Although Wal-Mart Stores East, L.P. (Walmart) had unlimited light duty positions available under its Temporary Alternate Duty (TAD) Policy, and although the company offered TAD to 89 occupationally injured employees at Distribution Center #6025 during the relevant timeframe, Walmart refused to provide TAD to numerous similarly situated pregnant workers, forcing them to take leave instead, often unpaid. This Court should rule that the Equal Employment Opportunity Commission (EEOC) prevails outright or that EEOC has asserted sufficient evidence to reach a jury under the modified burden-shifting framework articulated in Young v. United Parcel Service, Inc., 575 U.S. 206, 135 S. Ct. 1338 (2015). Walmart failed to meet its burden of production at step 2 of the Young inquiry because its justifications did not explain “why, when [it] accommodated so many, could it not accommodate pregnant women as well?” Young, 135 S. Ct. at 1355. Moreover, at step 3, the record demonstrates that Walmart’s TAD Policy significantly burdened pregnant employees, and that Walmart’s stated justifications for excluding those employees cannot justify this burden.

Also, the district court abused its discretion by dismissing two claimants as a sanction for EEOC’s inadvertent discovery delays and by denying EEOC certain discovery related to Walmart’s TAD Policy.

I. This Court has jurisdiction to review the district court’s order denying EEOC’s motion for summary judgment.

Walmart incorrectly asserts that “[t]his court lacks jurisdiction under 28 U.S.C. § 1291 to review the district court’s denial of summary judgment.” Walmart Br. 1; Walmart Am. Statement of Jurisdiction 1. Walmart relies on Via v. LaGrand, 469 F.3d 618 (7th Cir. 2006), but Via, a qualified immunity case, stands for the proposition that “this court lacks interlocutory jurisdiction to review the district court’s conclusion [that] genuine issues of material fact exist” and its order denying summary judgment on that basis. Id. at 622 (emphasis added). Contrary to Walmart’s erroneous assertion, “[a]n appeal from the final judgment brings up all antecedent issues.” Reichman v. U.S. Fire Ins. Co. (In re Kilgus), 811 F.2d 1112, 1115 (7th Cir. 1987). Accordingly, this Court routinely reviews district courts’ final decisions on cross-motions for summary judgment. See, e.g., Markel Ins. Co. v. Rau, 954 F.3d 1012, 1016 (7th Cir. 2020).

II. EEOC should prevail under Young’s modified burden-shifting framework.

As the district court concluded, and as Walmart does not contest, Walmart Br. 35 n.5, EEOC established a prima facie case of pregnancy discrimination under Young by, among other things, highlighting evidence that Walmart accommodated non-pregnant workers who were “similar” to EEOC’s claimants “in their ability or inability to work.” EEOC Br. 28-30; see Young, 135 S. Ct. at 1343 (quoting 42 U.S.C. § 2000e(k)). The district court nevertheless granted Walmart’s summary-judgment motion and denied EEOC’s summary-judgment motion. This was error because (1) Walmart failed to meet its burden of production at step 2 and (2) even if Walmart satisfied step 2, EEOC marshalled sufficient evidence at step 3 to establish—or to allow a jury to find—that Walmart’s stated reasons for denying TAD to pregnant workers were “not sufficiently strong to justify” “a significant burden on pregnant workers.” Young, 135 S. Ct. at 1354.

A. Walmart did not meet its burden at step 2 of Young’s modified burden-shifting framework.

1. Young articulated a more specific burden of production at step 2.

Walmart disputes that Young established a “more specific” or “heightened” burden at step 2, Walmart Br. 38-46, but this argument fails. Young articulated a modified version of the burden-shifting framework set forth in McDonnell Douglas Corp. v. Green, 411 U.S. 792 (1973), for failure-to-accommodate suits under Clause 2 of the Pregnancy Discrimination Act (PDA), Pub. L. No. 95-555, 92 Stat. 2076 (1978) (codified at 42 U.S.C. § 2000e(k)). Under this framework, the “[u]ltimat[e]” question is “whether the nature of the employer’s policy and the way in which it burdens pregnant women shows that the employer has engaged in intentional discrimination.” Young, 135 S. Ct. at 1344. Therefore, as EEOC argued, EEOC Br. 30-35, the employer must explain why it did not accommodate pregnant employees to provide sufficient information at step 2 about “the nature of the [challenged] policy.” See Young, 135 S. Ct. at 1355 (courts must consider “why, when the employer accommodated so many, could it not [also] accommodate pregnant women”).[1]

The PDA’s “basic objective” of “overturn[ing]” General Electric Co. v. Gilbert, 429 U.S. 125 (1976), Young, 135 S. Ct. at 1353-54, further underscores that Clause 2 suits demand a more specific burden of production than courts require under the traditional McDonnell Douglas framework. Employers “normally” cannot discharge their burden of production by “simply” asserting “that it is more expensive or less convenient” to accommodate similarly situated pregnant workers. Young, 135 S. Ct. at 1354. Likewise, an employer’s justification explaining why it conferred a benefit on non-pregnant employees—but not why it excluded pregnant employees—is the type of justification that “the employer in Gilbert could … have made,” Young, 135 S. Ct. at 1354, and it is therefore insufficient. To be sure, the only two other appellate cases analyzing Young have not adopted this more specific version of step 2. See Walmart Br. 42. But this Court must follow Young, which rejected “simply” cost or convenience as legitimate non-discriminatory reasons, 135 S. Ct. at 1354, although they would suffice under a traditional McDonnell Douglas analysis. See, e.g., Furnco Constr. Corp. v. Waters, 438 U.S. 567, 577 (1978) (under traditional McDonnell Douglas framework, the employer’s step 2 burden involves “merely” establishing “that [it] based [the] employment decision on a legitimate consideration”). Moreover, in those two other appellate decisions, the parties and the court focused on the other steps of the Young inquiry, so the decisions did not consider the precise question presented here as to the step 2 burden. See Durham v. Rural/Metro Corp., 955 F.3d 1279, 1281, 1287 (11th Cir. 2020) (per curiam); Legg v. Ulster Cnty., 820 F.3d 67, 74-75 (2d Cir. 2016).

Walmart contends that EEOC’s characterization of step 2 improperly shifts the “ultimate burden” to defendants. Walmart Br. 39-40. EEOC agrees that plaintiffs bear the “[t]he ultimate burden of persuading the trier of fact that the defendant [engaged in] intentional[] discriminat[ion].” Tex Dep’t of Cmty. Affairs v. Burdine, 450 U.S.248, 253 (1981). However, although Young requires that employers provide more at step 2 in Clause 2 PDA suits than courts typically demand in traditional disparate treatment suits, that requirement does not alter the “ultimate burden.” If the defendant satisfies its burden of production at step 2, the burden shifts back to the plaintiff at step 3.

Along similar lines, in asserting that “decades-old” Title VII precedent applying the traditional McDonnell Douglas framework “does not require a defendant ‘to introduce evidence which … would persuade the trier of fact that the employment action was lawful,’” Walmart Br. 41-42 (quoting Burdine, 450 U.S. at 257-58), Walmart ignores that Young adapted the McDonnell Douglas framework to answer a particular question: how to analyze evidence to assess whether the employer’s provision of accommodations to non-pregnant employees “similar in their ability or inability to work” amounted to intentional discrimination under the PDA. Young, 135 S. Ct. at 1353-54. Young’s modification of McDonnell Douglas is consistent with courts’ adaptation of that framework on countless occasions to a wide range of different employment discrimination scenarios. See, e.g., Caraballo-Caraballo v. Corr. Admin., 892 F.3d 53, 59 (1st Cir. 2018) (“a principle established in one case” applying McDonnell Douglas “will not always translate to another”).

Nor does EEOC’s characterization of step 2 conflict with Young itself. Walmart claims that the language from Young on which EEOC relies does not “address[] the question of whether defendants must make a heightened showing at step two.” Walmart Br. 39; see EEOC Br. 31. But Walmart fails to grapple with Young’s admonition that certain justifications are insufficient at step 2. See 135 S. Ct. at 1354. EEOC never argued that “[t]he PDA’s overruling of Gilbert … require[s] the wholesale rejection of neutral justifications for limiting benefit policies to certain (as opposed to only) non-pregnant employees with a special claim to the accommodations at issue.” Walmart Br. 45. Instead, we explained that Young’s skepticism of justifications like those “the employer in Gilbert” could have asserted, 135 S. Ct. at 1354, supports EEOC’s characterization of step 2. EEOC Br. 30-35. After all, employers possess particular knowledge respecting the “nature of [their] polic[ies]” and the answer to the question why they did “not accommodate pregnant [workers]” “when [they] accommodated so many.” See Young, 135 S. Ct. at 1344, 1355. Step 2 is the appropriate juncture for employers to articulate that knowledge.

Contrary to Walmart’s assertion, EEOC’s characterization of step 2 does not contravene Young by “granting an unconditional ‘most-favored-nation’ status to pregnant employees that would require employers who provide some workers with an accommodation to also provide similar accommodations to all pregnant workers, irrespective of other criteria.” Walmart Br. 45 (citing Young, 135 S. Ct 1349-50). Walmart is correct that Young said the PDA does not confer “‘most-favored-nation’ status” on pregnant workers, but Walmart misunderstands the Supreme Court’s conception of that status. Young explained that when “an employer provides one or two workers with an accommodation—say, those with particularly hazardous jobs, or those whose workplace presence is particularly needed, or those who have worked at the company for many years, or those who are over the age of 55,” the PDA does not require the employer to “provide similar accommodations to all pregnant workers (with comparable physical limitations).” 135 S. Ct. at 1349-50. EEOC’s articulation of step 2 does not impose such a requirement; instead, EEOC’s view of step 2 is consistent with the idea that “the nature of [pregnant employees’] jobs, the employer’s need to keep [pregnant workers] working,” or “other criteria” may be relevant. Id. at 1350.

Walmart’s misunderstanding of step 2 is evident again when it asserts that “[e]xplaining ‘why [a defendant] provided a benefit to non-pregnant employees’ and ‘why it excluded pregnant employees from the benefit’ are two sides of the same coin.” Walmart Br. 43 (quoting EEOC Br. 30) (alteration in Walmart’s brief). Walmart cites Burdine, 450 U.S. at 254, in support, Walmart Br. 43, but the citation refers to the employer’s production burden under the traditional McDonnell Douglas framework, while, again, PDA failure-to-accommodate suits require a more specific showing at step 2. Along similar lines, Walmart says its reasons for accommodating occupationally injured employees establish a “clear” “implication” that Walmart excluded other employees from TAD “because associates without occupational injuries did not expose Walmart to the same liability and costs.” Walmart Br. 43. Again, it is not clear that accommodating one group meant Walmart could not accommodate others, especially when Walmart had unlimited light duty positions.

2. Walmart did not meet Young’s more specific burden of production.

a. As EEOC explained, EEOC Br. 32-35, because Walmart failed to satisfy step 2, summary judgment for EEOC is appropriate (or at least EEOC’s claims should go to a jury). During discovery and at summary judgment, Walmart consistently relied on the TAD Policy itself as a purported legitimate, nondiscriminatory justification. But the TAD Policy explains only why the company provided TAD to occupationally injured workers, not why Walmart refused to offer TAD to pregnant workers.

Walmart’s failure to explain why it did not provide TAD to pregnant employees is striking given record evidence that the company could have provided unlimited TAD positions at Distribution Center #6025, and Walmart’s inability to identify evidence that offering TAD to pregnant employees would have imposed any operational impact. In particular, EEOC pointed to witness testimony that there was no limit on the number of TAD assignments available to occupationally injured employees at Distribution Center #6025, and that extending TAD to pregnant employees in 2017 did not affect the Distribution Center’s ability to provide TAD to occupationally injured employees. EEOC Br. 34 (citing R.154 at 10, 29; R.155 at 8, 35). These witnesses were well-equipped to testify on this topic, contrary to Walmart’s suggestion otherwise. See Walmart Br. 52. Both were human resources employees with more than 15 years of tenure at Distribution Center #6025: (1) the center’s Human Resources Manager, who said he was the person at the center who was most knowledgeable about the TAD Policy (and whose declaration Walmart submitted at summary judgment), and (2) an employee whose responsibilities included “monitor[ing]” “TAD days” and “produc[ing] … job offers” for employees on TAD. R.150; R.154 at 3-5, 7-8; R.155 at 4-6. Walmart does not point to any record evidence contradicting these witnesses.

We noted that at least one of the considerations the TAD Policy cites as a reason for offering light duty to occupationally injured employees—increasing morale and loyalty—could also apply to pregnant employees. See EEOC Br. 33. Walmart claims that the “workplace-morale justification” does not apply to pregnant workers “with the same force,” but Walmart concedes that offering light duty to pregnant workers can boost morale and loyalty. Walmart Br. 50. Walmart attempts to undermine EEOC’s morale point by arguing that Federal Rule of Evidence 407 renders inadmissible (as a “subsequent remedial measure[]”) a document stating that Walmart extended TAD to pregnant employees in 2017 “to create a positive working environment for pregnant” employees. Walmart Br. 51 (quoting R.150-1 at 3). But even without Walmart’s document, the common-sense conclusion that providing TAD to pregnant employees would boost morale and increase loyalty still stands. Further, Rule 407 does not prevent EEOC from offering the document to show the “feasibility” of providing TAD to pregnant employees during the relevant timeframe. Cf. Adams v. City of Chicago, 469 F.3d 609, 612 (7th Cir 2006) (in disparate impact case, Rule 407 did not prohibit use of subsequently adopted promotion procedures as evidence of available procedures when employer made disputed promotions). Also, such evidence is admissible to undermine Walmart’s asserted reasons for excluding pregnant employees from TAD, not as an example of a “subsequent measure” “that would have made an earlier injury or harm less likely to occur,” or “to prove culpability.” Fed. R. Evid. 407.

b. Walmart attempts to rebut EEOC’s arguments that the company’s reliance on its TAD Policy did not provide a sufficient justification at step 2. First, Walmart now argues on appeal that it “owes a special duty” “to accommodate associates injured in the course of their employment,” which “explains both why Walmart limited TAD eligibility to occupationally injured associates … and why Walmart excluded all associates without work injuries (pregnant or not) during the relevant period.” Walmart Br. 46, 50. But at summary judgment, Walmart did not claim to have a “special duty” vis-à-vis occupationally injured employees that explained why Walmart excluded pregnant employees from TAD. See, e.g., R.151 at 31-33; R.169 at 11-18. In any event, Walmart provides no legal authority for a “special duty” of accommodating only occupationally injured employees that would satisfy step 2. Young specifically rejected one reason Walmart proffers for having such a duty—“costs”—while, as discussed, “low morale” and combating “high employee turnover” apply equally to pregnant employees.

Second, although Walmart states that the TAD Policy “clearly referred to the costs and liability that Walmart faces under state worker’s compensation laws,” Walmart Br. 47, employers have a responsibility at step 2 to offer a clear articulation of their justifications at the time for excluding pregnant workers, rather than rely on courts and plaintiffs to infer their true justification. Moreover, as explained, EEOC Br. 33-34, to the extent Walmart asserts that it limited TAD to occupationally injured workers to comply with state workers compensation law, see Walmart Br. 36-38, 47-48, this justification would not be sufficient at step 2. An employer’s obligation to comply with workers compensation law does not preclude granting pregnant workers light duty, and that obligation accordingly does not explain why the employer did not also “accommodate pregnant [employees].” 135 S. Ct. at 1355. We acknowledged in our opening brief that the Second Circuit in Legg accepted “compliance with a state workers’ compensation scheme” as “a neutral reason for providing benefits to employees injured on the job but not pregnant employees,” 820 F.3d at 75; EEOC Br. at 33-34, and, as Walmart notes, the Young employer “relied on [a workers compensation] justification,” Walmart Br. 48. But the parties in Young did not raise the question whether compliance with workers compensation regimes constitutes a sufficient justification, and Legg did not take into account Young’s admonition that certain rationales will not satisfy step 2 given the PDA’s purpose of overruling Gilbert. See EEOC Br. 31-32.

Finally, Walmart’s asserted justifications at step 2 are inadequate as a matter of law under EEOC v. Target Corp., 460 F.3d 946 (7th Cir. 2006), because Walmart failed to provide enough information “to frame an issue of fact so that the court and [the plaintiff] can identify what evidence might” satisfy step 3 of the inquiry. EEOC Br. 35 (quoting Target, 460 F.3d at 958); see also Figueroa v. Pompeo, 923 F.3d 1078, 1088 (D.C. Cir. 2019) (plaintiffs “cannot be expected to disprove [employer’s] reasons unless they have been articulated with some specificity”) (citations omitted). Walmart says Target is distinguishable because, there, the employer offered a “vague and generic” explanation for declining to hire an applicant, while here, “Walmart identified a series of specific justifications for limiting TAD accommodations to occupationally injured associates.” Walmart Br. 53-54. But as explained supra pp. 8-11, Walmart did not offer any reasons for excluding pregnant workers from the TAD program, as Young demands.

B. EEOC marshalled sufficient evidence at step 3 to establish that Walmart’s stated reasons for limiting its TAD policy were not “sufficiently strong” to “justify” the “significant burden” “imposed” on pregnant workers, or at least to allow a reasonable jury to conclude this.

EEOC argued that it offered sufficient evidence at step 3 of Young’s framework to justify summary judgment for EEOC, or, at a minimum, to create a factual question for a jury. EEOC Br. 35-45. The record shows that (1) Walmart’s TAD policy “impose[d] a significant burden on pregnant workers” and (2) Walmart’s proffered reasons for excluding pregnant workers from TAD were “not sufficiently strong to justify [that] burden.” Young, 135 S. Ct. at 1354. Walmart’s counterarguments misinterpret Young and mischaracterize the record.

1. Walmart’s TAD Policy imposed a “significant burden” on pregnant workers.

a. EEOC pointed to comparative evidence illustrating that Walmart’s policies “significant[ly] burden[ed]” pregnant workers by “accommodat[ing] a large percentage of nonpregnant workers while failing to accommodate a large percentage of pregnant workers.” EEOC Br. 36-41 (quoting Young, 135 S. Ct. at 1354). As Walmart does not dispute, during the relevant time period at Distribution Center #6025, it (1) denied light duty to 100% of pregnant workers who were “similar in their ability or inability to work” to occupationally injured employees and (2) granted TAD to 100% of workers—89 total over three years—who requested light duty for occupational injuries (and who could work). See EEOC Br. 37. Walmart’s “categorical[] fail[ure] to accommodate pregnant employees” who required light duty, Young 135 S. Ct. at 1354, while it provided TAD to numerous similarly situated occupationally injured workers, is enough on its own to establish a “significant burden” on pregnant workers. See Legg, 820 F.3d at 76 (“adverse[] [e]ffect[]” on sole pregnant employee “undoubtedly imposed a significant burden”).

b. Walmart repeats the district court’s error in asserting that EEOC was required to provide statistical analysis respecting other groups of employees who were ineligible for TAD. Walmart Br. 57-59. According to Walmart, EEOC misapplied Young by incorrectly comparing the percentage of accommodated pregnant women (0%) to the percentage of accommodated employees with on-the-job injuries (100%), instead of to the total percentage of employees who were accommodated (including employees injured on the job and employees with disabilities). Walmart Br. 57-58.

But, as explained, EEOC Br. 37-38, Young did not hold that plaintiffs are required to compare percentages of employees accommodated across different groups to establish significant burden, nor did Young demand any specific statistical proof. The comparative evidence here is not precisely the same kind of evidence present in Young, but the Young framework, like the McDonnell Douglas framework, is meant to accommodate a range of factual circumstances that arise in different cases. See EEOC Br. 38 (explaining that Legg, 820 F.3d at 76, deemed sufficient plaintiff’s evidence of “significant burden” even though record was “unclear” whether employer “accommodated a large percentage of non-pregnant employees”). Again, Young’s inquiry is whether the record—including comparative evidence contrasting pregnant and non-pregnant workers—establishes that “the employer’s policies impose a significant burden on pregnant workers.” Young, 135 S. Ct. at 1354. Here, therefore, it suffices that Walmart denied TAD to 100% of pregnant employees while providing TAD to a substantial group of non-pregnant workers.[2]

Finally, Walmart is wrong that accepting EEOC’s comparative evidence would (1) nullify the significant-burden inquiry because step 3 would “come out in plaintiffs’ favor” in every Clause 2 failure-to-accommodate suit or (2) improperly confer “most-favored nation status.” Walmart Br. 57-58. Although Young and Legg suggest that excluding pregnant employees from an accommodation will often impose a significant burden, see Young, 135 S. Ct. at 1354-55; Legg, 820 F.3d at 76-77, not every such policy will do so, see, e.g., Young, 135 S. Ct. at 1349-50 (providing a tailored accommodation to a small number of non-pregnant employees with hazardous jobs may not implicate the PDA). Here, of course, the denied accommodation particularly burdened pregnant workers, who could have remained at the Distribution Center performing TAD instead of missing weeks or months of work (and pay). See EEOC Br. 9-10. And again, Walmart misunderstands that “most-favored nation status” refers only to the rejected proposition that an employer must accommodate similarly situated pregnant workers if it accommodates “one or two workers.” Young, 135 S. Ct. at 1349-50; see supra pp. 6-7. Here, by contrast, Walmart provided TAD to 89 occupationally injured workers, or 100% of those eligible.

c. As EEOC argued, the district court erred in disregarding evidence that employees with disabilities were eligible for accommodations on the rationale that Walmart’s Accommodation Policy covered all employees. EEOC Br. 38-39. To be sure, Walmart’s Accommodation Policy covers “pregnancy-related disabilities” (as required by the Americans with Disabilities Act) and “medical conditions,” R.148-4 at 2-3. But the “accommodations” that Walmart asserts pregnant workers received, Walmart Br. 12-15, were (typically unpaid) leaves of absence or intermittent leave, not other paid duty or modified tasks that would have permitted the claimants to continue working. EEOC Br. 30-40 (citing record).

Moreover, “a plaintiff may rebut an employer’s proffered justifications by showing how a policy operates in practice.” Young, 135 S. Ct. at 1355. In practice, although several claimants were eligible for reassignment under the Accommodation Policy, they remained on leave because no suitable positions were available before their pregnancies ended. See EEOC Br. 39-40 (citing record). In practice, “reduced schedules” that supposedly accommodated “lifting restrictions,” Walmart Br. 14, did no such thing, as claimants on reduced schedules were forced on full-time leave after receiving lifting restrictions. See, e.g., R.157-5 at 19-20; R.157-16 at 13-14, 38-39. Alyssa Gilliam’s part-time position lasted a single weekend before Walmart forced her to start leave due to a lifting restriction. R.157-1 at 22-24, 35-36.

Walmart asserts that, in describing how the Accommodation Policy operated in practice vis-à-vis pregnant workers, EEOC “offers no more than baseless speculation and inadmissible hearsay,” but Walmart does not identify which record evidence it considers “speculation” or “hearsay.” Walmart Br. 62. The authorities Walmart cites state only that Federal Rule of Civil Procedure 56(e)(1) requires that affidavits offered at summary judgment must be based on personal knowledge and set forth facts admissible into evidence. See Walmart Br. 62-63. These authorities are inapposite; Walmart does not point to any affidavit on which EEOC relies, and EEOC in fact cited documentary evidence and deposition testimony showing that pregnant employees did not, in practice, receive transfers or other accommodations. EEOC Br. 39-40 (citing record). To the extent Walmart suggests that certain statements by Distribution Center personnel are hearsay, this is wrong because they are statements by a party-opponent. See Fed. R. Evid. 801(d)(2). Moreover, the statements are admissible to show the burden imposed on Walmart’s pregnant employees who were forced to take leave, see EEOC Br. 40-41, which does not implicate “the truth of the matter asserted.” Fed. R. Evid. 801(c)(2).

d. In addition to EEOC’s comparative evidence summarized above (enough on its own), EEOC offered other evidence illustrating the “significant burden” on pregnant employees. Young explained that comparative evidence may offer “circumstantial proof” undermining the employer’s asserted justifications for disparate treatment, 135 S. Ct. at 1355, but Young left open that other kinds of evidence may also show a “significant burden” on pregnant employees. See, e.g., Legg, 820 F.3d at 75-76 (relying on evidence other than comparative evidence in step 3 analysis).

First, EEOC offered evidence that Walmart’s denial of TAD to pregnant employees imposed “significant” personal “burden[s]” on claimants. Among other things, claimants experienced financial stress, some feared that Walmart would force them to take leave if their medical providers imposed lifting restrictions, and at least one claimant did not return after taking leave because she wanted to avoid repeating her experience. EEOC Br. 40-41 (citing record). Walmart claims that, if such evidence could establish a “significant burden,” “every pregnancy discrimination plaintiff would prevail under Young.” Walmart Br. 63. But “significant burden” is only one part of a balancing inquiry that also considers the strength of the employer’s justification. Moreover, EEOC does not argue that the claimants’ emotional burdens established a “significant burden” on their own; instead, EEOC also relies on comparative evidence showing that Walmart categorically denied TAD to pregnant workers while granting TAD to 89 occupationally injured workers.

Second, we highlighted statements of discriminatory animus toward pregnant workers at Distribution Center #6025, including comments that pregnant employees did not belong, could not work, or did not want to work at the Distribution Center. See EEOC Br. 13 (citing record). Walmart appears to challenge this evidence on the rationales that TAD was a nationwide policy and that the statements were “anecdotal.” See Walmart Br. 64. But it is commonsense that anti-pregnancy statements connected to an employer’s denial of accommodations to pregnant workers—like the statements EEOC highlighted here—could impose harm (that is, a burden) on pregnant workers. And a factfinder could readily interpret such statements as reflecting how Walmart’s national policy operated in practice.

2. Walmart’s stated reasons for excluding pregnant women from TAD were not “sufficiently strong to justify the burden imposed.”

Even accepting Walmart’s stated justifications as meeting step 2, they are not “sufficiently strong to justify” the “significant” “burden imposed” on pregnant workers. See EEOC Br. 41-45.

First, Young explicitly disapproved employer justifications predicated “simply” on cost and convenience, see Young, 135 S. Ct. at 1354, but Walmart cannot even show a valid cost or convenience rationale. Even if offering TAD to occupationally injured workers saved Walmart money, see e.g., Walmart Br. 2-3, 37, that does not explain why Walmart did not offer TAD to pregnant workers, and there was no limit on the number of light duty assignments available at Distribution Center #6025. See EEOC Br. 6, 34. Second, several of Walmart’s reasons for providing TAD would apply equally to pregnant employees, namely allowing employees to remain productive and promoting loyalty, morale, and a positive working environment. R.149-1 at 2 (A.102); R.150-1 at 3. Third, to the extent Walmart invokes state workers compensation requirements, Walmart Br. 65, those obligations did not “prevent[] [Walmart] from offering the same accommodation to pregnant employees.” Legg, 820 F.3d at 77.

The district court and Walmart make several fundamental missteps in analyzing Walmart’s justifications. The court mistakenly assumed that assessing the strength of Walmart’s justifications improperly “shift[s] the burden” to the employer at step 3. See R.204 at 22 (SA.34). But the employer’s justification is central to step 3, which involves balancing the “strength” of the employer’s reason against the burden on pregnant employees. Young, 135 S. Ct. at 1355. The district court further said EEOC did not identify evidence “call[ing] into question [Walmart’s] motives” for not offering TAD to pregnant employees, see R.204 at 24 (SA.36), and Walmart suggests the same (Walmart Br. 66), but the court apparently misunderstood that Young’s balancing test is meant to prove the employer’s intent. In this case, the pertinent inquiry is whether the employer’s stated reasons justify excluding pregnant employees, and EEOC does not need to make an additional showing that those reasons are not credible, the typical analysis under the traditional McDonnell Douglas framework. See Young, 135 S. Ct. at 1354-55; EEOC Br. 44-45. Walmart also argues that its justifications are stronger than the employer’s justifications in Young. In particular, Walmart states that it may rely on a “special duty” to occupationally injured workers, while the Young employer could not, because Walmart provided TAD only to workers injured on the job, while the Young employer offered light duty to occupationally injured workers and other groups of non-pregnant workers. Walmart Br. 65-66. But as explained, Walmart’s “special duty” claim does not succeed, and in any event this argument cannot be reconciled with Walmart’s assertion that EEOC must provide comparative evidence identical to the evidence in Young. See supra pp. 13-14.

III. This Court should reverse the district court’s ruling striking EEOC’s claimants.

EEOC argued that the district court abused its direction by striking claimants Shannon Sonnentag and Leah Hayworth as a discovery sanction for EEOC’s inadvertent belated production of records. EEOC Br. 45-51. This sanction amounted to outright dismissal of Sonnentag’s and Hayworth’s claims, as Walmart recognizes. See EEOC Br. 45; Walmart Br. 67-69. But neither the district court nor Walmart has identified a single authority (1) authorizing dismissal of an aggrieved individual’s claim in an EEOC enforcement suit as a sanction for EEOC’s discovery conduct or (2) deeming dismissal a proportionate sanction for inadvertent discovery conduct. Although this Court recognizes a high bar for reversal of district court sanctions, see Walmart Br. 67 (citing Maynard v. Nygren, 372 F.3d 890, 893 (7th Cir. 2004), the dismissal sanction here was “fundamentally wrong” and “clearly unreasonable,” Maynard, 372 F.3d at 893, as the district court made no finding of fault, acknowledged EEOC’s “inadvertent” conduct, never found prejudice, and imposed a disproportionate sanction.

As Walmart acknowledges, Walmart Br. 67-68, district court are “require[d] … to find that the responsible party acted or failed to act with a degree of culpability that exceeds simple inadvertence or mistake” when imposing “dismissal as a sanction for discovery violations.” Ramirez v. T&H Lemont, Inc., 845 F.3d 772, 776 (7th Cir. 2016) (emphasis added). Inherent-authority sanctions are permitted only for “willful[] abuse [of] judicial process” or “litigation in bad faith,” while Federal Rule of Civil Procedure 37(b)(2)(A) authorizes dismissal only for discovery conduct involving “willfulness, bad faith, or any fault.” Ramirez, 845 F.3d at 776. Contrary to Walmart’s suggestion otherwise, see Walmart Br. 68, EEOC did explain that “fault” encompasses “extraordinarily poor judgment or gross negligence rather than mere mistake or carelessness.” EEOC Br. 46 (quoting Ramirez, 845 F.3d at 776 (cleaned up)).

EEOC argued that the district court failed to make the “require[d]” finding that EEOC’s conduct met the inherent-authority or Rule 37(b)(2)(A) standards for dismissal. EEOC Br. 46. Instead, the district court “accept[ed] [EEOC’s] assertion” that its omissions respecting Hayworth’s and Sonnentag’s records “were inadvertent,” R.142 at 11 (SA.11), and, we argued, an “inadvertent” discovery delay does not constitute “culpability that exceeds simple inadvertence or mistake.” Ramirez, 845 F.3d at 776 (emphasis added). Other facts further underscore that EEOC’s error did not meet the dismissal standard. As Walmart fails to dispute, EEOC promptly cured its omissions upon discovering them and the volume of delayed documents relating to Hayworth and Sonnentag was relatively minor in light of the enormous document production in this lawsuit. See EEOC Br. 48.

Walmart does not contest that the district court made no explicit finding of requisite culpability, and does not assert that EEOC engaged in “willful[]” or “bad faith” conduct as required for inherent-authority sanctions. Walmart Br. 68-69. Therefore, Walmart presumably argues only “fault” under Rule 37(b)(2)(A), in contending that the district court implicitly found that EEOC’s purported “history of repeated discovery-order violations” established “extraordinarily poor judgment” or “gross negligence.” Walmart Br. 68; see also EEOC Br. 48-49 (rebutting the proposition that the court’s recapitulation of the history of discovery amounted to an implicit finding of sufficient culpability). According to Walmart, the “court’s reference to EEOC’s late productions being ‘inadvertent’ does not undermine the basis for sanctions” because EEOC engaged in a “pattern of noncompliance.” Walmart Br. 69 (quoting Newman v. Metro. Pier & Exposition Auth., 962 F.2d 589, 591 (7th Cir. 1992)). But Newman—where the plaintiff committed serious intentional misconduct by failing to appear for her deposition multiple times and neglecting to complete interrogatory responses—does not stand for the broad proposition that a pattern of inadvertent discovery noncompliance generally justifies dismissal. See 962 F.2d at 590-91. Also, although the district court warned EEOC that it might strike claimants for discovery conduct, see Walmart Br. 21-22, neither the district court nor Walmart has identified any authority justifying such warnings or holding that EEOC’s discovery conduct warrants dismissal of claimants.

To be sure, this Court reviews sanctions “in light of the entire procedural history of the case,” Walmart Br. 69 (quoting e360 Insight, Inc. v. Spamhaus Project, 658 F.3d 637, 643 (7th Cir. 2011) (cleaned up)). But considering the “entire procedural history” requires taking into account, among other things, Walmart’s own discovery delays; EEOC’s sound reasons to handle collection of claimants’ medical records (to avoid deterring claimants who worried about disclosure of sensitive but irrelevant information); the ambiguity of the magistrate judge’s original order setting a thirty-days-before-depositions production deadline (effectively giving Walmart control over production deadlines based on when it noticed depositions); that EEOC’s prior production delays largely stemmed from slow responses by third-party records companies; and—as the district court acknowledged, R.97 at 4-5, 7 (A.81-A.82, A.84)—that EEOC promptly corrected its prior production delays. See EEOC Br. 13-16, 21-24, 50 (citing record).

Walmart erroneously implies that EEOC’s allegedly “late” addition of claimants “triggered a series of discovery disputes in which EEOC ignored deadlines and failed to disclose relevant documents in a timely fashion.” Walmart Br. 16. But when EEOC identified new claimants in December 2019 and January 2020, Walmart had not even requested a claimant-identification deadline. R.12 at 1-6; R.67 at 1-3. Instead, Walmart engaged in delays respecting discovery related to claimant identification. Beginning in 2015, during its administrative investigation, EEOC sought information needed to identify potential claimants, but Walmart did not comply despite repeated requests and an administrative subpoena. R.67 at 2. In this lawsuit, EEOC served discovery in February 2019 requesting the same information; the magistrate judge granted in part EEOC’s motion to compel that information after Walmart failed to respond; and Walmart finally provided responses more than two months after the magistrate judge’s order. R.36 at 36-38; R.67 at 2. And although Walmart claims EEOC committed a “discovery-order violation,” Walmart Br. 21, when EEOC initially did not produce pre-2014 claimant medical records, the magistrate judge deemed EEOC’s time-limited production only a “technical[]” breach because, at that juncture, there was no “clear ruling” on the appropriate temporal scope of medical records. R.86 at 19 (A.63).

Finally, Walmart only briefly addresses the argument that the dismissal sanction was not “proportionate to the circumstances.” Donelson v. Hardy, 931 F.3d 565, 569 (7th Cir. 2019); see EEOC Br. 49-51; Walmart Br. 70. According to Walmart, Newman and this Court’s unpublished opinion in Jennings v. Principi, 114 F. App’x 224 (7th Cir. 2004), support dismissal. But Newman itself warns that judicial sanctions “must be guided by the norm of proportionality,” 962 F.2d at 591, and the discovery violations in both cases (plaintiffs’ repeated failures to appear for depositions) stand in stark contrast to EEOC’s inadvertent delays here. Id.; Jennings, 114 F. App’x at 226-27. Moreover, “the extent of [any] misconduct,” Donelson, 931 F.3d at 569, was minor. EEOC quickly corrected its “inadvertent” delay, the volume of delayed documents was small in the context of the enormous document production, and delayed productions by both parties were commonplace (not “misconduct” justifying dismissal). Nor does the record support the idea that “lesser sanctions” would have been “ineffective[],” id., especially where Walmart itself requested alternative relief short of dismissal. See Walmart Br. 25. Again, the court failed to cite any authority establishing that (1) dismissal is proportionate for inadvertent discovery conduct or (2) dismissal of EEOC claimants is an appropriate sanction for EEOC’s discovery conduct.

Finally, any “harm,” Donelson, 931 F.3d at 569, fell short of “prejudice that would justify a sanction like dismissal.” Wallace v. McGlothan, 606 F.3d 410, 427 (7th Cir. 2010). As EEOC explained, EEOC Br. 50, and as Walmart does not contest, the district court’s orders resolving Walmart’s motions to strike did not conclude that Walmart suffered prejudice. See R.97 at 3-7 (A.80-A.84); R.142 at 8-11 (SA.8-SA.11). Nor, for the reasons explained, EEOC Br. 50, does the record support such a determination.

IV. The district court abused its discretion by limiting the scope of discovery pertaining to Walmart’s TAD Policy.

Finally, EEOC argued, EEOC Br. 51-53, that if this Court reverses the district court’s ruling granting summary judgment to Walmart, it should also reverse the court’s ruling denying EEOC’s motion to compel discovery “inquir[ing] into who was responsible for the changes in [Walmart’s TAD] policy or what training may have occurred after Young was decided.” R.142 at 3-4 (SA.3-SA.4). Although the court allowed EEOC to obtain documents (1) “relating to why [Walmart’s TAD] policy originally excluded pregnant women” and (2) “explain[ing] why Walmart changed its TAD policy,” the court blocked EEOC from seeking non-document discovery on the very same topics. R.106 at 2 (A.90); R.142 at 4 (SA.4).

EEOC argued that the district court abused its discretion by limiting discovery based on an “erroneous conclusion of law” respecting Young. EEOC Br. 52 (quoting Sherrod v. Lingle, 223 F.3d 605, 610 (7th Cir. 2000)); see Walmart Br. 71-72 (acknowledging that EEOC asserted a “legal argument”). The district court failed to recognize the importance of evidence that may shed light on whether Walmart adequately explained why it excluded pregnant employees from TAD and whether Walmart’s justifications are sufficient. See EEOC Br. 30-32, 41-44. Contrary to Walmart’s suggestion otherwise, Walmart Br. 71, this error is relevant to step 2 and step 3 of the Young framework, because the strength of Walmart’s justification and its reasons for excluding pregnant workers from TAD inform the analysis at both steps. See EEOC Br. 32-35, 41-45.

Moreover, the discovery EEOC sought is pertinent to the availability of punitive damages, as we explained. EEOC Br. 53. Walmart claims EEOC did not previously assert this argument, Walmart Br. 72, but EEOC has sought punitive damages from the outset of this lawsuit, see R.1 at 6, and Walmart sought summary judgment on the issue of punitive damages, see R.151 at 33-34. EEOC’s briefs in support of its motion to compel explained that the discovery EEOC sought “may provide an alleged defense to Walmart’s practices during the relevant time period,” “may shed light on any alleged burden Walmart considered for the original exclusion and subsequent allowance of the participation of pregnant women in its TAD program,” and may permit EEOC to pose deposition questions to “personnel responsible for the change in the scope of employees eligible for TAD.” R.102 at 2-5; R.114 at 3, 6-8. All those inquiries are directly relevant to the question whether Walmart knew it “may be violating federal law,” Cooke v. Stefani Mgmt. Servs., Inc., 250 F.3d 564, 568 (7th Cir. 2001), between 2015 (when Young was decided) and 2017 (when Walmart implemented Pregnancy TAD).

Contrary to Walmart’s suggestion otherwise, see Walmart Br. 72, EEOC’s explanations of why EEOC sought this discovery also demonstrate why documentary evidence alone was insufficient and why the ruling prejudiced EEOC. See EEOC Br. 52-53. For example, the challenged ruling prevented EEOC from seeking testimony from decisionmakers who declined to amend the TAD Policy after Young, or who revised the policy to include pregnant workers in 2017. And the ruling denied EEOC access to discovery that may illuminate whether Walmart was ever justified in excluding pregnant employees from TAD, and/or whether that exclusion stemmed from anti-pregnancy animus. Finally, Walmart asserted (without explaining why) that the discovery EEOC sought would be inadmissible under Federal Rule of Evidence 407, Walmart Br. 72, but that argument fails for the reasons explained supra pp. 9-10.

CONCLUSION

For the foregoing reasons and for the reasons identified in our opening brief, this Court should reverse the district court’s summary-judgment, claimant-dismissal, and discovery rulings.

Respectfully submitted,

CHRISTOPHER LAGE

Deputy General Counsel

JENNIfer s. goldstein

Associate General Counsel

ANNE NOEL OCCHIALINO

Acting Assistant General Counsel

/s/ Anne W. King

ANNE W. KING

Attorney, Appellate Litigation Services

Office of General Counsel

Equal Employment Opportunity
Commission

131 M St. NE, Fifth Floor

Washington, DC 20507

(202) 921-2748

anne.king@eeoc.gov

February 10, 2022

[1] To clarify, EEOC does not advance two versions of step 2, as Walmart claims. Walmart Br. 38. The questions (1) “why, when the employer accommodated so many, could it not [also] accommodate pregnant women” and (2) “why [the employer] excluded pregnancy-related medical conditions from the accommodations policy at issue, rather than merely explaining its reasons for limiting eligibility to a subgroup of non-pregnant employees,” see id., capture the same standard expressed in slightly different ways.

[2] Moreover, although the district court and Walmart claim that EEOC must present evidence comparing pregnant workers with other groups of “non-pregnant” employees who did not receive TAD, the record does not establish that any members of these groups—workers with disabilities or occupational injuries—were similar to pregnant employees in their “inability to work.” See EEOC Br. 38-39.