Stockwell et al v. City and County of San Francisco et al

Filing 117

ORDER by Judge Hamilton Denying 107 Renewed Motion for Class Certification (pjhlc2, COURT STAFF) (Filed on 10/11/2011)

1
2
3

UNITED STATES DISTRICT COURT

4

NORTHERN DISTRICT OF CALIFORNIA

5
6
JUANITA STOCKWELL, et al.,
7
Plaintiffs,

No. C 08-5180 PJH

8
v.

ORDER DENYING RENEWED
MOTION FOR CLASS CERTIFICATION

9
CITY AND COUNTY OF SAN
FRANCISCO,

11
For the Northern District of California

United States District Court

10

Defendant.
_______________________________/

12
13

Plaintiffs’ renewed motion for class certification came on for hearing before this court

14

on July 20, 2011. The parties appeared through their counsel. Having reviewed the

15

papers and considered the argument of counsel and the relevant legal authority, the court

16

DENIES plaintiffs’ renewed motion for class certification for the reasons stated at the

17

hearing and as set forth below.

18
19

RELEVANT FACTS
In the order denying plaintiffs’ first motion for class certification, the court detailed

20

the factual background of this matter and set forth the elements required for class

21

certification. On plaintiffs’ renewed motion for class certification, the court focuses on

22

plaintiffs’ current arguments to establish commonality in light of the Supreme Court’s

23

decision in Wal-Mart v. Dukes, — U.S. —, 131 S.Ct. 2541 (2011). The court notes that

24

although Dukes was decided on June 20, 2011, after plaintiffs filed notice of the instant

25

motion, and only two days before defendant filed its opposition, both parties addressed

26

Dukes in their papers in support of or in opposition to the present motion.

27
28

The Second Amended Class Action Complaint (SAC) alleges:

1

From 2007 through 2009, the CITY AND COUNTY OF SAN
FRANCISCO (“CCSF”), through its San Francisco Police Department
(“SFPD”) ignored an active Q-35 Inspector Eligibility List and instead
exclusively utilized a Q-50 Sergeant Eligibility List to award
promotions to approximately fifty-five (55) investigative positions
which had traditionally been held by Q-35 Assistant Inspectors.
Plaintiffs and the members of the putative class were highly qualified
but were not even considered for these promotions. The officers on
the Q-50 Sergeant List were on average much younger and
less-qualified than the Plaintiffs and putative class members all of
whom were eligible for promotion from the Q-35 Inspector List.

2
3
4
5
6
7
8
9

SAC ¶ 2. The plaintiffs who seek to be named class representatives (Juanita Stockwell,

11
For the Northern District of California

United States District Court

10

Terrye Ivy, Jacklyn Jehl, Mike Lewis and Vince Neeson) were all on the 1998 Q-35

12

Eligibility List for promotion. The remaining named plaintiffs were also placed on the Q-35

13

eligibility list.
The Q-35 list at issue was based on an assistant inspector exam administered in

14
15

April and May 1998. Doc. no. 64 (Rolnick Decl. I), Ex. D (October 1998 stipulation re

16

selection procedure for Q-35 list). February 2006 was last time the City promoted

17

individuals from the Q-35 assistant inspector list. Doc. no. 64, Ex. S ¶¶ 7-8 (Heather Fong

18

Declaration). In 2005 and 2006, Police Chief Fong announced a new staffing plan which

19

indicated that SFPD would no longer promote officers from the 1998 Q-35 list or administer

20

a new Q-35 exam, and announced a Q-50 Sergeant promotional examination. Id. ¶ 12-17.

21

In February 2007, the City adopted a Q-50 eligible list based on the results of the 2006

22

exam. Id. ¶ 19.
Plaintiffs were eligible to take the Q-50 promotional exam; plaintiffs Stockwell and

23
24

Jehl took the exam and plaintiffs Ivy, Neeson and Lewis did not. Doc. no. 60 (Villagomez

25

Decl. I) ¶ 3 and Ex. A; doc. no. 64 (Rolnick Decl. I), Exs. I-M (plaintiffs’ responses to RFAs).

26

In August 2007, Chief Fong appointed 51 sergeants from the Q-50 eligible list. Doc. no. 60

27

¶ 3.

28

Plaintiffs purport to represent the following class:
2

1

Class of police officers aged forty years and older and were
employed by CCSF at any time since October 18, 2006, who were on
the Q-35 Assistant Inspector Eligibility List, and could have been
selected for a promotion from that list had the SFPD properly
considered them for promotions to investigative positions from 2007
through 2009. Plaintiffs and members of the Class they seek to
represent were, on the basis of their age, denied promotional
opportunities and were thereby substantially adversely impacted in
their rank and compensation.

2
3
4
5
6
7
8

SAC ¶ 12.
Plaintiffs previously moved for class certification with respect to their FEHA claim

9

only and the court denied that motion by order dated August 27, 2010. In the order

11
For the Northern District of California

United States District Court

10

denying the motion for class certification, the court found that plaintiffs had satisfied the

12

FRCP 23(a) requirements of numerosity, typicality and adequacy, but that they failed to

13

establish commonality. The court further found that under Rule 23(b)(3), plaintiffs had

14

failed to provide the Court with any foundation to undertake a predominance analysis, and

15

further failed to establish that class action treatment was superior to individual adjudication.

16

Subsequently, plaintiffs amended their complaint to dismiss all claims under a

17

disparate treatment theory and to dismiss all plaintiffs whose rank on the Q-35 list was

18

below the cut-off point of their previously proposed class. Plaintiffs allege age

19

discrimination claims under the Age Discrimination in Employment Act 29 U.S.C § 621 and

20

the California Fair Employment and Housing Act, Cal. Gov’t Code § 12900 et seq. Plaintiffs

21

now seek class certification under Rule 23(b)(3).

22

I.

23

LEGAL STANDARD
Federal Rule of Civil Procedure 23(a) provides four requirements for class

24

certification: (1) numerosity (“the class is so numerous that joinder of all members is

25

impracticable”); (2) commonality (“there are questions of law or fact common to the class”);

26

(3) typicality (“the claims or defenses of the representative parties are typical of the claims

27

or defenses of the class”); and (4) adequacy of representation (“the representative parties

28

will fairly and adequately protect the interests of the class”). FRCP 23(a)(1)-(4). In addition,
3

1

the court must find that one of the requirements of Rule 23(b)(1), (b)(2), or (b)(3) is

2

satisfied. Wal-Mart Stores, Inc. v. Dukes, –– U.S. ––––, 131 S.Ct. 2541, 2548 (2011).

3

Rule 23(b)(3) requires a finding by the court “that questions of law or fact common to class

4

members predominate over any questions affecting only individual members, and that a

5

class action is superior to other available methods for fairly and efficiently adjudicating the

6

controversy.” FRCP 23(b)(3). Courts refer to the requirements of Rule 23(b)(3) as its

7

“predominance” and “superiority” requirements. Amchem Prods., Inc. v. Windsor, 521 U.S.

8

591, 615 (1997).

certification must affirmatively demonstrate his compliance with the rule - that is, he must

11
For the Northern District of California

“Rule 23 does not set forth a mere pleading standard. A party seeking class

10

United States District Court

9

be prepared to prove that there are in fact sufficiently numerous parties, common questions

12

of law or fact, etc.” Dukes, 131 S.Ct. at 2551 (emphasis in original). This requires a

13

“rigorous analysis” which frequently “will entail some overlap with the merits of the plaintiff's

14

underlying claim.” Id.

15

Dukes states the commonality requirement as follows: “Commonality requires the

16

plaintiff to demonstrate that the class members ‘have suffered the same injury.’” Dukes,

17

131 S.Ct. at 2551 (quoting Gen. Tel. Co. of the Southwest v. Falcon, 457 U.S. 147, 157

18

(1982)). The class members' “claims must depend on a common contention,” and that

19

common contention must be “of such a nature that it is capable of classwide resolution -

20

which means that determination of its truth or falsity will resolve an issue that is central to

21

the validity of each one of the claims in one stroke.” Id.

22

II.

23

COMMONALITY
Dukes clarified that “proof of commonality necessarily overlaps with respondents'

24

merits contention that Wal–Mart engages in a pattern or practice of discrimination. That is

25

so because, in resolving an individual's Title VII claim, the crux of the inquiry is ‘the reason

26

for a particular employment decision.’” Dukes, 131 S.Ct. at 2552 (quoting Cooper v.

27

Federal Reserve Bank of Richmond, 467 U.S. 867, 876 (1984)) (footnote omitted).

28

Plaintiffs’ renewed motion fails to cure the deficiencies of its original motion for class
4

1

certification and fails to meet the standard for commonality articulated in Dukes.

2

Plaintiffs allege that the following questions are common to the whole class:

3

(1)

whether SFPD’s policies or practices with respect to promotion of qualified

4

applicants for investigative positions from 2007 through 2009 had a

5

discriminatory impact on officers over forty who were otherwise qualified for

6

those positions;

7

(2)

8
9

policies and practices to prevent discrimination against older employees;
(3)

For the Northern District of California

United States District Court

10
11

14

whether SFPD’s policies and practices violated the Age Discrimination ADEA
and/or the California FEHA; and

(4)

12
13

whether SFPD’s Human Resources Department has failed to implement

whether damages and declaratory relief for the Class members are
warranted.

SAC ¶ 14.
In their initial motion for class certification, plaintiffs offered the following evidence of

15

commonality: (1) the statistical evidence prepared by Dr. Harry, and (2) excerpts from the

16

deposition testimony of Assistant Chief Tabak, to the effect that Q-35 eligible listed officers

17

were qualified to do the work of those assigned “to the officers in 2007-09;" that sergeants

18

performed the same work traditionally done by Assistant Inspectors; and that a rank on a

19

Q-35 exam was a good indication of an officer’s ability to become an effective inspector.

20

In the order denying class certification, the court determined that Dr. Harry’s

21

statistical evidence was not sufficient to show disparate impact, pointing out deficiencies in

22

the statistical evidence: failure to identify a well-accepted analytical methodology, failure to

23

identify the relevant comparison pools, and failure of Mr. Harry’s analysis to adequately

24

account for and explain away any alternative variables that could interfere with any

25

inference of discrimination. Doc. no. 75 at 13.

26

Plaintiffs’ renewed motion for class certification offers a revised expert report that

27

purports to eliminate the issues concerning methodology and comparison pools that the

28

court raised in denying the earlier motion for class certification. Putting aside the parties’
5

and the experts’ competing analyses, the fundamental problem with plaintiffs’ reliance on

3

Harry’s statistical evidence is that it does not show that plaintiffs were denied appointment

4

to investigative positions because of their age. Even without the benefit of the expert

5

analysis, it would be reasonable to expect that the City’s policy of hiring from the Q-50 list,

6

rather than the Q-35 list, would result in impacting applicants over the age of forty because

7

the Q-35 list contained applicants who had taken the qualifying test 8 years before the Q-50

8

test was administered, and the Q-35 applicants who had taken the test 8 years before were

9

likely to have reached the age of 40 by the time the Q-50 test was administered; compared

10

to the Q-35 list, which was composed almost entirely of over-40 aged applicants, the Q-50

11
For the Northern District of California

arguments detailing the differences between Harry’s earlier report and the revised report

2

United States District Court

1

list would necessarily cover a wider age-range which includes younger applicants who

12

would not have taken the Q-35 qualifying test 8 years earlier. Thus, the most that the

13

statistical evidence shows is that over-40 applicants were impacted when the City switched

14

to the Q-50 list, but the evidence does not establish that this correlation was “on account

15

of” or because of plaintiffs’ membership in a protected class. See Dukes, 131 S.Ct. at

16

2553.

17

In Dukes, the Supreme Court recognized two ways to bridge the gap between an

18

individual’s claim of bias and a classwide claim that shares common questions of law or

19

fact: First, if the employer “used a biased testing procedure to evaluate both applicants

20

for employment and incumbent employees, a class action on behalf of every applicant or

21

employee who might have been prejudiced by the test clearly would satisfy the

22

commonality and typicality requirements of Rule 23(a).” Second, “[s]ignificant proof that

23

an employer operated under a general policy of discrimination conceivably could justify a

24

class of both applicants and employees if the discrimination manifested itself in hiring and

25

promotion practices in the same general fashion, such as through entirely subjective

26

decisionmaking processes.” 131 S.Ct. at 2553 (quoting Falcon, 457 U.S. at 159)

27

(emphasis added).

28

Here, plaintiffs do not allege that the Q-50 testing procedure was biased under the
6

1

first Dukes/Falcon test, or even that they were denied the opportunity to apply for

2

investigative positions through the Q-50 list. Rather, they rely on statistical evidence to

3

establish under the second Dukes/Falcon test that the City “operated under a general

4

policy of discrimination.” Here, as in Dukes, that significant proof is “entirely absent.” 131

5

S.Ct. at 2553.

6

Assuming that the statistical evidence shows that the decision in 2007 to appoint

applicants, the evidence in the record demonstrates that such impact was directly caused

9

by the fact that most applicants on the Q-35 list had taken the qualifying exam 10 years

10

earlier and almost all had reached the age of 40 by 2007. Plaintiffs offer no evidence to

11
For the Northern District of California

investigators from the Q-50 list rather than Q-35 list had a disparate impact on over-40

8

United States District Court

7

show that this disparate impact was caused by a discriminatory policy. See Dukes, “Even if

12

they are taken at face value, these studies are insufficient to establish that respondents'

13

theory can be proved on a classwide basis.” 131 S.Ct. at 2555. The Supreme Court

14

further reasoned:

15
16
17
18
19
20
21
22

There is another, more fundamental, respect in which respondents'
statistical proof fails. Even if it established (as it does not) a pay or
promotion pattern that differs from the nationwide figures or the
regional figures in all of Wal–Mart's 3,400 stores, that would still not
demonstrate that commonality of issue exists. Some managers will
claim that the availability of women, or qualified women, or interested
women, in their stores' area does not mirror the national or regional
statistics. And almost all of them will claim to have been applying
some sex-neutral, performance-based criteria—whose nature and
effects will differ from store to store.

23
Id. Here, even assuming that the statistical evidence shows that the abandonment of the
24
Q-35 list had disparate impact on over-40 applicants, there is no commonality of issues
25
because plaintiffs do not show that all the members of the Q-35 list had sought to be
26
qualified for the Q-50 list and were denied, or that all Q-35 list members had taken the Q27
50 qualifying exam and were appointed to fewer investigative positions than under-40
28
7

1
2

applicants on the Q-50 list.
As defendant argues, “The fact that most of the remaining eligibles on the 1998 Q-

3

35 list are over 40 is a product of the age of the eligible list and does not establish a

4

disparate impact claim.” Doc. no. 113 at 26 (citing Katz v. Regents, 229 F.3d 831, 836 (9th

5

Cir. 2000)). In Katz, plaintiffs were participants in the Public Employees Retirement System

6

(PERS), a statewide retirement system administered by the State of California. Beginning

7

in 1961, the University of California Retirement Plan (UCRP) was offered to new

8

employees, and no other lab employees became members of PERS.” As part of a

9

downsizing effort, the university offered an early retirement incentive program to UCRP
members, but not to PERS members. To be eligible for the early retirement program, an

11
For the Northern District of California

United States District Court

10

employee had to be at least 50 years old with at least five years of service. As one might

12

expect, the evidence showed the average age of the PERS members was greater than that

13

of the UCRP members. The evidence also established the average age of potentially

14

eligible PERS members was 60.51 years. The allegedly disadvantaged PERS members

15

brought suit for age discrimination under both the Age Discrimination in Employment Act

16

(29 U.S.C. § 621 et seq.) and FEHA, arguing the early retirement plan had a disparate

17

adverse impact upon PERS members. The Ninth Circuit determined that the Katz plaintiffs

18

failed to show a causal link for their disparate impact claim:

19
20
21
22
23
24
25
26
27
28

While the plaintiffs can certainly show that they were treated
differently on the basis of their membership in the PERS program,
their evidence demonstrates that only 238 of the 895 employees at
the laboratories age 60 or over (roughly 27 percent) were adversely
impacted by the University's decision. Given the legitimate reason
advanced for the University's decision, the plaintiff's statistical
evidence is insufficient to raise an inference that the disparate impact
fell upon employees by virtue of their membership in a protected age
group. See Rose v. Wells Fargo & Co., 902 F.2d 1417, 1424 (9th
Cir.1990) (concluding that plaintiff's statistical evidence was
insufficient to raise inference of causation where other evidence
demonstrated that workforce after employment action was applied
was older than the workforce beforehand). As in Rose, the statistical
8

1
2
3
4
5
6
7
8

evidence of disparate impact upon employees age 60 or over is
minimal compared to the number of employees of that age group not
adversely affected by the University's decision.
...
At best, plaintiffs have demonstrated only that the average age of the
PERS members was 5 years older than the average age of eligible
UCRP members, and plaintiffs concede that numerous UCRP
employees older than the average PERS member were offered
VERIP III. The factor that determines an employees' eligibility to
participate in VERIP III is thus not an employee's age itself, but
instead whether that employee is a member of UCRP.

9

11
For the Northern District of California

United States District Court

10

229 F.3d at 836 (affirming dismissal prior to trial of disparate impact claim).
Plaintiffs argue that their statistical showing is much stronger here than in Katz, but

12

do not address the fundamental causation problem. Doc. no. 13. What the statistical

13

evidence shows is that there’s a correlation between plaintiffs’ age and the age of the Q-35

14

list, but it does not demonstrate that there’s a common issue of age discrimination to

15

explain why the plaintiffs did not get promoted to an investigative position. This issue of

16

causation is not a matter to be deferred to the merits phase because, as the Supreme

17

Court has recognized, the issue of commonality involves a “rigorous analysis” which

18

frequently “will entail some overlap with the merits of the plaintiff's underlying claim.”

19

Dukes, 131 S.Ct. at 2551-52.

20

Turning to plaintiffs’ statistical evidence and their expert opinions, Dr. Harry

21

contends that his analysis “illustrates that CCSF’s abandonment of the Q-35 list after 2006

22

and use of the Q-50 list from 2007-2009 caused a drastic decrease in the selection rate of

23

older qualified officers to investigative positions from a rate of 1.37 times down to only .69

24

times the selection rate of younger reachable officers. The finding is consistent whether

25

the selection rates are computed using aggregate or weighted data.” Doc. no. 107-2 at 16

26

and Chart #6 (infra). As the court has discussed earlier, this finding is credible because

27

abandonment of the Q-35 list necessarily affected over-40 individuals who almost

28

exclusively comprised the Q-35 list by 2007. The inference of causation, however, is
9

1

lacking. If anything, the data shows that fewer over-40 officers were “reachable” under the

2

Q-50 qualifying examination in 2006 than under the Q-35 examination in 1998 (728 vs.

3

189).

4

CHART #6

5

Age <40

Age 40+

Total

6

Q-35 List 2000-06

7

Total appointees

29

89

118

8

Aggregate reachable individuals

328

728

1,056

9

Appointment rate - raw data

8.84%

12.23%

Appointment rate - weighted

18.83%

25.72%

11
For the Northern District of California

United States District Court

10

12
13

by number appointed by appt
date
Selection rate - raw

1.38

Selection rate - weighted

1.37

14
15

Q-50 List 2007-09

16

Total investigative appointees

39

16

55

17

Aggregate reachable individuals

214

189

403

18

Appointment rate - raw data

18.22%

8.47%

Appointment rate - weighted

51.55%

35.34%

19
20
21
22

by number appointed by appt
date
Selection rate - raw

0.46

Selection rate - weighted

0.69

23
24
25
26
27
28

Doc. no. 107-2 (Harry Revised Report) at 16, Chart #6.
The question of how many over-40 individuals were reachable on each list seems to
be disputed. But even assuming that Harry’s estimate of the pool of reachable individuals
is correct, the difference in reachable over-40 individuals is likely to be directly correlated to

10

1

how many over-40 individuals chose to take the Q-50 exam in 2006. What the statistics fail

2

to show is whether the job-related testing procedure unfairly favored those applicants who

3

were not part of the protected class, or otherwise demonstrates that the disparate impact

4

was due to membership in a protected class. See Griggs v. Duke Power Co, 401 U.S. 424

5

(1971). If plaintiffs were asserting that theory, then an analysis comparing test scores and

6

the age of the test-takers would be relevant. But here, Harry’s report only demonstrates an

7

undisputed fact, that the abandonment of the Q-35 list impacted over-40 individuals

8

because the Q-35 list was eight years old and most of the individuals on the Q-35 list had

9

reached age 40 by that time.
In light of Dukes, statistical evidence of disproportionate impact, standing alone, is

11
For the Northern District of California

United States District Court

10

insufficient to establish that plaintiffs’ age discrimination claims can be proved on a

12

classwide basis. 131 S.Ct. at 2555-56. To establish their age discrimination claims,

13

plaintiffs must show causation: that the challenged employment policy or practice, while

14

facially neutral, has a disparate impact on certain employees “because of their membership

15

in a protected group.” Katz, 229 F.3d at 829. The City’s abandonment of the Q-35 list

16

admittedly affected over-40 individuals because they all took the qualifying test in 1998

17

whereas the Q-50 examination was administered 8 years later. Plaintiffs have not shown

18

that the impact of the policy was directly related to an employee's age because over-40

19

officers were not prohibited from taking the Q-50 examination in 2006.

20

Defendant also challenges Dr. Harry’s methodology on several grounds, including

21

Dr. Harry’s underlying factual assumptions that are not supported by facts in the record, Dr.

22

Harry’s use of the Chi-square method of analysis, and Dr. Harry’s failure to take into

23

account other factors that might demonstrate a nondiscriminatory effect of the employer’s

24

practices. Doc. no. 113 at 28-29. On the present motion, the court addresses defendant’s

25

third challenge to Dr. Harry’s analysis, which was previously raised by the court, for failure

26

“to adequately account for and explain away alternative variables that could interfere with

27

any inference of discrimination.” Doc. no. 75 at 13. Plaintiffs contend that regression

28

analyses are not applicable here because, unlike Dukes, defendant fails to identify specific
11

1

factors that might weaken or eliminate the showing of age discrimination. Doc. no. 107 at

2

20; doc. no. 114 at 12. See doc. no. 107-2 at 24-28 (Harry revised report explaining why

3

regression analysis is not warranted in this case). However, the court specifically identified

4

independent variables that would require regression analyses, such as examination scores,

5

job performance, seniority, experience and/or education. Doc. no. 75 at 13. The court

6

finds that plaintiffs’ failure to address this deficiency, in addition to the lack of causation and

7

other factors identified above which fail to establish commonality, supports denial of the

8

renewed motion for class certification.

requirement pursuant to FRCP 23(a), the court need not reach the question whether

11
For the Northern District of California

Having determined that plaintiffs have failed to demonstrate the commonality

10

United States District Court

9

plaintiffs have satisfied the requirements of FRCP 23(b)(3).

12
13
14

CONCLUSION
For the reasons set forth above, plaintiffs’ renewed motion for class certification is
DENIED.

15
16

IT IS SO ORDERED.

17

Dated: October 11, 2011
______________________________
PHYLLIS J. HAMILTON
United States District Judge

18
19
20
21
22
23
24
25
26
27
28
12