In the
United States Court of Appeals For the Seventh Circuit ____________________ No. 24-2823 KARA MITCHELL, Plaintiff-Appellant, v.
EXXON MOBIL CORPORATION, Defendant-Appellee. ____________________
Appeal from the United States District Court for the Northern District of Illinois, Eastern Division. No. 1:21-cv-06876 — Elaine E. Bucklo, Judge. ____________________
ARGUED MAY 13, 2025 — DECIDED JULY 14, 2025 ____________________
Before EASTERBROOK, BRENNAN, and PRYOR, Circuit Judges. BRENNAN, Circuit Judge. Kara Mitchell worked as a labora- tory technician for Exxon Mobil Corporation for a little over a year before she was terminated in 2020. ExxonMobil says it fired her because she compared unfavorably to peers in the company’s annual employee assessment process. Mitchell says ExxonMobil fired her because she is a woman. 2 No. 24-2823
She sued the company for sex discrimination under Title VII of the Civil Rights Act of 1964. The district court granted summary judgment to ExxonMobil because Mitchell failed to provide sufficient evidence for her claim. We agree and af- firm. I. Background A. ExxonMobil’s Employee Assessment Process ExxonMobil employs tens of thousands of employees worldwide, including about 20,000 in the United States. Every year, the company conducts a complex, multi-step perfor- mance assessment to evaluate the achievements and progress of its large workforce. This process, described in an internal company manual available to employees, is at the center of Mitchell’s challenge. At the beginning of every performance assessment period, employees complete a one-page summary of their accom- plishments, strengths, and development opportunities during the previous 12 months. They also select co-workers to evalu- ate their performance. Supervisors rely on the self-assess- ment, feedback from co-workers, and their own judgment to assign preliminary assessment categories to employees. These categories may be adjusted up or down at a conclusive assess- ment group meeting. Assessment groups comprise employees in similar roles with similar experiences. Nevertheless, employees in the group may have different job titles, report to different super- visors, and work in different locations across the country. During the group meeting, each employee is represented by a member of management, usually the employee’s direct super- visor. ExxonMobil relies on this relative assessment process No. 24-2823 3
because comparing employees with their peers “meets [] prin- ciples of meritocracy, performance differentiation, and con- tinuous improvement.” The assessment process differs based on an employee’s re- sponsibilities. Mitchell, an “operational, clerical, and admin- istrative” employee, was assessed by a process that assigns performance categories A, B, C, or D to employees in the as- sessment group. 1 Guidelines set quotas for the number of em- ployees who can be assigned each category. Per the manual, this ensures that all assessment groups “will meet a standard distribution, and each group will stand on its own.” Once supervisors reach consensus about employees’ rela- tive performance and assign final performance categories that accord with the quotas, they communicate the final category, and the reasons for that result, to employees. In turn, employ- ees are expected to incorporate that feedback into their indi- vidual development plans. Employees who are assigned category D are placed on ExxonMobil’s Management of Lower Relative Performance program. That program gives employees the option either to continue working under a performance improvement plan or to resign and receive a base salary and outplacement services until a specified date. Those who choose the latter, also called
1 Another assessment process applies to “managerial, professional,
and technical” employees. Operational, clerical, and administrative employees are ranked as outstanding (including a subcategory of outstanding with distinction); ex- cellent; very good; good; and in need of improvement (including a subcat- egory for in need of significant improvement). 4 No. 24-2823
the “Pay in Lieu” option, agree to sign a separation agreement and release claims. During the 2019–2020 performance assessment cycle at is- sue here, the performance improvement plan option was not available to employees who fell into the Management As- signed Category (“MAC”), meaning they were a new hire with three or more years of relevant experience before joining ExxonMobil. MAC-designated employees assigned category D could choose either to receive Pay in Lieu or be terminated without the benefits of Pay in Lieu. B. Mitchell’s Assessment ExxonMobil operates a fuel and lubricants plant in Cicero, Illinois. Mitchell began work at the testing laboratory there in March 2016 as a contractor while employed by a staffing com- pany. As a lab technician, she tested samples of the products entering and leaving the facility. In 2019, after approximately three years as a contractor, colleagues Jeffrey Hayes and Karen Hasberger encouraged Mitchell to apply for a full-time position with ExxonMobil at the Cicero plant. She was hired on April 22, 2019. Her title and work responsibilities did not change when her employer be- came ExxonMobil, and they remained the same until her ter- mination on August 28, 2020. Mitchell worked with two other lab technicians, Victor Aguirre and Daniel Pitts. The parties dispute who supervised these employees. The company maintains that although Hayes was based in the Paulsboro, New Jersey facility, he su- pervised all the lab technicians at Cicero. Mitchell counters that Hasberger, a chemist at the lab, was her direct supervisor. After Hasberger left the company, Mitchell says the No. 24-2823 5
supervisory role passed to a male employee. Still, she claims, her performance was ultimately reviewed by three other su- pervisors: Cicero plant manager Raul Sanchez, Hayes, and their direct supervisor Louis Henry “Hank” Muller. Mitchell had her first and only performance assessment by ExxonMobil approximately one year into her full-time em- ployment. As the manual required, she drafted a one-page summary of her accomplishments, strengths, and areas for development. She also selected five co-workers to evaluate her, including her two colleagues from the Cicero lab, Aguirre and Pitts. Mitchell met with Hayes to discuss her performance. Mitchell offered conflicting statements as to whether Hayes told her which preliminary assessment category she was as- signed. She said first that she learned she was assigned cate- gory B during that conversation. But later in her deposition, when asked what Hayes “conveyed” to her about her perfor- mance during that conversation, she responded, “I don’t re- call.” A June 9, 2020, email Hayes sent to Shannon McGuire in human resources assigned category B to all three lab techni- cians. On June 15, Muller emailed the cluster of supervisors he managed to discuss the preliminary assessment categories for eleven employees in their departments. Muller explained that the categories assigned to the operational, clerical, and ad- ministrative employees had to meet a new “required distribu- tion.” 2 That meant, he instructed, that category A must be
2 Muller’s email contained a table displaying the expected distribution
of performance categories within an assessment group. Employees as- signed the letter A for outstanding contributor were expected to make up 6 No. 24-2823
assigned to two employees and category C to two other em- ployees. To satisfy that distribution, Muller lowered Pitts to cate- gory C but kept Aguirre and Mitchell at category B. Muller indicated the proposed categories were not final and asked the supervisors he managed to “reply back with agreement or a different proposed category and rationale.” The same day Muller emailed only Sanchez, explaining he did not “see [Mitchell], [Pitts], or [Aguirre] enough to judge their initial placement” and asking for “[a]ny thoughts” about their letter assignments. The next day, Hayes wrote to Sanchez with new preliminary letter assignments for the three lab technicians. The email explained that Hayes and Muller discussed having to assign a category D. It was suggested to keep Aguirre at B, leaving how to apportion the C and D as- signments between Pitts and Mitchell. Attaching self-assessments from each of the lab techni- cians, Hayes concluded: “I would place [Pitts] as the C and unfortunately, [Mitchell] MAC’d at the D, if need be.” Hayes asked Sanchez to “review and confirm or realign my think- ing.” There is no evidence that Sanchez responded to the emails he received on June 15 and 16. He testified he did not believe he responded to either email. Mitchell claims Sanchez was in- volved in her performance assessment process “as early as June 15” when his name first appeared in an email recipient list from Muller.
20–30% of the assessment group; B for strong contributor, 40–50%; C for contributor, 20–30%; and D for needs improvement, 8–10%. No. 24-2823 7
At the final stage of the performance assessment process, Mitchell was assigned to a group based on her position and level of experience. As a lab technician, she was considered an operational, clerical, and administrative employee. She was also considered a MAC employee because she had three or more years of relevant experience as a contractor before start- ing full-time at ExxonMobil. Her assessment group consisted of three other first-year employees in the MAC category, two men and one woman: Gilmore Kwon, Louis Palombo, and Terry Davis. Even though the others were not based in the Cicero lab, like Mitchell they worked on fuels and lubricants operations. On June 17, two different groups were discussed at the as- sessment group meeting. Hayes represented Mitchell, while three other supervisors represented Kwon, Palombo, and Da- vis. Muller testified he was there to represent employees in the second group. But Mitchell disputes that Muller was pre- sent and challenges his testimony about what was discussed there. According to Muller, the supervisors at the meeting deter- mined that the other employees in Mitchell’s assessment group “tended to go ‘above and beyond’ in performing their job responsibilities, while [Mitchell] did not.” Mitchell was “simply not a ‘stand out’ in comparison to her peers.” At the conclusion of the meeting, Mitchell was assigned a final per- formance assessment category of D, while the other three em- ployees were assigned final categories of B. Ordinarily Mitchell would have received feedback about her assigned category from her direct supervisor—Hayes. But Hayes also fared poorly during the performance assessment for the 2019–2020 cycle and opted to retire. So, Muller 8 No. 24-2823
approached Sanchez, requesting he deliver feedback to Mitchell, Pitts, and Aguirre in Hayes’s stead. As Sanchez was not present at the assessment group meeting, Muller in- formed him about the discussions that took place and the rea- sons for Mitchell’s final relative assessment category. To Mitchell, Sanchez’s explanation for her assessment was founded on Muller’s “self-serving” declaration and was thus pretextual. On August 21, Sanchez met with Mitchell to deliver the re- sults of her performance assessment. Sanchez explained to Mitchell that her assessment group was “competitive” and that her assigned category resulted from a “lack of ‘stand outs’” when compared with her peers. He also provided Mitchell with a Separation Agreement explaining her options as a MAC employee assessed in the D category. Consistent with company policy and practice, assigning category D to Mitchell meant she was placed in the Manage- ment of Lower Relative Performance program. But as she was also a MAC employee, the policy during the 2019–2020 per- formance assessment cycle dictated she could choose either the Pay in Lieu option or resign immediately without the ben- efits of Pay in Lieu and without relinquishing her right to sue. After meeting with Sanchez, Mitchell emailed human re- sources several times between August 24 and 27 to inquire about the terms of her termination and the reasons for her as- sessment category. A human resources manager addressed Mitchell’s questions about the terms of her termination but directed her to consult Sanchez about her assessment process. Sanchez spoke again to Mitchell on August 24 about the as- sessment process. No. 24-2823 9
Mitchell ultimately decided not to sign the separation agreement and her employment with the company ended. She later learned a male contract employee assumed some of her responsibilities at the lab. As of early 2024, ExxonMobil asserts that no employee was hired to fill Mitchell’s position. Mitchell brought this suit against ExxonMobil alleging sex discrimination in violation of Title VII. 42 U.S.C. § 2000e- 2(a)(1). Following discovery, which the parties requested be extended numerous times, ExxonMobil moved for summary judgment on Mitchell’s sex discrimination claim. The district court granted the motion, and Mitchell appeals. II. Evidentiary Rulings Before reaching the merits of Mitchell’s sex discrimination claim, we consider her appeal of two evidentiary issues, which we review for abuse of discretion. Igasaki v. Ill. Dep’t of Fin. & Pro. Regul., 988 F.3d 948, 955 (7th Cir. 2021). First, Mitchell disputes whether Muller was present at her assessment group meeting and thus his declaration about the reasons she was assigned category D. She characterizes his testimony as “self-serving” and submits it should not be re- lied on at summary judgment. “[S]elf-serving statements in affidavits without factual support in the record carry no weight on summary judgment.” Butts v. Aurora Health Care, Inc., 387 F.3d 921, 925 (7th Cir. 2004). Rather than self-serving, Muller’s declaration about Mitchell’s category D assignment fell within his new respon- sibility. The ExxonMobil manual requires supervisors to dis- cuss the results of the performance group meeting with em- ployees to explain the reasons for their final assessments. 10 No. 24-2823
Hayes’s retirement transferred that responsibility to Muller as Hayes’s immediate supervisor. Even assuming Muller’s declaration was self-serving, suf- ficient facts show that he attended the meeting. He stated he was there as a representative of an employee in the other group discussed that day. Muller was also included in an email McGuire sent to “Assessment Meeting Participants.” No proof contradicts this evidence, so the district court did not abuse its discretion when it considered Muller’s declaration. Second, Mitchell asked the district court to consider evi- dence from a different case, Ontiveros v. Exxon Mobil Corp., No. 21 C 2335, 2024 WL 3823172 (N.D. Ill. Aug. 14, 2024). 3 She sought to use that evidence to support her claim of a pattern or practice of discrimination against female employees at the Cicero plant. The district court ruled inadmissible the evi- dence from the other case because the “defendant ha[d] had no opportunity to test [it] in the context of this case.” Mitchell asks us to reconsider the district court’s eviden- tiary decision because the other case involved the same defendant, the same attorney, and many of the same supervi- sors, but only a different female plaintiff. The evidence in that case, she argues, is “probative, relevant, and a matter of public record.” To introduce evidence from another case, as Mitchell re- quests, would violate Federal Rules of Civil Procedure 26(a) and 37(c)(1). Rule 26(a) requires a party to disclose the name of “each individual likely to have discoverable information—
3 A panel of this court heard argument in that case on May 14, 2025
(Appeal No. 24-2645). The same attorney represented the plaintiff in that case. No. 24-2823 11
along with the subjects of that information—that the disclos- ing party may use to support its claims or defenses.” A party who fails to do so is not allowed to use undisclosed infor- mation or witnesses “to supply evidence on a motion … un- less the failure was substantially justified or is harmless.” FED. R. CIV. P. 37(c)(1). Mitchell did not disclose her intent to rely on any docu- ments or depositions from the Ontiveros case before including them in her district court submission. Permitting their use at such a late stage would have prejudiced ExxonMobil, which had no opportunity to test that evidence. The district court thus did not abuse its discretion when it excluded evidence from Ontiveros, and that evidence is not included in our re- view of the district court’s summary judgment ruling. III. Title VII Discrimination We turn now to the merits of Mitchell’s discrimination claim. We review a district court’s grant of summary judgment de novo, construing the record in the light most fa- vorable to Mitchell, the nonmoving party, and drawing all reasonable inferences in her favor. Napier v. Orchard Sch. Found., 137 F.4th 884, 891 (7th Cir. 2025). “[G]eneralized and unsupported allegations cannot create a genuine dispute” precluding summary judgment. Anderson v. Street, 104 F.4th 646, 651 (7th Cir. 2024). So, ExxonMobil “may prevail by showing an absence of evidence to support” Mitchell’s claim. Wince v. CBRE, Inc., 66 F.4th 1033, 1040 (7th Cir. 2023). Mitchell contends that ExxonMobil discriminated against her based on her sex when it terminated her. Title VII makes it unlawful for an employer to “discharge any individual … because of such individual’s … sex … .” 42 U.S.C. § 2000e- 12 No. 24-2823
2(a)(1). In discrimination cases, “[w]hen a defendant moves for summary judgment, the ‘singular question’ for the district court is whether the plaintiff has introduced evidence that would ‘permit a reasonable factfinder to conclude that the plaintiff’s race, ethnicity, sex, religion, or other proscribed fac- tor caused the discharge or other adverse employment ac- tion.’” Igasaki, 988 F.3d at 957 (quoting Purtue v. Wis. Dep’t of Corr., 963 F.3d 598, 602 (7th Cir. 2020)). Like many plaintiffs alleging discrimination, Mitchell chose to carry her evidentiary burden by invoking the frame- work of McDonnell Douglas Corp. v. Green, 411 U.S. 792, 802– 03 (1973). To make a prima facie case under McDonnell Doug- las, Mitchell must show that she (1) belongs to a protected class, (2) was qualified for the job, (3) suffered an adverse employment action, and (4) was treated less favorably than similarly situated employees outside of her protected class. Igasaki, 988 F.3d at 957. Most plaintiffs find meeting “the prima facie burden is ‘not onerous.’” Ames v. Ohio Dep’t of Youth Servs., 145 S. Ct. 1540, 1545 (2025) (quoting Tex. Dep’t of Cmty. Affs. v. Burdine, 450 U.S. 248, 253 (1981)). Indeed, the Su- preme Court has repeatedly explained that the “precise re- quirements of a prima facie case can vary depending on the context and were ‘never intended to be rigid, mechanized, or ritualistic.’” Id. at 1546 (quoting Swierkiewicz v. Sorema N. A., 534 U.S. 506, 512 (2002)). If Mitchell successfully makes her prima facie showing, the “burden shifts to [ExxonMobil] to articulate a legitimate reason for the adverse action.” Partin v. Baptist Healthcare Sys., Inc., 135 F.4th 549, 559 (7th Cir. 2025) (quoting Elkharwily v. Mayo Holding Co., 823 F.3d 462, 470 (8th Cir. 2016)). This too is No. 24-2823 13
a light burden. See Stockwell v. City of Harvey, 597 F.3d 895, 901 (7th Cir. 2010). If ExxonMobil articulates a “legitimate, nondiscrimina- tory reason for its decision, the presumption of discrimination falls away.” Id. At that point, the burden “shifts back to [Mitchell] to show that the employer’s stated reason was a pretext.” Igasaki, 988 F.3d at 957 (citation omitted). Mitchell also urges us to consider her claim through the holistic lens articulated in Ortiz v. Werner Enterprises, Inc., 834 F.3d 760 (7th Cir. 2016). There, this court said that a plaintiff does not have to satisfy the McDonnell Douglas framework to succeed on a discrimination claim under Title VII. “Although there are many tests and rubrics for viewing discrimination claims, it is important to recall that, at the end of the day they are all merely convenient ways to organize our thoughts.” Brooks v. Avancez, 39 F.4th 424, 433 (7th Cir. 2022). This court said in Brooks that district courts should evalu- ate direct and indirect evidence together as a whole to assess “the only question that matters:” whether a reasonable fact- finder could conclude that a plaintiff suffered the adverse em- ployment action because of her membership in a protected class. Id.; see also Vega v. Chi. Park Dist., 954 F.3d 996, 1004 (7th Cir. 2020) (“What matters is whether she presented enough evidence to allow the jury to find in her favor.”). Consistent with Mitchell’s arguments, we evaluate the ev- idence using both approaches. See Napier, 137 F.4th at 891 (“McDonnell Douglas is entirely consistent with our holding in Ortiz, and it ‘remains an efficient way to organize, present and assess evidence in discrimination cases.’” (quoting Reives v. Ill. State Police, 29 F.4th 887, 892 (7th Cir. 2022))). 14 No. 24-2823
A. McDonnell Douglas Framework Neither party disputes that Mitchell satisfies the first and third elements of the McDonnell Douglas framework. As a woman she belongs to a protected class, and she suffered ad- verse action when she was terminated from her employment. In dispute are elements two and four. If on this record no rea- sonable jury could find that Mitchell was similarly situated to her chosen comparators, she cannot satisfy the fourth element of the framework. “To prevail by showing a similarly situated employee was treated differently, a plaintiff must show the purported comparator was ‘directly comparable to her in all material re- spects’ so as to ‘eliminate other possible explanatory varia- bles.’” Gamble v. County of Cook, 106 F.4th 622, 626 (7th Cir. 2024) (quoting Downing v. Abbott Lab’ys, 48 F.4th 793, 805 (7th Cir. 2022)). The purpose of the “similarly situated inquiry is to eliminate other possible explanatory variables, … [to] iso- late the critical independent variable—discriminatory ani- mus.” Smith v. City of Janesville, 40 F.4th 816, 823 (7th Cir. 2022) (citation modified). Mitchell identifies the two male lab technicians she worked with at the Cicero lab—Pitts and Aguirre—as com- parators. She contends they were similarly situated to her be- cause they held the same title, worked in the same location, responded to the same local manager (first to Hasberger and then to Sanchez), and were assigned preliminary assessment categories by the same three managers (Sanchez, Hayes, and Muller). Despite these similarities, Mitchell maintains that she was assigned to a lower final assessment category, which would allow a jury to infer sex discrimination. No. 24-2823 15
But finding the appropriate employees who are similarly situated to Mitchell is not so straightforward. ExxonMobil’s annual performance assessment process uses a particular method for comparing employees. It assigns employees to as- sessment groups based on similar roles and with similar ex- perience, even if employees have different job titles, report to different supervisors, and work in different locations across the country. One element of the company’s process is to des- ignate newly hired employees with three or more years of rel- evant experience before being hired full-time at ExxonMobil as MAC employees. Recall that Mitchell, a MAC employee, was assigned to a group of other MAC employees with operational, clerical, and administrative responsibilities. The employees included in her assessment group were one woman, Davis, and two men, Kwon and Palombo. Pitts and Aguirre, neither of whom were MAC employees, were placed in different assessment groups. Thus, under ExxonMobil’s assessment process, Mitchell, Pitts, and Aguirre were not similarly situated employees. Mitchell does not complain that ExxonMobil’s assessment process is facially discriminatory. Nor does she take issue with being categorized as a MAC employee, with the assess- ment group assignment process, or with the three individuals to whom she was compared. To Mitchell, the relative ranking that takes place at the assessment group meeting is beside the point. Rather, she contends discrimination occurred in the steps her supervisors took leading up to that meeting. 1. Mitchell alleges process errors in her evaluation. Mitchell first maintains that “key evidence” shows that as- sessment rankings are not “finalized” during the assessment 16 No. 24-2823
group meeting. She claims instead that preliminary assess- ments by local supervisors, as much as their input during the assessment group meeting, play a critical role in determining final categories. As proof, Mitchell explains that her prelimi- nary assessment category changed from a B to a D after Hayes, Muller, and Sanchez discussed the need to assign that category to one of the Cicero lab technicians. She relies on the fact that her assessment category did not change at the final assessment group meeting as proof that the preliminary pro- cess was as important as the final group meeting, if not more so. This is unconvincing for two reasons. First, Mitchell can- not point to any evidence showing that the final assessment group meeting was a pretense. The fact that her preliminary assessment category did not change does not discredit the as- sessment process. In any event, the evidence undermines Mitchell’s position. The record shows that Hayes initially assigned category B to each of Mitchell, Pitts, and Aguirre. Then, Muller changed the category for Pitts to C. And, after further discussion among and contemplation by supervisors, Hayes changed Mitchell’s category to D. From this timeline, the performance assess- ment process appears dynamic, involving supervisors at dif- ferent levels in the hierarchy discussing the readily changing rankings. This means that the preliminary assessment cate- gory could have just as easily changed at the final assessment meeting as it did in the preliminary assessment stage. Second, Mitchell claims that preliminary ranks are deter- minative of the final outcome because the categories did not change for her preferred comparators—Pitts and Aguirre—or her actual comparators—Davis, Kwon, and Palombo. The No. 24-2823 17
record does not support her argument, though. Pitts and Aguirre were initially assigned categories C and B, respec- tively, but evidence did not reveal their final assessment cate- gories. Davis, Kwon, and Palombo each received a final as- sessment category of B, yet their initial assessment categories are not part of the record. Without more, there is no support for Mitchell’s claim that preliminary categories did not change for any of the comparable employees. Mitchell also complains that in the preliminary assessment stage, she was ranked relative to Pitts and Aguirre in violation of the assessment process described in ExxonMobil’s manual. That portion of the manual is vague as to what goes into the preliminary assessment categorization. Supervisors are ex- pected to discuss their perspectives on an employee’s perfor- mance with the aid of the employee’s self-assessment and the feedback of the employee’s colleagues. Inherent in a supervi- sor’s analysis of an employee’s performance is a comparison of the employee to his other direct reports. Hayes cannot be expected to blind himself to the performance of other lab tech- nicians in his assessment of Mitchell. In any case, as already explained, Mitchell has not proved that the final assessment meeting merely rubber-stamped her preliminary category. So, even if some deviation from the manual occurred in the preliminary assessment stage, Mitch- ell has not established that she was prejudiced by it. 2. Mitchell’s alleged comparators are not similarly situated. The district court was correct that Pitts and Aguirre are not similarly situated to Mitchell—but not only because they were assigned to different assessment groups. The court ex- plained that, even assuming Pitts and Aguirre are proper 18 No. 24-2823
comparators, Mitchell did not offer evidence from which to draw a comparison relevant to the performance assessment context. To establish Pitts and Aguirre as proper comparators, Mitchell would have to show that they were assessed unfa- vorably in comparison to others in their assessment groups but were not assigned lower categories than their assessment group peers and were not placed in the Management of Lower Relative Performance program (like Mitchell was). Mitchell resists the probative value of this evidence. She coun- ters that this requirement is “overly stringent,” especially as she presented evidence that her supervisors “tainted the pro- cess for her” with a low preliminary assessment. The requirements for a prima facie case do vary by context and are not subject to a “rigid, mechanized, or ritualistic” in- ventory of facts. Ames, 145 S. Ct. at 1546. But the type of evi- dence the district court said would be proper to make this comparison is not overly stringent. The record does not con- tain any evidence about the final assessment categories as- signed to Pitts and Aguirre. Without any evidence about how they compared in the annual assessment process, the district court did not err when it concluded that Mitchell’s contention that she was treated less favorably based on sex is “mere spec- ulation.” 3. Mitchell cannot establish that ExxonMobil’s decision was pretextual. Even if we assume that Mitchell established a prima facie case, her sex discrimination claim fails because ExxonMobil’s proffered reason for termination—that her performance com- pared poorly relative to others in her comparator group—is No. 24-2823 19
nondiscriminatory. The burden thus shifts back to Mitchell to show that ExxonMobil’s proffered reason is pretextual. Stock- well, 597 F.3d at 901. A pretext is “a lie, specifically a phony reason for some action, not just faulty reasoning or mistaken judgment.” Napier, 137 F.4th at 892 (citation modified). Mitch- ell has also failed to meet that burden. She argues that no evidence supports the rationale that one of the Cicero lab technicians had to be placed in the D category before being ranked in their peer assessment group. In an email to his cluster of supervisors, though, Muller ex- plained that ExxonMobil has a required distribution for rank- ings and instructed the supervisors to keep that distribution in mind when recommending preliminary assessment catego- ries. This means that one or more employees in the cluster would have to be ranked in the D category. Why a Cicero lab technician had to receive that rank, and how many other employees in the cluster received it, is not part of the record. Yet answers to these questions would not help Mitchell. Whether for business reasons or arbitrarily, a business may decide to downsize a department without sub- sequent judicial review. See Galvan v. Indiana, 117 F.4th 935, 939 (7th Cir. 2024) (“[T]he court ‘is not a super personnel de- partment that second-guesses employers’ business judg- ments.’” (quoting Grant v. Trs. of Ind. Univ., 870 F.3d 562, 570 (7th Cir. 2017))). Mitchell also argues that the reasons articulated for her termination were pretextual because there is no evidence that her performance fell so short of her peers as to justify her cat- egory D assignment. But Mitchell did not fare well in compar- ison to the other three MAC employees in her assessment group. Muller testified based on his observations of the 20 No. 24-2823
assessment group meeting that the process was “competi- tive.” The managers discussed how Davis, Kwon, and Pal- ombo “tended to go ‘above and beyond’ in performing their job responsibilities, while [Mitchell] did not.” Mitchell was “simply not a ‘stand out’ in comparison to her peers.” Apart from Muller’s testimony, Mitchell does not cite any evidence to show that her performance was comparable to that of the employees in her assessment group. The record does not contain self-assessment summaries from the three comparators to weigh against Mitchell’s accomplishments. When interviewed, Mitchell admitted to not having met her comparators, not having observed their job performance, not knowing how their job performance compared to hers, and not knowing what final assessment they received during the previous annual review. True, there is also no evidence that Mitchell performed poorly or had any negative reviews prior to the assessment group meeting. Yet that meeting differenti- ated the employees and ranked them according to their rela- tive performances under ExxonMobil’s protocols. Mitchell was ranked last in that unchallenged process. Mitchell also argues that Sanchez has a history of discrim- inating against women at the Cicero plant and that he dis- played the same animus when ranking her performance at the preliminary ranking stage. This argument too is unpersua- sive. As already discussed, any purported evidence of Sanchez’s discriminatory behavior against women at Cicero comes from deposition testimony from a different case and is thus inadmissible here. What is more, Mitchell’s deposition testimony about Sanchez tells a different story: Q: Did you ever have any negative interactions with Mr. Sanchez? No. 24-2823 21
A: No. Q: Did Mr. Sanchez ever do anything during your ExxonMobil employment that caused you to believe that he was biased against women in some fashion? A: No. Q: Have you ever heard him make any deroga- tory comments about women in any fashion? A: No. As for Sanchez’s involvement in the preliminary assess- ment process, Hayes and Muller did ask for his input. Yet there is no evidence that Sanchez responded to those emails or provided his input in any way. Per the record, Sanchez be- came involved in Mitchell’s assessment process in August when Muller tasked him with filling in for Hayes and deliv- ering the assessment results. In sum, Mitchell fails to carry her burden of proving that ExxonMobil’s proffered reason for her termination is pre- textual. Other than the fact that she is a member of a protected class, there is no evidence in the record from which a reason- able juror could infer that she was terminated because of her sex. B. Ortiz’s Holistic Approach Mitchell also invokes Ortiz’s holistic approach, pointing to other ways in which she was treated less favorably than her male counterparts. The totality of the evidence approach ar- ticulated in Ortiz instructs us to ask whether Mitchell experi- enced sex discrimination, “eschewing any framework or for- mula.” Igasaki, 988 F.3d at 958 (citing Ortiz, 834 F.3d at 765). 22 No. 24-2823
She argues she presented sufficient evidence to establish a pattern or practice of sex discrimination at the Cicero plant. Pattern or practice claims require “showing that an employer regularly and purposefully discriminates against a protected group.” Puffer v. Allstate Ins. Co., 675 F.3d 709, 716 (7th Cir. 2012) (quoting Council 31, Am. Fed’n of State, Cnty. & Mun. Emps. v. Ward, 978 F.2d 373, 378 (7th Cir. 1992)). Mitchell must prove that “discrimination ‘was the company’s standard op- erating procedure—the regular rather than the unusual prac- tice.’” Id. (quoting Int’l Bhd. of Teamsters v. United States, 431 U.S. 324, 336 (1977)). Mitchell claims that four other female employees at the Cicero plant—Amee Patel, Hasberger, Kathleen Collier, and Margaret Jeziorski—were terminated around the time she was. To Mitchell, this constitutes a pattern of Sanchez using the Management of Lower Relative Performance program to systematically remove female employees from Cicero. Her contention fails in two ways. First, this evidence is in- admissible. As described above, deposition testimony taken in a different case from witnesses who were not disclosed in this case prejudices the defendant and cannot be considered here. Second, even if admissibility was not at issue, this evi- dence does not advance Mitchell’s pattern or practice claim. None of the female employees Mitchell discusses were di- rectly supervised by Sanchez. And, even if he played a role in their termination, there is no evidence that he had a meaning- ful role in Mitchell’s assessment process. The circumstances of these female employees also differed from Mitchell’s. Patel and Hasberger were managerial, pro- fessional, and technical employees, so they were subject to a different performance assessment process. Patel chose to No. 24-2823 23
retire before participating in the assessment process. And Col- lier was a contract employee, so she would not have been a part of the company’s performance assessment process. Mitchell has not offered evidence of a pattern or practice of sex discrimination by ExxonMobil. The district court cor- rectly concluded that Mitchell’s pattern or practice claim “boils down to her assertion that ‘several’ women in various positions at the Cicero plant were ‘let go by Sanchez through the [Management of Lower Relative Performance] process.’” But this assertion, without context and “at this high level of generality,” is not actionable. IV. Conclusion Mitchell fails to present sufficient facts to support an infer- ence of sex discrimination. We agree with the district court that no reasonable jury could find that ExxonMobil subjected Mitchell to sex discrimination. Mitchell’s Title VII discrimina- tion claim thus fails under both the McDonnell Douglas bur- den-shifting framework and Ortiz’s holistic approach. We therefore AFFIRM the district court’s grant of summary judg- ment to ExxonMobil.