CHICAGO TEACHERS UNION, LOCAL NO. 1, American Federation of Teachers, AFL-CIO, Plaintiffs-Appellants, v. BOARD OF EDUCATION OF the CITY OF CHICAGO, Defendants-Appellees, 797 F.3d 426

ROVNER, Circuit Judge. In the ongoing pursuit to improve the quality of the Chicago Public Schools (CPS), the Chicago Board of Education (“Board”) has implemented various systems and processes to improve the quality of education for children. One process involves reconstituting schools that the Board deems to be deficient. Such a reconstitution or “turnaround,” as it is known colloquially, involves removing and replacing all administrators, faculty, and staff from a selected school and relieving the local school council of certain duties.

Then, the Board either contracts with a third to operate the school, assigns the school to the Board’s Office of School Improvement, or turns it over to one of the nineteen geographic networks that make up the next layer of leadership in the Chicago School Board system. I. The Illinois School Code provides that a school may be subject to turnaround if it has been on probation for at least one year and has failed to make adequate progress in correcting deficiencies. 105 ILCS 5/34-8.3(d)(4).

Pursuant to the collective bargaining agreement between the Chicago Teachers Union and the Board, tenured teachers affected by reconstitution are placed in a reassigned teachers’ pool where they continue to receive a full salary and benefits for one school year. If a tenured teacher does not find a new position within that year, she is honorably terminated unless her time in the pool is extended.

Probationary appointed teachers, other teachers, and para-professionals are not placed in the reassigned teachers’ pool but are eligible for the cadre pool where they can receive substitute assignments for which they are paid per assignment. Tenured teachers who are not reassigned within a year are also eligible for the cadre pool. Teachers in the cadre pool continue to receive health benefits for one year and receive a higher rate of payment than those in the ordinary substitute pool. Between 2004 and 2011, the Board reconstituted sixteen CPS schools.

In autumn 2011, the Board began considering which schools would be subject to a new round of reconstitution. Oliver Sicat, the head of CPS’ portfolio office, led the process, at the end of which the CPS CEO, Jean Claude Brizard, made final recommendations to the Board, all of which were accepted. The CEO initially identified 226 schools that had been on probation for at least one year — the baseline eligibility for turnaround under Illinois law.

He then reduced the list to seventy-four schools by removing schools that met the objective criteria of a composite Illinois Standard Achievement Test (ISAT) score above the network average for elementary schools or a five-year graduation rate above network average for high schools. Brizard was responsible for selecting the final ten schools for turnaround and presenting those selections to the Board for a vote.

The district court described this process as “qualitative” and the Board asserted that the CEO used “subjective criteria.” According to Ryan Crosby, the Manager of School Performance at the relevant time, the decisions were not made on the basis of a written policy or on one particular set of factors.

Nevertheless, Crosby testified that the CEO and other participants in the decision-making considered factors such as academic performance, performance trends, leadership, whether the school was over or under utilized, proximity to and effect on other schools, school culture, facilities, safety, parent and community input, and input from CPS staff.

The meeting participants who analyzed each school in sessions called “deep dives” included CEO Brizard, Chief Portfolio Officer Sicat, Network Chiefs, the Chief Academic Officer, Noemi Dono-so, and Board staff responsible for areas such as safety, transportation, facilities, academic performance and special education. R. 63-3, pp. 54, 62 (ID# 869, 877); R. 69-3, Declaration of Denise Little, app. ex. 4, pp. 2-3 (ID# 1201-02); R. 69-3, Declaration of Harrison Peters, app. ex. 3, pp. 2-3 (ID# 1196-97). Some of the factors considered in evaluating a school’s candidacy for turnaround are decidedly objective.

A school’s academic trends, for example, are measured by its performance points score. Performance points are calculated by considering, among other things, standardized test scores, school attendance rates, academic progress, and improvement over time in comparison with other schools in the same geographic network. For high schools, the dropout rate, “freshman on track” rate, and success in advanced placement programs are also included in the performance points. The Board gave particular weight to improvements trends. A school that was on probation but improving was much less likely to be selected.

Individual employees’ performance ratings, years of service, and performance of students in a teacher’s individual classroom were not taken into account. At a February 2012 Board briefing, the CEO recommended ten schools for turnaround — two. high schools and eight elementary schools. The briefing set forth the detailed rationale for selecting each school and included the factors listed above. Some schools received even more detailed attention.

Casals, which was considered a “priority school” was slated for turnaround because it had an overall low performance, and student achievement was growing at a slower pace when compared with similar students at other schools, despite having received much assistance during its five years on probation. The briefing also set forth CPS’s response to community feedback it had received in opposition to the proposed turnaround at Casals. The Board voted to authorize the reconstitution of all ten schools as recommended. On June 30, 2012, the Board terminated all' teachers and staff from those, ten schools.

The ten schools were located exclusively on the south and west sides of Chicago where African Americans' make up 40.9% of tenured teachers. No schools were selected for turnaround on the north side, where only 6.5% of tenured teachers are African American. Of the tenured teachers displaced because of reconstitution, 51% were African American, despite comprising just 27% of the overall teaching population within CPS. In hard numbers, 213 African-American employees were displaced. The racial demographics at the ten reconstituted schools varied as shown in the table below.

School % African-American teachers Smith 88.6 Woodson 85 Stagg 83.7 Fuller 81 Herzl 75.6 Chicago Vocational 75 Tilden 57.4 Piccolo 39.1 Marquette 26.7 Casals 26.7 Board’s brief, p. 13. Plaintiffs Donald J. Garrett Jr., Robert Green, and Vivonell Brown, Jr., three African-American tenured teachers affected by the turnarounds, and the Chicago Teachers Union, Local No. 1, filed suit against the Board, alleging that the Board’s decision to reconstitute these ten schools was racially discriminatory.

Plaintiffs sought to certify a class of: All African American persons employed by the Board of Education of the City of Chicago as a teacher or para-professional staff, as defined in the labor agreement between the Chicago Teachers Union and the Board of Education, in any school or attendance center subjected to reconstitution, or “turnaround,” on or after the 2012 calendar year. R. 63, p. 2 (ID# 817). The proposed class consists of African-American staff in the following positions: 32 para-professionals, 11 probationary appointed teachers, 163 tenured teachers, and 7 teachers with no tenure status.

As of the briefing for this appeal, half of the 32 para-professionals displaced by the 2012 turnarounds were currently active employees, 7 of the 11 probationary appointed teachers were current employees, and 122 of the 163 tenured teachers were currently active CPS teachers. Board’s brief, pp. 11-12. African-American teachers and para-professionals displaced in the 2012 turnarounds also include teachers who have retired, who are on leaves of absence, and those no longer employed by the Board. The named plaintiffs sought class certification under Federal Rules of Procedure 23(b)(2), (b)(3) and/or (c)(4).

The district court denied class certification on May 27, 2014. Although it found that the class met the requirements for numerosity, typicality, and adequacy of representation, the district court found that the plaintiffs had not met their burden of establishing a common issue by a preponderance of the evidence. It also found that plaintiffs had „not adequately shown that common questions of law or fact predominated over individual claims as required by 23(b)(3), and that there was no basis for issue certification under Federal Rule of Civil Procedure 23(c)(4). II.

The purpose of class action litigation is to avoid repeated litigation of the same issue and to facilitate prosecution of claims that any one individual might not otherwise bring on her own. The district court’s task below was to determine if the plaintiffs-appellants presented a scenario in which judicial efficiency would be served by allowing their claims to proceed en masse through the medium of a class action rather than through individual litigation. Our analysis is not free-form, but rather has been carefully scripted by the Federal Rules of Civil Procedure.

For this reason, the civil procedure rules on class actions are the best place to begin. Before we turn to those rules, however, we note that this case comes to us from a district court order denying the certification of the class. Chicago Teachers Union, Local 1 v. Bd. of Ed., No. 12 C 10311, 301 F.R.D. 300, 304 (N.D.Ill. May 27, 2014), hereinafter “Order.” Our review of such a decision is deferential. CE Design Ltd. v. King Architectural Metals, Inc., 637 F.3d 721, 723 (7th Cir.2011). “We review class certification orders for abuse of discretion.

Abuse of discretion results when a district court commits legal error or makes clearly erroneous factual findings.” Reliable Money Order, Inc. v. McKnight Sales Co., Inc., 704 F.3d 489, 498 (7th Cir.2013). Deferential review can and must also be exacting. “A class may only be certified if the trial court is satisfied, after a rigorous analysis, that the prerequisites” for class certification have been met. CE Design, 637 F.3d at 723. The decision to certify a class or not can cause a considerable tilt in the playing fields of litigation and therefore is not one to take lightly. See id.

The party seeking certification bears the burden of demonstrating that certification is proper by a preponderance of the evidence. Messner v. Northshore Univ. HealthSystem, 669 F.3d 802, 811 (7th Cir.2012). A. Because a class action is an exception to the usual rule that only a named party before the court can have her claims adjudicated, the class representative must be part of the class and possess the same interest and suffer the same injury. Wal-Mart Stores v. Dukes, — U.S. -, 131 S.Ct. 2541, 2550, 180 L.Ed.2d 374 (2011). The general gate-keeping function of Federal Rule 23(

a) ensures that they are. All class actions, no matter what type, must meet the four explicit requirements of Federal Rule of Civil Procedure 23(a): (1) the class is so numerous that joinder of all members is impracticable (numer-osity); (2) there are questions of law or fact common to the class (commonality); (3) the claims or defenses of the representative parties are typical of the claims or defenses of the class (typicality); and (4) the representative parties will'fairly and adequately protect the interests of the class (adequacy of representation). Fed.R.Civ.P. 23(a) (parentheticals ours).

In addition to meeting these requirements of Rule 23, a class action must meet the requirements of one of the four categories in Rule 23(b). Rule 23(

b) sets forth the various requirements for class actions depending on, among other things, the type of relief sought. In this case, the plaintiffs sought certification under Rule 28(b)(2), (b)(3), and/or (c)(4), the requirements of which we will discuss after addressing the threshold requirements of 23(a). On appeal, the only remaining contested factor from Rule 23(

a) is commonality — whether “there are questions of law or fact common to the class.” Fed. R.Civ.P. 23(a)(2). Although a court need only find a single common question of law or fact (Wal-Mart, 131 S.Ct. at 2556), the mere occurrence of all plaintiffs suffering as a result of a violation of the same provision of law is not enough. Id. at 2551; Suchanek v. Sturm Foods, Inc., 764 F.3d 750, 755 (7th Cir.2014). The claims must depend upon a common contention that is capable of class-wide resolution. Wal-Mart, 131 S.Ct. at 2551.

In this context, class-wide resolution means that determining the truth or falsity of the common contention will resolve an issue that is central to the validity of each claim. Id. at 2551. The majority in Wal-Mart summed this up by stating: What-matters to class certification ... is not the raising of common ‘questions’— even in droves — but, rather the capacity of a classwide proceeding to generate common answers apt to drive the resolution of the litigation. Dissimilarities within the proposed class are what have the potential to impede the generation of common answers.

Id. at 2551 (emphasis in original) (citing Nagareda, Class Certification in the Age of Aggregate Proof, 84 N.Y.U.L.Rev. 97,131-132 (2009)). In Wal-Mart, a proposed class of all of the 1.5 million women who work or worked at the company alleged that the company discriminated against them on the basis of gender by denying them equal pay or promotions.

The Supreme Court reversed the certification of the class, finding that the plaintiffs could not bear the burden of demonstrating commonality when the employment decisions complained of by the plaintiffs resulted from millions of individual decisions made by low-level decision-makers who had been given full discretion over such matters. “Without some glue holding the alleged reasons for all those decisions together, it will be impossible to say that examination of all the class members’ claims for relief will produce a common answer to the crucial question why was I disfavored.” Id. at 2552 (emphasis in original).

That “glue,” the Wal-Mart majority explained, could be something such as a biased employment testing procedure or a general policy of discrimination established by top managers, but the facts of the case provided neither. Id. at 2553. To the contrary, as the court noted, the only relevant corporate policy was one forbidding discrimination and a policy of delegating employment decisions to local managers. Id. at 2554.

The Board argues that the facts here align with those in Wal-Mart — that is that the decision to reconstitute the schools was not made pursuant to a central uniform policy or even by a single decision-maker, but rather was based on “subjective, qualitative factors that were not uniformly applied.” Board’s brief, p. 19. And indeed the district court found that the “turnaround policy, to the extent there was one, was not well-defined or uniformly applied,” and therefore, “Plaintiffs’ proposed class fail[ed] to meet the commonality requirement.” Order, p. 11.

The district court concluded that if the turnaround decision had been made based solely on an objectively measurable criteria applied across the board, it could find a common issue, but because the decisions were made using qualitative, subjective, case-by-case review, commonality failed. Order, p. 10-11. Before we delve into the questions of whether first, the review was really case-by-case and second, whether subjective review dooms commonality, we should unpack the process through which a school was selected for reconstitution.

Recall that the process of identifying schools for reconstitution consisted of three steps. First, the CEO identified all of the schools eligible by state law for reconstitution due to poor past performance, that is, the school had been on probation due to low academic performance for at least one year. 105 ILCS 5/34 — 8.3(d). Then the CEO reduced that list of 226 schools to 74 schools by removing those that met the objective criteria of a composite ISAT score above the network average for elementary schools, or a five-year graduation rate above network average for high schools.

The third step is the one that the district court focused on most: in this step the CEO and other high-level board members attended a series of meetings in which they discussed the types of information that the group would consider concerning schools eligible for reconstitution, and then analyzed that information. The first question we ask, therefore, is if the latter subjective steps (assuming they are indeed subjective and individualized) destroy the alleged commonality created by the first clearly-objeetive steps. The Board and the district court’s reasoning assume that they do. But this cannot be so.

Suppose hypothetically that after the objective first and second steps, all of the schools remaining on the list had 100% African-American teachers, and no schools with white teachers remained on the list. We could undoubtedly conclude that the objective factors had a disparate impact on African-American teachers. Suppose that the Board went on to evaluate those 74 schools with all African-American teachers in a subjective, case-by-case manner to narrow the list from 74 to 10 — all of which still were made up of African-American teachers.

The introduction of subjective, case-by-case criteria would not alleviate the disparate impact of the initial objective criteria. Surely we would say that the plaintiffs could allege that there was sufficient commonality to establish a class. Every one of those teachers could answer the question, “why was I disfavored?” by pointing to the initial objective criteria that impacted only African-American teachers. This is why the plaintiffs point to Connecticut v. Teal to argue that a discriminatory intermediate step taints the entire process. Id., 457 U.S. 440, 102 S.Ct. 2525, 73 L.Ed.2d 130 (1982).

In Teal, the employer required those seeking a promotion to take a test. Id. at 443-44, 102 S.Ct. 2525. Although objective on its face, the test eliminated far more African-Americans than white candidates. Ultimately, the employer (faced with the lawsuit, it seems) promoted a disproportionately high number of African Americans to supervisor positions. The court determined that despite the fact that the bottom-line result was non-discriminatory, the plaintiffs established a prima facie showing of a discriminatory impact. Id. at 455-56,102 S.Ct. 2525.

It is true that Teal was not a class certification case, but to the extent the Board asks us to ignore the impact of the objective steps of the test, it is directly on point, particularly because “class determination generally involves considerations that are enmeshed in the factual and legal issues comprising the plaintiffs cause of action.” Comcast Corp. v. Behrend, — U.S. -, 133 S.Ct. 1426, 1432, 185 L.Ed.2d 515 (2013).

Teal helps to answer the question of whether a class can be certified where the alleged class of plaintiffs claims they were all harmed similarly in an early step of the process even if, under Wal-Mart, they cannot point to sufficient glue to bind their claims under a later part of the process. Teal instructs that an early discriminatory process can taint the entire process, and indeed our hypothetical demonstrates why this must be so.

And it certainly is more efficient to answer the question “did these early discriminatory processes have a disparate impact on race” just one time rather than over and over again in multiple separate lawsuits. In short, if the plaintiffs allege that the objective criteria in the first two steps narrowed the pool in such a way as to have a disparate impact on African-American teachers (and indeed they do), then this is the glue that binds the claims together without regard to the later, subjective step.

But even if, when evaluating the propriety of class certification, we were to ignore these initial objective steps in deciding which schools would be reconstituted, we would still have to conclude that the district court erred in applying the law of the Wal-Mart case to these facts. The Wal-Mart decision simply does not preclude class certification where subjective decision-making and discretion is alleged.

The district court, however, seemed to read Wal-Mart to say that certification of a class is not possible when the acts complained of are based on subjective discretionary factors made by multiple decision-makers. Our post-Wal-Mart decision in McReynolds, however, makes clear that this is not so. McReynolds v. Merrill Lynch, Pierce, Fenner & Smith, Inc., 672 F.3d 482 (7th Cir.2012).

In McReynolds, 700 African-American brokers accused Merrill Lynch of racial discrimination in pay by structuring team work and account distribution policies in such a way that had a disparate negative impact on African-American brokers. The “teaming” policy allowed brokers to form teams to share clients and commissions. Once formed, the team could decide whom to admit as a new member. Brokers could still work alone, but membership in a team was an undisputed advantage.

Under the account distribution policy, when a broker left the employ of Merrill Lynch, the other brokers within the branch office could compete for the accounts left behind by the exiting broker. According to company policy, thq managers were to award the accounts based on the competing brokers’ records of revenue generated for the company and the number and investments of clients retained. It turned out that team members tended to choose other team members who were most like themselves, and thus white brokers (who were the vast majority) seldom chose African-American colleagues for their teams.

And without the help of the teams, African Americans did not generate as much revenue or attract and retain as many clients as white brokers, thus reducing their chances of winning account distribution competitions. McReynolds, 672 F.3d at 488. Merrill Lynch, like Wal-Mart, delegated discretion over decisions that influence compensation — including decisions involving the teaming and account distribution policies — to 135 lower-level directors. On its face, these facts sound similar to those in the Wal-Mart case where the Supreme Court found no commonality in the claims.

This court found, however, that although the local lower-level managers had a measure of discretion with regard to teaming and account distribution, the exercise of that discretion was influenced by the two company-wide policies — one authorizing brokers rather than managers to form and staff teams, and the other basing account distributions on past success — that allegedly exacerbated racial discrimination. Id. at 489.

We held that this established sufficient commonality for a class certification such that the question as to whether these policies created a disparate impact on African Americans could be resolved most efficiently in one claim. Id. at 491. In doing so, we noted that if, instead, Merrill Lynch had delegated to local management the decision to allow teaming, the case would more closely resemble Wal-Mart. Id. at 489-90. In contrast, just a few months later in Bolden v. Walsh Constr.

Co., we reversed a grant of class certification where the facts fell on the other side of the line— reflecting discretionary decisions more in line with the Wal-Mart decision rather than McReynolds. Id. 688 F.3d 893 (7th Cir.2012). In Bolden, twelve African-American plaintiffs alleged that Walsh Construction tolerated racial discrimination in assigning overtime work and in working conditions. Id. at 894-95. They asked the district court to certify two different classes of African-American employees, covering all of Walsh’s 262 projects in the Chicago area going back several years.

This court overturned the certification of the class finding that the sites all had different superintendents, different policies, different working conditions, and ranged in the amount, if any, of discriminatory practices. Id. at 896, 898. Just as in Wal-Mart, Walsh had a company-wide non-discrimination policy and granted discretion to superintendents to assign work and address discrimination that occurred on the site. Id. at 898.

Thus the Supreme Court’s Wal-Mart decision and ours in McReynolds and Bolden together demonstrate that a company-wide practice is appropriate for class challenge even where some decisions in the chain of acts challenged as discriminatory can be exercised by local managers with discretion — at least where the class at issue is affected in a common manner, such as where there is a uniform policy or process applied to all. The Fourth Circuit (relying heavily on this Circuit’s

interpretation of Wal-Mart) summed it up well by noting that Wal-Mart did not set out a per se rule against class certification where subjective decision-making or discretion is alleged. Rather, where subjective discretion is involved, Wal-Mart directs courts to examine whether all managers exercise discretion in a common way with some common direction. Thus, to satisfy commonality, a plaintiff must demonstrate that the exercise of discretion is tied to a specific employment practice, and that the subjective practice at issue affected the class in a uniform manner. Scott v.

Family Dollar Stores, Inc., 733 F.3d 105, 113 (4th Cir.2013) (internal citations omitted), cert. denied, — U.S. -, 134 S.Ct. 2871, 189 L.Ed.2d 848 (2014). And indeed, even the district court acknowledged that “if a general policy that is enforced at the corporate level rather than by individual supervisors is claimed to be discriminatory, even if some discretion exists, commonality may be found.” Order p. 9.

In short, subjective, discretionary decisions can be the source of a common claim if they are, for example, the outcome of employment practices or policies controlled by higher-level directors, if all decision-makers exercise discretion in a common way because of a company policy or practice, or if all decision-makers act together as one unit. The Board maintains that no single criteria was used in the third step to narrow the field of seventy-four schools to ten, but this is not an entirely accurate description.

More precisely, one could say that each of the twenty-six schools chosen for reconstitution was chosen after being considered individually. This does not mean that a different selection criteria was used. For example, suppose a company has decided to reduce its workforce by cutting the lowest performing 25% of workers. To evaluate performance, it looks to sales, evaluations, work ethic, and peer reviews. The CEO terminates one worker because her sales numbers are low, another because her evaluations from her supervisor are sub-par, and yet another because of high absenteeism.

Although it is true that each employee was terminated for different reasons, it is not true that a different set of criteria were used for each. In fact, the employer implicitly considered each factor for each employee, even if only some of the performance criteria ultimately determined the employee’s fate. In this case, the Board tells us that after the objective, numerical calculations in steps one and two, it considered a number of factors. Those factors were discussed in a series of meetings that included a small group of key people with information about the various factors considered.

The group included the Board Chief Academic Officer, the Chief Portfolio Officer, Network Chiefs, and representatives from Board departments in charge of transportation, facilities, safety, and special education. In its brief, the Board describes the numerous factors considered in the various schools, but they could be boiled down to the following broader categories: academic performance, performance trends, leadership, whether the school was over or under utilized, proximity to and effect on other schools, school culture, facilities, safety, parent and community input, and input from CPS staff.

See Board’s brief, pp. 5-11. We know that this small group of decision-makers, even during the third and subjective stage of decision-making, used these same criteria to assess each school because they told us so again and again.

See, e.g., Id. at p. 4 (“Selecting the schools for turnaround in 2012 involved a lengthy recommendation process that considered the academic performance of schools that were eligible for turnaround, whether those schools’ performance improved over time, and whether measures that had been implanted in the school were working.”); Id. at p. 6 (“selecting the schools that would be reconstituted from those 74 schools was a qualitative process guided by subjective criteria that various stakeholders were asked to consider.

For example, ... transportation, facilities, safety and special education ... planned school actions such as closures and phase-outs.”); Id. at p. 7 (“These discussions included not only the academic performance of schools ... but also issues such as leadership and the culture of a school, gang boundaries, overall performance, the condition and utilization of facilities and the observable teaching in a particular building.”); Id. at p. 8 (committee considered improvement while on probation and school culture); Id. at p. 9 (“The briefing noted that the selection process considered information involving school culture, safety, facility quality, community feedback and targeted input from CPS staff.”)- See also R. 53-2, Deposition of Ryan Crosby, p. 28 (ID# 859). (“There was not one set of factors that necessarily made a — each—in every school that was recommended for reconstitution and appropriate candidate [sic] but things such as the academic culture of the school, whether or not quality instruction was being provided, whether or not there was good leadership in the school, the — in general as I said, the academic trends of the school, the quality of implementation of programs that were in existence.”).

Id. at pp. 28-29 (ID# 859-60) (describing academic trends as comprised of academic standardized test scores, the attendance rate, dropout rate, “freshman on track” record, enrollment and success in advanced placement classes, and a standardized academic progress assessment); Id. at p. 62 (ID# 877) (“input from community members and the chiefs — the network chiefs of schools based on their feedback provided to the portfolio office.”); Id. at 75 (ID# 878) (enrollment and utilization data); Id. at p. 79 (ID# 882), (location was one of the factors considered); R. 741, Crosby Dep. p. 71-72 (ID# 1604-05) (“talking with Network Chiefs, in talking with community members about what was going on in the schools to identify from that list of 80 what were a likely set of possible actions.”); R. 69-3, Declaration of Denise Little, app. ex. 4, p. 3 (ID# 1202), (factors considered included “academic performance ... leadership at the schools, the culture of a school, gang boundaries, overall performance, the condition of and underutilization of facilities and the observable teaching in a particular building”); R. 69-3, Declaration of Harrison Peters, app. ex. 3, p. 3 (ID# 1197) (factors considered included “academic performance ... leadership at the schools, the culture of the school, gang boundaries, overall performance, the condition of and utilization of facilities and the observable teaching in a particular building,” and input from parents); R.. 69-3, February Board Member Briefing, p. 4 (ID# 1208) (“school culture, climate, safety, facility quality, community feedback and targeted information from CPS staff.”).

The Board goes on to state that there was a “specific, unique rationale for each turnaround decision” (Board’s brief, p. 10), but the examples they offer come from the same set of criteria that they identified as applicable to all schools. We can boil these criteria down to the following ten categories: (1) academic performance, (2) performance trends, (3) leadership, (4) whether the school was over or under utilized, (5) proximity to and effect on other schools, (6) school culture, (7) facilities, (8) safety,- (9) parent and community input, and (10) input from CPS staff.

For example, the Board states that Fuller and Woodson were selected to provide support for a nearby school that was closing— criteria # 5 on our list. At Smith, the local school council had asked for better options — criteria # 9 on our list. The Board chose Casals because of its culture of complacency and poor quality instruction — criteria # 6 and # 3. We could continue through each school, but need not. It is clear that the Board applied the same set of criteria to all of the schools evaluated for reconstitution.

In this way, the scenario in this case is worlds away from that in Wal-Mart where a court could have no way of knowing why each of the thousands of individual managers made distinct decisions regarding promotions and pay in millions of employment decisions. Likewise, in Jamie S. the task of identifying disabled students who might need educational services fell to countless school district employees making highly individualized decisions about the need for services in individual students. Jamie S. v. Milwaukee Pub.

Schs., 668 F.3d 481, 496 (7th Cir.2012); but see Id. at 504 (Rovner, J. dissenting) (“I believe that notwithstanding the inherently child specific nature of child-find inquiries, a class action based on a truly systemic child-find failure may be viable.”) Here we have one decision-making body, led by a CEO with ultimate authority to recommend schools to the Board, using one set of factors to analyze -the need for turnaround in each school.

When a small group of decision-makers sits together in a room comparing and contrasting the success of schools in order to evaluate their ultimate fate, the concept of a uniform criteria and single-decision maker merge. They are of one mind, using one process. In short, we do not have myriad actions of individual managers. Here we have one decision-making body, exercising discretion as one unit, with the ultimate decision in the hands of one single person, CEO Brizard. R. 53-2, p. 62 (ID# 877).

Decisions by myriad low-level managers are different than decisions made by a single lead decision-maker or a few concentrated top-level managers as lower-level employees do not set policies for the entire company; whereas, when high-level personnel exercise discretion, resulting decisions affect a much larger group, and depending on their rank in the corporate hierarchy, all the employees in the company.

Consequently, discretionary authority exercised by high-level corporate decision-makers, which is applicable to a broad segment of the corporation’s employees, is more likely to satisfy the commonality requirement than the discretion exercised by low-level managers in Wal-Mart. Scott, 733 F.3d at 114. The plaintiffs have demonstrated commonality by asserting that a uniform employment practice (the set of criteria used to evaluate the school) used by the same decision-making body to evaluate schools was discriminatory. Wal-Mart, 131 S.Ct. at 2551, 2554.

See also, Bolden, 688 F.3d at 899, (“Walmart observes that it may be possible to contest, in a class action, the effect a single supervisor’s conduct has on many employees.”).

And in fact, the district court noted the same thing, when it said that “if a general policy that is enforced at the corporate level rather than by individual supervisors is claimed to be discriminatory, even if some discretion exists, commonality may be found.” Order, p. 9 (citing McRey-nolds, 672 F.3d at 488-91, and Scott, 733 F.3d at 114.) Yet the district court lost track of this principle when finding that the plaintiffs had not met their burden of establishing commonality because the selection process was qualitative and lacked uniformity. Order, p. 10.

The district court erred, therefore, when it stated that “[t]he Court could not resolve whether the Board’s turnaround policy was discriminatory as applied to all class members ‘in one stroke,’ for it would have to examine the rationale behind the decision to turn around each of the ten schools and compare those reasons to the decisions not to pursue the remaining sixty-three.” Order, p. 11. This is not so. The court need only resolve whether the “same conduct or practice by the same defendant gives rise to the same kind of claims from all of the class members.” Suchanek, 764 F.3d at 756.

And just as in McReynolds, whether employment practices “cause racial discrimination ... are issues common to the entire class and therefore appropriate for class-wide determination.” McReynolds, 672 F.3d at 489. B. Having concluded that the plaintiffs demonstrated sufficient commonality to fulfill the threshold requirements for a class action elucidated in Federal Rule 23(a), we now turn to the plaintiffs request for certification under Federal Rule 23(b)(2).

Rule 23(b)(2) permits class certification if “the party opposing the class has acted or refuses to act on grounds that apply generally to the class, so that final injunctive relief or corresponding declaratory relief is appropriate respecting the class as a whole.” Fed.R.Civ.P. 23(b)(2); Lewis v. City of Chicago, 702 F.3d 958, 962 (7th Cir.2012). Colloquially, 23(b)(2) is the appropriate rule to enlist when the plaintiffs’ primary goal is not monetary relief, but rather to require the defendant to do or not do something that would benefit the whole class.

Not surprisingly, “civil rights eases against parties charged with unlawful, class-based discrimination are prime examples” of Rule 23(b)(2) classes. Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 614, 117 S.Ct. 2231, 138 L.Ed.2d 689 (1997). In this case, the plaintiffs sought a declaratory judgment that the Board’s turnaround policies violated Title VII, 42 U.S.C. §§ 1981, 1983, and prospective in-junctive relief including a moratorium on turnarounds and the appointment of a monitor to evaluate and oversee any new turnaround process. R 63-1, p. 21 (ID# 841).

The 23(b)(2) class does not seek any money or individual relief. The district court held that a 23(b)(2) class could not be certified because “[although Plaintiffs’ request for a declaration that the turnaround policy violates federal law would apply class-wide, it would merely be a prelude to further relief, which would be inherently individualized.” Order, p. 16.

The order pointed out that no single injunction could provide relief without establishing a system for providing individualized relief to each class member “either placing class members in specific jobs based on their qualifications and openings or providing them with back pay and front pay if no position was available.” Id. at 17. The district court erred, however, by misunderstanding the nature of the relief sought. The proposed 23(b)(2) class did not seek individual relief such as reinstatement or individually calculated damages in the form of back pay and front pay.

It asked only that the court issue a declaration that the Board’s turnaround practice violated Title VII and 42 U.S.C. §§ 1981 & 1983, and for prospective injunctive relief including a • moratorium on turnarounds and the appointment of a monitor to evaluate and oversee any new turnaround process. We agree with the district court that to the extent that “each individual class member would be entitled to a different injunction or declaratory judgment against the defendant,” 23(b)(2) certification would not be appropriate. Johnson, 702 F.3d at 369-70 (emphasis in original).

But the 23(b)(2) plaintiffs here seek the same declaratory and injunctive relief for everyone. This class-wide relief is different from the individual equitable and monetary relief the plaintiffs seek through their Rule 23(b)(3) class action, including reinstatement and front pay. The Board replicated the district court’s error in its briefing before this court, spending several paragraphs describing the complexities required for providing individualized relief. See Board’s brief, pp. 26-27 (describing the difficulties in reinstating teachers with various experience, certifications, and damages).

But this is all frolic and detour. An order enjoining the board from reconstituting schools would provide the exact relief that the 23(b)(2) class requests. A moratorium would prevent a recurrent violation (see Milwaukee Police Ass’n v. Jones, 192 F.3d 742, 747 (7th Cir.1999)) and would provide prospective relief against an allegedly discriminatory practice. Wal-Mart, 131 S.Ct. at 2552, n. 7. Group relief is particularly appropriate because the Board did not individually assess any of the putative class members in the process of reconstituting the school and displacing the teachers.

Each was displaced because of the Board’s uniform reconstitution policies and practices. Moreover, the fact that the plaintiffs might require individualized relief does not preclude certification of a class for common equitable relief. Pella Corp. v. Saltzman, 606 F.3d 391, 395 (7th Cir.2010); Arreola v. Godinez, 546 F.3d 788, 801 (7th Cir.2008); Allen v. Int’l Truck and Engine Corp., 358 F.3d 469, 471-72 (7th Cir.2004). “It is routine in class actions to have a final phase in which individualized proof must be submitted.” Suchanek, 764 F.3d at 756.

See also Johnson, 702 F.3d at 369 (In a 23(b)(2) class action, “a declaration is a permissible prelude to a claim for damages.”). The district court conceded that “[p]laintiffs’ request for a declaration that the turnaround policy violates federal law would apply class-wide.” Order, pp. 16-17. This should have ended the matter and convinced the court to certify the 23(b)(2) class. But the district court became distracted by the issue of individual relief for teachers and staff— matters that can be resolved in a 23(b)(3) proceeding.

See Johnson, 702 F.3d at 371 (“Once declaratory relief is ordered, all that is left is a determination of monetary relief, and that is the type of proceeding for which (b)(3) is designed.”). In McReynolds, for example, when the court certified a 23(b)(2) class of African-American financial advisors, it did so because it concluded that it would be more efficient to evaluate the plaintiffs’ claims regarding the disparate impact of the policies on a class-wide basis rather than in 700 individual lawsuits. McReynolds, 672 F.3d at 490-91.

This was true despite the fact that if the claims of disparate impact prevailed, it might be necessary for the court to hold hundreds of separate trials to determine which class members were actually adversely affected by one or both of the practices and if so what loss each class member sustained. Id. at 491. “But at least,” the court concluded, “it wouldn’t be necessary in each of those trials to determine whether the challenged practices were unlawful.” Id. This case is no different.

It may be necessary to hold separate hearings to determine to what relief each class member or sub-class is entitled (both in terms of reinstatement and money damages), but the question as to whether the reconstitution process discriminates against African Americans, either by disparate impact or treatment, can be adjudicated class-wide. Likewise, a declaratory order that the turnaround process did or did not violate federal law would resolve the issue for all class members. And a moratorium on turnaround would also provide relief for all class members. For this reason, the Kartman v. State Farm Mut. Auto. Ins.

Co., 634 F.3d 883 (7th Cir.2011) ease to which the defendant points does not help. In Kartman, the plaintiffs dressed up what was really a claim for money damages (in the form of insurance payments) in injunctive clothing by asking that the court order the insurance company to evaluate their hail-damaged roofs under a uniform and objective standard. Id. at 889. This court found that the insurance company’s “approach to hail-damage estimating (if it was inconsistent) might be evidence tending to show that the insurer underpaid some hail-damage claims.

But it does not independently establish liability or support a separate injunctive remedy.” Id. at 891. In contrast, a determination of liability in this case (i.e. a finding that the reconstitution practice discriminated against African Americans) might require later determinations of individual relief, but would resolve all questions of liability. Indeed, this case follows the exact contours of the Wal-Mart decision which con-scribed the boundaries of 23(b)(2) as follows: Rule 23(b)(2) applies only when a single injunction or declaratory judgment would provide relief to each member of the class.

It does not authorize class certification when each individual class member would be entitled to a different injunction or declaratory judgment against the defendant. Similarly, it does not authorize class certification when each class member would be entitled to an individualized award of monetary damages. Wal-Mart, 131 S.Ct. at 2557 (emphasis in original). Here we have a proposed Rule 23(b)(2) class asking for the same injunction and declaratory relief for all.

By refusing to certify the class, the district court erred in its assessment of the legal requirements of Rule 23(b)(2) and its assessment of the 23(b)(2) class’s request. C. As we just described, a 23(b)(2) class cannot seek money damages unless the monetary relief is incidental to the injunctive or declaratory relief. Wal-Mart, 131 S.Ct. at 2557. The plaintiffs siphoned that portion of the complaint that requested monetary relief and individual remedies into a request for 23(b)(3) class certification.

Federal Rule 23(b)(3) allows for class certification when “questions of law or fact common to the class members predominate over any questions affecting individual members” and “when a class action is superior to other available methods for fairly and efficiently adjudicating the controversy.” Fed.R.Civ.P. 23(b)(3). The latter superiority requirement is not at issue here.

The district court instead found that common claims did not predominate, as “[t]he selection process involved a qualitative review, and the [c]ourt would need to delve into how each of the ten schools was evaluated in comparison to the other schools considered but not selected.” Order, p. 18.

To some extent the question of commonality that we dissected at length above, and the question of predominance overlap: To gain class-action certification under Rule 23(b)(3), the named plaintiff must demonstrate, and the District Court must find, that the questions of law or fact common to class members predominate over any questions affecting only individual members. This predominance requirement is meant to test whether proposed classes are sufficiently cohesive to warrant adjudication by representation, but it scarcely demands commonality as to all questions.

In particular, when adjudication of questions of liability common to the class will achieve economies of time and expense, the predominance standard is generally satisfied even if damages are not provable in the aggregate. Comcast Corp. v. Behrend, — U.S. -, 138 S.Ct. 1426, 1436-37, 185 L.Ed.2d 515 (2013) (internal citations omitted).

Our earlier discussion of commonality leads us to the conclusion that the district court also erred when determining that the plaintiffs failed to meet their burden of proving predominance when it concluded that the process of choosing schools to reconstitute was different for each school. The lower court reasoned that “there were specific facts and issues as to why each of the ten schools was selected for turnaround in 2012.” Order, p. 18.

This is true, but as we discussed at length above, however, each school was evaluated for its performance under the same set of criteria, analyzed by the same committee, and ultimately subject to the decision-making authority of one person. As the plaintiffs point out, they all suffered the same injury at the same time as the result of the same selection process by the same central decision-maker. Common issues of fact and law predominate in particular when adjudication of questions of liability common to the class will achieve economies of time and expense.

See Comcast Corp., 133 S.Ct. at 1437. “Rule 23(b)(3), however, does not require a plaintiff seeking class certification to prove that each element of her claim is susceptible to classwide proof. What the rule does require is that common questions predominate over any questions affecting only individual class members.” Amgen Inc. v. Connecticut Retirement Plans and Trust Funds, — U.S. -, 133 S.Ct. 1184, 1196, 185 L.Ed.2d 308 (2013) (internal citations omitted).

In this case, the key question upon which all of the litigation rises or falls can be answered for every plaintiff: was the selection process discriminatory?

This is a good time to issue the reminder that “Rule 23(b)(3) requires a showing that questions common to the class predominate, not that those questions will be answered, on the merits, in favor of the class.” Id., — U.S. -, 133 S.Ct. 1184, 185 L.Ed.2d 308 (2013). “[T]he office of a Rule 23(b)(3) certification ruling is not to adjudicate the case; rather, it is to select the ‘method’ best suited to adjudication of the controversy ‘fairly and efficiently.’ ” Id. at 1191.

Consequently, we can take no position as to whether the plaintiffs will be able to demonstrate that the selection process was indeed discriminatory either in treatment or impact. The only answer we provide today is that it will certainly be efficient and fair to answer the question once for all plaintiffs rather than in piecemeal litigation. If the selection process is determined to be discriminatory, individualized remedies and damages may have to be determined for each plaintiff or perhaps for subclasses of plaintiffs, such as tenured teachers, non-tenured teachers and the like.

But as we noted above, this does not prevent certification of the class. As the district court correctly noted “the fact that damages may be individualized in this case would not preclude certification.” Order, p. 18, citing Butler v. Sears Roebuck & Co., 727 F.3d 796, 801 (7th Cir.2013). Given these considerations, the plaintiffs have met the requirements for certification of the class under Rule 23(b)(3).

One single question would trigger a liability finding for both the 23(b)(2) and 23(b)(3) class: did the policies and process behind the 2012 reconstitution unlawfully discriminate against African-American teachers and staff? And the answer to this question would eliminate the need for repeat adjudication of this question for determinations of damages or individual injunctive relief. D. Finally, Rule 23(c)(4) permits the court to certify particular issues for resolution as a class action.

Because we conclude that the class can be certified under both Rule 23(b)(2) and 23(b)(3), we have no need to consider whether the district court should have considered certification of one particular issue. Nor must we consider the Board’s argument that plaintiffs Garrett, Green, and the Chicago Teacher’s Union are not appropriate class representatives, as the Board failed to appeal from the district court’s finding of adequacy of representation. For the foregoing reasons, the district court order is reversed and remanded for further consideration consistent with this opinion. .

District-run schools in CPS are organized into 19 geographic networks, which provide administrative support, strategic direction, and leadership development to the schools within each Network. Each network is headed by a Chief of Schools, also called a Network Chief. . The district court referred to 226 schools eligible for turnaround in 2012. On appeal, the Board clarified that there were 226 schools rated at the lowest academic level, level three, and thus eligible for turnaround in 2012.

There were also, however, an additional twenty-four schools rated at academic level two that had been on probation for a year or more and thus also were eligible for turnaround under Illinois law. The Board eliminated all but one of these level two schools from consideration for turnaround. We will continue to use the number 226 for simplicity. . In 2008, the school district began measuring the freshman on-track rate, a measurement developed by the University of Chicago.

The measurement looks at course grades and credits in the first year of high school and students are considered on-track at the end of their freshman year if they accumulated at least five course credits and failed no more than one semester course in a core subject during the school year. http://cps.edu/News/ Press_releases/Pages/PRl_08_27_2014.aspx . To avoid confusion, our references are to the district court docket cites with both individual record page numbers, and for ease of location, a page identification number (ID#) from the continuously paginated district court record. .

The defendants also claim that the plaintiffs waived this argument by failing to raise it below. We conclude that the argument was not waived, but rather a relevant response to the district court’s conclusion that the subjective criteria in the latter steps of the process defeated commonality. Once the district court separated the steps and determined that the subjective one doomed class certification, the plaintiffs were entitled to direct the court’s attention back to the objective aspects of the process, and demonstrate how a discriminatory step in a chain of events can affect the ultimate outcome. .

The Board voted to approve all recommendations for reconstitution. . There is some confusing language in the plaintiffs’ initial brief requesting class certification in the district court in which, after asking for declaratory and injunctive relief only, the plaintiffs make an offhanded and unexplained comment that "the assessment of backpay for these individuals is 'generally applicable to the class.’ ” R. 63-1, p. 18 (ID# 842).

The plaintiffs' reply brief in the district court, however, makes clear that its 23(b)(2) class seeks declaratory and injunctive relief and that "[a]ny additional relief to the (b)(2) class will be incidental to, and flow from, the declaratory relief sought. Calculating this relief will be ‘mechanical, formulaic' — and thus appropriate for a 23(b)(2) class,” R. 83, pp. 19-21 (ID# 1780-1782), citing Johnson v. Meriter Health Servs. Emp. Ret. Plan, 702 F.3d 364, 372 (7th Cir.2012). See also, footnote 8, infra. .

To the extent that any monetary relief is "incidental to the injunctive or declaratory relief” it could be included in a Rule 23(b)(2) class, if "it appear[s] that the calculation of monetary relief will be mechanical, formulaic, a task not for a trier of fact but for a computer program.” Johnson, 702 F.3d at 372.