Equal Employment Opportunity Commission v. Mavis Discount Tire, Inc. et al

Filing 110

OPINION AND ORDER re: 94 MOTION for Summary Judgment on Liability and Record-keeping violation and Seeking Bifurcated Trial filed by Equal Employment Opportunity Commission: For the foregoing reasons, the EEOC's motion for summary judgment on its pattern-or-practice claim and its record-keeping claim is DENIED. The EEOC's motion for a bifurcated trial is GRANTED IN PART and DENIED IN PART. Trial shall proceed in two phases, with punitive damages to be awarded ( if at all) during the second phase. The parties shall appear for a pretrial conference on October 2, 2015, at 11:00 a.m. in Courtroom 618 of the Thurgood Marshall Courthouse, 40 Foley Square, New York, New York. By Thursday of the week prior to th at conference, the parties shall file a joint letter, not to exceed three pages, raising any issues the parties would like to address at the pretrial conference and providing any information that the parties believe may assist the Court in advancing the case to settlement or trial. (Signed by Judge Katherine Polk Failla on 9/11/2015) (tn)

UNITED STATES DISTRICT COURT
SOUTHERN DISTRICT OF NEW YORK
------------------------------------------------------X
:
EQUAL EMPLOYMENT OPPORTUNITY
:
COMMISSION,
:
:
Plaintiff,
:
:
v.
:
:
MAVIS DISCOUNT TIRE, INC., et al.,
:
:
Defendants. :
:
----------------------------------------------------- X

USDC SDNY
DOCUMENT
ELECTRONICALLY FILED
DOC #: _________________
DATE FILED: September 11, 2015
______________

12 Civ. 741 (KPF)
OPINION AND ORDER

KATHERINE POLK FAILLA, District Judge:
Plaintiff Equal Employment Opportunity Commission (the “EEOC”),
initiated this action in January 2012, alleging that Defendants Mavis Discount
Tire, Inc., d/b/a Mavis Tire Supply Corp., Mavis Tire NY, Inc., and Cole
Muffler, Inc. (collectively, “Defendants” or “Mavis”) engaged in a pattern or
practice of sex discrimination in violation of Title VII of the Civil Rights Act of
1964, 42 U.S.C. §§ 2000e to 2000e-17. The EEOC alleges that Defendants
discriminated against the charging party, Nicole Haywood, and other similarly
qualified female applicants during the 2008 to 2012 time period, in favor of
hiring less qualified men for positions in their branch stores. The EEOC also
alleges that Defendants failed to comply with the record-keeping requirements
of Section 709(c) of Title VII, 42 U.S.C. § 2000e-8(c). The EEOC has now moved
for summary judgment on its pattern-or-practice and record-keeping claims,
and separately has moved to bifurcate the trial in this action in the event
summary judgment is not granted. For the reasons set forth below, the

EEOC’s motion for summary judgment is denied in its entirety, and the EEOC’s
motion for a bifurcated trial is granted in part and denied in part.
BACKGROUND1
A.

Factual Background
1.

Mavis’s Business

Mavis is a family-owned, independent tire dealer. (Def. 56.1 ¶¶ 207, 211).
In addition to tire sales, Mavis store locations also offer automotive parts and
services including brakes, alignments, suspension, shocks, struts, oil changes,
battery replacement, and exhaust work. (Id. at ¶ 212).
In 1987, Mavis was comprised of approximately 11 store locations; it has
grown significantly since then. (Def. 56.1 ¶ 213). Most notably, in 2008, Mavis
acquired approximately 50 stores from Cole Muffler, Inc. (Id. at ¶ 214).2 Today,
Mavis operates more than 140 store locations in New York, Connecticut,
Massachusetts, and Pennsylvania. (Id.).
Stephen Andre, currently the Vice President of Operations for Mavis, first
joined the Company in 1987. (Def. 56.1 ¶ 213). Andre was ultimately
responsible for all hires at Mavis during the relevant time period, having either
made or approved every offer of employment. (Pl. 56.1 ¶ 159). But although

1

The facts stated herein are drawn from the parties’ submissions in connection with the
instant motion, including the EEOC’s 56.1 Statement (“Pl. 56.1” (Dkt. #96)), and
Defendants’ responses thereto (“Def. 56.1 Response” (Dkt. #102-03)); Defendants’ 56.1
Counterstatement (“Def. 56.1” (Dkt. #103)), and the EEOC’s responses thereto (Pl. 56.1
Response” (Dkt. #107)). References to individual deposition transcripts will be referred
to as “[Name] Dep.” For convenience, the EEOC’s opening brief will be referred to as “Pl.
Br.” (Dkt. #95); Defendants’ opposition brief as “Def. Opp.” (Dkt. #104); and the EEOC’s
reply brief as “Pl. Reply” (Dkt. #106).

2

Defendants note that not a single female employee was employed in any of the 50 Cole
stores acquired by Mavis. (Def. 56.1 ¶ 215).

2

Andre retained the final authority for hiring, he did not make Mavis’s personnel
decisions alone. In 2008, in response to company growth, Mavis created the
position of Regional Training Manager (“RTM”). (Id. at ¶ 217). RTMs are
responsible for overseeing between eight and fifteen stores that fall within their
assigned regions. (Id. at ¶ 218; Pl. 56.1 Response ¶ 218). For each store within
a particular region, an RTM controls the payroll, trains the store managers,
ensures profitability, performs human resources functions, and recruits
qualified applicants. (Id. at ¶ 219).
Every application submitted at a store location is given to the appropriate
RTM for review. (Def. 56.1 ¶ 233). If the RTM determines there is a need for a
particular position, the RTM determines whether to interview any of the
applicants. (Id. at ¶ 235). If the decision is made to interview, the RTM checks
the applicant’s references and confirms that the information provided in the
application is accurate. (Id. at ¶ 237). If, after conducting the interview,
checking the references, and confirming the accuracy of the information set
forth in the application, the RTM decides to offer the position to the applicant,
the RTM makes a recommendation to Andre to hire the applicant. (Id. at
¶ 238; Andre 30(b)(6) Dep. 223).3 Based on his review of the information
presented by the RTM as well as Mavis’s budget, Andre decides whether to hire

3

Andre was deposed in his individual capacity, and as Mavis’s 30(b)(6) witness. See Fed.
R. Civ. P. 30(b)(6) (“[A] party may name as the deponent a … private corporation …. The
named organization must then designate one or more … persons who consent to testify
on its behalf.”). References to Andre’s 30(b)(6) deposition will be designated accordingly.

3

the applicant. (Def. 56.1 ¶ 239).4 The Mavis personnel involved in hiring
testified that they do not actively seek to hire a certain group of applicants;
rather, Mavis seeks to hire the most qualified applicants. (Id. at ¶ 242).
There are four available positions at each Mavis store location: (i) Store
Manager; (ii) Assistant Manager; (ii) Mechanic; and (iv) Tire Installer/Alignment
Technician (“Technician”). (Def. 56.1 ¶ 216).5 From 2008 to 2012, Mavis hired
80 Store Managers, 655 Mechanics, and 1,688 Technicians. None of these
individuals was female. (Pl. 56.1 ¶¶ 46-51). During this same period of time,
Mavis hired 288 Assistant Managers, only one of whom was female. (Id.). After
the EEOC filed its lawsuit in January 2012, Mavis hired its first female Store
Manager (id. at ¶ 47), and its second female Assistant Manager (id. at ¶ 49).
2.

Unsuccessful Female Applicants
a.

Nicole Haywood

Nicole Haywood unsuccessfully applied on multiple occasions to work at
Mavis. (Pl. 56.1 ¶ 112; Def. 56.1 Response ¶ 112). At the time of her initial
application in 2008, Haywood had approximately 15 years of experience
working at Sears Automotive as a customer service representative; for
approximately nine of those years, she performed managerial duties, including

4

For applications submitted at store locations, Andre considers only those applicants
who have been recommended for hire by the RTM (Def. 56.1 ¶ 240). For applications
submitted by email, Andre makes the initial determination as to whether the applicant
is worth pursuing. (Andre 30(b)(6) Dep. 84). He then forwards the application to RTMs
who may need to fill positions in the applicant’s area of expertise. (See Def. 56.1
Response ¶ 188).

5

Tire Installer and Alignment Technician are technically separate positions at Mavis, but,
for the purposes of the parties’ statistical analyses and the instant motion for summary
judgment, they are treated as one.

4

scheduling technicians, delegating their work, handling customer complaints,
and receiving shipments. (Pl. 56.1 ¶¶ 113-14). In 2008, Mavis hired three
male Sears Automotive employees. (See id. at ¶¶ 115-22). Significantly, all
three had been trained by Haywood while at Sears Automotive. (See Haywood
Dep. 33, 81, 132-33). After she learned of these hires, Haywood attempted to
renew her application to Mavis at various stores and through various channels.
(See Pl. 56.1 ¶¶ 116-22).
Eventually, Haywood was interviewed by Andre for 15 to 30 minutes. (Pl.
56.1 ¶¶ 123-24; Def. 56.1 Response ¶ 124). According to Andre, the interview
went poorly. (See generally Andre Dep. 73-76). He testified that Haywood
arrived 15 minutes late for the interview; displayed a poor attitude; and
disparaged her then-current employer. (Def. 56.1 Response ¶ 124). Following
the interview, Andre contacted Jim Dubois, a Store Manager for Mavis who had
previously supervised Haywood at Sears Automotive, to ask whether Dubois
would recommend offering Haywood a position with Mavis. (Id.). Based on
both his observations during the interview and Dubois’s comments, Andre
decided not to extend an offer of employment to Haywood. (Id.).
b.

Krystal Murnane

Krystal Murnane, who at the time of her 2009 application to Mavis had
approximately six years’ experience performing basic maintenance on cars, was
told by Mavis employees at its Bayshore, New York location that she would not
be hired by Mavis because she was a woman. (See Murnane Dep. 79 (testifying
that a Mavis employee said: “[D]on’t get your hopes up, you’re a chic[k]. These

5

guys are old school.”); id. at 80 (testifying that a Mavis employee told her, “a
girl can’t do a guy’s job”); id. (“[T]he manager [at Mavis] told me that all the
guys would stare at me.”)). Murnane applied to work at Mavis on three or four
occasions between 2009 and 2010. (Pl. 56.1 ¶ 132). Mavis did not hire here
and, indeed, never contacted her. (Id.).
c.

Pyper Braly

Pyper Braly, a tire technician with two years’ experience changing tires in
a pit crew during NASCAR races and another nine months’ experience as a
mechanic, applied to work at Mavis’s Huntington Station, New York store, but
when she handed in her application, the male manager who took the
application was rude and dismissive. (Pl. 56.1 ¶ 125). Mavis never called
Braly in for an interview and did not hire Braly. (Id. at ¶ 126).
d.

Amber Luby

Amber Luby, who had worked for roughly six months as a tire technician
with Goodyear, applied to work for Mavis at the East Greenbush, New York
store in October 2009. (Pl. 56.1 ¶ 127). When Luby returned to Mavis to follow
up on her initial application, she was told that Mavis did not have a record of
her application. (Id. at ¶ 128). She completed a second application, but never
heard back from Mavis and was not hired. (Id.). In the summer of 2012, Luby
applied a third time to Mavis, but again was not hired. (Id. at ¶ 130).
e.

Julie Griffin

Julie Griffin worked as a Head Service Advisor at a Honda dealership for
four years when she first applied to work at Mavis. (Pl. 56.1 ¶ 133). She was

6

interviewed by a Mavis Store Manager, who told her that he wanted to hire her,
but needed approval from his RTM. (Id. at ¶ 134). Griffin eventually learned
that the RTM decided to hire the Store Manager’s friend for the position. (Def.
56.1 Response ¶ 135).
f.

Ashley Jackson

Ashley Jackson, who, at the time of her application to Mavis had 17
months of automotive experience, applied twice to work for Mavis. (Pl. 56.1
¶ 136). Jackson called Mavis after each of her two job applications to follow up
on her applications. She testified that she was “brushed off,” and told that
Mavis was not hiring. (See id. at ¶ 137; Def. 56.1 Response ¶ 137).
g.

Barbara Evans

Barbara Evans, who at the time of her application to Mavis had managed
a repair shop for over six years, submitted an online application to work at
Mavis. (Pl. 56.1 ¶ 139). Evans contacted Mavis to follow up on her online
application and was told that if she submitted an application online, Mavis
would have received it. Although she provided Mavis with her contact
information, she never heard anything back either confirming receipt of her
application or advising her about its status. (Id. at ¶ 140).
h.

May Menawi

May Menawi applied for employment with Mavis in 2008 to be an
Assistant Manager at its Tarrytown, New York location. (Pl. 56.1 ¶ 142).
At that time, she worked as a service writer at Sears Automotive in White
Plains. (Id. at ¶ 143). Days after Menawi completed an application at the

7

Tarrytown location, the Store Manager interviewed Menawi for approximately
30 to 60 minutes. (Def. 56.1 Response ¶ 143). Following the interview,
Menawi was not hired. (Id.).
i.

Miranda Cooper and Adeline Van Dyke

In similar fashion, Miranda Cooper and Adeline Van Dyke applied to
Mavis but were not hired. (Pl. 56.1 ¶¶ 144-48). At the time Cooper applied,
she had a Board of Cooperative Educational Services (“BOCES”) certification for
general automotive repairing, and a work history at various automotive shops.
(Id. at ¶¶ 144-45). Van Dyke had automotive training and knowledge and had
performed supervisory responsibilities at her husband’s auto body shop. (Id. at
¶¶ 149-50).
In total, there are 42 female applicants whose “credentials on paper”
make them well-qualified candidates, according to the EEOC, for Mavis
employment. (Pl. 56.1 ¶ 91). The EEOC has also identified a number of male
applicants Mavis hired during the 2008 to 2012 time period who had equal or
less experience than the unsuccessful female applicants. (See id. at ¶¶ 15158).
3.

Mavis’s Record-Keeping

The EEOC has also expressed concerns about Mavis’s record-keeping
practices. By way of background, Mavis contends that it had an application
retention process in place during 2008 through 2012. (Def. 56.1 ¶¶ 220-24).
Store Managers were told to collect the applications received in-person at store
locations in a manila envelope; RTMs were responsible for collecting those

8

applications from each store and providing them to Human Resources on a
weekly basis. (Pl. 56.1 Response ¶ 220). This application retention process
was not memorialized or communicated to store management in writing. (Id.).
Instead, RTMs conveyed this policy to Store Managers verbally. (Id.; Def. 56.1
¶ 229). Applications that were submitted via email were forwarded to Andre
and retained in electronic form indefinitely. (See Def. 56.1 ¶ 226-29).
Each of the unsuccessful female applicants identified by the EEOC
during the 2008 to 2012 time period testified that she submitted application
forms for employment, but Defendants have been unable to locate all of these
applications. (See Pl. 56.1 ¶¶ 164-76). The EEOC contends that the inability
to locate these applications, along with testimony from Mavis employees,
demonstrates that Defendants’ record-keeping policy was not effective at
retaining all applications. (See id. at ¶¶ 177-200). Defendants counter that,
although their employees “could not testify under oath that every application
submitted to [their] 140+ locations were kept and delivered to Human
Resources, [they] did not cherry pick applications and throw them away.” (E.g.,
Def. 56.1 Response ¶ 187).
4.

The Experts’ Statistical Analyses

The EEOC’s expert economist, Dr. Marc Bendick, Ph.D., issued an initial
report (the “Bendick Report” (Dkt. #97-1)) on October 25, 2013. (Pl. 56.1 ¶ 5).
In relevant part, Dr. Bendick opined that:


Because of Mavis’s incomplete retention of applications
and the possible chilling of female applicants, it was not
appropriate to rely on Mavis’s records of applications

9

received to determine the expected representation of
women in Mavis’s hiring. (Bendick Report ¶ 7(a)).


Based on widely-accepted public data from the U.S.
Census Bureau, the EEOC, and similar sources, a
reasonable, conservative estimate of the expected
representation of women in Mavis’s hiring for in-store
positions is 13.3% for Store Managers, 12.4% for
Assistant Managers, 2.2% for Mechanics, and 5.9% for
Technicians. (Id. at ¶ 7(b)).



Throughout 2008 to 2012, the number of women
employed by Mavis was substantially lower than the
expected employment of women. This shortfall was
present in each job category in every year examined.
(Id. at ¶ 7(c)).



The shortfall in female employees observed between
2008 and 2012 included employment of zero women in
all years among Technicians; zero women in all years
among Mechanics; zero women among Store Managers;
and one female Assistant Manager. (Id. at ¶ 7(f)).



Comparing these figures with those that would be
expected in Mavis’s workforce, the number of Mavis’s
“missing” female employees can be calculated as
between: 7.4 to 16.3 Store Managers per year; 11.5 to
22.3 Assistant Managers per year; 1.9 to 4.0 Mechanics
per year; and 13.0 to 26.9 Technicians per year. (Id. at
¶¶ 38-41).



Throughout 2008 to 2012, the difference between the
expected employment of women in Mavis stores and
their actual employment was far too large to have arisen
by chance alone. With the exception of the underrepresentation of women among Mechanics, where the
expected number of female employees was too small for
meaningful calculation of statistical significance, the
under-representation of women in every job category in
every year from 2008 through 2012 was statistically
significant at levels ranging from 2.9 standard
deviations to 5.4 standard deviations. These numbers
of standard deviations mean that the probability that
the observed under-representation arose by chance
alone varied from less than one chance in 100 to less

10

than one chance in ten million. (Id. at ¶ 7(d)).6
Defendants retained their own expert economist, Paul F. White, Ph.D., to
review the Bendick Report, discuss any concerns with Dr. Bendick’s
methodology and conclusions, and conduct an analysis of Mavis’s hiring
practices for the 2008 to 2012 time period. (Def. 56.1 ¶ 245). Dr. White issued
his report (the “White Report” (Dkt. #97-4)) on January 17, 2014. In relevant
part, Dr. White opined that the Bendick Report was flawed in three ways:


Dr. Bendick should have compared Mavis’s hiring to a
universe no broader than that of “Tire Dealers.”
Instead, Dr. Bendick inappropriately compared the
Mavis workforce to that found in other industries whose
primary purpose is different from that of a tire retailer.
Because some of the industries included in Dr.
Bendick’s analysis had higher proportions of female
workers, his methodology resulted in an overstatement
of the number of “missing” females at Mavis. (See White
Report 1-2, 12-16).7



Dr. Bendick did not analyze the number of “hires” at
Mavis during 2008 to 2012. Instead, he compared the
number of females he believes should have been present
each year in the Mavis workforce with Mavis’s
incumbent workforce for that year. (Id. at 1, 16-17)



Dr. Bendick did not analyze female applicant
availability among the applications that Mavis
maintained. His decision not to analyze the applicant
flow data was based on a mere “possibility” that some
applications may not be accounted for and would result

6

Dr. Bendick had also opined as to the amount of damages female applicants suffered as
a result of Mavis’s hiring practices. Defendants’ expert offers a competing damages
analysis. Neither of these analyses is implicated by the instant motions.

7

For example, Dr. Bendick took into account the percentage of females employed as
“Wheeled Vehicle Repairer[s]” by the U.S. Army (5.8%), and as “Auto Maintenance
Warrant Officer[s]” by the U.S. Air Force (5.1%), in estimating the expected percentage
of Mavis’s female Mechanics. (Bendick Report ¶ 34). Notably, the percentage of female
employees in Dr. White’s preferred pool of “Craft Workers” employed by “Tire Dealers,”
is only 2.3%.

11

in a distortion of the female availability rate. Dr.
Bendick should have examined the applicant flow data
in order to (at the very least) provide a range of potential
availability measures. (Id. at 2, 24).
Dr. White performed his own statistical analysis, using the number of
individuals hired for the four positions and comparing the number of females
hired with the number of females expected to be hired by companies
considered Tire Dealers. (White Report 27-29). Notably, Dr. White looked at
the number of “missing” female hires (and calculated the standard deviation)
on a year-by-year basis rather than for the entire five-year period. (See, e.g., id.
at 28). In doing so, Dr. White concluded that the shortfall in female hiring per
position, per year, was 0.82 to 2.08 for Store Managers; 2.18 to 8.51 for
Assistant Managers; 2.23 to 3.57 for Mechanics; and 5.32 to 8.47 for
Technicians. (See id. at 27-29). Using the shortfalls in hiring on a per-year
basis, Dr. White calculated the statistical significance to be between 0.95
standard deviations and 3.07 standard deviations. (Id.).
Dr. Bendick issued a rebuttal report (the “Bendick Rebuttal” (Dkt. #977)) on February 18, 2014. (Pl. 56.1 ¶ 5). First, Dr. Bendick responded to Dr.
White’s criticism of the use of incumbent rather than hiring data. As it turns
out, he claimed, the EEOC had requested hiring data from Defendants early on
during discovery, but Defendants had not provided this data until after the
Bendick Report was issued (and, indeed, not until after the White Report had
been completed). (See Bendick Rebuttal ¶¶ 8-10). With the hiring data now at
his disposal, Dr. Bendick modified his statistical analysis to focus on the

12

employees hired by Mavis during 2008 to 2012, rather than on the number of
incumbent employees. (Id. at ¶ 10). Notably, Dr. Bendick concluded that “the
shortfall results in the ‘hiring analysis’ are generally more statistically
significant than those in the ‘incumbent employees’ analysis because the
number of hiring transactions over 2008-2012 is larger than the number of
incumbent employees in individual years during that period.” (Id. at ¶ 11; see
also id. at ¶ 10 (reporting standard deviations ranging from 3.5 to 10.3)).
Second, the Bendick Rebuttal addressed Dr. White’s opinion that the
correct comparator pool consists of “Tire Dealers” rather than a broader set of
employers. (Bendick Rebuttal ¶¶ 13-20). Dr. Bendick disagreed with White’s
assessment, but nonetheless repeated his analysis using Dr. White’s suggested
pool of Tire Dealers. Dr. Bendick concluded that even relying on “this
inappropriate estimate of availability, statistically significant shortfalls of
women’s hiring during 2008-2012 are still universally documented.” (Id. at 19;
see also id. (reporting standard deviations ranging from 2.8 to 5.7)).
Third, Dr. Bendick addressed his rejection of Defendants’ applicant flow
data as unreliable, a rejection that Dr. White had criticized. (Bendick Rebuttal
¶ 21). Specifically, using Dr. White’s tabulation of the number of applications
from female applicants in Mavis’s possession, Dr. Bendick concluded that the
number of applications was substantially lower than the number of
applications from female candidates that one would expect. He concluded this
using both his own estimate of the percentage of females in the workforce and
Dr. White’s estimated percentage of female workers at Tire Dealers. (Id. at

13

¶¶ 25-26). Using his own estimates, Dr. Bendick concluded that Mavis’s
application records contain: (i) 39.8% as many female applicants as expected
for the Store Manager position; (ii) 69.3% as many female applicants as
expected for the Assistant Manager position; (iii) 68.2% as many female
applicants as expected for the Mechanic position; and (iv) 20.3% as many
female applicants as expected for the Technician position. (Id. at ¶ 25). In
contrast, using Dr. White’s estimated percentage of females in the workforce,
Dr. Bendick concluded that Mavis’s application records contain: (i) 58.2% as
many female applicants as expected for the Store Manager position; (ii) 86.9%
as many female applicants as expected for the Assistant Manager position; (iii)
65.2% as many female applicants as expected for the Mechanic position; and
(iv) 63.2% as many female applicants as expected for the Technician position.
(Id. at ¶ 26). Dr. Bendick concluded that these discrepancies make “it is
extremely likely that [an] undercount [of female applicants] actually exists” in
Mavis’s application records. (Id. at ¶ 28 (emphasis in original)).
Finally, putting aside his reservations about the reliability of the
applicant flow data, Dr. Bendick performed a statistical analysis based on the
applicant flow data and the number of hires per year. (Bendick Rebuttal ¶ 29).
In other words, Dr. Bendick assumed for the sake of analysis that the
percentage of Mavis’s actual female applicants would equal the percentage of
female employees Mavis hired for each position, and multiplied the number of
hires by the percentage of female applicants to come up with the expected
number of female hires. (See id.). Dr. Bendick concluded that “even …

14

apply[ing] availability figures computed from Mavis’s incomplete application
records, statistically significant shortfalls of women’s hiring during 2008-2012
are still universally documented.” (Id. (reporting standard deviations ranging
from 2.1 to 5.0)).
One key difference between the experts’ approaches is that Dr. Bendick
aggregated the hiring data before computing statistical significance, as
represented in the number of standard deviations, while Dr. White looked at
the data in each year separately. (Compare Bendick Rebuttal ¶¶ 12, 20
(“combining five years’ hiring data before computing statistical significance”),
with White Report 27 (calculating the standard deviation based on “the
shortfall number of … hires for each job during each year”)). This difference
was addressed in the Bendick Rebuttal, as well as in the depositions of the
experts. Dr. Bendick explained that he initially utilized a year-by-year analysis
when he performed a comparison based on incumbent employee data “only
because the continued employment of individuals from one year to the next
makes it statistically inappropriate to add together all years’ workforces as
though each year’s workforce is independent of the workforce in previous
years.” (Bendick Rebuttal ¶ 11 n.10). However, Bendick opined that “no such
issue of lack of independence arises when hiring decisions are analyzed, so that
it is appropriate to analyze all years’ hiring decision together, not separately for
individual years.” (Id. (emphasis added); see also Bendick Dep. 329 (“The
hypothesis I was assigned to test was that there was a pattern and practice
applying across [a] five-year period of under hiring women compared to their

15

expected representation. The statistical test that corresponds to that
hypothesis is the one that … look[s] at the total corpus of hires across the five
years[.]”); id. at 332 (“The hires in individual years are independent essentially
of hires in other years.”)).
Dr. White was questioned extensively during his deposition on his
decision to analyze the data and report statistical significance on a year-byyear basis. In his answers, Dr. White disagreed with Dr. Bendick’s decision to
report only aggregate data for the entire time period. (White Dep. 30).
Specifically, Dr. White contended that “[b]ecause … the hiring decisions from
one year can affect the demand for employees the next year, … they’re [not]
totally independent decisions .… It affects the number of hires from one year to
the next. It can.” (Id. at 30-31). When pressed to explain the practical import
of analyzing the data in the aggregate, or on a year-by-year basis, Dr. White
offered the following summary: When the data is aggregated, “for each of the
four jobs across all years there is a statistically significant finding. For some of
the years it’s not statistically significant on a year-to-year basis, and for some
of the jobs it’s … statistically significant on a year-to-year basis.” (Id. at 45).
Dr. White acknowledged that “[w]ith smaller numbers of observations in [a
year-by-year] analysis it takes a larger difference between the actual and
expected outcome to generate a statistically significant finding.” (Id. at 33). He
testified, however, that in his experience, “the most common approach is to
report [statistical significance] both ways” — that is, in the aggregate and on a

16

year-by-year basis — “[a]nd to then let attorneys and the court decide the
implications from it.” (Id. at 32).
B.

Procedural Background
On January 26, 2009, Nicole Haywood filed an EEOC charge of

discrimination, alleging that Mavis failed to hire her because of her sex. (Pl.
56.1 ¶ 162). On February 9, 2009, the EEOC provided notice to Mavis of
Haywood’s charge of discrimination. (Id. at ¶ 163). On November 3, 2010, the
EEOC issued a request for information to Mavis, asking that it provide a list of
all employees identified by their name, sex, as well as other information; all
applications for employment and resumes submitted to Mavis; and a list of all
persons hired by Mavis. (Id. at ¶ 164). Thereafter, on August 5, 2011, the
EEOC issued a letter of determination that stated, in part, that the EEOC had
determined that: (i) “there [wa]s reasonable cause to believe that [Mavis]
violated Title VII by failing to … hire [Haywood] and a class of females because
of their sex”; and (ii) Mavis “fail[ed] to preserve employment applications for one
year and, once the subject charge was filed, … fail[ed] to preserve application[s]
until the final disposition of the charge.” (Id. at ¶ 165).
The EEOC brought this action on January 31, 2012. (Dkt. #1).
Defendants filed their Answer on April 1, 2012. (Dkt. #7). At the close of
discovery, the EEOC and Defendants indicated their intent to file cross-motions
for summary judgment. (Dkt. #65, 66). On November 25, 2014, the Court held
a pre-motion conference to discuss the anticipated motions for summary
judgment, as well as the EEOC’s motion to bifurcate the trial. (See November

17

25, 2015 Conference Transcript (“Nov. 25 Tr.”) (Dkt. #75)). Although both
parties argued that the trial in this action should be bifurcated and proceed in
two phases, the parties did not agree as to whether the issue of punitive
damages should be decided by a jury during the liability phase (“Phase I”) or
during the remedial phase (“Phase II”). (See id. at 6-10, 14-16; see generally
Discussion Sec. A(2), infra (describing the liability and remedial phases of a
pattern-or-practice case)). Following the pre-motion conference, the EEOC filed
its motion for summary judgment and for bifurcation of the trial on February
13, 2015. (Dkt. #94-97). Defendants, opting not to file the summary judgment
motion they previously anticipated, filed their opposition on March 27, 2015.
(Dkt. #101-04). The motions were fully briefed upon the filing of the EEOC’s
reply on April 13, 2015. (Dkt. #106-07).
DISCUSSION
A.

Applicable Law
1.

Summary Judgment Motions Generally

Under Federal Rule of Civil Procedure 56(a), summary judgment may be
granted only if all the submissions taken together “show that there is no
genuine issue as to any material fact and that the moving party is entitled to
judgment as a matter of law.” Celotex Corp. v. Catrett, 477 U.S. 317, 322
(1986) (internal citation and quotation marks omitted); see also Anderson v.
Liberty Lobby, Inc., 477 U.S. 242, 247-48 (1986). The moving party bears the
initial burden of demonstrating “the absence of a genuine issue of material
fact.” Celotex, 477 U.S. at 323. A fact is “material” if it “might affect the

18

outcome of the suit under the governing law,” and is genuinely in dispute “if
the evidence is such that a reasonable jury could return a verdict for the
nonmoving party.” Anderson, 477 U.S. at 248; see also Jeffreys v. City of New
York, 426 F.3d 549, 553 (2d Cir. 2005) (citing Anderson). The movant may
discharge this burden by showing that the nonmoving party has “fail[ed] to
make a showing sufficient to establish the existence of an element essential to
that party’s case, and on which that party will bear the burden of proof at
trial.” Celotex, 477 U.S. at 322; see also Selevan v. N.Y. Thruway Auth., 711
F.3d 253, 256 (2d Cir. 2013) (finding summary judgment appropriate where the
non-moving party fails to “come forth with evidence sufficient to permit a
reasonable juror to return a verdict in his or her favor on an essential element
of a claim” (internal quotation marks omitted)).
If the moving party meets this burden, the nonmoving party must “set
out specific facts showing a genuine issue for trial” using affidavits or
otherwise, and cannot rely on the “mere allegations or denials” contained in the
pleadings. Anderson, 477 U.S. at 248, 250; see also Celotex, 477 U.S. at 32324; Wright v. Goord, 554 F.3d 255, 266 (2d Cir. 2009). The nonmoving party
“must do more than simply show that there is some metaphysical doubt as to
the material facts,” Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S.
574, 586 (1986), and cannot rely on “mere speculation or conjecture as to the
true nature of the facts to overcome a motion for summary judgment,” Knight v.
U.S. Fire Ins. Co., 804 F.2d 9, 12 (2d Cir. 1986) (quoting Quarles v. Gen. Motors
Corp., 758 F.2d 839, 840 (2d Cir. 1985)). Furthermore, “[m]ere conclusory

19

allegations or denials cannot by themselves create a genuine issue of material
fact where none would otherwise exist.” Hicks v. Baines, 593 F.3d 159, 166
(2d Cir. 2010) (quoting Fletcher v. Atex, Inc., 68 F.3d 1451, 1456 (2d Cir. 1995))
(internal quotation marks and citations omitted); see also Vargas v. Transeau,
514 F. Supp. 2d 439, 442 (S.D.N.Y. 2007) (observing that “the mere existence
of a scintilla of evidence in support of the [non-movant’s] position will be
insufficient” to defeat summary judgment (internal quotation marks and
citations omitted)), aff’d sub nom. Vargas v. Pfizer, Inc., 352 F. App’x 458 (2d
Cir. 2009) (summary order).
“When ruling on a summary judgment motion, the district court must
construe the facts in the light most favorable to the non-moving party and
must resolve all ambiguities and draw all reasonable inferences against the
movant.” Dallas Aerospace, Inc. v. CIS Air Corp., 352 F.3d 775, 780 (2d Cir.
2003). However, in considering “what may reasonably be inferred” from
witness testimony, the court should not accord the non-moving party the
benefit of “unreasonable inferences, or inferences at war with undisputed
facts.” Berk v. St. Vincent’s Hosp. & Med. Ctr., 380 F. Supp. 2d 334, 342
(S.D.N.Y. 2005) (citing County of Suffolk v. Long Island Lighting Co., 907 F.2d
1295, 1318 (2d Cir. 1990)).
2.

Title VII Pattern-or-Practice Discrimination

“Title VII … prohibits various forms of employment discrimination on the
basis of race, color, religion, sex, or national origin.” E.E.O.C. v. Bloomberg
L.P., 967 F. Supp. 2d 802, 809 (S.D.N.Y. 2013) (quotations marks and citation

20

omitted). Title VII prohibits both intentional discrimination — known as
disparate treatment — and unintentional discrimination practices that have a
disproportionately adverse effect on a protected class — known as disparate
impact. United States v. City of New York, 713 F. Supp. 2d 300, 316 (S.D.N.Y.
2010) (citing Ricci v. DeStefano, 557 U.S. 557, 577 (2009)). This is a disparate
treatment case for sex discrimination in Defendants’ hiring practices.
A pattern-or-practice claim is a particular vehicle to bring a Title VII
case. “Pattern-or-practice disparate treatment claims focus on allegations of
widespread acts of intentional discrimination against individuals. To succeed
on a pattern-or-practice claim, plaintiffs must prove more than sporadic acts of
discrimination; rather, they must establish that intentional discrimination was
the defendant’s ‘standard operating procedure.’” Robinson v. Metro-N.
Commuter R.R. Co., 267 F.3d 147, 158 (2d Cir. 2001) (quoting Int’l Bhd. of
Teamsters v. United States, 431 U.S. 324, 336 (1977)), abrogated on other
grounds by Wal-Mart Stores, Inc. v. Dukes, 131 S. Ct. 2541 (2011). In light of
all the circumstances of the case, discrimination must be proven by a
preponderance of the evidence to be the defendant’s “regular” policy.
Teamsters, 431 U.S. at 336; accord United States v. City of New York, 717 F.3d
72, 88 (2d Cir. 2013).
There is a “manifest” difference between claims of individual
discrimination and claims of a pattern or practice of discrimination. Cooper v.
Fed. Reserve Bank of Richmond, 467 U.S. 867, 876 (1984). In part, this is
because the Supreme Court has cautioned that isolated or individual instances

21

of discrimination, even if true, should not be construed to turn every Title VII
case into “a potential companywide class action.” Gen. Tel. Co. of Sw. v. Falcon,
457 U.S. 147, 159, (1982). Indeed, “courts considering what evidence is
necessary to show that an employer routinely and purposely discriminated
have also required substantial proof of the practice.” King v. Gen. Elec. Co.,
960 F.2d 617, 624 (7th Cir. 1992); see also In re W. Dist. Xerox Litig., 850 F.
Supp. 1079, 1085 (W.D.N.Y. 1994) (“[T]he burden of establishing a pattern or
practice of discrimination is not an easy one to carry.”).
Pattern-or-practice cases proceed in two phases: a liability phase and a
remedial phase. Robinson, 267 F.3d at 158. In the liability phase, with which
the instant motion for summary judgment is concerned, “plaintiffs must
produce sufficient evidence to establish a prima facie case of a policy, pattern,
or practice of intentional discrimination against the protected group.” Id. The
Second Circuit has recognized that Supreme Court precedent “sets a high bar
for the prima facie case the Government or a class must present in a patternor-practice case: evidence supporting a rebuttable presumption that an
employer acted with the deliberate purpose and intent of discrimination against
an entire class.” City of New York, 717 F.3d at 87.8

8

As noted, the summary judgment motion concerns only the liability phase of the
EEOC’s pattern-or-practice case. If a plaintiff succeeds in proving liability, the court
may fashion classwide injunctive relief, and then, in the second phase — known as the
relief phase — individual plaintiffs may avail themselves of a rebuttable inference of
discrimination in litigating a particular adverse employment decision to obtain
individual relief. During this second phase, “the burden then rests on the employers to
demonstrate that the individual applicant was denied an employment opportunity for
lawful reasons.” Int’l Bhd. of Teamsters v. United States, 431 U.S. 324, 362 (1977)
(citation omitted). “This rebuttable inference arising at the relief stage, after proof by
preponderance of the evidence at the liability stage of the existence of a pattern or

22

To establish liability, plaintiffs’ cases alleging a pattern or practice of
discrimination are characterized by a “heavy reliance on statistical evidence.”
Reynolds v. Barrett, 685 F.3d 193, 203 (2d Cir. 2012) (quoting Robinson, 267
F.3d at 158 n.5); see also Attenborough v. Constr. & Gen. Bldg. Laborers’ Local
79, 691 F. Supp. 2d 372, 388 (S.D.N.Y. 2009) (“[S]tatistical evidence is critical
to the success of a pattern-or-practice disparate treatment claim.”). “[T]he
liability phase is largely preoccupied with class-wide statistical evidence
directed at establishing an overall pattern or practice of intentional
discrimination.” Robinson, 267 F.3d at 168. Statistics are so central to
pattern-or-practice cases that they “alone can make out a prima facie case of
discrimination if the statistics reveal a gross disparity” in employee treatment.
Id. at 158 (internal quotation marks and citation omitted); see also Hazelwood
Sch. Dist. v. United States, 433 U.S. 299, 307-08 (1977) (“Where gross
statistical disparities can be shown, they alone may in a proper case constitute
prima facie proof of a pattern or practice of discrimination.”).
Statistics, however, “are not irrefutable,” and they must be viewed in
light of all of the circumstances of the case. Teamsters, 431 U.S. at 339. A
pattern-or-practice claim usually includes “a combination of strong statistical
evidence of disparate impact coupled with anecdotal evidence of the employer’s
intent to treat the protected class unequally.” Mozee v. Am. Comm’l Marine

practice of intentional discrimination, should not be confused with the rebuttable
presumption arising at the threshold of the liability stage, after presentation of only a
prima facie case of such a pattern or practice.” United States v. City of New York, 717
F.3d 72, 88 n.15 (2d Cir. 2013).

23

Serv. Co., 940 F.2d 1036, 1051 (7th Cir. 1991); see also Robinson, 267 F.3d at
158.
Anecdotal evidence normally serves a distinct purpose: It brings “the
cold numbers convincingly to life.” Teamsters, 431 U.S. at 339. “To the extent
that evidence regarding specific instances of alleged discrimination is relevant
during the liability stage, it simply provides ‘texture’ to the statistics.”
Robinson, 267 F.3d at 168; see O’Donnell Constr. Co. v. Dist. of Columbia, 963
F.2d 420, 427 (D.C. Cir. 1992) (“Anecdotal evidence is most useful as a
supplement to strong statistical evidence[.]”). Therefore, “[w]hile anecdotal
evidence may suffice to prove individual claims of discrimination, rarely, if ever,
can such evidence show a systemic pattern of discrimination.” Middleton v.
City of Flint, 92 F.3d 396, 405 (6th Cir. 1996) (quoting O’Donnell, 963 F.2d at
427) (emphasis in Middleton).
Once a prima facie case has been established, the burden then shifts to
the employer to rebut the prima facie case, that is, to demonstrate that
plaintiffs’ “proof is either inaccurate or insignificant.” City of New York, 717
F.3d at 85 (quoting Teamsters, 431 U.S. at 360) (emphases omitted).
Defendants may “assault ... the source, accuracy, or probative force” of
plaintiffs’ proof. Id. at 86 (citation omitted). In this setting, the employer’s
burden is one of production — not persuasion. See Reeves v. Sanderson
Plumbing Prods., Inc., 530 U.S. 133, 142 (2000) (“This burden is one of
production, not persuasion; it can involve no credibility assessment.” (internal
quotation marks and citation omitted)); St. Mary’s Honor Ctr. v. Hicks, 509 U.S.

24

502, 511 (1993) (“[T]he Title VII plaintiff at all times bears the ultimate burden
of persuasion” (internal quotation marks and citation omitted)); Texas Dep’t of
Cmty. Affairs v. Burdine, 450 U.S. 248, 255 (1981) (“If the defendant carries
this burden of production, the presumption raised by the prima facie case is
rebutted[.]”); City of New York, 717 F.3d at 85 (“[T]he rebuttal burden in [a
pattern-or-practice case] is one of ‘production.’” (collecting cases)).
Finally, the Second Circuit has admonished that
[while] it is always open to a[n] [employer] to meet its
burden of production by presenting a direct attack on
the statistics relied upon to constitute a prima facie
case[,] … the rebuttal need not be so limited. A
defendant may rebut the inference of a discriminatory
intent by accepting a plaintiff’s statistics and producing
non-statistical evidence to show that it lacked such an
intent.
City of New York, 717 F.3d at 85; see also id. at 90 (“[T]he employer may satisfy
its burden of production by ‘provid[ing] a nondiscriminatory explanation for the
apparently discriminatory result.’” (quoting Teamsters, 431 U.S. at 360)).
B.

Analysis
1.

The EEOC May Bring Its Pattern-or-Practice Claim Under
Section 706

As a threshold matter, Defendants argue that the EEOC does not have
the authority to bring a “pattern or practice” claim under Section 706 of Title
VII, 42 U.S.C. § 2000e-5(b). (Def. Opp. 3-8). Instead, they argue that Section
707 of Title VII, 42 U.S.C. § 2000e-6, which explicitly references “pattern or
practice” suits, provides the EEOC’s only avenue for maintaining such a claim.
(See Def. Opp. 7). Compare Section 706 (authorizing the “[EEOC] to

25

eliminate … unlawful employment practice by informal methods of conference,
conciliation, and persuasion,” and providing that where “the [EEOC] has been
unable to secure from the respondent a conciliation agreement …, the [EEOC]
may bring a civil action against [the] respondent”), with Section 707
(authorizing the EEOC to “bring a civil action” against a private entity when it
“has reasonable cause to believe that any person or group of persons is
engaged in a pattern or practice of resistance to the full enjoyment of any of the
rights secured by this subchapter”).
Whether the EEOC may bring a pattern-or-practice claim under Section
706 is significant, Defendants argue, because of (i) the different burden-shifting
frameworks between the two provisions and (ii) the different remedies available.
(See Def. Opp. 7-8). On the issue of burden-shifting, Defendants argue that
Section 707 claims must proceed under the Teamsters burden-shifting
framework, see 431 U.S. at 336, whereas Section 706 claims proceed under the
burden-shifting framework of McDonnell Douglas Corp. v. Green, 411 U.S. 792
(1973). (See Def. Opp. 5-6). With respect to the available remedies,
Defendants note that Section 707 provides for equitable relief only, whereas
Section 706 also provides for compensatory and punitive damages. (Id. at 6-7).
In sum, Defendants claim that “the EEOC is attempting to avail itself of the
more advantageous Teamsters burden-shifting framework applicable to
[Section] 707 claims and seek compensatory and punitive damages, which are
only available under [Section] 706.” (Id. at 7).

26

The Second Circuit has not addressed the specific question of whether
the EEOC can bring a pattern-or-practice claim under Section 706. Indeed,
the parties have not cited (and this Court could not find) a single district court
within the Second Circuit to weigh in on this particular issue. However, before
addressing the merits of the parties’ arguments as to whether the EEOC can
bring a pattern-or-practice claim under Section 706, the Court must briefly
address Defendants’ reliance on two particular district court cases to support
their argument.
Defendants asseverate that: (i) “[Section] 706 does not permit claims
alleging a ‘pattern or practice’ of discrimination” (Def. Opp. 5 (emphasis in
original)); and (ii) a “[Section] 706 and 707 ‘hybrid action’ is simply not
permitted under Title VII” (id. at 7). As support for these propositions,
Defendants cite the district court’s opinions in EEOC v. Bass Pro Outdoor
World, LLC, No. 11 Civ. 3425 (KPE) (“Bass Pro II”), 2013 WL 1124063, at *2 n.3
(S.D. Tex. Mar. 18, 2013), and EEOC v. Bass Pro Outdoor World, LLC (“Bass Pro
III”), 35 F. Supp. 3d 836, 841 (S.D. Tex. 2014). (See Def. Opp. 5, 7). As the
EEOC points out in its reply (Pl. Reply 3), Defendants’ citation to Bass Pro II
and III for support is highly misleading, since the district court in that case
ultimately rejected the arguments Defendants now espouse.
The district court in Bass Pro III explicitly endorsed the EEOC’s
maintenance of a pattern-or-practice claim under Section 706 and granted the
EEOC’s motion for reconsideration on this very issue — thereby reversing
course from its earlier pronouncements in EEOC v. Bass Pro Outdoor World,

27

LLC (“Bass Pro I”), 884 F. Supp. 2d 499, 520 (S.D. Tex. 2012), and Bass Pro II.
(See Pl. Reply 2-3). Indeed, the district court in Bass Pro III could not have
been clearer on this point, indicating that “[u]pon exhaustive review of Title
VII’s statutory scheme and the cases interpreting it, … [the court] was wrong to
summarily conclude in [Bass Pro I] that § 706 claims cannot be proven using
the Teamsters framework.” Bass Pro III, 35 F. Supp. 3d at 859 (emphasis
added).9 Notably, the portion of Bass Pro III that Defendants quote in their
opposition is not native to Bass Pro III; it was originally in Bass Pro I, and was
repeated in Bass Pro III only as part of a recap of the procedural history of the
litigation. (See Def. Opp. 7 (quoting Bass Pro III, 35 F. Supp. 3d at 841)).
Ordinarily, the Court would chalk up Defendants’ reliance on Bass Pro II
and Bass Pro III to oversight.10 However, given Defendants’ citation to another
obsolete district court opinion, it may represent more of a strategic choice on
their part. Defendants chastise several district courts for “somehow ignor[ing]”
the “critical distinction[s]” between Section 706 and 707. (Def. Opp. 7
(referring to the EEOC’s position as “untenable” given “long-recognized
jurisprudence” to the contrary)). To that end, Defendants cite Serrano v. Cintas

9

The Court notes that the district court subsequently granted the employer’s motion for
an interlocutory appeal to the Fifth Circuit. See EEOC v. Bass Pro Outdoor World, LLC
(“Bass Pro IV”), No. 11 Civ. 3425 (KPE), 2014 WL 6453606, at *4 (S.D. Tex. Nov. 17,
2014) (“Because the question is close and because an immediate appeal will give the
Court and the parties needed certainty as to how the case should proceed to trial, the
Court will grant Defendants’ motion for interlocutory appeal of the Motion to Reconsider
order.”).

10

If, after the EEOC filed its reply, Defendants realized that they had cited in error cases
that had been reconsidered or vacated, they would have been well-served to send a
letter to the Court explaining their inadvertence.

28

Corp. (“Serrano I”), 711 F. Supp. 2d 782 (E.D. Mich. 2010), as an example of an
opinion in which a court stood fast against such obliviousness and, in the
process, “den[ied] the EEOC the right to utilize the Teamsters burden-shifting
framework in a § 706 case.” (Def. Opp. 7). What Defendants fail to
acknowledge, and what the EEOC is understandably quick to point out (Pl.
Reply 3), is that the district court’s decision in Serrano I was promptly vacated
and remanded by the Sixth Circuit. Serrano v. Cintas Corp. (“Serrano II”), 699
F.3d 884, 896 (6th Cir. 2012), cert. denied, 134 S. Ct. 92 (2013). Once again,
the district court decision Defendants cite was explicitly overruled, and the very
argument Defendants advance before this Court was rejected. Id. (“[W]e hold
that the district court erred in concluding that the EEOC may not pursue a
claim under the Teamsters pattern-or-practice framework, pursuant to its
authority vested in § 706 of Title VII.”).11
The Court draws attention to Defendants’ citation to defunct iterations of
Bass Pro and Serrano to explain why, if reliance on these decisions was indeed
a strategic choice on Defendants’ part, it was a poor one. Where, as here, there
is no Second Circuit precedent that controls this Court’s decision, Defendants
had an opportunity to persuade the Court that Bass Pro III, Serrano II, and

11

Indeed, the decision in Serrano II should have been doubly hard for Defendants to
ignore, as it featured prominently in the district court’s reconsideration opinion in Bass
Pro III. See EEOC v. Bass Pro Outdoor World, LLC (“Bass Pro III”), 35 F. Supp. 3d 836,
847 (S.D. Tex. 2014) (“This Court finds Serrano [II] persuasive …. The result in Serrano
[II] is consistent with the statutory text and purpose, and it is supported by precedent
from the Supreme Court, several courts of appeals, and other district courts”); id. at
850 (“Recent district court opinions confirm that Serrano [II] was rightly decided.”); id. at
852 (“[T]he Court fully agrees with Serrano [II] that, even if Teamsters can be used in
suits brought pursuant to § 706, there remain important distinctions between § 706
and § 707.”).

29

their progeny, were wrongly decided. By vociferously championing decisions
that are no longer good law, Defendants have squandered that opportunity.
Turning to the merits of the debate, the Court notes the clarity of Second
Circuit law that a pattern-or-practice case “is not a separate and free-standing
cause of action ..., but is really merely another method by which disparate
treatment can be shown.” City of New York, 717 F.3d at 83 (quoting Chin v.
Port Authority of N.Y. & N.J., 685 F.3d 135, 148-49 (2d Cir. 2012)). As such,
the pattern or practice method of proof has been found to be available in some
types of discrimination actions, but not others. Compare Franks v. Bowman
Transp. Co., 424 U.S. 747, 772 (1976) (holding that class action plaintiffs may
“carr[y] their burden [by] demonstrating the existence of a discriminatory hiring
pattern and practice”), Teamsters, 431 U.S. at 362 (permitting the Government
to rely on “proof of the pattern or practice” to “support[] [the] inference” of
discrimination under Section 707), and Burgis v. N.Y.C. Dep’t of Sanitation,
— F.3d —, No. 14-1640-cv, 2015 WL 4590507, at *3 (2d Cir. July 31, 2015)
(finding that “in the context of a putative class action alleging employment
discrimination under § 1981 and/or the Equal Protection Clause,” “in certain
circumstances[,] ... statistics alone may be sufficient” “to warrant a plausible
inference of discriminatory intent if they show a pattern or practice that cannot
be explained except on the basis of intentional discrimination”), with Chin, 685
F.3d at 150 (holding that “the pattern-or-practice method of proof is not
available to nonclass, private plaintiffs” (emphasis added)).

30

The Court finds the opinions in Serrano II and Bass Pro III to be
persuasive on the issue of whether the EEOC can use the pattern or practice
method of proof in suits brought under Section 706.12 To the extent the Court
draws liberally from these opinions, it is because they are thorough and wellreasoned. Reliance on these decisions also makes a great deal of practical
sense, as the courts in Serrano II and Bass Pro III considered and rejected each
of the arguments that Defendants propound here, and, as noted supra,
Defendants have not given the Court any reason to think they were wrongly
decided.
First, permitting the EEOC to bring pattern-or-practice cases under
Section 706 does not “render § 707 meaningless or … superfluous[.]” (Def.
Opp. 7). As the court in Bass Pro III explained, there are at least three
circumstances under which the EEOC might initiate an action under Section

12

As the Court in Bass Pro IV noted, “in general, opportunities for courts of appeals to
address this question are rare” because “[s]ettlement pressures typically end large-scale
employment discrimination cases before the entry of an appealable final judgment.”
2014 WL 6453606, at *3. That is not to say that Serrano II stands alone in extending
the “pattern or practice” method of proof to Section 706 cases brought by the EEOC.
Rather, the decision in Bass Pro III represents a growing trend of district courts that
have adopted the Sixth Circuit’s logic, and permitted the EEOC to bring Section 706
pattern-or-practice cases in the wake of Serrano II. See EEOC v. PMT Corp., No. 14 Civ.
599 (DSD) (TNL), 2015 WL 5026029, at *3 (D. Minn. Aug. 24, 2015) (“Whether the
EEOC proceeds under its authority in § 706 or § 707, however, does not restrict the
EEOC’s ability to assert a pattern-or-practice claim.”); EEOC v. Rosebud Restaurants,
Inc., No. 13 Civ. 6656 (ARW), 2015 WL 1594067, at *2 (N.D. Ill. Apr. 7, 2015) (“Because
widespread discriminatory actions are within the category of discrimination prohibited
by Title VII, they are also within the authority granted to the EEOC by § 706.” (internal
citation omitted)); David v. Signal Int’l, LLC, 37 F. Supp. 3d 814, 820 (E.D. La. 2013)
(“[T]his Court finds the Sixth Circuit decision in Serrano [II] persuasive and clarifies its
[previous order] by specifying that the EEOC may bring its pattern and practice claims
under both §§ 706 and 707 using the Teamsters framework[.]”); EEOC v. Pitre, Inc., 908
F. Supp. 2d 1165, 1174 (D.N.M. 2012) (“Section 706 does not prevent the EEOC from
seeking relief for individuals and the general public under a pattern-or-practice
theory.”).

31

707 instead of under Section 706: (i) where the EEOC “has ‘reasonable cause’
to believe such a suit necessary” but where no individual has “file[d] a charge
against the employer,” 35 F. Supp. 3d at 852 (citations omitted); (ii) where the
EEOC wishes to avoid “intervention as of right by parties aggrieved,” id. at 85253; and (iii) where the EEOC intends “to request a three-judge district court in
[a] case[] of ‘general public importance,’ and to appeal a ruling by such a court
directly to the Supreme Court,” id. at 853. These differences vitiate the
argument that Section 707 will be rendered superfluous if the EEOC can
maintain a pattern-or-practice claim under Section 706. See generally Skilling
v. United States, 561 U.S. 358, 413 n.45 (2010) (“Overlap with other federal
statutes does not render [a statute] superfluous.”).
Second, and more fundamentally, the Court does not view an
interpretation that results in statutory overlap as running contrary to
Congress’s intent. (See Def. Opp. 7-8). The Court finds it perfectly rational
that, “in an effort to ensure that the EEOC could prevent unlawful employment
practices, Congress opted to give the EEOC broad and overlapping authority.”
EEOC v. Pitre, Inc., 908 F. Supp. 2d 1165, 1173 (D.N.M. 2012) (internal citation
omitted). The conclusion that “Sections 706 and 707 clearly overlap, providing
the EEOC with multiple routes to bring employers who engage in unlawful
discrimination to justice,” id. at 1173-74, is unsurprising. Indeed “legislative
enactments in [the] area [of employment discrimination] have long evinced a
general intent to accord parallel or overlapping remedies against
discrimination.” Alexander v. Gardner-Denver Co., 415 U.S. 36, 47 (1974); see,

32

e.g., Anderson v. Conboy, 156 F.3d 167, 180 (2d Cir. 1998) (“[W]ith respect to
race discrimination, Title VII and Section 1981 coexist despite partially
overlapping coverage.” (citation omitted)).
Third, the Court disagrees with Defendants that any unfairness inures to
employers from the application of the Teamsters burden-shifting framework
rather than the McDonnell Douglas framework. (See Def. Opp. 7 (arguing that
the EEOC is seeking to “avail itself of the more advantageous Teamsters
burden-shifting framework applicable to § 707”)). As the Sixth Circuit has
explained,
This argument is based on a mistaken premise. The
Teamsters framework is not an inherently easier
standard of proof; it is simply a different standard of
proof. Indeed, under Teamsters, the plaintiff’s initial
burden to make out a prima facie case is
heightened .… [U]nder Teamsters the plaintiff must
demonstrate the existence of a discriminatory
procedure or policy. This is no simple task, as the
plaintiff must prove that discrimination was the
company’s standard operating procedure — the regular
rather than the unusual practice .… Thus, the EEOC
must always weigh the risks — as well as the benefits
— of proceeding under the Teamsters framework, for
doing so involves a greater chance of losing at the prima
facie stage.
Serrano II, 699 F.3d at 896 (internal citations and quotation marks omitted);
see also City of New York, 717 F.3d at 87 (“Teamsters sets a high bar for the
prima facie case the Government or a class must present in a pattern-orpractice case[.]”). This argument also glosses over the fact that the Teamsters
framework is routinely used in pattern-or-practice cases under Section 706
brought as private class actions. See Chin, 685 F.3d at 148 (noting that the

33

pattern or practice “method of proof … originated in the class action context, in
Franks”); see also Bass Pro III, 35 F. Supp. 3d at 849 (finding “no principled
reason why a private class action may employ Teamsters and an EEOC class
action may not”).
Finally, the Court rejects Defendants’ argument that Congress, in
amending Title VII to make compensatory and punitive damages available
under Section 706 but not Section 707, “expressly chose to leave equitable
relief the only remedy for the EEOC pattern[-]or[-]practice claims.” (Def.
Opp. 6-7). As the court in Bass Pro III noted, “[t]his argument misses the
mark”:
Congress did intend to make compensatory and
punitive damages available to victims of a
discriminatory pattern or practice, it just required that
they — or the EEOC — seek them in a § 706 suit.
Simple logic bears this out. Wherever there is a pattern
or practice of discrimination, it is axiomatic that there
are individual victims who have been discriminated
against. These are the same individuals that could
bring their own § 706 suit or whose rights the EEOC
could seek to vindicate in its own § 706 action.
Consequently, to hold … that the EEOC cannot
seek … damages for victims of a discriminatory pattern
or practice would be to place these individuals in a
worse position than those who are discriminated
against in a series of isolated incidents.
[It] is
impossible to believe [that is what] Congress intended.
Bass Pro III, 35 F. Supp. 3d at 853 (emphases in original, footnote and internal
quotation marks omitted). Accordingly, the Court finds that the EEOC may

34

maintain a pattern-or-practice claim under Section 706 as well as under
Section 707.13
2.

Summary Judgment on the EEOC’s Pattern-or-Practice Claim
Is Denied

Turning now to the merits, the EEOC has moved for summary judgment
on its pattern-or-practice claim, contending that its expert’s findings of
Defendants’ statistically significant hiring shortfalls of female employees, the
presence of an “inexorable zero” of female hires, and the anecdotal evidence
from female applicants “conclusively show that Mavis engaged in a pattern or
practice of discrimination against hiring women.” (Pl. Br. 19). Defendants
respond that genuine issues of material fact remain, such that summary
judgment is inappropriate. (See Def. Opp. 14). Specifically, Defendants
maintain that Mavis’s “decentralized hiring practice … diminishe[s] [the]
significance of any statistical shortfalls” (id. at 16); the “disagreement between
the parties’ respective experts” regarding whether to aggregate data before

13

Defendants half-heartedly argue that the EEOC failed adequately to plead a pattern-orpractice claim under Section 706. (See Def. Opp. 3 (arguing in passing that “Plaintiff[]
fail[ed] to plead a claim for pattern or practice of sex discrimination”)). Defendants are
mistaken. (See Compl. ¶ 8(b) (“Defendants’ hiring system disproportionately excludes
women from employment and constitutes a pattern or practice of discrimination against
women.”)). Regardless, even if the EEOC had failed to specify that it intended to utilize
a pattern-or-practice method of proof in prosecuting its case, it is not at all clear that its
pleading would have been deficient. See EEOC v. Port Auth. of N.Y. & N.J., 768 F.3d
247, 254 (2d Cir. 2014) (“[A] discrimination complaint need not allege facts establishing
each element of a prima facie case of discrimination to survive a motion to dismiss[.]”)
(citing Swierkiewicz v. Sorema N. A., 534 U.S. 506, 510-11 (2002) (“The prima facie case
under McDonnell Douglas … is an evidentiary standard, not a pleading requirement .…
This Court has never indicated that the requirements for establishing a prima facie case
under McDonnell Douglas also apply to the pleading standard that plaintiffs must
satisfy in order to survive a motion to dismiss.”)); Serrano II, 699 F.3d at 897
(“Swierkiewicz compels the conclusion that a plaintiff is not required to plead whether
she intends to employ the McDonnell Douglas or the Teamsters burden-shifting
evidentiary framework.”).

35

computing statistical significance “confirms that issues of fact … exist” (id. at
21); and anecdotal evidence of those who applied to Mavis “only highlight[s] the
multitude of disputed issues of fact” because at least some of the claimants
“candidly testified that they did not believe gender had anything to do with why
they were denied employment with Mavis until the EEOC contacted them” (id. at
23 (emphasis in original)). In reply, the EEOC argues that Defendants have
failed to rebut the EEOC’s prima facie case grounded in statistical evidence.
(Pl. Reply 8-9). Although this presents a very close case, the Court is
persuaded that Defendants have (albeit barely) met their burden of production
under the Teamsters framework. Summary judgment is therefore
inappropriate, and the issue of Defendants’ liability must go to trial.
As a preliminary matter, the Court notes that both sides would have
been better served making arguments more directly aimed at their respective
burdens under the Teamsters framework. Defendants, while understandably
trying to draw attention to the existence of genuine issues of material fact,
largely ignore the burden they face on summary judgment — namely that of
rebutting the EEOC’s prima facie case. The EEOC, while justifiably confident
that it has stated a prima facie case, fails to argue why precisely it believes that
Defendants have failed to meet their countervailing burden of production under
Teamsters.
Within the Teamsters framework, the EEOC has presented an
incontestably strong prima facie case. As with most pattern-or-practice cases,
the Court begins with an analysis of the EEOC’s statistical evidence of hiring

36

disparities. The Court is mindful that “[n]ot all disparities … are probative of
discrimination. Before a deviation from a predicted outcome can be considered
probative, the deviation must be ‘statistically significant.’” Ottaviani v. State
Univ. of N.Y. at New Paltz, 875 F.2d 365, 371 (2d Cir. 1989). “A finding of two
standard deviations corresponds approximately to a one-in-twenty, or five
percent, chance that a disparity is merely a random deviation from the norm,
and most social scientists accept two standard deviations as a threshold level
of ‘statistical significance.’” Id. While courts have rejected the finding of two
standard deviations as “a formal ‘litmus’ test for assessing the legitimacy of
Title VII claims[,] … [i]t is certainly true that a finding of two to three standard
deviations can be highly probative of discriminatory treatment.” Id. at 372
(citations omitted).14
Here, the EEOC has presented statistical evidence that demonstrates the
disparities in hiring female employees by Defendants is exceedingly unlikely to
have been caused by chance. Dr. Bendick frequently refers to the disparities
he identifies as “statistically significant,” and he is correct. In each of Dr.
Bendick’s analyses, for each of the four positions, he calculates a standard
deviation of two or more. When Dr. Bendick looked at Defendants’ incumbent
employee data (before Defendants had produced their hiring data), he reported

14

Care should be exercised, when looking at the number of standard deviations, not to
ascribe any overt significance to the value of the integer. As noted previously, “two
standard deviations corresponds roughly to a 1-in-20 chance that the outcome is a
random fluctuation. Three standard deviations corresponds to approximately a 1-in384 chance of randomness. Finally, a range of four to five standard deviations
corresponds to a probability range of 1 chance in 15,786 to 1 chance in 1,742,160.
Ottaviani v. State Univ. of N.Y. at New Paltz, 875 F.2d 365, 372 n.7 (2d Cir. 1989).

37

standard deviations of between 2.9 and 5.4; when he focused on their hiring
data, he found disparities even more severe, between 3.5 and 10.3; when he
limited his approach to consider the workforce only of “Tire Dealers,” as per Dr.
White’s preference, he still found standard deviations ranging from 2.8 to 5.7;
and, finally, when he considered Defendants’ actual applicant flow data, he
reported standard deviations ranging from 2.1 to 5.0. This statistical evidence,
which has been “finely tuned to the relevant labor pool,” would likely be
sufficient to establish a prima facie case without other evidence. Segar v.
Smith, 738 F.2d 1249, 1278 (D.C. Cir. 1984) (“When … statistical evidence
is … finely tuned to the relevant labor pool, gross disparities need not be shown
to permit an inference of discrimination.”); see also Hazelwood, 433 U.S. at
307-08 (“[G]ross statistical disparities … may in a proper case constitute prima
facie proof of a pattern or practice of discrimination.”); City of New York, 717
F.3d at 84 (“[I]nstances of discrimination against particular employees are
relevant to show a policy of intentional discrimination, [but] they are not
required; a statistical showing of disparate impact might suffice[.]”).
To the extent there was any doubt as to the sufficiency of the EEOC’s
prima facie case, however, the statistical analysis is buttressed by the
undisputed fact that Defendants hired zero female Store Managers, Mechanics,
or Technicians from 2008 to 2012, during which time they hired 80 male Store
Managers, 655 male Mechanics, and 1,688 male Technicians. “[F]ine tuning of
the statistics could not have obscured the glaring absence” of female employees
hired for these positions. Teamsters, 431 U.S. at 342 n.23; see, e.g., Victory v.

38

Hewlett-Packard Co., 34 F. Supp. 2d 809, 823-24 (E.D.N.Y. 1999) (finding that
plaintiff had raised “inference of gender discrimination” based in part on “the
troubling fact that nary a single female sales representative ha[d] ever been
promoted to a managerial position”).
The Court is hesitant to embrace fully the EEOC’s argument regarding
“the inexorable zero” of female hires. See Teamsters, 431 U.S. at 342 n.23
(“[T]he inference of discrimination came not from a misuse of statistics but
from ‘the inexorable zero.’”); e.g., Capaci v. Katz & Besthoff, Inc., 711 F.2d 647,
662 (5th Cir. 1983) (“To the noble theoretician predicting the collisions of
weightless elephants on frictionless roller skates, zero may be just another
integer, but to us it carries special significance in discerning … policies and
attitudes.”); United States v. City of New York, 713 F. Supp. 2d 300, 318
(S.D.N.Y. 2010) (“[A] court cannot help but be circumspect where a municipal
department in the country’s largest city repeatedly selects only applicants of
one sex for job vacancies — after all, zero is not just another number.” (internal
quotation marks and citation omitted)). Here, it is undisputed that Defendants
did hire one female Assistant Manager. However, the power of the raw
numbers (and the inference these numbers support) is dissipated only slightly,
if at all, by Defendants’ hiring of a single female Assistant Manager during a
span of five years. There is still, as the Court in Teamsters put it, “a glaring
absence” of female employees. See EEOC v. Andrew Corp., No. 81 Civ. 4359
(EEB), 1989 WL 32884, at *14 (N.D. Ill. Apr. 3, 1989) (noting that, despite the
occasional hiring of minority applicants, the data was nonetheless “dominated

39

by the ‘inexorable zero’ and [could not] be ‘explained away’”). After all, the
inference of discrimination stemming from Defendants’ hiring of zero female
Store Managers, Mechanics, and Technicians remains strong; notably, the
“inexorable zero” of female hires is paired with evidence — both in the
anecdotal accounts of claimants and in Defendants’ application records — of
actual female applicants for each of the positions. Cf. Ortiz-Del Valle v. Nat’l
Basketball Ass’n, 42 F. Supp. 2d 334, 338 n.1 (S.D.N.Y. 1999) (noting that the
inference created by the “inexorable zero” “may be weak [where there is] scant
evidence of other women who applied and were rejected”). Moreover, although
the EEOC cannot claim that Defendants failed to hire a single female Assistant
Manager during this time period, it is noteworthy that, looking at Defendants’
applicant flow data, the disparity in hiring of female Assistant Managers is the
most statistically significant among all four positions. (See Bendick Rebuttal
¶ 29 (reporting a standard deviation of 5.0)). Put another way, Defendants’
hiring of only one Assistant Manager is even more conspicuous (and even more
supportive of the inference of discrimination) than “the inexorable zero” found
in the positions of Store Managers, Mechanics, or Technicians.
Finally, the Court finds that the EEOC has bolstered its prima facie case
with anecdotal evidence that “provides ‘texture’ to the statistics,” Robinson, 267
F.3d at 168, and further lends credence to the inference of discrimination. For
example, Haywood was not hired for a position, even though three of her male
colleagues — individuals she had trained to perform work at another
automotive store — were. (See Haywood Dep. 132-33). Murnane similarly was

40

told by a Mavis employee, “don’t get your hopes up, you’re a chic[k]. These
guys are old school.” (Murnane Dep. 79). She was also told by a manager at
Mavis that she would not be hired because the “guys would stare” at her
during work. (Id. at 80). Other female applicants, such as Jackson and Braly,
were treated dismissively during the application process and never contacted.
(See Pl. 56.1 ¶¶ 125, 137). Taking the EEOC’s proffered statistical and
anecdotal evidence together, the Court finds no difficulty in finding that the
prima facie burden has been met.
Under Teamsters, the burden thus shifts to Defendants to “satisfy [their]
burden of production by ‘provid[ing] a nondiscriminatory explanation for the
apparently discriminatory result.’” City of New York, 717 F.3d at 90 (quoting
Teamsters, 431 U.S. at 360)). The Court must allow that Defendants, in part
because they fail to press the import of their burden as one of production under
the Teamsters framework, come perilously close to failing to rebut the prima
facie case to any degree. That Defendants’ burden is one of production — and
how courts have interpreted this burden in pattern-or-practice cases —
ultimately saves them from summary judgment.
“In the nature of things, the determination that a defendant has met its
burden of production (and has thus rebutted any legal presumption of
intentional discrimination) can involve no credibility assessment. For the
burden-of-production determination necessarily precedes the credibilityassessment stage.” St. Mary’s Honor Ctr., 509 U.S. at 509. “If the defendant
fails to rebut the plaintiff’s prima facie case, the presumption arising from an

41

unrebutted prima facie case entitles the plaintiff to prevail on the issue of
liability and proceed directly to the issue of appropriate relief.” Id. “An
employer facing [the] serious accusation [of intentional discrimination] must
have a broad opportunity to present in rebuttal any relevant evidence that
shows that it lacked such an intent.” City of New York, 717 F.3d at 87. That
said, “[a]t a minimum the employer must raise a genuine issue of fact as to the
veracity of plaintiffs’ proof, and ‘[t]o accomplish this, the defendant must
clearly set forth, through the introduction of admissible evidence, the reasons
for’ the observed disparity.” Segar, 738 F.2d at 1287 (quoting Burdine, 450
U.S. at 254-55)).
Defendants’ rebuttal consists of an alternative statistical analysis, an
attack on the probative value of the EEOC’s statistical analysis, and various
challenges to the anecdotal evidence. (See Def. Opp. 16-26). The “central issue
in the pending case is what showing an employer must make to satisfy its
burden of production in a pattern-or-practice case.” City of New York, 717
F.3d at 85. Conscious of the fact that the burden Defendants face now is not
the burden they will face at trial, the Court is compelled to deem Defendants’
rebuttal to be sufficient. See EEOC v. Joint Apprenticeship Comm. of Joint
Indus. Bd. of Elec. Indus., 186 F.3d 110, 122 (2d Cir. 1999) (warning, in a
disparate impact case, against “plac[ing] too high a burden on the defendant,
which bears only the burden of production, not of proof” and “infring[ing] upon
the province of the fact-finder, whose job it is to weigh the persuasiveness of
defendant’s evidence”); EEOC v. Sears, Roebuck & Co., 839 F.2d 302, 309 (7th

42

Cir. 1988) (“[An employer does] not have a burden of producing more
compelling evidence than … the EEOC.”); EEOC v. Freeman, 961 F. Supp. 2d
783, 799 (D. Md. 2013) (“The burden is not on Defendant to conduct its own
analysis to rebut the results produced by the EEOC’s flawed report. It is
sufficient for Defendant to point out the numerous fallacies in [the EEOC’s]
report, which raise the specter of unreliability.”), aff’d, 778 F.3d 463 (4th Cir.
2015).
Defendants have not provided the Court with “a nondiscriminatory
explanation for the apparently discriminatory result.” Teamsters, 431 U.S. at
360 n.46. Implicitly, perhaps, they argue that they always hire the most
qualified applicants, but they have not endeavored to show in any categorical
fashion that the female applicants rejected were less qualified than the males
they hired. Instead, they rebut the prima facie case by arguing that the
observed statistical disparities are not as statistically significant as they appear
to be. (See Def. Opp. 19). Indeed, Dr. White, who looked at the data on a yearby-year basis, found disparities generally in the one-to-three standard
deviation range. (See White Report 27-29).
For example, for the Store Manager position, Dr. White found a yearly
shortfall of between 0.82 and 2.08 female hires, resulting in a standard
deviation of between 0.95 and 1.51 per year. According to Dr. White’s
preferred methodology, analysis of Defendants’ hiring of female Store Managers
would fail to reach the widely-accepted threshold of a statistical significance
represented by two standard deviations. Although Dr. White’s contrary finding

43

would not be sufficient to defeat Plaintiff’s prima facie case — in the sense that
Defendants could not hope to prevail on a motion for summary judgment of
their own — the Court cannot say that it offers no rebuttal to Plaintiff’s prima
facie case.15 Indeed, given the reluctance of courts to imbue any particular
statistical threshold with outcome-determinative power in the context of
evaluating a plaintiff’s establishment of a prima facie case, see, e.g., Ottaviani,
875 F.2d at 373 (“[T]here simply is no minimum threshold level of statistical
significance which mandates a finding that Title VII plaintiffs have made out a
prima facie case[.]”), a court should be equally (if not more) circumspect in
letting a particular level of statistical significance decide, as a matter of law, the
critical issue of a defendant’s liability.
The EEOC objects to Defendants’ tack of raising an issue of fact by
presenting expert analysis that looks at the data in a different fashion, yet this
does not strike the Court as unusual. A battle of statisticians is to be expected
in a pattern-or-practice case. What the Court does find unusual is the EEOC’s
attempt to raise the specter of a Daubert motion now in its reply brief. (See Pl.
Reply 6-7 (citing Daubert v. Merrell Dow Pharmaceuticals, 509 U.S. 579

15

The precise weight of a defendant’s burden of production on rebuttal was a point of
dissension in the City of New York case, with the majority opting not to impose the
heavy burden that the dissent felt was warranted. Compare City of New York, 717 F.3d
at 107 (Pooler, C.J., dissenting) (“I think the City was required both under the law of
this Circuit and the Supreme Court to show more. At the outset, the majority
recognizes that ‘[i]n a pattern-or-practice case, the plaintiff’s initial burden is heavier,’
however, it fails to recognize that a similarly heavy burden also exists for the defendantemployer under Teamsters.”), and id. at 110 (“It is well recognized among other Circuits
that the employer must defeat the Government’s statistical proof.” (emphasis added),
with id. at 84 (majority opinion) (“‘[D]efeat’ might be thought to imply something
stronger that ‘rebut,’ but the [Supreme] Court’s language indicates that the Court
means the same thing in both contexts.”).

44

(1993))).16 The EEOC was aware, as of the filing of Dr. White’s report in
January 2014, that he approached the data in a manner different than Dr.
Bendick. The debate over the year-by-year analysis featured prominently in
both experts’ depositions, which took place in October 2014, well before the
parties submitted their motion papers. Yet neither party opted to challenge the
opposing expert’s approach as unreliable by filing a Daubert motion.17
Although the Court need not address this issue, because it was raised
only in reply, see, e.g., Evangelista v. Ashcroft, 359 F.3d 145, 156 n.4 (2d Cir.
2004), it will briefly respond here in recognition of its unique gatekeeping role
when it comes to evidentiary matters, see, e.g., Brenord v. Catholic Med. Ctr. of
Brooklyn & Queens, Inc., 133 F. Supp. 2d 179, 188 n.4 (E.D.N.Y. 2001)
(engaging in a sua sponte inquiry into qualifications of an expert under
Daubert) (collecting cases). Here, Dr. White gave explanations during his
deposition for why he thought a year-by-year breakdown was appropriate (viz.,
that hiring from past years influences hiring decisions of the next year). This
explanation is open to challenge (see Pl. Br. 15-16), but the Court cannot say
Dr. White’s opinions are inadmissible. “A minor flaw in an expert’s reasoning
or a slight modification of an otherwise reliable method” does not itself require
exclusion under Daubert; exclusion is only warranted “if the flaw is large

16

Although the EEOC challenged Dr. White’s approach in its opening brief (Pl. Br. 15-16),
it did not cite Daubert or suggest that Dr. White’s findings would be inadmissible.

17

Rule 4(G) of this Court’s Individual Rules of Practice in Civil Cases provide that all
“motions to exclude testimony of experts must be made by the deadline for dispositive
motions and should not be treated as motions in limine.” Accordingly, a Daubert motion
should have been filed concurrently with the EEOC’s motion for summary judgment.

45

enough that the expert lacks good grounds for his or her conclusions.”
Amorgianos v. Nat’l R.R. Passenger Corp., 303 F.3d 256, 267 (2d Cir. 2002)
(citation and quotation marks omitted). This is because “our adversary system
provides the necessary tools for challenging reliable, albeit debatable, expert
testimony.” Id. “[V]igorous cross-examination, presentation of contrary
evidence, and careful instruction on the burden of proof are the traditional and
appropriate means of attacking shaky but admissible evidence.” Id. (quoting
Daubert, 509 U.S. at 596).
Dr. White further testified that, in his experience, both aggregated and
year-by-year approaches are used in cases such as this. And that is precisely
what the jury will see and hear in this case, with each party presumably
attempting to undermine the other’s approach during cross-examination. The
jury will then decide whether to credit Dr. Bendick’s approach or Dr. White’s
approach. See Port Auth. Police Asian Jade Soc. of N.Y. & N.J. Inc. v. Port Auth.
of N.Y. & N.J., 681 F. Supp. 2d 456, 466-67 (S.D.N.Y. 2010) (“It is the jury’s
role to decide between competing conclusions based on the data and
analysis.”); Victory, 34 F. Supp. 2d at 824 (denying summary judgment in a
pattern-or-practice case with conflicting statistical evidence and noting that the
“[r]esolution of the battle of experts is a matter best suited for the trier of fact”).
Additionally, Defendants have challenged the probative nature of the
statistical evidence because of their “decentralized” hiring system. (Def.
Opp. 16). “Where uncoordinated and independent employment decisions are
made by different persons, statistics as to hiring by the overall entity may be

46

less significant in demonstrating bias than where a single office makes all
employment decisions.” Coser v. Moore, 739 F.2d 746, 750 (2d Cir. 1984). An
“assault on the … probative force” of statistical evidence is undoubtedly a valid
way for an employer to meet its burden of production on rebuttal. Again, the
Court cannot and need not say that a jury will accept this argument, and, if it
does, that this evidence can overcome the sheer power of the EEOC’s statistical
evidence. It matters not; Defendants’ burden at this stage is merely one of
production. See Dothard v. Rawlinson, 433 U.S. 321, 338-39 (1977)
(Rehnquist, J., concurring) (“If the defendants in a Title VII suit believe there to
be reason to discredit plaintiffs’ statistics that does not appear on their face,
the opportunity to challenge them is available to the defendants just as in any
other lawsuit. They may endeavor to impeach the reliability of the statistical
evidence, they may offer rebutting evidence, or they may disparage in
arguments or in briefs the probative weight which the plaintiffs’ evidence
should be accorded.”).
To that same end, Defendants offer evidence rebutting the EEOC’s
anecdotal evidence regarding individual applicants’ experiences. Although
neither side’s presentation of anecdotal evidence will be dispositive of classwide liability for a pattern or practice of discrimination, the Court cannot say
that Defendants failed to satisfy their burden of production with respect to
meeting anecdotal evidence the EEOC supplied. For all of these reasons, and
even while recognizing the strength of the EEOC’s case, the Court will deny its
motion for summary judgment on the pattern-or-practice claim.

47

3.

Summary Judgment on the EEOC’s Record-Keeping Claim Is
Denied

In addition to seeking summary judgment on its charge of discrimination
under Title VII, the EEOC also moves the Court to grant summary judgment on
its Title VII record-keeping claim. (See Pl. Br. 20-24). In sum, the EEOC
argues that there is no genuine issue of material fact remaining as to whether
Defendants retained personnel records (including applications for employment),
as required by Section 709(c) of Title VII, 42 U.S.C. § 2000e-8(c), and 29 C.F.R.
§ 1602.14. (Pl. Br. 20). Summary judgment is warranted, according to the
EEOC, because Defendants had no formal system for maintaining these
records, their employees conceded that not all applications were retained, and
several of the claimants’ applications could not be located. (Id. at 21). The
Court must disagree.
Section 709(c) provides that: “[e]very employer ... shall [i] make and keep
such records relevant to the determinations of whether unlawful employment
practices have been or are being committed, [ii] preserve such records for such
periods, and [iii] make such reports therefrom as the Commission shall
prescribe by regulation[.]” 42 U.S.C. § 2000e-8(c). In turn, the applicable
regulation requires that:
Any personnel or employment record made or kept by
an employer (including but not necessarily limited to …
application forms submitted by applicants and other
records having to do with hiring … ) shall be preserved
by the employer for a period of one year from the date
of the making of the record or the personnel action
involved, whichever occurs later …. Where a charge of
discrimination has been filed, … the respondent
employer shall preserve all personnel records relevant
48

to the charge or action until final disposition of the
charge or the action.
29 C.F.R. § 1602.14. No monetary sanctions are provided for violations of
Section 709(c) and 29 C.F.R. § 1602.14; the only relief available is equitable in
nature. See EEOC v. AutoZone, Inc., No. 06 Civ. 1767 (PCT) (PGR), 2008 WL
4418160, at *6 n.14 (D. Ariz. Sept. 29, 2008).
Defendants’ retention procedures are not immune from criticism.
However, the Court cannot say, as a matter of law, that Defendants failed to
retain applications in violation of Section 709(c). Mavis retained more than
19,400 pages of applications and resumes, and produced these records to the
EEOC. (See Def. Opp. 9). See EEOC v. LA Weight Loss, 509 F. Supp. 2d 527,
540 (D. Md. 2007) (summary judgment was inappropriate where the employer
“produce[d] evidence demonstrating that it … maintained a policy of retaining
documents, resumes, and interview notes …, and … attempted to make
applicant data from internet job-posting services available to the EEOC”).
While the claimants’ testimony regarding the number of times they applied is
relevant circumstantial evidence, and might convince a jury that Mavis’s
retention must have been imperfect, this testimony is not uncontroverted.
Defendants claim that there was a retention policy in place, and that despite
being unwritten, this policy worked. (See Def. 56.1 Response ¶¶ 184-200). Cf.
AutoZone, 2008 WL 4418160, at *6 (finding summary judgment appropriate
where the employer “specifically admit[ted] that the specified documents …
existed and that they related to [the claimant], that the specified documents
had been in the possession of [the employer], and that it ha[d] not produced to
49

the EEOC any of the specified documents”). They also claim that they did not
selectively delete or throw away applications from certain candidates. (See
Andre 30(b)(6) Dep. 353-54 (“I can tell you what didn’t happen. We didn’t
cherry pick them and throw them away.”)). Cf. EEOC v. Oak-Rite Mfg. Corp.,
No. 99 Civ. 1962 (DFH), 2001 WL 1168156, at *17 (S.D. Ind. Aug. 27, 2001)
(awarding summary judgment on record-keeping claim where employer
“testified that applications of rejected applicants were being thrown away”).
Although the jury may well conclude, after hearing from witnesses, that one
side offers more credible evidence, the Court is not permitted to make such a
credibility determination on summary judgment. Genuine issues of material
fact remain.
Having determined that this issue too will go to trial, the Court need not
reach the EEOC’s request for a finding of spoliation. However, given the
extraordinary remedy the EEOC proposes — namely, “that the Court … issue
an adverse inference instruction to the jury instructing that all damages it
awards in the form of out-of-pocket losses (lost wages and benefits) and
interest, and punitive damages, should be based on the shortfall of women in
the positions at issue [as identified by the EEOC’s expert] and not just on the
number of claimants” (Pl. Opp. 27) — the issue of spoliation merits some
discussion.
In the context of a Section 709(c) claim, in order to make a finding of
spoliation, the Court does not need evidence of the employer’s “bad faith”; still,
there must be “evidence of intentional destruction.” Byrnie v. Town of

50

Cromwell, Bd. of Educ., 243 F.3d 93, 109 (2d Cir. 2001); accord Zimmermann v.
Assocs. First Capital Corp., 251 F.3d 376, 383 (2d Cir. 2001). Here, at the
summary judgment stage, issues of fact remain as to whether Defendants
intentionally destroyed any applications. (See Def. Opp. 30-33 (noting that
some applications may have been submitted before Defendants’ duty to retain
was triggered and arguing that there was inconsistent testimony from the
claimants as to the number of applications they submitted)). Cf. Byrnie, 243
F.3d at 109 (finding spoliation where the employer “admitted its policy to
destroy … records soon after the hiring process had been completed and that
the records in this case were so destroyed” and “[a]t no point … asserted that
their destruction was merely accidental”).
Perhaps more significant than the request for a finding of spoliation is
the EEOC’s request that this Court instruct the jury to award monetary
damages (both compensatory and punitive) based on a hypothetical number of
female applicants whose applications may have been discarded. Although the
Court is generally aware that the EEOC takes an aggressive posture toward
vindicating the rights of discrimination claimants, the Court is not aware of a
similar attempt to, essentially, shoehorn monetary damages into Section
709(c), which provides only for injunctive relief, or otherwise artificially
augment the monetary damages allocated to victims of discrimination through
an adverse inference. At the risk of hyperbole, it appears to the Court that the
EEOC has attempted to shoot the moon here. While the Court can, after a
finding of intentional destruction of documents relevant to discrimination

51

plaintiffs’ claims, imagine an adverse inference instruction that permitted a
jury to infer that lost documents would have further supported the liability of
an employer for discriminatory practices, see, e.g., Byrnie, 243 F.3d at 110
(finding “enough circumstantial evidence exists to permit a reasonable trier of
fact to conclude that the destroyed documents would show unlawful
discrimination”), the Court cannot imagine (and the EEOC has not cited) a case
in which a jury is instructed to award damages, both compensatory and
punitive, based on the inference that additional claimants exist.
While the EEOC twice refers to this as a “modest remedy” (Pl. Br. 27, 28),
the Court finds the proposed remedy to be ill-conceived from both a practical
and constitutional standpoint. The proposed “compensatory” damages based
on hypothetical applicants either would not be compensatory, leading to a
windfall for the claimants, or would necessitate a “post-trial process whereby
EEOC would be allowed to identify additional class members to share in any
recovery.” (Pl. Br. 27 n.5). Both scenarios are unpalatable, and the latter
scenario is sketched out in such minimal detail that the fraught logistics barely
emerge. For example, it is entirely unclear to the Court whether and how
Defendants would be permitted to rebut the inference of discrimination and
challenge the individualized relief for each new claimant that emerges posttrial. Yet, under the Teamsters framework, they surely have that right.
The proposed “punitive” damages to be awarded as a result of such an
adverse inference instruction under the EEOC’s proposal fare no better: at the
time of the award, such damages would be entirely untethered to the harm

52

suffered by any individual and would be presumptively unreasonable. See
Johnson v. Nextel Commc’ns Inc., 780 F.3d 128, 148 (2d Cir. 2015) (“[P]unitive
damages awards must be tethered to compensatory damages in order to
comply with due process[.]”); In re Simon II Litig., 407 F.3d 125, 139 (2d Cir.
2005) (noting that “[h]armful behavior that is not ‘correlatable’ with class
members and the harm or potential harm to them [is] precluded” from the
punitive damages calculus).
The motion for summary judgment on the record-keeping violation, and
the concomitant request for a finding of spoliation, must be denied.
4.

The EEOC’s Motion for Bifurcation of the Trial Is Granted in
Part and Denied in Part

Finally, as the Court has denied the EEOC’s motion for summary
judgment, the Court must determine whether and how to bifurcate the trial in
this action.18 Specifically, if the trial is to be bifurcated, the Court must decide
whether punitive damages will be addressed during Phase I or Phase II.
Under Title VII, a jury may award punitive damages if it finds that the
employer “engaged in a discriminatory practice or discriminatory practices with
malice or with reckless indifference to the federally protected rights of an
aggrieved individual.” 42 U.S.C. § 1981a(b)(1). For an employer of Mavis’s size,
“[t]he sum of the amount of compensatory damages … and the amount of

18

According to the EEOC, “th[e] motion regarding a bifurcated proceeding and punitive
damages would [have] become moot if the EEOC’s summary judgment motion on
pattern-or-practice liability [were] granted.” (Pl. Br. 29 n.6).

53

punitive damages awarded … [can]not exceed, for each complaining
party … $300,000.” Id. § 1981a(b)(3)(D).
The EEOC argues that, “[b]ecause of the overlap in the evidence to be
presented to establish liability based on a pattern or practice of discrimination
and to determine class-wide punitive damages, the issue of punitive damages
should be addressed during [Phase] I.” (Pl. Br. 29). Defendants counter that
“punitive damages — like compensatory damages — are an individual, not a
class-wide remedy” and thus should be reserved for Phase II. (Def. Opp. 9).
Moreover, Defendants argue that awarding punitive damages before
compensatory damages are determined is bound to run afoul of the rule that
punitive damages awarded must be “reasonable and proportionate to the
amount of harm to the [claimant] and to the general damages recovered.” (Def.
Opp. 10 (quoting State Farm Mut. Auto. Ins. Co. v. Campbell, 538 U.S. 408, 426
(2003))). The EEOC responds to this argument by noting that the statutory
caps provided by Title VII preclude an unreasonable punitive damages award.
(See Pl. Br. 32-33). Having considered the parties’ arguments on this issue, the
Court agrees with Defendants that the issue of punitive damages should be
decided during Phase II.
Taking the easiest (inasmuch as it is uncontested) part of this motion
first, the Court finds that bifurcation of the liability and remedial portions of
the trial is appropriate. Although Teamsters does not explicitly require a
bifurcation of trials into two phases, it certainly suggests this approach.
Teamsters, 431 U.S. at 361 (“[A] district court must usually conduct additional

54

proceedings after the liability phase of the trial to determine the scope of
individual relief.”); see also Wal-Mart, 131 S. Ct. at 2561 (“[The] procedure for
trying pattern-or-practice cases that gives effect to these statutory
requirements” includes a second “phase, [in which the burden of proof
… shift[s] to the company”); Brown v. Nucor Corp., 576 F.3d 149, 159 (4th Cir.
2009) (“Bifurcation of Title VII class action proceedings for hearings on liability
and damages is now commonplace.” (citation omitted)); Robinson, 267 F.3d at
168 (“[L]itigating the pattern-or-practice liability phase [first] reduce[s] the
range of issues in dispute and promote[s] judicial economy.”). Accordingly,
Phase I of the trial in this action will concern liability, and Phase II will concern
individual relief.
Turning to the issue of when to address punitive damages, the Court is
mindful that “there are procedural and substantive constitutional limitations
on [punitive damages] awards,” and that “[t]he Due Process Clause of the
Fourteenth Amendment prohibits the imposition of grossly excessive or
arbitrary” awards. State Farm, 538 U.S. at 416. Moreover, “courts must
ensure that the measure of [punitive damages] is both reasonable and
proportionate to the amount of harm to the plaintiff and to the general
damages recovered.” Id. at 420. However, “[c]ourts are divided … on
whether … the issue of punitive damages, can be determined on a class-wide
basis in the first phase” of a bifurcated pattern-or-practice trial. PMT Corp.,
2015 WL 5026029, at *4 (collecting cases). On this issue, the Court finds
instructive the Second Circuit’s treatment of class-wide punitive damages

55

awards outside the context of pattern-or-practice claims, specifically in In re
Simon II Litigation, 407 F.3d 125 (2d Cir. 2005), and in Johnson v. Nextel
Communications Inc., 780 F.3d 128 (2d Cir. 2015).
In Simon II, the Second Circuit expressed concern with the district
court’s class certification order, which sought “an assessment of punitive
damages prior to an actual determination and award of compensatory
damages.” 407 F.3d at 138. The Court’s concern, animated by the Supreme
Court’s decision in State Farm, was that the anticipated trial plan “would fail to
ensure that a jury will be able to assess an award that, in the first instance,
will bear a sufficient nexus to the actual and potential harm to the plaintiff
class, and that will be reasonable and proportionate to those harms.” Id. As
the Second Circuit admonished, in assessing punitive damages, the “conduct
relevant to the reprehensibility analysis must have a nexus to the specific harm
suffered by the plaintiff, and [can]not be independent of or dissimilar to the
conduct that harms the plaintiff.” Id. at 139 (citing State Farm, 538 U.S. at
422-23)).
In Johnson, the Second Circuit addressed the prudence of having a
punitive damages “ratio” established in an early phase of trial, which would
then be applied in a subsequent phase. Because the Court vacated the district
court’s certification of the class, it did “not decide the constitutional question
whether such a punitive damages ratio [would be] inconsistent with due
process.” 780 F.3d at 148-49. However, precisely because the Circuit

56

“remand[ed] the case for further proceedings,” it also “tarr[ied] to express
concern about the district court’s adoption of [the] trial plan.” Id. at 148.
The trial plan in that case was somewhat more complicated than the
plan proposed by the EEOC here. Specifically, it called for three phases:
In Phase I, a jury would decide the common liability
issues. In Phase II, the same jury would determine the
compensatory damages for the named plaintiffs. The
jury would then determine if the class was entitled to
punitive damages, and, if so, it would determine a
punitive-to-compensatory damages “ratio” based on
defendants’ conduct toward the entire class. Finally, in
Phase III, different juries would conduct abbreviated
individualized damages trials for the class members, in
which individualized defenses would be considered.
After the Phase III jury determined the amount of
compensatory damages for each individual class
member, the court would apply the punitive-tocompensatory damages ratio set by the Phase II jury
and award the resulting punitive damages, if any, while
retaining the discretion to make an independent
assessment of whether the total award was
inappropriate for any particular class member.
Johnson, 780 F.3d at 136. In offering guidance to the district court on remand,
the Second Circuit expressed concern that the “Phase II jury would lack any
conception of the actual damages to the members of the class to whom the
ratio would subsequently be applied.” Id. at 149. This proposed “one-size-fitsall punitive damages ratio,” the Court explained, “would be no more tethered to
compensatory damages than the lump sum [of punitive damages] disapproved
of in Simon II.” Id.
Here, by contrast, the EEOC has suggested “an aggregate [punitive
damages] award that … would [be] distribute[d] to those in the class who are
entitled to relief on a pro rata basis.” (Nov. 25 Tr. 8). In other words, as in
57

Simon II, a lump-sum punitive damages award would be decided “prior to an
actual determination and award of compensatory damages.” 407 F.3d at 138.
This proposal, however, engenders the same concerns articulated in Simon II
and Johnson. The EEOC’s proposed Phase I jury would be wholly ignorant as
to the actual damages suffered by any of the individual claimants, a scenario
that the Second Circuit has disfavored. See Simon II, 407 F.3d at 139 (rejecting
lump-sum punitive damages award because the “conduct relevant to the
reprehensibility analysis must have a nexus to the specific harm suffered by
the plaintiff, and … it [can]not be independent of or dissimilar to the conduct
that harms the plaintiff”). Accordingly, the Court is persuaded that the
prudent course is to address the issue of punitive damages during Phase II.
See David v. Signal Int’l, LLC, 37 F. Supp. 3d 814, 821-22 (E.D. La. 2013)
(“After a jury determines whether Signal engaged in a pattern or practice of
discrimination in Phase I, Phase II will address issues concerning each
individual employees’ claim for relief, including … compensatory and punitive
damages.”).19
Even accepting the EEOC’s argument that the $300,000 cap in damages
prevents any individual claimant’s punitive damages award from being

19

The Court also notes that, although Johnson did not involve a pattern-or-practice claim,
the Second Circuit cited the Fifth Circuit’s decision in Allison v. Citgo Petroleum Corp.,
151 F.3d 402 (5th Cir. 1998), which held, in a pattern-or-practice case, that “punitive
damages must be determined after proof of liability to individual plaintiffs at the second
stage ..., not upon the mere finding of general liability to the class at the first stage.”
Johnson v. Nextel Commc’ns Inc., 780 F.3d 128, 149 n.26 (2d Cir. 2015) (quoting Allison,
151 F.3d at 418). Coupled with its aversion in Johnson and Simon II to classwide
punitive damages awards, the Second Circuit’s tacit approval of Allison provides added
comfort to the Court that punitive damages should be determined during Phase II of a
pattern-or-practice trial.

58

constitutionally offensive, the Court disagrees with the foundational premise of
the EEOC’s argument for Phase I determination of punitive damages. That is,
the EEOC presumes, as do Defendants, that there will be two juries. But the
EEOC’s entire argument as to why punitive damages should be tried during
Phase I falls apart if the very same jury hears Phase II. (See Pl. Opp. 30
(arguing that “the jury hearing the evidence and making the finding on
systemic discrimination” should decide the “entitlement to punitive damages”)).
See EEOC v. New Indianapolis Hotels, LLC, No. 10 Civ. 1234 (WTL) (DKL), 2012
WL 364052, at *2 (S.D. Ind. Feb. 1, 2012) (“The Court notes that the use of one
jury also addresses the EEOC’s arguments for determining punitive damages in
phase one. The EEOC correctly points out that, if there are two juries, deciding
punitive damages in the second phase would be unduly wasteful: the parties
would have to represent evidence presented at the liability phase again in order
to assist the trier in making its punitive damages determination. Yet if there is
one jury, the parties need not revisit the evidence produced in the first
phase.”).
Although the Court teed this issue up during the pre-motion conference,
explaining to the EEOC that the “argument about judicial economy works
better if there is, in fact, an entirely separate jury … impaneled for purposes of
[Phase] II” (Nov. 25 Tr. 5), neither party has provided a satisfactory explanation
as to why two juries would be necessary for this case. As the EEOC conceded
during the pre-motion conference, while “[t]here usually is a gap [between
phases in pattern-or-practice cases], … that doesn’t preclude the Court from

59

proceeding” directly into Phase II after a Phase I liability verdict in favor of the
EEOC. (Id. at 4). For their part, Defendants simply argued that there should
“be a separate proceeding” because, after Phase I, “you then have a finding,
[and] there then is … preparation” required “to show in each of th[e]
[claimant’s] cases that there were reasons other than this pattern or practice
against hiring women that resulted in their nonhiring.” (Id. at 13-14).
However, despite the interest the Court expressed on this topic during
the pre-motion conference, the parties have not cited cases or pointed to
specific, practical reasons that explain why this particular action must be tried
to two separate juries. (Cf. Nov. 25 Tr. 13 (“In some … cases, your Honor, there
are things that are done in between [Phase I and Phase II]. There are some
things that weren’t done here[.]”)). Unlike pattern-or-practice cases that
proceed under a bifurcated discovery plan, with a pattern-or-practice liability
trial coming at the close of liability discovery and before discovery regarding
individualized relief, discovery in this action is closed. That is, costs associated
with Phase II discovery are sunk, and witnesses have already been
inconvenienced and deposed. Similarly, in some cases the number of
claimants or class members is so large as to necessitate two (or more) separate
proceedings. See, e.g., Robinson, 267 F.3d at 169 n.13 (“[G]iven the [1,300]
members in the putative class, the district court is likely to try the remedial
phase of each class member’s claim before a separate jury from the one that
considers the liability phase[.]”). Not so here, where there is a relatively modest

60

number of claimants. (See Pl. Opp. 27 n.5 (noting that there are “15 testifying
class members”)).
It is decidedly not the Court’s place to conjure up reasons why two juries
need to hear this case. It is enough for now to say that the case will be tried in
two phases. See New Indianapolis Hotels, 2012 WL 364052, at *2 (“[T]rial of
the EEOC’s claims will proceed before one jury with a liability phase and if
necessary, a second, damages phase following immediately on the heels of the
first phase.”); Rosen v. Reckitt & Colman, Inc., No. 91 Civ. 1675 (LMM), 1994
WL 652534, at *5 (S.D.N.Y. Nov. 17, 1994) (“If by bifurcation the Plaintiff
means separate trials, with distinct juries, the Court would not expect
significant economies of scale …. The Court will instead conduct a single trial,
in two stages, which addresses Plaintiffs’ fear of repetitively presenting their
pattern and practice evidence, and Defendants’ concern of an unfair
presumption resulting from evidence that a second jury has not had an
opportunity to hear.”). Although the Court will not foreclose the parties from
moving in limine to have Phase I and Phase II tried to two separate juries, the
Court will not endorse such a bifurcation absent good reason. See State Farm,
538 U.S. at 416 (noting that “compensatory and punitive damages … [are]
usually awarded at the same time by the same decisionmaker”).20

20

Because the Court is of the mind that this case can be tried to a single jury, it does not
reach the EEOC’s alternative argument that a Phase I jury should decide entitlement to
punitive damages, while a Phase II jury decides the amount of punitive damages. (See
Pl. Br. 31 n.7 (citing, inter alia, Pitre, 908 F. Supp. 2d at 1178)). The Court notes,
however, that the approach to bifurcation adopted in Pitre has not been universally
accepted. See, e.g., EEOC v. Performance Food Grp., Inc., 16 F. Supp. 3d 576, 583 (D.
Md. 2014) (“A Pitre Phase II jury bound by a Pitre Phase I jury’s determination that [the
employer] acted with malice or reckless indifference in regard to the individual that the

61

CONCLUSION
For the foregoing reasons, the EEOC’s motion for summary judgment on
its pattern-or-practice claim and its record-keeping claim is DENIED. The
EEOC’s motion for a bifurcated trial is GRANTED IN PART and DENIED IN
PART. Trial shall proceed in two phases, with punitive damages to be awarded
(if at all) during the second phase.
The parties shall appear for a pretrial conference on October 2, 2015, at
11:00 a.m. in Courtroom 618 of the Thurgood Marshall Courthouse, 40 Foley
Square, New York, New York. By Thursday of the week prior to that
conference, the parties shall file a joint letter, not to exceed three pages, raising
any issues the parties would like to address at the pretrial conference and
providing any information that the parties believe may assist the Court in
advancing the case to settlement or trial.

Pitre Phase II jury was considering would not be basing its exercise of discretion upon
its own finding of malice or reckless indifference. Nor would the Pitre Phase II jury be
able to consider, in regard to its discretionary decision, the degree of malice or reckless
indifference pertinent to the individual issue. Hence, there would, inevitably, be a
serious issue whether, under the Pitre procedure, [the employer] would be deprived of
its Seventh Amendment right to trial by jury.”); see generally Johnson, 780 F.3d at 150
n.27 (“To the extent the trial plan here contemplates that the Phase III jury (or juries) or
the district court would have to reassess the reprehensibility of Nextel’s conduct in
order to determine the appropriateness of a punitive damages award, such a plan could
run afoul of the Seventh Amendment.”). But see EEOC v. Cintas Corp., No. 04 Civ.
40132 (SFC), 2015 WL 1954476, at *6 (E.D. Mich. Apr. 29, 2015) (“[W]hile the
availability of punitive damages shall be adjudicated at the Phase I Trial, any
determination of the aggregate amount and individual distribution of punitive damages
shall be reserved for the Phase II Trial.”); Ellis v. Costco Wholesale Corp., 285 F.R.D.
492, 543 (N.D. Cal. 2012) (“[T]rying these potentially overlapping issues of liability and
entitlement to punitive damages before a single jury ensures compliance with the
Seventh Amendment’s prohibition on reexamination. Conversely, separating all
classwide liability determinations (of liability and availability of punitive damages) from
any quantification of damages would preserve separate questions for separate fact
finders and is thus consistent with the Seventh Amendment.” (collecting cases)).

62

SO ORDERED.
Dated:

September 11, 2015
New York, New York

__________________________________
KATHERINE POLK FAILLA
United States District Judge

63