Free Reply to Response to Motion - District Court of Colorado - Colorado

Case 1:03-cv-02485-MSK-PAC

Document 380-7

Filed 03/29/2006

Page 1 of 8

EXHIBIT F

Case 1:03-cv-02485-MSK-PAC

Document 380-7

Filed 03/29/2006

Page 2 of 8

LEX

SEE 1994 U.S. DIST. LEXIS 16511

SHEILA ROSEN, individually and on behalf of all other persons similarly situated, Plaintiffs, v. RECKITT & COLMAN INC. and BOYLE-MIDWAY INC., Defendants.
91 Civ. 1675 (LMM)
UNITED STATES DISTRICT COURT FOR THE SOUTHERN DISTRICT OF NEW YORK

1994 U.S. Dist. LEXIS 16511; 77 Fair Empl. Prac. Cas. (BNA) 370

November 10, 1994, Decided
November 17, 1994, Filed

CASE SUMMARY:

OUTCOME: The court denied the employees' motions
for bifucation and a protective order and granted defen-

PROCEDURL POSTUR: Plaintiff former employees brought an action against defendants, their employer
and the company that acquired the employer's business,

dants' motion to compel discovery with limitations on the form of some of the testimony.
LexisNexis(R) Headnotes

in the employees' action under the Fair Labor Standards
Act, 29 u.s.es. § 216(b), as incorporated by the Age

Discrimination in Employment Act (ADEA), 29 u.s.es.

§ 621 et seq. The employees fied motions for bifucation and a protective order, and defendants fied a motion to compel discovery.

Governments:; Courts :; Rule Application & Interpretation
Civil Procedure:; Discovery Methods

(HN1) The cours have a preference for the liberal con-

OVERVIEW: The employees claimed that defendants engaged in a pattern or practice of age discrimiation in employment. The affected class of employees consisted
of persons who were termnated when the employer's

strction of discovery rules.

business was sold. The employees sought bifurcation of the trial into a single "macro" trial on common issues of liability and a series of "micro" trials focusing on issues of liability and damages as they pertained to the individ-

Governments :; Courts :; Rule Application & Interpretation
Civil Procedure:; Discovery Methods
(HN2) While discovery provides many opportities for

abuse, this concern must be balanced against a defendant's need to prepare a meaningful defense.

ual members of the affected class of employees. The
court found that bifucation was inappropriate because

the employees failed to show that they were similarly situated and that their claims were not just a mass of in-

dividual claims. The cour found no compelling reasons
to bifurcate the triaL. The cour also rejected the employ-

Evidence:; Writings & Real Evidence:; Summaries Civil Procedure:; Discovery Methods (HN3) Cours have often been forced to confront the

ees' argument that limiting defendants' depositions to a small percentage of the employees would provide a sufficient sample size. The size of the affected class was small and presented a problem with achieving statistical

problem of stability of samples from small sets. It is axiomatic in statistical analysis that the precision and
dependability of statistics is directly related to the size of the sample being evaluated.

stability. Additionally, anecdotal evidence might be
sparsely distributed. Although the cour refused to limit the number deposed, the court only permtted defendants one day to depose each class member.

Labor & Employment Law :; Discrimination :; Age
Discrimination:; Defenses & Exceptions

Labor & Employment Law :; Discrimination :; Age
Discrimination:; Proof of Discrimination

Case 1:03-cv-02485-MSK-PAC

Document 380-7

Filed 03/29/2006

Page 3 of 8
Page 2

1994 U.S. Dist. LEXIS 16511, *; 77 Fair Empl. Prac. Cas. (BNA) 370

(HN4) An employer is clearly entitled to seek anecdotal evidence that may be used to support or negate the pattern or practice claim of an affected class of employees. Without significant individual testimony to support statistical evidence, cours have refused to find a pattern or
practice of discrimnation. Pattern or practice actions

and in requiring that which is to be brought to the jury's

attention to be done so in a manner least likely to confse

that body. Fairness requires that if the number of witnesses is limited, the limitation should be applied equally
to all partes, to the same issue of fact, and only after

notice to the parties.

under the Age Discrimiation in Employment Act re-

quire both statistical and anecdotal evidence.

JUDGES: (* 1) McKENNA
OPINIONBY: LAWRENCE M. McKENNA

Labor & Employment Law :; Discrimination :; Age Discrimination:; Enforcement Civil Procedure:; Trials:; Separate Trials
(HN5) Bifucation of an Age Discrimination in Employment Act action is likely to be inappropriate if the
plaintiffs are not similarly situated.

OPINION:
MEMORANDUM AND ORDER
McKENNA, D.l
Plaintiffs brought this action against Defendants,

Labor & Employment Law :; Discrimination :; Age Discrimination:; Enforcement
Civil Procedure:; Class Actions:; Prerequisites Civil Procedure:; Trials:; Separate Trials (HN6) The Teamsters model would be inappropriate for what are, in effect, individual claims. A requirement for a class action to proceed under the Age Discrimination in Employment Act is that the employees must be similarly
situated.

Reckitt & Colman Inc. and Boyle-Midway Inc. (jointly referred to herein as "R&C"), pursuant to the Fair Labor Standards Act, 29 Us. C. § 216(b) (1965 & Supp. 1994), as incorporated by the Age Discrimination in Employment Act, 29 Us.c. § 621, et seq., (1985) ("ADEA").

Presently before the Cour are Plaintiffs' motions for bifucation and a protective order, and Defendants' motion to compel discovery. For the reasons stated below, Plaintiffs' motions are denied and Defendants' motion is
granted.

I. The Instant Motions

Evidence:; Witnesses:; Examination & Presentation of Evidence

Labor & Employment Law :; Discrimination :; Age Discrimination:; Enforcement (HN7) Pretrial planning should be given to the proper
presentation of "anecdotal" evidence regarding the individual experiences of various employees, union stew-

Plaintiffs claim that Defendants engaged in a pattern or practice of age discrimination in employment. The affected class n1 consists of certain persons formerly employed by Boyle-Midway, who were termnated when that company was acquired by Reckitt & Colman.

ards, supervisors, and managers. Such evidence may be

n1 While resembling a class action under
Fed. R. Civ. P. 23, this is a representative, opt-in action -- unique to the ADEA. References to the

offered by the plaintiffs or defendants to provide ilustrative support for their respective positions and for the

studies conducted by their experts and, indeed, may be critical to a decision on the merits of the claim of class discrimination. Some limits, however, may be needed on the number of such persons the parties may call as witnesses.
(*2)

plaintiff "class" are for ease of reference only, and do not imply the relevance of Rule 23 jurisprudence to this action.

Plaintiffs seek bifurcation of the trial into two
Evidence:; Witnesses:; Examination & Presentation of Evidence Evidence:; Relevance :; Confusion, Prejudice & Waste
of

stages: a single "macro" trial, concerned with common

Time

(HN8) While infequent, prospective limitations have been placed on the testimony of percipient witnesses to limit witnesses to a representative sample. The district
cour retains considerable latitude even with admittedly relevant evidence in rejecting that which is cumulative

issues of liability; and a series of "micro" trials focusing on issues ofliability and damages as they pertain to individual class members. Plaintiffs suggest that bifucated trials are utilized in virally all "pattern or practice" employment discrimination cases because the two stage trial
scheme is a highly effcient means of conducting such

multi-party litigation.

Case 1:03-cv-02485-MSK-PAC

Document 380-7

Filed 03/29/2006

Page 4 of 8
Page 3

1994 U.S. Dist. LEXIS 16511, *; 77 Fair Empl. Prac. Cas. (BNA) 370

R&C argues that trial bifucation would violate their due process right to present defenses to this action at a "meaningful time." In essence, R&C's argument is that deferring the presentation of individual defenses until the micro trials increases the likelihood of a finding of a pattern of discrimination in the macro triaL. This finding,

Oct. 15, 1991). Pursuant to that Order the Cour wil not permt additional plaintiffs to join the class.

they contend, creates a presumption of liability that wil
unfairly prejudice them in each micro triaL. R&C also

n3 There does not appear to be a statute of limitations governng the opt-in procedure in an ADEA action. Kelley v. Alamo, 964 F.2d 747,
750 (8th Cir. 1992). In Kelly, the court argued

maintains that the motion to bifurcate is prematue, as
the parties are in the midst of discovery, which may alter the optimal trial strategy.

that cutoff dates could be established in the absence of Congressional direction to the contrary,

Also before the Cour are cross discovery motions. On April 26, 1994, R&C sent notices of its intent to take

so long as plaintiffs were given the maximum latitude consistent with fairess to the defendant,
and so long as the cutoff date was set prospec-

the depositions (*3) of twenty of the fift members of
the plaintiff class. Plaintiffs' counsel refused to produce these individuals, on the ground that this number of
depositions was unnecessary and overly burdensome.

tively. Id.; see also Roshto v. Chrysler Corp., 67 F.R.D. 28, 29 (E.D. La. 1975) (cutoff dates have been ordered in many § 216 cases).
(*5)

Plaintiffs, assuming the Court wil bifucate the trial,
have offered to produce only those few class members
they intend to call as witnesses in the macro triaL.

III. Discovery
The Cour next considers the cross motions concerning the number of class members that R&C may depose. R&C maintains that their discovery efforts to date have been thwarted by Plaintiffs' counsel, who, inter alia, pro-

R&C argues that it is entitled to discover all evi-

dence, particularly anecdotal evidence, relevant to the claims and their defenses in this action. It objects to discovery limited to a sample of the class hand-picked by
Plaintiffs' counseL. R&C has moved the Court to compel

vided class members with draft answers to many of
R&C's interrogatories. R&C seeks more effective discovery through depositions of each member of the class, in order to prepare both common and individual defenses to this action.

the production of each member of the class for depositions, and Plaintiffs have cross-moved for a protective
order.

II. Establishing a Cutoff Date for Opt-ins
As a preliminary matter, the Court addresses the issue of the size of the plaintiff class. The class curently contains the named Plaintiff, Sheila Rosen, and 49 optins. Plaintiffs have suggested that the number of opt-ins
may rise to over 100.

The partes devote much of their arguments toward the proper characterization of ths action as one or more
of "disparate impact," "disparate treatment," or "pattern

or practice." Such characterization may alter the initial or shifting burdens of proof, n4 which may determne, op-

timal or permissible trial strategies, which in tu may affect how discovery is conducted. For example, in a
Title VII "pattern or practice" action bifucated under the Teamster model, it might be desirable to limit discovery for the macro trial to a small percentage of the plaintiff class, deferring discovery of the entire class to any later micro trials.

R&C argues that the class should now be closed, since Sheila Rosen notified all potential opt-ins of ths
action almost three years ago, and only 58 individuals

(*4) chose to respond. n2 R&C maintains that it is entitled to know its exposure in this action, and to begin to
prepare all of

its individual defenses.
n4 See, e.g., International Bhd. of Teamsters v. United States, 431 Us. 324, 521. Ed. 2d 396,

n2 In its Order of March 17, 1994, the Court dismissed 9 members of the class who failed to respond to R&C's discovery requests.
The parties appear to have forgotten that on Plain-

97 S. Ct. 1843 (1977) (defining the "Teamster"

bifucation model for Title VII class actions under Rule 23); Rosen v. Reckitt & Colman Inc.,
No. 91 Civ. 1675, 1992 Us. Dist. LEXIS 18039,

tiffs' suggestion, the Court set a bar date for joinder of 90 days from the date of the mailing of notice to potential class members. n3 Rosen v. Reckitt & Colman Inc., No.
91 Civ. 1675, 1991 Us. Dist. LEXIS 14663 (S.D.N.Y.

at *5-9 (S.D.N.Y. Nov. 24, 1992).
(*6)

The Cour does not view this discovery scheme,
suggested by the plaintiff class, to be an especially eff-

Case 1:03-cv-02485-MSK-PAC

Document 380-7

Filed 03/29/2006

Page 5 of 8
Page 4

1994 U.S. Dist. LEXIS 16511, *; 77 Fair Empl. Prac. Cas. (BNA) 370

cient one. First, as discussed below, if the claim is ultimately viewed to be one of pattern or practice, then R&C may require a substantial amount of additional discovery to frame its defense. It might require a relatively large number of depositions for statistical purposes, and a still

Plaintiffs' argument is fatally flawed as a matter of elementary statistics, because the size of the instant class is small. A simple example demonstrates the problem of

larger number to discover anecdotal evidence for this common issue. Second, in the event that the macro trial is followed by micro trials, R&C would at that time be entitled to detailed discovery from each class member.
Third, in the event that the macro trial fails to establish a pattern or practice, plaintiffs are likely to pursue individual actions, in which case R&C would again be entitled

achieving statistical "stability" in sampling small sets. Suppose we are interested in determning the average income of the residents in a particular geographical region. If that region is the United States, with a population
of 250 millon people, a randomly chosen one percent

population sample, containng 2.5 million individuals, would given an accurate estimate of the average income of the entire country. This estimate would be highly "stable" in the sense (*9) that the fortitous inclusion of

to discovery from each plaintiff. Fourh, R&C argues
that it cannot conduct meaningful settlement negotiations

highly unusual individuals, such as lottery winners,

would be unlikely to affect the accuracy of the answer.
n5

until it knows its exposure, which wil depend in large
part on the results of individual depositions, including

recent informtion on the mitigation of damages. Fifth,
(HNl) the Cour takes notice of

the preference for liberal

constrction of discovery rules. See Hickman v. Taylor,

329 Us. 495, 507-08, 91 1. Ed. 451, 67 S. Ct. 385
discovery provides many opportities for abuse, this
concern must be balanced against a defendant's need to

(1947). (*7) Sixth, (HN2) while the Cour is aware that

n5 Since the averaging process would involve totalling the income of all 2.5 million individuals, and then dividing by 2.5 million, the effect of a few unusually large incomes would be strongly "diluted" by the incomes of millions of more tyical individuals.

prepare a meaningful defense.
While each of these factors supports R&C's position, the Cour fmds two of them to be partcularly significant:

If, on the other hand, the geographical region was a hamlet with population 100, the same one percent population sample would consist of a single individuaL. This

the likelihood that R&C will ultimately be entitled to depose each member of the class; and R&C's right to
obtain statistically significant evidence in support of its defense.

result is clearly "unstable" in the sense that the average income would change dramatically depending on the particular individual chosen.
(HN3) Cours have often been forced to confront the problem of stability of samples from small sets. See, e.g., Moultrie v. Martin, 690 F.2d 1078, 1083 & nn. 7-10 (4th Cir. 1982) ("It is axiomatic in statistical analysis that the precision and dependability of statistics is directly related
to the (*10) size of the sample being evaluated.");

The parties have vigorously debated the issue of a suffcient sample size. R&C cites to several employment discrimination cases in which courts permtted at least as many depositions as they seek in this action. See, e.g., Defs.' Sureply Mem. at 7 (claiming that the cour in Pines v. State Farm General Ins. Co., 1992 Us. Dist.

LEXIS 6972 (C.D.Cal. Feb. 25, 1992), authorized 40
depositions out of a 960 member class); Id. (claiming
that the cour in Sperling authorized 90-100 depositions);
Kaas v. Pratt & Whitney, 1991 Us. Dist. LEXIS 11177,

at *14-16 (S.D. Fla. Mar. 18, 1991) (approximately 100 depositions); Lusardi v. Xerox Corp., 118 F.R.D. 351, 354 (D.N.J. 1987) (*8) (discovery of 51 members of 1312 member class and depositions of 10 of 13 named plaintiffs.); Brooks v. Farm Fresh, Inc., 759 F. Supp. 1185, 1188 (E.D. Va. 1991) (127 depositions permtted to determne if plaintiffs were all similarly situated).
Plaintiffs argue that in each of these cases, the number of depositions, while large in an absolute sense, constitutes a small percentage of the number of plaintiffs. Plaintiffs suggest that the same percentage, say 5%, be permtted here, which would amount to 2 or 3 authorized depositions.

United States v. La Chance, 788 F.2d 856 (2d Cir.) (explainig the dependence of statistical "fluctuations" upon sample size), cert. denied, 479 Us. 883 (1986); Paul G. Hoel, Introduction to Mathematical Statistics 130 (5th ed. 1984) (the precision of estimating an average value increases as the square root of the increase in the size of
the sample). In the action at hand, the "sample space"

number that is considered "small" for many statistical puroses. Hoel, supra, at 158-59. The Court therefore

(the potential pool of study) contains 50 individuals, a

rejects Plaintiffs' suggestion of limiting discovery to a
small percentage of the class.

The argument for permtting relatively many depositions in the instant action is strengthened by the fact that R&C seeks not only evidence that all class members wil have, such as evidence relating to damages, but also anecdotal evidence, which is likely to be very sparsely distributed throughout the class.

Case 1:03-cv-02485-MSK-PAC

Document 380-7

Filed 03/29/2006

Page 6 of 8
Page 5

1994 U.S. Dist. LEXIS 16511, *; 77 Fair Empl. Prac. Cas. (BNA) 370

(HN4) R&C is clearly entitled to seek anecdotal evidence that may be used to support or negate Plaintiffs'

pattern or practice claim. King v. General Elec. Co., 960 F.2d 617, 624 (7th Cir. 1992) (*11) ("Without significant individual testimony to support statistical evidence, courts have refused to fmd a pattern or practice of discrimination."); see also Sperling v. Hoffman-La Roche, Inc., No. 85 Civ. 2138 (D.N.J. Aug. 30, 1994) (unpublished opinion of Special Master recogning that "pat-

The opt-in plaintiffs present what amounts
to a series of more than thirteen hundred
individual cases. The members of the proposed class come from different depart-

ments, groups, organizations, suborganizations, units and local offces

tern or practice" actions under the ADEA require both statistical and anecdotal evidence); Manual for Complex Litigation, Second, § 33.53.
The sparsity of anecdotal evidence presents a somewhat different statistical problem from the previously considered issue of general estimation stability. The problem is best explained through the example of a state
lottery, a process driven by the exceptional outcome,
rather than the tyical one. Consider the estimation of the

within the Xerox organization. The opt-in plaintiffs perform different jobs at different geographic locations and were subject

to different job actions concernng reductions in work force which occurred at

various times as a result of various deci-

sions by different supervisors made on a
decentralized employee-by-employee basis.

percentage of New Yorkers who have :"on the gra~d prize in the lottery. Since there are relatively few wllners, averaging the number of winners in a small sample, say 100 individuals, would almost certainly produce the
result that exactly 0% of the population won the lottery.

!d. at 465. After decertification of the class, individual

actions were brought in the Western District of New
York, and the plaintiffs sought consolidation of the cases

and a trial of liability under the Teamsters modeL. In
granting summry judgment for Xerox, the Cour relied

While this numerical value is not far wrong in an absolute sense (i.e., the actual percentage is close to zero), it

in part on the weakness of plaintiffs' evidence, and the
findings in Lusardi, (*14) but also noted that regardless

is precisely the anecdotal evidence (*12) of the rare winner that is responsible for over $ 1 bilion in annual
lottery sales in New York State. n6

of the strength of plaintiffs' evidence, (HN6) the Team-

sters model would be inappropriate for what were, in effect, individual claims. In re Western Dist. Xerox Litig., 850 F. Supp. 1079, 1083 (WD.N.Y. 1994); see
also Sperling v. Hoffman-La Roche, Inc., 862 F.2d 439, 444 (3d Cir. 1988) (a requirement for a class action to proceed under the ADEA is that the employees must be similarly situated), affd, 493 Us. 165, 107 1. Ed. 2d
480, 110 S. Ct. 482 (1989).

n6 David Firestone, New York Lottery: Debating the Fate of a Payoff Not as Magical as It Seemed, N.Y. Times, October 4, 1994, at B4.

The anecdotal evidence that R&C seeks may be sparsely distributed among members of the plaintiff
class, yet it may also be highly probative of the issue of a
"pattern or practice" of discrimination. Were the Cour to

In the action at bar, the Court has not made a determination that Plaintiffs are similarly situated, nor has it
resolved any other substantive issue. n7 Such determa-

limit R&C to 10 or 20 depositions, anecdotal exculpatory

information might go entirely undiscovered. Fairness therefore mandates that R&C be granted an opportty to depose each member of this relatively small plaintiff
class. In order to prevent these depositions from over-

tions would be inappropriate before the substantial completion of discovery. Even if such a determation was made, the Cour is not obliged to bifurcate the trial, and finds no compelling reasons to do so in its discretion.

burdening the class, the Court wil impose a time limit on each deposition, discussed in Section V.
iv. Bifurcation

n7 The Court has not, as Plaintiffs claim "noted the suffciency of Plaintiffs' statistical
proofs." (PIs.' Reply Mem. at 33.) In its Order of

Plaintiffs have moved the Cour for an Order bifucating the trial on the Teamster modeL. However (HN5)

bifurcation is likely to be inappropriate if the Plaintiffs
are not (*13) similarly situated. In Lusardi v. Xerox

November 24, 1992, the Cour only held that the conficting statistical evidence presented by the parties raised an issue of material fact suffcient
to defeat R&C's motion for summary judgment.
(* 15)

Corp., 122 F.R.D. 463 (D.N.J. 1988), a conditionall~ certified opt-in class action under the ADEA was decertified after substantial discovery, because, inter alia:

Case 1:03-cv-02485-MSK-PAC

Document 380-7

Filed 03/29/2006

Page 7 of 8
Page 6

1994 U.S. Dist. LEXIS 16511, *; 77 Fair Empl. Prac. Cas. (BNA) 370

Plaintiffs have argued that bifurcation wil result in a
great saving of the parties' and the Cours' resources,

because common issues need only be tried in a single
macro trial, while individual issues can be tried in subse-

quent micro trials. Defendants more plausibly suggest
that bifurcation is most likely to result in additional effort -- following an initial "pattern or practice" trial, the court wil need to conduct trials for individual claims of liability and damages, even if the initial trial fails to establish a pattern of discrimination.

crimination or merit to individual causes of action, the same jury wil then hear evidence on the relevant individual defenses and damages. n8 Testimony for this second phase of the trial wil be strictly by deposition. Since the trial wil proceed directly from liability to damages, Defendants must use their limited time for each deposition wisely, as they wil not have a second opportnity to
depose class members with regard to individual issues.

If by bifucation the Plaintiff means separate trials,
with distinct juries, the Court would not expect signifi-

cant economies of scale, however this action is ultimately characterized. The Cour 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 jur has not had an

n8 In some complex cases, the jury, after deciding certain issues, may be called upon to consider other issues. Additional arguments and instrctions may be given if all the evidence was previously presented, or the cour may entertain
additional evidence, whether immediately after the first verdict or after some recess.

opportity to hear.
V. Trial Management
In the interest of fairness, conservation of the partes'

Manual for Complex Litigation, Second, §
22.45; see also Gries v. Zimmer, Inc., 1991 Us. App. LEXIS 16729 (4th Cir. May 9, 1991) (two phase trial with single jury for reduction in force

resources, and judicial convenience, the Court wil employ the following scheme for proceeding (* 16) with discovery and trying this case. R&C wil be permtted one day to depose each member of the plaintiff class. It may first require the individual class members to produce documents and information (e.g., relating to employment subsequent to termnation by R&C) by interrogatory and notice to produce. A Magistrate Judge wil be appointed to assist, if need be, in the coordination of depositions within geographic areas, so as to minimie
the parties' deposition-related expenses.

ADEA action); Jeffries v. Harleston, 820 F.
Supp. 741 (S.D.N.Y. 1993) (juries fact findings followed by further deliberations on individual liability and punitive damages.)
(* 18)

R&C may then move for the establishment of any rational subclasses. The trial (or trials, should subclasses be established) wil be conducted in two stages. The first stage wil be concerned with liability. Each side may

The Court recognizes that prospective limitations are infrequently placed on the number of percipient witnesses (witnesses testifying as to perception of ultimate facts). More common are limitations on the number of expert or character witnesses, where the number of available witnesses is "staggering." 22 Charles A. Wright & Kenneth W. Graham, Jr., Federal Practice and Procedure, § 5220 (1978). However in a class or opt-in action, the

present the oral testimony of a maximum of ten class
members with regard to issues of liability only. Oral tes-

number of available witnesses, if not "staggering" can certainy be Ulnageably large, and the need for limits
has been recognized:

tirnony may be used for the presentation of anecdotal
evidence, or to establish the factual basis underlying sta-

tistical evidence. On written application and for good cause shown the Cour may alter, in a symetrical fashion, the precise number of class member witnesses allowed to present such oral testimony.

(HN7) Pretrial planning should, however, also be given to the proper presentation of
"anecdotal" evidence regarding the individual experiences of various employees,

In this first phase, each side may also offer the testirnony by deposition (* 17) of other class members with regard to common or individual issues of liability. Such testimony is limited only by the usual constraint on cumulative evidence established by Rule 403 of the Federal
Rules of Evidence.

union stewards, supervisors, and managers. Such evidence may be offered by the plaintiffs or defendants to provide ilustrative support for their respective positions

and for the studies conducted by their experts and, indeed, may be critical to a decision on the merits of the claim of class
discrimination. Some limits, however,

The jur wil then deliberate on the question of liability. Should the jury fmd a pattern or practice of dis-

may be needed on the number of such

Case 1:03-cv-02485-MSK-PAC

Document 380-7

Filed 03/29/2006

Page 8 of 8
Page 7

1994 U.S. Dist. LEXIS 16511, *; 77 Fair Empl. Prac. Cas. (BNA) 370

persons the parties may call as witnesses. .

Jr., Federal Practice and Procedure, § 5220 (1978) (fairness requires that if the number of witnesses is limited, the limitation should be applied equally to all parties, to the same issue of fact and only after notice to the parties).

Manual for Complex Litigation, Second, (*19) § 33.53
(emphasis supplied).

have been placed on the testimony of percipient wit-

(HN8) While infrequent, prospective limitations

In any event, no prospective limitations are being placed on the amount of (*20) testimony generally, but

nesses, see, e.g., Donovan v. Burger King Corp., 672 F.2d 221,225 (1st Cir. 1982) (affirngjudge's decision in pretrial conference to limit witnesses in Fair Labor
Standards Act action to representative sample that would "give me a feeling for what is going on in these places.");
Hamling v. United States, 418 Us. 87, 129, 411. Ed. 2d

only on the form of some of the testimony. The Cour recognizes the preference for oral testimony over testimony by deposition, but anticipates that, should liability be established, damage testimony by deposition would
not prove an area of great contention.

590, 94 S. Ct. 2887 (1974) ("The District Cour retains considerable latitude even with admittedly relevant evidence in rejecting that which is cumulative, and in requiring that which is to be brought to the jury's attention
to be done so in a manner least likely to confuse that

Dated: November 10, 1994 New York, New York
SO ORDERED.

LAWRENCE M. McKENNA
U.S.D.J.

body."); 22 Charles A. Wright & Kenneth W. Graham,