Jacinta Downing v. Abbott Laboratories

48 F.4th 793

• Court: Court of Appeals for the Seventh Circuit
• Decided: September 12, 2022
• Docket: 21-2746
• Status: Published

Opinion

In the

United States Court of Appeals For the Seventh Circuit ____________________ No. 21-2746 JACINTA DOWNING, Plaintiff-Appellant, v.

ABBOTT LABORATORIES and ABBOTT MOLECULAR, INC., Defendants-Appellees. ____________________

Appeal from the United States District Court for the Northern District of Illinois, Eastern Division. No. 1:15-cv-05921 — John J. Tharp, Jr., Judge. ____________________

ARGUED JUNE 3, 2022 — DECIDED SEPTEMBER 12, 2022 ____________________

Before SYKES, Chief Judge, and FLAUM and BRENNAN, Cir- cuit Judges. BRENNAN, Circuit Judge. Jacinta Downing worked for many years as a sales manager and then a sales executive at Abbott Molecular, Inc. Over time, that company faced finan- cial difficulties. The company said that because of reductions in its sales force and Downing’s work performance, it ended her employment. Downing claims the company racially dis- criminated and retaliated against her, so she sued. 2 No. 21-2746

Many of her claims survived summary judgment, but after trial a jury found for Abbott Molecular. On appeal Downing challenges several of the district court’s decisions, including evidentiary rulings, the exclusion of her expert witness, the jury instructions, and the testimony of her former manager, on which she moved for a mistrial. Downing argues that these errors, individually and cumulatively, denied her a fair trial. She also appeals the grant of summary judgment to the com- pany on her disparate-impact claim, for which she contends she had sufficient evidence. We conclude that as to each deci- sion, the district court ruled correctly or did not abuse its dis- cretion, so we affirm. I. Background Our description of the relevant facts comes from the jury trial transcript and other district court records. A. Factual Jacinta (or Jay) Downing, an African-American woman, had many years of sales experience when she was hired in 2002 by Abbott Molecular, Inc., a subsidiary of Abbott Labor- atories. Downing’s first job was Area Sales Manager. Her su- pervisor was Chris Jowett, a white man, who in 2009 arranged for her to be promoted to be one of four Regional Sales Man- agers. Each supervised a team of sales representatives who sold millions of dollars of healthcare products to hospitals, commercial laboratories, and clinics, and who negotiated ser- vice contracts. According to Downing, she “really liked working for” Jo- wett, who was “very inspirational” and mentored her. She considered him a man of great integrity and honesty. In March 2011 Jowett reviewed Downing and gave her an No. 21-2746 3

overall performance rating of “achieved expectations.” Ac- cording to that review, one of her main challenges was fore- casting future business. Downing also fell somewhat short in developing relationships with decisionmakers at key ac- counts, and she was not comfortable with some of Abbott Mo- lecular’s product categories. In 2012, Jowett gave Downing a performance rating of “ex- ceeded expectations,” the highest rating available. He wrote, “Jay is a strong leader for the Abbott Molecular sales organi- zation and can be counted on to deliver results in spite of chal- lenges. I look forward to Jay’s continued success at Abbott Molecular.” Jowett was pleased with Downing’s improve- ment over the previous year, and by giving her the “exceeded expectations” rating Jowett said he was encouraging Down- ing. Shortly after completing that performance review, Jowett accepted a new role at Abbott. Abbott Molecular came under financial pressure in early 2012. Medicare cut its reimbursement rates for a key Abbott product, which significantly impacted the company’s mar- gins. Multiple competitors also entered the market with up- dated versions of products Abbott sold. In response, Abbott changed personnel in the Abbott Molecular division and re- duced its work force. Mark Bridgman, a white man, was transferred from another Abbott division to fill Jowett’s for- mer role in Abbott Molecular. Abbott also created the position of National Sales Director for the U.S. to “coach and guide the sales managers and reps on a day-to-day basis a bit more beyond what [Bridgman] was able to do.” In October 2012, Peter Farmakis, a white man, was hired to fill the role. Farmakis oversaw one man, Mike Kohler 4 No. 21-2746

(who is white), and three women: Jean Gray (who is white), Charlotte Jones (who is African-American), and Downing. Almost immediately, Farmakis had issues with Downing, as well as with some of the other managers who reported to him. Downing had a conflict with a customer, in which she withheld a software key that could have possibly disrupted patient care, in October 2012. Farmakis was displeased, and he instructed Downing not to disrupt patient care going for- ward. Two months later, Farmakis was upset after he learned Downing had unilaterally informed another customer that Abbott would forgive a termination fee of $177,000. By Janu- ary 2013, Farmakis had a list of concerns about Downing’s performance, including oversights in end-of-year sales fore- casting, which impacted other parts of Abbott’s business. Farmakis discussed his concerns about Downing’s perfor- mance with Sarah Longoria, Abbott Molecular’s human re- sources director in February 2013. The next month, Farmakis gave Downing her performance review. Although Downing’s overall rating was “achieved expectations,” Farmakis in- cluded some detailed criticisms of her performance and iden- tified areas for improvement. In July 2013, Farmakis had a conference call with his four direct reports: Kohler, Gray, Jones, and Downing. According to Downing, Farmakis was “shouting and screaming” at the three women, whom he accused of “throwing in the towel.” Downing reported the incident to Abbott’s Employee Rela- tions Department. She relayed her belief that Farmakis was discriminating against her because of her race and gender. About the same time, Gray also complained to Employee Re- lations about Farmakis’s behavior. Gray, Jones, and Downing No. 21-2746 5

discussed their opposition to Farmakis’s management in emails with each other. Abbott investigated these complaints against Farmakis. An Employee Relations specialist sent a climate survey to the four managers who reported to him. In August 2013, Gray, Jones, and Downing gave very negative responses, which fo- cused primarily on Farmakis’s management style. Gray wrote that “[Farmakis is] especially hard on Jay. He embarrasses her and calls her out in calls or emails, part of the unfairness. Her numbers are good so don’t understand why he calls her out, makes her feel stupid, don’t know why he’s doing that to her.” The Employee Relations specialist removed the identify- ing information from the survey responses, deleted some of the comments, and sent them to Longoria. The anonymized feedback was shared with Farmakis, who Bridgman then coached to improve his management style. Throughout 2013, Abbott Molecular’s business continued to falter, resulting in layoffs. In January 2014, Abbott realigned its sales teams. Sales representatives who had pre- viously reported to Downing were assigned to new teams. During the same month, Abbott also placed Downing on a performance improvement plan, the last step before termina- tion. Downing then retained legal counsel and gave notice that she intended to file discrimination claims against the company. Abbott later cut Downing’s stock award in March 2014, which left her unhappy. That fall Downing filed a discrimination charge with the EEOC, which she later amended. Throughout 2014 Abbott’s business had not improved, so the company instituted 6 No. 21-2746

another reduction in force in January 2015. All four Regional Sales Managers, including Downing, lost their jobs when that position was eliminated. Farmakis, the National Sales Direc- tor, was also terminated. At the same time Abbott terminated Downing’s employ- ment as part of the reduction in force, the company invited her to apply for the position of Regional Commercial Director. Downing understood this position to be essentially identical to her previous job. Keith Chaitoff, who replaced Bridgman in Abbott Molecular’s leadership, testified his expectations for directors were “[d]ramatically” different. To him, the director role “needed people that understood business more holisti- cally and understood the financial drivers because we had to focus more on profit.” Abbott selected ten candidates, including Downing, to in- terview for the director position. A process was used to rate the candidates and to extend offers. Under that process, two African-American candidates (Downing and Jones), received the lowest ratings of the candidates for the position. Abbott did not select Downing, Jones, Kohler, or Farmakis, and ultimately extended offers for the position to four white can- didates. One of those candidates declined, and an African- American man, Eron Butler, was hired. B. Procedural In June 2015 Downing filed suit claiming discrimination under Title VII and 42 U.S.C. § 1981. She later amended her complaint to allege: (1) racial discrimination under § 1981; (2) retaliation under § 1981; (3) racial discrimination under Title VII; (4) sexual discrimination under Title VII; and (5) retalia- tion under Title VII. The parties proceeded through No. 21-2746 7

discovery, at the close of which Abbott moved for summary judgment. Abbott’s summary judgment motion was “largely denied” because, the district court concluded, “the record in this case is replete with material factual disputes.” For example, the facts did not establish whether Downing had a history of sig- nificant performance problems. To the court, Downing had “adduced enough evidence to support a jury verdict in her favor on her claim that Farmakis retaliated against her for complaining about discrimination by giving her negative per- formance evaluations and placing her on performance man- agement plans.” Likewise, the court considered “evidence of more subtle differential treatment” and an allegedly racially charged remark made by Farmakis. There was also statistical evidence that “Farmakis’s tenure coincided with a dramatic decrease in black representation at Abbott Molecular.” The court reasoned, therefore, that a jury could find racial consid- erations motivated the coaching plan and performance im- provement plan. Further, a reasonable jury could find “Farmakis’s animus was the proximate cause of Downing’s termination.” So, the district court decided that the discrimi- nation and retaliation claims predicated on Abbott’s failure to hire Downing as a director should proceed to trial. The district court did grant Abbott’s motion for summary judgment in two respects. First, the court ruled that “Down- ing’s performance management claim can be based only on race discrimination, not gender discrimination, as only race based discrimination was mentioned in the initial EEOC charge.” Second, Downing could not proceed on her claim of disparate impact in the hiring process for the director 8 No. 21-2746

position, as “a sample size of 10 individuals is simply too small to be statistically meaningful.” 1 At the same time, the district court granted Abbott’s mo- tion to exclude the proposed expert testimony of Dr. Destiny Peery, a legal academic with a background in social psychol- ogy. Dr. Peery intended “to opine that there is evidence in the record that is consistent with how stereotyping and biases (either implicit or explicit) manifest and affect people in em- ployment settings.” But the court concluded that Dr. Peery’s testimony would not help the jury because she disavowed any conclusion about what role, if any, stereotypes and biases played in Abbott’s treatment of Downing. The court also de- cided that Dr. Peery employed an unreliable methodology. Before and during trial, the district court granted Abbott’s motions to exclude certain evidence Downing proffered. The court excluded: • Proposed testimony from two of Downing’s former subordinates, who would have testi- fied they held high opinions of her manage- ment ability and character (“pretext evi- dence”); • Evidence of internal complaints made about two other Abbott employees (“comparator evidence”); • Portions of the climate survey responses that Jones and Gray submitted; and

1 Downing does not appeal the district court’s grant of summary judg-

ment to Abbott on her hostile work environment or sex discrimination claims. No. 21-2746 9

• Statistical evidence of a decline in the num- ber of African-American employees at Ab- bott Molecular between 2012 and 2015. A two-week jury trial took place in August 2021. From the beginning the trial was hard fought. Downing’s counsel raised numerous objections, including “a general concern to the entire way that the opening statement is being presented.” At the close of Abbott’s opening statement, its counsel told the jury: “[T]here are real people with real families being ac- cused of race discrimination in this case … I’m going to ask you to deliver a verdict for Abbott and, in the process, vindi- cate these people and restore their reputations.” During trial Downing presented evidence that she had been a high performer at Abbott, and that Farmakis had an- tipathy towards her and others. Abbott, on the other hand, offered evidence about Downing’s performance problems, the business downturn that led to the reductions in force and Downing’s termination, and the considerations that drove its hiring decisions for the director position. Toward the end of trial, Abbott called Jowett as a witness. He discussed Downing’s performance in detail, including some of her issues. Jowett also testified Downing was “not really open to feedback.” He explained that while he had re- cruited other former Abbott employees to join another com- pany he worked at between 2018 and 2021, he did not recruit Downing because he did not believe she could “take on the complexity of the sales position.” According to Jowett, he was “flabbergasted” upon learning that Downing had identified him in an interrogatory response as someone who discrimi- nated against African-Americans. Downing’s counsel then moved for a mistrial, contending that Abbott’s counsel had 10 No. 21-2746

unduly influenced Jowett by telling him Downing had called him a racist. The court denied the motion, explaining that Downing’s interrogatory response could fairly be read to im- ply Jowett was a racist. The disputes at trial extended to the jury-instruction con- ference. Downing’s counsel proposed a series of instructions beyond those in the Seventh Circuit’s pattern instructions, but the district court rejected most of them, including: • A description of and quotations from the civil-rights statutes under which Downing sued; • A list of the various types of circumstantial evidence that a plaintiff in a workplace dis- crimination case may use; • A statement that making an adverse em- ployment decision because of racial stereo- types is a form of race discrimination; and • An instruction on spoliation for Abbott’s failure to preserve a survey that Farmakis had referenced in his testimony. The district court largely used the Seventh Circuit’s pattern jury instructions. During deliberations the jury posed written questions, two of which are relevant on appeal. The first question was: “Can we award punitives without finding ‘yes’ on any of the claims?” The parties’ counsel agreed to respond “no,” and the district court gave that response. The second question read: “To be considered a ‘protected activity,’ does it need both op- posing and reporting, or are one out of two sufficient.” To the No. 21-2746 11

court, this second question asked whether one action or both was required. After discussion with the parties’ lawyers, the court responded by submitting to the jury the written defini- tion of an unlawful employment practice at 42 U.S.C. § 2000e– 3. Following deliberations the jury returned a verdict for the defense. The jury found that Downing did not prove she was subject to any adverse employment action because of her race. It also found that Downing did not prove any of her retalia- tion claims. After the district court entered judgment, Downing ap- pealed, raising a litany of issues. She challenges many of the court’s evidentiary rulings; the exclusion of her expert wit- ness; the instructions to the jury; and the testimony of Jowett, as well as the court’s denial of the attendant mistrial motion. Our appellate review of these decisions is deferential, so Downing faces a demanding task. She also objects to the dis- trict court’s grant of summary judgment to Abbott on her dis- parate-impact claim, which we review de novo. II. Evidentiary Rulings Downing first contends that the district court excluded ev- idence at trial about her discrimination and retaliation claims on which the court had previously relied to deny Abbott sum- mary judgment. That court ultimately granted 10 of Abbott’s 14 motions in limine, which to Downing resulted in a “funda- mentally unfair trial.” Because “decisions regarding the admission and exclusion of evidence are peculiarly within the competence of the dis- trict court,” Pittman ex rel. Hamilton v. County of Madison, 970 F.3d 823, 829 (7th Cir. 2020), they are reviewed for abuse of 12 No. 21-2746

discretion. Under that standard, “the district court’s decision is to be overturned only if no reasonable person would agree with the trial court’s ruling.” Aldridge v. Forest River, Inc., 635 F.3d 870, 875 (7th Cir. 2011); accord Antrim Pharms. LLC v. Bio- Pharm, Inc., 950 F.3d 423, 430 (7th Cir. 2020) (same). In addi- tion, for reversal to be warranted, the error must have “likely affected the outcome of the trial.” Wilson v. Wexford Health Sources, Inc., 932 F.3d 513, 522 (7th Cir. 2019) (citation omit- ted). Title VII of the Civil Rights Act prohibits an employer from taking an adverse employment action against an indi- vidual “because of such individual’s race, color, religion, sex, or national origin.” 42 U.S.C. § 2000e-2(a). A plaintiff may use circumstantial evidence to prove discrimination through a chain of inferences. Diaz v. Kraft Foods Glob., Inc., 653 F.3d 582, 587 (7th Cir. 2011). Our court has recognized three categories of circumstantial evidence in Title VII cases: “(1) ambiguous statements or behavior towards other employees in the protected group; (2) evidence, statistical or otherwise, that similarly situated employees outside of the protected group systematically receive better treatment; and (3) evidence that the employer offered a pretextual reason for an adverse em- ployment action.” Id. Downing contends that the district court improperly ex- cluded evidence about pretext, comparators, the climate sur- vey responses, and demographic statistics. A. Pretext Evidence Testimony was excluded which Downing argues would have shown that Abbott acted adversely to her based on pre- text. She submits that two of her subordinates, Michael No. 21-2746 13

Cerney and Anecia Thedford, would have testified—contrary to Abbott’s assertions—that Downing was an exemplary manager. According to Downing, Cerney and Thedford found her to be inspirational and highly knowledgeable about products and pricing. Abbott responds that Downing was al- lowed to present those witnesses’ assessments of her manage- ment, although they were properly limited to those opinions conveyed to decisionmakers. To prevail on a Title VII racial-discrimination claim, a plaintiff must provide evidence that the decisionmaker acted because of her race. Schandelmeier-Bartels v. Chicago Park Dist., 634 F.3d 372, 378–79 (7th Cir. 2011). Pretext does not exist “if the decisionmaker honestly believed the nondiscriminatory reason” given by an employer for an adverse employment ac- tion. Stockwell v. City of Harvey, 597 F.3d 895, 902 (7th Cir. 2010) (citation omitted). So, in evaluating pretext, the focus is on what the decisionmakers knew, and their perceptions are “controlling.” Id. at 903. Downing asserts this principle does not apply here. To Downing, pretext exists in two ways: Abbott’s assertion that Farmakis believed she was not knowledgeable about the com- pany’s products or skilled in front of customers, and in Farmakis’s statements that members of Downing’s team re- layed that message to him. Even assuming her proposed ex- ception exists to the rule limiting the inquiry to what was shared with decisionmakers, this argument does not succeed. Contrary to Downing’s portrayal, neither Farmakis nor Ab- bott maintained or suggested that Cerney or Thedford leveled these criticisms against Downing. Rather, per Farmakis and Abbott, other employees under Downing’s supervision made the statements in question, which were consistent with 14 No. 21-2746

Farmakis’s firsthand observations. Because several employ- ees other than Cerney and Thedford reported to Downing, those two individuals could not have rebutted the testimony that certain subordinates shared the criticisms with Farmakis. Cerney was permitted to offer substantial testimony about Downing as a “motivational,” supportive, and kind manager. More to the point, the district court ruled narrowly that Downing’s counsel could elicit testimony from Downing’s subordinates that Farmakis did not seek out their opinions about her management style. But counsel could not ask what the subordinates would have told Farmakis about Downing’s character or expertise. The district court emphasized that the relevant opinions were those held by Farmakis or another de- cisionmaker. The district court properly concentrated this inquiry by limiting the testimony of Downing’s subordinates to what they had communicated to Farmakis. Whether Cerney and Thedford agreed with Farmarkis’s poor opinion of Down- ing’s capabilities is not relevant to her Title VII claims, and such testimony would have prejudiced Abbott. To evaluate pretext, the evidence is what the decisionmakers knew and believed. Stockwell, 597 F.3d at 903. Downing’s contention about what these two subordinates thought of her therefore does not engage with the district court’s reasoning or the rel- evant case law. The court was within its discretion to exclude Downing’s additional evidence purporting to show pretext. B. Comparator Evidence Downing argues next that the district court abused its dis- cretion by not admitting evidence relating to two purported comparators: her co-manager Mike Kohler, and Kirk Mason, No. 21-2746 15

an employee in another department. To Downing, the “com- plaints about, investigation of, and suit against Kohler were paradigmatic comparator evidence.” She asserts Kohler had the same job as she did, they were subject to the same stand- ards, and he engaged in similar but more serious conduct than she did. Abbott responds that Kohler’s situation in- volved different decisionmakers, job responsibilities, allega- tions, and time periods. To prevail by showing a similarly situated employee was treated differently, a plaintiff must show the purported com- parator was “directly comparable to her in all material respects” so as to “eliminate other possible explanatory vari- ables.” Williams v. Off. of Chief Judge of Cook Cnty., 839 F.3d 617, 626 (7th Cir. 2016) (citation omitted); see also Barbera v. Pearson Educ., Inc., 906 F.3d 621, 629 (7th Cir. 2018). The two employ- ees who are purportedly similarly situated must deal with the same supervisor, be subject to the same standards, and “have engaged in similar conduct without such differentiating or mitigating circumstances as would distinguish their em- ployer’s treatment of them.” Barbera, 906 F.3d at 629 (citations omitted). The differences between Kohler’s conduct and Downing’s conduct overwhelm the similarities. Kohler was accused of creating a hostile work environment by intimidating his di- rect reports. Those allegations were unlike the performance problems Abbott cited when Downing’s employment was ter- minated. Further, Kohler was a National Molecular Physician Specialist, a different position than Downing. Kohler also re- ported to a different supervisor—Bridgman, not Farmakis. So, as a matter of law, Kohler was not a comparator relative to Downing. See id. The district court thus did not abuse its 16 No. 21-2746

discretion in excluding the evidence about Kohler’s alleged workplace misconduct. Downing also claims the district court abused its discre- tion by not admitting evidence about Mason, even though Abbott opened the door to such evidence by implying a com- plaint Downing made against him was baseless. Abbott responds that Downing effectively forfeited this argument be- cause she failed to address the two reasons the district court gave to not permit Downing to testify about Mason: She lacked personal knowledge of the investigation into the alle- gations against him, and such testimony would confuse the jury. Mason, a manager in the Contracts and Pricing Depart- ment, was referenced in Abbott’s opening statement because he did not authorize Downing to forgive the $177,000 debt to a customer. Though Downing is correct that Abbott implied she had fabricated charges of racism against Mason, the alle- gations against Mason that Downing was precluded from introducing at trial were largely unrelated. The Employee Re- lations Department found Mason had engaged in yelling, speaking in a condescending manner, using inappropriate language, and referring to colleagues as his “work wife.” None of that was relevant to Downing’s job performance or her charge of racial discrimination. Setting aside Downing’s lack of personal knowledge about the outcome of the investi- gation into Mason’s wrongdoing, the district court was within its discretion to conclude that such evidence would have con- fused the jury and fallen short under Federal Rule of Evidence 403. No. 21-2746 17

C. Climate Survey Responses Downing asserts that because the climate survey re- sponses showed that Farmakis treated other members of her protected class poorly, the district court should not have ex- cluded them. According to Downing, “Jones’s and Gray’s contemporaneous, written complaints about Farmakis’s dis- criminatory behavior, memorialized in their climate survey responses, were extremely relevant” and were “inextricably part of Downing’s circumstances and theory of the case.” Ab- bott points out that the district court admitted into evidence, over Abbott’s objections, the PowerPoint summary of the cli- mate survey that Employee Relations received and reviewed. To Abbott, the district court was correct not to “allow Down- ing to submit portions of other people’s responses that had nothing to do with race discrimination or her interactions with Farmakis.” Jones and Gray offered responses in the climate survey that they believed Farmakis was hostile to them because they were women. But those survey responses did not discuss race, so they had limited relevance to the jury’s consideration of Downing’s racial discrimination claim. The district court had previously granted Abbott summary judgment on Downing’s sex discrimination claim, and she fails to acknowledge that was not an issue at trial. And, as the district court reasoned, there was a substantial risk of prejudice inherent in “having a bunch of complaints made by other people … offered for their truth.” The district court did permit Downing to introduce one statement from Gray’s climate survey response: “[Farmakis is] especially hard on Jay [Downing]. He embarrasses her and calls her out in calls or emails, part of the unfairness. Her 18 No. 21-2746

numbers are good so don’t understand why he calls her out, makes her feel stupid, don’t know why he’s doing that to her.” This response was important to Downing’s case, and her counsel referenced it during her closing argument. Consider- ing the district court allowed Downing to introduce this state- ment, despite its prejudice to Abbott, Downing fails to show the court abused its discretion by excluding other, less proba- tive excerpts of the survey responses. Even more, Downing overlooks that the district court admitted and the jury heard the most important survey-response evidence. Longoria—the decisionmaker in Employee Relations—testified she saw the PowerPoint presentation that summarized the climate survey responses. We cannot say that “no reasonable person would agree with the trial court’s ruling.” Aldridge, 635 F.3d at 875. The ex- clusion of the individual survey responses thus was not an abuse of discretion. And any error on this topic likely did not affect the trial’s outcome, as would be necessary for a reversal on this ground. Cf. Wilson, 932 F.3d at 522. D. Statistical Evidence Between 2012 and 2015, the number of African-American employees at Abbott Molecular decreased from 32 (4.7% of the company’s work force) to 16 (3%). After consideration, the district court excluded these statistics because they did not ac- count for employees who, rather than being forced out or ter- minated, left for other reasons, retired, or transferred to other positions in Abbott. Downing argues these statistics are probative of racial dis- crimination and that their exclusion was an abuse of discre- tion. To Downing, Longoria was a decisionmaker responsible No. 21-2746 19

for the disproportionate attrition of African-American employees at Abbott Molecular. Downing also contends the statistical evidence should have been permitted to rebut as- sertions by Abbott’s witnesses that they sought to increase di- versity at the company. Abbott responds that these data were not probative. Not only do they not identify which employees left for reasons other than termination, but they do not reveal which depar- tures, if any, were connected to the decisionmakers in this case—Farmakis and Longoria. For Abbott, that means the sta- tistics are insufficiently probative and unfairly prejudicial, and the district court properly excluded them. Statistical evidence may help an individual employee with a Title VII claim show that racial discrimination was an em- ployer’s standard operating procedure, but those statistical comparisons must involve a proper group. Matthews v. Waukesha Cnty., 759 F.3d 821, 829–30 (7th Cir. 2014). To sup- port her position, Downing relies primarily on two cases: Coleman v. Donahoe, 667 F.3d 835 (7th Cir. 2012), and Vega v. Chicago Park Dist., 954 F.3d 996 (7th Cir. 2020). But each dis- cusses statistical evidence in a Title VII setting only briefly. In Coleman, this court noted that a Title VII plaintiff may use “evidence, but not necessarily rigorous statistical evidence, that similarly situated employees were treated dif- ferently.” 667 F.3d at 860 (citation omitted). Nothing in that statement suggests statistical evidence must be admitted when it does not show whether employees were, in fact, treated differently. Similarly, in Vega we concluded that a jury’s verdict for a plaintiff had evidentiary support because the plaintiff showed that “no Caucasian park supervisors were fired [during the relevant period], while 17.6% of the 20 No. 21-2746

Park District’s Hispanic park supervisors were fired during that same period.” 954 F.3d at 1005. So, statistical comparisons must involve a proper group. At issue here is how many of the African-American em- ployees who left Abbott Molecular between 2012 and 2015 did so involuntarily. Downing failed to show how many of the employees were fired or otherwise forced out of Abbott. And the record established that at least some of them left to pursue careers elsewhere. On appeal, Downing admits she cannot identify how many African-American employees left the company to pursue other positions, including positions within Abbott. This concession persuades us that the district court did not abuse its discretion in excluding the statistical evidence. E. Harmless Error To Downing, these various alleged evidentiary errors had the individual and collective effect of denying her a fair trial. In support, Downing cites the jury’s first question: “Can we award punitive [damages] without finding ‘yes’ on any of the claims?” But the inference Downing advances—that the jury wanted to find for her but could not because of the eviden- tiary rulings—is conjecture. Speculation as to the evidence the jury did or did not rely on in reaching its verdict is just that, considering the jury never saw or heard the excluded evi- dence. By rule, with exceptions not relevant here, inquiry into a jury’s mental processes concerning a verdict is precluded. See FED. R. EVID. 606(b). Indeed, Downing prevailed in the discussion among the district court and counsel on how to respond to this first ques- tion. After the court read it to the parties’ counsel, Downing’s No. 21-2746 21

attorney spoke first, and her proposed answer—“No”—was adopted and relayed to the deliberating jury. The excluded evidence that Downing contends should have been admitted is also cumulative of other evidence the jury heard. “[E]rrors in admitting evidence that is merely cu- mulative of properly admitted evidence are harmless.” Jordan v. Binns, 712 F.3d 1123, 1138 (7th Cir. 2013) (citations omitted); see also Mason v. S. Ill. Univ. at Carbondale, 233 F.3d 1036, 1047– 48 (7th Cir. 2000) (same). The excluded pretext evidence was cumulative of Cerney’s extensive testimony, and the excluded portions of the climate survey responses were cumulative of the more probative portions that were admitted at trial. The proposed comparator evidence was weak, as neither Kohler nor Mason was even plausibly a comparator, so no harmless- error analysis is necessary on that point. And the statistical evidence had limited probative value because it did not ac- count for the circumstances of any individual employee’s de- parture from Abbott Molecular. Even if we were to determine that the statistical evidence should have been admitted, there is no reason to believe any error likely affected the outcome of the trial, so reversal is not warranted. See Wilson, 932 F.3d at 522. III. Expert Testimony We consider next whether the district court erred by ex- cluding the proposed opinion testimony of Dr. Destiny Peery about stereotyping and racial bias. Expert testimony is admissible when: (1) “the expert’s sci- entific, technical, or other specialized knowledge will help the trier of fact to understand the evidence or to determine a fact in issue; (2) “the testimony is based on sufficient facts or 22 No. 21-2746

data”; (3) “the testimony is the product of reliable principles and methods”; and (4) “the expert has reliably applied the principles and methods to the facts of the case.” FED. R. EVID. 702. Under Rule 702 and Daubert v. Merrell Dow Pharmaceuti- cals, Inc., 509 U.S. 579 (1993), the district court acts as the gate- keeper for expert evidence, C.W. ex rel. Wood v. Textron, Inc., 807 F.3d 827, 834 (7th Cir. 2015), evaluating the proffered ex- pert’s qualifications, the reliability of the expert’s methodol- ogy, and the relevance of the expert’s testimony. Kirk v. Clark Equip. Co., 991 F.3d 865, 872 (7th Cir. 2021). The party seeking to introduce expert witness testimony has the burden to show, by a preponderance of the evidence, that the testimony meets the Daubert standard. Gopalratnam v. Hewlett-Packard Co., 877 F.3d 771, 782 (7th Cir. 2017). We re- view de novo whether the district court properly applied the Daubert framework. The district court here did so: In ruling on Abbott’s motion to exclude Dr. Peery’s opinions, it ex- pressly relied on Rule 702 and Daubert. It then considered whether Dr. Peery’s proposed testimony was consistent with those authorities. The ultimate decision “to exclude or admit the expert wit- ness testimony” is reviewed “for an abuse of discretion only.” Id. (citation omitted); see also Textron, 807 F.3d at 835. A district court abuses its discretion when no reasonable person can agree with its decision. Antrim Pharms. LLC, 950 F.3d at 430. The district court considered two aspects of Dr. Peery’s testi- mony—the reliability of her methodology, and its helpfulness to the jury. Reliability. Downing has not offered an argument on ap- peal as to reliability. Even if she has not waived this point, the No. 21-2746 23

district court correctly determined that Dr. Peery’s methodol- ogy was not reliable. In assessing reliability, a district court may consider sev- eral factors, including the known or potential rate of error and the “existence and maintenance of standards controlling the technique’s operation.” Kirk, 991 F.3d at 873 (citation omitted). A court has “considerable leeway in deciding in a particular case how to go about determining whether particular expert testimony is reliable.” Manpower, Inc. v. Ins. Co. of Pa., 732 F.3d 796, 806 (7th Cir. 2013) (citation omitted). As this court has noted, a proposed expert must “bridge the analytical gap” by showing a “rational connection” between the data and the ex- pert’s contested conclusion. Gopalratnam, 877 F.3d at 786 (cit- ing Manpower, 732 F.3d at 809). Dr. Peery’s report begins by reviewing literature concern- ing stereotyping and discrimination. It then pivots, opining that negative assessments of Downing’s performance at Ab- bott are consistent with the possibility of stereotyping or bias. Dr. Peery does not substantively discuss the methodology she used. The connection between the data in the report and Dr. Peery’s opinions exists “only by the ipse dixit” of her as the witness. Id. at 781. The district court reasonably concluded that Dr. Peery’s methodology was unreliable, so her opinions were properly excluded. See id. at 786–87 (affirming the exclu- sion of expert opinion evidence on reliability grounds); Kirk, 991 F.3d at 877 (same); Textron, 807 F.3d at 837–38 (same). Helpfulness to the Jury. Dr. Peery’s proposed testimony was also excluded because it would not help the jury determine any fact at issue. Downing suggests this was not only errone- ous but an abuse of discretion that warrants reversal. Abbott responds that the district court correctly excluded Dr. Peery’s 24 No. 21-2746

opinions because she could not assess the probability of bias in any way. In a Daubert inquiry, the district court must evaluate the relevance of the expert’s testimony to a particular case. Kirk, 991 F.3d at 872; Gopalratnam, 877 F.3d at 779. For expert testi- mony to be admissible, the expert must have something “use- ful to say” about the particular circumstances at issue. Kunz v. DeFelice, 538 F.3d 667, 676 (7th Cir. 2008). To the district court, Dr. Peery opined “only that stereo- types ‘may,’ ‘might,’ or ‘could have’ played a role in Abbott’s decision making.” She did not offer a definite opinion as to whether Abbott discriminated against Downing. The court’s conclusion precisely tracks Dr. Peery’s putative testimony. For her opinions to help the jury, she needed to speak to whether Abbott’s decisionmakers racially discriminated against Downing. That requirement was not satisfied, the court correctly concluded, when Dr. Peery could not opine— with even five percent certainty—that anyone at Abbott made a decision about Downing’s employment on the basis of racial bias. See id. (affirming exclusion of expert witness’s testimony because his generalized testimony did not have anything use- ful to say about the particular facts in dispute). Dr. Peery’s testimony also would be unfairly prejudicial, as the court ruled, because jurors might incorrectly conclude she was of- fering an opinion on the ultimate liability question. For these reasons, the district court did not abuse its discretion by ex- cluding Dr. Peery’s opinion testimony. IV. Jury Instructions The district court also denied Downing’s requests for cer- tain jury instructions. We review those decisions for abuse of No. 21-2746 25

discretion. E.E.O.C. v. AutoZone, Inc., 809 F.3d 916, 921–22 (7th Cir. 2016); Aldridge, 635 F.3d at 876. To the extent “the case turns on a question of law,” our review is de novo. Kuberski v. Rev Recreation Grp., Inc., 5 F.4th 775, 779 (7th Cir. 2021) (cita- tion omitted). We will reverse “only if the instructions in their entirety so thoroughly misled the jury that they caused preju- dice.” Farnik v. City of Chicago, 1 F.4th 535, 544 (7th Cir. 2021) (internal quotation marks and citations omitted); accord Auto- Zone, 809 F.3d at 922. Statutory Instruction. Downing claims the district court abused its discretion by denying her proposed jury instruc- tion 2, which quoted portions of Title VII and 42 U.S.C. § 1981. This denial, according to Downing, resulted in the jury lack- ing knowledge of “what the anti-discrimination laws pro- hibit.” In declining to give this instruction, the district court stated, “I am inclined to agree with the defense view that we don’t need to include this statement of the purpose of the civil rights laws.” What was relevant, in the court’s view, was “to tell the jury what they need to find in order to render a verdict on the claims that are submitted in this case.” On this point, the jury was instructed: Ms. Downing must prove by a preponderance of the evidence that Abbott terminated her em- ployment because of her race. To determine that Abbott terminated her employment because of her race, you must decide that Abbott would not have terminated her employment had Ms. 26 No. 21-2746

Downing not been African American but every- thing else had been the same. We read this instruction to sufficiently convey what the anti- discrimination laws prohibit and what the jury was required to find to rule for Downing. The verdict form also asked the jury to determine whether Downing was (1) placed on performance-management measures; (2) subjected to termination; or (3) not rehired into a different role “because of her race.” Adverse employment actions, taken because of an employee’s race, are exactly what Title VII prohibits. See 42 U.S.C. § 2000e-2(a)(1); Barbera, 906 F.3d at 628. Even more, in response to the second jury ques- tion, the district court adopted Downing’s counsel’s sugges- tion to give the jury a copy of the statutory definition of an unlawful employment practice. Rather than being misled, we conclude that the jury was sufficiently informed about the ap- plicable law. See AutoZone, 809 F.3d at 921–23; Aldridge, 635 F.3d at 876. In view of the instruction that was given, the wording of the verdict form, and the response to the jury’s second question, the refusal to give Downing’s proposed jury instruction 2 was not an abuse of discretion. Circumstantial evidence instruction. Next, Downing believes the district court should have given her proposed jury instruc- tion 12. That provided a plaintiff “may prove discrimination or retaliation by offering different types of evidence,” includ- ing ambiguous statements toward other African American employees or evidence that Downing failed to receive desired treatment for which she was qualified. In its denial the court told Downing’s counsel: “You can tell [the jury], make your arguments about the probative value of the evidence and how it proves or does not prove discrimination or retaliation. No. 21-2746 27

We’re not going to build that into the jury instructions.” Ab- bott contends Downing failed to preserve this challenge and that the court’s instruction as to how the jury should weigh evidence was sufficient. The district court did not abuse its discretion here because it gave the circumstantial evidence instruction. After the close of evidence and argument, the court instructed the jury: You may have heard terms—the terms “direct evidence” and “circumstantial evidence.” Di- rect evidence is evidence that directly proves a fact. Circumstantial evidence is evidence that indirectly proves a fact, that is, evidence that re- quires an inference … You are to consider both direct and circumstantial evidence. The law does not say that one is better than the other. It is up to you to decide how much weigh to give to any evidence, whether it is direct or circum- stantial. Downing does not acknowledge or account for this instruc- tion, which the jury could have used to determine whether discrimination or retaliation had occurred. Stereotyping Instruction. Per Downing, the district court also abused its discretion by refusing to give proposed jury instruction 17: “Making an adverse employment decision be- cause of racial stereotypes is a form of race discrimination.” This prejudiced her, Downing submits, because the court precluded her theory of stereotyping for lack of an adequate evidentiary foundation, even though there was “ample evi- dence.” 28 No. 21-2746

Abbott responds that Downing forfeited this challenge, as on appeal she fails to contest the district court’s ruling that the instruction was unnecessary. She was also permitted to argue that Abbott’s actions were based on stereotyping, and Abbott sees no record support for this instruction. Downing is incorrect that the record is replete with evi- dence of stereotyping. She claims that Abbott engaged in bla- tant racial stereotyping because it termed her “defensive,” “negligent,” and lacking in “executive presence.” At oral ar- gument, her attorney made the puzzling suggestion that a manager who does not communicate well in front of custom- ers is a “horrible racial stereotype.” 2 Yet such descriptors are legitimate criticisms of an employee’s job performance. Downing provides no admissible evidence or authority that these labels are prevalent racial stereotypes. For this point she relies on Dr. Peery’s testimony, but that was appropriately ex- cluded under the Daubert standard. Given that Downing did not otherwise develop a factual record that would render such an instruction relevant, the district court did not abuse its discretion by denying it. See AutoZone, 809 F.3d at 922–23. “[A] judge need not deliver instructions describing all valid legal principles.” Id. at 923 (citation omitted). Instead, a trial court should generally allow a litigant to argue that a jury should draw a certain inference from the evidence. Id. The district court appropriately permitted Downing’s counsel to argue to the jury that Abbott based its adverse employment actions on racial stereotypes. For these reasons, the denial of Downing’s request for this jury instruction was not an abuse of discretion.

2 Oral Arg. at 10:45. No. 21-2746 29

Spoliation Instruction. Downing’s final challenge to the jury instructions involves spoliation. Farmakis testified at trial that he sent a survey to everyone in Abbott Molecular’s sales divi- sion to figure out what was going on within the business. Downing objected, arguing that the survey was not produced in discovery. Abbott responded that although the company could not locate the survey, Farmakis could still testify to his personal knowledge that he had conducted it and reviewed its results. So, the district court overruled the objection. Near the end of the jury-instruction conference, Downing requested “something along the lines of a spoliation instruc- tion” that would have told the jury that it could draw a nega- tive inference from Abbott’s failure to produce the document Farmakis had referenced during his testimony. The district court agreed with Abbott that there was an insufficient basis for a spoliation instruction. Downing contends that was an abuse of discretion. To obtain an adverse inference related to spoliation, a plaintiff must demonstrate that a defendant intentionally de- stroyed documents in bad faith. Perez v. Staples Cont. & Com. LLC, 31 F.4th 560, 569 (7th Cir. 2022); Norman-Nunnery v. Madison Area Tech. Coll., 625 F.3d 422, 428 (7th Cir. 2010). “The crucial element in a spoliation claim is not the fact that the documents were destroyed but that they were destroyed for the purpose of hiding adverse information.” Norman-Nun- nery, 625 F.3d at 428. Downing has not shown that anyone at Abbott destroyed the survey, which was taken in 2012—nine years before the trial—much less that any such destruction was for the purpose of hiding adverse information. Rather than an abuse of discretion, the district court’s refusal to give the jury a spoliation instruction was correct. 30 No. 21-2746

The district court properly relied on this circuit’s pattern jury instructions for Downing’s claims, and it did not abuse its discretion by rejecting the four jury instructions discussed above. V. Jowett Testimony and Mistrial Motion Downing further argues that the district court should have limited the trial testimony of her former supervisor Chris Jo- wett, or declared a mistrial based on his testimony. As noted above, we review a trial court’s decision to admit or exclude evidence for an abuse of discretion. Henderson v. Wilkie, 966 F.3d 530, 534 (7th Cir. 2020). A party “seeking to overturn the district court’s evidentiary ruling bears a heavy burden because a trial court’s balancing of probative value and unfair prejudice is highly discretionary.” Id. (cleaned up). An error in admitting evidence will not be a ground for a new trial unless justice requires otherwise, meaning the error must have prejudiced the aggrieved party’s substantial rights. Fed. R. Civ. P. 61; Stegall v. Saul, 943 F.3d 1124, 1127–28 (7th Cir. 2019). A district court’s denial of a mistrial is likewise reviewed for an abuse of discretion. Farnik, 1 F.4th at 542. In this context, we consider whether the district court committed an error of law or made a clearly erroneous finding of fact. Id. (citing Christmas v. City of Chicago, 682 F.3d 632, 638 (7th Cir. 2012)). The ultimate question is whether the appellant was denied a fair trial. Id. According to Downing, Abbott elicited testimony from Jo- wett as part of a strategy of “race baiting.” Downing sees a recurring theme at trial of Abbott appealing “to the jury’s emotions by asking it to vindicate the people Downing had No. 21-2746 31

accused of racism.” Based on this claim, Downing sought to preclude Jowett from testifying about his experiences with her, except for what he shared with decisionmakers. The district court stated it would not make a “blanket ruling” lim- iting Jowett’s testimony, although it would “judge the rele- vance in the context of the specific questions.” Abbott points out that after Downing’s initial objection and the court’s ruling, she did not renew her motion to ex- clude this testimony. Nor did she object to Jowett’s testimony about her skills and capabilities, his perception that her claims of racial discrimination lacked merit, and his feelings about her accusation that he perpetrated racial bias while at Abbott. Abbott submits that Downing therefore waived or forfeited the argument she now advances. Downing does not respond to Abbott’s contention that she failed to preserve this chal- lenge. “If the district court admits the contested evidence, the op- ponent must make a timely objection or motion to strike, stat- ing the specific ground of objection, if the specific ground was not apparent from the context.” Jimenez v. City of Chicago, 732 F.3d 710, 719 (7th Cir. 2013) (citations omitted); see also Griffin v. Foley, 542 F.3d 209, 218–19 (7th Cir. 2008). Downing’s failure to object to specific portions of Jowett’s testimony precludes her challenge on appeal. Further, arguments for excluding testimony, when made on appeal, are not preserved by mo- tions in limine that reference different arguments. Jimenez, 732 F.3d at 720. Downing argued to the district court that Jowett should be “limited to testify to what he shared with the deci- sionmakers.” But on appeal, she claims Jowett should not have been permitted to testify, despite his extensive personal knowledge of Downing’s performance, because he was part 32 No. 21-2746

of Abbott’s strategy of “race baiting” the jury. This gap be- tween Downing’s objection at trial and her argument on ap- peal means this challenge has not been preserved. See Jimenez, 732 F.3d at 720. Even if Downing had preserved the challenge to Jowett’s testimony, it would not form a basis to grant a new trial. A trial court’s decision to admit evidence as more probative than prejudicial is highly discretionary. Henderson, 966 F.3d at 534. Downing relays how she perceives Jowett’s testimony fit into the trial’s narrative. She also offers a conclusory state- ment that “Abbott’s conduct in baiting the jury” deprived her of a fair trial. But Downing cites no authority that a witness in Jowett’s managerial position, with his knowledge of her per- formance and capabilities, should not have been allowed to testify. As Abbott contends, Jowett’s testimony was properly ad- mitted in response to Downing’s testimony that his 2012 “ex- ceeded expectations” rating was evidence of her exemplary performance. That is the primary reason the district court gave for allowing Jowett to testify about his assessment of Downing’s performance. Downing does not respond to the basis for this ruling, so she has not shown it was an abuse of discretion. Instead, Downing claims “race baiting,” which she con- tends required a mistrial. Downing asserts that Jowett, when asked, testified that Abbott’s counsel told him that Downing had accused him of racial bias in his former position at Abbott. After Downing’s counsel moved for a mistrial on this basis, Abbott’s counsel stated, “I literally read to him the in- terrogatory response, full stop. That’s what I did.” In the in- terrogatory at issue, Abbott asked Downing to “identify the No. 21-2746 33

following Abbott executives who discriminated against Afri- can Americans,” to which Downing responded with a list that included Jowett. Downing’s counsel argued that Abbott had tainted the jury by asking witnesses, including Jowett, whether they were aware that she accused them of racism. The district court denied the motion for a mistrial, although it permitted Downing’s counsel to ask Jowett what Abbott’s at- torney had said to him. The district court made no error of law or clearly errone- ous finding of fact, so the denial of the motion for a mistrial must stand. See Farnik, 1 F.4th at 542. It was not unduly prej- udicial for Abbott to ask witnesses in a trial about racial dis- crimination if they were aware that the plaintiff had accused them of racial discrimination. In addition, the district court accepted the representation that Abbott’s counsel had only read the interrogatory response to Jowett. Downing offers no reason why doing so was not within the court’s discretion. It is not surprising that Jowett interpreted the interrogatory re- sponse as Downing accusing him of racism, because that is what it implies. Nothing about Jowett’s testimony or the cir- cumstances leading up to it deprived Downing of a fair trial. Relatedly, Downing contends the “race baiting” was com- pounded by the district court’s refusal to give proposed jury instruction 18. That instruction would have stated in part: “You have heard reference in opening statements that your role is to vindicate people and restore their reputations. You must disregard that statement, which is not evidence, nor is it an accurate statement of the law or your role or duty in this case.” The district court rejected this instruction for two reasons. First, it mischaracterized what counsel for Abbott said during 34 No. 21-2746

his opening statement, which was that a verdict for Abbott would vindicate the former Abbott employees who had been accused of discriminating against Downing. Second, other jury instructions, including those that told the jury not to con- sider sympathy, were sufficient to cure any prejudice. Both reasons are correct. Downing would not have been unfairly prejudiced by a comment that a defense verdict would vindicate Abbott employees whom Downing accused of racial bias. Juries are permitted to rely on their collective common sense. See Stragapede v. City of Evanston, Ill., 865 F.3d 861, 866 (7th Cir. 2017). This would include that a verdict for the company would reflect well on its employees. The district court also gave instructions that were sufficient to eliminate any possible prejudice. The jury was told: “Do not let sympa- thy, prejudice, fear, or public opinion influence you.” The court also instructed, “[the jury’s] concern is only whether Ms. Downing has proved that in taking the challenged employ- ment action Abbott discriminated against Ms. Downing be- cause of her race and/or retaliated against her for complaining about discrimination.” So, contrary to Downing’s contentions on appeal, the jury was instructed to disregard any effort to convince it to decide the case based on the feelings of the for- mer Abbott employees whom she accused of racial bias. For these reasons, a new trial is not warranted. Downing did not preserve her objections to Jowett’s testi- mony, and in any event the district court was within its dis- cretion not to exclude the testimony because it was relevant. No. 21-2746 35

Any error on the admission of Jowett’s testimony did not war- rant a mistrial. VI. Disparate-Impact Claim Last, Downing appeals the district court’s grant of sum- mary judgment to Abbott on her disparate-impact claim. We review the district court’s grant of summary judgment de novo, construing facts in the light most favorable to Down- ing and drawing all reasonable inferences in her favor. Khun- gar v. Access Cmty. Health Network, 985 F.3d 565, 572 (7th Cir. 2021) (citations omitted). “An inference is not reasonable if it is directly contradicted by direct evidence provided at the summary judgment stage, nor is a ‘conceivable’ inference nec- essarily reasonable at summary judgment.” MAO-MSO Re- covery II, LLC v. State Farm Mut. Auto. Ins. Co., 994 F.3d 869, 876 (7th Cir. 2021) (citing Cont’l Cas. Co. v. Nw. Nat. Ins. Co., 427 F.3d 1038, 1041 (7th Cir. 2005)). Summary judgment is appropriate “if the movant shows that there is no genuine dis- pute as to any material fact and the movant is entitled to judg- ment as a matter of law.” FED. R. CIV. P. 56(a). An employment practice may be unlawful under Title VII based on its disparate impact. 42 U.S.C. § 2000e-2(k). “Under a disparate impact theory, an employer is held liable when a facially neutral employment practice disproportionately im- pacts members of a legally protected group.” Farrell v. Butler Univ., 421 F.3d 609, 616 (7th Cir. 2005). A plaintiff must first show that the employment practice had an adverse impact on employees with a protected characteristic, such as race. Ernst v. City of Chicago, 837 F.3d 788, 796 (7th Cir. 2016). If the em- ployee makes such a showing, then the burden shifts to the employer to show its employment practice “is job-related for 36 No. 21-2746

the employee’s position and consistent with business neces- sity.” Id. (quoting 42 U.S.C. § 2000e-2(k)(1)(A)(i)). The district court concluded that the sample size of 10 ap- plicants for the director position was too small to be statisti- cally meaningful, so it granted Abbott summary judgment on Downing’s disparate-impact claim. Downing challenges the process Abbott used to rate candidates and extend offers dur- ing the rehiring process for the director position. Under that process, two African-American candidates—Downing and Jones—received the lowest ratings of ten candidates for the position. Downing argues she presented ample evidence that the subjective rating system Abbott used disparately impacted African-Americans. This included those white can- didates without managerial experience all received higher ratings than Downing and Jones. Abbott defends the dispar- ate-impact ruling because the sample size is so small that no inference of discriminatory impact would be proper. A plaintiff cannot make a prima facie case of adverse impact where the affected group is “too small for any valid statistical comparisons,” which “immunize[s] most single de- cisions from disparate impact challenges.” Council 31, Am. Fed’n of State, Cnty. & Mun. Emps., AFL-CIO v. Ward, 978 F.2d 373, 378 (7th Cir. 1992); see also Watson v. Fort Worth Bank & Tr., 487 U.S. 977, 996–97 (1988) (noting that “small or incom- plete data sets” prevent a plaintiff from making a prima facie case of adverse impact). Accord Morgan v. Harris Tr. & Sav. Bank of Chicago, 867 F.2d 1023, 1028 (7th Cir. 1989) (“Where the sample size or alleged effect is so statistically insignificant that no inference of discriminatory impact is proper, plaintiff fails to present a prima facie case.”); Vitug v. Multistate Tax Comm'n, 88 F.3d 506, 514 n.3 (7th Cir. 1996) (stating an No. 21-2746 37

employer’s hiring for four positions presented a “sample size” that was “too small for any meaningful statistical com- parison”). Because only 10 individuals applied for the direc- tor position, the district court reached that conclusion here. Downing asserts the district court ruled, incorrectly, that dis- parate-impact liability was per se unavailable because the sample size was too small. But she cites no authority for her assertion, and the case law holds otherwise. In response to this ruling, Downing points to her dispar- ate-treatment evidence. She highlights what she sees as dis- crepancies between the interview results and the applicants’ performance records. But Downing did not offer this argu- ment in the district court, and a contention that an employer did not fairly assess an individual applicant for a position is not an allegation of disparate impact. See Farrell, 421 F.3d at 617. The district court correctly granted summary judgment to Abbott on Downing’s disparate-impact claim. VII. Conclusion On each point Downing raises, the district court either ruled correctly or it did not abuse its discretion. Accordingly, we AFFIRM the district court’s judgment in all respects.