs for employees who took Covered Leave, Kaufman asserted: “It is reasonable to assume that a share of leave is used for medical reasons, and it is reasonable to expect that employees with a need for medical leave also insure [sic] medical costs. As a self-insured firm, it is reasonable to expect that some of these costs would pass directly on to Spirit though [sic] Spirit’s insurance coverages. Spirit was concerned about high medical costs related to age and certain medical conditions. … While reasonable, this assumption is not needed to support my findings. My report finding is that there is a ranking and termination disparity with respect to covered leave, not with respect to health care costs. However, I do note that this disparity is not present for non-covered leave, and that this is consistent with an interpretation that medical expense was a factor. This secondary finding of consistency does not require an assumption that covered leave resulted in medical costs to Spirit.” (Doc. 1042-3 at 16.) 65 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 66 of 83 no actual part in the retention rankings given by individual Spirit managers or in the RIF selections at issue in this case. Kaufman measured certain age disparities relating to 2012 PM ratings and 2013 retention ratings but stated he did not do so for the purpose of connecting these disparities to the RIF. (Doc. 1042-3 at 13.) Rather, he said the purpose was to “demonstrate that age disparities exist in the retention ranking process that is incremental to the disparity present in the performance management process.” (Id. at 6.) Absent a connection to termination decisions, however, the relevance of these other asserted disparities is not clear. He also limited his inquiries in significant aspects. For example, on PM ratings he measured age disparities only for employees who had rating reductions from 2010 to 2012 (and did not measure overall age disparities in the assignment of ratings) and also excluded the impact of 2010 PM ratings, where older workers may have had (and apparently did have) significant rating disparities in their favor. On the 2013 retention ratings, he measured disparities in C rankings that were assigned only to “Low-Rated” employees (i.e., Meets Some, Unacceptable), excluding a number of employees who were given Cs despite having PM ratings in the middle category (i.e., Meets). He asserted that he did so because Spirit’s documents showed it was seeking to identify low rated employees, and “[t]his leads to a reasonable conclusion that Spirit’s termination practices for low rated employees differ from Spirit’s termination practices for employees assigned a ‘Meets’ performance rating,” which is “borne out by differences in termination rates across rating scores,” and as such it was “necessary to control for performance rating when evaluating terminations.” (Doc. 1042-3 at 14.) Thus, he said, the “Meets” category was left out to control for performance rating when evaluating terminations. (Id.) But unlike performance ratings, retention ratings and selection for the RIF took into account factors other than performance that were not controlled by this analysis. So, an unidentified number of 66 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 67 of 83 workers may have been selected for the RIF although their performance was fair (i.e., “Meets”) because their skills were not deemed critical or they were less versatile than others in the group. Dr. White’s analysis asserted that if age differences were examined looking at all three performance categories from which terminations actually occurred, including Meets, there was no significant age disparity. Whether or not these methodological issues warrant exclusion under Daubert, they weaken the significance of Dr. Kaufman’s findings that there were age disparities in the RIF. Leaving aside resolution of any Daubert issues concerning Dr. Kaufman’s analysis, the court concludes that Dr. Kaufman’s opinions, under all of the circumstances and considered together with Plaintiffs’ other evidence, do not reasonably show that Spirit had a regular procedure or policy of age discrimination. A disparity of 2.05 standard deviations in this context, or even the 3.03 standard deviations identified by Kaufman in his initial report, is not, under all of the circumstances, a gross disparity of the sort that would reasonably support a finding that Spirit employed a pattern or practice of age discrimination. In Apsley, for instance, the Tenth Circuit rejected a pattern or practice claim despite expert testimony about disparities of more than 5 standard deviations in the number of older workers predicted to be recommended for hire versus the number actually recommended. Id. Here, Dr. Kaufman expressed termination disparities in terms of standard deviations, but the bulk of his analysis appears tangential or even irrelevant to the termination issue and his reports do not explain or make clear the actual number of older workers he believes would have been included in the RIF had the selection been random versus how many actually were selected. Dr. White, using the SPEEA workers on the OWBPA list and controlling for skill management code and job code, determined that about 178 workers age 40 or over would be expected to be selected for the 67 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 68 of 83 RIF if it were a neutral selection free of age discrimination, while 185 were in fact selected, a difference of less than 8 persons, which he concluded was not statistically significant (1.32 SD). (Doc. 1041-120 at 8.) This was based on 185 of 1,832 older SPEEA workers on the list being selected for the RIF and 36 of 464 younger SPEEA workers being selected. His number changed only slightly (to a difference of about 11 persons (1.80 SD)) when the list of all OWBPA workers was considered. (Id. at 10.) Dr. Kaufman applied a different analysis, of course, included different controls, and as indicated above ultimately found a variance of about 2 or perhaps 3 standard deviations from a neutral selection. As indicated, Kaufman did not discuss the predicted versus actual number of persons selected for the RIF, instead discussing the “odds ratio” of older versus younger workers being selected, the number of standard deviations involved, and whether the finding was statistically significant. (See Doc. 1042-2 at 12.) But regardless, it is clear that the absolute difference between the number of older workers who were actually selected in the RIF and the number that would have been expected in a neutral selection is quite small, perhaps as few as 7-11 persons out of the 1,832 SPEEA workers considered for the RIF and of 221 who were ultimately selected. When considered in the context of the ranking process actually used by Spirit, which involved rating decisions by a multitude of different managers, and the unmeasured impact of rating criteria such as criticality and versatility, the disparities identified by Plaintiffs’ expert are not the type of gross disparities that reasonably show Spirit had in place a pattern or practice of age discrimination within the meaning of Teamsters. The uncontroverted facts show it was rumored that Spirit was going to target older and sicker workers in the RIF. Plaintiffs have constructed a narrative around that theory based largely on an expert opinion showing at most a moderate disparity in the proportion of older versus younger workers selected for the RIF, and an assortment of corporate cost assessments and 68 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 69 of 83 projections relating to employees. Plaintiffs claim a “standard operating procedure” of discrimination at Spirit when, out of a workforce numbering in the thousands, Spirit ultimately included perhaps 11 more older workers than would have been predicted from a neutral selection. The practical insignificance of this type of disparity is due in part to the unknown impact of rating criteria such as criticality and versatility – which managers were specifically told to factor into their retention ratings – that could easily account for a modest disparity. A number of individuals selected for the RIF may have been equal to others in performance but were thought by their managers to lack a broad range of skills that made other employees slightly more valuable to the group. Other variables not assessed by Kaufman included the particular retention groups into which employees fell and differences in the managers who assigned ratings. Such variables could account for the modest statistical variations shown by Plaintiffs. Cf. Rea v. Martin Marietta Corp., 29 F.3d 1450, 1456 (10th Cir.1994) (“[F]or statistical evidence to create an inference of discrimination, the statistics must ... eliminate nondiscriminatory explanations for the disparity.”) (citation omitted.) Aside from this statistical evidence, there is a complete lack of evidence showing that Spirit has any history, recent or otherwise, of engaging in age discrimination against its workers. No individuals acts of age discrimination, let alone a pattern of such conduct, is cited in the evidence. Based on the uncontroverted facts, Defendants are entitled to summary judgment on Plaintiffs’ collective claims that their terminations were attributable to a pattern and practice of intentional age discrimination at Spirit. 2. Termination Claim - Disparate Impact. Plaintiffs’ disparate impact termination claim, according to their summary judgment response, is premised upon Spirit’s “adoption of a bell curve for PM ratings.” (Doc. 1070 at 55.) This policy, which was newly adopted or enforced by Spirit in connection with its 2012 PM ratings 69 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 70 of 83 process, forced managers to rate performance across five specified levels (Exceptional, Exceeds, Meets, Meets Some, Unacceptable) within a never-before-imposed 15%-70%-15% distribution. This required that a total of 15% of employees be rated in the lowest two categories. Plaintiffs contend Dr. Kaufman’s analysis shows the policy had a disparate impact on older workers, because older workers were more likely to be downgraded from the 2010 PM ratings to the 2012 PM ratings. Plaintiffs assert Dr. Kaufman further found lower-ranked older employees were twice as likely to receive a C retention rating (as compared to younger lower-ranked workers), and “that the resulting C rankings were directly correlated with layoff decisions.” (Doc. 1070 at 56.) According to Plaintiffs, “[t]his proof that a specific piece of the PM process and a specific 2013 RIF component cause an age-disparate impact satisfies Plaintiffs’ prima facie case.” (Id.) The Tenth Circuit outlined the framework for disparate impact ADEA claims in Fulghum v. Embarq Corp., 785 F.3d 395 (10th Cir. 2015): Disparate impact claims are grounded in the premise that “some employment practices, adopted without a deliberately discriminatory motive, may in operation be functionally equivalent to intentional discrimination.” Watson v. Fort Worth Bank & Trust, 487 U.S. 977, 987, 108 S.Ct. 2777, 101 L.Ed.2d 827 (1988). Accordingly, “a claim for disparate impact [does not] require proof of intentional discrimination.” Cinnamon Hills Youth Crisis Ctr., Inc. v. Saint George City, 685 F.3d 917, 922 (10th Cir.2012). A plaintiff asserting a claim of disparate impact discrimination can make out a prima facie case by demonstrating the challenged employment practice caused a disparate impact on the protected group. Tabor v. Hilti, Inc., 703 F.3d 1206, 1220 (10th Cir.2013). “Statistical evidence is an acceptable, and common, means of proving disparate impact.” Id. at 1222 (quotation omitted). The framework applied to ADEA disparate impact claims differs from that applied to Title VII disparate impact claims because the “scope of disparate-impact liability under ADEA is narrower than under Title VII.” Smith, 544 U.S. at 240, 125 S.Ct. 1536. This is so because the ADEA “contains language that significantly narrows its coverage by permitting any ‘otherwise prohibited’ action ‘where the differentiation is based on reasonable factors other than age.’” Id. at 233, 125 S.Ct. 1536 (quoting the ADEA). Thus, although a Title VII defendant has the burden of producing evidence of a “business necessity” for the challenged employment practice, an ADEA disparate-impact defendant need only produce evidence the practice is based on “reasonable factors other than age” (“RFOA”). Id. at 241–43, 70 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 71 of 83 125 S.Ct. 1536; see also id. at 238–39, 125 S.Ct. 1536 (noting the RFOA provision is inapplicable when an ADEA plaintiff proceeds under a disparate treatment theory). “Unlike the business necessity test, which asks whether there are other ways for the employer to achieve its goals that do not result in a disparate impact on a protected class, the reasonableness inquiry includes no such requirement.” Id. at 243, 125 S.Ct. 1536. At trial, the ADEA defendant must persuade the factfinder its reasonableness “defense is meritorious.” Meacham v. Knolls Atomic Power Lab., Inc., 554 U.S. 84, 101, 128 S.Ct. 2395, 171 L.Ed.2d 283 (2008). (Id. at 418.) As Defendants point out, it was not at all clear prior to the filing of Plaintiffs’ summary judgment response that their disparate impact claim was premised on Spirit’s adoption of a bell curve distribution policy for the 2012 PM ratings.14 Plaintiffs now contend that the adoption of this policy had a disparate impact on older SPEEA workers being selected for termination in the RIF because Dr. Kaufman’s analyses “suggest that the presence of age disparity in termination begins with disparately low performance ratings, is magnified by disparities in retention rankings, then flows through to termination outcomes.” (Doc. 1042-3 at 18.) This “daisy chain” theory of causation, as Defendants describe it, raises potential hurdles, including Defendants’ argument that Plaintiffs failed to timely exhaust this particular claim before the EEOC, as well as whether Plaintiffs’ proof has isolated the bell curve requirement in 2012 PM ratings as the cause of disparity in terminations, as opposed to other factors such as the mandatory distribution of retention rankings under the CBAs. After considering the nature of Plaintiffs’ claim in light of the evidence and arguments presented, the court concludes that this claim fails as a matter of law for two reasons. First, even assuming Plaintiffs have timely exhausted the claim, Plaintiffs have failed to cite evidence isolating the mandatory distribution of 2012 PM ratings as the cause of the alleged disparity in 14 Neither the First Amendment Complaint nor the pretrial order identified the bell curve distribution policy as the basis for the ADEA collective disparate impact claim. 71 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 72 of 83 dismissal of older workers in the RIF. Second, Defendants are entitled to summary judgment because the record shows that Spirit’s use of a bell curve distribution in performance evaluations was a “reasonable factor other than age.” “To establish a prima facie case of disparate impact discrimination on the basis of age, plaintiff must show that a specific identifiable employment practice or policy caused a significant disparate impact on the protected group.” Leo v. Garmin Int'l, No. 09-CV-2139-KHV, 2009 WL 3122502, at *4 (D. Kan. Sept. 24, 2009) (citing Carpenter v. Boeing Co., 456 F.3d 1183, 1187 (10th Cir.2006)). “Thus, an employee must point to both a significant disparate impact and a particular policy or practice that caused the disparity.” Id. (citing Pippin Burlington Res. Oil & Gas, 440 F.3d 1186, 1200 (employee must isolate and identify specific employment practices allegedly responsible for observed statistical disparities)). See also Smith v. City of Jackson, Miss., 544 U.S. 228, 241 (2005) (in disparate impact case, the employee is “‘responsible for isolating and identifying the specific employment practices that are allegedly responsible for any observed statistical disparities.’”) (citation omitted.) Plaintiffs’ reliance on Dr. Kaufman’s analysis fails to satisfy their burden of isolating and showing that the bell curve distribution required in 2012 PM ratings was the cause of statistical disparities in the RIF. As an initial matter, Dr. Kaufman’s identification of a statistical disparity in reduced PM ratings for older workers between 2010 and 2012 does not appear to specifically identify use of the 15-70-15 mandatory distribution as the specific cause of any disparity. Moreover, it was the 2013 retention exercise that actually identified the C-rated workers selected in the RIF, not the 2012 PM ratings, and it is uncontroverted that three factors were considered by managers in assigning those retention rankings: performance (both 2012 PM ratings and 2013 performance), versatility, and criticality. As indicated previously, Dr. Kaufman did not consider 72 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 73 of 83 or address any disparity in terminations caused by consideration of 2013 performance, by versatility, or by criticality. His opinion that age disparities in 2012 PM ratings changes “flow[ed] through” to termination outcomes does not attempt to specifically isolate the impact of the bell curve distribution in 2012 PM ratings on terminations.15 Pippin, 440 F.3d at 1200 (“it is not enough to simply allege that there is a disparate impact on workers, or point to a generalized policy that leads to such an impact.”) (citation omitted); Meacham, 554 U.S. 84, 100–01 (2008) (plaintiff must isolate and identify the specific employment practice responsible for any observed statistical disparity, otherwise employers could be liable for “the myriad of innocent causes that may lead to statistical imbalances.”) (quoting Smith, 544 U.S. at 241) (citation omitted.) Plaintiffs have failed to cite evidence from which a jury could reasonably conclude, without resorting to speculation, that adoption of a mandatory 15-70-15 distribution for 2012 PM ratings caused a disparate impact in the selection of workers for the RIF. Even if Plaintiffs had made a prima facie showing of causation, the record shows beyond reasonable dispute that requiring a bell curve distribution in performance ratings was a reasonable factor other than age. Spirit has accordingly shown it is entitled to judgment based on that factor. See Meacham v. Knolls Atomic Power Lab., 554 U.S. 84 (2008) (RFOA is affirmative defense that defendant must establish). As noted above, it is not unlawful for an employer to adopt a policy with a disparate impact “where the differentiation is based on reasonable factors other than age….” 29 U.S.C. § 623(f)(1). Plaintiffs argue Defendants have not shown that its use of a mandatory distribution in PM ratings was reasonable because “Spirit fails to identify a specific business purpose” for the policy, fails to explain how it relates to a goal of retaining its best workers, fails to show it provided sufficient training on how to rate performance rather than relying on age-based 15 As Defendants point out, about 85% of employees who experienced lowered PM ratings in 2012 did not have a low PM or C retention rating. (See Doc. 1041 at 12-13.) 73 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 74 of 83 stereotypes, it “heard from managers and HR personnel regarding concerns about impacts of the new PM policy,” and it fails “to produce any evidence showing it assessed the impact of the bell curve based on age or took any steps to reduce the harm it caused to older workers.” (Doc. 1070 at 57-58.) Spirit has cited evidence that at times in the past, including in 2010, its PM ratings suffered from “grade inflation,” with a large percentage of workers rated in the top categories. Spirit cites evidence that it adopted the policy of requiring a 15/70/15 distribution in PM ratings to force managers to identify lower-performing employees. HR training materials provided to managers for the 2012 PM ratings emphasized the need to accurately rate performance and specifically identified common mistakes such as grade inflation, “nice guy effect” (rating on personality rather than performance), and “central tendency” (rating people in the middle because it is easier than identifying the top and bottom performers). (Doc. 1080-8 at 14-18.) The adoption of a mandatory ranking distribution under these circumstances was entirely reasonable. Ranking employees based on their relative job performance is a legitimate business objective and requiring a mandatory distribution of ratings to differentiate between top, middle, and lower performers is clearly a reasonable method of furthering that objective. Cf. Pippin, 440 F.3d at 1201 (“Corporate restructuring, performance-based evaluations, retention decisions based on needed skills, and recruiting concerns are all reasonable business considerations.”) Moreover, in the context of Spirit’s efforts to reduce headcount at its Wichita facility as a way to control costs, Spirit had a legitimate business imperative in identifying its lower-performing workers, and the mandatory distribution requirement for PM ratings furthered that objective. Courts have recognized that “relying on prior performance ratings and the determination of which employees have the skills most useful to the company going forward are reasonable criteria for any company to use in 74 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 75 of 83 deciding which employees to keep and which to let go in a RIF.” Pippin, 440 F.3d at 1201. Contrary to Plaintiffs’ suggestions, the mandatory distribution was reasonable in light of these business objectives and Spirit in fact instructed managers on the need to avoid unlawful discrimination in the assignment of PM ratings. Plaintiffs fault Spirit for failing to produce evidence that it considered the potential disparate impact of the 15/70/15 policy on older workers before adopting it, but even if true, such a fact does not render this policy unreasonable in light of the other factors noted above, including the obvious relation of the policy to a legitimate business objective. Fitzpatrick v. Newmont Mining Corp., No. 13-CV-03409-PAB-CBS, 2015 WL 1809981, at *6 (D. Colo. Apr. 17, 2015) (noting no case authority that the RFOA provision requires an employer to consider the impact of the challenged practice on employees over 40 years of age or that the failure to do so renders a stated rationale unreasonable). The court notes that Plaintiffs offer no alternative method by which Spirit could have ensured that its top, middle, and lower-tier performers were identified in PM ratings. No doubt there were other possibilities available. But under the ADEA, which only requires that Spirit’s chosen method not be based on age and be “reasonable,” the use of such a required distribution in performance rankings had a reasonable basis and therefore was not unlawful. Cf. Smith, 544 U.S. at 243 (“While there may have been other reasonable ways for the City to achieve its goals, the one selected was not unreasonable. Unlike the business necessity test, which asks whether there are other ways for the employer to achieve its goals that do not result in a disparate impact on a protected class, the reasonableness inquiry includes no such requirement.”) Defendants’ motion for summary judgment as to Plaintiffs’ collective termination claim for disparate impact under the ADEA is accordingly GRANTED. 75 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 76 of 83 3. Pattern or Practice Hiring Claim – Disparate Treatment Plaintiffs’ final collective claim alleges that Spirit intentionally discriminated against them in rehiring, on account of age, by deciding before the layoff not to rehire any laid off workers, by tagging laid-off worker applicants as “AP-Hold,” thereby “precluding their advancement through the hiring process,” and by telling laid off applicants to apply for jobs online while withholding the fact their applications would be classified as “AP-Hold.” (Doc. 1034 at 11-12.) Additionally, Plaintiffs contend a statistical analysis shows “the odds of staying in [AP-Hold] was 145 times as high for [the laid off] applicants,” and “applicants 40 and over had only 76% the odds of a successful initial screening as under 40 applicants.” (Id. at 12.) Defendants seek summary judgment on the claim, arguing the uncontroverted facts show that Spirit followed age-neutral hiring policies and processes, and that it has no discriminatory standard operating procedure to reject job candidates based on age. (Doc. 1041 at 52.) With regard to Plaintiffs’ statistical analysis, Spirit contends the analysis of Plaintiff’s expert Dr. Bardwell is inadmissible and without probative value. Even if Bardwell’s analysis is considered, Spirit contends it rebuts the suggestion of a discriminatory pattern or practice with its own analysis showing that Spirit selected older employees at statistically neutral rates. (Id. at 60.) Spirit additionally argues that even if Plaintiffs could establish a pattern or practice of age discrimination, Plaintiffs’ claims would still fail at the second stage of Teamsters, because Spirit can show the Hiring Plaintiffs would not have been hired for a variety of individual reasons. Finally, Spirit contends that some Plaintiffs’ claims have not been properly exhausted. Plaintiffs contend a “reasonable fact-finder could easily conclude that Spirit decided not to rehire the older workers subject to the 2013 RIF based on the same discriminatory animus that led to their initial terminations.” (Doc. 1070 at 48.) Among other things, Plaintiffs point to evidence 76 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 77 of 83 of Spirit’s comments indicating it did not want to rehire workers let go in the RIF, its reliance on the system that screened out former employees and put them in AP-Hold status, and other evidence allegedly showing that Spirit “implemented a policy and practice of age discrimination that its executives worked hard to conceal.” (Id. at 49.) Plaintiffs contend the “failure to hire or even consider Plaintiffs for these open positions was a fundamental part of the plan to ‘replace long tenured employees with new hires,’ the heart of the pattern or practice of discrimination here.” (Doc. 1070 at 50.) They contend their statistical expert has shown that Spirit’s post-RIF hiring “disadvantaged older applicants because a disproportionate share of employees terminated during the July [2013] RIF were age 40 and above.” (Id.) As the court noted previously, at the first stage of a pattern or practice claim a summary judgment motion “must focus solely on whether there is sufficient evidence demonstrating that defendants had in place a pattern or practice of [age] discrimination during the relevant limitations period.” Thiessen, 267 F.3d at 1109. On their claim that Spirit failed to rehire them as part of a pattern or practice of intentional discrimination based on age, Plaintiffs first invoke “the same discriminatory animus” involved in their initial terminations. (Doc. 1070 at 48.) That evidence, however, fails to show a pattern and practice of age discrimination largely for the reasons previously discussed. Plaintiffs have made no showing of a Spirit history of engaging in age discrimination, no evidence that it has used facially discriminatory policies, cited no individual instances of Spirit management engaging in age discrimination, while at the same time citing only a modest statistical disparity in the selection of older SPEEA workers in the RIF, with the selection resulting from a process in which hundreds of individual managers made the decisive rating determinations. 77 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 78 of 83 As evidence of age discrimination animus relating specifically to hiring (or rehiring), Plaintiffs cite a comment by one Spirit representative in February 2013 that, “Everyone we send out through layoff we will not want them back….” (Doc. 1070 at 48.) This stray comment – whose context is unclear16 – hardly shows an intent to discriminate on the basis of age. Along the same lines, Plaintiffs argue that evidence of Spirit’s resistance to rehiring laid off workers is evidence of age-based animus. The record, however, shows that Spirit was reluctant to rehire workers let go in the RIF because they had been identified as being in the lower echelon of performers. In VP Justin Welner’s deposition, for example, Welner explained why Spirit did not prohibit the rehire of employees from the RIF altogether but was nevertheless disinclined to rehire them: Q For people that were terminated in July of 2013, do you know what criteria were used to determine whether or not they were good employees that could be rehired? A So in 2013 when we did the reduction in July/August timeframe, that was the bottom 10 percent from a performance and capability standpoint. We used the process in the collective agreement to go through the retention process. And so that's the -- the process we used. Q So given that they were considered to be the bottom 10 percent, did you not want to rehire those individuals? A So we did not make them ineligible for rehire. In fact, we -- we actually had a conversation about how to apply that and chose not to go down that route. Q Can you explain what you mean by that? A Sure. So if you -- again, I know you're only talking about Wichita. But if you look at our release and severance agreements in Oklahoma, they included a not-eligible-for-rehire clause. We 16 The statement is part of an email discussing an alternative proposal to performance exits, including a layoff, in which the email author alludes to language in the proposal (which is not set forth in the email) and “wonder[s] if we should strengthen the language a bit in the feb. box where we indicate performance exits and all others remaining will be placed on performance plans. There is the possibility before we get through the entire retention process we still exit a few for performance earlier. Everyone we send out through layoff we will not want them back, does the designated candidate limit us at all, I don’t remember that piece of the discussion.” 78 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 79 of 83 chose not to do that in Wichita with this group. And the reason we chose not to do that was these groups are represented by a union, and my concern was if we included a not-eligible-for-rehire clause, that unions would work against us and encourage people not to accept the severance and not to sign the release. Q So that clause was left out in order to ensure that people signed the release of claims for the -- and received severance from the company. A It was included to give us the best chance of having people accept the severance offer and sign the release. Q Okay. But you indicated that because that reduction was the bottom 10 percent, that those are individuals that the company might not want to bring back, correct? A I just said they were the bottom 10 percent. But, you know, in theory, you're correct. It's the bottom 10 percent. We paid a -- a generous severance agreement to exit them. We generally would not be looking to bring them back. (Doc. 1070-34 at 2-3.) (See also Doc. 1041-111)(Spirit HR representative cautioning that before rehiring RIF’d workers, Spirit should perform “due diligence” and “vet the reason they were a designated C and [ensure] there are not past performance concerns,” because the company “paid a lot of money to get them out the door and we should not be passing up an opportunity to ‘upskill’ and get the right talent in the door.”) The evidence thus clearly shows that Spirit was reluctant to rehire Plaintiffs and other SPEEA workers let go in the RIF, but Plaintiffs cite no evidence that this reluctance was attributable to age-based animus, as opposed to arising from performancebased concerns. It is also clear from the uncontroverted facts that Spirit’s policy of applying AP-Hold status to any applications submitted by former employees, including Plaintiffs, was the principal reason their applications, with some ad hoc exceptions, went unreviewed and unprocessed. Plaintiffs do not dispute that the AP-Hold process had been in place since 2009, well before the events at issue 79 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 80 of 83 here, but contend evidence of “a written objective to ‘replace long-tenured employees with new hires,’” in light of subsequent events, provides “powerful evidence of a policy of age discrimination, stemming back to the goals of the July 2013 RIF and continuing through the AP Hold, which played a central role in the discriminatory acts here.” (Doc. 1070 at 52.) This attempt to show that application of the AP-Hold policy to their applications resulted from a plan by Spirit to discriminate against older workers is highly speculative and ultimately fails to meet Plaintiffs’ burden to establish that Spirit engaged in a pattern or practice of age discrimination. Spirit’s written comment about “replacing long-tenured employees with new hires” was, as examination of the underlying document and the portion of the quotation omitted by Plaintiffs shows, made in the context of an October 2012 Spirit HR proposal to reduce costs in part by offering early retirement incentives to Spirit’s employees (company-wide) over the age of 62: “Replace long/service/high cost employees with new hires by offering a Voluntary Separation Package.” (Doc. 1080-7 at 3) (underlining in original). The initiative also discussed several other ways to reduce costs without resorting to a layoff by “aggressively” managing overhead and salaried workforce, noting in an accompanying section that the company was currently “636 Heads over budget with 622 Open Requisitions.” (Id. at 4.) The initiative also discussed aggressively managing performance separations, stating Spirit should have “at minimum 2.2% or 138 employees not meeting expectations [on PM ratings] based on the standard, normal distribution,” while “most performance models would indicate we should have at least 5% (314EEs) with an additional 15% (942EEs) only partially meeting expectations.” (Id. at 2.) Against that background, the initiative noted the company had opened only 89 performance coaching plans, generating only “38 terminations including retirements,” which had affected only 0.06% of employees. (Id.) The failure of the plan to use retirement incentives and performance separations to reduce headcount, of 80 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 81 of 83 course, led to Spirit’s decision to conduct a RIF. But its contemplation of replacing long-service employees with new hires in the context of a voluntary retirement incentive, as indicated above, provides no evidence of an intent to unlawfully discriminate on the basis of age. Plaintiffs’ attempt to spin this into evidence that Spirit had an ongoing pattern or practice of engaging in age discrimination rests more on speculation than evidence. On their termination claim, Plaintiffs failed to show a causal connection between Spirit’s purported plan to discriminate and the actual selection of employees for the RIF through the ratings of individual managers. Here, Plaintiffs fail to demonstrate any discriminatory intent on Spirit’s part in its application of APHold status to job applications submitted by any former employees, including Plaintiffs. The policy itself is facially age-neutral and no evidence is cited that Spirit applied it to Plaintiffs as a result of age-based animus. Spirit has cited evidence that the policy of putting terminated employees on a hold status was related to legitimate business objectives, both because it ensured that any previously-terminated applicants would be vetted before being rehired and, as Welner noted, it furthered the company’s interest in obtaining high-performing workers, because the persons let go in the RIF had been identified as the bottom ten percent of performers. Noting Spirit’s assertion that it made logistical sense to apply this policy to former employees, Plaintiffs argue this assertion is “implausible” and forms a fact question that should be resolved by a jury. (Doc. 1070 at 53.) But it is Plaintiffs’ burden initially to establish that engaging in age discrimination was Spirit’s “standard operating procedure,” and Plaintiffs have failed to cite any evidence showing a plan or practice of discrimination by Spirit generally, or any evidence specifically showing that application of the AP-Hold policy to RIF’d employees was the product of some discriminatory intent or practice. 81 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 82 of 83 Against this dearth of other evidence, Plaintiffs put forth the analysis of a statistician, Dr. Robert A. Bardwell, Ph.D., who was “retained … to evaluate the applications for positions with Spirit submitted by employees separated in the July RIF.” (Doc. 1046-2 at 3.) Plaintiffs point out that Dr. Bardwell concluded Plaintiffs were “disadvantaged by being stranded in the Applicant Held Pool” resulting in statistically significant reduced chances of “success in Spirit’s hiring process” and that Spirit’s post-RIF hiring “also disadvantaged older applicants because a disproportionate share of employees terminated during [the RIF] were age 40 and above.” (Doc. 1070 at 50) (quoting Doc. 1046-2 at 4). It is undisputed on the record that for some years prior to 2013, every Spirit applicant had to complete an online application and that any applicant identifying as a “former employee” was designated as being in AP-Hold status. That designation meant the application was on hold unless it was manually reviewed, such as when a manager requested help filling a particular requisition. There is no evidence this policy was adopted due to discriminatory animus or that it was applied with a discriminatory purpose to Plaintiffs. It was obviously facially neutral; it applied to all SPEEA workers who were terminated in the RIF, young and old alike. Dr. Bardwell’s analysis concluded, and it is undoubtedly true, that being placed in AP-Hold status disadvantaged Plaintiffs in hiring relative to other applicants who were not former employees and whose applications were not stayed in this manner. (See Doc. 1046-2 at 14) (“Spirit employees terminated in the July RIF were not provided the same opportunity to be rehired as other Spirit employees or outside applicants,” and this “disparity resulted from placing July Riffed employees disproportionately in the Applicant Hold Pool….”) But Dr. Bardwell’s analysis, if not entirely “irrelevant” as argued by Defendants, is insufficient to permit any inference that Spirit had a pattern or practice of age discrimination. The adoption of an age-neutral policy prior to the contemplation of the RIF, which 82 Case 6:16-cv-01282-JWB Document 1115 Filed 05/16/23 Page 83 of 83 was applied to all former Spirit employees who left the company for whatever reason, and which allegedly had an adverse impact on older workers only because they were let go in the RIF in a somewhat greater proportion than younger workers, does not reasonably suggest that Spirit had a policy and practice of intentional age discrimination in hiring. And for reasons previously discussed, the statistical analysis in combination with Plaintiffs’ other evidence likewise fails to establish the prima facie case of discrimination required by Teamsters. IV. Conclusion Plaintiffs’ motion for final certification of a collective action (Doc. 1035) is GRANTED; Defendants’ motion to strike the collective claims and to decertify (Doc. 1044) is DENIED. Defendants’ motion to exclude testimony of Dr. Toni Locklear (Doc. 1036) is GRANTED. Defendants’ motion to exclude testimony of Elizabeth Pendo (Doc. 1037) is GRANTED IN PART and DENIED IN PART. Defendants’ motion to exclude testimony of Dr. Lance Kaufman (Doc. 1042) is DENIED AS MOOT. Defendants’ motion to exclude testimony of Dr. Kevin Cahill (Doc. 1043) is GRANTED and its motion to strike (Doc. 1090) is GRANTED. Defendants’ motion to exclude testimony of Dr. Robert Bardwell (Doc. 1046) is DENIED AS MOOT. Defendants’ motion for partial summary judgment (Doc. 1040) is GRANTED. Plaintiffs’ collective ADEA claims are DISMISSED on the merits. Plaintiffs’ individual claims are not at issue in this ruling and remain pending. IT IS SO ORDERED this 16th day of May, 2023. _____s/ John W. Broomes__________ JOHN W. BROOMES UNITED STATES DISTRICT JUDGE 83

Some case metadata and case summaries were written with the help of AI, which can produce inaccuracies. You should read the full case before relying on it for legal research purposes.

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.