Carlos Bell v. Civil Service Commission

If this opinion indicates that it is “FOR PUBLICATION,” it is subject to revision until final publication in the Michigan Appeals Reports.

STATE OF MICHIGAN

COURT OF APPEALS

CARLOS BELL and All Others Similarly Situated, UNPUBLISHED July 16, 2020 Plaintiffs-Appellees,

v No. 347929 Wayne Circuit Court CIVIL SERVICE COMMISSION and STATE LC No. 17-003861-CZ PERSONNEL DIRECTOR,

Defendants-Appellants. Before: TUKEL, P.J., and SERVITTO and BECKERING, JJ.

PER CURIAM.

Defendants, the Michigan Civil Service Commission and the State Personnel Director, appeal by leave granted1 the trial court’s order granting plaintiffs’ motion for class certification under MCR 3.501(A), with plaintiff Carlos Bell as the class representative.2 Defendants argue that the trial court improperly certified two classes in a single suit, contrary to MCR 3.501(B)(3)(d)(ii), and clearly erred by finding that plaintiffs had met the numerosity and commonality requirements for class certification. Finding no error, we affirm.

I. BACKGROUND

Defendants use an Entry Level Law Enforcement Examination (“ELLE”) as a “screening tool” for certain classified law-enforcement positions with the Michigan State Police (MSP) and the Department of Natural Resources (DNR). Two versions of the ELLE are at issue in this case, the former version of the Elle (“2002 ELLE”), in use from March 2002 through December 2014, and the current version of the ELLE (“2014 ELLE”), which replaced the 2002 ELLE in December 2014. The 2002 ELLE consisted of three sections, two of which, the Human Interaction and Reading Ability sections, contained multiple-choice questions, while the third, Incident

1 Bell v Civil Serv Comm, unpublished order of the Court of Appeals, entered June 26, 2019 (Docket No. 347929). 2 References to plaintiff in the singular signify Bell as the movant for class certification.

-1-

Observation Report Writing, required applicants to watch a video and hand write a report in a test booklet describing their observations. Applicants had to pass each section with a minimum score, and achieve a total minimum score to pass the exam. The 2014 ELLE also consists of three sections: (1) Human Relations and Judgment; (2) Report Writing; and (3) Reading. The first and third sections are similar, respectively, to the Human Interaction and Reading Ability sections of the 2002 ELLE. The Report Writing section of the 2014 ELLE requires applicants to watch a video, hand write a report about their observations, and then answer multiple choice questions about the report. The handwritten report is never scored; all test questions are multiple choice and scoring of the 2014 ELLE is “fully automated.”

Plaintiff Bell, an African-American male, took and failed the 2002 ELLE twice, and the 2014 ELLE once. He filed the present class-action complaint on March 3, 2017, alleging that both ELLE examinations have a disparate adverse impact on African-American applicants. Plaintiff alleged violation of the Elliot-Larsen Civil Rights Act, MCL 37.2101 et seq., on two theories: (1) disparate impact, and (2) intentional discrimination arising from defendants’ failure to monitor the adverse impact of the examinations on African-American applicants.

Plaintiff moved for class certification on June 29, 2018. He asked for certification for two classes, separating those who took the 2002 ELLE and the 2014 ELLE:

[A]ll African-American applicants who took the Entry-Level Law Enforcement examination at any time on or after February 15, 2014 through November 2014, and failed the reading test, writing test, or overall exam score; and

[A]ll African-American applicants who took the Entry-Level Law Enforcement examination at any time on or after December 1, 2014 continuing to the present and beyond, and failed the reading or writing test or both.

The trial court granted plaintiff’s motion for class certification, but without providing any analysis regarding the class certification factors under MCR 3.501(A). Defendants subsequently filed an emergency application for leave to appeal and, in lieu of granting the motion, this Court vacated the trial court’s order and remanded for more precise articulation of each class certification requirement.3

On remand, the trial court followed our instructions and issued a written opinion granting plaintiff’s motion. The trial court explained that plaintiff had asked to certify two classes, one concerning each version of the test. The court then addressed the various prerequisites to class certification set forth in MCR 3.501(A)(1), generally referred to as numerosity, commonality, typicality, adequacy, and superiority. See Henry v Dow Chem Co, 484 Mich 483, 488; 772 NW2d 301 (2009). With regard to numerosity, the court explained that plaintiff had asserted that over 600 applicants failed the 2002 ELLE and 98 failed the 2014 ELLE, that precise numbers were not required, and that common sense dictated that with hundreds of African-American applicants

3 Bell v Civil Serv Comm, unpublished order of the Court of Appeals, entered January 23, 2019 (Docket No. 346562).

-2-

having failed the examinations, numerosity was established. The court also found that plaintiff established the existence of common questions of fact and law, explaining:

Here, a common question regarding the ELLE exam exists and individualized questions concerning how much damage, if any, each class member has suffered also exist. The individualized determinations of damages will not predominate over the common question of whether the ELLE exams had a disparate impact on African-American applicants. Furthermore, plaintiff’s proposed method of implementing a two-stage trial plan may be employed to resolve any individual issues concerning class member relief, making class action even more plausible.

The court noted that the parties did not dispute typicality or adequacy, and then moved to the final factor, superiority; i.e., whether a class-action suit was “superior to other available methods of adjudication in promoting the convenient administration of justice.” MCR 3.501(A)(1)(e). Unconvinced by defendants’ argument that the class-wide treatment of damages would be impracticable or impossible, the court noted that the same common questions were at issue, and stated that “individual damages may be addressed in a second-stage proceeding, if necessary.” The court was also concerned that requiring individual actions would create a risk of inconsistent adjudications, and believed that proceeding as a class action was the “superior way to resolve this dispute.” Finding all of the prerequisites in MCR 3.501(A)(1) to be established, the trial court granted plaintiff’s motion for class certification. The court subsequently entered a stipulated order that stayed a deadline for defendants to produce a list of contact information for class members pending the resolution of an expected application for leave to appeal. As expected, defendants sought leave to appeal the trial court’s order which, as noted above, was granted by this Court.

II. ANALYSIS

A. STANDARD OF REVIEW

“The proper interpretation and application of a court rule is a question of law, which we review de novo.” Henry v Dow Chem Co, 484 Mich 483, 495; 772 NW2d 301 (2009). Principles of statutory construction apply when interpreting a court rule. Id. The first step is to consider “the plain language of the court rule in order to ascertain its meaning.” Id. “The intent of the rule must be determined from an examination of the court rule itself and its place within the structure of the Michigan Court Rules as a whole.” Id. (quotation marks and citation omitted).

In the context of a decision on class certification, a trial court’s factual findings are reviewed for clear error, and the trial court’s discretionary determinations are reviewed for an abuse of discretion. Id. at 495-496. Under the clear error standard, this Court “will overturn the trial court’s decision only if [the Court is] left with the definite and firm conviction that a mistake has been made.” Adams v West Ottawa Pub Sch, 277 Mich App 461, 465; 746 NW2d 113 (2008). “The trial court abuses its discretion when its decision is outside the range of reasonable and principled outcomes.” Heaton v Benton Constr Co, 286 Mich App 528, 542; 780 NW2d 618 (2009).

-3-

B. LEGAL STANDARDS

Members of a class may only sue or be sued as representatives of all class members if they meet the requirements of MCR 3.501(A)(1). Henry, 484 Mich at 496. MCR 3.501(A)(1) allows a suit to proceed as a class-action suit only if all of the following requirements are met:

(a) the class is so numerous that joinder of all members is impracticable;

(b) there are questions of law or fact common to the members of the class that predominate over questions affecting only individual members;

(c) the claims or defenses of the representative parties are typical of the claims or defenses of the class;

(d) the representative parties will fairly and adequately assert and protect the interests of the class; and

(e) the maintenance of the action as a class action will be superior to other available methods of adjudication in promoting the convenient administration of justice.

As previously indicated, these requirements are generally referred to as numerosity, commonality, typicality, adequacy, and superiority. Henry, 484 Mich at 488. At issue here is whether plaintiffs have met the requirements of numerosity and commonality. “[S]trict adherence to the class certification requirements is required[,]” and a “party seeking class certification must meet the burden of establishing each prerequisite before a suit may proceed as a class action.” Henry, 484 Mich at 500. This is accomplished by “provid[ing] the certifying court with information sufficient to establish that each prerequisite for class certification in MCR 3.501(A)(1) is in fact satisfied.” Id. at 502. “Because there is limited case law in Michigan addressing class certifications, this Court may refer to federal cases construing the federal rules on class certification.” Neal v James, 252 Mich App 12, 15; 651 NW2d 181 (2002), overruled on other grounds by Henry, 484 Mich 483 (2009). While this Court may look to federal court decisions as persuasive authority, such decisions are not binding on this Court. See Abela v Gen Motors Corp, 469 Mich 603, 607; 677 NW2d 325 (2004) (“Although lower federal court decisions may be persuasive, they are not binding on state courts.”).

C. CLASS CERTIFICATION

1. SEPARATE CLASSES

Defendant first argues the trial court improperly certified two separate classes in one suit, contrary to MCR 3.501(B)(3)(d)(ii). We disagree.

Under MCR 3.501, the trial court may order that “a proposed class be divided into separate classes with each treated as a class for purposes of certifying, denying certification, or revoking a certification.” MCR 3.501(B)(3)(d)(ii). Defendants contend that the trial court is prohibited from certifying multiple classes in a single case unless each originates from a single class that meets all certification requirements under MCR 3.501(A). Defendants support their position with reference to a 2006 law review article analyzing a similar provision in the federal rules of civil procedure.

-4-

In that article, the author posits two theories of certifying separate classes: “the replacement theory, which posits that subclasses can be certified without regard to the certifiability of the global class action, and the contingency theory, which requires any subclass to be a part of a certified global class.” Scott Dodson, Subclassing, 27 Cardozo L Rev 2351 (2006). Despite the article’s conclusion that “the replacement theory is the best interpretation of the [federal] subclassing provision,” id., defendants urge this Court to adopt the contingency theory, and to hold that the trial court erred by certifying two classes apart from the existence of a certified global class. We decline to do so.

“[T]he plain language of MCR 3.501(A) provides sufficient guidance for class certification decisions in Michigan.” Henry, 484 Mich at 502. Professor Dodson’s observation regarding the language of the similar provision in FR Civ P 23 is equally applicable here: “[t]The term ‘class’ is used throughout [MCR 3.501] detached from connotations of certifiability.” Subclassing, 27 Cardozo L Rev at 2372. For example, MCR 3.501(A)(1) uses the term “class” to refer to a pre- certified, and potentially uncertifiable, group of plaintiffs when it states, “[o]ne or more members of a class may sue or be sued as representative parties on behalf of all members in a class action” if certain criteria are met. In addition, MCR 3.501(B)(3)(d)(ii) specifically states that the court may divide “a proposed class” into separate classes. Like the use of “class” in MCR 3.501(A)(1), to refer to a pre-certified or possibly uncertifiable group of plaintiffs, the use of “proposed class” in MCR 3.501(B)(3)(d)(ii) evokes a pre-certified or potentially uncertifiable class that is divided into separate classes, “with each treated as a class,” for one of three purposes: “certifying, denying certification, or revoking a certification,” MCR 3.501(B)(3)(d)(ii). Thus, we conclude that nothing in the plain language of MCR 3.501 prohibits a trial court from certifying two or more classes in a single lawsuit, where appropriate.4

Defendants agree in principle that the trial court could certify separate classes in a single lawsuit, but argue that doing so was not appropriate here. They hypothesize that if the trial court’s “separate certification approach” prevails in this case, future plaintiffs could bring any number of claims against the State in one action and simply ask for separate classes. Claims challenging the hiring practices of the MSP could be brought with claims challenging the hiring practices for data analysts at the Department of Treasury. Even more farfetched, challenges to the treatment of inmates by the Department of Corrections could be combined with challenges to the maintenance of roads by the Department of Transportation. The problem with defendants’ hypotheticals is that, while it is true that “at a sufficiently abstract level of generalization, almost any set of claims . . . could display commonality,” there must nevertheless be “a common issue the resolution of which will advance the litigation.” Sprague v General Motors Corp., 133 F3d 388, 397 (CA 6, 1998). Regarding the non-hypothetical case at bar, resolving the question of whether the ELLE tests violated the ELCRA arguably would advance the litigation, and defendants have not suggested otherwise. For the reasons stated above, we conclude that the trial court did not improperly divide

4 We also note that an advisory comment to the 1966 amendment of the federal rules states with regard to FR Civ P 23: “Two or more classes may be represented in a single action. Where a class is found to include subclasses divergent in interest, the class may be divided correspondingly, and each subclass treated as a class.”

-5-

the proposed class into separate classes and treat each one as a separate class within a single action.5

2. NUMEROSITY

Next, defendants argue that the trial court erred by finding the numerosity requirement met because plaintiff failed to show or to provide a means of showing how many of the potential class members suffered an actual injury as a result of failing the examinations. Defendants argue that if applicants would have been ineligible for those positions for a different reason, such as requirements concerning their age, citizenship status, education, or criminal record, “the act of failing the Exam would have had no impact on that person’s ability to seek those jobs.” Defendants’ argument is unavailing.

“[N]o particular minimum number is necessary to meet the numerosity requirement, and the exact number of members need not be known as long as general knowledge and common sense indicate that the class is large.” Zine v Chrysler Corp, 236 Mich App 261, 288; 600 NW2d 384 (1999). “[T]he plaintiff must adequately define the class so potential members can be identified and must present some evidence of the number of class members or otherwise establish by reasonable estimate the number of class members.” Duskin v Dep’t of Human Servs (On Remand), 304 Mich App 645, 653; 848 NW2d 455 (2014) (quotation marks and citation omitted). “The proponent must establish that a sizeable number of class members have suffered an actual injury.” Id.

In support of their position, defendants rely on Duskin, an appeal involving claims of unemployment discrimination based on disparate treatment. The proposed plaintiff class in Duskin was “comprised of all ‘minority’ male employees” of the Department of Human Services (DHS), including over 500 “African-American, Hispanic, Arab, and Asian males in various departments and offices throughout the state.” Id. at 648. The plaintiffs maintained that, “since 2003, fewer minority males have been promoted within the DHS to the positions of program manager, district manager, county director, and first line supervisor because of ‘department wide cultural deficiencies regarding minority males.’ ” Id. They asserted that some of the class members applied for but were denied promotions or training opportunities for which they were qualified, while “frustration with some of the [Department’s] supervisory and management employees’ discriminatory attitudes and practices involving racial and gender bias directed against minority males . . .” rendered others “too discouraged to apply.” Id. The trial court ultimately granted the plaintiffs’ motion for class certification. The trial court found that the plaintiffs established numerosity because, “while not all class members had applied for promotions, all class members had ‘an interest in making sure that they are not discriminated against if they do.’ ” Id. at 650.

5 Further, defendants do not identify any prejudice that they would suffer as a result of the trial court’s certification of two separate classes. Were we to agree with defendants’ position, the ultimate remedy would be to sever the classes so they could each proceed in two separate actions. Thus, principles of efficiency and judicial economy are served by litigating the claims of the two classes together in one proceeding.

-6-

On appeal, this Court held that the trial court had clearly erred by finding that the plaintiffs had established numerosity. The Court explained that minority employees who did not apply for promotions out of fear of discrimination could not be included in the class because “membership must be based on objective criteria[,]” and that “while the minority males established an estimate of the number of class members, they did not provide an adequate statement of basic facts to support that a sizeable number of those class members suffered an actual injury.” Id. at 653-654.

Unlike the class members in Duskin, the class members in the present case have suffered actual injury as determined by objective criteria. Plaintiffs have alleged that defendants’ preemployment tests have a disparate impact on African-American candidates. All of the class members took and failed the tests; thus, each member who took and failed a test allegedly suffered the actual injury of discrimination on the basis of race, regardless of their other qualifications for the law-enforcement positions. Defendants have not cited, nor has our research revealed, any authority requiring plaintiffs in a disparate-impact discrimination case to also show that class members would have met all other employment criteria as part of establishing numerosity. See Williams v Ford Motor Co, 187 F3d 533, 535 (CA 6, 1999) (involving a certified class comprised of all African-American applicants for unskilled hourly employment with Ford from 1989 to the present who were excluded from unskilled employment with Ford in Ohio because they scored “low” on the unskilled pre-employment test); Easterling v Conn Dep’t of Corrections, 265 FRD 45, 48-51 (D Conn, 2010), modified, 278 FRD 41 (D Conn, 2011) (finding numerosity in an action challenging a component of the department’s preemployment physical examination, even though applicants who passed the physical examination would be subject to additional preemployment screening requirements). Accordingly, we cannot say that the trial court clearly erred by finding that plaintiff met the requirement for numerosity. Defendants’ argument that some potential class members would have been ineligible for employment because they did not meet minimum threshold requirements such as age, citizenship status, education, and absence of prior felony convictions relates more properly to damages than to numerosity. 6

Lastly, defendants contend that treating unscored writing sections as failing scores improperly inflates the number of members in the first class, i.e., those who took the 2002 ELLE.

6 “A court should generally limit relief ‘only to those class members who would have filled vacancies had there been no discrimination.’ ” Easterling, 278 FRD 41, 48 (D Conn, 2011), quoting Ingram v Madison Square Garden Center, Inc, 709 F.2d 807, 812–13 (CA 2, 1983). The class-wide calculation and pro rata distribution of back pay is appropriate when “the number of qualified class members exceeds the number of openings lost to the class through discrimination, and identification of the individuals entitled to relief would drag the court into a quagmire of hypothetical judgments.” Easterling, 278 FRD at 48, (quoting Catlett v Mo Highway and Transp Comm'n, 828 F2d 1260 (CA 8, 1987). See also United States v City of New York, 276 FRD 22, 44 (“Because it is impossible to determine exactly which non-hire victims would have received job offers . . . the court must first determine the aggregate amount of individual relief to which the subclasses are entitled and then distribute that relief pro rata to eligible claimants.”). Under these principles, those members of the class who may have been ineligible for employment because they failed to meet minimal requirements, such as age, citizenship status, education, or had a past felony conviction are excluded from any employment-related damages, such as lost wages and benefits.

-7-

In order to be part of the first class, an applicant had to fail the reading section, fail the writing section, or fail the overall examination. As previously explained, an applicant could fail the 2002 ELLE examination by failing to obtain a minimum score on any of the three sections or, by passing all three sections, but failing to achieve an overall minimum score. Based on the way the 2002 ELLE was scored, defendants argue in their brief to this Court that “[m]embership in the class for the [2002 ELLE], then, is expressly defined to include only the following three groups: (1) those who failed the Reading section; (2) those who failed the Writing section; and (3) those who passed all three sections, but failed to achieve the required overall exam score” (defendants’ emphasis). Accordingly, applicants who failed only the video section of the 2002 ELLE would not be class members because they would not meet any of the three criteria: they did not fail the reading section, it is not known if they failed the writing section because it would not have been scored, and they would not fall into the category of those who failed the overall test because, in order to fall into that category, applicants first had to pass all three sections. Nevertheless, these applicants are included as members of the first class because unscored writing tests are considered failing scores. Defendants contend that treating unscored tests as failing tests constructs a class on pure conjecture, and that the remedy would be to include only those whose writing sections were scored but who failed to achieve a passing score.

We agree that an unscored writing section is not necessarily a failing score. Nor is it a passing score. But where the only two options are pass or fail, an unscored section seems more like a failing score than a passing one. All this aside, defendants base their argument on the premise that membership in the first class is “expressly defined” as including, inter alia, “those who passed all three sections, but failed to achieve the required overall exam score” (defendants’ emphasis). However, the definition of the class that the trial court certified does not contain the italicized words. On the contrary, applicants are members of the class if they failed the reading section, if they passed the reading section and the video section but failed the writing section, or if they failed to achieve the overall minimum score, regardless of whether they passed all three sections. Applicants whose writing sections were not scored might not have been able to achieve the overall minimum score, and thus arguably would be members of the class by virtue of falling into the latter category, along with those who may have passed all three sections, but did not score a passing combined score. Although the definition of the first class may add to the class more people than defendants’ proposed interpretation, it does not follow that the definition is overly broad.

A class definition can be overly broad if it captures as potential members those not intended by the complaint. See Mich Ass’n of Chiropractors, 300 Mich App at 591. Plaintiffs alleged in their complaint that defendants discriminated against African-American applicants by developing and using tests that “have a disproportionately disparate impact on African Americans compared with Caucasians . . . .” They described the tripartite structure of the 2002 ELLE, and explained that it was scored on a pass/fail basis, that the report writing section would be scored only if an applicant passed both the Human Interaction and Reading Ability sections, separately and in combination, and that “there was a total minimum score expectation.” It was the written test as a

-8-

whole that plaintiffs alleged in their complaint was discriminatory, not merely portions of it.7 Nothing about the definition of the first class suggests that it overreaches the intent of the complaint to obtain relief for African-American applicants who took and failed the tests used by defendants as part of the hiring process for the law-enforcement jobs at issue.

3. COMMONALITY

Next, defendants argue that plaintiffs failed to demonstrate commonality because individualized questions predominate over any common questions. Defendants claim this is the case because the writing portion of the 2002 ELLE was subjectively scored, and because any individual issues of damages will require hundreds of “mini-trials.”

“To establish commonality, the proponent of certification must establish that issues of fact and law common to the class predominate over those issues subject only to individualized proof.” Duskin, 304 Mich App at 654 (quotation marks and citation omitted). “The common contention . . . must be of such a nature that it is capable of classwide resolution—which means that determination of its truth or falsity will resolve an issue that is central to the validity of each one of the claims in one stroke.” Id. (quotation marks and citation omitted). Commonality “does not require all issues in the litigation to be common . . . .” Hill v Warren (On Remand), 276 Mich App 299, 311; 740 NW2d 706 (2007).

Defendants present only a cursory argument with regard to their claim that the subjective nature of scoring the Writing Section of the 2002 ELLE defeats commonality. They contend only that the trial court did not address the issue and they fail to demonstrate how a different result would be warranted on the basis of this argument. In addition, they do not contest plaintiffs’ position that graders were guided by a standard scoring template and that use of the template limited the graders’ discretion so that the limited amount of discretion allowed in grading the Writing Section did not defeat commonality. See Wal-Mart Stores, Inc v Dukes, 564 US 338, 356; 131 S Ct 2541; 180 L Ed 2d 374 (2011) (suggesting that commonality would be satisfied where, while discretionary issues were at issue, a “common mode of exercising discretion” had been shown). It seems that here, the use of a template to score the examinations would clearly show that subjective scoring of the Writing Section provided a “common mode of exercising discretion.” Accordingly, defendants have failed to persuade us that the discretion involved in scoring the Writing Section of the 2002 ELLE defeated commonality.

Defendants next argue that the trial court should not have certified the classes with respect to damages because individual questions predominate. We cannot say that we necessarily share defendants’ concerns.

According to plaintiffs’ proposed trial strategy, the stages of the trial would proceed as follows:

7 In fairness, plaintiffs’ expert opined that the reading and writing portions of both tests had a disparate impact on African-American candidates.

-9-

At the First Stage, the following issues are decided: (1) whether the ELLE exam had an adverse impact on African-American applicants; (2) whether using the ELLE exam is justified by business necessity and, if so, whether there was a less discriminatory alternative; (3) whether Defendants are liable to the name Plaintiff, as well as his damages; and (4) resolution of class-wide damages issues. At the Second Stage, to the extent that all damages cannot be resolved in the First Stage, individual hearings are held on back pay, and other available damages and relief.

Pursuant to this strategy, only those damages appropriate to the class as a whole would be decided at the first stage of trial, and the trial court would be left with the discretion and flexibility to address any individual issues at the second stage. The trial court seemed sanguine about this approach and, contrary to what defendants argue, it would not involve over 600 mini-trials on individual damages.8 At this point, we see no reason to dictate to the trial court how to handle any potential relief to which class members may be entitled.

III. CONCLUSION

For the reasons stated above, we conclude that the trial court’s factual findings as to the requirements for class certification were not clearly erroneous and that the court did not abuse its discretion by certifying the classes at issue.

Affirmed.

/s/ Jonathan Tukel /s/ Deborah A. Servitto /s/ Jane M. Beckering

8 In fact, as indicated in footnote (4), federal courts faced with determining damages in similar disparate impact discrimination cases have eschewed any attempt to “second-guess” who would have been hired and have opted to calculate class-wide awards and distribute them pro rata to those entitled to receive them. See Easterling, 278 FRD at 48; United States v City of New York, 276 FRD at 44.

-10-