United States District Court, D. Colorado
OPINION AND ORDER ON MOTIONS FOR SUMMARY JUDGMENT AND
MOTIONS TO RESTRICT
S. KRIEGER, CHIEF UNITED STATES DISTRICT JUDGE
MATTER comes before the Court on cross Motions for Summary
Judgment: Defendants Columbine Health Systems
(“Columbine”) and The Worthington, Inc. d/b/a New
Mercer Commons Assisted Living Facility's (“New
Mercer, ” and collectively with Columbine,
“Defendants”) Motion for Summary Judgment on the
Only Claim Involving Marlene Hoem (#63); and
Plaintiff Equal Employment Opportunity Commission's
(“EEOC”) Motion for Partial Summary Judgment
(#64). Responses (##76, 77)
and Replies (##82, 83) were filed to both
motions. Also before the Court are three unopposed Motions to
Restrict (##71, 81, 86), which were filed by
EEOC asserts claims under Title VII of the Civil Rights Act
of 1964 (42 U.S.C. § 2000e et seq.). Federal
question jurisdiction exists pursuant to 28 U.S.C. §
following is a summary of the relevant facts viewed most
favorably to the non-movants. More detail will be provided as
needed in the Court's analysis.
manages several senior citizen facilities, including New
Mercer, an assisted living facility in Fort Collins,
Colorado. The parties dispute whether Columbine and New
Mercer are part of a single integrated enterprise, which -
for the reasons explained below - is not an issue that is
necessary to resolve in conjunction with this motion.
However, in construing the facts most favorably to the
non-movant, the Court will refer to Columbine and New Mercer
to 2008, the Defendants employed four Black as Personal Care
Providers (“PCPs”) (the claimants) at New Mercer.
All came from Africa. Kiros Aregahgn is from Ethiopia, and
Mohamed Osman Mahgoub, Sawson Ibrahim, and Hanaa Gual are
from Sudan. Marlene Hoem, a Caucasian, was their supervisor.
mid-2008, the Defendants hired Paula Lewis to oversee New
Mercer. Shortly after Ms. Lewis was hired, she had a
conversation with Ms. Hoem about the claimants. According to
Ms. Hoem, Ms. Lewis told her that New Mercer “had to
get rid of ‘these people, ' because they just
can't speak English.” Ms. Hoem believed these
comments were discriminatory and expressed her disagreement
with them to Ms. Lewis. In September 2008, Ms. Lewis
suggested that Ms. Hoem demote one of the African employees,
but Ms. Hoem refused to do so. Defendants terminated Ms.
Hoem's employment a week later, citing her failure to
comply with this directive as one - although not the only -
reason for terminating her employment.
early 2009, Defendants imposed a new requirement that PCPs
complete a training course and pass a written examination
(the “PCP Exam”). The training course and
examination were conducted in English. The four African
claimants completed the course, but each received a score
below 75 percent on the examination. As a result in May 2009,
the Defendants terminated their employment.
claimants timely filed EEOC complaints, the EEOC conducted an
investigation and brought this lawsuit. Its Amended Complaint
(#18) asserts three claims: (1) unlawful
discrimination by disparate treatment based on race and/or
national origin; (2) unlawful discrimination by disparate
impact based on race and/or national origin; and (3) unlawful
retaliation in termination of Ms. Hoem's employment. Both
parties filed motions for partial summary judgment.
Standard of Review
of the Federal Rules of Civil Procedure facilitates the entry
of a judgment only if no trial is necessary. See White v.
York Intern. Corp., 45 F.3d 357, 360 (10th Cir. 1995). A
trial is required if there are material factual disputes to
resolve, thus entry of summary judgment is authorized only
“when there is no genuine dispute as to any material
fact and the movant is entitled to judgment as a matter of
law.” Fed.R.Civ.P. 56(a); Savant Homes, Inc. v.
Collins, 809 F.3d 1133, 1137 (10th Cir. 2016).
is material if, under the substantive law, it is an essential
element of the claim. See Anderson v. Liberty Lobby,
Inc., 477 U.S. 242, 248 (1986). A dispute is genuine if
the conflicting evidence would enable a rational trier of
fact to resolve the dispute for either party. Becker v.
Bateman, 709 F.3d 1019, 1022 (10th Cir. 2013).
law governs which facts are material and what issues must be
determined. It also specifies the elements that must be
proved for a given claim or defense, sets the standard of
proof, and identifies the party with the burden of proof.
See Anderson v. Liberty Lobby, Inc., 477 U.S. 242,
248 (1986); Kaiser-Francis Oil Co. v. Producer's Gas
Co., 870 F.2d 563, 565 (10th Cir. 1989). A factual
dispute is “genuine” if the evidence presented in
support of and in opposition to the motion is so
contradictory that, if presented at trial, a judgment could
enter for either party. See Anderson, 477 U.S. at
248. When considering a summary judgment motion, a court
views all evidence in the light most favorable to the
non-moving party, thereby favoring the right to a trial.
See Tabor v. Hilti, Inc., 703 F.3d 1206, 1215 (10th
Cir. 2013); Garrett v. Hewlett Packard Co., 305 F.3d
1210, 1213 (10th Cir. 2002).
movant has the burden of proof on a claim or defense, the
movant must establish every element of its claim or defense
by sufficient, competent evidence. See Fed. R. Civ.
P. 56(c). Once the moving party has met its burden to
establish a genuine dispute, the responding party must
present competent and contradictory evidence as to a material
fact. See Celotex Corp. v. Catrett, 477 U.S. 317,
322-23 (1986); Perry v. Woodward, 199 F.3d 1126,
1131 (10th Cir. 1999); Bacchus Indus., Inc. v. Arvin
Indus., Inc., 939 F.2d 887, 891 (10th Cir. 1991).
the moving party does not have the burden of proof on the
pertinent issue, it may point to an absence of sufficient
evidence to establish a claim or defense that the non-movant
is obligated to prove. Once the movant has done so, the
respondent must come forward with sufficient competent
evidence to establish a prima facie claim or defense
to justify a trial. If the respondent fails to produce
sufficient competent evidence to establish its claim or
defense, the claim or defense will be dismissed as a matter
of law. See Celotex, 477 U.S. at 322-23.
case involves motions for summary judgment filed by both
sides. Because the determination of whether there is a
genuine dispute as to a material factual issue turns upon
which party has the burden of proof and whether adequate
evidence has been submitted to support a prima facie
case or establish a genuine dispute as to material fact, each
motion is evaluated independently. Atl. Richfield Co. v.
Farm Credit Bank of Wichita, 226 F.3d 1138, 1148 (10th
Cir. 2000); Buell Cabinet Co. v. Sudduth, 608 F.2d
431, 433 (10th Cir. 1979); In re Ribozyme Pharms., Inc.,
Sec. Litig., 209 F.Supp.2d 1106, 1112 (D. Colo. 2002).
EEOC's Motion for Partial Summary
EEOC seeks summary judgment on 1) its claim for unlawful
discrimination based on disparate impact; 2) on a discrete
issue - whether Columbine and New Mercer are part of an
“integrated enterprise” with over 500 employees;
and 3) on four of Defendants' affirmative defenses. The
Court addresses each in turn.
Disparate Impact Claim
VII forbids not only intentional racial discrimination but
also “practices that are fair in form, but
discriminatory in operation, ” which is often described
as “disparate impact” discrimination. Hilti,
Inc., 703 F.3d at 1220 (citing Lewis v. City of
Chicago, 560 U.S. 205, 211 (2010)); see 42
U.S.C. § 2000e-2(k). The disparate impact theory seeks
to remove employment obstacles which create “built-in
headwinds and freeze out protected groups from job
opportunities and advancement, ” unless those obstacles
are required by business necessity. Hilti, 703 F.3d
at 1220 (citing EEOC v. Joe's Stone Crab, Inc.,
220 F.3d 123, 1274 (11th Cir. 2000)).
establish a prima facie claim for disparate impact,
a plaintiff must come forward with evidence that shows that
(i) an employer's employment practice (ii) disparately
impacted a protected group of employees. Hilti, 703
F.3d at 1220. If the plaintiff makes this prima
facie showing, the burden shifts to the defendant to
demonstrate that the challenged practice is “job
related for the position in question and consistent with
business necessity.” 42 U.S.C. §
2000e-2(k)(1)(A)(i); Hilti, 703 F.3d at 1220. If
such showing is made, the burden returns to the plaintiff to
demonstrate that an available alternative employment practice
would serve the employer's legitimate business needs and
cause a less severe disparate impact. Ricci v.
DeStefano, 557 U.S. 557, 578 (2009); see 42
U.S.C. § 2000e-2(k)(1)(A)(ii).
the first element, there is no dispute that Columbine and/or
New Mercer first required passage of the PCP examination as a
condition of employment. This undoubtedly qualifies as an
employment practice. The question becomes whether there is
evidence that it had a disparate impact on a protected group.
facie showing of disparate impact is
“essentially a threshold showing of a significant
statistical disparity… and nothing more.”
Ricci, 557 U.S. at 587; Carpenter v. Boeing
Co., 456 F.3d 1183, 1196 (10th Cir. 2006). Although no
specific mathematical formulation is required, statistical
disparities between non-protected and protected groups must
be substantial to raise an inference of causation. Watson
v. Fort Worth Bank and Trust, 487 U.S. 977, 994-95
(1988). To determine whether the Plaintiffs' statistical
evidence is sufficient, the Court considers three factors:
(1) the size of the disparity between the pass/fail rates of
different groups of test takers; (2) the statistical
significance of the disparity; and (3) whether the
statistical evidence isolates the challenged employment
practice as the cause. Hilti, 703 F.3d at 1222.
size of the disparity between (a) the employees of the
protected group enjoying a job or job benefit; and (b) the
total composition of the employees enjoying that job or
benefit must be significant. Carpenter, 456 F.3d at
1193, 1202. The EEOC guidelines suggest that a disparity of
20 percent or more in selection rate will be considered
evidence of adverse impact in a disparate impact claim. 29
C.F.R. § 1607.4(D). Although not controlling, this
guideline often acts as a general rule of thumb. See
Hilti, 703 F.3d at 1223; see also Watson, 487
U.S. at 995.
second factor, statistical significance, measures the
likelihood that the disparity between the groups is random.
Hilti, 703 F.3d at 1223. Statistical significance is
expressed in terms of standard deviations. The Supreme Court
has recognized that a disparity of more than two or three
standard deviations makes it unlikely that the disparity
occurred randomly. Id. (quoting Hazelwood Sch.
Dist. v. United States, 433 U.S. 299, 308 n.14 (1977))
third factor is whether the statistical evidence isolates the
specific employment practice as the cause of the disparity.
Watson, 487 U.S. at 994. A plaintiff must show that
the challenged practice results in the disparate impact by
eliminating factors other than the challenged practice that
might contribute to the disparity. See Wards Cove Packing
Co. v. Atonio, 490 U.S. 642, 657 (1989). This becomes
necessary when the employer uses the challenged practice
along with subjective criteria to make the employment
decision. Hilti, 703 F.3d at 1224.
Defendants contend that the EEOC has not come forward with
sufficient statistical evidence to show that the PCP Exam had
a disparate impact on employees of a protected race
(Black/African-American) or of a protected national origin
(African). Under the circumstances of this case, it
is not necessary to distinguish between race and national
origin because all four claimants are both Black and of
African descent. The EEOC's statistical evidence includes
test results for the four claimants and six other individuals
identified as African and/or Black by Dr. Thomas Haladyna in
his report. A spreadsheet shows the PCP Exam results for all
African exam-takers. Four out of seven exam-takers of African
origin, and four out of nine exam-takers who are Black,
failed the PCP Exam. In comparison, only one of 138
exam-takers who were White, Hispanic, or belonged to other
races failed the PCP Exam.
Haladyna initially opined that the selection rate for African
exam-takers (which he referred to as the “boundary
rule”) was less than four-fifths (or eighty percent) of
the rate for all exam-takers. In a supplemental report he
specified a pass rate of 42.8 percent for African
exam-takers, 55.6 percent for Black exam-takers, 99.3 percent
for White exam-takers, and 100 percent for Hispanic
exam-takers. Dr. Haladyna found the disparity in results of
6.78 standard deviations between the groups to be
as true, this evidence is sufficient to establish a prima
facie disparity in test results for the protected
groups. The Defendants contend, however, that Dr.
Haladyna's conclusions are unreliable because that the
statistical sample used was too small. For this proposition,
they rely on the opinion of Dr. Harpe.
Harpe does not contradict Dr. Haladyna's mathematical
findings. She does not contest that, in terms of the size of
the disparity, the statistical data fails the four-fifths
guideline nor that Dr. Haladyna's miscalculated the
standard deviation. Instead, Dr. Harpe criticizes the small
sample size for Black or African PCPs, as well as Dr.
Haladyna's decision to include or exclude various
test-takers from the analysis. For example, Dr. Harpe notes
that 39 test-takers were excluded from analysis because not
enough was known about their race or ethnicity; she asserts
that given the small sample size of the “African”
subgroup, if any of those excluded individuals did fall
within that category, they could significantly affect the
prominently, Dr. Harpe invokes the so-called
“flip-flop” rule, which is federal regulatory
guidance promulgated (in part) by the EEOC that purportedly
holds that in situations involving extremely small sample
sizes, if a single hypothetical individual is subtracted from
the group with the higher selection rate and added to the
group with the lower rate, and the hypothetical recalculated
standard deviation results in the reversal of an adverse
impact determination (i.e., it is less than two or
three standard deviation), there is a relatively high
likelihood that the difference in selection rates is a random
one. See: Howe v. City of Akron, 789 F.Supp.2d 786,
801 (N.D. Ohio 2010); see also Bazile v. City of
Houston, 858 F.Supp.2d 718, 739-40 (S.D. Tex. 2012). Her
analysis under this principle essentially consists of two
Dr. Harpe opines that the scores of two of the African
exam-takers, Mr. Mahgoub and Ms. Ibrahim, should be excluded.
She relies on a memo drafted by Penny Rubala, Defendants'
Director of Clinical Education, in which Ms. Rubala recorded
that “during all three days of PCP class [including the
quiz] Mohamed [Mahgoub] and Sawsan [Ibrahim] were cheating.
This was demonstrated by [Ibrahim] holding her hand up by her
face and then looking at her husband[']s quiz as he was
completing answers and further by [Mr. Mahgoub] whispering
under his breath in his native language to [Ms.
Ibrahim].” Therefore, Dr. Harpe removes their scores
from the data and makes calculations as to the size of the
disparity and the statistical significance based on these
removals from the populations in question (i.e., two
failures out of five African, and seven Black, PCPs). Those
calculations show standard deviations between two and three
for both groups - in other words, right on the borderline of
Dr. Harpe invokes the flip-flop rule to show that if one
removes a single failure and adds a single hypothetical
individual to the passing category, and recalculates the
standard deviations, those standard deviations drop below two
for both the African and Black categories. As such, according
to her opinion, the sample size is too small for Dr.
Haladyna's statistical results to be statistically
significant (or at least statistically useful). The EEOC does
not address the flip-flop rule specifically but instead
broadly contends that Dr. Haladyna's methodology is
perfectly appropriate, and that courts commonly accept
similarly small population sizes when undertaking the
disparate impact analysis. The EEOC cites a number of cases