United States District Court, D. Maryland
[Copyrighted Material Omitted]
[Copyrighted Material Omitted]
[Copyrighted Material Omitted]
EEOC, Plaintiff: Debra Michele Lawrence, Ronald Lynn
Phillips, LEAD ATTORNEY, Maria Kate Boehringer, U.S. Equal
Employment Opportunity Commission, Baltimore, MD; Keyana
Capri Laws, Melanie Marie Peterson, Philip Matthew Kovnat,
U.S. Equal Employment Opportunity Commission, Philadelphia,
Freeman, Defendant: W Randolph Teslik, LEAD ATTORNEY, Donald
R Livingston, John T Koerner, Akin Gump Strauss Hauer and
Feld LLP, Washington, DC.
U.S. Office of Personnel Management (" OPM" ),
Intervenor: Alex S Gordon, LEAD ATTORNEY, Office of the
United States Attorney, Baltimore, MD.
W. TITUS, UNITED STATES DISTRICT JUDGE.
poker expert Kenny Rogers once sagely advised, "
You've got to to know when to hold 'em. Know when to
fold 'em. Know when to walk away."  In the Title
VII context, the plaintiff who wishes to avoid paying a
defendant's attorneys' fees must fold 'em once
its case becomes so groundless that continuing to litigate is
unreasonable, i.e. once it is clear it cannot have a winning
hand. In this case, once Defendant Freeman revealed the
inexplicably shoddy work of the EEOC's expert witness in
its motion to exclude that expert, it was obvious Freeman
held a royal flush, while the EEOC held nothing. Yet, instead
of folding, the EEOC went all in and defended its expert
through extensive briefing in this Court and on appeal. Like
the unwise gambler, it did so at its peril. Because the EEOC
insisted on playing a hand it could not win, it is liable for
Freeman's reasonable attorneys' fees.
as a regular part of its hiring process, conducted criminal
background checks on all applicants who were offered a
position, and conducted credit background checks on
applicants who were offered financially sensitive
positions. EEOC v. Freeman, 961
F.Supp.2d 783, 787 (D. Md. 2013). Importantly, applicants
were not turned away for any negative information.
Rather, Freeman limited in scope the type of negative
information that would disqualify an applicant. For example,
as to the criminal background check, Freeman generally did
not consider arrests, but only convictions that had occurred
within the past seven years. Id. at 788.
Furthermore, Freeman did not consider all
convictions, but only those for certain crimes. Id.
Similarly, with regard to credit checks, only certain
negative items would exclude an applicant from being hired.
Id. at 789. The Equal Employment Opportunity
Commission (the " EEOC" ) alleged that
Freeman's use of background checks was discriminatory.
Id. at 790.
being rejected by Freeman based on information in her credit
report, Katrina Vaughn filed a charge of discrimination with
the EEOC. Id. at 789. Based on this charge, the EEOC
filed a complaint in this Court alleging that Freeman's
use of background checks had a disparate impact on
African-American, Hispanic, and male job applicants.
Id. On April 27, 2010, the Court dismissed all
claims relating to hiring decisions made before March 23,
2007, the date 300 days before the original charge of
discrimination was filed on January 17, 2008. ECF No. 19. On
January 31, 2011, the Court granted Freeman's partial
motion for summary judgment and dismissed all claims relating
to hiring decisions based on criminal background checks made
prior to November 30, 2007, the date 300 days before
September 25, 2008, when the EEOC first notified Freeman that
it was expanding its investigation to include race
discrimination based on the use of criminal background
checks. ECF No. 43. On August 24, 2012, the EEOC voluntary
dismissed with prejudice its claim that Freeman unlawfully
discriminated against Hispanics. ECF No. 95.
18, 2012, the EEOC served Freeman with the expert report of
Dr. Kevin R. Murphy. ECF No. 186 at 9. After Freeman pointed
out various errors in the report, the EEOC served a second
expert report that purported to fix those errors on July 26,
2012. ECF No. 186-17. On December 18, 2012,
Freeman moved to exclude Dr. Murphy's report and
the report of a second expert, Dr. Beth Huebner, arguing that
a significant number of errors and omissions within the
reports made them unreliable, and thus
inadmissible. ECF No. 108. On December 21, 2012,
Freeman moved for summary judgment, arguing that, without
reliable statistical evidence of a disparate impact, the EEOC
could not make out a prima facie case of
discrimination. ECF No. 114. The EEOC responded to each
motion. ECF Nos. 121, 126. In its opposition to the motion to
exclude the expert reports, the EEOC submitted a supplemental
expert report from Murphy that purported to moot
Freeman's criticisms. ECF No. 121 at 7. The EEOC also
contended that Freeman was the source of any errors in the
expert report. Id. at 13. The EEOC also proffered
yet another report by Murphy at the motions hearing on June
19, 2013. Freeman, 961 F.Supp.2d at 797.
August 9, 2013, the Court excluded the EEOC's experts and
granted Freeman's motion for summary judgment. ECF No.
150. Freeman subsequently moved for attorneys' fees. ECF
No. 154. After filing its appeal, ECF No. 166, the EEOC moved
to stay proceedings related to Freeman's motion for
attorneys' fees. ECF No. 172. The Court granted that
motion over Freeman's opposition. ECF No. 176. The EEOC
also moved to amend the record for appeal with Murphy's
report tendered at the motions hearing, ECF No. 177, and the
Court granted that motion, again over Freeman's
opposition. ECF No. 181.
Fourth Circuit affirmed the Court on February 20, 2015.
EEOC v. Freeman, 778 F.3d 463 (4th Cir. 2015).
Freeman submitted a renewed motion for attorneys' fees
that included those fees incurred on appeal, ECF No. 186, and
has further supplemented its request to account for fees
incurred defending its fee request. ECF No. 197.
EEOC has, understandably, taken a keen interest in
employers' use of background checks to make hiring
decisions. " Because of the higher rate of
incarceration of African-Americans than Caucasians,
indiscriminate use of criminal history information might have
the predictable result of excluding African-Americans at a
higher rate than Caucasians." Freeman, 961
F.Supp.2d at 786. Similarly, because blacks typically fare
worse in many indicators of economic health, such as income
and poverty rate, than whites, credit checks may
disproportionately exclude blacks. The EEOC is certainly
entitled to attempt to police the use of background checks
through litigation, and to attempt to use litigation to
challenge whether an employer's use of background checks
is " job-related for the position in question and
consistent with business necessity." Id. at
before the EEOC can get to the question of business
necessity, it must actually make out a prima facie
case of disparate impact, and before it can make out a
prima facie case of disparate impact, it must
actually produce reliable statistical evidence showing that a
particular employment practice has a disparate impact on a
protected class. Id. at 791. The EEOC is
not entitled to require an
employer to demonstrate business necessity when it is
consistently unable to produce any reliable evidence of a
disparate impact. See EEOC v. Propak Logistics,
Inc., 746 F.3d 145, 157 (4th Cir. 2014) (" The
reference to statutory goals and missions, however, cannot be
divorced from the manner in which those purposes are
implemented." ) (Wilkinson, J., concurring). Continuing
to litigate once the faults in the EEOC's case were laid
bare was unreasonable, and Freeman is entitled to reasonable
attorneys' fees from that point forward.
The Availability of Attorneys' Fees Under Title
VII provides that " the court, in its discretion, may
allow the prevailing party...a reasonable attorney's fee
(including expert fees) as part of the costs, and the
Commission and the United States shall be liable for costs
the same as a private person." 42 U.S.C. §
2000e-5(k). By its terms, this provision allows either a
prevailing plaintiff or a prevailing defendant to recover
attorneys' fees. However, the award of attorneys'
fees to a prevailing plaintiff involves different
considerations than an award to a prevailing defendant. The
prevailing plaintiff is acting as a " private attorney
general" in vindicating an important federal interest
against a violator of federal law, and therefore "
ordinarily is to be awarded attorney's fees in all but
special circumstances." Christiansburg Garment Co.
v. EEOC, 434 U.S. 412, 416-17, 98 S.Ct. 694, 54 L.Ed.2d
648 (1978). The opposite is true of a prevailing defendant. A
prevailing defendant not only is not vindicating any
important federal interest, but the award of attorneys'
fees to prevailing defendants as a matter of course would
undermine that interest by making it riskier for "
private attorney generals" to bring
claims. Id. at 422. Accordingly,
before a prevailing defendant may be awarded fees, it must
demonstrate that a plaintiff's claim was "
frivolous, unreasonable, or groundless, or that the plaintiff
continued to litigate after it clearly became so."
Id. Importantly, however, this standard does not
require a plaintiff to have acted in bad faith. Id.
at 421; see also Arnold v. Burger King
Corp., 719 F.2d 63, 66 (4th Cir. 1983) (" A
plaintiff acting in good faith may nevertheless be assessed
fees if his claim is groundless or frivolous" ).
assessing whether to assess attorneys' fees against a
plaintiff, it is important that a court not engage in minute,
post hoc questioning of a plaintiff's claims and
conduct in light of plaintiff's failure on the merits.
EEOC v. Great Steaks, Inc., 667 F.3d 510, 517 (4th
Cir. 2012). However, there is " neither a precise test
to be used, nor a specific quantum of proof required, in
determining whether a plaintiff's claim was
unreasonable." Propak Logistics, 746 F.3d at
151 . A decision to award fees is committed to the discretion
of the Court, which is in the best position to assess the
considerations relevant to the conduct of litigation.
Freeman is Entitled to Reasonable Attorneys'
The EEOC Needed to Present Reliable Statistical Evidence that
Freeman's Policies had a Disparate Impact to Make Out
Facie Case of
black letter Title VII law that, in order to make out a
prima facie case of
disparate impact, a plaintiff must present " reliable
and accurate statistical analysis performed by a qualified
expert." Freeman, 961 F.Supp.2d at 786. This
evidence is necessary because a disparate impact claim
alleges that an otherwise neutral policy has a negative
effect on the ability of a protected group to get hired.
Accordingly, there must be a statistical disparity
" between the number of protected class members in the
qualified applicant group and those in the relevant segment
of the work-force." Id. at 791.
reciting this black letter law meets several of the
EEOC's arguments in its opposition to an award of
attorneys' fees. First, the EEOC asserts that its case is
not factually groundless. ECF No. 191 at 5. The basis for
this argument appears to be that it may have been
possible for the EEOC to show evidence of a disparate impact,
and it just so happened to fail here. Id. at 5-7.
But the possible existence of evidence somewhere to back up a
claim does not constitute reasonable grounds for continuing
to litigate that claim. Cases are litigated based on evidence
actually produced, not hunches, however reasonable those
hunches may be. Without reliable statistical evidence
produced by an expert, which is entirely absent here, the
EEOC simply could not make out its prima facie case.
The factual lack of reliable expert testimony showing a
disparate impact was fatal to the EEOC's disparate impact
claim in this particular case, even if in some
alternate universe (or simply by hiring a better expert), the
EEOC could have made out a claim.
EEOC also could not reasonably rely on general population
statistics to make out its prima facie case. It is
true that, in some circumstances, general population
statistics will suffice to show that a particular policy has
a disparate impact. See Griggs v. Duke Power
Co., 401 U.S. 424, 430-32, 91 S.Ct. 849, 28 L.Ed.2d 158
(1971). However, an examination of Griggs, the first
Title VII disparate impact case, puts the folly to the
EEOC's argument that it could have done so here. In
Griggs, an employer had a facially neutral
requirement that an applicant be a high school graduate.
Id. at 427-28. The Supreme Court reversed the
district court's dismissal of the complaint, reasoning
that, despite the lack of discriminatory intent, the fact
that African-Americans had lower high school graduation rates
in the employer's state meant that the policy had a
discriminatory impact, and that Title VII allowed such
claims. Id. at 429-32. The difference between
Griggs and this case is obvious. In Griggs,
the challenged policy, a high school graduation requirement,
and the general population statistic, high school graduation
rates, were the exact same. Griggs simply
does not apply to a case where the challenged policy is not
congruent with the general population statistic being offered
as proof of a disparate impact.
addition to Griggs, the EEOC cites several cases for
the proposition that " [l]ower federal courts have also
accepted external population proof in a variety of contexts
despite arguments that they did not perfectly match the facts
of the case." ECF No. 191 at 10. An examination of each
of these cases reveals that none supports this argument, even
Peightal v. Metropolitan Dade County, 26 F.3d 1545
(11th Cir. 1994), a white applicant challenged, on equal
protection grounds, a fire department's affirmative
action policies. The Eleventh Circuit noted that, for a
government employer to show a compelling interest in
combating discrimination that would justify using affirmative
action, " statistical comparisons between the
employer's work force and the composition of the relevant
population are probative of a pattern of
discrimination." Id. at 1553.
The Eleventh Circuit did allow the defendant fire department
to present statistical disparities between the general
relevant population of Hispanics and the number of Hispanics
who passed a qualifications exam, notwithstanding that those
statistics did not account for supposed disparities in
graduation rates for Hispanics, graduation from high school
being a prerequisite to take the exam. Id. at 1556
n.17. It accepted this statistic, however, because (a) it
determined that the relevant statistical refinement was not
available, as there was no data on Hispanic high school
graduation rates in the area, and (b) the plaintiff had not
shown beyond speculation that it would matter. Id.
Moreover, the Eleventh Circuit also noted the numerous
refinements the fire department had made to ensure
that the external population statistics reasonably matched
the relevant applicant pool. Id. at 1553-57. In
fact, the case had previously been remanded for those
refinements precisely because of " the relative
undesirability of general population figures."
Id. at 1554.
Carpenter v. Boeing Co., 456 F.3d 1183, 1197 (10th
Cir. 2006), the Tenth Circuit noted that " the
population selected for statistical analysis need not
perfectly match the pool of qualified persons." However,
the EEOC could not have reasonably relied on that case.
First, the plaintiff's expert in that case did attempt to
set out a specific methodology for constructing a
representative sample which, as explained below, the
EEOC's expert here never did. Id. at 1194-95
(describing expert's process for defining similarly
situated employees and measuring the appropriate statistics
to determine disparate impact). Moreover, in that case, the
Tenth Circuit still determined that the expert's
conclusions were unreliable. Id. at 1198-99. It
is curious that the EEOC would rely on a case that is
directly contrary to its position to support the notion that
its position was legally reasonable.
cases cited by the EEOC to support the reasonableness of its
reliance on general population statistics are simply
inapplicable. They support the proposition that general
population statistics will be allowed where those statistics
are completely congruent with an employment criteria, as in
Griggs, or where a limited number of data points, in
an otherwise refined sample, do not perfectly align with the
general population data. The cases do not support the
proposition that a plaintiff is entitled to rely on general
population statistics divorced from any reference to the
defendant's employment practices without any attempt to
control for those practices.
been clear since at least 1989, when the Supreme Court
decided Wards Cove Packing Co., Inc. v. Atonio, 490
U.S. 642, 109 S.Ct. 2115, 104 L.Ed.2d 733 (1989), that
refined, statistical comparisons are necessary to make out a
disparate impact claim. Id. at 650-51. The
EEOC's attempt to use general population statistics
illustrates the rationale underlying Wards Cove,
which is that an employer should not be held liable for
disparities that are not caused by its employment practices.
Id. at 651 (refusing to hold employer liable for
disparities caused by " reasons that are not [the
employer's] fault" ). The EEOC focused on general
background statistics such as arrest and incarceration rates,
and the racial disparity in those with " good" or
" bad" credit histories. E.g. ECF No. 121
at 15-16; ECF No. 186-14 at 23. But Freeman was not generally
concerned about arrests at all, and was not concerned about
all convictions, or about whether those convictions led to
incarceration. Instead, Freeman was concerned about only
particular types of convictions. Similarly, Freeman was not
concerned about whether an applicant had " good" or
" bad" credit. Rather, Freeman was concerned about
only certain indicators of bad credit. The EEOC failed to
produce any general population statistics that approximated
the criteria used by Freeman to make hiring decisions. The
use of general population statistics also failed to take into
account the subjective component of Freeman's policy,
which gave employees who failed a background check an
opportunity to explain the circumstances before Freeman
denied them a position. ECF No. 108-4 at 15. And, indeed, the
EEOC should have been aware that any reliance on general
population statistics would be flawed, since population
statistics demonstrated adverse outcomes for Hispanics
generally, yet its expert failed to find a statistically
significant disparate impact on Hispanics from Freeman's
specific use of background checks. Freeman, 961
F.Supp.2d at 798-99.
general population statistics were appropriately tailored in
this case, the EEOC overlooks a key issue, which is noted in
the cases cited by the EEOC, yet ignored by it: in order to
use general population statistics that do not match the
particular employment criteria instead of statistics from the
employer, the missing statistics must be impossible or
extremely difficult to obtain. See
Carpenter, 456 F.3d at 1199 (noting, in excluding
statistics that were insufficiently tailored, that "
Plaintiffs have not established that the data necessary to
establish the impact on CBA-qualified workers were
unavailable" ); Peightal, 26 F.3d at 1556 n.17
(noting that district court properly included
undifferentiated statistics where " no differentiated
population statistics exist based upon the level of education
of minorities" ); see also Wards Cove,
490 U.S. at 651 (" Alternatively, in cases where such
labor market statistics will be difficiult if not
impossible to ascertain, we have recognized that certain
other statistics...are equally probative for this
it was not impossible for the EEOC to obtain the relevant
data from Freeman that it needed to do the appropriate
statistical analysis. The data necessary to determine the
impact of Freeman's use of background checks was
available, because it was provided by Freeman in discovery.
Freeman, 961 F.Supp.2d at 793 (noting that Freeman
" clearly demonstrated that it properly supplied the
EEOC during discovery with complete background check logs for
the entire period covered by the complaint." );
Freeman, 778 F.3d at 469 (Agee, J., concurring)
(same). Accordingly, the EEOC could not have reasonably hoped
to rely on general population statistics in the event that
its flawed expert reports were excluded, because it could not
have possibly demonstrated that the relevant data was
upshot, then, is that the EEOC could not reasonably pursue
its claims unless
it could produce an arguably reliable expert report with
statistics showing a disparate impact caused by Freeman's
particular employment practices. As explained in the next
section, the EEOC wholly failed to do this.
It was Unreasonable for the EEOC to Continue to Litigate on
the Basis of Obviously Flawed Expert Reports
clear in the opinions of this Court and the Fourth Circuit,
Murphy's report was severely flawed. His conclusions were
drawn from an unrepresentative sample, even though data was
available from which a representative sample could be drawn.
Freeman, 961 F.Supp.2d at 793-94. He excluded data
from much of the relevant time period, and excluded data from
most of Freeman's branches. Id. at 794-96. He
cherry-picked data. Id. at 795. Worse yet , his
reports contained a " mind-boggling number of
errors," most of which were introduced by Murphy.
Id. at 796. The question is not whether Murphy's
reports were flawed, because they clearly were. The question
is if they were so flawed that it was unreasonable for the
EEOC to litigate in reliance on them.
The EEOC should have been aware of the unreliability of
Murphy's analysis no later than after ...