Correspondence from Rodney to Counsel; Report of Silas Lee, III

    95-9022L
IN THE UNITED STATES COURT OF APPEALS 

FOR THE SECOND CIRCUIT

Docket Nos. S5-9022L & 95-9024XAP
AMERICAN FEDERATION OF STATE, COUNTY AND MUNICIPAL EMPLOYEES, 
AFL-CIO (AFSCME); the Civil Service Employees Association, 
Inc., Region I/Local 1000, (CSEA); Rita Wallace, Rachel 
Braver, Dorothy Garage, Linda Kelly, and Lois Whitely on 
behalf of themselves and all others similarly situated,

Plaintiffs-Counter-Defendants-Appellants-Cross-Appellees,
Odessa Colvin; Erna Fluhr; Laurie Gillibertie; Stephan 

Goldberg; Fred Jordan,
Plaintiffs-Counter-Defendants,

v.
COUNTY OF NASSAU; its County Executive; the Comptroller; 

the Members of the County Board of Supervisors; 
and the Members of the Civil Service Commission,

Defendants-Counter-Claimants-Appellees-Cross-Appellants.

ON APPEAL FROM THE UNITED STATES DISTRICT COURT 
FOR THE EASTERN DISTRICT OF NEW YORK

BRIEF OF PLAINTIFFS-APPELLANTS

Paul M. Smith 
Susan R. Podolsky 
Sean H. Donahue 
JENNER & BLOCK 
601 13th St., N.W. 
Washington, D.C. 20005 
(202) 639-6000
Attorneys for

Plaintiffs-Appellants



RULE 26.1 STATEMENT

Appellant American Federation of State, County and 

Municipal Employees, AFL-CIO ("AFSCME") has no parent 
companies or subsidiaries. Appellant Civil Service Employees 

Association, Inc., Region I/Local 1000, an affiliate of 

AFSCME, also has no parent companies or subsidiaries.



LAW O F F IC E S

J e n n e r  &  B l o c k

CHICAGO, IL 60611 
(312) 222-9350

CHICAGO OFFICE 
ONE IBM PLAZA

A PARTNERSHIP INCLUDING PROFESSIONAL CORPORATIONS

601 T H IR T E E N T H  S T R E E T ,  N.W. 
S U IT E  1200

W A SH IN G T O N , D.C. 2 0 0 0 5

MIAMI OFFICE 
ONE BISCAYNE TOWER 

MIAMI, FL 3313! 
(305) 530-3535 

(305) 530-0008 FAX

( 2 0 2 )  6 3 9 - 6 0 0 0  

( 2 0 2 )  6 3 9 - 6 0 6 6  F A X

LAKE FOREST OFFICE 
ONE WESTMINSTER PLACE 
LAKE FOREST, IL 60045 

(708) 295-9200 
(708) 295-7810 FAX

November 17, 1995

Charles Steven Ralston 
David T. Goldberg 
Paul K. Sonn 
NAACP Legal Defense and 

Educational Fund, Inc. 
Suite 1600 
99 Hudson Street 
New York, New York 
10013

Re: AFSCME, et al. v. County of Nassau, et al.,
Nos. 95-9022 & 95-9024

Dear Steve, David, and Paul:
Enclosed, a copy of our brief in this case. Thank you 

all very much for supporting AFSCME with such an excellent 
brief and for orchestrating things with your fellow amici.

Sincere.! v

£aul M. Smith



LAW O F F IC E S

J e n n e r  &  B l o c k
CHICAGO OFFICE 
ONE IBM PLAZA 

CHICAGO, i L 60611 
(312) 222-9350

A P A R T N E R S H I P  I N C L U D I N G  P R O F E S S I O N A L  C O R P O R A T I O N S MIAMI OFFICE
ONE BISCAYNE TOWER 

MIAMI, FL 33131 
(305) 530-3535 

(305) 530-0008 FAX

601 T H IR T E E N T H  S T R E E T ,  N.W. 
S U IT E  1200

W A SH IN G T O N , D.C. 2 0 0 0 5
LAKE FOREST OFFICE 

ONE WESTMINSTER PLACE 
LAKE FOREST, IL 60045 

(70S) 295-9200 
(708) 295-7810 FAX

( 2 0 2 )  6 3 9 - 6 0 0 0  

( 2 0 2 )  6 3 9 - 6 0 6 6  F A X

November 17, 1995

Charles Steven Ralston 
David T. Goldberg 
Paul K. Sonn 
NAACP Legal Defense and 

Educational Fund, Inc. 
Suite 1600 
99 Hudson Street 
New York, New York 
10013

Re: AFSCME, et al. v. County of Nassau, et al,,
Nos. 95-9022 & 95-9024

Dear Steve, David, and Paul:

Enclosed, a copy of our brief in this case. Thank you 
all very much for supporting AFSCME with such an excellent 
brief and for orchestrating things with your fellow amici.

£aul M. Smith



LAW O F F IC E S

CHICAGO OFFICE 
ONE IBM PLAZA 

CHICAGO, IL 60611 
(312) 222 - 8350 

(312) 527-0484 FAX

J e n n e r  &  B l o c k
A P A R T N E R S H I P  I N C L U D I N G  P R O F E S S I O N A L  C O R P O R A T I O N S

6 0 !  T H IR T E E N T H  S T R E E T ,  N.W. 
S U IT E  1200

W A SH IN G T O N , 0 .C .2 O O O 5
(202) 633-6000 

(202) 639-6066 FAX

MIAMI OFFICE 
ONE BISCAYNE TOWER 

MIAMI, FL 33131 
(305) 530-3535 

(305) 530-0006 FAX

LAKE FOREST OFFICE 
ONE WESTMINSTER PLACE 
LAKE FOREST, IL 60045 

(708) 295-9200 
(706) 295-7810 FAX

November 17, 1995

Charles Steven Ralston 
David T. Goldberg 
Paul K. Sonn 
NAACP Legal Defense and 

Educational Fund, Inc. 
Suite 1600 
99 Hudson Street 
New York, New York 
10013

Re: AFSCME, et al. v. County of Nassau, et al..
Nos. 95-9022 & 95-9024

Dear Steve, David, and Paul:

Enclosed, a copy of our brief in this case. Thank you 
all very much for supporting AFSCME with such an excellent 
brief and for orchestrating things with your fellow amici.

Sincerely,
/

/
^aul M. Smith



TABLE OF CONTENTS
PAGE

TABLE OF AUTHORITIES . ................................... ii
STATEMENT OF JURISDICTION . . . . . . . .  ........  . . . .  1

STATEMENT OF ISSUES........ ............ ...................2
STATEMENT OF THE C A S E ...................................... 2
SUMMARY OF ARGUMENT ........... . . . . . . . . . . . . .  15

ARGUMENT ...............  . . . . . . . . . . . . . . .  17
I. Under Well-Settled Law Governing Attorney Fee 

Awards against Civil Rights Plaintiffs, No
Award was Justified Here .........................17

A. The District Court Applied an Incorrect
Legal Standard.......... .................. 17

1. The Christiansburg Standard . . . . .  17

2. The Standard Applied H e r e .............. 20
B. This Case was Far from Frivolous or

Unreasonable.........................   29
II. The Award of Expert Fees was Unauthorized. . . 39

III. The Amount of Fees Awarded Was an Abuse of
D i s c r e t i o n ............   45

A. No Fees Should have been Awarded for 
Phases of the Case Prior to the Point 
When it Became Unreasonable for 
Plaintiffs to Pursue their Claims
Further ..............................   46

B. The District Court Erred by Assessing 
Fees Against Plaintiffs Relating to 
Claims on Which Plaintiffs Prevailed and
to Claims Not Found to be Frivolous . . .  48

CONCLUSION 50



TABLE OF AUTHORITIES

CASES
Badillo v. Central Steel & Wire Co., 717 F.2d 1160

(7th Cir. 1983) ......................................... 24

Bradley v. School Board, 416 U.S. 696 (1974) ............. . 42

Carrion v. Yeshiva University. 535 F.2d 722 (2d Cir.
1976) ..................... ....................... .. 25

Christiansburg Garment Co. v. EEOC, 434 U.S. 412
(1978) ........................ . . . . . .  ........  passim

Coates v. Bechtel. 811 F.2d 1045 (7th Cir. 1987) . . . . . .  19
Coleman v. McLaren. 631 F. Supp. 763 (N.D. 111. 1986) . . . .  47 

Colombrito v. Kelly. 764 F.2d 122 (2d Cir. 1985) . . 19, 23, 47 
EEOC v. Bruno's Restaurant. 13 F.3d 285 (9th Cir. 1993) . . .  22

EEOC v. Jordan Graphics. Inc.. 769 F. Supp. 1357
(W.D.N.C. 1991) ............. . . . . . . . . . . . . .  47

EEOC v. Reichhold Chems., Inc.. 988 F.2d 1564 (11th
Cir. 1993) .......................... . . . . . . . . . .  19

EEOC v. Sears Roebuck and Co.. 1987 U.S. Dist. LEXIS
4281 (N.D. 111. 1987)    27

EEOC v. Tarrant Distributors. Inc.. 750 F.2d 1249
(5th Cir. 1984)    26

Eastway Const. Corp. v. New York, 762 F.2d 243 (2d
Cir. 1 9 8 5 ) ..........................................   25

Edwards v. Interboro Institute, 840 F. Supp. 222
(E.D.N.Y. 1994) ........  . ...............  . . . . . .  20

Ekanem v. Health & Hosp. Corp.. 724 F.2d 563 (7th
Cir. 1 9 8 3 ) .....................     26

Faraci v. Hickey-Freeman Co., 607 F.2d 1025 (2d Cir.
1979)        25

Fort v. Roadway Express. Inc.. 746 F.2d 744 (11th
Cir. 1 9 8 4 ) .........................     26

Foster v. Mydas Assoc., Inc,, 943 F.2d 139 (1st Cir.
1991)    19

- I X -



Freidrich v. City of Chicago, 888 F.2d 511 (7th Cir.
1989)     41

General Camera Corp, v. Urban Dev. Corp.. 734 F.2d
468 (2d Cir. 1984)    25

Gerena-Valentin v. Koch. 739 F.2d 755 (2d Cir. 1984) . . . .  25

Glvmph v. Spartanburg General Hosp., 783 F.2d 476
(4th Cir. 1986) ....................................  19, 26

Greenberg v. Hilton International Co.. 870 F.2d 926
(2d Cir. 1 9 8 9 ) ................... ................... 24, 47

Harbulak v. County of Suffolk. 654 F.2d 194 (2d Cir.
1981)  25

Hensley v. Eckerhart, 461 U.S. 424 (1983) ................ 19, 49

Hughes v. Rowe, 449 U.S. 5 (1980) .................. 19, 23

James v. Stockham Valves & Fitting Co., 559 F.2d 310 
(5th Cir. 1977), cert, denied. 434 U.S. 1034
(1978)       5

Jane L. v. Bangerter, 61 F.3d 1505, 1513-17 (10th
Cir. 1995) ...................... ............ .. 19, 26

Jones v. Continental Corp., 789 F.2d 1225 (6th Cir.
1986)    45

Kaimowitz v. Howard. 547 F. Supp. 1345 (E.D. Mich.
1982)    49

Landgraf v. USI Film Prods., 114 S. Ct. 1483 (1994) . . . passim

Lane v. Sotheby Parke Bernet. Inc.. 758 F.2d 71 (2d
Cir. 1985)       47

Le Beau v. Libbev-Owens-Ford Co.. 799 F.2d 1152 (7th
Cir. 1986), cert, denied, 484 U.S. 815 (1987) . . . 21, 26

Lewis v. Coughlin. 801 F.2d 570 (2d Cir. 1986) . . . . . . .  24

Marquart v. Lodge 837. Intern. Ass'n of Mach. & Aero,
Workers. 26 F.3d 842, 853-54 (8th Cir. 1 9 9 4 ) ..........  19

Milwe v. Cavuoto. 653 F.2d 80 (2d Cir. 1981) . . . . . . . .  23

Mitchell v. Office of Los Ang. Cty. Super, of Sch.,
805 F .2d 844 (9th Cir. 1986), cert, denied. 484
U.S. 858 (1987) .............................. 19, 21

Mylett v. Jeane. 910 F.2d 296 (5th Cir. 1990) . . ........  . 19
- iii-



Newman v. Piggie Park Enterprises. Inc., 390 U.S. 400
(1968) ........... . . . . . . . .  .....................  17

Noyes v. Channel Products. Inc.. 935 F.2d 806 (6th
Cir. 1991) .   19

Parks v. Watson. 716 F.2d 646 (9th Cir. 1983) . . .  ........  23

Prate v. Freedman. 583 F.2d 42 (2d Cir. 1 9 7 8 ) ............... 25
Rivers v. Roadway Express, Inc,. 114 S. Ct. 1510 (1994) . 40, 41

Roadway Express. Inc, v. Piper. 447 U.S. 752 (1980) . . . . .  18

Rounseville v. Zahl. 13 F.3d 625, 632 (2d Cir. 1994) . . 18, 19
Shipes v. Trinity Indus., 31 F.3d 347 (5th Cir. 1994) . . . .  43

Sobel v. Yeshiva Univ,, 839 F.2d 18 (2d Cir. 1988) ........  33

Thomas v. First Federal Sav. Bank, 659 F. Supp. 421
(N.D. Ind. 1 9 8 7 ) ..................................   26

Vernon v. Cassadaga Valiev Cent. Sch, Dist., 49 F.3d
886 (2d Cir. 1 9 9 5 ) ........ .............................  44

West Virginia Univ. Hosps.. Inc, v. Casey, 499 U.S.
83 (1991) .................................. .. 40, 41

STATUTES
28 U.S.C. § 1291 .............................................  2
28 U.S.C. § 1331 ........... .............................  . . 1

42 U.S.C. § 1 9 8 1 ..........   41
42 U.S.C. § 1988 . . . . .  ............. . . . . . . . . .  19, 20

42 U.S.C. § 2000e-5 (k)  ..............................17, 21
Civil Rights Act of 1991, § 113, 105 Stat. 1079 . . . . .  passim 

Fed. R. Civ. P. 23(e) . . . .  ................................ 28

- iv-



MISCELLANEOUS
Annotation, Right of Defendant in Civil Rights Case 

to Receive an Award of Attorney's Fees under 
Civil Rights Attorney's Fees Awards Act of 1976 
(42 U.S.C. S 1988). 104 A.L.R. Fed. 14 (1991) . . . .  . 20

Mary F. Derfner & Arthur D. Wolf, Court Awarded Attorney's
Fees (1995).......................................... 19, 24

E. Richard Larson, Federal Court Awards of
Attorney's Fees (1981) . . . . . . . . . . . .  ........  19

H.R. Rep. No. 40(11), 102d Cong., 1st Sess. (1991) . . . . . .  19

H.R. Rep. No. 485, 101st Cong., 2d Sess. (1990)............. 19

- v-



IN THE UNITED STATES COURT OF APPEALS 
FOR THE SECOND CIRCUIT

Nos. 95-9022 & 95-9024
AMERICAN FEDERATION OF STATE, COUNTY AND MUNICIPAL EMPLOYEES, 
AFL-CIO (AFSCME); the Civil Service Employees Association, 
Inc., Region I/Local 1000, (CSEA); Rita Wallace^ Rachel 
Braver, Dorothy Garage, Linda Kelly, and Lois Whitely on 
behalf of themselves and all others similarly situated,
Plaintiffs-Counter-Defendants-Appellants-Cross-Appellees,

Odessa Colvin; Erna Fluhr; Laurie Gillibertie; Stephan 
Goldberg; Fred Goldberg,
Plaintiffs-Counter-Defendants,

v.
COUNTY OF NASSAU; its County Executive; the Comptroller; 

the Members of the County Board of Supervisors; 
and the Members of the Civil Service Commission,

Defendants - Counter-Claimants-Appellees -Cross- 
Appellants.

ON APPEAL FROM THE UNITED STATES DISTRICT COURT 
FOR THE EASTERN DISTRICT OF NEW YORK

BRIEF OF PLAINTIFFS-APPELLANTS
This is an appeal from orders entered by the 

Honorable I. Leo Glasser of the United States District Court 

for the Eastern District of New York. The order awarding 
attorneys' fees to defendants is reported at 825 F. Supp. 468 

(1993). The order awarding expert fees to defendants, dated 

May 31, 1995, is not reported. The district court's merits 

decision is reported at 799 F. Supp. 1370 (1992).
STATEMENT OF JURISDICTION 

The district court exercised federal-question 

jurisdiction under 28 U.S.C. § 1331. This Court's
- 1 -



jurisdiction is invoked under 28 U.S.C. § 1291. The district 

court entered a final judgment covering the fee issues 
presented here on October 2, 1995, and appellants filed a 

notice of appeal on October 10, 1995.
STATEMENT OF ISSUES

1. Whether the district court acted properly in 

awarding more than $1.5 million in attorneys' and expert fees 

to the defendants in this complex Title VII class action.
2. Whether, even assuming an attorneys' fee award 

was proper, the district court was correct in applying 

section 113 of the Civil Rights Act of 1991 retroactively to 

authorize an expert fee award in this case.
3. Whether it was proper to award fees relating

(1) to claims that had never been attacked as frivolous, and

(2) to phases of the case prior to the time when it is 

alleged that plaintiffs should have known that their primary 

claim was frivolous.
STATEMENT OF THE CASE

This is an appeal from orders awarding over $1.5 

million in attorney's fees, expert fees and expenses to 
Nassau County as the prevailing defendant in this Title VII 

class action. Appellants -- the American Federation of 
State, County and Municipal Employees (AFSCME), its affiliate 

representing most of Nassau County's work force, and several 

individual county employees -- filed this case in 1984, 
alleging that Nassau County had intentionally discriminated 

in setting wages for jobs performed primarily by women.

- 2 -



After a lengthy trial in 1989-90, the court ruled against 

appellants on most of their claims. 799 F. Supp. 1370.- In 
1993, the court granted defendants' motion for attorneys' 

fees and costs, 825 F. Supp. 468 (A102), but it did not enter 

judgment until 1995, when it determined that this award 
should include expert fees, see Memorandum and Order of May 

31, 1995 (A146). Appellants challenge both the 1993 and the 

1995 orders.
At trial, appellants' primary claim was that jobs 

performed mainly by women were intentionally given pay grades 

lower than they would have received based on a neutral 

application of the County's own wage-setting criteria. See 

799 F. Supp. at 1411-12 (A95-A96) (summarizing the case law 

authorizing such a claim). Those criteria were adopted in 
1967 when the County conducted a complete reformulation of 

its compensation system orchestrated by the consulting firm 

of Cresap, McCormick and Paget ("Cresap"). In the Cresap 

process, an entirely new set of job titles and job 
descriptions was developed, and salary grades were assigned 

to these jobs based on four factors: "(1) the knowledge and 

the skills required by the job; (2) the complexity and the 

variety of the duties of the job; (3) the responsibility for 

independent action; and (4) the responsibility for 

supervision." Id. at 1375 (A59). Since 1967, "[t]he Cresap

- The court's opinion is reprinted in the Opinion and Order 
Volume of the parties' Joint Appendix at page A54. In this 
brief, citations to pages in the two Transcript Volumes of 
the joint appendix are preceded by "T"; citations to pages in 
the two Exhibit Volumes are preceded by "E".

-3-



system has continued to be the primary framework for the 
classification and for the evaluation of new and existing 

jobs in Nassau County," with most of the employees remaining 

in job titles and grades established at that time. Id. at 
1377 (A61). Thus, the "present compensation system of Nassau 

County is a direct descendant of a comprehensive job and 
salary evaluation process conducted on behalf of the County" 

by Cresap. Id. at 1374 (A58).
In 1983, AFSCME conducted a study of the Nassau 

County wage system and concluded that there was a clear 

pattern of pay disparities that could not be squared with the 
Cresap criteria. See "The Wages of Inequality" (1983) (Att.

A to Pis. Motion for Class Certification). Appellants filed 

their complaint the next year, alleging that Nassau County 

discriminated on the basis of gender in salaries paid to 

employees in job classifications with more than 70 percent 

female occupants. In addition to this broad-based "pay 
equity" claim, the complaint also asserted several "equal 

pay" claims -- i .e ., claims that women in particular job 
classifications were doing the same work as that performed by 

men in other, higher-paid classifications. One example was 
the claim that there was no meaningful difference between the 

virtually all-female job of Police Communications Operator 

and the all-male job of Fire Communications Technician.

Defendants moved to dismiss the complaint for 
failure to state a claim, but this motion was denied in 1985.

-4-



609 F. Supp. 695 (A25) .2/ Two years later, the court 
certified a class of all women who, since July 28, 1982, had 

worked in county jobs that had 70 percent or more female 
incumbents. 664 F. Supp. 64 (A48). Between 1985 and 1989, 

the parties conducted extensive discovery focusing on (1) the 

nature of the Cresap process job classification and wage­

setting process, (2) the basis of any adjustments in the 

Cresap job titles and grades that had occurred between 1967 

and the 1980s, and (3) evidence pertinent to the more 

specific "equal pay" claims.
At a 16-day judge trial in 1989-90, plaintiffs 

relied primarily on two expert witnesses. The first was Dr. 

Stephan Michelson, a labor economist with extensive 

experience testifying as an expert in employment- 
discrimination cases. See E139-E154. He conducted a major 

statistical study of current wages in Nassau County, which, 

on average, were about $8000 dollars higher for employees in 

male-dominated job titles than they were for employees in 
female-dominated job titles. Using "multiple regression" 

techniques,- Dr. Michelson sought to determine how much of

- The court concluded "as a matter of law, that 
[plaintiff's complaint] state[d] claims for intentional 
discrimination under the disparate treatment theory of Title 
VII." 609 F. Supp. at 711 (A42).
- "Regression analysis is a statistical method that permits 
analysis of a group of variables simultaneously as part of an 
attempt to explain a particular phenomenon, such as earnings 
disparities between blacks and whites. The method attempts 
to isolate the effects of various factors on the phenomenon." 
James v. Stockham Valves & Fitting Co., 559 F.2d 310, 332 
(5th Cir. 1977), cert, denied, 434 U.S. 1034 (1978).

-5-



this differential could be explained by 78 gender-neutral 

variables reflecting characteristics of the jobs. He 
concluded that, while consideration of these job 

characteristics did serve to reduce the apparent salary 
differential attributable to gender, there remained a gap of 

approximately $3000, or about two salary grades, between the 

salaries of equivalent male- and female-dominated jobs. See 

799 F. Supp. at 1396 (A80).
Plaintiffs' other expert was Donald Treiman, a 

professor of sociology at UCLA and a leading expert on pay 
equity for women. See E406-E424. He studied the 1967 

Cresap job evaluation process, using several different 

methodologies. First, Dr. Treiman looked at the salary 

grades assigned by Cresap in 1967 to male- and female- 
dominated jobs that were equivalent with respect to (1) the 

years of training and experience required and (2) whether 

they involved supervision of others.- Second, he asked 

graduate students to apply the Cresap four-factor point 
system using the original Cresap descriptions of several 

hundred jobs. The results of these studies were very 
consistent: on average, the actual grades given to female-

dominated jobs in 1967 were 2-3 salary grades lower than the 

grades given to other jobs that had the same training and

- Treiman used these criteria because they could be 
objectively measured, using the Cresap job descriptions, and 
because, in his judgment, they constituted reasonable proxies 
for the Cresap job-evaluation factors. See E433.

- 6 -



experience/supervisory requirements or had been rated as 

equivalent by the graduate students. See E437, E440.
Finally, Dr. Treiman undertook a statistical 

analysis of the point scores for jobs listed on plaintiffs' 

exhibit 616 (E1-E33) -- a document that was generated during

the Cresap job-evaluation process in the 1960s and was 
identified at trial by one of the participants in the Cresap 

process as reflecting the actual conclusions reached by the 

Cresap team about the application of the point factors to 
each job title. See T591. Dr. Treiman converted the points 
on exhibit 616 to grades and compared the results with the 

actual grades assigned in 1967. He found that the actual 
grades assigned were much more favorable to male-dominated 

jobs. As a result, the average grade differential between 

female-dominated and other jobs with equivalent training and 

experience requirements and supervisory duties increased by 

approximately 1.6 salary grades. See E505. In sum, Dr. 
Treiman's three studies of the Cresap process corresponded 

closely to Dr. Michelson's extensive analysis of current 

salaries, with both experts testifying that there was a pay 

differential between predominantly female jobs and 
predominantly male jobs, in the range of 1.5 to 2.5 salary 

grades, that (1) could not be explained by job 
characteristics and (2) apparently had its origin in the 
discriminatory application of the Cresap criteria for setting 

pay based on job characteristics.

-7-



The defendants countered this evidence with two

expert witnesses of their own -- Dr. David P. Jones, an 

industrial psychologist, and Dr. Joan Haworth, a labor 

economist and statistician. These experts, while raising 
some questions about the Treiman and Michelson methodologies, 

never disputed the fact that, if one looked only at the 
intrinsic characteristics of jobs, female-dominated jobs in 

Nassau County were not paid on the same basis as male- 
dominated jobs. Indeed, their own rating system (using 
information about job characteristics derived from the 

Dictionary of Occupational Titles)), itself produced the 
conclusion that all-female jobs in Nassau County were paid on 

average $1900 less than otherwise equivalent all-male jobs.- 

Instead, defendants' main statistical defense 

involved the addition to the regression equations of 

variables said to represent the "market," which had the 

effect of eliminating significant differentials between 

salaries paid to male- and female-dominated jobs in the 
various regression equations presented by all four experts.-

- This result is shown on the table as a value of 19 for 
the variable "Pet." in the 1986 regression equation in 
Defendants' Exhibit GGG-1, Table 14 (E605). See also id. 
Table 13 (E602-E603) (replicating Dr. Treiman's study using 
training and experience and supervisory responsibility as 
variables and finding a 1986 differential of $2300 between 
equivalent all-male and all-female jobs); Def. Exh. KKK-12 
(E935) (summarizing all of these results, including 
defendants' replication of Dr. Michelson's analysis, and 
producing similar results).

- Defendants' experts also added other variables, 
representing "working conditions," hours worked per week, and 
jobs' salary grades in the pre-1967, pre-Cresap grade

(continued...)
- 8 -



These measures of the market were derived in two ways. In 
some of their studies of current salaries, the "market" rate 

was the average salary paid for comparable jobs by 
Westchester County, Rockland County, Suffolk County and New 

York City. For all studies of the Cresap results, as well as 

some other studies of current salaries, Jones and Haworth 
"estimated" the market rate for each job, using formulae in 

which the only variable was the salary assigned to that job 

by Nassau County.-

In addition to the expert testimony summarized 

above, the evidence at trial focused on (1) the details of 
the Cresap process, (2) the County's wage-setting criteria 

and responsiveness to market forces in the years after the 

Cresap study, and (3) the County's past history of 
segregating some jobs by gender and the adequacy of its 

affirmative efforts to integrate its work force. Plaintiffs 

also presented evidence concerning the equal pay claims, 
attempting to show that female employees in several job

- (...continued)
structure. These variables, however, either proved to be 
statistically insignificant or were rejected by the district 
court as representing factors the County never actually 
considered in wage-setting, see 799 F. Supp. at 1391, 1401 
(A75, A85) .
- In other words, they drew the conclusion that the Cresap 
salaries were tied to the "market" by using a market rate 
that was itself derived from those same Cresap salaries. See 
T998-T1000 (Dr. Jones); Def. Exh. KKK-1, at 9, 43 (E867,
E901). And the same essentially circular method of 
calculating a market variable based on the Nassau grade was 
used in all replications of Dr. Michelson's regressions 
except for regressions relating to the 79 job titles for 
which actual salary data from other jurisdictions was 
available. See T802-T803 (Dr. Haworth).

-9-



titles did essentially the same work as male employees in 

higher-paid titles.
In 1992, the district court issued a lengthy 

opinion, ruling for the defendants on the broad pay-equity 

claim and on all of the equal pay claims except one - - 
involving female "police detention aides" and male police 

officers. The court did not question plaintiffs' legal 
theory, and was at times critical of the defenses mounted by 

Nassau County, but it reserved its strongest criticism for 
plaintiffs' expert witnesses -- labelling them, at various 

points, as "evasive" and as "partisans." It found 
methodological flaws in Dr. Treiman's "training and 
experience" and "graduate student" studies that, in its view., 

rendered them unreliable.- With respect to Dr. Treiman's 

analysis of Exhibit 616, the court refused to credit the 

testimony at trial that this document reflected the results 

of the point evaluation of jobs in the Cresap process. It 
concluded instead that the exhibit "plainly captures one -- 

or perhaps many -- of the moments in that process; but the 
court has no satisfactory basis on which to conclude that 

[the exhibit] is the last moment -- or even a significant 

moment of that process." 799 F. Supp. at 1388 (A72)
(emphasis in original). The court also expressed serious

- The court simultaneously saw no basis for relying on Dr. 
Jones' responses to these Treiman studies -- i.e., the 
analyses where he took versions of Dr. Treiman's regression 
equations and added variables supposedly representing the 
market or hours worked. 799 F. Supp. at 1385-86 (A68-69).
It pointed out that there was no evidence that these factors 
played any role in the Cresap process. Id.

- 10 -



reservations about the methods used by Dr. Treiman to select 
data from this exhibit, concluding that these methods were so 
deficient that they fatally undercut his analysis.-

With respect to Dr. Michelson's work, the court 
noted a variety of concerns, but gave primary emphasis to Dr. 

Michelson's failure to include in his analyses a variable or 
variables reflecting the "market" wage for each job. The 

court credited Dr. Haworth's testimony that "there is indeed 
a gap between the salaries of men and women in Nassau County 
that is not explained by job specifications alone," but that 
"the sex-related salary differential identified by Dr. 

Michelson is reduced to virtual insignificance when a market 
variable is introduced." Id. at 1401 (A85).

Turning to the non-statistical evidence, the court 
found that Nassau County's work force was in fact highly sex- 

segregated but that formal barriers to integration had been 

eliminated by the 1970s and that the "better explanation" for 

the current situation is that men and women prefer different 
jobs. Id. at 1404 (A88). While the County had created an 

affirmative action committee, it had invested the committee 
with neither the "resources nor the authority to undertake 

genuinely affirmative action." Id. at 1406 (A90). But the 
court added that this "apparent indifference" was not

Here again, however, the court gave little or no credence 
to_defendants' experts' responses to the exhibit 616 study, 
which had attempted to explain the grade changes at issue by 
reference to "market" data and the pre-Cresap grades of jobs. 
The problem, once more, was the complete absence of evidence 
that anyone in 1967 actually relied on such factors in 
adjusting grades. See 799 F. Supp. at 1391 (A75).

- 11 -



probative on the issue of intentional discrimination. Id.
Moreover, the court was not persuaded that plaintiffs had 
proved disparate treatment with respect to upgrades of jobs 
since 1967. While more male-dominated jobs than female- 

dominated jobs had received upgrades (reflecting changed 

duties, recruitment problems, or both), other efforts had 

been made to augment salaries in female-dominated jobs when 
these same factors arose.—

Thus, while acknowledging that plaintiffs had 
articulated a legally valid claim, the court held that they 
had not presented sufficient evidence to carry their burden 

of proof with respect to the broad claim of discrimination 
against all workers in predominantly female job titles. The 

court did, however, rule that plaintiffs had proved 

intentional discrimination based on gender with respect to 
one of their three "equal pay" claims. It held that the 

duties of employees in the entirely female job title of 

police detention aide and the duties of male police officers 

assigned as "turnkeys" were "substantially similar and that 

the salary disparity between these two groups is attributable 
to intentional discrimination." Id. at 1409 (A93). Because 

of a large wage disparity and the time period involved, this 

ruling led to the award of more than $1.6 million to 17 women

- When the County had difficulty with recruitment or 
retention in female-dominated jobs in the 1970s and 1980s, it 
tended to take steps more limited than an upgrade of a job 
title --steps such as temporarily hiring into a higher 
"step" within the established salary grade or agreeing to 
promote a group of incumbents into a higher job title. See 
799 F. Supp. at 1392-95 (A76-A79).

- 12 -



who have served as police detention aides. See A109-A113; 

A117-A122. But the court rejected plaintiffs' other two 

equal-pay claims: the police communications operator/fire 
communications technician claim and another claim involving 

female correctional center clerks and male correction 

officers assigned to clerical work. 799 F. Supp. at 1407-08, 
1409-10 (A91-A92; A93-A94).

After this decision, the defendants moved for an 
award of attorneys fees and costs. They argued that, as of 

the time the case was ready for trial, plaintiffs should have 
known their claims lacked merit and should not have proceeded 
further. On this basis, defendants argued that they had 

satisfied the standard for fee awards to Title VII defendants 
set forth in Christiansbura Garment Co. v. EEOC. 434 U.S. 412 

(1978), requiring, as a predicate for any fee award to a 

prevailing defendant, a finding that the case was "frivolous, 
unreasonable, or groundless."

In 1993, the district court granted defendants' 
motion. 825 F. Supp. 468 (A102). In its opinion, the court 

began with a critique of the Christiansbura standard, 

suggesting that the Supreme Court had departed from the 

"plain meaning of the statute derived from the unambiguous 
words" when it "engrafted" on the statute a distinction 

between the standard applicable to plaintiffs' fee awards and 

that governing defendants' fee awards. Id. at 469-70 (A103- 
■̂■1(-)4) . See also id. at 472 (A106) . The court then proceeded 

to quote at length from its own prior critique of plaintiffs'

-13 -



experts and concluded that, overall, plaintiffs' evidence 
"fell considerably short of the mark." id.

In determining whether the case was "frivolous" 
under Christiansburg, the court drew a distinction between 

(1) a case brought by a "modestly salaried employee or group 

of employees" against a wealthy defendant, and (2) this case, 
involving a major union which invested in costly expert 

studies and had "very able and experienced counsel." id. at 

473 (A107). Then, recognizing its obligation to avoid "post 
hoc reasoning" in the application of the Christiansburg 

standard, the court nevertheless stated that it was necessary 
to base its ruling on a "kaleidoscopic view of the entire 

case which was possible only from the vantage point of time." 

id. From that vantage point, the court concluded that a fee 

award would serve the purposes of the statute, and awarded 

all of the $982,407.23 in attorneys fees requested by the 

defendants, covering all of their work since the complaint 
was filed in 1984.

The court deferred decision on the issue of expert 
fees in 1993, because the only basis for an award of expert 
fees was section 113 of the Civil Rights Act of 1991, the 

retroactivity of which was then pending before the Supreme 
Court. Subsequently, in Landqraf v. USI Film Procis.. 114 

S.Ct. 1483 (1994), the Supreme Court held that other key 

provisions of the Civil Rights Act of 1991 apply only 

prospectively. Nevertheless, on May 31, 1995, the court 

issued an order distinguishing Landqraf and applying section

- 1 4 -



113 of the Act retroactively so as to entitle defendants to 
an additional award of $550,974.66 in expert fees. (A146).

SUMMARY OF ARGUMENT

1. The decision below was an egregious departure 
from settled law governing attorney's fee awards in Title VII 

cases. Title VII plaintiffs may not be assessed fees unless 

their claims are "frivolous, unreasonable, or groundless." 
Christiansburg Garment Co. v. EEOC. 434 U.S. 412 (1978). In 

applying this standard, moreover, courts are directed not to 
engage in "post hoc reasoning." The district court, however, 

assessed plaintiffs' case from an avowedly after-the-fact 
perspective. Moreover, it improperly gave great weight to 

the fact that the lead plaintiff was a large national union 

with substantial resources. The court made no effort to 
square its decision with the large body of case law 

elaborating the Christiansburg rule. And it ultimately chose 
to award fees based on its determination that plaintiffs' 
experts were too "partisan" -- which is not the relevant 
inquiry under Christiansbura.

If the district court had applied the correct 
standard, it could not have awarded fees to the prevailing 

defendants here. Plaintiffs presented substantial evidence 
in support of claims that the court itself found to be 

legally valid. Plaintiffs' primary contention -- that the 

wage disparity between "women's jobs" and "men's jobs" could 

not be explained based on the duties and requirements of the

-15



job themselves -- went essentially undisputed at trial.

While the court found methodological flaws in the work of 

plaintiffs' experts, these flaws could not have been designed 
to "slant" the results. Indeed, the only criticism of the 
experts' methods that was shown to have affected their 

conclusions was Dr. Michelson's omission of a "market" 

variable from his regressions - - a  decision that was 

certainly reasonable in a case where the defendant purported 
to have based its wage scale on job characteristics rather 
than the market. In sum, this was a case in which, had the 
court been willing to draw certain entirely plausible 

inferences, a judgment for plaintiffs would have been 

justified. It did not even approach the kinds of frivolous, 

abusive lawsuits that have been found to justify awards of 
fees to defendants.

2. The district court also erred when it held 
plaintiffs liable for expert fees under Section 113 of the 

Civil Rights Act of 1991, which became law nearly two years 
after the completion of the trial in this case. The court 
failed to apply the Landgraf presumption against 

retroactivity and failed to follow the Supreme Court's 
teachings on new procedural measures and retroactive 
application of legal sanctions.

3. Even assuming the district court correctly 
deemed plaintiffs' presentation at trial to be frivolous, the 
amount of fees awarded was an abuse of discretion. The court 

should not have assessed fees attributable to phases of the

-16-



litigation completed before plaintiffs reasonably could have 

known of the insufficiency of their claims. As defendants 
acknowledged below, the defects in plaintiffs' claims became 

apparent, if at all, only after the close of discovery. The 

court also erred when it assessed fees attributable to parts 
of plaintiffs' case that were demonstrably not frivolous, 
including a claim on which plaintiffs prevailed.

ARGUMENT

I. Under Well-Settled Law Governing Attorney Fee Awards
against Civil Rights Plaintiffs, No Award was Justified Here.
A * The District Court Applied an Incorrect Legal 

Standard.

1. The Christiansburg Standard. Title VII of the
Civil Rights Act of 1964 gives district courts the discretion 

to "allow the prevailing party . . .  a reasonable attorney's 

fee." 42 U.S.C. § 2000e-5(k). Plaintiffs who have proven out 
their claims of employment discrimination are "ordinarily" 
entitled to recover their attorney's fees under this 

provision. See Newman v. Piggie Park Enterprises. Inc.. 390 
U.S. 400, 402 (1968).

In Christiansburg Garment Co. v, EEOC. 434 U.S. 412 
(1978), the Supreme Court explained that "two strong 

equitable considerations" favoring fee awards to a prevailing 
Title VII plaintiff are "wholly absent in the case of a 

prevailing Title VII defendant." Id. at 418. First, a Title 
VII plaintiff is "the chosen instrument of Congress to 

vindicate 'a policy that Congress considered of the highest
-17-



priority.'" Id. (quoting Newman. 390 U.S. at 402). Second, 
"when a district court awards counsel fees to a prevailing 
plaintiff, it is awarding them against a violator of federal 

law." Id. Because of these considerations, a Title VII 

plaintiff "should not be assessed his opponent's attorney's 

fees unless a court finds that his claim was frivolous, 

unreasonable, or groundless, or that the plaintiff continued 

to litigate after it clearly became so." Id. at 424, This 
test is designed to "encourage suits by victims of 
discrimination while deterring frivolous litigation."
Roadway Express. Inc, v. Piper. 447 U.S. 752, 762 (1980).

The inquiry under Christiansbura "generally turns 
on whether the claim itself is clearly meritless."

Rounseville v. Zahl, 13 F.3d 625, 632 (2d Cir. 1994). The 

Christiansbura Court specifically warned against the 

"understandable temptation" of a trial court that has ruled 

against a Title VII plaintiff on the merits to use "hindsight 
logic" and "post hoc reasoning" in deciding whether the 

plaintiff's suit was frivolous. 434 U.S. at 421-22. Such 
reasoning, the Court explained, would "discourage all but the 

most airtight claims" and "undercut the efforts of Congress 
to promote the vigorous enforcement of Title VII." id. at 
422. Instead, courts should examine whether it was 

reasonable for the losing plaintiff to have seen promise in 

his claims "at the outset," and should recall that some cases 

do not reveal themselves as weak until " [d]ecisive facts" 
emerge at discovery or at trial. Id. Even when the legal or

-18-



factual foundation for the lawsuit initially appears

"questionable or unfavorable," the plaintiff may have had an 
"entirely reasonable ground for bringing suit." Id. at 422.

The Christiansburg standard has become a firmly 
established part of federal law.11' Mindful that the test 

for assessing fees against civil rights plaintiffs "is, and 

should remain, difficult to meet," Foster v, Mydas Assoc.. 

Inc., 943 F.2d 139, 145 (1st Cir. 1991), courts of appeals 
have frequently relied upon Christiansburg to reverse fee 
awards to prevailing defendants.— Federal courts are

— See Hensley v. Eckerhart, 461 U.S. 424, 429 n.2 (1983) 
(extending Christiansburg standard to civil rights cases 
covered by 42 U.S.C. § 1988); Hughes v. Rowe. 449 U.S. 5, 14 
(1980) (per curiam). Congress has repeatedly and 
specifically approved Christiansburg's "dual standard" of fee 
eligibility for civil rights cases. See, e,g.. H.R. Rep. No. 
40(11), 102d Cong., 1st Sess. 32 (1991) (House Report to 
Civil Rights Act of 1991, citing Christiansburg with 
approval); H.R. Rep. No. 485, 101st Cong., 2d Sess. 73 & n.7.7 
(1990) (incorporating Christiansburg standard as guide to 
fee-shifting provision of Americans with Disabilities Act). 
See also E. Richard Larson, Federal Court Awards of 
Attorney's Fees 86-87 (1981) (reviewing legislative history 
of § 1988); Hensley. 461 U.S. at 429 n.2 (same).

- See, e.g., Jane L. v, Bangerter. 61 F.3d 1505, 1513-17 
(10th Cir. 1995); Marquart v. Lodge 837, Intern. Ass'n of 
Mach. & Aero. Workers. 26 F.3d 842, 853-54 (8th Cir. 1994); • 
Rounseville v. Zahl, 13 F.3d 625, 632 (2d Cir. 1994); EEOC v. 
Reichhold Chems, , Inc.. 988 F.2d 1564, 1571 (nth Cir. 1993); 
Noyes v. Channel Products. Inc.. 935 F.2d 806, 810 (6th Cir. 
1 9 9 D ; Mylett v. Jeane, 910 F.2d 296, 299 (5th Cir. 1990); 
Coates v. Bechtel, 811 F.2d 1045, 1049 (7th Cir. 1987); 
Mitchell v. Office of Los Ana. Ctv, Super, of Sch.. 805 F.2d 
844, 847-48 (9th Cir. 1986), cert, denied. 484 U.S. 858
(1.9 87) ; Glymph v. Spartanburg General Hosp.. 783 F.2d 476,
479 (4th Cir. 1986); Colombrito v. Kelly. 764 F.2d 122, 132 
(2d Cir. 1985). See also 1 Mary F. Derfner & Arthur D. Wolf, 
Court Awarded Attorney's Fees f 10.04 at 90 (1995) (noting 
frequency of such reversals); Marquart. 26 F.3d at 848-49 
(noting rarity of fee awards against plaintiffs under 
Christiansburg).

- 19 -



properly "hesitant" to assess attorney's fees against civil 

rights plaintiffs, see Rounseville. 13 F.3d at 632; in their 
extensive experience applying Christiansburg. they have 
limited such fee awards to "truly egregious cases."—7

2. The Standard Applied Here. The district court 
discussed the Christiansburg standard, but the test it 

actually applied bore little resemblance to the one 

prescribed by Congress and elaborated upon by the Supreme 

Court. Indeed, the court below devoted long sections of its 

1993 opinion to criticizing Christiansburg as an illegitimate 
departure from the "plain meaning" of the s t a t u t e . W h e n  

it came time to follow the dictates of Christiansburg. the 

court proved to be far more receptive to a fee award for 

prevailing defendants than the Supreme Court or the scores of 

other federal courts that have applied its teachings. The

- Annotation, Right of Defendant in Civil Rights Case to 
Receive an Award of Attorney's Fees under Civil Rights 
Attorney's Fees Awards Act of 1976 (42 U.S.C. 5 1988). 104 
A.L.R. Fed. 14, 26 (1991).

- _ The court opined that the Supreme Court "engrafted" a 
plaintiff-favoring standard onto neutral statutory language, 
825 F. Supp. at 469 (A103), and that "[i]f Congress did not 
intend that the prevailing party -- whether it be plaintiff 
or defendant -- could recover attorney's fees, it would have 
authorized fees for one or the other," id. at 470 (A104).
But cf. Christiansburg. 434 U.S. at 418 (Title VII's text 
"does not even invite, let alone require" that construction). 
The district court described the Christiansburg holding as an 
example of "judicial legislation," see T1167, and a departure 
f rom^ Title VII's ’’plain meaning," 825 F. Supp. at 472 (A106) . 
Despite its affinity for "plain meaning," the court below had 
no trouble_rejecting what it called "the lexicographer's 
understanding" of the operative words of the Christiansburg 
standard. See 825 F. Supp. at 472 (A106). See also Edwards 
vh-I.nterboro Institute, 840 F. Supp. 222, 231 (E.D.N.Y. 1994) 
(Glasser, J.) _ (calling for "reexamination of Christiansburg 
m  the revealing light of experience").

- 20 -



end result was to convert 42 U.S.C. § 2000e-5(k) into a 

"general penalty provision" for unsuccessful Title VII 

plaintiffs. See Mitchell v. Office of Los Angeles Ctv. Sup.

of.Schools, 805 F .2d 844, 847 (9th Cir. 1986), cert, denied.
484 U.S. 858 (1987).

First, and perhaps most strikingly, the district 
court disregarded the Supreme Court's admonitions against 
"post hoc reasoning." See 434 U.S. at 421-22. In order "to 

bring the law into accordance with experience and justice," 

the court rejected what it called "the inflexible, sterile 
application of the 'post hoc' pronouncement." 825 F. Supp. 
at 473 (A107). Concluding that "in a case as complex as 

this, an intelligent appraisal of the Christiansburg 

standards can only be made after the dust of the litigation 

battle has settled and after the smoke from the defense 

salvos has cleared," the court assessed fees based upon "a 

kaleidoscopic view of the entire case which was possible only 
from the vantage point of time." Ibid. This was just the 

opposite of the straightforward, objective inquiry that 
Christiansburg prescribes.

The district court repeatedly relied "on the facts 
as found at trial to support a finding that the suit should 

not have been brought." Le Beau v. Libbev-Owens - Ford Co..

799 F .2d 1152, 1160 (7th Cir. 1986), cert, denied. 484 U.S. 

815 (1987). It awarded fees because plaintiffs presented 

expert witnesses who, in the court's view, were not credible, 
fair-minded witnesses. But the court never attempted to

- 21 -



explain when, how and why reasonable persons in plaintiffs' 

position would have perceived before trial that the eminent 

witnesses they had retained, and the complex statistical 

analyses those experts prepared, would prove so unpersuasive. 

If courts are required to avoid "post hoc reasoning," it 

hardly makes sense to award fees based on how credible 

plaintiffs' experts appeared to a particular judge during 

cross-examination. See EEOC v. Bruno's Restaurant. 13 F.3d 
285, 288 (9th Cir. 1993).

Second, in determining that a defendant's fee award 
"would not frustrate the purpose of the Act but would further 

it," the district court emphasized that the "real plaintiff" 

in this case was not a "modestly salaried employee or group 

of employees," but "a major union" and a "dominant force on 

the American labor scene," which "was not economically or 

otherwise disadvantaged" and "was represented by very able 

and experienced counsel who impressed the court with their 

exquisite grasp of statistics, regression analyses, and 

masses of complex computer-generated tables." 825 F. Supp. 

at 472-73 (A107-A108). Indeed, these factors apparently were 
a linchpin of the district court's decision, in view of the 

court's acknowledgement that a fee award against "modestly 
salaried" employees suing a large defendant "would, 

understandably, frustrate the purpose of the Act."

But these factors are plainly irrelevant and 
improper under Christiansbura. Neither AFSCME's size nor its 

perceived economic status have any bearing on whether its

- 22 -



claims were, objectively speaking, "clearly meritless," 

Rounseville, 13 F.3d at 632. See, e.g. . Parks v. Watson. 716 

F.2d 646, 664-65 (9th Cir. 1983) (rejecting as unsupported 

and unsound the argument espoused by the court below)

The district court's fee liability criterion would punish the 

filing of novel and challenging "impact" cases by the class 
of plaintiffs most able to bring them. No less than 

individual plaintiffs, organizations that pursue civil rights 
violators serve "a policy that Congress considered of the 

highest priority." See Christiansbura. 434 U.S. at 418. Cf. 
Colombrito v, Kelly, 764 F.2d 122, 133 (2d Cir. 1985) (citing 
NAACP v. Button. 371 U.S. 415 (1963)).^;

For similar reasons, the quality of appellants' 

counsel cannot be a relevant factor. A plaintiff's pro se 

status may militate against an assessment of fees, see Hughes

v._Rowe. 449 U.S. 5, 16 (1980), but the fee liability

standard for represented plaintiffs does not rise and fall 
with the perceived ability of their lawyers. See Milwe v. 

Cavuoto, 653 F.2d 80, 83 (2d Cir. 1981) (high quality of

- In Christiansburg itself, the Court rejected the 
suggestion that the EEOC as plaintiff should be subject to a 
less forgiving standard of fee liability because of its 
substantial resources and its statutory obligation to bring 
antidiscrimination claims. See 434 U.S. at 423 n.20.
A prevailing party's economic resources similarly do not 
affect its entitlement to fees. See Milwe v. Cavuoto. 653 
F .2d 80, 83 (2d Cir. 1981).

- _ The district court's approach, moreover, would make fee 
eligibility determinations turn on the fortuity of whether an 
organization chose to sue in its own name or instead simply 
organized a suit brought in the name of a handful of 
"modestly salaried" members.

-23-



prevailing plaintiff's counsel "irrelevant" to eligibility 

for fees). A sliding frivolity scale would eliminate the 

"objective" character of the Christiansburg standard, see 

Greenberg v. Hilton International Co.. 870 F.2d 926, 940 (2d 

Cir. 1989), and give civil rights plaintiffs a incentive to 

hire mediocre lawyers. But cf. Lewis v. Coughlin. 801 F.2d 

570, 576 (2d Cir. 1986) (fee-shifting should be administered 

so as to allow plaintiffs to "attract competent counsel").

Third, the district court, while expressing 

perplexity about the meaning of the "frivolous, unreasonable 
or without foundation" standard, 825 F. Supp. at 470, 472 

(A104, A106), neither discussed nor cited a single case 

upholding a defendant's fee award under Christiansburg. Had 

the court compared this case to those in which defendant's 
fee awards have been upheld, it would have seen that fee 

awards to defendants have generally been restricted to rather 
extreme episodes of misconduct by plaintiffs. See, e.g,. 
Badillo v. Central Steel & Wire Co.. 717 F.2d 1160, 1163-64 

(7th Cir. 1983) (reviewing case law and concluding that 

awards under Christiansburg are "limited to situations where 

plaintiff's conduct was abusive, or merely a disguised effort 

to harass or embarrass the defendant"; two characteristic 

fact-patterns are frivolous relitigations and cases in which 
the plaintiff is "aware with some degree of certainty of the 
factual or legal infirmity of his claim"); 1 Mary Frances 

Derfner & Arthur D. Wolf, Court Awarded Attorney Fees,

1 10.04 at 100-101 (1995) ("In general, a prevailing

-24-



defendant cannot recover attorney fees" under Christiansbura 

"unless the plaintiff produces no evidence to support her 

claim, or unless the plaintiff has no colorable legal 

theory."). Caselaw in this circuit follows the same 
pattern.—

Indeed, we have been unable to find a case in which 

a prevailing defendant won a fee award in circumstances even 

remotely similar to those presented here -- where plaintiffs' 
legal theory was upheld by the district both before and after 

the trial, 609 F. Supp. at 711 (A42); 799 F. Supp. at 1411-12 

(A95-A96), and the plaintiffs presented a substantial amount 

of evidence supporting that theory. Compare 825 F. Supp. at 

473 (plaintiffs' establishment of a prima facie case is 

"meaningless"), with Le Beau. 799 F.2d at 1159 (finding this

— The cases in which this Court has approved fee awards 
under Christiansburg have involved extreme episodes of 
litigation abuse, including relitigation of claims resolved 
adversely to the plaintiff in prior proceedings, see Eastway 
Const. Corp. v. New York. 762 F.2d 243 (2d Cir. 1985);
Gerena-Valentin v. Koch. 739 F.2d 755 (2d Cir. 1984); Prate 
v. Freedman. 583 F.2d 42, 46-47 (2d Cir. 1978); pursuit of an 
"entirely groundless" discrimination claim in the face of 
"uncontradicted evidence" showing both an absence of 
discrimination and a valid cause for the defendant's actions, 
Faraci v. Hickey-Freeman Co.. 607 F.2d 1025, 1027 (2d Cir. 
1979); pursuit of damages claims against a state 
notwithstanding a clear Eleventh Amendment bar, General 
Camera Corp. v. Urban Dev. Corp.. 734 F.2d 468 (2d Cir. 1984) 
(affirming fee award for claims dismissed in No. 81-civ-4080 
(S.D.N.Y. Oct. 13, 1982) (slip op. on LEXIS database)), and a 
plaintiff's pursuit of the "absurd claim" that a police 
officer violated his right to privacy by reaching into his " 
car to serve a summons, Harbulak v. Countv of Suffolk. 654 
F.2d 194, 196 (2d Cir. 1981). See also Carrion v. Yeshiva 
University. 535 F.2d 722, 728-29 (2d Cir. 1976) (pre- 
Christiansbura decision upholding fee award against plaintiff 
whose own testimony was "'an unmitigated tissue of lies'" and 
whose lawsuit was "motivated by malice and vindictiveness") 
(quoting district court).

-25-



factor determinative); EEOC v. Tarrant Distributors, Inc.,
750 F.2d 1249, 1251 (5th Cir. 1984) (same), and Thomas v. 

First Federal Sav. Bank. 659 F. Supp. 421 (N.D. Ind. 1987) 

(finding this factor highly relevant).— Perhaps the best 

illustration of how far the district court strayed from the 

mainstream is Glymph v. Spartanburg General Hosp., 783 F.2d 

476, 479-80 (4th Cir. 1986), in which the Fourth Circuit 

reversed a fee award against plaintiffs because the trial 
court had denied the defendants' dispositive motion; allowed 

the case to proceed to trial; taken a few weeks to rule on 

the merits, and issued a fifteen-page opinion. Here, the 

court awarded fees after having the case under submission on 

the merits for more than two years and then issuing a merits 

opinion covering 47 pages of the Federal Supplement. If 

plaintiffs' claims had been "truly frivolous, the district 
court would have had no need to engage in prolonged and fact- 

specific inquiries." Jane L. v. Banaerter. 61 F.3d 1505,

1514 (10th Cir. 1995) (citing Hughes v . Rowe. 449 U.S. 5, 15- 

16 (1980)). The sheer lack of precedent for a fee award in 

remotely similar circumstances is sufficient to make it clear 

that the district court's understanding of the Christiansburg

— The district court gave scant or no weight to other 
features of this case that ordinarily would militate strongly 
against a finding of frivolousness or unreasonableness, 
including its ruling, see 664 F. Supp. 64 (A48), certifying 
the plaintiff class, e.g,, Ekanem v. Health & Hosp. Corp..
724 F.2d 563, 574 (7th Cir. 1983), and the "novelty and the 
difficulty of the issues" presented in this case, Fort v. 
Roadway Express. Inc.. 746 F.2d 744, 748 (11th Cir. 1984).

-26-



standard differed dramatically from the well settled test 

that has been applied in hundreds of other cases.
Finally, the district court indicated that it saw 

"an award of reasonable attorney's fees" as a "remedy" for 

the "abuses" of expert witnesses who testified as "partisans" 
rather than objective scholars. 825 F. Supp. at 472 (A106). 

But the court below did not cite, and counsel for plaintiffs 
have not discovered, any other case in which a civil rights 

plaintiff has been assessed with attorney's fees based upon a 

court's conclusion that the plaintiffs' expert was 

insufficiently objective.— The reason is that such an 

approach bears little relationship to Christiansburg. After 

all, in almost any case where plaintiffs rely on statistics 

and the court ultimately rules against them, that ruling will 

reflect the conclusion that the plaintiffs' experts presented 

a slanted view of reality, omitting or mischaracterizing

^  In EEOC v. Sears Roebuck and Co.. 1987 U.S. Dist. LEXIS 
4281, *40 (N.D. 111. 1987), the defendant moved for fees 
against the EEOC, which had relied on statistical evidence as 
circumstantial proof of intentional discrimination. After 
recalling Christiansbura's warnings against the perils of 
hindsight, id. at *40, the court rejected the defendant's 
argument in language that is equally applicable here:

The EEOC produced statistical evidence of 
disparities in both commission sales hiring and 
promotion and checklist compensation. Only after 
hearing Sears' extensive evidence clearly refuting 
EEOC's statistics did the court ultimately find 
EEOC's statistical analyses flawed and its evidence 
of discrimination insufficient.........

EEOC engaged in extensive discovery and statistical 
analysis in preparation for trial. The defects in 
its analyses were not fully revealed until after it 
had presented its case in chief. Id. at *40-*41.

-27-



facts that would have demonstrated to an "objective" observer 

that the defendant had not discriminated. For that reason, 
if plaintiffs are not to be unduly deterred from pursuing 

substantial but risky claims, something more must be shown 

before a defendant's verdict can become the basis of a fee 
award: there must be a showing that the plaintiffs

themselves, prior to trial, could not reasonably have 
believed that they had a substantial case.

The district court's approach would cause 
particularly severe problems in class actions like this one. 

The court held, in effect, that once plaintiffs' experts 

completed their studies, plaintiffs should have scrutinized 

the expert reports, identified key flaws, and voluntarily 

dismissed the case prior to trial. But here, AFSCME and the 

other class representatives would have been required to 

provide notice to class members of their right to object to 

the dismissal and to satisfy the court that a dismissal with 

prejudice was a fair and adequate resolution of the class 

claims. See Fed. R. Civ. P. 23(e). Any objectors could have 
pointed to reports filed by reputable experts who were 

prepared to testify that the statistical evidence supported 

the plaintiffs' claims. Plaintiffs' counsel, in turn, would 
have been forced to attack the reports prepared by their own 

experts. If the court then rejected the dismissal as unfair 

to the class, counsel would have been duty-bound to try to 

resuscitate a case that they had just assailed as groundless.

-28-



This quandary further illustrates why the Christiansburcr 
standard should not be watered down.

B • This Case was Far from Frivolous or 
Unreasonable.

If the court had applied the correct standard, it 
could never have reached the conclusion that a fee award to 

the prevailing defendants was justified in this case. Far 

from preparing a "frivolous" case for trial, plaintiffs 

marshalled a great deal of evidence supporting their 

contention that Nassau County had intentionally disfavored 
jobs performed primarily by women in setting its pay grades.

To begin with, it bears emphasis that plaintiffs 
succeeded in persuading the district court that the wages 
paid for one such job title were intentionally 

discriminatory. They won over $1.6 million in relief for 17 

female police detention aides who had been hired into a 

deliberately sex-segregated job title in which they performed 

duties identical to those of male police officers who were 
paid much more. 799 F. Supp. at 1416 (A100); A109-A113; 
A117-A122. That fact alone suggests that it was not 

"frivolous" for appellants to claim that gender played a role 

in other wage-setting decisions by the same County personnel.

Moreover, while the court ultimately was not 
persuaded by plaintiffs' broader claim of systematic 

discrimination, the deficiencies in their case identified by 

the court do not even approach the kinds of problems that 

would justify a defendant's fee award. In a case of this 
kind, the bulk of the evidence supporting plaintiffs' claim

-29-



of discrimination against all predominantly female jobs 

necessarily was statistical in nature. The court, without 

questioning the eminent qualifications of plaintiffs' two 

statistical experts, found them to be singularly unpersuasive 

as witnesses. But the reality is that any flaws in these 

experts' work identified by the court either were immaterial 

(in the sense that they did not affect to outcome of the 
studies) or else involved highly debatable questions about 

the relevance of the "market" as a factor explaining the 

County's conduct. Moreover, there is no reason to believe 

that plaintiffs, their counsel, or their experts acted with 

anything less than complete good faith in putting together 

their case. In such a situation, it is insupportable to 

suggest that plaintiffs should bear the defendants' 

attorneys' fees because they made the decision to proceed to 
trial with their claims.

Plaintiffs' experts had excellent credentials. 
Professor Treiman was a tenured professor of sociology at 

UCLA, where he taught courses both in statistical method and 

in the social value attributed to particular occupations. He 

had spent two years leading a National Academy of Sciences 

study of issues relating to "comparable worth" and coediting 

a book entitled Women, Work, and Wages: Equal Pay for Jobs 

Pf Equal Value, published by the National Academy of Sciences 
in 1981. He had lectured on these issues worldwide and 

served as consultant to the State of Michigan and the 
Province of Ontario in their efforts to implement

-30-



nondiscriminatory wage scales for their public employees.

See E406-E424. For his part, Dr. Michelson was a leading 

expert on statistical analysis of employment discrimination 

and had testified as an expert for both plaintiffs and 

defendants in many prior cases. See E139-E154

As for the work of these experts, their primary 

joint conclusion was undisputed at trial. Drs. Treiman and 

Michelson testified, based on various regression analyses, 

that the differential in pay between male-dominated jobs and 
female-dominated jobs in Nassau County could not be fully 

explained on the basis of the characteristics of the jobs.

Dr. Treiman arrived at this conclusion based on studies of 

the results of the Cresap process in 1967, while Dr.

Michelson studied current salaries using 78 different 

variables drawn from job descriptions. Both found that, even 

after consideration of job characteristics, there remained a 

$2000-$3000 differential between the pay of equivalent male- 

dominated and female-dominated jobs.

Far from disproving this conclusion, defendants' 

experts confirmed it. They reached essentially the same 

results when they ran their own analyses (1) using variations 
on the Treiman and Michelson methodologies and (2) using 

their own system for evaluating job characteristics (based on 

the Dictionary of Occupational Titles). Summarizing the

— Defendants raised no objection to Dr. Michelson's 
qualifications and stipulated to Dr. Treiman's. T7, T291.

-31-



results of all of these studies, defendants' expert Dr. Joan
Haworth concluded that

no matter what you did to look at the 
specifications of a job based on what's written 
on the job description, you'll get some 
relationship, a fairly strong relationship 
between those job specifications and the salary 
that's being paid.

There's also still in all of these specifications 
a gap between men and women's salaries that has 
not been explained by the job specification 
variables. T734 (emphasis added).

See also 799 F. Supp. at 1401 (A85) (noting Dr. Haworth's

confirmation "that there is indeed a gap between the salaries

of men and women in Nassau County that is not explained by
job specifications alone").

It is with this fact in mind that one must evaluate 

the significance of the particular criticisms of plaintiffs' 

experts contained in the district court's 1992 merits 

opinion. For example, the court labeled Dr. Michelson an 

"evasive" witness, and repeatedly made an issue out of the 

fact that he had revised his report after it was initially 

submitted. It criticized him for not knowing precisely how 

his staff had "coded" a particular phrase that appears in 

some of the job specifications, as well as for a handful of 

particular coding decisions that the court found problematic. 

But none of these criticisms in any way undercuts the 

consensus at trial that characteristics of jobs expressed in 
the job specifications could not explain the gender gap in 
Nassau County salaries.

-32-



Similarly, the court expressed concerns about the 
omission from Dr. Michelson's analysis of two variables -- 

reflecting the influence of collective bargaining and 

"working conditions" on wages. Such omissions, however, are 

only significant to the extent that the omitted factors 

constitute, or are a proxy for, "actual determinants of 

salary" and it is shown that the inclusion of the omitted 

variable would have affected the apparent gender disparity. 

Sobel v. Yeshiva Univ,. 839 F.2d 18, 34-35 (2d Cir. 1988). 

Here, defendants did not and could not show that collective 
bargaining played a significant role in establishing the 

relative wage rates for particular jobs.— Nor did their 

experts even attempt to show that any such effect altered the 

overall bottom-line gender disparity. As for "working 

conditions," although the court did find that such conditions 

were considered in setting wage rates, see 799 F. Supp. at 

1399 (A83), and defendants' experts studied this issue, see 

E513-E528; E530-E544, defendants never argued that the

— The evidence showed that most job titles had their 
grades set in the 1967 Cresap process (long before County 
workers had the right to collective bargaining), that newer 
job titles were given grades based on the Cresap grades of 
similar jobs, and that only a few jobs had ever had their 
grades changed through union negotiations. Moreover, while 
it was true that police personnel had different salary 
schedules negotiated by a different union, Dr. Michelson's 
analyses included a "police" variable, which had the effect 
of isolating any wage differential caused the representation 
of police employees by a union other than AFSCME.

-33-



omission of this factor from Dr. Michelson's analyses 
affected his results.—

Ultimately, therefore, the court's dismissal of Dr. 
Michelson's study as insignificant was based on one critique 

-- the study's failure to "account for the effect of market 

forces on the present salary disparity in Nassau County."
799 F. Supp. at 1401 (A85). As the court itself 

acknowledged, id. at 1401-02 (A85-86), it was only the 

insertion of a variable supposedly representing the "market" 

that served to eliminate the large salary disparity between 
male- and female-dominated jobs in the County. It is, 

however, highly debatable whether Dr. Michelson was properly 

criticized for excluding a "market" variable from his 
regressions.

The reason Dr. Michelson did not do so was that 

there was very little basis for supposing that the market was 

an "actual determinant" of the grades of specific jobs in 

Nassau County. See T270-T271. Plaintiffs' theory was that 
the County had selected a system for assigning job grades 

based on job characteristics -- and had then selectively 
departed from the results produced by their avowed pay­

setting system. This theory was supported by several key 

facts. First, the market played no role in grade-setting in

- _Defendants' experts added to their regression equations
variables measuring the frequency of workers compensation 
claims in particular jobs and whether the job involved work 
outdoors or indoors. See, e.g.. T800 (Dr. Haworth). They 
never testified, however, that the addition of these 
variables altered the gender differential.

-34-



the 1967 Cresap process. See 799 F. Supp. at 1385 (A69)

("the court is unable to conclude that the Cresap team 

actually considered market data in setting individual salary 

grades"); id. at 1391 (A75) (finding "no satisfactory basis 

on which to conclude that the Cresap process entailed the use 

of labor market data to set specific salary grades"). 

Moreover, in the 1970s and 1980s, most jobs remained in their 

Cresap grades, id. at 1377 (A61), new jobs were assigned 

grades based on rules of thumb derived from the Cresap grade 

structure (rather than the market), id., and only a very 

small proportion of jobs (almost all of them male-dominated) 

were upgraded due to market forces - - i .e., recruitment 

and/or retention problems, id. at 1393-95 (A77-A79)

To be sure, for any job, there is a minimum salary 

that must be offered "in the marketplace" if the job is to be 

filled. But that does not mean that Dr. Michelson lacked a 

reasonable basis for omitting variables representing "the 

market" from his study. Indeed, defendants' own data 

demonstrate that any existing market constraints left Nassau 

County with ample room to discriminate. Those data show that 

salaries for the "same" jobs in Nassau and four neighboring 

jurisdictions (New York City and Suffolk, Westchester and

— Concomitantly, there was no evidence that any job had 
ever been downgraded because Cresap had pegged it above the 
market. The district court found that market data had been 
considered in decisions not to upgrade jobs. 799 F. Supp. at 
1401 (A85). But plaintiffs had a reasonable basis for 
concluding that any such consideration of market data was at 
best haphazard and was itself given disparate weight 
depending on whether the employees requesting an upgrade were 
predominantly male or predominantly female.

-35-



Rockland Counties) varied in the early 1980s by an average of 

$5000 (or 27 percent of the lowest salary for each job) --an 

average variation that was approximately twice the gender 

differential found by Dr. Michelson. See E548. There was 

more than enough "play in the joints" of the regional public 

labor market to allow Nassau County to favor male-dominated 

jobs over equivalent female-dominated jobs by $20Q0-$3000.— 

The question thus becomes what to make of the fact 

that, when defendants' experts calculated an average "going 

rate" using these widely disparate figures and plugged that 
variable into the Michelson regressions, the market variable 

was shown to be highly correlated with Nassau salaries and 

the gender variable was reduced to insignificance. This fact 

plainly does not show that Nassau actually based its salaries 

on the average salaries paid by these other four 

jurisdictions (and in fact it did not). Instead, it shows 

that other public employers, using whatever criteria they may 

have applied, also tended to overvalue traditionally male job 

titles vis-a-vis traditionally female jobs that rate as 

equivalent based on their inherent characteristics. In this 
light, it is unsurprising that the gender variable was 

reduced to insignificance. All that the so-called "market" 

variable was then measuring was the extent to which Nassau's

— In the civil service context, as long as the wages 
offered for all jobs are above the minimum required to 
recruit or retain qualified personnel, there is no mechanism 
by which market forces would constrain an employer from 
discriminating in favor of male-dominated jobs.

-36-



discriminatory treatment of predominantly female job titles 

exceeded the discriminatory treatment that existed elsewhere.
The district court's critique of Dr. Treiman's 

studies of the Cresap process was somewhat different. The 

court did not identify any variable that was omitted from his 

studies and could have served to "explain" the apparent 

gender differential he identified. Indeed, the court firmly 

rejected defendants experts' efforts to portray the Cresap 

grade-setting process as one in which something other than 

job characteristics -- the market, or pre-1967 wage rates in 
the County -- played a role. 799 F. Supp. at 1385 (A69). 

Instead, the court held that the data and methodologies used 

by Treiman were sufficiently problematic that his results 

were simply unreliable. Id. at 1385, 1390 (A69, A74).

Thus, the court was unpersuaded by Dr. Treiman's 

two efforts to "replicate" the Cresap job evaluation 

process -- finding that he had not given enough time and 

information to the graduate students who rated the jobs using 

Cresap's factors and that his study of wages of jobs with 

equivalent training and experience requirements and 

supervisory responsibilities was too limited in scope.—

- See 799 F. Supp. at 1380-85 (A64-A69). The court also 
questioned the identification of male- and female-dominated 
jobs in 1967, which, due to limitations in recordkeeping, had 
to be accomplished by looking at the first names on lists of 
incumbents. See E456 n.6. But any concerns in this regard 
were largely formalistic, for nearly all jobs in Nassau 
County (nurses, clericals, laborers, mechanics, etc.) were 
either over 90-percent male or over 90-percent female. 
Defendants never attempted to show that even a single job was 
miscategorized in the Treiman studies.

-37-



But these studies were only intended to supplement Dr. 

Michelson's comprehensive study of the relationship between 

gender, job characteristics, and wages in Nassau County.

Their results were entirely consistent with those produced by 

Dr. Michelson. And, most importantly, there is no indication 
that the flaws in these studies reflected an effort to 

distort the results in plaintiffs' favor. For example, Dr. 

Treiman's study of training and experience and supervisory 

responsibilities was limited to those factors because they 
were objectively measurable, not because consideration of 

other factors would have lessened the gender differential. 
Similarly, while it might have been advisable to give the 

graduate students more information and more time to complete 

their scoring of jobs, the design of this study can hardly 

have been affected by a desire to slant the results.

As for Dr. Treiman's study of exhibit 616 -- the 

document that plaintiffs believed to reflect the County's 

"neutral" evaluation of jobs in 1967 -- the court rejected 

that study primarily because it concluded that the exhibit 

was not "a master compilation of the final scores from the 

Cresap job evaluation process." 799 F. Supp. at 1387 (A71). 

But, at trial Vito Competiello (a participant in the Cresap 

process and still a senior County official) flatly testified 

that the exhibit did summarize the results of the Cresap job 

evaluation process. See T591. Thus, the court's finding 

about the significance of exhibit 616 was hardly something 
that the plaintiffs had an obligation to anticipate.

-38-



Moreover, while the court questioned the methods that 

plaintiffs' experts used for taking data from exhibit 616, 

there was never any indication that these methods affected 

the experts' conclusions in any way.— Here again, 

defendants' experts countered by pointing to the market or 

pre-Cresap grades to explain the County's actions, and the 

court again rejected those studies as lacking any basis in 

fact. 799 F. Supp. at 1391 (A75).

Plaintiffs' statistical analyses constituted a 

good-faith effort to prove a difficult case. While none was 

perfect, the analyses produced strikingly consistent results. 
Whatever their flaws, there is no reason to conclude that the 

studies were deliberately skewed to produce those results. 

Every step in the process was entirely open to the 

defendants, who used these data to put together what turned 

out to be effective responses. This was the adversarial 
system in action. It was not, in any sense, an abuse of that 

system by appellants or their experts.

II. The Award of Expert Fees was Unauthorized.
Even assuming the court was correct in awarding 

attorney's fees to the defendants, it still erred in awarding

— Thus, the court considered it improper that data were 
key-entered in plaintiffs' counsel's office, not Dr. 
Treiman's. 799 F. Supp. at 1389 (A73). But those data were 
supplied to defendants, who never even attempted to show any 
errors. The court also questioned the inclusion of point 
totals for jobs where the final total had been crossed out on 
the exhibit. Id. at 1387 (A71). But it was reasonable to 
include the best data available for each job, and again there 
was no showing that this decision affected the result.

-39-



over $500,000 in expert fees. The sole basis for this award 
was a section of the Civil Rights Act of 1991, passed nearly 

two years after the trial ended, that altered prior law to 

allow recovery of expert fees by prevailing parties in Title 

VII litigation. This provision should not have been applied 
retroactively in this case.

The trial in this case lasted from November 27,

1989 to May 11, 1990. At that time, prevailing parties were 

not entitled to an award of expert witness fees. See West 
Virginia Univ. Hosps., Inc, v. Casev. 499 U.S. 83 (1991) 

(interpreting 42 U.S.C. § 1988). In Section 113 of the Civil 

Rights Act of 1991, 105 Stat. 1079, which became law on 

November 21, 1991, Congress amended Title VII and 42 U.S.C.

§ 1988 to bring expert, witness fees within the category of 

compensable attorney's fees. In its 1993 ruling on 

attorney's fees, the district court noted that defendants 

would not be entitled to attorney's fees unless Section 113 

were applied "retroactively," and that under existing circuit 

precedent it would not be so applied. See 825 F. Supp. at 
474 (A108) (citing Wisdom v. Intrepid Sea-Air Space Museum. 

993 F.2d 5 (2d Cir. 1993)). The court deferred decision on 

defendants' request for expert fees pending the Supreme 

Court's resolution of the retroactivity issue.

On April 26, 1994, the Supreme Court announced its 
decisions in Landgraf, 114 S.Ct. 1483, and Rivers v. Roadway 

Express, Inc., 114 S.Ct. 1510. In Landgraf. the Court 

reaffirmed the longstanding judicial "presumption against

-40-



statutory retroactivity," 114 S. Ct. at 1508, and held that 

the new damages and jury trial provisions of the 1991 Act did 

not apply to cases arising before the statute's effective 

date. In Rivers, it held that the expansion of 42 U.S.C.

§ 1981 effected by the 1991 Act also does not apply 

retroactively.—
Notwithstanding the holdings in Landgraf and 

Rivers, the district court ultimately did award expert fees 
under the 1991 Act. See Memorandum and Order of May 31, 1995 

(A146). It reasoned that Section 113 was distinguishable 

from the provisions at issue in Landgraf because it was 
procedural, and that an expert fee award was proper because 

plaintiffs "surely knew" when they filed this suit that 

defendants would need to retain experts. See A153-A154.

The district court's ruling on expert fees was 

wrong on multiple grounds. Section 113 would plainly operate 

"retroactively" if applied against plaintiffs here, for it 

became law long after any possibly relevant reference point 

in this case, whether it be the allegedly discriminatory 

conduct at issue, the filing of the suit, discovery, or 

trial. In Landgraf, the Supreme Court noted that it had 

never, absent a clear statement from Congress, "read a 

statute substantially increasing the monetary liability of a

— Rivers. 114 S.Ct. at 1517-1519, precludes any argument 
here based upon lower court expert fee shifting decisions 
prior to West Virginia Univ. Hosps., Inc, v. Casey. 499 U.S. 
83 (1991). But cf. Memorandum and Order of May 31, 1995, 
A151 n.3, (citing Freidrich v. City of Chicago, 888 F.2d 511 
(7th Cir. 1989)) .

-41-



private party to apply to conduct occurring before the 

statute's enactment." 114 S.Ct. at 1507. The $550,974.66 

expert fee award here is a substantial liability that was not 

authorized when plaintiffs filed and litigated this action. 

The court below offered no reason to read a statutory text 

that the Landgraf Court found too indeterminate to authorize 
retroactivity, id. at 1493-96, to call for retroactive 

operation of the specific provision at issue here. Under the 

presumption against retroactive application of statutes that 

was reaffirmed in Landgraf. the expert fee award here should 
be set aside.—

The district court's reasons for distinguishing 

Landgraf rested in large part upon its misconstruction of 

dicta in the Landgraf opinion about the application of new 

"procedural" provisions to pending cases. The Landgraf Court 

observed that, because new procedural statutes "regulate 

secondary rather than primary conduct," they "may often be 

applied in suits arising before their enactment without 

raising concerns about retroactivity." 114 S. Ct. at 1502. 

The Court cautioned, however, that the applicability of new- 

procedural provisions "depends upon the posture of the

- Bradley v. School Board. 416 U.S. 696 (1974), in which 
the Court approved the application of an intervening 
attorney-fee provision in favor of prevailing school 
desegregation plaintiffs, depended on circumstances entirely 
absent here, including the Bradley plaintiffs' adjudged 
entitlement to fees under preexisting law. See Landgraf. 114 
S. Ct. at 1503. In contrast to this case, application of the 
new fee statute in Bradley simply "'did not impose an ... 
additional burden'" on the party held liable for fees. Id. 
(quoting Bradley. 416 U.S. at 721).

-42-



particular case," and that procedural provisions as well as

substantive ones can raise retroactivity concerns. Id, at 
1502-1503 n.29. Thus, while a new jury trial provision would 

normally apply to cases tried after its effective date (even 

ones that arose before that date), its enactment "would 

ordinarily not warrant retrial of cases that had previously 
been tried to a judge." Id. at 1505 & n. 34.

Assuming that Section 113 is "procedural" for 

retroactivity purposes, under Landgraf it is not dispositive 

that this case arose before Section 113 became law. See id. 
at 1499.- However, and as the district court failed to 

recognize, the retroactivity problem here is not that this 

case arose before enactment of the 1991 Act; it is that the 

trial of this action (and with it all activities involving 

expert witnesses) ended long before that point. The Landgraf 

dicta on procedural retroactivity yield a plain answer to 

this problem, one that requires reversal here. The expert 

testimony in this case was developed and completed "under the 

old regime," Landgraf. 114 S. Ct. at 1502 n.29; see id̂ _ at 
n.34, a regime in which expert fees awards were not 

authorized. Retroactivity principles demand that defendants' 
motion be adjudicated under that regime.- The Landgraf

~ If Section 113 is "substantive," then it does not apply 
here because this case arose long before the enactment of the 
1991 Act.

C_f. Shipes v. Trinity Indus. . 31 F.3d 347 (5th Cir.
1994) (holding that, under Landgraf. applicability of new 
procedural rules depends on the posture of the case; refusing 
to apply § 113(b) to case in which expert fee application had

(continued...)
-43-



Court's overall concern with fair notice, id. at 1497-1501, 
together with its discussion of intervening procedural 

provisions, id. at 1505 n.34, make plain that Section 113 
does not apply in the circumstances of this case.

This Court's decision in Vernon v. Cassadaaa Valley 
Cent.. Sch. Dist.. 49 F.3d 886 (2d Cir. 1995), further 

demonstrates that the district court was wrong. In Vernon. 

the Court upheld application of a new statute of limitations 
to bar a claim that had accrued before the statute's 

enactment. The plaintiffs, however, had an opportunity to 

file their complaint after enactment of the new limitations 
period, and indeed had been given specific notice of the need 

to satisfy the new, tighter time limit. See 49 F.3d at 889- 

90. This "secondary conduct," the Court concluded, was the 

proper reference point from which to judge whether 

application of the new law would offend retroactivity 

principles. Id. at 890. In the present case, all "secondary 
conduct" relating to the retention and use of expert 

witnesses occurred before enactment of the 1991 Act.

The court below overlooked the Landgraf Court's 
categorical disapproval of retroactive application of legal 
sanctions. See, 114 S. Ct. at 1497. See also id. at 1505

— (...continued)
been denied before enactment of 1991 Act). The timing of a 
trial court's order has no special significance under 
Landgraf; rather, it is the parties' conduct that implicates 
the_"familiar considerations of fair notice, reasonable 
reliance, and settled expectations," 114 S. Ct. at 1499, 
animating retroactivity doctrine. The Shipes court simply 
had no need to take account of this distinction.

-44-



(Court would "'hesitate to approve the retrospective 
application of liability on any theory of deterrence'") 

(citation omitted). Fee awards against civil rights 

plaintiffs are designed to "deter the bringing of lawsuits 
without foundation." Christiansburg. 434 U.S. at 421 

(internal quotations and citations omitted). See also, e.g.. 

Jones v. Continental Corp.. 789 F.2d 1225, 1232 (6th Cir. 

1986) (an attorney's fee awards against civil rights 

plaintiffs is "an extreme sanction").— Such provisions may 

not be applied retroactively absent an "explicit command" 

from Congress, one that is lacking in the 1991 Act.

Landgraf. 114 S. Ct. at 1506. For this reason as well, the 

district court's expert fee award was improper.

III. The Amount of Fees Awarded Was an Abuse of Discretion.
The amount of fees that the district court assessed 

would have been a gross abuse of discretion even if its 

erroneous ruling on fee eligibility, see Part I, supra. had 

been correct. As far as plaintiffs have been able to 

determine, the fee award in this case was, by far, the 

largest ever assessed against a civil rights plaintiff under 

the Christiansburg standard. This huge award resulted from 

the court's decision to award fees for all work performed by 

defendants' counsel -- even work that (1) was performed long'

—' Contrary to the district court's statement in its
memorandum, see A154, retroactive application of an expert 
fee-shifting provision would not deter frivolous filings.
See Landgraf. 114 S.Ct. at 1506 n.35 (noting that deterrence 
requires a prior opportunity to conform to the law).

-45-



before the time when it is claimed that plaintiffs should 

have dismissed the case or (2) related to claims that have 
never been labelled as frivolous by defendants or the 

district court.

A . No Fees Should have been Awarded for Phases of the 
Case Prior to the Point When it Became Unreasonable 
for Plaintiffs to Pursue their Claims Further

The court below did not suggest that plaintiffs'

claims were frivolous when filed. Indeed, it upheld the

legal sufficiency of those claims in memorandum orders

denying defendants' motion to dismiss and approving
plaintiffs' motion for class certification. Defendants

argued only that plaintiffs acted unreasonably by opting to

go to trial once discovery had (in defendants' view) revealed

fatal flaws in their principal claims; even they did not

contend that plaintiffs' actions prior to the close of

discovery were unreasonable.— It follows that their

— See. e.g.. Defendants' Memorandum on Fee Issue at 2
(A126) ("Plaintiffs' decision to press this matter to trial 
was unreasonable."); Defendants' Reply Memorandum at 2 (A136) 
("the quantum of statistical and anecdotal evidence known to 
plaintiffs through discovery was so inadequate that it was 
unreasonable for them to proceed to trial at all"; "it is 
primarily because plaintiffs put defendants to the enormous 
expense of a trial that this motion is brought") (emphasis in 
original); id. at 5 (A139) ("In short, the flaws in 
plaintiffs' statistical case were so fundamental that the 
County should have been spared the expense of exposing them 
at trial."); Transcript of Argument on Fee Issue at 3-4 
(T1143-T1144) ("I . . . have attempted to dissect for the
Court . . . the point in the litigation where we believe that 
the plaintiff and their attorneys certainly should have known 
that they did not have the minimum quantum to present this 
evidence, and that it was in 1988 when they completed receipt 
of their reports from their experts.").

-46-



recovery of fees should have been linked to work performed at 
or after trial.

"An award of fees need not cover the entire period

of a case but may be limited to the period after events

demonstrated that the case was frivolous, unreasonable or

groundless." Greenberg v. Hilton Int'l Co.. 870 F.2d 926,

940 (citing Christiansburg and Hermes v. Hein. 742 F.2d 350,

358 (7th Cir. 1984)), reh'g granted in part on other grounds.

875 F .2d 39 (2d Cir. 1989) .— As said in Hermes,

the district court should determine whether and 
when it should have become clear, in light of the 
plaintiffs' failure to uncover necessary facts from 
discovery already taken, that further discovery 
would not have produced sufficient evidence of a 
genuine issue of material fact. It is only at such 
a juncture that the plaintiffs' continued 
litigation might have become frivolous so as to 
justify attorneys' fees for subsequent work. 742 
F. 2d at 358 (emphasis added).

Assessing fees only from the point at which claim 

became frivolous is consistent with the Court's observations 

in Christiansburg that "the course of litigation is rarely 

predictable," and that "[djecisive facts" revealing the

- See also, e.g., EEOC v. Jordan Graphics, Inc., 769 F. 
Supp. 1357 (W.D.N.C. 1991) (plaintiffs' claims were non- 
frivolous when filed, but plaintiffs should have recognized 
absence of factual support for claims at close of discovery; 
assessing only portion of attorney's fees that was incurred 
after close of discovery); Coleman v. McLaren. 631 F. Supp. 
763, 765-67 (N.D. 111. 1986) (assessing attorney's fees 
accruing after handing down of appellate decisions that 
rendered plaintiff's claims frivolous). Cf. Colombrito v. 
Kelly, 764 F.2d 122, 132 (2d Cir. 1985) (declining to award 
fees for allegedly unreasonable litigation decision that 
"added no additional testimony or expense to the trial"); 
Lane v. Sothebv Parke Bernet, Inc.. 758 F.2d 71, 73 (2d Cir. 
1985) (remanding for findings "as to whether [plaintiff] 
should have continued this action after completion of 
discovery").

-47-



frailty of the plaintiff's case "may not emerge until 
discovery or trial." 434 U.S. at 422. If each decision to 

proceed further subjects a plaintiff to liability for an 

award covering the entire case (no matter how sound it 

appeared at the outset), the already substantial risks 
attending major civil rights and antidiscrimination 

litigation would become intolerable.
The court below never identified any juncture at 

which plaintiffs' case, although reasonable when filed, 

became unreasonable to pursue. Its own negative assessment 
of plaintiffs' case became possible "only . . . after the 

smoke from the defense salvos" had cleared. 825 F. Supp. at 

473 (A107). Nevertheless, the court held plaintiffs liable 

for all fees incurred by defendants since the very beginning 

of the case. This was an stark abuse of discretion. Even if 

the district court's harsh appraisal of plaintiffs' proof at 

trial was correct, the court should have confined defendants' 

award to the fees accrued after plaintiffs' conduct became 

unreasonable. As defendants themselves recognized, see supra 

n. 32, plaintiffs' conduct became unreasonable, if at all, 
only when they failed to withdraw their claims on the eve of 

trial.

B. The District Court Erred bv Assessing Fees Against 
Plaintiffs Relating to Claims on Which Plaintiffs 
Prevailed and to Claims Not Found to be Frivolous)

The district court's awards of attorney's fees in

this case covered the County's expenditures in defending
against all of plaintiffs' claims, even though the court

-48-



never found or even suggested that all of these claims were 

frivolous. Most strikingly, the fee award to defendants 

included attorneys' fees attributable to plaintiffs' equal 

pay claims on behalf of the 17 police detention aides, even 

though plaintiffs prevailed on these claims, netting more 

than $1.6 million in backpay and interest and entitling 

themselves to an award of fees. See 799 F. Supp. at 1425-16 

(A99-A100); A109-A113, A117-A121. It also covered fees 

defendants incurred in prosecuting unsuccessful counterclaims 

against AFSCME. In addition, the fee award to defendant 

embraced fees related to plaintiffs' "equal pay" claims on 

behalf of police communications operators and clerical 
workers at the Nassau County Detention Center. While the 

district court ultimately ruled against the plaintiffs on 

these claims, e.g. . 799 F. Supp. at 1415-16 (A99-A100) 

(concluding that jobs of fire dispatchers and police 

dispatchers are not "substantially similar"), it did not find 

these claims frivolous or unreasonable. It was error to 

assess fees against plaintiffs in connection with these 

distinct and undisputedly non-frivolous claims. See. e.g.. 

Hensley v, Eckerhart, 461 U.S. 424, 435 n.10 (1983) (claim-

by-claim approach that governs prevailing plaintiffs' fee 

awards applies to prevailing defendants as well); Kaimowitz 

v. Howard. 547 F. Supp. 1345, 1351-52 (E.D. Mich. 1982) 

(assessing attorney's fees against plaintiff only for the 

subset of his claims that was found to be frivolous).

-49-



Conclusion

For the foregoing reasons, the district court's 

orders awarding attorney's and expert fees to defendants 

should be reversed.

Busan R. Podolsky 
Sean H. Donahue 
JENNER & BLOCK 
601 13th St., N.W. 
Washington, D.C. 20005 
(202) 639-6000

Attorneys for
Plaintiffs-Appellants

-50-