Joseph Simpson v. Thomas J. Dart

23 F.4th 706

• Court: Court of Appeals for the Seventh Circuit
• Decided: January 6, 2022
• Docket: 21-8028
• Status: Published

Opinion

In the

United States Court of Appeals For the Seventh Circuit ____________________ No. 21-8028 JOSEPH SIMPSON, et al., Plaintiffs-Petitioners, v.

THOMAS J. DART, et al., Defendants-Respondents. ____________________

Petition for Permission to Appeal from the United States District Court for the Northern District of Illinois, Eastern Division. No. 1:18-cv-00553 — Sharon Johnson Coleman, Judge. ____________________

SUBMITTED SEPTEMBER 27, 2021 — DECIDED JANUARY 6, 2022 ____________________

Before EASTERBROOK, WOOD, and SCUDDER, Circuit Judges. SCUDDER, Circuit Judge. Joseph Simpson applied to work as a Correctional Officer at the Cook County Department of Corrections four separate times between 2014 and 2017. The Department declined to hire him each time. Simpson believed the hiring practices underlying these rejections violated his rights—and those of other unsuccessful Black appli- cants—under Title VII of the Civil Rights Act. Invoking 2 No. 21-8028

theories of both disparate treatment and disparate impact, Simpson’s class action complaint alleged that, through the use of a five-step hiring process for correctional officers, the De- partment of Corrections both intended to discriminate against Black applicants and succeeded in producing that discrimina- tory result. The district court denied Simpson’s motion for class certi- fication, finding that none of his proposed classes—a general class of all unsuccessful applicants and five subclasses of can- didates dismissed at each step of the hiring process—satisfied Rule 23(a)(2)’s requirement that they present “questions of law or fact common to the class.” But the district court’s anal- ysis seems at times to have conflated and merged Simpson’s disparate impact claims with his disparate treatment claims for intentional discrimination. While disparate treatment claims may require a more searching commonality inquiry along the lines of that performed by the district court, disparate impact claims most often will not: the common questions are whether the challenged policy has in fact disparately impacted the plaintiff class and, if so, whether that disparate impact is jus- tified by business necessity. Because the district court’s analysis did not clearly deline- ate its reasoning for declining to certify three of Simpson’s subclasses on a disparate impact theory, we grant Simpson’s petition for Rule 23(f) review, vacate the denial of class certi- fication as to these three subclasses, and remand to allow the district court to reconsider whether they should be certified. No. 21-8028 3

I A The Cook County Sheriff’s Office oversees all operations at the Cook County Department of Corrections, including the hiring of correctional officers. That hiring authority is dele- gated in large part to an administrative body within the Sher- iff’s Office known as the Cook County Sheriff’s Merit Board, although the Sheriff’s Office itself makes the final hiring deci- sion. Both entities—the Sheriff’s Office (through Sheriff Thomas J. Dart in his official capacity) and the Merit Board— are defendants in this case, as is Cook County itself. The Department of Corrections considers applicants by employing a five-step hiring process. Applicants may be elim- inated from contention at any step. The Merit Board controls the first four steps—(1) an initial written exam; (2) a written situational exam; (3) a physical fitness test; and (4) a more dis- cretionary “final review,” which itself appears to consist of a background check, drug testing, and multiple interviews. Ap- plicants who successfully complete each step are certified by the Merit Board as “eligible for hire” and proceed to step (5), a discretionary “file review” (and a polygraph test) con- ducted by the Sheriff’s Office, which then makes the final hir- ing decision. Simpson’s complaint alleged that the defendants insti- tuted and implemented this multistep process to discriminate against Black applicants. See 42 U.S.C. § 2000e-2(a)(1) (impos- ing liability on employers who “fail or refuse to hire … any individual … because of such individual’s race, color, reli- gion, sex, or national origin”). The complaint alternatively al- leged that, regardless of the defendants’ intent, the policies 4 No. 21-8028

did in fact disparately impact Black applicants, who were pur- portedly hired at significantly lower rates than white appli- cants. See id. § 2000e-2(k)(1)(A)(i) (prohibiting the use of “a particular employment practice that causes a disparate im- pact on the basis of race, color, sex, or national origin” and that is not “job related for the position in question and con- sistent with business necessity”). In discovery, Simpson pro- duced statistical evidence that he believes proves that, at each of the five steps of the hiring process, Black applicants re- ceived rejections more often than white applicants. B Simpson first moved to certify just one class of all unsuc- cessful Black applicants dating back to March of 2015. He later sought to add five subclasses for candidates rejected at each of the five challenged steps of the hiring process. The district court permitted Simpson to amend his complaint to add these proposed subclasses before it considered the motion for class certification. The defendants had a full and fair opportunity to oppose class certification. The district court acknowledged that Simpson’s proposed classes alleged violations based on theories of both “disparate impact and discriminatory intent.” In denying the certifica- tion motion “in its entirety,” the court’s analysis focused solely on Rule 23(a)(2)’s “commonality” requirement—that there must exist “questions of law or fact common to the class.” Fed. R. Civ. P. 23(a)(2). As to each of Simpson’s pro- posed classes under both his disparate impact and disparate treatment theories, the court determined that no such com- mon questions existed—a finding preventing certification. At no point, though, did the district court differentiate or No. 21-8028 5

separate its analysis of Simpson’s disparate impact claims from its consideration of his disparate treatment claims. Simpson has not sought review of the district court’s re- fusal to certify either the original combined class or the sub- classes for steps (4) and (5) of the hiring process—the Merit Board’s final review and the Sheriff’s Office’s file review. He instead seeks interlocutory review only as to the subclasses for steps (1), (2), and (3)—the initial written exam, the written situational exam, and the physical fitness test. We refer to these together as the exam subclasses. The district court analyzed steps (1) and (2), the two writ- ten exams, together and found that the class plaintiffs had “made little effort to establish that these standardized tests are racially biased.” The court observed that these tests had “been validated in other jurisdictions and by agencies throughout the country.” And while the plaintiffs had presented statisti- cal evidence indicating that Black applicants were rejected at higher rates than white applicants at both steps, the district court discredited this evidence because it “did not control for any racially neutral factors.” The district court’s analysis of step (3), the physical fitness test, was much the same. The court found that the plaintiffs had failed to “present evidence that the physical ability test was administered in a racially- biased manner,” and instead pointed only to statistical dis- parities in the pass rates for white and Black applicants. With his motion for class certification denied in its en- tirety, Simpson invoked Federal Rule of Civil Procedure 23(f) and sought our review. He contends that the district court erred in finding that, as applied to the three exam subclasses, his disparate impact claims did not present a common ques- tion capable of classwide determination. 6 No. 21-8028

II Promulgated by the Supreme Court pursuant to 28 U.S.C. § 1292(e), Rule 23(f) grants courts of appeals discretion to “permit an appeal from an order granting or denying class- action certification.” Fed. R. Civ. P. 23(f). We have often exer- cised that discretion where “deciding the appeal would clar- ify class action law.” Driver v. AppleIllinois, LLC, 739 F.3d 1073, 1076 (7th Cir. 2014) (citing Blair v. Equifax Check Servs., Inc., 181 F.3d 832, 835 (7th Cir. 1999) (explaining that the exercise of Rule 23(f) review is appropriate where “an appeal may facili- tate the development of the law”)). This is such a case. So we grant Simpson’s petition and vacate the district court’s judg- ment as to the three exam subclasses, making two observa- tions to help inform the district court’s analysis on remand. A First, Rule 23(a) enumerates four—and only four—re- quirements for class certification: numerosity, commonality, typicality, and adequacy of representation. See Fed. R. Civ. P. 23(a)(1)–(4). Likelihood of success on the merits is not among them. See Payton v. County of Kane, 308 F.3d 673, 677 (7th Cir. 2002). From there a class must fall within one of three permis- sible “types of class actions” set out in Rule 23(b). And if a class plaintiff satisfies all the requirements of Rule 23(a) and Rule 23(b), the class must be certified, even if it is sure to fail on the merits. See Shady Grove Orthopedic Assocs., P.A. v. All- state Ins. Co., 559 U.S. 393, 399–400 (2010) (explaining that, “in each and every case where the Rule’s criteria are met,” the plaintiff “may bring his claim in a class action if he wishes”) (cleaned up). Put another way, properly certified classes can lose as well as win. See Bennett v. Dart, 953 F.3d 467, 469 (7th Cir. 2020); Schleicher v. Wendt, 618 F.3d 679, 686 (7th Cir. 2010). No. 21-8028 7

To be sure, the Supreme Court has recognized that the “rigorous analysis” required at the class certification stage may “entail some overlap with the merits of the plaintiff’s un- derlying claim.” Wal-Mart Stores, Inc. v. Dukes, 564 U.S. 338, 351 (2011). But the Court has also taken care to emphasize that “[m]erits questions may be considered to the extent—but only to the extent—that they are relevant to determining whether the Rule 23 prerequisites for class certification are satisfied.” Amgen Inc. v. Conn. Ret. Plans & Tr. Funds, 568 U.S. 455, 466 (2013). Put most simply, “Rule 23 grants courts no license to engage in free-ranging merits inquiries at the certification stage.” Id. Front and center here is the second of Rule 23(a)’s four re- quirements—commonality. A plaintiff may represent a class only if “there are questions of law or fact common to the class.” Fed. R. Civ. P. 23(a)(2). A Title VII plaintiff can estab- lish liability on a disparate impact theory by pointing to an employment policy that “causes racial discrimination” and “is not justified by business necessity.” McReynolds v. Merrill Lynch, Pierce, Fenner & Smith, Inc., 672 F.3d 482, 489 (7th Cir. 2012); see also Ricci v. DeStefano, 557 U.S. 557, 577–78 (2009). To certify a disparate impact class, then, the plaintiff’s claims as to those three elements must be based on “common con- tention[s]” that are “capable of classwide resolution.” Wal- Mart, 564 U.S. at 350. The first element—identifying a discrete employment pol- icy to challenge—often presents difficulties for proposed dis- parate impact classes. In Wal-Mart, where the only “policy” the plaintiffs identified was one of “allowing discretion” in hiring decisions, the Court recognized the need for some mer- its inquiry at the class certification stage: a court must know 8 No. 21-8028

how that discretion is exercised to know whether there exists “a common mode of exercising discretion that pervades the entire company.” Id. at 355–56. In the end, the Court con- cluded that because the plaintiffs could not show that Wal- Mart supervisors exercised their challenged discretion in a common way, “it [would] be impossible to say that examina- tion of all the class members’ claims for relief will produce a common answer to the crucial question why was I disfavored.” Id. at 352 (emphasis in original). This difficulty may have doomed certification of Simp- son’s subclasses (4) and (5), as both the Merit Board’s final re- view and the Sheriff’s Office’s file review processes seem to involve multiple decisionmakers exercising discretion in var- ious unknown ways. We need not decide the issue, because Simpson has limited his petition to subclasses (1), (2), and (3)—the three exam subclasses. And those subclasses, it seems, concern challenges not to the exercise of discretion by multiple actors but to the uniform administration of standard- ized tests to each putative class member. This sort of “single [companywide] policy was the missing ingredient in Wal- Mart.” Bolden v. Walsh Constr. Co., 688 F.3d 893, 898 (7th Cir. 2012); see McReynolds, 672 F.3d at 490. But it may not be miss- ing here. Where, as here, a plaintiff identifies a discrete employment policy that allegedly results in discrimination, Title VII dis- parate impact claims are well suited for classwide adjudica- tion: the policy either disparately impacted the plaintiff class or it did not. To put it even more in terms of this case, once Simpson has identified the three challenged exams, the ensu- ing Rule 23 analysis requires no inquiry into the merits to de- termine that the remaining two elements of his disparate No. 21-8028 9

impact claims almost necessarily present questions common to the class—whether those exams in fact caused a disparate impact and, if so, whether the use of the exams was justified by business necessity. See McReynolds, 672 F.3d at 489 (ob- serving that “whether [the challenged policy] causes racial discrimination and whether it nonetheless is justified by busi- ness necessity are issues common to the entire class and there- fore appropriate for class-wide determination”). Accordingly, the district court’s analysis was not “limited to those aspects of the merits that affect the decisions essential under Rule 23.” Dancel v. Groupon, Inc., 949 F.3d 999, 1005 (7th Cir. 2019) (quoting Schleicher, 618 F.3d at 685). To assess whether Simpson’s claims presented common questions, the district court did not need to consider whether Simpson’s sta- tistical experts controlled for racially neutral factors or whether the tests had been validated for use at other depart- ments. These considerations go not to commonality but in- stead to whether the exam subclasses can ultimately succeed on the merits. We cannot know at this stage. But what we can say is that, as to each subclass, the answer seems likely to be the same for each class member. That is all that Rule 23(a)(2) requires. See Wal-Mart, 564 U.S. at 352; In re Allstate Corp. Se- curities Litig., 966 F.3d 595, 604 (7th Cir. 2020) (“At class certi- fication, the issue is not whether plaintiffs will be able to prove these elements on the merits, but only whether their proof will be common for all plaintiffs, win or lose.”). B That brings us to our second observation. Some of the con- fusion in the district court’s opinion appears to stem from not separating its analysis of Simpson’s disparate impact claims from its assessment of his disparate treatment claims. 10 No. 21-8028

The claims are not one and the same. Indeed, today they root themselves in different provisions of Title VII. See 42 U.S.C. § 2000e-2(a)(1) (disparate treatment); id. § 2000e-2(k)(1)(A)(i) (disparate impact). Commonality may be harder to establish for disparate treatment claims, as those claims require “proof of intentional discrimination, which is not an element of a disparate impact claim.” McReynolds, 672 F.3d at 483. Simpson has not peti- tioned for review of the class certification decision as to his disparate treatment claims. Accordingly, we can assume without deciding that it was reasonable for the district court to find that Simpson had not shown that the question of intent was a common one that would yield a common answer across each proposed class. See Messner v. Northshore Univ. HealthSystem, 669 F.3d 802, 811 (7th Cir. 2012) (“Plaintiffs bear the burden of showing that a proposed class satisfies the Rule 23 requirements … by a preponderance of the evi- dence.”) (citations omitted). Even so, class certification is not an all-or-nothing propo- sition. Certification may be appropriate as to some of the class’s claims but not others. This observation flows from the text of Rule 23 itself: an order certifying a class “must define the class and the class claims, issues, or defenses.” Fed. R. Civ. P. 23(c)(1)(B) (emphasis added). If a class’s definition was syn- onymous with its claims, this language would be superfluous. Instead, the “class” takes its definition from the people in it and the period of time during which the relevant conduct oc- curred, and the “class claims” are the legal claims for which the class satisfies all of Rule 23’s certification requirements. See Wachtel ex rel. Jesse v. Guardian Life Ins. Co. of America, 453 F.3d 179, 187–88 (3d Cir. 2006) (holding that a Rule 23(c)(1)(B) No. 21-8028 11

certification order must include two elements: “(1) a readily discernible, clear, and precise statement of the parameters de- fining the class or classes to be certified, and (2) a readily dis- cernible, clear, and complete list of the claims, issues or de- fenses to be treated on a class basis”). Sometimes those requirements may be satisfied as to some of the class representative’s claims but not others. Only the former become “class claims.” See Fed. R. Civ. P. 23(c)(4) (“When appropriate, an action may be brought or maintained as a class action with respect to particular issues.”). So the fact that the district court found that Simpson could not show that, as to the three exam subclasses, his disparate treatment claims were “capable of classwide resolution,” Wal-Mart, 564 U.S. at 350, does not mean that the same must be true for his dispar- ate impact claims. For their part, the defendants insist that the district court’s failure to separate the disparate impact and disparate treat- ment claims is Simpson’s own fault. Simpson, they say, con- flated his disparate impact and disparate treatment theories and focused his efforts primarily on the latter. The defendants therefore ask us to find the disparate impact claims waived. We decline to do so. Simpson’s disparate impact claims have been in this case since its inception, and he briefed them in his motion for class certification. Nothing prevents a Title VII plaintiff from alleging both that a challenged policy caused a disparate impact and that it was intended to do so. And the district court recognized that Simpson had done just that, calling his claim one for “disparate impact and discrim- inatory intent.” 12 No. 21-8028

But the district court’s analysis did not reflect that recog- nition. When conducting class certification analyses, district courts should assess the four Rule 23(a) factors as to each of the putative class claims. See Santiago v. City of Chicago, 19 F.4th 1010, 1017 (7th Cir. 2021) (vacating a class certification order where the district court “analyzed the proposed classes under each provision [of Rule 23], but never made clear which claims the analyses refer to”). A one size (or one claim) ap- proach is at odds with the “rigorous analysis” required at the class certification stage. Wal-Mart, 564 U.S. at 351. Instead, as we recently explained, a district court should begin by identi- fying the elements of the plaintiff’s various claims: “[o]nly by properly circumscribing the claims and breaking them down into their constituent elements can a district court decide which issues are common, individual, and predominant.” Santiago, 19 F.4th at 1018. * * * We GRANT Simpson’s petition, VACATE the denial of class certification in part, and REMAND for the district court to reconsider certification of the three exam subclasses in light of our observations and the remaining Rule 23 factors. Whether Simpson’s three exam subclasses will ultimately merit certification is a question for the district court to answer in the first instance.