Not what you're
looking for? Try an advanced search.
Buy This Entire Record For
MEACHAM v. KNOLLS ATOMIC POWER LAB.
February 13, 2002
CLIFFORD B. MEACHAM; THEDRICK L. EIGHMIE; AND ALLEN G. SWEET, ALL INDIVIDUALLY AND ON BEHALF OF ALL OTHER PERSONS SIMILARLY SITUATED; AND JAMES R. QUINN, PHD; DEBORAH L. BUSH; RAYMOND E. ADAMS; WALLACE ARNOLD; WILLIAM F. CHABOT; ALLEN E. CROMER; PAUL M. GUNDERSEN; CLIFFORD J. LEVENDUSKY; BRUCE E. PALMATIER; NEIL R. PAREENE; WILLIAM C. REYNHEER; JOHN K. STANNARD; DAVID W. TOWNSEND; AND CARL T. WOODMAN, PLAINTIFFS,
KNOLLS ATOMIC POWER LABORATORY, AKA KAPL, INC.; LOCKHEED MARTIN, INC.; AND JOHN J. FREEH, INDIVIDUALLY AND AS AN EMPLOYEE OF KAPL AND LOCKHEED MARTIN, DEFENDANTS. JAMES R. QUINN, PHD, PLAINTIFF, V. KNOLLS ATOMIC POWER LABORATORY, INC., AKA KAPL, INC.; LOCKHEED MARTIN, INC.; AND JOHN J. FREEH, INDIVIDUALLY AND AS AN EMPLOYEE OF KAPL AND LOCKHEED MARTIN, DEFENDANTS.
The opinion of the court was delivered by: David R. Homer, United States Magistrate Judge.
MEMORANDUM-DECISION AND ORDER
The twenty-six plaintiffs, all over age forty, were employed by
defendant Knolls Atomic Power Laboratory, Inc. ("KAPL") until 1996 when
their employment was terminated in an involuntary reduction-in-force
(IRIF). Plaintiffs then commenced this class action*fn1 against KAPL,
Lockheed Martin, Inc., KAPL's parent company, and John J. Freeh
("Freeh"), KAPL's President and General Manager, alleging that their
terminations violated the Age Discrimination in Employment Act (ADEA),
29 U.S.C. § 621 et seq., and the New York Human Rights Law (HRL),
N.Y. Exec. Law § 290 et seq. (McKinney 2001).*fn2 Plaintiffs allege
that the violations resulted from both disparate treatment and disparate
impact. Following trial, a jury returned verdicts in favor of the
plaintiffs on their disparate impact theory, in favor of defendants on the
disparate treatment theory, and awarded damages totaling $5,077,285.33.
Judgment was entered on the verdicts. Am. J. (Docket No. 133).
Presently pending are (1) defendants' motion for judgment as a matter
of law, a new trial and remittitur (Docket No. 137; (2) plaintiffs'
motion for prejudgment interest, postjudgment interest and an upward
adjustment of damages for increased taxes (Docket No. 138); and (3)
plaintiffs' motion for attorneys' fees and costs (Docket No. 197). For the
reasons which follow, defendants' motion for judgment as a matter of law
is denied and their motion for a new trial and remittitur is granted in
part and denied in part. Plaintiffs' motion for prejudgment interest is
granted in part and denied in part, their motion for postjudgment
interest is granted and their motion for an upward adjustment of damages
for increased taxes is denied. Finally, plaintiffs' motion for attorneys'
fees and costs is granted.
KAPL operates a laboratory to design, build and test prototype naval
nuclear reactors and to train United States Navy personnel in their
operation and maintenance. KAPL employs approximately 2,700 individuals
principally at sites in Niskayuna and West Milton, New York, both within
forty miles of Albany. KAPL operates the laboratory under a "cost-plus"
contract between the United States Department of Energy (DOE) and
Lockheed Martin by which KAPL receives reimbursement for expenses it
incurs plus a percentage of those expenses. The contract is overseen by
DOE's Schenectady Naval Reactors Office (SNR).
The twenty-six prevailing plaintiffs were employed in exempt, or
salaried, positions in various of KAPL's sixteen sections.*fn3 The
sixteen sections were headed by managers who reported directly to Freeh.
Each section was also subdivided into sections headed by subsection
managers. Certain subsections were further subdivided into units headed
by unit managers.
In March 1994, KAPL began formulating a plan to reduce its workforce by
approximately 140 positions in accordance with budget projections
coordinated between KAPL and SNR. KAPL conceived and adopted a "Workforce
Adjustment Plan" (WAP) to achieve the desired reduction in two stages.
The first was the "Voluntary Separation Plan" (VSP) by which KAPL offered
early retirement to employees with at least twenty years of service in
return for the payment of $20,000. The second was the IRIF. In the Fall
of 1995, KAPL offered the VSP to its employees. Ultimately, 107 employees
applied and were approved for the VSP.
In November 1995, KAPL proceeded with the IRIF.*fn4 First, KAPL
identified the positions within the sixteen sections which were to be
eliminated. This "excess skills analysis" sought to determine those
positions which KAPL could terminate with the least adverse effect on
KAPL's operations. These included, for example, specialized positions in
reduced demand. After the employees in those excess positions were
identified, the section, subsection and unit managers were directed to
evaluate the individuals in those positions by a prescribed process which
resulted in the assignment of a numerical score for each employee subject
to the IRIF.
First, the managers assigned scores of from zero to ten points to each
such employee in four categories: length of service, work performance,
criticality to KAPL's operations, and the flexibility of their talents.
For length of service, each employee subject to the IRIF was given one
point for every two years of service to a maximum of ten points.*fn5 For
work performance the subject employees were assigned scores based on
their previous two performance evaluations.*fn6 For criticality and
flexibility, subject employees were assigned ratings points in each
category by their managers.
At the time of the IRIF, KAPL employed 2,063 individuals in exempt
positions after the VSP. Of those, 1,203, or fifty-eight percent, were
forty years of age or older [hereinafter "older employees"]. Of those,
245 were considered for the IRIF. Of the 245, 179, or seventy-three
percent, were older employees. Of the thirty-one exempt employees selected
for the IRIF, thirty, or ninety-seven percent, were older employees.*fn9
The terminated employees received certain benefits, including severance
totaling one week of pay for every year of service to a maximum of
twenty-six weeks; continuation of benefits for one year; and the free use
and services of an employment counseling service outside KAPL.
Plaintiffs commenced this action on January 6, 1997.*fn10 Defendants
KAPL and Lockheed Martin were named in the ADEA cause of action and all
three defendants were named in the HRL cause of action. Plaintiffs
alleged two theories of liability under both the ADEA and the HRL
— disparate treatment and disparate impact. Compl. (Docket No. 1).
Defendants' motion to bifurcate the trial between liability and damages
was granted, Docket No. 59, and the trial on liability commenced before a
jury on June 20, 2000. On July 26, 2000, the jury returned a verdict
finding under plaintiffs' disparate impact theory that the defendants had
discriminated against the twenty-six exempt employees and that such
discrimination had been willful. The jury found in favor of defendants on
plaintiffs' disparate treatment theory of discrimination. Docket No. 86.
Prior to the commencement of the damages phase of the trial, eight
plaintiffs*fn11 settled their claims with defendants. See Docket Nos.
111-17 & 121. The trial on damages commenced before the same jury on
September 19, 2000. Evidence was presented in four groups of plaintiffs
with four or five plaintiffs in each group and with the jury returning a
separate verdict for each group. The verdict for the final group was
returned on November 22, 2000. The jury's verdicts on damages awarded the
remaining eighteen prevailing plaintiffs total damages as follows:
Raymond E. Adams $411,823.13
Wallace Arnold $526,825.81
Deborah L. Bush $246,509.50
William R. Chabot $119,734.85
Allen E. Cromer $109,386.97
Thedrick L. Eighmie $91,445.00
Paul M. Gundersen $160,218.71
Clifford J. Levendusky $190,605.55
Clifford B. Meacham $125,000.00
Bruce E. Palmatier $332,126.09
Christine A. Palmer*fn12 $78,812.50
Neil R. Pareene $283,783.95
James R. Quinn $68,921.04
William C. Reynheer $1,115,357.58
John K. Stannard $258,189.40
Allen G. Sweet $128,953.46
David W. Townsend $397,482.91
Carl T. Woodman $432,108.88
Docket Nos. 119, 124, 126 & 128. Judgment was entered on the verdicts.
Docket No. 133. These motions followed.
Defendants move for an order (1) granting judgment as a matter of law
pursuant to Fed.R.Civ.P. 50(b), or in the alternative (2) granting a new
trial pursuant to Fed. R. Civ. P. 59, or (3) granting remittitur on
certain of plaintiffs' damage awards.
A motion for judgment as a matter of law under Rule 50 should be
granted when "there is no legally sufficient evidentiary basis for a
reasonable jury to find for [the moving] party on that issue."
Fed.R.Civ.P. 50(a)(1). The standard under Rule 50 "mirrors" that for a
motion for summary judgment under Fed.R.Civ.P. 56. Anderson v. Liberty
Lobby, Inc., 477 U.S. 242, 250-51 (1986). On a motion under Rule 50, a
court must consider all evidence in the record and not simply the
evidence favorable to the nonmovant. Reeves v. Sanderson Plumbing
Products, Inc., 530 U.S. 133, 149 (2000); Tolbert v. Queens Coll.,
242 F.3d 58, 70 (2d Cir. 2001). "In doing so, however, the court must
draw all reasonable inference in favor of the nonmoving party, and it may
not make credibility determinations or weigh the evidence." Reeves, 530
U.S. at 149 (citations omitted). Thus, in reviewing the entire record, a
court should consider only that evidence favorable to the nonmoving party
and any evidence supporting the moving party which is uncontradicted and
The standard for granting a new trial under Rule 59 is less demanding.
"[U]nlike a motion for judgment as a matter of law, a trial judge
considering a motion for a new trial is free to weigh the evidence
himself and need not view it in the light most favorable to the verdict
winner." United States v. Landau, 155 F.3d 93, 104 (2d Cir. 1998)
(internal quotations and citation omitted); see also Funk v. F & K
Supply, Inc., 43 F. Supp.2d 205, 224 (N.D.N.Y. 1999 (McAvoy, J.). Thus, "
`a motion for a new trial may be granted even if there is substantial
evidence to support the jury's verdict.'" Caruolo v. John Crane, Inc.,
226 F.3d 46, 54 (2d Cir. 2000) (quoting Landau, 155 F.3d at 104). A court
should grant a new trial if "convinced that the jury has reached a
seriously erroneous result or that the verdict is a miscarriage of
justice." Caruolo, 226 F.3d at 54 (internal quotations and citation
A motion for judgment as a matter of law and for a new trial following
a jury verdict must both be filed within ten days of the entry of
judgment. Fed.R.Civ.P. 50(b), 59(b). The original judgment in this case
was filed on December 5, 2000. Docket No. 131.*fn14 Thus, excluding
Saturdays, Sundays and legal holidays as required by Fed.R.Civ.P. 6(a),
defendants' motion was required to be filed on or before December 19,
2000. On December
19, 2000, defendants filed the instant notice of motion
and a supporting affidavit. Docket No. 137. In an order filed the same
day, defendants were granted until February 1, 2001 to file and serve
supplemental pleadings in support of their motion. Docket No. 136.
Defendants timely filed a memorandum of law in support of their motion on
February 1, 2001. Docket No. 196. As a threshold matter, plaintiffs now
contend that defendants' motion is untimely because their memorandum of
law was not filed within the ten days required by the rules. Pls. Mem. of
Law (Docket No. 203) at 2-6.
The ten day time limit for this motion is jurisdictional and could not
be extended. See Lichtenberg v. Besicorp Group, Inc., 204 F.3d 397, 401
(2d Cir. 2000) (describing time limit as "uncompromisable"); Weissman v.
Dawn Joy Fashions, Inc., 214 F.3d 224, 230 (2d Cir. 2000); Fed.R.Civ.P.
6(a). If a motion is not filed within the ten day period, the district
court is divested of jurisdiction to consider the motion. See Weissman,
224 F.3d at 230. There is no dispute that defendants completed the filing
and service of their notice of motion and supporting affidavit within the
ten day period. Plaintiffs' contention here thus requires a determination
of when a motion is deemed filed for purposes of Rules 50 and 59.
Generally, a motion is deemed filed when it is delivered to the Clerk's
Office. See Wight v. BankAmerica Corp., 219 F.3d 79, 85 (2d Cir. 2000).
Defendants' notice of motion and affidavit were delivered to the Clerk's
Office and served within the ten day period. Plaintiffs argue that such
filing was not effective for purposes of Rules 50 and 59, however,
because the papers filed failed to include a memorandum of law as
required by N.D.N.Y.L.R. 7.1(a)(1) and 7.1(g). Plaintiffs also argue that
allowing defendants to file their memorandum of law six weeks after their
notice of motion was filed circumvents the purpose of the ten day limits
in Rules 50 and 59.
Plaintiffs arguments fail for at least two reasons. First, defendants'
motion was accepted for filing by the Clerk and the Court on December
19, 2000 and for that reason alone should be deemed filed as of that
date. See Wight, 219 F.3d at 85 (holding that the district court's
treatment of a motion filed under Rule 59 as timely supported the
conclusion that the motion timely filed even though supplemental
pleadings were filed after the ten day period to cure technical defects in
the original pleadings). Second, the requirement in the local rules that
a memorandum of law accompany any motion may be excused or delayed if
"good cause is shown." N.D.N.Y.L.R. 7.1(b)(3). Here, given the length and
complexity of the trial and the fact that a transcript of the trial was
not completed until January 9, 2001 (Docket Nos. 143-95), good cause
existed for permitting defendants additional time to file and serve their
memorandum of law. Thus, the delayed filing of the memorandum of law did
not vitiate the timeliness of defendants' motion under the local rules.
Moreover, a district court possesses the inherent power to authorize
departures from the requirements of its local rules. See Somlyo v. J.
Lu-Rob Enters., 932 F.2d 1043, 1048-49 (2d Cir. 1991) ("it is the
business of the district court to determine whether fairness demands that
noncompliance [with local rules] be excused"); see also Wight, 219 F.3d
at 85-86 (approving the district court's "`tailor[ing] the Local Rules to
best achieve a just outcome.'") (quoting Somlvo, 932 F.2d at 1049).
Accordingly, defendants' motion here was timely filed and plaintiffs'
contention to the contrary must be rejected.
A plaintiff may establish violations of the ADEA and the HRL under
theories of disparate treatment and disparate impact. Maresco v. Evans
Chemetics, Div. of W.R. Grace & Co., 964 F.2d 106, 115 (2d Cir. 1992);
see also Smith v. Xerox Corp., 196 F.3d 358, 363 n. 1 (2d Cir. 1999)
(holding that claims under the ADEA and the HRL "are analyzed
identically"); Gonzalez v. City of New York, 135 F. Supp.2d 385, 399-400
& nn. 9, 11 (E.D.N.Y. 2001) (holding that disparate impact age
discrimination claim may be asserted under the HRL and analyzed
identically to those brought under the ADEA). Disparate treatment
requires evidence that an employer intentionally discriminated against a
person forty years of age or older while "[d]isparate impact . . . results
from the use of employment practices that are facially neutral in their
treatment of different groups but that in fact fall more harshly on a
protected group and cannot be justified by business necessity." Maresco,
964 F.2d at 115 (internal quotation and alteration omitted); see also
District Council 37, AFSCME v. New York City Dep't of Parks &
Recreation, 113 F.3d 347, 351 (2d Cir. 1997). Disparate impact age
discrimination does not require proof of discriminatory intent. Smith,
196 F.3d at 364 (citing Griggs v. Duke Power Co., 401 U.S. 424, 432
(1971)). Rather, the disparate impact theory addresses employment
practices "that are fair in form but discriminatory in impact." Griggs,
401 U.S. at 431. Although courts of appeals are divided over the
viability of the disparate impact theory under the ADEA, this theory of
proof remains available in the Second Circuit. See Smith, 196 F.3d at 367
& n. 6.
To recover under a disparate impact claim, a plaintiff must first
establish a prima facie case "by identifying a specific employment
practice which, although facially neutral, has had an adverse impact on
her as a member of a protected class." Smith, 196 F.3d at 364 (citing
Watson v. Fort Worth Bank & Trust, 487 U.S. 977, 994 (1988)); Knighton
v. City of Syracuse Fire Dep't, 145 F. Supp.2d 217, 224 (N.D.N.Y. 2001)
(Scullin, C.J.); Gonzalez, 135 F. Supp.2d at 399; Hogan v. General Elec.
Co., 109 F. Supp.2d 99, 104 (N.D.N.Y. 2000) (Hurd, J.). The burden then
shifts to the employer to offer a business necessity for the challenged
employment practice. Smith, 196 F.3d at 365; Hogan v. Metromail,
167 F. Supp.2d 593, 595 (S.D.N.Y. 2001). If an employer offers such a
business necessity, the burden then returns to the plaintiff to "show
that the employer's proffered reason was merely a pretext for
discrimination." Id.; Knighton, 145 F. supp.2d at 224. Defendants contend
on this motion that plaintiffs failed at each of these steps of their
claim and as well failed to establish that the defendants' conduct was
willful or that defendants were liable under the HRL.
1. Plaintiffs' Prima Facie Case
Defendants contend that plaintiffs' evidence failed to establish a
prima facie case as a matter of law because (a) plaintiffs failed to
identify a facially neutral employment practice, (b) the statistical
evidence was insufficient, and (c) the testimony of plaintiffs'
statistical expert witness was improperly admitted and insufficient.
Defs. Mem. of Law at 8-16.
First, however, the issue presented here follows a trial on the
merits. "It is well-established that once a [discrimination] case has
been fully tried on the merits, the question whether the plaintiff has
established a prima facie case is no longer relevant." Ottaviani v. State
Univ. of New York, 875 F.2d 365, 373 (2d Cir. 1989)
(citing Mitchell v. Baldrige, 759 F.2d 80, 83
(D.C. Cir. 1985)) (internal quotations
omitted).*fn15 Thus, "the only issue to be decided at that point is
whether the plaintiffs have actually proved discrimination." Bazemore v.
Friday, 478 U.S. 385, 398 (1986). Because this case has been fully
tried, defendants' motion on this ground is denied since the only
cognizable issue at this stage is whether plaintiffs proved
discrimination. In the alternative, however, defendants' arguments
concerning plaintiffs' prima facie case are considered on their merits.
a. Facially Neutral Employment Practice
Plaintiffs' initial burden under the disparate impact theory required,
inter alia, that they identify "a specific employment practice which,
although facially neutral, has had an adverse impact on" older
employees. Smith, 196 F.3d at 364; Gonzalez, 135 F. Supp.2d at 399.
Plaintiffs identified "the implementation of the exempt employee [IRIF]
portion of the WAP." Pls. Mem. of Law (Docket No. 203) at 11-13.
Defendants contend that this employment practice was insufficiently
specific to satisfy plaintiffs' burden.
The requirement that a plaintiff specify the employment practice which
he or she alleges caused the discrimination serves to insure that an
employer not be held liable for discrimination simply because a result
appeared discriminatory numerically. See Smith, 196 F.3d at 367-68; see
also Wards Cove Packing Co. v. Atonio, 490 U.S. 642, 656-58 (1989).
Thus, "a plaintiff generally cannot rely on the overall decision-making
process . . . [as] a specific employment practice." Smith, 196 F.3d at
Plaintiffs contend that because the implementation of the guidelines
for the IRIF was a component of defendants' WAP, which included other
components such as the VSP, employee retraining and transfers, their
specification of the "implementation" of the IRIF satisfied its burden of
identifying a specific employment practice. Such identification,
however, actually specifies the decision-making process by which exempt
employees were selected for the termination of their employment. The
implementation identified here by plaintiffs retains all the elements of
overall decision-making regarding the IRIF, which itself was comprised of
various elements. The fact that defendants' implementation of the IRIF
was a subpart of the larger WAP does not of itself render it a "specific"
employment practice sufficient to satisfy plaintiffs' initial burden.
Nevertheless, a general decision-making process such as that identified
by plaintiffs here may serve as the specified employment practice "if the
plaintiff can show that the elements of the employer's decision-making
process are not capable of separation for analysis." Smith, 196 F.3d at
368 (citing 42 U.S.C. § 2000e-2(k)(1)(B)(i)). The trial testimony
established that exempt employees were selected for the IRIF pursuant to
"Salaried Employee Reduction-in-Force Guidelines." Those guidelines
established four criteria by which managers were to select exempt
employees for the IRIF. Employees considered for the IRIF were given up
to ten points in each for: company service, performance, flexibility and
criticality. Points were totaled for each exempt
employee considered for
the IRIF and the exempt employees with the lowest point totals were
selected for the IRIF. These guidelines were facially neutral, were
followed by all managers in selecting exempt employees for the IRIF, and
led directly to the selection of plaintiffs for the IRIF.
The criterion for service was objective on its face, but because it
denied credit to employees for any period of employment beyond twenty
years, it arguably had a disproportionate impact on employees with
lengthy periods of employment with defendants, all of whom were over
forty years of age.*fn16 The remaining three criteria required the
application of objective standards in various categories.*fn17 The points
assigned for the four criteria were then totaled to determine which
employees would be selected for the IRIF.
Thus, various factors were considered for each criterion and the four
criteria were considered as a group to determine selection for the IRIF.
The evidence adduced by plaintiffs at trial, principally that of KAPL
personnel regarding the guidelines and plaintiffs' statistical expert
witness, sufficed to establish that no particular factor and no
particular criterion caused the disparate impact on older employees.
Rather, the evidence established that, as asserted by plaintiffs, such
impact resulted from the implementation of the guidelines and that the
various factors and criteria, the elements of defendants'
decision-making, reasonably were "not capable of separation for
analysis." Smith, 196 F.3d at 368. Thus, in the circumstances of this
case, plaintiffs' identification of the implementation of the guidelines
for selecting exempt employees for the IRIF satisfied plaintiffs' initial
burden of proof.
b. Sufficiency of Plaintiffs' Statistical Evidence
Defendants next contend that plaintiffs failed to satisfy their initial
burden of demonstrating that any statistical disparity against plaintiffs
was caused by age discrimination. Defs. Mem. of Law at 11-14. Plaintiffs
respond that the statistical evidence offered through their expert
witness sufficed to meet this burden. Pls. Mem. of Law at 13-19.
After a plaintiff specifies the employment practice responsible for the
adverse impact on older employees, a plaintiff must then demonstrate,
generally through statistical data, that the employment practice caused a
significant disparity in outcome between older employees and younger
employees. See Smith, 196 F.3d at 364-65 (citing Watson, 487 U.S. at
994-95). Where a plaintiff relies on a statistical disparity to meet this
burden, that disparity "must be sufficiently substantial to raise an
inference of causation." Id.; see also Watson, 487 U.S. at 994 (holding
that statistical evidence at this stage must be "of a kind and degree
sufficient to" infer discrimination); Gonzalez, 135 F. Supp.2d at 399.
No bright line rules have been established to determine the sufficiency
of statistical evidence at this stage. See Ottaviani, 875 F.2d at 373
("in accordance with Supreme Court pronouncements, we must reject
appellants' suggestion that this court announce a rule of law with respect
to what level of statistical significance automatically gives rise to a
presumption of discrimination."); see also Smith, 196 F.3d at
365 ("no bright line rules exist"). However, courts generally require a
plaintiff to meet two requirements. First, a "plaintiff must identify the
correct population for analysis." Smith, 196 F.3d at 368. Generally, this
population will be those employees who were subject to the employment
practice in question. See id.; Lander v. Montgomery County Bd. of
Commissioners, 159 F. Supp.2d 1044, 1060 (S.D.Ohio 2001); Shah v. New
York State Dep't of Civil Serv., No. 94 CIV 9193 RPP, 2001 WL 839986, at
*7 n. 7 (S.D.N.Y. July 25, 2001).
Here, plaintiffs identified four possible populations through the
testimony of their expert witness, Dr. Janice Fanning Madden. Included in
the four groups were all exempt employees at KAPL and those 245 exempt
employees who were actually considered for the IRIF. Madden Tr. (Docket
No. 154) at 17-18. There was sufficient evidence from which the jury
could reasonably find that either group was the correct population for
purposes of comparison. As to the population of exempt employees, the
implementation of the IRIF commenced with that population within which
those employees with excess skills were identified. Moreover, in its own
internal analysis of the statistical impact of the IRIF on older
employees, KAPL itself chose the population of 2.063 exempt employees for
purposes of analysis. See, e.g., Madden Tr. at 34; see also Fed.R.Evid.
801(d)(2) (admission by party-opponent). Thus, the record contained
sufficient evidence from which the jury could conclude that the correct
population for purposes of comparison was KAPL's entire population of
There was also evidence from which the jury could find that the correct
population was comprised of the 245 exempt employees whose positions were
identified during the excess skills analysis as subject to the IRIF. It
was this more limited group who were placed on the matrices and whose
numerical scores were compared according to the procedures prescribed in
the WAP. Thus, it was this group of exempt employees who were actually
and universally subject to all aspects of defendants' implementation of
the guidelines for selecting exempt employees for the IRIF — the
employment practice identified by plaintiffs in their prima facie case.
Accordingly, the record also contained sufficient evidence from which the
jury could conclude that the correct population for purposes of
comparison was the exempt employees who were placed on the matrices,
i.e., those actually considered for the IRIF. See Smith, 196 F.3d at 368
("The [correct] population in a reduction-in-force situation consists of
workers subject to termination."). The second requirement is that a
plaintiff demonstrate a statistically sufficient comparison between the
predicted result for older employees in the population selected and the
actual result. Ottaviani, 875 F.2d at 371; see also Smith, 196 F.3d at
366 (referring to "expected result" and "obtained result" rather than
"predicted result" and "actual result"); Hogan, 109 F. Supp.2d at 104.
This comparison serves to determine whether there is a statistically
significant disparity between the predicted and actual results. For this
case a predicted result is determined by the percentage of older
employees in the population of employees subject to the IRIF. The actual
result constitutes the percentage of older employees actually selected
for the IRIF in the population subject to the IRIF. A comparison of the
predicted and actual results must be made to determine if a statistically
significant disparity exists. Smith, 196 F.3d at 365-66; Ottaviani, 875
F.2d at 371.
"Statistical significance" measures "the probability that a disparity
is simply due to chance rather than any other identifiable factor."
Ottaviani, 875 F.2d at 371. One measure of statistical significance is
standard deviation. See Smith, 196 F.3d at 365-66; Ottaviani, 875 F.2d at
371. Standard deviation is a "unit of measurement used to express the
probability that an [actual] result is merely a random deviation from a
predicted result." Ottaviani, 875 F.2d at 371. Generally, "`[t]he greater
the number of standard deviations, the less likely it is that chance is
the cause of any difference between the [predicted] and [actual]
results.'" Id. (quoting Coates v. Johnson & Johnson, 756 F.2d 524, 536
(7th Cir. 1985)). If the actual result varies from the predicted result
by two standard deviations,*fn18 "[courts generally consider this level
of significance sufficient to warrant an inference of discrimination."
Smith, 196 F.3d at 366; see also Ottaviani, 875 F.2d at 371-72.
Here, as to the population of 2.063 exempt employees, there were
1,203, or fifty-eight percent, who were older employees. The predicted
result from this population for the thirty-one exempt employees whose
employment was terminated in the IRIF was thus fifty-eight percent of
thirty-one, or eighteen older employees. The actual result was that
thirty of the thirty-one employees, or ninety-seven percent, whose
employment was terminated in the IRIF were older employees. According to
Dr. Madden, there was one chance in 348,000 that this actual result would
occur. Madden Tr. at 20. This constitutes between four and five standard
As to the population of 245 exempt employees placed on the matrices and
actually considered for the IRIF, there were 179, or seventy-three
percent, who were older employees. The predicted result from this
population for the thirty-one exempt employees whose employment was
terminated in the IRIF was thus seventy-three percent of thirty-one, or
twenty-three older employees. The actual result was thirty, or
ninety-seven percent. According to Dr. Madden, there was one chance in
1,260 that this actual result would occur. Madden Tr. at 25-27. This
constitutes between three and four standard deviations.*fn19
However, this conclusion is further supported by testimony offered by
Dr. Madden during her redirect examination. The statistical evidence
described above assessed the evidence to determine the probability that
the statistical disparity between the predicted and actual results
deviated from the norm by chance or for some other reasons, such as age
discrimination. According to Dr. Madden, she also performed multiple
regression analyses — "a statistical test which identifies factors,
called independent variables, that might influence the outcome of an
observed phenomenon, called a dependent variable." Smith, 196 F.3d at 363
n. 3. By this method, a statistician seeks to identify legitimate
factors, other than age discrimination here, which may have influenced or
caused the statistical deviation from the norm. Id.
Dr. Madden testified that she performed multiple regression analyses to
determine if the statistical disparity could be explained by the
individual components of the employees' numerical ratings on the matrices
— years of service, performance appraisals, flexibility and
criticality. Dr. Madden concluded that there was no significant
correlation between an employees' years of service and the likelihood
that he or she would be selected for the IRIF. Madden Tr. at 88. As to
the other three components, Dr. Madden found that there existed
statistically significant age effects in that older employees were more
likely than younger employees to have experienced a decline in their
performance appraisals after KAPL began planning for the IRIF and
generally in the ratings for flexibility and criticality. Id. at 88-89.
This evidence further supported the jury's finding that plaintiffs'
satisfied their burden of proof to establish a prima facie case.
Accordingly, defendants' contention that there was insufficient evidence
to establish a prima facie case must be rejected.
c. Evidence of Multiple Regression Analyses
Defendants contend that Dr. Madden's testimony regarding multiple
regression analyses was erroneously admitted in evidence. Defs. Mem. of
Law at 14-16. Dr. Madden made no mention of multiple regression analyses
in her expert witness reports, her deposition before trial or during her
direct examination at trial. On cross-examination at trial, however, Dr.
Madden was asked by defendants' counsel at least three times about
additional analyses she had conducted. Madden Tr. at 38-39, 56-58,
76-77. In none of his questions did defendants' counsel limit Dr. Madden
to analyses conducted prior to her reports or her deposition, and Dr.
Madden testified each time that she had in fact conducted such analyses
following her deposition and prior to trial. Id. In the most detailed
colloquy, Dr. Madden was questioned as follows:
Q. You did not do a regression analysis for
criticality, did you?
Q. And did you do one for flexibility?
Q. And for company service?
Madden Tr. at 58. On redirect examination, plaintiffs' counsel asked Dr.
the results of her multiple regression analyses. Id. at 83.
Defendants' counsel objected that such analyses had not previously been
disclosed to defendants. Id. The objection was overruled on the ground
that the questions of defendants' counsel had "opened the door" and
unless plaintiffs' counsel was permitted to ask Dr. Madden about these
analyses, the jury could infer that her analyses were unfavorable to
plaintiffs. Id. at 83-87. Dr. Madden then testified to the results of her
multiple regression analyses. Id. at 87-89. A trial court is accorded
"substantial deference" in making evidentiary rulings and those rulings
are reviewed only for clear abuse of discretion. Reilly v. Natwest Mkts.
Groups, Inc., 181 F.3d 253, 266 (2d Cir. 1999); see also Healey v.
Chelsea Res., Ltd., 947 F.2d 611, 619-20 (2d Cir. 1991). "Further, even
an erroneous evidentiary ruling will not lead to reversal unless
affirmance would be `inconsistent with substantial justice.'" Perry v.
Ethan Allen, Inc., 115 F.3d 143, 150 (2d Cir. 1997) (quoting
Fed.R.Civ.P. 61). Here, the ruling on defendants' objection was proper.
Defendants' questions to Dr. Madden did not limit her answers to analyses
she performed prior to her reports or her deposition. Defendants elicited
testimony from Dr. Madden that she had performed multiple regression
analyses but left unanswered the results of those analyses. If plaintiffs
had not been permitted to elicit those results, the jury could reasonably
have inferred that Dr. Madden's analyses were adverse to plaintiffs.
Thus, the testimony of Dr. Madden on redirect examination was properly
admitted and was not an abuse of discretion. Cf. Caruolo v. A C & S,
Inc., No. 93 CIV. 3752 (RWS), 1999 WL 147740, at *14 (S.D.N.Y. Mar. 18,
1999) (holding that cross-examination by defendants of plaintiffs' expert
witness opened the door to the admission of additional testimony on
redirect examination of tests conducted by another expert and reported in
a treatise), aff'd in part and rev'd in part on other grounds, 226 F.3d 46
(2d Cir. 2000).
Moreover, while the testimony of Dr. Madden in this regard supported
plaintiffs' claims, that testimony was limited, taking no more than two to
three minutes (less than two pages of transcript) in a trial that lasted
five weeks during the liability phase alone. Its secondary importance is
also underscored by the limited time devoted to it by counsel during
their summations. Accordingly, even if the admission of this evidence was
erroneous, any such error was harmless. See Fed.R.Evid. 61. Defendants'
contention here must, therefore, be rejected.
2. Defendants' Burden of Production
Plaintiffs having satisfied their burden of establishing a prima facie
case, the burden then shifted to defendants to offer a business
justification for the challenged employment practice. See Griggs, 401
U.S. at 432; Smith, 196 F.3d at 365. The jury here found that defendants
failed to meet this burden. Special Verdict (Docket No. 86) at 11.
Defendants contend that this finding was erroneous as a matter of law.
Defs. Mem. of Law at 17-21. Plaintiffs disagree. Pls. Mem. of Law at
An employer's burden at this stage is to demonstrate "a legitimate
business justification" for the challenged employment practice. Wards
Cove, 490 U.S. at 643. The employer's burden is one of production, not
persuasion, as the burden of persuasion remains with the plaintiff
throughout the case. See id.; Watson, 487 U.S. at 108; EEOC v. Joint
Apprenticeship Comm., 186 F.3d 110, 120 (2d Cir. 1999). An employer meets
this minimal burden by articulating a business necessity or reason for
employment practice. See Watson, 487 U.S. at 108.
The jury found that defendants had not met this burden. Defendants
first argue that the jury was asked the "wrong question." Defs. Mem. of
Law at 17. The verdict form asked the jury whether any defendant "has
articulated a business justification for selecting the plaintiffs for
termination of their employment in the reduction-in-force?" Special
Verdict at 11. Defendants failed to object to this question prior to its
submission to the jury. Charge Conference Tr. (Docket No. 163) at 44-48.
Thus, the question must rise to the level of plain error to be
cognizable. See Shah v. Pan Am. World Servs., Inc., 148 F.3d 84, 96 (2d
Cir. 1998) (applying plain error review where party failed to object to
proposed special verdict form). Defendants do not claim plain error ...