ROSS et al v. LOCKHEED MARTIN CORP.

Filing 16

MEMORANDUM OPINION denying 4 Motion for Preliminary Approval of Class Action Settlement Agreement. See attached document for details. Signed by Judge Ketanji Brown Jackson on July 28, 2017. (lckbj2)

UNITED STATES DISTRICT COURT
FOR THE DISTRICT OF COLUM B IA

)
VERNON ROSS and DEBRA JOSEY, )
o n b eh alf o f themselves and all others )
simila rly situated,
)
)
P laintiffs,
)
)
v.
)
)
LOCKHEED MARTIN CORP.,
)
)
Defendant.
)
)

No. 16-cv-2508 (KBJ)

M EM ORANDUM OPINION
Named P laintiffs Vernon Ross and Debra Josey (“P laintiffs”) allege that
Defendant Lockheed Martin Corporation (“Lockheed”) has “engaged in a pattern or
practice of employment discrimination” that is “manifest[] in Lockheed Martin’s
performance appraisal system.” (Compl., ECF No. 1, ¶ 11.) According to Ross and
Josey, Lockheed’s annual employee performance review process is “flawed in both
design and implementation” (id.) because “African-American non-represented, salaried
employees below the level of Vice P resident have received lower overall ratings on
their annual performance reviews as compared to similarly situated white employees”
(id . ¶ 15), and the lower ratings have, in turn, resulted in “lower salaries, raises, [and]
bonuses, [lower] long term incentive awards, fewer promotions, and a lower retention
rate” for African Americans at Lockheed (id . ¶ 67). P laintiffs’ three-count complaint
claims that Lockheed’s performance review process has been systemically injurious in a
manner that amounts to both intentional race discrimination (see id. ¶¶ 65– 68 (Count I))

and disparate impact race discrimination (see id. ¶¶ 70–73 (Count II)). P laintiff Ross
further contends, solely on his own behalf, that Lockheed retaliated against him “for
filing a Charge of Discrimination . . . and/or complaining to senior executives at the
Company of racial discrimination faced by him and other African-American
employees.” (Id. ¶ 78 (Count III).)
Critically, Ross and Josey seek to prosecute the race discrimination claims on
behalf of the following class of plaintiffs:
[all] salaried non-represented African-American employees below
the level of Vice P resident who received at least one performance
evaluation between January 1, 2013 and February 29, 2016, with an
overall rating below ‘significantly exceeded commitments’ while
employed at Lockheed Martin.
(Id . ¶ 1.) The complaint contends that the discrimination claims are susceptible to
class-action treatment because, under Lockheed’s performance review process, there is
an “absence of measurable indicators” of achievement, which has allegedly “resulted in
inadequate safeguards against bias in the assessment of African American employees.”
(Id . ¶ 18; see also id. ¶ 21 (resting the complaint’s systemic discrimination allegations
on the fact that “[m]anagers’ comments on employee performance have not consistently
relied on specific, measurable, time-sensitive measures of employees’ performance” and
“[a]s a result, similar or even identical performance could garner different ratings under
different supervisors”).)
P laintiffs have filed their putative class action complaint along with a proposed
Settlement Agreement; therefore, this case has come to this Court fully formed. (See
Compl.; Settlement Agreement, ECF No. 4-1.) One key feature of the resolution that
P laintiffs have negotiated with Lockheed (in addition to a $22.8 million settlement fund

2

and certain changes to Lockheed’s performance appraisal process) is the class members’
agreement to release a broad swath of potential legal claims against the company,
including claims that have nothing whatsoever to do with Lockheed’s performance
review procedures. (S ee, e.g., Settlement Agreement at 22 (stating that the class
members agree to release “any and all racial employment discrimination claims of
whatever nature, known or unknown,” including but not limited to all “claims of alleged
racial discrimination in employment or in the provision of employee benefits claims
under Title VII, Section 1981, ERISA[,]” and “any other federal, state, or local” law).) 1
Also noteworthy is what is not featured in the proposed Settlement Agreement:
ho w mu ch money each class member can expect to receive in exchange for releasing any
and all race discrimination claims “that were or could have been” asserted against
Lockheed. (Id .) P laintiffs’ counsel insists that no class member’s expected recovery
can be determined, or even estimated, up front; rather, each class member must first
formally accept the terms of the Settlement Agreement and complete a detailed form
that discloses—for the first time—the potential race discrimination and/or benefits
claims that she has already agreed to release. (See Tr. of Oral Arg. at 69.) In
operation, then, a putative class member must decide whether to opt out of the
Settlement Agreement before knowing (1) the nature and value of the potential legal
claims that she might otherwise have brought against Lockheed based on her
employment history, or (2) the amount that she is likely to receive for participating in
the settlement and relinquishing all of her (previously undisclosed) claims.

1

Pag e-number citations t o documents t hat t he p arties have filed refer to t he p age n umbers t hat t he
Co u rt ’s electronic filin g system automatically assigns.

3

Before this Court at present is P laintiffs’ motion for preliminary certification of
this case as a class action and preliminary approval of the Settlement Agreement. (S ee
P ls.’ Mot. for P reliminary Approval of Class Action Settlement Agreement (“Mot.”),
ECF No. 4.) In the motion, P laintiffs request that this Court make a preliminary
determination that the complaint satisfies the requirements of a viable class action
under Federal Rule of Civil P rocedure 23, and P laintiffs also seek preliminary approval
of the Settlement Agreement so that the class-wide notice and detailed claim forms can
be distributed. (S ee g enerally Mot.; Settlement Agreement.) P laintiffs have
consistently maintained that their putative class and proposed settlement satisfy all of
the necessary criteria for certification and approval under Rule 23 such that this case
should be permitted to cruise right into the notice and hearing stages of the class-wide
settlement process. However, for the reasons explained fully below, this Court has
reluctantly concluded that it has no choice but to pump the brakes.
In brief, P laintiffs have failed to demonstrate that the commonality prerequisite
for Rule 23 class certification can be adequately established, because they have not
explained how it is that Lockheed’s performance appraisal process systematically
discriminates against African-Americans such that it qualifies as either a “general
policy of discrimination” or a “testing procedure or other companywide evaluation
method” that gives rise to discrimination claims that are susceptible to common proof.
S ee Wa l-M art S to res, In c. v. Dukes, 564 U.S. 338, 353 (2011). This Court also sees
several fairness-related red flags when it considers the terms of the proposed Settlement
Agreement, including an egregious imbalance between the particular claims alleged in
the complaint and the scope of the class members’ release; a draconian set of

4

consequences that results from a putative class member’s mere failure to respond to the
class-wide notice; and a dearth of crucial information about the potential expected
recovery in relation to the claims being released, as is required for adequate assessment
of each putative class member’s individual settlement position.
Consequently, not only is this Court unable to find that P laintiffs have defined a
certifiable class for the purpose of Federal Rules of Civil P rocedure 23(a) and 23(b), the
Court must also conclude that the proposed Settlement Agreement cannot be
preliminarily approved as “fair, reasonable, and adequate” under Rule 23(e)(2).
Accordingly, the instant motion for preliminary approval of the Settlement Agreement
must be DENIED. A separate order consistent with this memorandum opinion will
follow.
I.

B ACKGROUND
A.

Dis parate Tre atment And Dis parate Impact Claims Under Title VII

Claims of employment discrimination under Title VII may proceed under both
“disparate treatment” and “disparate impact” theories. Ricci v. DeS tefano, 557 U.S.
557, 577 (2009); see a lso Griggs v. Duke Power Co., 401 U.S. 424, 431 (1971). 2
“Disparate treatment occurs when ‘[t]he employer simply treats some people less
favorably than others because of their race, color, religion, sex, or national origin.’”
An d erson v. Zubieta, 180 F.3d 329, 338 (D.C. Cir. 1999) (alteration in original)
(quoting Int’l Bhd. of Teamsters v. Un ited States, 431 U.S. 324, 335 n.15 (1977)); see
2

Th e Griggs d ecision recognized that , even t hough t he Civ il Rig hts A ct o f 1964 as o riginally enacted
“d id n ot include an exp ress p rohibition o n policies o r pract ices t hat p roduce a disparate imp act[,]”
claims o f disparate imp act were n evertheless available under t hat st atute. Ri cci, 557 U.S. at 577.
Co n g ress later amended Title VII t o codify t he h olding of Griggs and exp ressly provide for d isparateimp act claims. S ee i d. at 578 (citing Civ il Rights Act of 1991, Pub. L. No . 102-166, 105 St at. 1071
(co d ified at 42 U.S.C. § 2000e-2(k)(1)(A))).

5

a lso 42 U.S.C. § 2000e-2(a)(1) (making it unlawful for an employer “to fail or refuse to
hire or to discharge any individual, or otherwise to discriminate against any individual
with respect to his compensation, terms, conditions, or privileges of employment,
because of such individual’s race, color, religion, sex, or national origin”). To make
out a p rima f acie case of disparate treatment discrimination, a plaintiff must prove that
“(i) [he or she] suffered an adverse employment action (ii) because of [his or her] race,
color, religion, sex, or national origin.” Bra d y v. Of fice o f S ergeant a t Arms, 520 F.3d
490, 493 (D.C. Cir. 2008); see also id. at 493 & n.1 (citing M cDonnell Douglas Corp. v.
Green, 411 U.S. 792, 802 (1973)). “P roof of discriminatory motive is critical for
[disparate treatment] claims.” Anderson, 180 F.3d at 338 (internal quotation marks and
citation omitted).
By contrast, a disparate impact claim arises when “policies or practices that are
neutral on their face and in intent . . . nonetheless discriminate in ef f ect against a
particular group.” Id. at 339 (emphasis added; internal quotation marks and citation
omitted); see a lso 42 U.S.C. § 2000e-2(k)(1)(A)(i) (providing that a plaintiff may
establish a prima f acie disparate impact violation by “demonstrat[ing] that a respondent
uses a particular employment practice that causes a disparate impact on the basis of
race, color, religion, sex, or national origin”). 3 “When presenting a disparate impact
claim, a plaintiff must generally ‘demonstrate with statistical evidence that the practice
or policy has an adverse effect on the protected group.’” Greater New Orleans Fair

3

A n employer may d efend against a d isparate imp act claim by “demo nstrat[ing] that the challenged
p ractice is job relat ed for t he position in q uestion and consistent wit h business necessity[.]” 42 U.S.C.
§ 2000e-2(k)(1)(A)(i); see a lso Ricci, 557 U.S. at 578. “Ev en if t he employer meets t hat b urden,
h o wever, a p laintiff may still s ucceed b y showing t hat t he employer refuses to adopt an available
alt ernative employment p ractice t hat h as less disparate imp act and s erves t he employer’s legitimate
n eeds.” Ricci, 557 U.S. at 578 (cit ing 42 U.S.C. § 2000e-2(k)(1)(A)(ii), (C)).

6

Ho u sing Action Ctr. v. U.S. Dep’t of Housing and Urban Dev., 639 F.3d 1078, 1085– 86
(D.C. Cir. 2011) (quoting Garcia v. Johanns, 444 F.3d 625, 633 (D.C. Cir. 2006)).
Furthermore, and notably, a plaintiff must demonstrate that she was personally injured
as a result of an employment policy with a disparate impact in order to be able to
challenge the policy under Title VII. See 42 U.S.C. § 2000e-5(f) (conferring a cause of
action under Title VII on a “person claiming to be aggrieved” by an unlawful
employment practice); see a lso M elendez v. Ill. Bell Tel. Co ., 79 F.3d 661, 668 (7th Cir.
1996); Yo ung v. Covington & Burling LLP, 736 F. Supp. 2d 151, 158– 59 (D.D.C. 2010).
Both disparate treatment and disparate impact theories are available to an injured
plaintiff who seeks to challenge discrimination that results from an employer’s policy
of delegating employment decisions to individual supervisors based on subjective or
discretionary criteria. See Watson v. Fort Worth Bank & Trust, 487 U.S. 977, 991
(1988). A plaintiff challenging such a policy under a disparate impact theory is
relieved from having to prove that discriminatory intent motivated either the policy of
delegation or the particular adverse employment decision that affected her, see id. at
986– 87; however, crucially, “the plaintiff’s burden in establishing a prima facie case
goes beyond the need to show that there are statistical disparities in the employer’s
work force.” Id. at 994 (plurality opinion). 4 This is because “[i]t is completely
unrealistic to assume that [employers’] unlawful discrimination is the sole cause of
. . . statistical imbalances in the composition of their work forces.” Id. at 992 (plurality
opinion). Rather, to support a claim for disparate impact, “the plaintiff must offer

4

A majo rit y of t he Supreme Co urt expressly adopted t he reasoning of Justice O’Connor’s p lurality
o p in ion in Wa tson d uring t he fo llowing Term. S ee Wa rds Cove Packing Co. v. Atonio, 490 U.S. 642,
656– 58 (1989), su perseded b y statute o n other g rounds, 42 U.S.C. § 2000e-2(k).

7

statistical evidence of a kind and degree sufficient to show that the practice in question
has caused the exclusion of applicants for jobs or promotions because of their
membership in a protected group.” Id. at 994 (plurality opinion). P ut another way,
“[a]s a general matter, a plaintiff must demonstrate that it is the application of a
specific or particular employment practice that has created the disparate impact under
attack.” Wards Cove Packing Co., 490 U.S. at 657; see also 42 U.S.C. § 2000e2(k)(1)(B)(i) (providing that, unless a disparate-impact plaintiff can establish that
making such a showing is impossible, she must “demonstrate that each particular
challenged employment practice causes a disparate impact”).
B.

Plaintiffs ’ Alle gations In This Case

On December 23, 2016, P laintiffs filed the complaint in the instant case, which
alleges that Lockheed, “one of the largest aerospace, defense, and technology
companies in the world” (Compl. ¶ 8), engaged in race discrimination against its
African-American employees through the operation of the performance appraisal system
that the company uses to evaluate all of its employees (see id. ¶ 11). According to the
complaint, over the course of a three-year period between January of 2013 and February
of 2016 (see id . ¶ 1), Lockheed’s performance appraisal system “produced a disparate
impact in performance ratings, and consequently in the promotions, compensation, and
retention of salaried African-American employees below the level of Vice P resident”
(id . ¶ 11). The two named plaintiffs, Vernon Ross and Debra Josey, allege that they
were injured by the performance appraisal system and seek to certify a class of nearly
all current and former African-American Lockheed employees below the rank of Vice

8

P resident who were evaluated pursuant to the performance appraisal system during the
relevant three-year period. (See id. ¶¶ 11, 57.)
1.

The P erformance Appraisal System

According to the complaint, between 2013 and 2016, Lockheed “evaluated all
non-represented [i.e., non-union], salaried employees under essentially the same
performance appraisal system, although with some changes over time.” (Id. ¶ 12.)
Lockheed used this performance appraisal system “in all of its business areas—
Aeronautics, Rotary and Missions Systems, Missiles and Fire Control, Space Systems,
and Enterprise Operations” (id. ¶ 13)—and the system consisted of the following basic
components. First of all, every year (at the beginning of the performance appraisal
cycle) each salaried employee was required to identify “commitments” that reflected the
employee’s individual goals for her work at Lockheed during the upcoming year, with
manager approval. (See id. ¶¶ 17– 18.) Then, at the end of the calendar year, the
employee would receive a performance review in which a manager evaluated that
employee’s work, using both written comments and numerical ratings. (See id. ¶ 20.)
The written comments addressed the degree to which the employee had lived up to each
commitment, and the numerical ratings reflected the employee’s performance (on a 5.0
scale) with respect to meeting both her individual commitments and a series of desired
“behaviors” that Lockheed had prescribed across the entire company. (Id. ¶¶ 17, 20,
36.) An employee’s performance review also aggregated each of these numerical
rankings to calculate a single “composite score.” (Id. ¶ 20.)
Once all managers completed their performance reviews of the individual
employees in their purview, groups of managers then gathered for “calibration” sessions

9

that were held in order to determine an “ultimate tier placement” for the employees
under their supervision. (Id. ¶¶ 22– 23.) A tier placement reflected an overall
assessment of an employee’s performance vis-à-vis the identified commitments; for
example, the highest tier was called “significantly exceeds commitments[.]” (Id. ¶ 57.)
Notably, these calibration sessions operated on a curve: there were four possible tiers
(see id . ¶ 35), and the manager groups placed “predetermined percentages of [the]
employees” that were under consideration in a given session into each of the four tiers
(id . ¶ 22). In doing so, the managers purportedly took account of each employee’s
performance review as well as commentary from other participants in the calibration
session, including participants who had not reviewed that employee’s individual
performance appraisal results. (S ee id. ¶ 23.) Ultimately, an employee had the
opportunity to appeal her performance review and overall tier placement under the
appraisal system, but “[v]ery few appeals [were] brought or [were] successful in
altering overall ratings.” (Id. ¶ 24.)
According to the complaint, the overall tier placement that an employee received
after her individual performance review and the group calibration session played a
prominent role in Lockheed’s subsequent decisions regarding base compensation, bonus
payments, and promotions. (S ee id. ¶¶ 27– 32.) Specifically, P laintiffs maintain that the
“standard percentage increase in salary” that Lockheed awards to employees each year
was “based on their tier and their place in the salary range for their position.” (Id.
¶ 28.) Lockheed also “bestow[ed] discretionary merit increases” in compensation, as
well as “a variety of monetary awards and bonus programs[,]” based in part on an
employee’s overall tier placement. (Id. ¶¶ 29– 30.) Finally, “[f]or certain positions at

10

Lockheed Martin, employees may advance through what are called growth
promotions[,]” which “represent movement to a higher level within the employees’
existing positions” and which were “based in part on employees’ . . . performance
evaluations.” (Id . ¶ 31.)
P laintiffs’ complaint alleges that, due to flaws in the design and implementation
of the performance evaluation system, Lockheed’s performance review process “has
produced a disparate impact based on race in evaluation ratings and, consequently, in
the compensation, promotion, and retention of African-American employees.” (Id.
¶ 14.) For example, with respect to the individual employee performance reviews, the
complaint asserts that Lockheed failed to prescribe “measurable indicators” for
managers to use when deciding how to rate an employee for a particular objective or
behavior, and therefore Lockheed “ha[s] not provided adequate safeguards against bias
in the assessment of African American employees.” (Id. ¶ 19; see also id. ¶ 18.) In
addition, according to the complaint, managers’ written comments “have not
consistently relied on specific, measurable, time-sensitive measures of employees’
performance[,]” which means that “similar or even identical performance could garner
different ratings under different supervisors.” (Id. ¶ 21.)
The complaint also identifies flaws with respect to the group calibration
sessions. P laintiffs allege that the “discussion of any one employee was cursory at
best”; that “employees holding different positions but at the same level were compared
against one another”; that “at times, employees were represented by managers who
knew little about their performance if the employees’ manager was unavailable to attend
a given meeting”; and that, “[a]s a result, employees may have been assigned to the

11

forced distribution tiers by persons with little if any direct knowledge of employees’
performance[.]” (Id. ¶ 23.) Furthermore, with respect to the appeals process, the
complaint alleges that the right to appeal “does not rectify the biased assessments
resulting from the Company’s performance appraisal system” because employees do not
have “an adequate right to appeal to a manager who was not involved in preparing the
review from which [the employee is] appealing.” (Id. ¶ 24.)
The complaint maintains that, as a consequence of these alleged flaws in the
performance appraisal process, Lockheed’s evaluation system “allow[s] for racially
biased assessments of employees.” (Id. ¶ 12.) The result, P laintiffs contend, is that
“African-American non-represented, salaried employees below the level of Vice
P resident have received lower overall ratings on their annual performance reviews as
compared to similarly situated white employees[,]” and P laintiffs also allege, “[u]pon
information and belief,” that this disparity “is statistically significant.” (Id. ¶ 15.) In
addition, P laintiffs assert that the disparity in overall performance ratings has had
corresponding effects “in the compensation, promotion, and retention of AfricanAmerican employees.” (Id. ¶ 14; see also id. ¶ 28 (stating that, “[u]pon information and
belief,” “discrimination in the performance appraisal process” caused disparities in
merit salary increases); id. ¶ 32 (performance appraisal system led to disparities in
growth promotions); id. ¶¶ 33– 34 (performance appraisal system led to disparities in
retention rates).) Thus, P laintiffs attribute the alleged disparities in compensation,
promotion, and retention to “flaws in the design and implementation of the Lockheed

12

Martin performance appraisal system, including the nature of the calibration and
validation process.” (Id. ¶ 16.) 5
2.

Allegations Regarding The Named P laintiffs

P laintiffs Debra Josey and Vernon Ross allege that they have been personally
injured as a result of Lockheed’s performance appraisal system. (See id. ¶¶ 35– 56.)
P laintiff Josey is a current Lockheed employee and has worked at the company and its
predecessors since 1983. (S ee id . ¶ 7.) She is currently a Software Engineer Manager
in Lockheed’s Rotary and Mission Systems division in Florida, and she previously
worked as an Engineering Leadership Development P rogram Manager in Lockheed’s
Information Systems and Solutions division in Maryland. (See id.) Josey alleges that
she “received lower ratings than similarly situated white employees during her
employment at Lockheed Martin despite her stellar performance throughout the [2013–
2016] period.” (Id. ¶ 35.) Josey contends that the effects of the performance appraisal
system on her are manifest when one compares the relatively low overall tier
placements she received during the relevant period—second tier (out of four) for her
work in 2012, third tier for 2013, and third tier for 2014—with the high numerical
ratings and positive written comments that she received in her individual performance
reviews during that same timeframe. (See id. ¶¶ 35– 37.) Josey alleges that the thirdtier ratings she received for 2013 and 2014 “put her below average[,]” notwithstanding

5

Th e complaint also asserts t hat, “[i]n addition t o t he flaws in t he p erformance appraisal system,
[cert ain] changes imp lemented b y former Senior Vice President o f Human Resources John Lu cas
in creased t he discriminat ion against African A merican employees” and “diminished Lo ckheed Martin’s
effo rts t o encourage racial diversity at t he Comp any.” (Co mpl. ¶¶ 25–26.) These Lu cas-related
alleg ations appear in two stray p aragraphs t hat stand alone in t he complaint (see i d.), and John Lu cas is
n o t mentioned further. No specifics are p rovided regarding h is conduct, t he “changes” that were made
u n d er his supervision, or t he alleged “discrimination” t hat A frican-American employees suffered d ue to
h is actions.

13

the fact she received composite numerical ratings of 4.1 and 3.9 (out of 5.0) on her
individual performance reviews in those two years—ratings that, she alleges, “show that
she substantially exceeded Lockheed Martin’s expectations.” (Id. ¶ 36.) Moreover, in
contrast to the third-tier rating that she received in 2014, Josey’s performance review
for that same year “did not have any negative comments concerning her commitments
or her behavioral objectives.” (Id. ¶ 37.) Josey maintains that the divergence between
her individual performance review and her overall tier rating is attributable to the fact
that, during group “calibration sessions[,]” she was compared to employees with
“markedly different roles” from hers. (Id. ¶ 38.) Josey appealed her tier placement for
2013, but her appeal was unsuccessful. (See id. at 39.)
In the complaint, Josey alleges that, “[a]s a result of her lower-tier performance
ratings, [she] has been paid less than her white counterparts with the same or less
experience” (id. ¶ 40); that she did not receive any bonuses or awards in 2013 or 2014
(see id .

¶ 41) ; and that her performance review ratings have “negatively impacted her

ability to be promoted within the company” (id . ¶ 42). With respect to promotions,
Josey alleges that she unsuccessfully applied for 55 positions between 2012 and 2015,
of which at least 37 would have constituted promotions, and that in the four cases in
which she knows the identity of the candidate who was selected for the position, three
of the successful candidates were white employees who Josey believes were less
qualified than she. (See id. ¶ 43.) In 2015, Josey was notified of her impending layoff
as part of a reduction in force (a development that she believes would not have taken
place “if she had received higher ratings”); however, she avoided being laid off by

14

accepting a different position at the same level that required her to relocate from
Maryland to Florida. (Id. ¶ 44.)
P laintiff Vernon Ross is a former Lockheed employee who worked for the
company from 1991 to 2015, most recently as the Director of STEM, Generations, and
Higher Education in Human Resources in Lockheed’s Enterprise Operations division.
(S ee id . ¶ 6.) Like Josey, Ross alleges that he “received ratings lower than those of
similarly situated white employees during his time at Lockheed Martin despite
commendable work performance.” (Id. ¶ 46.) Ross also contends that “[t]he contrast
between [his] written reviews and his [tier ratings] shows that ratings have a weak
relationship to performance.” (Id. ¶ 48.) Ross was placed in the second tier out of four
for 2012, the third tier for 2013, and the third tier for 2014 (see id. ¶ 47), and alleges
that these tier placements stand in contrast with the uniformly positive written
comments that he received on his performance review for 2014 (see id. ¶ 48). Ross
“does not know against whom he was compared at calibration sessions,” but “his direct
manager told him that he was calibrated with all Human Resources Directors, regardless
of the wide variation in duties.” (Id. ¶ 50.) Ross alleges that because of his “unique
positions at Lockheed Martin throughout the period from 2011 through 2015 posing
unusual challenges[,]” there were “no proper comparators for his position.” (Id.) Ross
appealed the results of his 2013 performance appraisal, and although Lockheed changed
two of the individual numerical ratings on his performance reviews during the appeals
process, Ross’s overall tier placement remained unchanged. (See id. ¶ 53.)
The complaint alleges that Ross’s relatively low tier ratings “prevented him from
advancing within the Company[,]” and that “he was repeatedly bypassed in promotions

15

to a Vice P resident position in favor of white employees who often had lesser
credentials than him.” (Id. ¶ 52.) Ross adds that he “received only one long term
incentive award” during his 24 years as a Lockheed Martin employee. (Id. ¶ 51.) In
2015, Lockheed notified Ross that it planned to terminate his position as part of a
reduction in force (see id . ¶¶ 55– 56), and Ross then “applied to over 40 jobs within
Lockheed Martin,” some at his then-current level and some at lower levels, but he did
not receive any offers (id. ¶ 56). Ross alleges that he “was unable to find a new job due
in significant part to the discriminatorily low appraisal ratings he had received.” (Id .)
Ross’s employment was terminated at the end of 2015 (see id.), and according to the
complaint, he received the initial notice of termination “[s]hortly after” he filed a
charge of discrimination with the EEOC against Lockheed (id . ¶ 55). Ross also alleges
that the EEOC charge was not his first formal complaint: in July of 2014, he allegedly
“complained internally of the discrimination that he and other African American
employees faced at Lockheed Martin.” (Id. ¶ 54.)
3.

The Class Definition, Allegations, And Claims

As mentioned above, Josey and Ross seek to bring this lawsuit as a class action
on behalf of a class that the complaint defines as follows:
[A]ll African-American salaried employees below the level of Vice
P resident employed by Defendant in the United States for at least one
day between January 1, 2013 and February 29, 2016, and who received
at least one performance evaluation during that period with an overall
rating below “significantly exceeds commitments[.]”
(Id . ¶ 57.) Several groups of African-American employees who might otherwise be
included within this definition are expressly excluded from the putative class:
(1) employees who signed release agreements, (2) union-represented
employees, (3) individuals who have asserted claims of race
discrimination against Lockheed Martin, which remain pending before
16

any local, state or federal agency or in any state or federal court as of
the date of preliminary approval, (4) individuals employed by Sandia
Corporation, and (5) individuals who became (or become) employees of
Lockheed Martin or one of its subsidiaries as a consequence of stock or
asset acquisitions consummated on or after January 1, 2012 including,
but not limited to, the following transactions: Industrial Defender,
Materion Assets, Deposition Sciences, Astrotech AssetsZeta Associates,
Sun Catalytix, Systems Made Simple, and Sikorsky.
(Id .) Moreover, and notably, as it relates to the allegedly discriminatory performance
review process, the proposed class definition does not contain any objective criteria that
permit identification of the particular African-American employees who allegedly
suffered concrete injury as a result of Lockheed Martin’s performance appraisal
system—i.e., those African-American employees who demonstrably should have
received either higher numerical rankings or a higher tier rating than Lockheed assigned
to them during the relevant period. (See id.)
The complaint alleges that, as defined, the putative class contains “over 5,500
members who worked in over 40 states across the United States” (id . ¶ 59), and that
Ross and Josey “are members of the Class they seek to represent” (id . ¶ 58). The
complaint also contains several allegations regarding the putative class that expressly
aim to demonstrate that the prerequisites to class certification set forth in Federal Rule
of Civil P rocedure 23 are satisfied, and that are discussed at greater length below. (S ee
id . ¶¶ 59– 64; see a lso infra P art III.A.)
With respect to the legal claims that P laintiffs seek to maintain on behalf of the
class, the complaint alleges, in two separate counts, that Lockheed’s performance
appraisal system subjected members of the proposed class to intentional and disparate
impact race discrimination, and a third count alleges that P laintiff Ross was retaliated
against as a result of his discrimination complaints. (See Compl. ¶¶ 65– 79.) To be

17

specific, as mentioned above, Count I of the complaint alleges that “Lockheed Martin
intentionally discriminated against P laintiffs Ross and Josey and members of the
proposed Class on the basis of their race by assigning African-American employees
lower ratings than other employees in the performance appraisal system,” in violation
of 42 U.S.C. § 1981. (Id. ¶ 67.) In Count II, the complaint alleges that “Lockheed
Martin has maintained a pattern or practice of employment discrimination against
African-American employees in performance appraisal ratings” (id . ¶ 72), and that
Lockheed Martin’s policies and practices “have had a disparate impact against AfricanAmerican employees in performance appraisal ratings” (id . ¶ 73), both in violation of
Title VII of the Civil Rights Act of 1964. Finally, Count III of the complaint alleges
that Ross “has suffered [an] adverse employment action because Defendant retaliated
against him for filing a Charge of Discrimination with the EEOC and/or complaining to
senior executives at the Company of racial discrimination faced by him and other
African-American employees.” (Id. ¶ 78.) P laintiffs’ complaint seeks declaratory,
monetary, and injunctive relief. (See id., P rayer for Relief, ¶¶ a–d.)
C.

The Proposed Se ttlement Ag reement

On December 23, 2016, P laintiffs filed the putative class action complaint along
with a proposed Settlement Agreement, and asked this Court for both preliminary
certification of the proposed class for settlement purposes under Federal Rule of Civil
P rocedure 23(a) and (b), and preliminary approval of the Settlement Agreement under
Federal Rule of Civil P rocedure 23(e). (See P ls.’ Mot. for P relim. Approval of Class
Action Settlement Agreement (“Mot.”), ECF No. 4; see also Settlement Agreement, Ex.

18

A to Mot., ECF No. 4-1.) 6 The proposed Settlement Agreement provides programmatic
and monetary relief to members of the class. The programmatic relief primarily
consists of the establishment of a “council” that will recommend that certain changes be
made to Lockheed’s performance evaluation process (including ending the calibration
session procedure), ensure that diverse slates of candidates are considered for open
positions, and review salaries to ensure equity. (See Settlement Agreement at 28–33.)
Lockheed also agrees to implement inclusion training, to improve its collection of data
regarding performance evaluations and employee advancement, and to fulfill certain
obligations to report this data to its board of directors and to Class Counsel. (See id . at
34– 40.)
As for monetary relief, according to the Settlement Agreement, Lockheed has
agreed to pay $22.8 million into a fund that is to be distributed among the members of
the class who affirmatively opt to participate in the settlement based on criteria that a
Claims Administrator will determine after gathering information from the class
members, as described below. (S ee id. at 41– 47.) Regardless, P laintiffs’ counsel will
be paid 28% of this settlement fund as attorneys’ fees, plus $225,000 per year for four
years as compensation for their role in supervising the Settlement Agreement’s
programmatic relief. (See id . at 53.) Counsel would also be reimbursed for their
litigation expenses to date, which are approximately $125,000. (S ee id.; Mot. at 39.)

6

A ft er t he Co urt’s mo tion hearing, Plaintiffs also filed a supplement al memorandum in support of t heir
ap p roval mo tion (see Pls.’ Su ppl. M em. in Supp. o f Pls.’ M ot. for Prelim. A pproval (“Suppl. M em.”),
ECF No . 13), as well as a No tice o f Supplemental Authority alerting the Co urt to a recent D.C. Circuit
d ecision affirming a class certification order (see Pls.’ Notice o f Suppl. Authority, ECF No . 14
(d is cussing DL v. District of C olumbia, 860 F.3d 713 (D.C. Cir. 2017))).

19

For the class members’ part, the Settlement Agreement contains a broad release
of legal claims that extends well beyond any claims that arise from the allegedly
discriminatory operation of Lockheed’s performance appraisal system, and also requires
that class members act affirmatively to obtain any monetary benefit in exchange for
releasing their claims. That is, per the agreement, all class members who do not
affirmatively opt out will automatically release “any and all racial employment
d iscrimination claims of whatever nature, known or unknown,” including “any and a ll
cla ims o f racial discrimination in emp loyment o r in th e p rovisio n o f emp loyee b enefits”
that have arisen in the context of their employment relationship with Lockheed at any
time up until the moment the class member signs his or her claim form. (Settlement
Agreement at 22; see a lso Tr. of Oral Arg. at 27 (counsel for P laintiffs acknowledging
that the release effectively includes “[a]ny and all race discrimination claims[,]”
including a claim that “has nothing to do with the evaluations that [an employee]
receive[s] from [her] employer”); Suppl. Mem. at 12 (stating that “[u]nder the terms of
the settlement” in this case, class members “will release Lockheed Martin from a ll
types of racial discrimination claims” (emphasis added)).) 7 Yet, a class member who
wishes to receive any portion of the settlement fund as compensation for this broad
release of claims must go further, by undertaking to complete an extensive claim form
and timely submit it to the designated Claims Administrator. (See Settlement
Agreement at 44; see a lso Claim Form, Ex. 2 to Settlement Agreement, ECF No. 4-1.)
Thus, class members who do not respond to the notice of settlement in any fashion not

7
Th e n amed Plaintiffs also s pecifically agree t o release all claims t hey may h ave against Lo ckheed
M art in, whether o r n ot related t o race discrimination. (See Settlement A greement at 24.)

20

only have all of their discrimination and benefits claims extinguished, they also forfeit
entirely any opportunity to receive any monetary compensation for those extinguished
claims. 8
To facilitate award determinations, the claim form that class members are asked
to complete asks a wide-ranging series of questions that seek information not only
about the class members’ experiences in regard to Lockheed’s performance review
system (i.e., the subject matter of the complaint), but also about any instance of racebased discrimination that class members believe they may have suffered during their
employment at Lockheed. (S ee, e.g., Claim Form at 84 (asking about any form of race
discrimination); id. at 86 (asking about discriminatory non-promotions); id . at 87
(asking about discriminatory terminations).) The Settlement Agreement then leaves it
to the Claims Administrator to evaluate a class member’s individual legal claims and to
allocate the settlement fund accordingly. (See Settlement Agreement at 45– 47; Kovach
Decl. ¶¶ 11– 23; see a lso Kovach Decl. ¶ 15 (explaining that the Claims Administrator
“will also assign points based off of Claim Form responses”).)

8

Un d er t he p roposed Settlement Agreement, even those class members who d o comp lete t he claim fo rm
an d submit it t o t he Claims Admin istrator apparently are not automat ically entitled t o receive any
co mp ensation in exchange for releasing t heir claims. Du ring the hearing t hat t his Co urt held on April
24, 2017, Plain t iffs’ counsel o rally represented t hat “everyone will g et at least s omething[,]” b ut
co u nsel b ased t hat representation o n the fact that one o f the factors that the Claims A dministrator will
p u rportedly consider in d eciding how mu ch a p art icular class member will receive is “the number o f
weeks wo rked fo r Lockheed Martin d uring t he class p eriod[,]” and “there won’t b e anybody in the class
wh o h asn’t wo rked t here at least fo r some p eriod.” (Tr. o f Oral A rg. at 74 (d iscussing Settlement
A g reement at 45); see a lso Decl. of Lo ree Kovach (“Kovach Decl.”), Ex. 7 t o Su ppl. M em., ECF No.
13– 7, ¶ 11 (exp laining t hat t he length-of-emp loyment factor “will likely result in each Claimant
receiving at least some award”).) Ho wever, t he Sett lement A greement it self does not s pecify a
min imu m amo u nt o f compensation for each class member, and it n either inst ructs t he Claims
A d min ist rator regarding how to weigh each o f t he s ix lis ted factors n or n otifies class members of h ow
t h e Claims Admin istrator will g o about doing so. (See Settlement Agreement at 45.)

21

Significantly, the Settlement Agreement provides no guidance regarding how the
Claims Administrator will distribute the settlement fund, what the minimum recovery
amount will be for any class member who submits a claim form, or what the average
recovery will be (either for class members in general or for class members who indicate
on their claim form that they have experienced other types of discrimination). Nor does
the class notice provide any such information to the class members. (See generally
Notice of Class Action (“Class Notice”), Ex. 4 to Settlement Agreement, ECF No. 4-1.)
And because each person who is awarded money from the settlement fund “will be
required to keep the amount of their award confidential” (Settlement Agreement at 47),
it also appears that class members will never be able to discern this information. At the
motion hearing, P laintiffs’ counsel made clear that only after all class members have
decided whether or not to opt out and have returned their detailed claim forms will
P laintiffs’ counsel—with aid from the Claims Administrator—submit a final allocation
of the settlement fund to the Court for approval, and as a practical matter, this final
accounting might not take place until after the Court holds its final approval hearing.
(S ee Tr. of Oral Arg. at 86 (stating that “[t]he actual allocation I believe will still be in
process at th[e] time [of the final approval hearing]”).)
In anticipation of the process described above, the Settlement Agreement
contemplates that the Claims Administrator will send written notice of this lawsuit and
of the proposed settlement to all class members shortly after the Court issues a
preliminary approval order. (S ee Settlement Agreement at 43–44; see also Class
Notice.) The proposed notice offers each class member four options: (1) fill out and
return the attached claim form and participate in the settlement; (2) opt out of the

22

settlement; (3) submit an objection to the Court regarding the fairness of the settlement;
or (4) do nothing. (S ee Class Notice at 102.) Any opt-out request must include certain
identifying information, as well as a verbatim copy of the specific opt-out language that
appears in the class notice. (S ee id. at 110– 11.) 9 Individuals who do nothing—that is,
who neither follow the opt-out procedure nor submit a claim form—both forfeit any and
all race discrimination claims they may have against Lockheed and also lose the
opportunity to recover a portion of the settlement fund. (See Settlement Agreement at
24.) During the motion hearing, P laintiffs’ counsel estimated based on his experience
that approximately 30– 50% of class members will not respond to the notice in any
fashion. (S ee Tr. of Oral Arg. at 33– 34).
II.

LEGAL STANDARDS
The district court’s role in reviewing a proposed settlement agreement in a class-

action lawsuit follows a “three-stage process, involving two separate hearings[.]” 4
William B. Rubenstein, Newberg o n Cla ss Actions § 13:10 (5th ed. 2014); see also Fed.
Judicial Ctr., M anual f or Complex Litig ation § 21.632 (4th ed. 2004). “First, the parties
present a proposed settlement agreement to the court for so-called ‘preliminary
approval.’” 4 Newberg on Class Actions § 13:10 (emphasis omitted). “If the case is
presented for both class certification and settlement approval, the certification hearing
and preliminary fairness evaluation can usually be combined.” M anual f or Complex
Litig a tion § 21.632. “Second, if the court does preliminarily approve the settlement

9

If a cert ain n umber o f class members opt o ut of t he Settlement Agreement, Lo ckheed M artin will h ave
t h e o ption to either v oid the agreement o r retrieve fro m t he settlement fund $4,000 p er class member
wh o o pts o ut. (S ee Settlement Agreement at 19.) The Court g ranted the p arties’ joint mo t ion fo r leave
t o file t h e requisite n umber o f o pt-outs t hat t rigger t his provision u nder seal. (See Sealed M ot., ECF
No . 5; M in. Order o f Apr. 25, 2017.)

23

(and conditionally certify the class), notice is sent to the class describing the terms of
the proposed settlement” and explaining class members’ options with respect to the
settlement agreement, including the right to object to the proposed settlement. 4
Newb erg on Class Actions § 13:10 (emphasis omitted); see also Fed. R. Civ. P .
23(e)(1), (e)(5). Finally, the court holds a hearing after which “the court decides
whether or not to give ‘final approval’ to the settlement[,]” which “can also encompass
a decision certifying the class” if the court has not made that decision already. 4
Newb erg on Class Actions § 13:10.
The instant lawsuit is presently at the preliminary approval stage, and P laintiffs
seek an order that both preliminarily certifies the class and preliminarily approves the
Settlement Agreement. (S ee g enerally Mot.)
A.

Clas s Certification

P arties frequently seek to certify a class for settlement purposes, sometimes (as
in this case) because they have “settle[d] before . . . even a class action complaint has
been filed.” M anual f or Complex Litig ation § 21.132; see also Amchem Prods., Inc. v.
Win d so r, 521 U.S. 591, 618 (1997) (noting that “the ‘settlement only’ class has become
a stock device”). “When presented with a settlement-only class, a court must determine
whether the proposed class satisfies the requirements of Federal Rule of Civil P rocedure
23, with one exception: the court does not need to consider whether ‘the case, if tried,
would present intractable management problems[.]’” Alvarez v. Keystone Plu s Co nstr.
Co rp ., 303 F.R.D. 152, 159 (D.D.C. 2014) (quoting Amch em, 521 U.S. at 620). That
lone exception aside, the remaining class-certification requirements “demand undiluted,
even heightened, attention in the settlement context.” Amchem, 521 U.S. at 620.

24

A court may certify a class under Rule 23 only if it satisfies all of the
prerequisites set forth in Rule 23(a) and at least one of the three requirements of Rule
23(b). S ee Co mcast Corp. v. Behrend, 133 S. Ct. 1426, 1432 (2013). Under Rule 23(a),
the party seeking class certification must demonstrate that: “(1) the class is so numerous
that joinder of all members is impracticable; (2) there are questions of law or fact
common to the class; (3) the claims or defenses of the representative parties are typical
of the claims or defenses of the class; and (4) the representative parties will fairly and
adequately protect the interests of the class.” Fed. R. Civ. P . 23(a). Courts generally
refer to these prerequisites as “numerosity, commonality, typicality, and adequacy of
representation[.]” Amgen v. Conn. Retirement Plans & Trust Funds, 568 U.S. 455, 460
(2013). As relevant here, in order to satisfy the commonality requirement, “[class
members’] claims must depend upon a common contention[,]” and “[t]hat 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.” Wal-M art, 564 U.S. at 350.
Rule 23(b) lays out different requirements for three different “types of class
actions.” Fed. R. Civ. P . 23(b). P laintiffs here seek certification under Rules 23(b)(2)
and 23(b)(3), relying on Rule 23(b)(2) for purposes of their requested injunctive relief
and Rule 23(b)(3) for purposes of their requested monetary relief. (See Mot. at 27– 30.)
S ee a lso Eubanks v. Billington, 110 F.3d 87, 96 (D.C. Cir. 1997) (explaining that a
court “may adopt a ‘hybrid’ approach” under Rule 23(b), “certifying a (b)(2) class as to
the claims for declaratory or injunctive relief, and a (b)(3) class as to the claims for
monetary relief”); a ccord 2 Newberg o n Cla ss Actions § 4:1. Under Rule 23(b)(2), a

25

class action may be maintained if “the party opposing the class has acted or refused 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). Under Rule 23(b)(3), a class action may be maintained if “the court
finds that the questions of law or fact common to class members predominate over any
questions affecting only individual members, and that a class action is superior to other
available methods for fairly and efficiently adjudicating the controversy.” Fed. R. Civ.
P . 23(b)(3).
When certifying a class, the court “must define the class and the class claims,
issues, or defenses[.]” Fed. R. Civ. P . 23(c)(1)(B). “Defining the class is of critical
importance” because, among other things, “it identifies the persons . . . entitled to relief
[and] bound by a final judgment,” and as a result, “[t]he definition must be precise,
objective, and presently ascertainable.” M anu al f o r Co mp lex Litig ation § 21.222. “The
class definition should describe the operative claims, issues, or defenses, such as injury
resulting from securities fraud or denial of employment on account of race.” Id.; see
a lso Ko h en v. Pac. Inv. Mgmt. Co. LLC, 571 F.3d 672, 677 (7th Cir. 2009) (“[I]f the
definition is so broad that it sweeps within it persons who could not have been injured
by the defendant’s conduct, it is too broad.”); Th o rp e v. District o f Co lumb ia, 303
F.R.D. 120, 141– 42 (D.D.C. 2014) (assessing whether class definition was “fatally
overbroad” by virtue of including individuals who had not plausibly suffered an injury).
Notably, courts have taken divergent approaches to applying Rule 23 when a
party moves for a “preliminary” or “conditional” certification order and requests
preliminary approval of a class settlement agreement as a prelude to distributing notice

26

to the class. S ee 4 Newberg on Class Actions § 13:18 (describing the split in authority).
“Most courts” have held that “a less stringent standard applies at the preliminary
approval phase with regard to the requirements for class certification.” Id. (citing, e.g.,
S cho enbaum v. E.I. Du Pont de Nemours & Co., No. 4:05CV01108, 2009 WL 4782082,
at *5 (E.D. Mo. Dec. 8, 2009)). Under this approach, at the preliminary approval stage,
the court “make[s] a preliminary determination that the proposed class satisfies the
criteria set out in Rule 23(a) and at least one of the subsections of Rule 23(b).” M anual
f o r Co mp lex Litigation § 21.632. This approach “contemplates that formal class
certification will be combined with the [final] fairness hearing,” and thus, at the
preliminary approval stage, the judge merely scans for “obvious impediments to class
certification” so as to save the plaintiff the trouble of noticing the hearing on final
settlement approval in cases where the court foresees that the formal certification
motion will be a non-starter. Sch oenbaum, 2009 WL 4782082, at *5.
Other courts, however, “undertake a full certification analysis under Rule 23(a)
and (b) before addressing whether a proposed settlement should be p relimin arily
approved.” 4 Newberg o n Class Actions § 13:18 (emphasis in original) (citing, e.g.,
Ep h edra Prods. Liab. Litig., 231 F.R.D. 167, 170 (S.D.N.Y. 2005)). In explaining this
approach, the court in Eph edra Products observed that the 2003 amendments to Rule 23
deleted preexisting language that had provided for “conditional” class certification, and
the Ep h edra Products court inferred from that change that “Rule 23 must be rigorously
applied even at th[e] ‘preliminary’ stage.” 231 F.R.D. at 170. In addition, it appears
that the more rigorous approach finds support in the text and structure of Rule 23’s
provisions, because the settlement rule (Rule 23(e)) specifically confers upon “class

27

members” the right to receive notice and object to the settlement proposal—objections
that the court will consider at the final hearing, see Fed. R. Civ. P . 23(e)(1), (e)(5)—and
thus strongly suggests that the class must be formed and the class members identified
p rio r to the court’s issuance of the notice and convening of the settlement hearing.
In any event, at the very least, district courts appear to have discretion under the
Federal Rules to decide whether to subject a putative class to the full force of Rule 23
at the preliminary approval stage or merely screen for obvious impediments to class
certification and leave the formal certification decision for the final approval hearing.
Cf . La rio noff v. United S tates, 533 F.2d 1167, 1183 (D.C. Cir. 1976) (holding that
district courts have discretion regarding the timing of the certification decision); Fed.
R. Civ. P . 23(c)(1)(A) (providing that courts should decide whether or not to certify a
class “[a]t an early practicable time”); id . 2003 advisory committee note (enumerating
various “considerations [that] may affect the timing of the certification decision”);
LCvR 23.1(b) (providing that, “in the exercise of its discretion[,]” a court may extend
the deadline for a certification motion prescribed by the Local Rules). 10 In this Court’s
10

Vario u s contextual factors mig ht in fluence a court’s choice bet ween t hese t wo d ifferent approaches in
a p art icular case. S ee 7AA Ch arles Alan Wright, et al., Federal Practice and Procedure § 1785.3 (3d
ed . 2005) (“The t ime at wh ich the court fin ds it appropriate to make its class-action d etermination may
v ary with t he circumstances of t he p articular case.”). In some cases, fo r examp le, a court mig ht choose
t o d elay t he certification d ecision b ecause it is likely t hat in formation p ertinent t o Rule 23(a)’s
req u irements will come to lig ht d uring t he n otice and o bjection p rocesses t hat t ake p lace after t he
p relimin ary approval stage. In o ther cases, judicial economy mig ht b e best served b y deciding the
cert ification issue early (i.e., at t he p reliminary approval stage), so t hat t he notices required by Ru les
23(c)(2) (regarding certification) and 23(e)(1) (regarding s ettlement) can be combined. S ee Manual fo r
C o mplex Litigation § 21.633. In yet o ther cases, concerns wit h the class definition mig ht lead a court
t o resolve t he certification issue, and thereby “define t he class and t he class claims, issues o r defenses,”
Fed . R. Civ . P. 23(c)(1)(B), b efore the notice is issued and t he p utative class members are fo rced t o
ch o ose h ow to respond. S ee 7A Federal Practice and Procedure § 1760 (noting t hat “hav[ing] the class
d es cribed in some d etail . . . may p rove essential in actions under Ru le 23(b)(3) in o rder t o g ive class
memb ers t he n otice required by Ru le 23(c)(2), t hereby enabling them t o decide whether t o opt out o f
t h e action[,]” and early certification “is o f considerable value . . . in co nnection with t he n otice and
rev iew of a d ismissal o r compromise o f the action under Rule 23(e)”).

28

view, the expansive class definition proposed in P laintiff’s complaint warrants squarely
addressing class certification at the present juncture.
B.

Court Approval Of Clas s Settlement Ag reements

Once the court has certified a class—whether in the context of adversarial
litigation or for settlement purposes—the court must play an active role in any
agreement among the present parties that settles the case. See Fed. R. Civ. P . 23(e)
(“The claims, issues, or defenses of a certified class may be settled, voluntarily
dismissed, or compromised only with the court’s approval.”); see a lso M anual f or
Co mp lex Litigation § 21.61 (explaining that court approval under Rule 23(e) is required
regardless of “[w]hether a class action is certified for settlement or certified for trial
and later settled”). In this supervisory role, the court must scrutinize any proposed
settlement agreement so as to “protect[] unnamed class members ‘from unjust or unfair
settlements affecting their rights[.]’” Amchem, 521 U.S. at 623 (quoting 7B Fed eral
Pra ctice a nd Procedure § 1797). With this concern in mind, the court’s review of a
proposed settlement agreement “must be exacting and thorough[,]” M anual f or Complex
Litig a tion § 21.61; indeed, some courts “have gone so far as to characterize the court’s
role as akin to the high duty of care that the law requires of fiduciaries.” Wong v.
Accretive Health, Inc., 773 F.3d 859, 862 (7th Cir. 2014) (internal quotation marks and
citation omitted); see a lso 4 Newb erg on Class Actions § 13:40 (“[T]he court is said to
have a ‘fiduciary duty’ toward absent class members during the settlement of a class
suit.”). And the legal standard set forth in Rule 23(e) itself dictates that, if a proposed
settlement “would bind class members, the court may approve it only after a hearing
and on finding that it is fair, reasonable, and adequate.” Fed. R. Civ. P . 23(e)(2).

29

The text of Rule 23(e) does not expound upon what constitutes a “fair,
reasonable, and adequate” settlement agreement, but an advisory committee note to the
2003 amendments to that rule provides that “[f]urther guidance can be found in the
Manual for Complex Litigation.” Fed. R. Civ. P . 23(e) 2003 advisory committee’s note.
That resource provides as follows:
Fairness calls for a comparative analysis of the treatment of class
members vis-à-vis each other and vis-à-vis similar individuals with
similar claims who are not in the class. Reasonableness depends on an
analysis of the class allegations and claims and the responsiveness of
the settlement to those claims. Adequacy of the settlement involves a
comparison of the relief granted relative to what class members might
have obtained without using the class action process.
M a n ua l f o r Complex Litigation § 21.62. The Manual goes on to provide a long and
non-exhaustive list of factors that courts may consider in the context of the Rule 23(e)
inquiry, including “the advantages of the proposed settlement versus the probable
outcome of a trial on the merits of liability and damages”; “the fairness and
reasonableness of the procedure for processing individual claims under the settlement”;
and “the apparent intrinsic fairness of the settlement terms.” Id. The D.C. Circuit has
not announced any particular set of factors that guide the Rule 23(e) fairness inquiry,
but
courts in this circuit generally consider five factors: (1) whether the
settlement is the result of arm’s-length negotiations; (2) the terms of the
settlement in relation to the strength of plaintiffs’ case; (3) the status of
the litigation proceedings at the time of settlement; (4) the reaction of
the class; and (5) the opinion of experienced counsel.
Alva rez, 303 F.R.D. at 163. The context of each particular case determines “which
factors apply and what weight to give them.” M a n ual f or Co mp lex Litig ation § 21.62.

30

In the context of a motion for preliminary approval of a proposed settlement
agreement, a court “must make a preliminary determination on the fairness,
reasonableness, and adequacy of the settlement terms[.]” Id. § 21.632. “Rule 23 itself
does not outline” the preliminary settlement approval process, “nor, therefore, does it
provide district courts a standard by which to adjudicate the motion for preliminary
approval.” 4 Newberg o n Class Actio ns § 13:10. In this district, judges presented with
motions for preliminary approval of a class settlement agreement typically “consider
(1) whether the proposed settlement appears to be the product of serious, informed,
non-collusive negotiations, (2) whether it falls within the range of possible judicial
approval, and (3) whether it has any obvious deficiencies, such as granting unduly
preferential treatment.” Richardson v. L’Oreal USA, Inc., 951 F. Supp. 2d 104, 106– 07
(D.D.C. 2013) (internal quotation marks and citation omitted); see also 4 Newberg o n
Cla ss Actions § 13:10 (reporting that “courts in most circuits use some variation” of
this test). Although “the standard that governs the preliminary approval inquiry is less
demanding than the standard that applies at the final approval phase[,]” a court must
“not simply rubber-stamp a motion for preliminary approval [of a class settlement
agreement], and review is more than perfunctory.” 4 Newb erg o n Cla ss Actions § 13:13
(internal quotation marks and citations omitted). Ultimately, “[p]reliminary approval of
a proposed settlement to a class action lies within the sound discretion of the court.”
Rich a rdson, 951 F. Supp. 2d at 106 (internal quotation marks and citation omitted).
III.

ANALYSIS
P laintiffs have asked this Court to certify preliminarily that their discrimination

lawsuit against Lockheed Martin can proceed as a class action, and they also request

31

that this Court provide preliminary approval of the parties’ proposed Settlement
Agreement. As explained fully below, the Court is not persuaded that Rule 23(a)(2)’s
commonality requirement is satisfied, nor is it convinced that the significant fairness
issues that are apparent with respect to the terms of the Settlement Agreement can be
overcome. Therefore, this Court cannot assent to either of P laintiffs’ requests.
A.

Plaintiffs Have Failed To Demonstrate That The Proposed Class
Action Satis fies The Co mmonality Prerequisite Fo r Ce rtification
Unde r Rule 23

P laintiffs propose to represent a class of individuals that encompasses all current
and former African-American employees at Lockheed (subject to limited exclusions not
relevant here) who received a less-than-perfect performance evaluation on at least one
occasion during a three-year period. (See Compl. ¶ 57.) P laintiffs maintain that there
are more than 5,500 such individuals (see id. ¶ 59), and they argue that their proposed
class satisfies the commonality requirement set forth in Rule 23(a)(2) “because . . . each
member of the Class was subject to the same performance evaluation system at
Lockheed Martin.” (Mot. at 23; see also id. (“The discrimination in pay, promotions,
and retention experienced by each member of the proposed Class is bound together by
the common, unifying system that produced the unlawful disparities complained of.”).)
Unfortunately for P laintiffs, the Supreme Court’s landmark decision in Wal-M art
S to res, In c. v. Dukes, 564 U.S. 338 (2011), spoke directly to what is required in order to
satisfy the commonality criterion in the context of an employment discrimination class
action, and its reasoning makes abundantly clear that the commonality criterion is not
satisfied under the circumstances presented in this case.

32

1. The Supreme Court Has Held That, To Maintain An Employment
Discrimination Case As A Class Action, Rule 23(a) Commonality
Requires That The Employer Have Discriminated Against All Class
Members In The Same Way
The Supreme Court’s Wal-M art decision “changed the landscape” that a district
court must navigate when considering whether a putative class action satisfies Rule
23(a)’s commonality requirement. DL v. District o f Co lumb ia, 713 F.3d 120, 126 (D.C.
Cir. 2013). In Wal-Mart, three current and former female employees sued Wal-Mart
Stores, Inc.— “the Nation’s largest private employer[,]” 546 U.S. at 342—and alleged
that the company’s pay and promotion practices violated Title VII’s prohibition on sex
discrimination under both disparate-treatment and disparate-impact theories. See id . at
343– 45. Notably, “[p]ay and promotion decisions at Wal-Mart [were] generally
committed to local managers’ broad discretion, which [was] exercised ‘in a largely
subjective manner.’” Id. at 343 (quoting Duk es v. Wal-Mart Stores, Inc., 222 F.R.D.
137, 145 (N.D. Cal. 2004)). Certain promotions were available only to employees who
had “above-average performance rating[s]” and satisfied other objective criteria, but
even for those decisions, senior managers had discretion to choose among all employees
who satisfied those criteria. Id. The plaintiffs’ theory of the case was that “a strong
and uniform ‘corporate culture’ permit[ted] bias against women to infect, perhaps
subconsciously, the discretionary decisionmaking of each one of Wal-Mart’s thousands
of managers—thereby making every woman at the company the victim of one common
discriminatory practice.” Id. at 345. Accordingly, the three named plaintiffs sought to
litigate their case as a class action on behalf of all current and former female Wal-Mart
employees, a group that was approximately 1.5 million strong. See id . at 342.

33

When it analyzed whether the case could be certified as a class action under Rule
23, the Supreme Court first observed that “[t]he class action is an exception to the usual
rule that litigation is conducted by and on behalf of the individual named parties
only[,]” id . at 348 (emphasis added) (internal quotation marks and citation omitted),
and that Rule 23 only permits class-action treatment where the class claims are “fairly
encompassed by the named plaintiff’s claims[,]” id. at 348– 49 (internal quotation marks
and citations omitted). With respect to Rule 23’s commonality requirement—which the
Justices called “[t]he crux of th[e] case”—the Supreme Court cautioned that the Rule’s
“language is easy to misread” because, at first blush, virtually every class complaint
raises “questions of law or fact common to the class.” Id. at 349; see also id. at 349– 50
(explaining that overarching questions such as “Do our managers have discretion over
pay?” and “Is that an unlawful employment practice?” are certainly common to all class
members, but are insufficient to establish commonality because they “give[] no cause to
believe that all [class members’] claims can productively be litigated at once”). Beyond
the mere identification of basic facts or general principles that are common to the
members of the class, the Court explained, the commonality requirement demands that
all class members’ claims “must depend on a common co n tention . . . [that] is capable
of classwide resolution[,]” such that “determination of [that contention’s] truth or
falsity will resolve an issue that is central to the validity of each one of the claims in
one stroke.” Id. at 350 (emphasis added). Thus, “for example, the assertion of
discriminatory bias on the part of the same supervisor” might satisfy the commonality
requirement because, under that circumstance, a common contention that is capable of

34

common proof—i.e., that supervisor exercised discretion in a discriminatory fashion—
would lie at the center of each class member’s claim. Id. (emphasis added).
The Court also expounded on the commonality requirement as it applies in the
specific context of Title VII employment-discrimination class actions, explaining that
because the core of any individual plaintiff’s Title VII claim is “the reason for a
particular employment decision,” litigating many such claims (arising out of many
different employment decisions) at once is only productive if “some glue hold[s] the
alleged reasons for all those decisions together,” id . at 352 (emphasis in original)
(internal quotation marks and citation omitted). 11 P ut another way, the commonality
requirement is met in an employment discrimination case only where the plaintiff
demonstrates “that examination of all the class members’ claims for relief will produce
a common answer to the crucial question why was I disf avored.” Id. (emphasis in
original).
In other words, the “common contention” or “glue” that binds an employment
discrimination class together must bridge a “conceptual gap” between “an individual’s
claim that he has been denied a promotion [or higher pay] on discriminatory grounds,
and . . . the existence of a class of persons who have suffered the same injury as that
individual[.]” Id. at 352– 53 (quoting Gen. Tel. Co. of the Sw. v. Fa lcon, 457 U.S. 147,
157– 58 (1982)). And, notably, the Wal-M art decision recognized two possible methods
by which a plaintiff might establish the requisite common contention: a plaintiff must
either (1) identify a “testing procedure or other companywide evaluation method that

11

In t h e in stant case, t he Plaintiffs h ave pled one o f their t wo class claims u nder 42 U.S.C. § 1981
rat h er t han Title VII, b ut t he s tandards articulated in Wa l-Mart are “equally apposite t o claims u nder
t h [at] statute[.]” Burton v. District of C olumbia, 277 F.R.D. 224, 229 n .6 (D.D.C. 2011).

35

can be charged with bias[,]” or (2) provide “[s]ignificant proof that an employer
operated under a general policy of discrimination[.]” Id. at 353 (quoting Fa lco n, 457
U.S. at 159 n.15). Requiring that one of these avenues be pursued ensures that, when
evaluating the certification motion, the court does not lose sight of the need to
“‘identify[] the specific employment practice that is challenged’” as producing a
racially disparate outcome, id.at 357 (quoting Wards Cove Packing Co., 490 U.S. at
656), which is a key task that even individual plaintiffs must accomplish in a Title VII
case. Cf . Falcon, 457 U.S. at 159 n.15 (“[I]t is noteworthy that Title VII prohibits
discriminatory p ra ctices, not an abstract policy of discrimination.” (emphasis in
original)). (S ee a lso P art I.A, supra.)
Turning to the particular claims at issue in Wal-M art, the Supreme Court
observed that the named plaintiffs had not sought to identify a biased companywide
evaluation method, and had also failed in their attempt to provide “significant proof that
Wal-Mart operated under a general policy of discrimination.” Wal-M art, 564 U.S. at
353 (internal quotation marks omitted). The Court noted that the only policy the
plaintiffs had identified was the “‘policy’ of a llo win g d iscretion by local supervisors
over employment matters[,]” id. at 355 (emphasis in original), and it explained that,
because the simple fact of delegating decisions “should itself raise no inference of
discriminatory conduct,” id. (quoting Watson, 487 U.S. at 990), to establish sufficient
commonality, the plaintiffs needed to “identif[y] a common mode of exercising
discretion that pervades the entire company[,]” id . at 356. The plaintiffs failed to do so,
the Court held, because they had produced only regional and national disparity data
rather than the sort of granular, store-by-store data that could demonstrate truly

36

pervasive discrimination. S ee id . at 356– 57. And the Court made clear that even more
granular disparity data would not necessarily have cured the commonality defect,
because such data alone could not establish the necessary “specific employment
practice” (i.e., a “common mode of exercising discretion”) that produced a disparate
impact in a manner that was common to the class. Id. at 356– 57.
In the end, after analyzing facts and claims that are strikingly similar to those
that are presented in the instant case, the Supreme Court offered a summation for its
discussion of commonality in a Title VII employment class action that echoes
resoundingly in the case before this Court: “Merely showing that [an employer]’s policy
of discretion has produced an overall sex-based disparity does not suffice.” Id. at 357.
2. P laintiffs Have Not Established That Lockheed’s P erformance Review
System Causes Racially Disparate Outcomes In A Manner That Can Be
Established Through Common P roof
As explained, the plaintiffs in Wal-M art sought to litigate employment
discrimination claims on behalf of all female current and former Wal-Mart employees.
If Wa l-M a rt teaches anything, it is that plaintiffs can maintain such an employment
discrimination lawsuit as a class action only if they can identify a company-wide
“common contention” regarding the reasons that each member of the class has suffered
an injury. S ee 564 U.S. at 350. As discussed above, Wal-M art recognized two methods
by which a plaintiff can thread together many individuals’ employment discrimination
claims so as to demonstrate that they can productively be litigated at once: by pointing
to a “testing procedure or other companywide evaluation method that can be charged
with bias[,]” or by providing “[s]ignificant proof that an employer operated under a
general policy of discrimination[.]” Id. at 353 (quoting Fa lco n, 457 U.S. at 159 n.15).

37

In this Court’s view, P laintiffs in the instant case have done neither, and thus have
fallen far short of establishing a commonality of issues as Rule 23(a)(2) requires.
First of all, try as they might, P laintiffs have failed to identify a “testing
procedure or other companywide evaluation procedure that can be charged with bias.”
Id . (emphasis added). This is so notwithstanding P laintiffs’ vigorous efforts to
characterize Lockheed’s performance appraisal system as a “companywide evaluation
method” (Suppl. Mem. at 2 (internal quotation marks and citation omitted)), and to
assert that several common elements of that system are “flawed” (Compl. ¶ 11).
P laintiffs allege, for example, that the group “calibration sessions” at which multiple
managers were required to discuss many employees and assign them into one of four
performance “tiers” suffered from inadequate information about each employee under
consideration and often required apples-to-oranges comparisons between employees
with different positions. (S ee Compl. ¶ 23.) But the contention that the companywide
evaluation procedures often resulted in ratings that were poorly correlated with job
performance (see id . ¶¶ 17– 24), however plausible, does not supply an account of how
those procedures themselves resulted in the racially disparate outcomes that P laintiffs
have observed in Lockheed’s overall workforce. That is, in order make a plausible
“charge[]” that a companywide evaluation method is infected “with bias[,]” Wal-M a rt,
564 U.S. at 353, it is clear that a plaintiff must provide some “detail about how th[at]
examination[] operated in a biased way.” Burton, 277 F.R.D. at 229 (emphasis in
original); see also Wards Cove, 490 U.S. at 657 (explaining that in a disparate impact
case, “a plaintiff must demonstrate that it is the application of a specific or particular

38

employment practice that has created the disparate impact under attack”). P laintiffs
here have taken no steps whatsoever toward accomplishing this critical goal.
Nor have P laintiffs provided “significant proof” that Lockheed “operated under a
general policy of discrimination.” Wal-M art, 564 U.S. at 353. As was the case in Wa lM a rt, P laintiffs have not identified any affirmative policy that had a demonstrably
discriminatory intent or effect on African-American Lockheed employees; instead, the
only connection that P laintiffs have mustered between Lockheed’s policies and the
allegedly racially disparate performance-review outcomes is the allegation that “[t]he
absence of measurable indicators” in Lockheed’s individual performance reviews
“resulted in inadequate safeguards against bias in the assessment of African American
employees.” (Compl. ¶ 18; see also id. ¶ 19.) But just as in Wal-M art, that allegation
attributes racially disparate outcomes to the “policy” of committing unrestrained
discretion to individual supervisors, see 564 U.S. at 355, and such a discretionary
system can only amount to a policy of discrimination that is susceptible to common
proof if P laintiffs “identif[y] a common mode of exercising discretion that pervades the
entire company[,]” id. at 356. The mere fact that all of Lockheed’s supervisors used the
same allegedly ill-defined numerical rubric to grade employee performance (see Compl.
¶¶ 17– 21)—an aspect of the system that constrained supervisor discretion, even if
inadequately—says nothing about how individual supervisors exercised what discretion
was left to them, and after Wal-M art, it is clear beyond cavil that such silence isn’t
enough.
To state the point succinctly: in order to establish the requisite commonality with
respect to a discrimination challenge to an employee-review system that permits various

39

managers to exercise discretion, P laintiffs needed to demonstrate “that all managers
would exercise their discretion in a common way[,]” Wal-M art, 564 U.S. at 356, and
they have not done so. Other than pleading, “upon information and belief,” that “the
racial disparity in ratings [under Lockheed’s performance evaluation system] is
statistically significant” (Compl. ¶ 15), P laintiffs have pointed to no evidence of biased
decision making of any kind, and certainly not statistical evidence of the type that
demonstrates that the discretionary ratings decisions led to racially disparate outcomes
in a co mmon way. See Wal-Mart, 564 U.S. at 357; see also Watson, 487 U.S. at 994
(plurality opinion) (noting that in a disparate impact case, “the plaintiff must offer
statistical evidence of a kind and degree sufficient to show that the practice in question
has caused the exclusion of applicants for jobs or promotions because of their
membership in a protected group”).
Notably, even the anecdotal allegations of the two named plaintiffs fall short of
giving rise to any belief that a common mode of discretionary decision making resulted
in consistent race discrimination against the entire class. First of all, two anecdotes in a
class of over 5,500 almost certainly do not constitute “substantial proof” that any
commonalities between them are pervasive throughout the class. See Wa l-Mart, 564
U.S. at 358 (explaining that “roughly one account for every eight members of the class”
might substantiate an inference that discrimination was pervasive, but “about 1 for
every 12,500 class members” would not). Second, if anything, the allegations of the
two named plaintiffs are more notable for their differences than for their similarities.
P laintiff Josey points to a disparity between her high individualized numerical ratings
and her below-average tier rating as evidence of discrimination (see Compl. ¶ 36),

40

whereas P laintiff Ross alleges no such disparity (see id. ¶¶ 45– 56). P laintiff Ross
intimates that he was discriminated against by virtue of being lumped in with other
employees during calibration sessions even though “there are no proper comparators for
his position” (id. ¶ 50), whereas P laintiff Josey alleges that her calibration sessions
were suspect because she was compared against both managers and non-managers (see
id . ¶ 38). All in all, then, it is clear to this Court that P laintiffs have failed to
demonstrate that the challenged performance evaluation system resulted in race
discrimination against all class members through a mechanism that could be established
through common proof.
The cases that P laintiffs highlight in which courts have certified employmentdiscrimination classes on the basis of companywide discretionary evaluation systems all
involved some thread that tied the many discretionary decisions together—the very
thing that is absent here. For example, P laintiffs rely on M cReyn olds v. M errill Lyn ch,
Pierce, Fenner & Smith, Inc., a Seventh Circuit decision affirming certification of a
class of African-American financial advisors that sued their employer for racially
discriminatory compensation practices. See 672 F.3d 482, 483 (7th Cir. 2012),
a b ro gated on other grounds as recognized by Ph illips v. S h eriff o f Co ok Cty., 828 F.3d
541, 559 (7th Cir. 2016). (S ee Suppl. Mem. at 5– 6.) In that case, the defendant
“delegate[d] discretion over decisions that influence the compensation” of individual
employees to many different supervisors, but the plaintiffs pointed to a companywide
policy that plausibly infected those decisions with discrimination in a common manner.
Id . at 488. Specifically, the plaintiffs identified a “teaming policy” that enabled brokers
who were members of teams to increase their compensation, see id., and they alleged

41

that, because individual brokers were allowed to pick their own team members pursuant
to that policy, the teams became “little fraternities” in which “brokers choose as team
members people who are like themselves” along racial lines, id . at 489. Thus, the
companywide “teaming” policy explained the common and allegedly discriminatory
manner in which the brokers were exercising their discretion in a way that was central
to the plaintiffs’ disparate impact claims, and as a result, the court found that the
policy’s legality was susceptible to common proof. See id.
P laintiffs also highlight M oore v. Napolitano, a decision from this District in
which current and former employees of the United States Secret Service sought
certification for a putative class action alleging discrimination in the Secret Service’s
promotion practices. S ee 926 F. Supp. 2d 8, 11–12 (D.D.C. 2013), p et. f o r p ermission
to a p p eal d enied sub nom. In re Jo hnson, 760 F.3d 66 (D.C. Cir. 2014). (See Mot. at
29.) “[D]ifferent decision makers no doubt injected some subjectivity into the
evaluations of different class members” in that case; however, crucially, “every
promotion decision was ultimately made by the Director of the Secret Service.” In re
Jo h n son, 760 F.3d at 73; see also Wal-Mart, 564 U.S. at 350 (hypothesizing that “the
assertion of discriminatory bias on the part of the same supervisor” could satisfy the
commonality requirement). P laintiffs have not alleged any similarly centralized
performance evaluation process here.
Consequently, unlike in the cases P laintiffs cite, P laintiffs have not identified a
“common contention” that “is central to the validity of each one of the [class members’]
claims” and that “is capable of classwide resolution[.]” Wal-Mart, 564 U.S. at 350. It
cannot be overstated that P laintiffs’ mere reliance on an alleged racial disparity in

42

overall employment outcomes (see Compl. ¶ 15) does not establish commonality,
because “merely proving that [a] discretionary system has produced a racial or sexual
disparity is not enough[,]” Wal-Mart, 564 U.S. at 357 (emphasis in original). And it is
similarly insufficient to point to the performance appraisal system as a whole and
contend, as P laintiffs do, that it acted as “an identical or similar headwind against all
class members.” (Suppl. Mem. at 4.) Rather, in any disparate impact case, “a plaintiff
must demonstrate that it is the application of a specific or particular employment
p ra ctice that has created the disparate impact under attack.” Wards Cove, 490 U.S. at
657 (emphasis added). This important work is required in all disparate impact cases,
but it “is all the more necessary when a class of plaintiffs is sought to be certified”
because the court must satisfy itself that the class members’ claims are susceptible to
common proof. Wal-M art, 564 U.S. at 357.
With neither an account of how the common features of the performance
appraisal system led to racially disparate outcomes (other than by enabling
discretionary decisions by individual supervisors), nor “statistical evidence of a kind
and degree sufficient to show that the practice in question has caused” race
discrimination, see Wa tso n, 487 U.S. at 994 (plurality opinion), P laintiffs cannot assure
this Court that class members will actually share “a common answer to the crucial
question why was I disf avored.” Wal-Mart, 564 U.S. at 352 (emphasis in original).
Indeed, common answers are especially elusive under the circumstances presented here,
as P laintiffs apparently anticipated, given the broad class definition (which contains no
limitation that ensures all class members have suffered the same adverse action or have
otherwise been personally affected by Lockheed Martin’s performance appraisal

43

system), and the expansive series of probing questions on the proposed claim form.
P laintiffs’ expected class clearly encompasses individuals with widely varying
experiences of discrimination. (S ee Compl. ¶ 57 (proposing a class definition that lacks
any injury criteria and thus includes a broad and diverse swath of employees); see also
Suppl. Decl. of Cyrus Mehri, ECF No. 13-6, ¶ 4 (admitting that “[t]he Claim Form . . .
does not limit the compensation, promotion, and termination claims for which
Claimants will be compensated to those for which they could establish a link between
their performance appraisal ratings and the allegedly discriminatory actions”).) 12 Such
potential breadth of experiences and claims among the putative class members is not the
mark of a class that meets the commonality requirement of Rule 23(a).
B.

Plaintiffs Have Failed To M ake A Pre liminary Sho wing That The
Te rms Of The Pro posed Settlement Agreement Are “Fair, Reasonable,
And Ade quate”

Even if this lawsuit could be certified as a class action consistent with Rule 23,
its settlement could not be approved at this juncture without “a preliminary
determination on the fairness, reasonableness, and adequacy of the settlement terms[.]”
M a n ua l f o r Complex Litigation § 21.632; see also Fed. R. Civ. P . 23(e)(2) (“If [a
proposed settlement] would bind class members, the court may approve it only after a

12

Th e absence of any in jury-focused limit ation in Plaintiffs’ class d efinition is t elling, and it stands in
s t ark contrast with class d efinitions in similar cases in which classes h ave b een certified. S ee, e.g.,
Mo o re, 926 F. Su pp. 2d at 16 (certifying class o f “[a]ll current and fo rmer African-American Special
A g ents who b id for p romo tion to a GS-14 p osition from 1995–2004 and w ere not p romoted t o GS-14 on
t h e fi rst b id l ist o n which they bid” (emphasis added)); McReynolds v. S odexho Marriott Servs., Inc.,
208 F.R.D. 428 (D.D.C. 2002) (certifying class of “all A frican-Americans wh o are o r were s alaried
emp lo yees o f So dexho . . . wh o h ave held or sought t o obtain [management p ositions], and who h ave
b een, continue to b e, o r may in t he future b e a dversely impacted b y Sodexh o’s racially d iscrimin atory
p o licies and practices affecting p romo tions o r advancement” (emphasis added)).

44

hearing and on finding that it is fair, reasonable, and adequate.”). As explained above,
in this district, courts presented with motions for preliminary approval of class
settlement agreements consider such fairness-related factors as “(1) whether the
proposed settlement appears to be the product of serious, informed, non-collusive
negotiations, (2) whether it falls within the range of possible judicial approval, and
(3) whether it has any obvious deficiencies, such as granting unduly preferential
treatment.” Richardson, 951 F. Supp. 2d at 106– 07 (internal quotation marks and
citation omitted). These and other related considerations, which are discussed below,
compel this Court to conclude that the terms of P laintiffs’ proposed Settlement
Agreement are sufficiently problematic that the agreement cannot be approved, even
preliminarily.
To be sure, and to the parties’ credit, this Agreement “appears to be the product
of serious, informed, non-collusive negotiations.” Id. P laintiffs report that the parties
engaged in “eleven days of in-person and numerous telephonic mediation sessions[,]”
during which they discussed “statistical analyses, company policies and practices, and
anecdotal evidence in this case.” (Mot. at 12.) Lockheed apparently “produced
electronic human resource data[,]” and P laintiffs’ counsel “retained a labor economics
expert . . . to conduct statistical analyses of [that] data.” (Id.) These facts satisfy the
Court that the parties’ negotiations were vigorous and well-informed. See Richardson,
951 F. Supp. 2d at 106– 07. But that alone is not enough to establish that the resulting
agreement is f air to the absentee class members. And the terms of the Settlement
Agreement that bear on the other fairness considerations mentioned above plainly

45

indicate that the negotiating parties have given the absentee class members’ interests
short shrift in several respects.
First, and most notably, there is a gross imbalance between the claims that are
actually at issue in this case and the claims that the class members who participate in
the settlement would be required to release. See M anual f or Complex Litigation § 21.62
(explaining that the “[r]easonableness [of a settlement agreement] depends on an
analysis of the class allegations and claims and the responsiveness of the settlement to
those claims”). P laintiffs have consistently and unabashedly maintained that “the
primary focus of the Complaint is on flaws in the performance appraisal system”
(Suppl. Mem. at 13); yet, they also candidly acknowledge that, somehow, class
members are required to “release Lockheed Martin from a ll types of racial
discrimination claims” under the terms of the Settlement Agreement (id. at 12
(emphasis added) (citing Settlement Agreement at 22–23)). Plaintiffs have provided no
reasonable explanation for the Settlement Agreement’s mandate that a class member
relinquish her right to bring race discrimination claims that have nothing to do with the
performance review system in order to settle a case about allegedly discriminatory
performance reviews—and this Court sees none. And even more puzzlingly, the
Agreement’s proposed broad release of claims sweeps even further: although the class
period ended in 2016, class members agree release all race discrimination claims th a t
a rise a t a n y time up until the Court finally approves the Settlement Agreement (see
Settlement Agreement at 22), and not only do class members agree to release all race
discrimination claims regarding the terms and conditions of their employment, they also

46

assent to give up “claims of alleged racial discrimination . . . in th e p ro vision o f
emp lo yee b enefits” as well (id. (emphasis added)).
It is perhaps not surprising that Lockheed would assent to such an agreement—in
one fell swoop, this proposed settlement forecloses any and all potential race
discrimination claims that thousands of African-American employees might otherwise
have brought against the company. But it is shocking to this Court that counsel for the
putative class members would contend that a release this broad and consequential is a
“fair” bargain as it relates to the absent individuals whose potential legal claims are
effectively extinguished by it. The sheer breadth of the released claims alone, when
compared to the scope of the claims in the complaint, is enough to cause this Court to
have serious doubts about whether the proposed Settlement Agreement comes anywhere
close to “fall[ing] within the range of possible judicial approval[.]” Richardson, 951 F.
Supp. 2d at 107.
The opt-out procedure laid out in the Settlement Agreement provides another
source of serious concern. Under the Agreement, class members who do not respond to
the class notice in any manner—that is, who neither complete and submit the extensive
claim form nor comply with the prescribed opt-out procedures—will not only release all
of their race discrimination claims against Lockheed Martin but will also become
ineligible to recover any compensation from the settlement fund. (See Settlement
Agreement at 24.) That is exactly the sort of circumstance that raises legitimate
questions about the fairness of a settlement agreement. See 4 Newberg on Class Actions
§ 13:60 (noting that “[c]ompromising claims without compensation” is a “red flag in
proposed settlements”). Additionally, if P laintiffs’ counsel is correct that it is likely

47

that thousands of class members will not even respond to the class notice, much less
affirmatively opt out (see Tr. of Oral Arg. at 33– 34 (counsel’s assertion, based on past
experience, that the response rate is likely to be 30– 50% ), then this Settlement
Agreement effectively allows Lockheed to inoculate itself against any and all race
discrimination and race-related benefits claims by a huge swath of its African-American
employees for a price that hardly seems “adequate.”
This Court is also quite concerned about the miniscule amount of information
that an absent class member will have about (1) the legal claims that are implicated in
the settlement process, and (2) his or her anticipated recovery, at the time the opt-out
decision is made. See M anual f or Co mp lex Litig ation § 21 (listing “the fairness and
reasonableness of the procedure for processing individual claims under the settlement”
as a relevant factor under Rule 23(e)). That is, for all its detail, the class notice
provides no sense of what the minimum recovery is likely to be; no sense of what the
average recovery is likely to be; and no sense of how giving particular answers on the
claim form will likely influence the amount of a class member’s recovery. P laintiffs’
counsel explains that none of this information is discernable until the Claims
Administrator reviews all of the claim forms and develops information regarding the
number and type of race discrimination claims that may exist throughout the class. (See
Tr. of Oral Arg. at 69.) But that is exactly the defect that makes this settlement so
potentially damaging to the absent class members—if they had been participants in this
litigation from the start, P laintiffs’ counsel would have already investigated their
discrimination allegations and would be in a position to provide them with critical
information about the estimated value of the claims that the Settlement Agreement

48

releases. As things currently stand, this Court has no idea how P laintiffs’ counsel could
possibly have determined that the amount that Lockheed is offering to settle this case is
sufficient to redress the legal claims of the class members without having already
gathered information regarding the universe of claims at issue. And of course, if
P laintiffs’ counsel does have some knowledge of, and projections about, the scope of
the absentee class members’ discrimination claims, then the value of those claims can
be estimated, at least in general terms, and this recovery-related information should be
provided to the class as part of the notice. 13 What is more, it is clear to this Court that
no reasonable counsel advising an individual absent class member would encourage his
client to agree ex a n te to release legal claims that the lawyer has not yet evaluated in
exchange for nothing other than the vague hope that a claims official might
subsequently do the work of assessing the strength of his client’s case and paying a
worthy settlement amount. And yet, that appears to be precisely what P laintiffs’
counsel proposes for the absent class members in the context of the settlement at issue
in this case.
The bottom line is this: a putative class member needs to be able to make an
informed determination regarding whether or not to opt out of a class settlement, and
this is especially so when the proposed agreement requires him or her to give up
potentially viable legal claims in exchange for the promise of a monetary recovery.

13

Th is is t o say, Plaintiffs’ counsel cannot h ave it b oth ways. Eit her (1) h e h as n o clue about what t he
class memb ers will rep ort o n t he claim fo rm regarding t heir exp eriences with race d iscrimin ation, in
wh ich case n o in dividual recovery estimates can b e provided, but h e also cannot p ossibly have made a
rat io nal d etermination t hat t he $22.8 s ettlement fund is adequate t o cover all o f t he class memb ers’
claims , o r (2) he is sufficiently aware of t he scope and n ature of t he class memb ers’ p otential claims
t h at h e can assert with confidence t hat t he settlement fund amount is adequate, but if t his is so, then h e
s h ould b e able t o provide t he class memb ers with in formation about the ant icipated recovery if t hey
d ecide t o p articipate in t he set tlement.

49

Ordinarily, and ideally, when faced with such a choice, the class member would want to
know how much the claims he is releasing are likely to be worth, so that this figure can
be compared to the amount being offered in settlement of those claims. In the typical
case involving a settlement class, this math is easily done, because the claims in the
complaint are all that is at issue, and class counsel presents a settlement agreement with
a negotiated amount that relates in some discernable way to the legal claims that will be
released. S ee, e.g., In re APA Assessment Fee Litig ., 311 F.R.D. 8, 13, 19 (D.D.C.
2015) (approving settlement agreement that “represent[ed] close[] to 13% of plaintiffs’
best possible recovery” and under which “Claimants w[ould] receive a pro rata share of
the fund based on the amount of . . . fees they paid during the class period”); Alva rez,
303 F.R.D. at 157 (approving settlement agreement that “provide[d] for a monetary
payment to each of the[] class members consisting of 1.37 times the amount owed for
overtime during the relevant period”). But even if the overall value of the released
claims is uncertain, at a minimum, any informed opt-out determination requires
information regarding the estimated amount that a class member would be expected to
receive if she participates, and it is no answer to say, as P laintiffs do here, that ‘an
estimate is not possible because we do not yet know the scope of the claims being
released.’ (See, e.g., Tr. of Oral Arg. at 69.) Far from curing the problem, the fact that
P laintiffs themselves purport to have not yet fully evaluated the claims that class
members are relinquishing in this action underscores the fundamental fairness issues
described above, and suggests that, rather than doing what is necessary to ensure that
all class members’ rights are protected and advanced, P laintiffs have opted to bind their

50

fellow class members indiscriminately and rely on a claims administrator to do the
heavy lifting of sorting it all out on the back end.
In light of all of these deficiencies, which this Court has considered individually
and in the aggregate, the Court is compelled to conclude that P laintiffs have failed to
demonstrate that the proposed Settlement Agreement is sufficiently “fair, reasonable,
and adequate” to satisfy the standard for preliminary approval.
IV.

CONCLUSION
In this lawsuit, P laintiffs seek to certify a class for settlement purposes that, in

effect, is comprised of virtually every salaried African-American employee of Lockheed
Martin. P laintiffs say that the 5,500-plus members of their proposed class have
suffered race discrimination as a result of Lockheed’s performance review process, and
have been victimized in a manner that is susceptible to common proof, but P laintiffs do
not present any theory of how the performance appraisal system resulted in racially
disparate outcomes, much less evidence that the challenged system discriminates
against all class members in the same way, as the commonality element of Federal Rule
of Civil P rocedure 23(a) requires. Moreover, the broad release of legal claims in the
proposed Settlement Agreement strongly suggests that the parties have endeavored to
use this case to resolve the entire universe of race discrimination claims that these class
members might have against Lockheed, via a lawsuit that is purportedly much more
limited in scope. This imbalance creates serious fairness concerns with the breadth of
the release, the adequacy of the settlement fund, the prescribed opt-out procedures, and
the overall claims administration process. All this leads this Court to the firm but
reluctant conclusion that this case cannot be preliminarily certified as a class action

51

under Rule 23, and that the parties’ proposed Settlement Agreement is so potentially
unfair that it cannot be preliminarily approved. Accordingly, as set forth in the
accompanying Order, P laintiffs’ motion for preliminary certification and approval will
be DENIED.

DATE: July 28, 2017

Ketanji Brown Jackson
KETANJI BROWN JACKSON
United States District Judge

52