King v. Georgia Power Company Brief for Plaintiffs-Appellants
Public Court Documents
January 11, 1972
Cite this item
-
Brief Collection, LDF Court Filings. King v. Georgia Power Company Brief for Plaintiffs-Appellants, 1972. b2c28e0b-ba9a-ee11-be36-6045bdeb8873. LDF Archives, Thurgood Marshall Institute. https://ldfrecollection.org/archives/archives-search/archives-item/5c8b3170-892b-4e03-bc41-c9f683c2cd18/king-v-georgia-power-company-brief-for-plaintiffs-appellants. Accessed November 23, 2025.
Copied!
IN THE
UNITED STATES COURT OF APPEALS
FOR THE FIFTH CIRCUIT
NO. 71 - 3229
CHARLES KING, et al.,
and
WILLIE MOREMAN, et al.,
Plaintiffs - Appellants,
- vs -
GEORGIA POWER COMPANY, et al.,
Defendants - Appellees.
On Appeal From The United States District Court
For The Northern District Of Georgia
BRIE]?’ FOR PLAINTIFFS - APPELLANTS
HOWARD MOORE, JR.
ELIZABETH R. RINDSKOPF
75 Piedmont Avenue, N.E. Suite 1154
Atlanta, Georgia 30303
ALBERT J. ROSENTHAL GEORGE COOPER
HARRIETT RABB
435 West 116th Street New York, New York 10027
ISABEL GATES WEBSTER
75 Piedmont Avenue, N.E. Suite 1170
Atlanta, Georgia 30303
Of Counsel JACK GREENBERG
WILLIAM L. ROBINSON
MORRIS J. BALLER
10 Columbus Circle - Suite 2030 New York, New York
Attorneys for Plaintiffs-Appellants
IN THE
UNITED STATES COURT OF APPEALS
FOR THE FIFTH CIRCUIT
NO. 71 - 3229
CHARLES KING, et al.,
and
WILLIE C. MOREMAN, et al.,
Plaintiffs - Appellants,
- vs -
GEORGIA POWER COMPANY, et al.,
Defendants - Appellees.
CERTIFICATE
The undersigned counsel for plaintiffs-appellants King, Moreman,
et al. in conformance with Local Rule 13(a) certifies that the follow
ing listed parties have an interest in the outcome of this case.
These representations are made in order that Judges of this Court
may evaluate possible disqualification or recusal:
1. Charles King, Paul Brown, Ed Dulaney, Sammie L. Davenport,
Rufus Mitchell, and Willie C. Moreman, all plaintiffs.
2. The class of black employees of Georgia Power Company whom
plaintiffs represent.
3. Georgia Power Company, defendant.
4. International Brotherhood of Electrical Workers and Local
84 thereof, defendant.
I N D E X
Page
TABLE OF AUTHORITIES.................................... i
STATEMENT OF ISSUES PRESENTED FOR REVIEW................ x
STATEMENT OF THE C A S E .................................. 1
STATEMENT OF FACTS .................................. 4
A. Background information ...................... 4
B. Continuing patterns of discrimination . . . . 9
C. Testing practices............................ 14
D. Assignment practices ........................ 20
E. Recruitment practices ...................... 22
ARGUMENT
I. GEORGIA POWER COMPANY'S USE OF EMPLOYMENT TESTS
VIOLATES TITLE VII SINCE THOSE TESTS EXCLUDE
BLACKS FROM HIRING OR PROMOTION INTO BETTER
JOBS AND SINCE THE TESTS' JOB RELATEDNESS HAS
NOT BEEN ADEQUATELY DEMONSTRATED................ 24
A . The Equal Employment Opportunity Com
mission Guidelines on Employee Selection
Procedures Define the Appropriate Standards
of Test Use. Since Georgia Power's Testing
Practices Do Not Conform To These Standards,They Should Be Declared Unlawful ............ 26
B. Even If This Court Chooses To Develop Its Own Standard For Review Of Validity Data,
It Should Enjoin The Georgia Power Testing
Program. The Hite Study Fails To Demon
strate Job Relatedness Or Business Necessity
As Measured By Any Reasonable Standard . . . . 34
II. THE COURT BELOW ERRED IN FAILING TO FIND THAT
GEORGIA POWER COMPANY HAS CONTINUED TO ASSIGN
BLACK HIREES TO MENIAL JOBS ON THE BASIS OF RACE,
DESPITE PLAINTIFFS' UNREBUTTED STATISTICAL
EVIDENCE SHOWING THESE PRACTICES ................ 45
Page
III. GEORGIA POWER COMPANY VIOLATES TITLE VII
BY PLACING PRIMARY RELIANCE ON WORD-OF
MOUTH RECRUITMENT AND WALK-IN APPLI
CATIONS TO FILL JOB VACANCIES, IN LIGHT
OF ITS SUBSTANTIALLY SEGREGATED WORK FORCE ........
IV. THE COURT BELOW ERRED IN FAILING TO GRANT
FULL AND EFFECTIVE SENIORITY RELIEF, OR IN
SOME INSTANCES ANY RELIEF, TO BLACKS WHOSE
SENIORITY RIGHTS AND STATUS WERE ADVERSELY
AFFECTED BY DEFENDANTS' DISCRIMINATION ............
A. The Relief Granted By The District
Court Pertaining To Seniority Was
Inadequate And Ineffectual In ManyRespects ..............................
B. This Court Should Correct The Defi
ciencies Of The District Court's
Seniority Remedy In Keeping With Its Duty To Provide Full And Effective
Affirmative Relief From Racial
Discrimination..............................
V. THE COURT BELOW ERRED IN FAILING TO GRANT
ADEQUATE TITLE VII RELIEF IN THE NATURE OF BACK PAY........................................
A. The District Court Improperly Limited
The Amount Of Back Pay It Awarded To
The Named Plaintiffs In Allowing Them
Far Less Than Their Actual Loss Due
To Defendants' Discrimination ..............
B. The District Court Erred In Denying
Back Pay To Members Of Plaintiffs'
Class, Even Though The Court Found
That Class Members Had Suffered Severe
Economic Loss Resulting From Defendants' Discrimination..............................
VI. THE COURT BELOW ERRED IN LIMITING THE PERIOD
FOR ANY CLASS BACK PAY AWARD BY APPLICATION
OF AN INAPPROPRIATE AND UNDULY RESTRICTIVE STATUTE OF LIMITATIONS ............................
VII. THE COURT BELOW ERRED IN LIMITING ITS AWARD
OF ATTORNEYS' FEES TO PLAINTIFFS' COUNSEL BY
APPARENTLY FAILING TO COMPENSATE THEM FOR TIME SPENT AT AND AFTER TRIAL AND BY BASING THE AWARD
ON A SUGGESTED MINIMUM FEE SCHEDULE THAT WAS INAPPROPRIATE FOR THIS CASE..................
48
56
56
65
68
68
71
83
91
CONCLUSION 97
Page
APPENDIX A
APPENDIX B
?
: Note on Form of Citations..........A-l
: Materials on the Technical
Defects of the Georgia Power
Testing Program and Validity
Study.............................. B-l
CERTIFICATE OF SERVICE
TABLE OF AUTHORITIES
CASES Page
Allen v. Lockheed Corp. 61 LC
H 9333 (N.D. Ga. 1968).................... 95
Baker v. F & F Investment, 420 F.2d
1191 (7th Cir. 1970) .................... 90
Bankers Fidelity Life Insurance Co. v. Oliver,
106 Ga. App. 305, 126 S.E. 2d 887
(Ga. 1962) .............................. 89
Banks v. Lockheed-Georgia Corp., 46 FRD
442 (N.D. Ga. 1968)...................... 95
Baxter v. Savannah Sugar Refining Corp. 46
F.R.D. 56 (S.D. Ga. 1969)................ 77
Beard v. Stephens, 372 F.2d 685
(5th Cir. 1967).......................... 86
Bing v. Roadway Express, Inc. 444 F.2d
687 (5th Cir. 1971)...................... 46,54,65
Boudreaux v. Baton Rouge Marine Contracting
Corp., 437 F .2d 1011 (5th Cir. 1971). . . . 84
Bowe v. Colgate-Palmolive Company 416
F .2d 711 (7th Cir. 1969).................. 70,75,78
Brazier v. Cherry, 293 F.2d 401
(5th Cir. 1961).......................... 88
Brown v. City of Meridian, 356 F.2d 602
(5th Cir. 1966).......................... 85
Cape Cod Food Products, Inc. v. National
Cranberry Ass'n. 119 F.Supp.
242 (D. Mass. 1952)...................... 95
Carter v. Gallagher, ___ F.Supp. ___, 3 EPD
1(8205 (D. Minn. 1971), aff'd in
pertinent part, ___ F.2d ___, 3 EPDH8335 (8th Cir. 1971).................... 30
Clark v. American Marine Corp. 320 F.Supp.
709 (E.D. La. 1970), aff'd per curiam
437 F . 2d 959 (5th Cir. 1971).............. 82,90,94,95
CONT'D
Clark v. American Marine Corp., 304 F.Supp.
603 (E.D. La. 1969) .................... 51,52,84
Colbert v. H-K Corp., 3 EPD f8248 (5th
Cir. 1971).............................. 95
Culpepper v. Reynolds Metals Co. 421
F . 2 d 888 (5th Cir. 1970)................ 76,88,95
Dobbins v. Local 212, IBEW, 292 F.Supp.
413 (S.D. Ohio, 1968) .................. 93
Dobbins v. Local 212, IBEW, F.Supp. , 61
LC f 9327 (S.D. Ohio 1969)............... 94
EEOC Decisions
Decision 70 - 630 29
Decision 71 - 1418 29
Decision 71 - 1471 29
Decision 71 - 1525 29
Unnumbered Decision (Dec. 6 , 1966) . . . . 29
Evans v. I-T-E Corp., 313 F.Supp.
1354 (N.D. Ga. 1970)................... 95
Franklin v. City of Marks, 439 F.2d
665 (5th Cir. 1971)................... 86,90
Freeman v. Ryan, 408 F.2d 1204 (D.C. Cir. 1968). . 95
Glascoe v. Howell, 431 F.2d 863 (8th Cir. 1970). . 90
Griggs v. Duke Power Company, 401 U.S. 424
(1971).................................. 16,25,29,32,34,
44,51,53,55,62,
63,77,80
Hardin v. Kentucky Utilities Co., 390 U.S.
1 (1968)................................ 33
Hicks v. Crowa-Zellerbach Corp., 319 F.Supp.
314 (E.D. La. 1971)................... 16,30
Hodgson v. American Can Co., Dixie Products,
___ F . 2d ___, 3 EPD f 8171 (8thCir. 1971).............................. 91
Holliday v. REA Express, 306 F.Supp. 898
(N.D. Ga. 1969)........................ 95
Page
- ii -
CONT'D
Huson v. Oatis Engineering Corp., 430 F.2d
37 (5th Cir. 1970)........................ 88
In re Osofsky, 50 F.2d 925 (S.D.N.Y. 1931) ........ 96
Jenkins v. United Gas Corp., 400 F.2d 28
(5th Cir. 1968).......................... 77
Johnson v. Georgia Highway Express Co., 417
F.2d 1122 (5th Cir. 1969)................ 70,75,77,87,94
Jones v. Lee Way Motor Freight Co., 431 F.2d
245 (10th Cir. 1970), cert, denied 401U.S. 954 (1971).......................... 46
Jones v. Montag, 3 EPD H8243 (N.D. Ga. 1969) . . . . 95
Lazard v. Boeing Co., ___F.Supp. , 3 FEP
Cases 643 (E.D. La. 1971)................. 85,90
Lea v. Cone Mills Corp., 301 F.Supp. 97
(M.D.N.C. 1969) aff'd in pertinent part,
438 F.2d 86 (4th Cir. 1971).............. 52,55
Lea v. Cone Mills Corp., 438 F.2d 86 (4th Cir.1971).................................... 82
Lefton v. City of Hattiesburg, 333 F.2d 280
(5th Cir. 1964).......................... 85
Local 53, International Ass'n. of Heat & Frost
Insulators & Asbestos Workers v. Vogler,
407 F . 2d 1047 (5th Cir. 1969)............ 51,53,55,65,
67,79,84
Local 53, International Ass'n. of Heat & Frost
Insulators & Asbestos Workers v. Vogler,294 F.Supp. 368 (E. D. La. 1968)............ 54
Local 186 v. Minnesota Mining and Manufacturing
Co., 304 F.Supp. 1284 (N.D. Ind. 1969). . . 77
Local 189, United Papermakers and Paperworkers v.
United States, 416 F.2d 980 (5th Cir.
1969), cert, denied, 397 U.S. 919 (1970) . . 7,53,64,65,80
Long v. Georgia Kraft Co., F.2d ,4
EPD H7556 (5th Cir. 1971) .
Page
- iii -
95
CONT ' D
Long v. International Brotherhood, 60
LC f9306 (N.D. Ga. 1969)................ 95
Louisiana v. United States, 380 U.S. 145
(1965).................................. 67,79
Mclver v. Russell, 264 F.Supp. 22
(D. Md. 1967).......................... 90
Miller v. Amusement Enterprises, Inc.,
426 F .2d 534 (5th Cir. 1970)............ 82
Miller v. International Paper Co., 408
F . 2d 283 (5th Cir. 1969)................ 67,78,82
Morrow v. Crisler, ___ F.Supp. ___, 4 EPD
H7541 (S.D. Miss. 1971)................ 54
NAACP v. Button, 371 U.S. 415 (1963)............. 96
Newman v. Piggie Park Enterprises, Inc.
390 U.S. 400 (1968).................... 82,96
NLRB v. Mooney Aircraft, 366 F.2d 809
(5th Cir. 1966)........................ 81
NLRB v. Rutter - Rex Manufacturing Co.,
396 U.S. 258 (1969).................... 80
NLRB v. United Marine Division, Local 333,
417 F. 2d 865 (2nd Cir. 1969) .......... 81
Oatis v. Crown-Zellerbach Corporation, 398
F . 2d 496 (5th Cir. 1968)................ 78
Parham v. Southwestern Bell Telephone Co.,
433 F.2d 421 (8th Cir. 1970)............ 50,52
Pettway v. American Cast Iron Pipe Co.
411 F.2d 998 (5th Cir. 1969)............ 82
Potts v. Flax, 313 F.2d 284 (5th Cir. 1963). . . . 79
Power Reactor Co. v. Electricians, 367
U.S. 396 (1961)........................ 33
Quarles v. Philip Morris, Inc. 279
F.Supp. 505 (E.D. Va. 1968)............ 80
Page
IV
CONT'D
Red Lion Broadcasting Co. v. F.C.C. 395 U.S. 367(1969).................................... 32
Robinson v. Lorillard Corp., 444 F.2d
791 (4th Cir. 1971)...................... 75,76,82
Sanders v. Dobbs Houses, Inc., 431 F.2d
1097 (5th Cir. 1970), cert.denied,
3 EPD 18127 (1970)........................ 94,95
S.E.C. v. New England Electric System, 384
U.S. 176 (1966).......................... 33
Shultz v. Wheaton Glass Co., ___F.Supp. ___, 3
EPD 18270 (D.N.J. 1970), aff'd in pert
inent part, ___ F . 2d ___, 3 EPD 182 96
(3rd Cir. 1971).......................... 90
Simler v. Conner, 352 F.2d 138 (10th Cir. 1965). . . 95
Smith v. Cremins, 308 F.2d 187 (9th Cir. 1962). . . . 90
Sprogis v. United Air Lines, Inc., 444 F.2d
1194 (1971) cert, denied ___ U.S. ___,
4 EPD 17588 (1971)........................ 77
Sullivan v. Little Hunting Park, 396 U.S.
429 (1969)................................ 88
Trinity Valley Iron & Steel Co. v. NLRB
410 F.2d 1161 (5th Cir. 1969)............. 81
Udall v. Taliman, 380 U.S. 1 (1965)................ 32
United States v. Bethlehem Steel Corp., 446
F . 2d 652 (2nd Cir. 1971).................. 48,64,65,66
United States v. Hayes International Corp., 415
F . 2d 1038 (5th Cir. 1969)................ 46,65,82
United States v. Ironworkers Local 8 6, 315 F.Supp.
1201 (W.D. Wash. 1971), aff'd 433 F.2d 544
(9th Cir. 1971), cert, denied 4 EPD 57526
(1971).................................... 55
United States v. Ironworkers Local 392, 3 EPD 58063
(E.D. 111. 1970).......................... 54
Page
v
C O N T 'D
United States v. Jacksonville Terminal Company,
___, F .2d ___, 3 EPD f8324 (5th Cir.
1971).................................... 30,46,47,65
United States v. Plumbers Local 73, 314
F.Supp. 160 (S.D. Ind. 1969).............. 52,54
United States v. Sheet Metal Workers, Local 36,
416 F . 2d 123 (8th Cir. 1969)............ 46,52,54,65
United States v. Virginia Electric & Power Co.
4 EPD f7502 (E.D. Va. 1971)............... 54
United States v. West Peachtree Tenth Corp.,
437 F. 2d 221 (5th Cir. 1971).............. 45
United States v. Wood, Wire, and Metal Lathers,
Int. U., Local 46, ___ F.Supp. ___, 3
EPD 18204 (S.D.N.Y. 1971)............ .. . 77
Vogler v. McCarty, Inc., ___F.2d ___, 4 EPD
17581 (5th Cir. 1971)..................... 64
Wakat v. Harlib, 253 F.2d 29 (7th Cir. 1958) . . . . 90
Waters v. Wisconsin Steel Works of International
Harvester Co., 427 F.2d 476
(7th Cir. 1970).......................... 85
Statutes and Regulations
28 U.S.C. §1291..................................... 1
28 U.S.C. §1652 .................................... 85
42 U.S.C. § 1 9 8 1 ..................................... 85
42 U.S.C. §1982 85,90
42 U.S.C. §1983 85,90
42 U.S.C. §1985 .................................... 90
42 U.S.C. §1988
Page
. 84,85,86
C O N T 'D
Civil Rights Act of 1964, Title I I ................ 82
Civil Rights Act of 1964, Title VII,
42 U.S.C. §§2OOOe et seq.................. passim
42 U.S.C. §2000e-2 (a)..................... 52
42 U.S.C. §2000e-2 (a) ( 1)................ 50,52,55
42 U.S.C. §2000e-2 (a) (2) 51,52
42 U.S.C. §e000e-4(f) .................... 32
42 U.S.C. §2000e-5, - 8 .................. 31
42 U.S.C. §2000e-5 (d) .................. 83
42 U.S.C. §2000-5 (g) .................... 54,73,74,75,86
42 U.S.C. §2000e-5 (k).................... 91,92
42 U.S.C. § 2000e-6...................... 3
42 U.S.C. §2000e-6(a)(3).................. 73
42 U.S.C. §2000e-12 (b).................... 32
Equal Employment Opportunities Commission
Guidelines on Employee Selection Procedures ,
29 C.F.R. §§1607 et seq. (1970).......... 28
29 C.F.R. §1607.1(c)..................... 28
29 C.F.R. §1607.4 (c) ( 2)................. 34
29 C.F.R. §1607.5 (a) , (b) , (c) ............ 28,33,34,40
29 C.F.R. §1607.5(b) (3)-(4).............. B-9
29 C.F.R. §1607.5 (b) (5)................. B-ll
Page
- vii -
CONT'D
Paae_
Equal Pay Act,
29 U.S.C. §206 .......................... 90
Federal Rules of Appellate Procedure,
Rule 30(c)................................ 2
Federal Rules of Civil Procedure, Rule 23 ........ 5
Georgia Code Annotated
§3-704.................................... 83,85,86,89
§3-706 .................................. 83,85,89
National Labor Relations Act, 29 U.S.C.
§§151 et seq.............................. 79
29 U.S.C. §160 (c) (Section 10(c)..........
Office of Federal Contract Compliance (OFCC)
Regulations
79
Regulation implementing Executive Order
No. 11,246, 41 C.F.R.
Section 5-12.805-51 (b) (5)................ 50
Employee Testing and Other Selection
Procedures,
36 Fed. Reg. 19307 (1971) ................ 30
Pennsylvania Guidelines on Employee
Selection Procedures, a Pa. Bull.
2005 (1971) .............................. 30
Other Authorities
American Bar Association (A.B.A.) Code of
Professional Responsibility,
Disciplinary Rule 2-106 ................ 92
Disciplinary Rule 2-106 (B)(7)............ 94
Disciplinary Rule 2-106 (B) (8)............ 95
Annotation, 56 A.L.R. 2d 13 (1957)................ 92
Blumrosen, The Duty of Fair Recruitment Under the
Civil Riqhts Act of 1964, 22 Rutgers L.
Rev. 465 (1968) ..........................
I.D.J. Bross, Desiqn for Decision (1953)..........
49,50,51
40
- viii-
40
CONT'D
Page
Clark - Case Interpretative Memorandum on Title VII,
Congressional Record (Senate) April 8, 1964)... 80
Cooper and Sobol, Seniority and Testing Under Fair
Employment Laws, 82 Harv. L. Rev. 1598 (1969) . . 7
W.J. Dixon and F.J. Massey, Jr., Introduction to Stat
istical Analysis (2d ed. 1957).................. 42
A .L. Edwards, Statistical Methods (2d ed. 1967)........ 40
J.P. Guilford, Fundamental Statistics in Psychol
ogy and Education,4th ed. McGraw-Hill,
1965 App. B, p. 508, Table D.................... B-4
W. Hays, Statistics for Psychologists (1963)............ 40
McNemar, Quinn, Psychological Statistics. 4th ed.
John Wiley & Sons, Inc. (New York, 1969)........ B-5
C.C. Peters and W.B. Van Voorhis, Statistical Pro
cedures and Their Mathematical Bases (1940) . . . 40
M.H. Walker, Elementary Statistical Methods (1945) . . . . 40
STATEMENT OF ISSUES PRESENTED FOR REVIEW
1. Whether Georgia Power Company's use of employment
aptitude tests as a screening device for hiring and promotion
violates Title VII of the 1964 Civil Rights Act where such
tests had a substantially disproportionate impact on black
employees and applicants, and where the Company's purported
proof of the tests' job-relatedness does not comply with the
EEOC Guidelines on Employee Selection Procedures or any pro
fessionally accepted standards of test validation?
2. Whether the court below erred in refusing to find that
Georgia Power Company has continued to assign black hirees to
menial positions on the basis of race and without regard to
their qualifications as shown by plaintiffs' unrebutted
statistical evidence ?
3. Whether in light of its racially stratified work force
the Georgia Power Company's exclusive reliance on a word-of-
mouth employee referral system and walk-in-hiring violates
Title VII of the 1964 Civil Rights Act?
4. Whether the partial seniority relief granted by the district
court was inadequate as a matter of law in that it fails to pro
vide a full and effective remedy to plaintiffs and the members
of their class, with respect to both the necessary modification of
defendants' discriminatory seniority system and the adjustment of
the seniority positions of black employees whose seniority status
was adversely affected by defendants' discrimination?
x
5. Whether the district court erred in limiting its award of
compensatory back pay:
A. by limiting the back pay awards to named plaintiffs
to a minimal amount which was far less than these plaintiffs
would have earned absent defendants' racial discrimination.
B. by refusing to award any back pay at all to members
of plaintiffs' class, even though the court found that
class members had suffered a severe economic loss as a
result of defendants' discrimination?
6 . Whether the district court erred in its limitation of the
period of any class back pay award by applying an inappropriate
and unduly restrictive statute of limitations on back pay relief.
7. Whether the district court erred in its award of counsel
fees to plaintiffs' attorneys where the award did not give those
attorneys any compensation for participation in trial and post
trial proceedings, and where the award was calculated on the
basis of a suggested minimum fee schedule which fails to allow
for the special experience and skill required to litigate these
actions?
IN THE
UNITED STATES COURT OF APPEALS
FOR THE FIFTH CIRCUIT
NO. 71 - 3229
CHARLES KING, et al. ,
and
WILLIE MOREMAN, et al.,
Plaintiffs-Appellants,
- vs -
GEORGIA POWER COMPANY, et al.,
Defendants-Appellees.
BRIEF FOR PLAINTIFFS-APPELLANTS
STATEMENT OF THE CASE
This appeal involves two cases, each a broad class action
attacking across-the-board practices of employment discrimination
under Title VII of the Civil Rights Act of 1964, 42 U.S.C. §§2000e
et seq., which were consolidated for trial and decision below.
The appeal is from the final judgment in these actions of the
United States District Court for the Northern District of Georgia,
Smith C.J. This Court has jurisdiction of the appeal pursuant
to 28 U.S.C. §1291.
This complex litigation was instituted by complaints filed
on April 12, 1968 in No. 11723 and on October 28, 1968 in No. 1218b
1/(King complaint, Moreman complaint). Plaintiffs in both actions,
the appellants here, are six black employees of defendant Georgia
Power Company who at the time of filing suit all held laborer jobs.
(King complaint at 2; Moreman complaint at 3.) Defendants in both
actions were the Georgia Power Company, a privately owned utility
providing electrical power and ancillary services throughout Georgia
(Ic[.) , and Local Union No. 84, an affiliate of the International
Brotherhood of Electrical Workers, which represents all the unionized
employees of the Georgia Power Company in its Atlanta power-pro
duction facility (Id.).
The two complaints made essentially identical allegations
concerning the defendants' across-the-board policies and practices
of racial discrimination in employment. The defendants' answers
denied all the substantive allegations of the complaint (King
answers; Moreman answers). The court below on August 18, 1969 con
solidated the two private actions, and defined the class in both
actions to include "All presently employed Negro laborers at the
Atkinson-McDonough Plant of the Georgia Power Company in Atlanta,
Georgia" (Pre-trial Order at 1).
1/ Appellants have elected to proceed under the deferred appendix
system pursuant to Rule 30(c), Federal Rules of Appellate Procedure;
therefore appendix page citations are not available as this brief is filed. All citations are therefore to the original record on
appeal as compiled by the clerk of the district court. The forms
of citation used in this brief are explained in the Note on Form
of Citations, Appendix A hereto.
2
Subsequent to the filing of these private actions, the
United States on January 10, 1969 filed a pattern and practice suit
against the same defendants under §707 of Title VII, 42 U.S.C.
2/§§2000e-6 (U.S. complaint). The district court on June 23, 1970
ordered the United States suit consolidated with both private actions
for trial. From that date onward, many class aspects of the private
actions were litigated as part of the pattern-and-practice action.
On July 13, 1970 the United States filed a motion for pre
liminary injunction requesting relief for certain classes of black
employees (Motion for Preliminary Injunction). After an evidentiary
hearing, the court on September 22,1970 entered findings of fact,
conclusions of law, and an order on the United States' motion
(Preliminary Injunction). As modified by the court's order of
October 5, 1970 this preliminary order granted part of the relief
sought by that motion (Amended Preliminary Injunction).
On November 30, 1970 a third private individual action,
George Jones v, Georgia Power Company, C.A. No. 14182, was con-
2/solidated for trial. All the consolidated cases came on for
hearing on January 18, 1971; the trial lasted seven days.
2/ While the United States suit presented many essentially identi
cal questions of fact and law, it was broader in scope than the plaintiffs' private actions. United States v. Georgia Power Company,
et al., C.A. No. 12355, challenged defendants' practices statewide,
and joined as defendants seven local unions of the IBEW which represent Georgia Power Employees at various locations throughout
the state.
3/ No appeal has been taken from the court's final judgment
in No. 14182, and that matter is not before this Court on appeal
3
Following trial on the merits and the filing of various post
trial pleadings, the district court proceeded to enter its opinion
4/on the merits on June 30, 1971 (Opinion). This decision incorp
orated and modified the court's previous order on the motion for
preliminary injunction, and added additional findings and con
clusions. In general, the court held some but not all of defendants'
contested practices in violation of Title VII; and granted some but
not all of the relief plaintiffs requested from those practices
which it found discriminatory. The final decree in the private
actions was filed on October 22, 1971 (Private decree). This decree
incorporated by reference many aspects of the decree filed pre
viously in the United States case on September 27, 1971 (US decree).
On November 9, 1971 plaintiffs King, Moreman, et al. timely
filed their Notice of Appeal from the final judgment entered in the
private actions (Notice of Appeal). On this appeal plaintiffs
seek reversal both of the district court's findings that certain
of defendants' practices do not violate Title VII, and of its
failure to grant adequate relief from those practices which it did
find discriminatory.
STATEMENT OF FACTS
A . Background Information
These actions were instituted by complaints filed on April
12, 1968 and October 28, 1968 (King complaint, Moreman complaint).
Each complaint broadly alleged that defendants were engaged in a
4/ The decision is not as of this date officially reported; how
ever it appears in the labor law reports at 3 EPD f8318 (CCH
Reporter) and 3 FEP Cases 767 (BNA Reporter).
4
comprehensive policy and practice of racial discrimination with
respect to the terms, conditions, and benefits of employment (Id..) .
Each of the six named plaintiffs is a black man who worked
for Georgia Power in the job classifications of laborer and was a
member of one of the defendant unions at the time these actions
5/were instituted. Each plaintiff sought promotion or advancement
into a better and higher-paying job than laborer, but had been
denied such advancement pursuant to defendants' policies and
practices (King complaint at 2, Moreman complaint at 3). The
plaintiffs brought these actions as class actiorfeunder Rule 23,
Federal Rules of Civil Procedure, on behalf of all other similarly
situated black persons (King complaint at 3, Moreman complaint at 2)
Defendant Georgia Power Company (hereafter "Georgia Power"
or the "Company") is engaged in the business of production, dis
tribution and sale of electric power and related services through
out the state of Georgia, with its principal office at Atlanta
(Opinion at 6-7). The court found that Georgia Power employed
7,515 persons as of December 25, 1970, of whom 6,891 were white
and 524 were black [sic.] (Opinion at 7). Defendant Local Unions
are affiliated with the International Brotherhood of Electrical
Workers (hereafter "IBEW") and each represent workers in a separate
5/ Plaintiffs Charles King, Ed Dulaney, Sammie L. Davenport,Paul Brown, Rufus Mitchell, and Willie C. Moreman were all laborers
at Plant McDonough-Atkinson in Atlanta from the time these actions
were instituted until 1969 or later; all were members of defendant
Local 84.
5
geographical division of the Company. These unions together re
present 3,853 employees as their exclusive collective bargaining
agent with respect to rates of pay, wages, hours of employment,
working conditions, and other terms of employment (Opinion at 7).
All of these unions have entered into a single collective bar
gaining agreement with the Company (Id̂ .) The current bargaining
agreement became effective on April 29, 1969. It replaced the
prior agreement of July 1, 1965, and was to continue in effect
until June 30, 1971 and from year to year thereafter subject to
amendment by the parties (Opinion at 9).
All Georgia Power employees in bargaining unit jobs held
jobs in one of four organizational parts of the Company, each
vincluding one or more administrative units. Each of these * 2 3 4
6/
6/ Local 882 represents employees at Athens; Local 84, at Atlanta;
Local 923, at Augusta; Local 780, at Columbus; Local 876, at Macon;
Local 847, at Rome; and Local 511, at Valdosta (Opinion at 7). There
are no other unions which represent Georgia Power employees. Only
Local 84 was a defendant in the private actions. See n.2, supra.
7/ (1) The Production Department is a statewide administrative
unit for purposes of promotion, transfers, demotion and layoffs;
its function is the production of electrical energy at the Company's steam and hydroelectric generating plants.
(2) Seven geographic operating divisions with headquarters
at Athens, Atlanta, Augusta, Columbus, Macon, Rome and Valdosta
form separate administrative units for purposes of promotion,
transfer, demotion and layoffs; their function is the transmission,
distribution, and sale of electrical energy and appliances.
(3) The Construction Department is divided into three parts,
Plant Construction, Line Construction and Substation Construction,
each of which is a separate statewide administrative unit for pur
poses of promotion, transfer, demotion and layoffs.
(4) The General Sex-vices Department is a statewide admin
istrative unit whose functions include maintenance of many Company
vehicles and, at Atlanta, a variety of special services.
The majority of non-bargaining unit employees work in the Atlanta general headquarters. There is no seniority system with respect to these employees (Opinion at 7-8).
- 6 -
administrative units is a separate "seniority division" for pur
poses of determining employees 1 seniority rights under the collect
ive bargaining agreement (Opinion at 8).
The collective bargaining agreement sets forth each job
classification covered by the agreement and its rate of pay, and
sets forth the "sections" in which each job classification is con
tained (Opinion at 9). There are currently 19 such sections in the
Company, each containing one or more lines of progression comprised
of (usually) functionally related jobs through which employees pro
gress from entry level jobs to successively higher paying jobs
(Id.). The collective bargaining agreement also sets forth the
principles of seniority which govern employees' rights with respect
to promotion, transfer, demotion, and layoff. Employees may ex
ercise seniority only in the seniority division and section where
they work (Opinion at 9-10). The seniority system is generally of
8/
a rigid "job-seniority" type. Promotions are awarded (compe
tency being sufficient), and demotions due to lack of work are
allocated, on the basis of classification (job) seniority within
9/the employees 's seniority division and section (Iĵ .) .
8/ See Cooper and Sobol, Seniority and Testing Under Fair
Employment Laws, 82 Harv. L. Rev. 1598, 1602-1604 (1969), and Local
189, United Papermakers and Paperworkers v. United States, 416
F .2d 980, 983 (5th Cir. 1969).
9/ The court found: "First preference in filling a job vacancy
is given to the senior employee, competency being sufficient, in
the same classification in the same section and division where the
vacancy exists, requesting such transfer. If no such employee re
quests a lateral transfer, the vacancy is filled by the promotion
of the most senior employee, competency being sufficient, in the
next lower classification in the section and seniority division
requesting such promotion . . . The classification to which an
employee is demoted depends upon his combined seniority in his
classification and lower classifications at the time of demotion
as compared to the combined seniority of the employees in the lower classification" (Opinion at 10).
7
Cross-section transfer is permitted, but the transferring employee
is treated as a new employee in his new section, without carry-
over seniority or return rights to his former job (Opinion at 10).
The effect of this rigid job seniority system has been to dis
courage employee movement between classifications except within and
along the established lines of progression, as an inspection of
data in the record amply demonstrates (Gov. Ex. 7, 15, 17
[suppl.]).
This system of seniority is the route to promotion, and the
means by which employees of Georgia Power Company seek and obtain
successively higher paying jobs. Its significance in that regard
is self-evident. In the context of this case, however, the
seniority system has been used by Georgia Power Company in con
junction with a wide range of other devices, some subtle, and
others overt, to maintain a strict racial stratification of its em
ployees throughout the Company. While these devices are discussed
separately below, their cumulative impact has been to relegate the
Company's black employees to the lowest paying, most menial jobs
in the Company, while depriving them of all hope of advancement.
10/
10/ The court found: "Employees may transfer from one section
to another, but they may not transfer seniority gained in the
former section to the new section. While seniority retained in
a former section may not be used in his new section for any purpose,
in a case of layoff due to lack of work in his present section such
retained seniority may be used in the former section as his pro
tection against layoff, [citation omitted] If an employee has
made a cross-section transfer and is deemed to be incompetent in
his new classification he cannot exercise his seniority rights
for the purpose of regaining his former classification, but is
subject to possible discharge". (Opinion at 10).
8
B.
Prior to July 29, 1963, Georgia Power followed an open and
unvarying policy of limiting black employees to the most menial
jobs. Blacks in jobs covered by the collective bargaining agree
ment were relegated to the classifications of laborer, janitor,
porter and maid, the four lowest paying jobs in the bargaining
unit. These jobs were completely dead-end; they were in separate
lines of progression, and at least until that date transfer out
was absolutely prohibited. All of the other jobs, available only
to whites, paid more at the start and offered continued oppor
tunities for further advancement (Opinion at 12).
On July 29, 1963, three black laborers were for the first
time permitted to transfer to other jobs. The court below seems
to have treated this date as a watershed event, when discrimina
tion against blacks came to a halt (Opinion at 45). The court be
low made no specific finding of fact to this effect, however —
as indeed, on this record, it could not. Not only does the record
contain no evidence that opportunities to transfer were thence
forth made available to black employees on any significant scale;
now, over seven years later, the percentage of company employees
who are black has actually dropped, and the overwhelming majority
of black employees are still locked into the lowest paying, least
attractive jobs (Gov. Ex. 14).
We show below that Georgia Power continued to discriminate
against black employees after July 29, 1963 by its use of unlawful
employment tests and its practices of job assignment and recruit
ment. Each of these contentions was rejected by the district court
Continuing Patterns of Discrimination
9
and is in issue here. But, apart from these contested issues, the
record is replete with facts which were either found as true by the
district court or are not open to serious dispute by the parties,
and which clearly constitute examples of continued discrimination.
1. Until April 29, 1969, months after the institution
of these lawsuits, Georgia Power's black employees had no seniority
rights whatever which could be exercised for purposes of promotion,
advancement, or transfer. This situation was assured by the
singular status, under the 1965 and earlier collective bargaining
agreements, of Sections 17, 18, and 19. These sections consisted
of the job classifications of Laborers, Janitors, and Porters, and
Maids, respectively (Opinion at 11). These jobs, the lowest-pay-
11/ing in the Company (Opinion at 12), were and are black jobs.
Until April 29, 1969, all four jobs continued to be in separate
dead-end lines of progression, and employees in those job class
ifications had no seniority applicable to any other jobs in the
Company (Opinion at 11).
2. The collective bargaining agreement effective April
29, 1969 gave laborers, for the first time, limited seniority
rights for promotion and advancement in other lines of progression.
However, these black laborers were placed at the bottom of their
respective lines of progression and granted only job seniority in
11/ No whites have ever held the jobs of Porters, Janitors, and
Maids (Opinion at 13). Prior to July 29, 1963 all black employees
were in one of those job classifications or that of Laborer
(Opinion at 12). Prior to July 2, 1965, a great majority of white
employees were immediately assigned to higher-paying and more
responsible jobs than the four black jobs (Opinion at 12).
10
their entry-level jobs. They remained, therefore, not only
behind all their white contemporaries, but also behind all later-
hired whites who were assigned to entry level jobs in the non
laborers lines of progression at any time prior to April 29, 1969.
Defendants maintained this system until its modification was ordered
by the district court's decree of September 27, 1970 (US decree).
3. Even after July 29, 1963, defendants continued to
assign all newly-hired black employees to the job classifications
of Laborers, Janitors, Porters, and Maids. All 156 blacks hired by
Georgia Power in the period July 29, 1963 to June 9, 1969 were so
assigned, irrespective of their qualifications (Gov. Ex. 7,
Gov. Ex. 15 [suppl.]).
4. Janitors, porters and maids -- all of whom still are
black -- are still in separate "lines of progression". The ex-
12/
12/ The court found: "Under the terms of the current collect
ive bargaining agreement between the Defendant Company and
Defendant Unions, the Laborer's job classification (formerly
separate) has now been made the entry level job classification in
the lines of progression in 15 of 19 sections as that term is used
in the collective bargaining agreement, [citation omitted] The
term "entry level job" as used in this paragraph means simply the
job that pays lowest in a particular line of progression.
"As a result of the current collective bargaining agreement,
"Section XVII, Laborers in Any Department of the Company, in the
1965 Collective Bargaining Agreement" was abolished and persons
assigned to that section in which they were working as of April 29,
1969. The persons affected in this instance retained the section
seniority which they had accumulated in Section XVII and it became
their section seniority in the section to which they were assigned. . . " (Opinion at 10-11)
The former laborers' exercise of their Section XVII seniority
remained, of course, subject to defendants' other requirements
for promotion or transfer, most notably the tests.
11
pression is deceptive, since there is nowhere to progress to in
those lines.
5. The collective bargaining agreement effective April
29, 1969 created the new job classifications of "switchmen" and
"coal samplers" as purportedly nDn-laborer jobs. No whites have
ever held these jobs (Opinion at 14). In fact, they are laborers’
jobs, paying considerably less than the lowest paying white job13/(Gov. Ex. 2 ) .
6 . Earnings of black employees have been substantially
lower, when matched against those of their white contemporaries
and even against some later hired whites (Opinion at 41). Of the
Company's employees covered by the collective bargaining Agree
ment, hired between July 2, 1965 and September 6 , 1968 (after these
suits were instituted), 788 were white and 89 black. Their monthly
salaries as of September 6, 1968 break down, by race, as follows
(Opinion at 16) :
Monthly Salary White Black Percent Black
Over - $800 0 0 0$700 - $800 0 0 0$b00 - $700 15 0 0$500 - $600 517 0 0$400 - $500 236 1 0.4Under- $400 20 88 81.0
TOTAL 788 89 1 0 . 1
13/ Those two jobs were referred to in the proceedings below
by all parties and the court as "laborers" jobs. The reason is that
prior to April 29, 1969 the duties performed by switchmen and coal samplers were performed by laborers. On that date, the newly
agreed upon contract "reclassified" those laborers performing that work as switchmen and coal samplers, and provided for a token wage increase of $20 per month (Gov. Ex. 2).
12
7. Although some whites--albeit only a few— were
initially employed as laborers, the longest any such white who was
hired between July 29, 1963 and January 9, 1970 remained a laborer
was thirteen months, and the average time spent by whites as
laborers before promotion to a better job was three months and one
week (Gov. Exhs. 6 , 7, 15, 17 [suppL]). During the same period, how
ever, all of the blacks started as laborers, only 13 out of 95 have
been promoted since, and even they remained laborers an average
of two years and nine months (Id.).
8 . Apart from the forms of discrimination relating direct
ly to opportunities for transfer and promotion, the Company dis
criminated in other respects in its treatment of black applicants
and employees, causing the court below to enjoin continuation or
resumption of such practices. These include discrimination in re
imbursing employees for expenses incurred in working away from
home (Opinion at 23-24), and in treatment of applicants (Opinion
at 24-25) . As to hiring, the court noted that some of the Company's
"supervisory personnel have unilaterally engaged in acts to dis
courage, confuse, and by-pass blacks in favor of whites" (Opinion
at 46). Examples cited by the court include informing an apparent
ly qualified black that there were no openings and then hiring a
white shortly after; receiving a black's application when there is
no opening but not referring to it when one develops; the company's
"frequent practice of hiring white employees, when the applications
of equally or better qualified blacks are on file";its "practices
of informing black applicants that no openings are available when
in fact vacancies are available", "of informing blacks that
13
vacancies rarely occur when in fact they occur often", "of inform
ing blacks that they must begin as laborers when in fact a large
number of white employees begin at a higher classification", "of
telling black applicants that they will be called if a vacancy
occurs and failing to call them", and "of failing to inform black
applicants that they should apply at a different location for cer
tain jobs in which they express interest" (Opinion at 46-47).
The cumulative effect of so many types of discrimination in
so many parts of the company's operations serves, as a very minimum,
to dispel once and for all any notion that Georgia Power stopped
discrimination against blacks on July 29, 1963.
C. Testing Practices
On July 29, 1963, Georgia Power Company, for the first time
in its history, ostensibly ceased using race as a basis for restrict
ing all blacks to laborer category jobs and announced that black
employees would be allowed to advance into jobs above the laborer
classification (Opinion at 12). However, less than a month later,
on August 19, 1963, the Company imposed a new hiring and pro
motion requirement which had the effect of substantially rein
stituting the policy of excluding blacks from better jobs. This
new requirement required all incumbent employees seeking transfer
from formerly all-black job classifications (laborer, janitor,
porter, and maid), and all persons seeking new employment in job
classifications other than these classifications to take and
achieve a pre-determined passing score on each of one or more
aptitude tests. Incumbent employees in all other job classifica
tions -- the formerly all-white classificatinns -- had never been
and were not now required to take or pass any aptitude test
14
in order either to be promoted in their present lines of progression
or to be transferred to new lines of progression (Opinion at 13).
These new tests have effectively blocked any substantial upward
movement by blacks.
The Georgia Power Company's test requirement includes nine
different employment tests. Persons assigned to any job in the
"manual and hourly classifications" must pass both the verbal and
numerical components of the Personnel Tests for Industry (PTI-V
and PTI-N). In addition, assignment to several of these class
ifications also requires passing the Bennett Mechanical Compre
hension Test (Bennett). Jobs in the "office and clerical class
ifications" require passing of the Short Employment Tests (SET),
and applicants for several of these classifications must also pass
the General Clerical Test (GCT). All persons assigned to jobs in
the "technical classifications" must pass the PTI-V and the PTI-N.
In addition, several jobs within the "technical classifications"
also require the Bennett, and for several other jobs in technical
classifications persons must pass the Revised Minnesota Paper
Form Board Test along with the PTI. Finally, persons seeking jobs
in "miscellaneous classifications" — home economist and mer-
14/chandise salesmen -- must pass the Wonderlic Personnel Test
(Co. Ex. 1-B).
In every case the employee must pass every test required for
the job. Once an employee passes the entire test battery he be-
14/ The tests required for each position and the required scores
are set forth in Appendix B to this brief, p. B-3 .
15
comes eligible for promotion into all jobs in the line of pro
gression without ever being retested. If he fails to achieve the
passing score on any one of the tests required, he is denied the
job, regardless of his other qualifications or scores on the other
tests (Opinion at 29). This inflexible requirement applies to any
job outside the laborers' category, including the entry level
helper jobs in traditionally white lines of progression. Thus,
plaintiff Charles King, who successfully passed two of the tests
and failed the third by only one point, was ineligible to become
a helper (Co. Ex. 83) .
The Supreme Court has already acknowledged that two of the
tests used by Georgia Power, the Bennett and the Wonderlic, have a
grossly discriminatory impact on blacks. Griggs v. Duke Power
Co., 401 u.S. 424, 430 (1971); see also Hicks v. Crown-Zellerbach
Corp., 319 F.Supp. 314, 319 (E.D. La. 1971). The other tests have
similar discriminatory effects. The district court below found
that "as a general rule, black people score significantly lower
than do white people on aptitude and intelligence teste of the
type used by Georgia Power Company" (Opinion at 30). The full
extent of this discriminatory impact is not clear from the record,
since the only data available consisted of test scores of present
employees tested and hired since 1968. This group included very
few blacks. Blacks had until recently been discriminatorily
excluded from testing through the company practice of assigning
most blacks to laborer positions without testing and having them
16
sign "waiver of promotion" forms. Within this post-1968 group,
which is in no sense a random sample, blacks generally do far less well
16,than whites on the tests used by Georgia Power Company (Gov. Ex. 12).
Approximately one year after the adoption of these tests by
Georgia Power Company, John Hawkins revised the passing score re
quirements for the test. Hawkins, a political science major and
admittedly not a professional psychologist, worked alone. These
revised scores were based on a visual, and acknowledgedly non-
statistical, study of how test scores related to job performance
(6 Tr. at 203-213). Georgia Power undertook no further study until
the institution of this action. The test and passing score re
quirements as fixed by Hawkins have remained in effect to date,
15/
15/ These "waiver of promotion" forms were regularly required even
of high school graduate blacks, and were often requested without any
explanation of the implications, or opportunity to take the tests.
Prior to the filing of this suit, only one white had signed a similar
waiver. During the period from January 1, 1968 to March 20, 1970, for which complete data are available, 62 blacks and only 9 whites
executed waiver forms. (1 Tr. 45, 102, 127, 129, 179, 186, 189, 223,
231; 2 Tr. 43, 45, 121-122, 130, 221, 231; Gov. Ex. 6 , Co. Ex. C,Co. Ex. 6 , Gov. Ex. 12).
16/ The following shows the comparative rates of failure for blacks and whites within this sample group.
Test Blacks Whites
PTI-V 30.0% . 94%PTI-N 43.0% 1.25%Bennett 37.5% .85%SET 18.0% .003%GCT 56.0% .00%Minnesota no blacks tested 4.00%
Source: Gov. Ex. 12.
The district court also found, for this group, that blacks scored 9-12 points lower than whites on the Georgia Power tests
(Opinion at 30-31).
17
except fot minor modifications ordered by the district court
(6 Tr. 165, Co. Ex. 1-B, US decree at 3).
In December 1968, Georgia Power Company began a study of its
testing program under the direction of Dr. Lorain Hite, an indus
trial psychologist, which was completed just shortly before the trial
commenced (Opinion at 33; 6 Tr. 35). This study is reported as
"Validation Data and Procedures for the Employment Testing Program
of Georgia Power Company" (Co. Ex. 75). At trial Dr. Hite also
testified concerning a summary sheet of data derived during the
preparation of this study (Gov. Ex. 41) .
Dr. Hite's study is limited to validation data for only
thirteen of the company's over fifty job classifications. He did not
even attempt validation for the Minnesota test and the Wonderlic
test. The PTI tests, the Bennett test, the SET tests, the GCT tests
were studied, where appropriate, for only these thirteen job class
ifications. For at least one line of progression at the Company -
the serviceman line - no job within the line was included under the
study (Co.Ex. 75; Co. Ex. E).
Dr. Hite in the course of this study tried at least three
different statistical procedures (discriminant function analysis,
multiple regression analysis, and a statistical comparison of the
performance of tested versus non-tested personnel), in an attempt to
find some basis for justifying test use. However, he placed primary
reliance on discriminant function analysis (Co. Ex. 75, p. 1), in
implementing this analysis, as reported in the study, Dr. Hite used
statistical techniques to produce an optimal weighting of test scores,
which differed for each job. His study reports data only as to the
use of tests in accordance with that optimal weighting. In every
18
instance this weighting is inconsistent with the company's existing
use of the tests, and in many cases grossly inconsistent. Thus,
for the lineman classification, the optimal test weighting called
for elimination of the PTI-V score, contrary to the company's
practice. For seven of the twelve reported classifications, under
the discriminant function analysis — appliance servicemen, store
keepers, helpers, coal equipment operators, meter readers, switch
board operators, and garage mechanics — the optimum formulae
assigned negative weights to at least one test, thus suggesting that
persons scoring lower on the test should be given preference over
higher scorers, contrary to the company's present practice (Co. Ex. 75)
Even if the tests had been used according to Dr. Hite's op
timal weighting, his own summary data sheet (Gov. Ex. 41) reveals
that under the discriminant function analysis the results reported
satisfy standard measures of statistical significance only as to two
17/of thirteen job classifications reported upon. The multiple re
gression results are statistically significant for no job class-18/ —lfication (Gov. Ex. 41).
The District Court agreed that Dr. Hite's results indicated
that the tests were not of "significant help to the Company in pre
dicting job performance of applicants" for at least four jobs --
switchboard operator, garage mechanics, coal equipment operators,
and helpers (Opinion at 34). Based on this finding, the court en-
17/ See pp.40 - 43, infra, and Appendix B, pp. 4-5 , infra, for dis
cussion of the concept of significance and its application here.
18/ See p . 41 n,39, infra, and Appendix B, pp. 4- 5 , infra.
19
joined further use of the tests as to three of these four jobs
(US decree at 3). However, the tests remain in use, without any
change or modification from the company's longstanding practice, for
the fourth job (helper), and for all other jobs at the company. As
a practical matter this means that passage of the test battery is
a prerequisite to any significant promotional opportunity in the
Georgia Power Company.
D. Assignment Practices
As shown above, until July 29, 1963 Georgia Power Company
restricted all black employees to laborer category jobs as a matter
of policy (Opinion at 12, 41). On that date, the Company announced
a policy change with respect to allowing blacks to compete for pro
motion, but remained silent as to its policy with respect to the
initial assignment of blacks. The record in this latter regard
clearly demonstrates that no change in the Company's policy of
assigning black hirees to laborer jobs exclusively was contem
plated; or if such a change were contemplated, it was certainly
not implemented. Georgia Power continued to exclude virtually all
blacks, regardless of their qualifications, from assignment to jobs
higher than laborer until at least 1969, long after these actions
had all been filed.
In the period July 29, 1963 to June 9, 1969, all 156 blacks
hired by Georgia Power were intially assigned to laborer jobs, irre
spective of their qualifications (Gov. Ex. 15 {suppl.], Gov. Ex. 7).
Although most of these black hirees admittedly did not meet Company
20
standards for non-laborer jobs, this fact cannot explain the assign
ment of all these blacks to laborer positions. In most cases,
Georgia Power did not allow blacks an opportunity to qualify: 41
of these 156 persons had high school degrees (Gov. Ex. 17 [suppl.],
Gov. Ex. 6 , 7), but only seven of them were tested prior to hiring
(Gov. Ex. 15 [suppl.], Gov. Ex. 6 ). Two of those tested before hir
ing passed the tests, but both were nevertheless assigned to laborer
jobs (Gov. Ex. 15 [suppl.], Gov. Ex. 6 , Gov. Ex. 7). A total of 28
black post-July 29, 1963 hirees were fully qualified under the
Company's standards by January 16, 1970 (Opinion at 18, 44). Of
these 28 black employees, 23 were initially assigned to laborer
jobs and only five were initially assigned to higher jobs (Id.).
Even the latter finding of the trial court insufficiently sets
forth the facts, in that it fails to indicate that all five blacks
assigned to non-laborer jobs were employed after June 9, 1969, as
the record shows (Gov. Ex. 15 [suppl]. Gov. Ex. 14 [suppl.],
Gov. Ex. 7).
Georgia Power accorded white employees hired after July 29,
1963 markedly different treatment. The court below found that of
627 whites hired between July 29, 1963 and January 10, 1969 in the
Steam Plants, General Repair Shop, and Atlanta and Macon Operating
Divisions, 534 were assigned to jobs other than laborer
19/
19/ We argue elsewhere that these standards were unlawful, and
the Court so found with respect to the high school education requirement.
21
(Opinion at 19). The court further found that 542 of the 600 whites
hired between July 2, 1965 and January 16, 1970 in the Steam Plants
were assigned to non-laborer jobs (Opinion at 15). Moreover, white
hirees were given every opportunity to qualify for higher level
jobs. For example, of 624 whites hired into the Steam Plants and
Atlanta and Macon operating divisions between July 29, 1963 and
June 9, 1969, 594 were tested piror to their hiring; and of the 30
not tested and therefore not qualified for non-laborer jobs, 25
were initially assigned to higher level jobs anyway (Gov. Ex. 17
[suppl.], Gov. Ex. 14 [suppl.]).
21/E. Recruitment Practices
22/
As set forth above, virtually all Georgia Power's black em
ployees have been relegated to the lowest-paying jobs in the Company,
in the laborers category, while white employees monopolize jobs in
in the higher classifications. The court below found that Georgia
20/
2_0/ The court also found that the other 93 white employees re
mained as laborers for an average of only three months and one
week prior to being promoted into a higher job (Opinion at 19)
By way of contrast, education- and test - qualified blacks remained
as laborers for an average of two years and nine months before being
promoted (Gov. Ex. 4, 5, 12; Co. Ex. 7). The court below obscured this glaring contrast, but admitted that "black employees in the
laborer classification average a longer period of time than whites prior to promotion" (Opinion at 16).
21/ The court below stated that "'[r]ecruiting' relates to affirmative efforts by the company to attract applicants, while 'hiring'
relates to the processing of applicants regardless of their source."
(Opinion at 45-46). Appellants adopt this distinction in this part of the brief.
22/ See pp. 9-13, supra.
22
Power relies to a substantial extent on word-of-mouth recruitment
by incumbent employees as a method of informing persons of job
vacancies and of finding persons to fill those vacancies (Opinion
at 28). It also found that "many" employees of both races had
learned of job vacancies through incumbent employees; approx
imately 30 percent of the persons employed by the Company as of
January 16, 1970, had friends or relatives working for the Company
at the time of their employment (Ic3.) . The Company also places
substantial reliance on walk-in application as a method of obtaining
applicants for vacant positions (1 Tr. 167; 2 Tr. 8, 170; 4 Tr. 213).
The court found that most of the engineering colleges at
which the Company maintains recruitment programs have predominantly
white student bodies (Opinion at 29). The Company relies primarily
on these institutions, which are located throughout the South, as
sources of executive personnel (Id.).
The court below found no evidence of any affirmative program
of recruitment, other than that summarized immediately above. And
indeed, the entire record contains no evidence of any such program,
of whatever sort. Georgia Power's recruitment "policy" consists
of acceptance of the consequences of the situation it occupies as a
major employer throughout Georgia.
23
ARGUMENT
Introduction
The facts in this case show that the date of July 29, 1963
marked merely a transition from an open and declared rule that
blacks were to be confined to menial work without exception, to a
covert policy to permit a few token blacks to rise a notch or two
and employ a few token whites for a while in the lowest positions,
but nevertheless to adhere to the old rules to the greatest ex
tent that the defendants thought they could get away with. The
court below failed to recognize or correct several of the most
egregious practices maintained by Georgia Power pursuant to this
policy. In this argument, we show that the court below erred and
must be reversed with respect to those policies relating to test
ing, job assignment, and employee recruitment. Next, we urge
that the relief accorded by the court below was wholly inadequate,
and that this "relief" itself perpetuates the effects of race dis
crimination. In particular, we argue that the district court's
remedy was legally deficient with respect to the adequacy and
scope of compensatory seniority, and adequacy of back pay awards
to the named plaintiffs, the availability of class-wide back pay, the
applicable statute of limitations, and the adequacy of attorney's fees.
I. GEORGIA POWER COMPANY'S USE OF EMPLOYMENT
TESTS VIOLATES TITLE VII SINCE THOSE TESTS EXCLUDE BLACKS FROM HIRING OR PROMOTION
INTO BETTER JOBS AND SINCE THE TESTS' JOB
RSLATEDNESS HAS NOT BEEN ADEQUATELY
DEMONSTRATED.
In confronting the issue raised by Georgia Power's employ
ment testing program, this Court must reach questions not presented
24
or decided in the landmark case of Griggs v. Duke Power Company, 401
U.S. 424 (1971). Nevertheless, we submit, application of the
principles established by Griggs must lead this Court to hold
Georgia Power's tests unlawful under Title VII.
The Griggs decision laid down the courts' basic approach for
Title VII review of employment tests. Under Griggs, the employer's
"good intent or absence of discriminatory intent" is unimportant.
401 U.S. at 432.
Rather, in fhe words of the Supreme Court:
The touchstone is business necessity. If
an employment practice which operates to
exclude Negroes cannot be shown to be
related to job performance, the practice is prohibited.
401 U.S. at 431. And, the Court has made it clear that the employer
has the burden of proof in establishing this business necessity.
Congress has placed on the employer the
burden of showing that any given require
ment must have a manifest relationship
to the employment in question.
401 U.S. at 432. There can be no question that the tests used by
Georgia Power operate to exclude Negroes in the sense referred to23/
in the Griggs case, and the district court so found (Opinion at
30-31). Thus, on this appeal the only issue is whether Georgia
Power has satisfied its burden of proving business necessity
sufficient to justify its use of discriminatory tests.
Realizing its affirmative burden, the Company introduced a
"validity study", conducted by one Dr. Lorain Hite, which the
2 3/ See pp. 16-17 , supra .
25
Company claimed satisfied the requirements of Griggs. The dis
trict court concurred with the Company's position (Opinion at
51-52). This ruling was in error. The Hite study does not prove
business necessity under any reasonable standard, as our analysis
24/
shows. If the courts permit employers to satisfy their burden
of justification with a study such as the Hite document, the rule
of Griggs will be wholly frustrated. Employers will then merely
be required to go through a hollow statistical exercise to pro
duce a few meaningless numbers for cursory judicial inspection.
The Griggs decision means more than this. The courts must es
tablish standards which assure that the employer's proof of test
validity is meaningful. The present case presents this Court
with the task of determining those standards.
A. The Equal Employment Opportunity Commission
Guidelines on Employee Selection Procedures
Define the Appropriate Standards of Test Use.
Since Georgia Power's Testing Practices do
Not Conform to These Standards, They Should Be Declared Unlawful.
Full evaluation of Dr. Hite's validity study and assess
ment of its adequacy is a necessarily technical and difficult
job, because the study itself is highly technical. Although we
present this technical analysis below, we first urge that it was
inappropriate for the district court to attempt its own analysis
24/ Because of its numerous serious deficiencies, the Hite study
proves little or nothing. To the extent it proves anything, it
proves that Georgia Power's use of the tests is wholly unjustified
and irrelevant to job performance. See. pp. 35-44 , infra.
26
of the adequacy of the Company validity study. The degree of
confusion which the district court not surprisingly evidenced
shows why the courts are ill-equipped to undertake such technical
26/analyses. Indeed it is impossible to read the transcript of
the trial below without sympathizing with the district judge as
he struggled with psychological statistics to master the study's
technical details and establish standards for assessing its
2 7/
adequacy.
25/
25/ The district judge himself expressed his dissatisfaction at
being placed in a position which he did not have the technical ex
pertise to fill:...this has happened a lot in this EEOC business, that
everybody passes the buck around at the administrative
level and says we don't know the answer, let it go to
court and let the judge figure it out. We don't have
the expertise to do these things. That's what bothers
me, so now I have had about a seven hour course in
psychological testing and yet I may have to rule its
applicability to 8,000 people. (6 Tr. 142).
As we argue in the text, the court below erred in ignoring the
alternatives to his plight.
26/ The opinion and decree bear witness to the district court’s per plexity. Their remedial provisions are in many ways inexplicable in
terms of the court's findings of fact and what the Hite study showed
Thus, for example, the court permitted continued use of the tests for helper positions even though the court found the tests to be of
no "significant help to the company in predicting job performance
of applicants" for those positions (Opinion at 34). The court also
allowed continued use of the Wonderlic test— which was barred in
Griggs— even though the study did not even include that test and no
other evidence was introduced which would support its use.
27/ See 3 Tr. 20, 23-24, 25, 27-28, 40, 47, 51, 6 6, 70, 73, 139,
142, 144, 147, 184, and 6 Tr. 63, 64, 81, 91, 98, 101, 111, 124-125,
14 2, 143, 144, 157.
Perhaps the most poignant comment by the judge was the
following:
"I might be disqualified in this case. My daughter flunked
statistics in college and I am beginning to see why she did. Would
you all accept my disqualification at this point?" [not accepted]
(6 Tr. 157).
27
It was unnecessary for the district judge to become so en
meshed in the technical details of establishing a standard for
review of validity studies. The Equal Employment Opportunity
Commission (EEOC), the agency charged with enforcement of Title
VII, has published extensive Guidelines on Employee Selection
Procedures, 2 9 C.F.R. §L607 (1970), in CCH Empl. Prac. Guide f 16,904,
which provide a complete set of standards for reviewing test use.
The court below, however, refused to follow the Guidelines.
It rejected the Guidelines despite their stated purpose of pro
viding
a workable set of standards for employers,
unions, and employment agencies in deter
mining whether their selection procedures
conform with the obligations contained in Title VII of the Civil Rights Act of 1964 (29 C.F.R. §1607.1 (c) ) .
The court also declined to recognize that the Guidelines impose no
extraordinary requirements, but rather permit "any appropriate
validation strategy" in conformity with established professional
standards set forth in the Standards for Educational and Psycho
logical Tests and Manuals of the American Psychological Association
29 C.F.R. §1607.5 (a) - (b) . The court thought that the Guidelines
established an unnecessarily strict standard. Indeed, the district
judge appears to have mistakenly believed that no available test
could meet the EEOC Standard; he noted at one point that
the rather startling evidence offered by the
government was to the effect that there was no
test known to exist or yet devised which could meet such standards.
(Opinion at 33, 51 n.8) This finding was clearly erroneous and
wholly unsupported in the record. The government made no such claim
28
and if it had, the claim would have been false. The EEOC, acting
pursuant to its Guidelines, has frequently approved test use both
28/in published and unpublished decisions.
The district court erred in failing to follow the EEOC Guide
lines. The value of these Guidelines has been emphasized by a
number of courts. They formed the basis of the Griggs decision,
where the Supreme Court said.
Since the Act and its legislative history
support the Commission's construction, this
affords good reason to treat the Guidelines
as expressing the will of Congress.
401 U.S. at 435. This Court has also stressed the Guidelines'
28/ See e.g., unnumbered decision, Dec. 6 , 1966; Decision 71-
1471, March 19, 1971 (CCH Empl. Prac. Guide f6220); Decision 71-
1525, March 26, 1971 (CCH f6224); Decision 71-1418, March 17, 1971
(CCH f6223) ; Decision 70-630, March 17, 1970 (CCH 16136).
The district judge's mistaken belief is evidently the result
of an erroneous inference drawn from testimony of the government's
expert witness, Dr. James J. Kirkpatrick. Dr. Kirkpatrick testified
that a test having a discriminatory impact on blacks would meet Guide
lines standards only if "it was shown that Negroes really were poor on the job to the same degree they were poorer, lower on the tests"(3 Tr. 197) or if an adjustment were made in test scores to eliminate
the discriminatory impact (3 Tr. 196). This colloquy then ensued:
The Court: Do you know of any tests that the
Power Company could come by that would meet
this standard you have set? I don't believe
there is one?
The Witness: I don't think so either; it'sa question that has to be answered by re
search and not by pulling a test off the
shelf or trying it from the publisher's
catalogue (3 Tr. 197-198).
The court may have understood Dr. Kirkpatrick to mean that he knew
of no test which met his standards. But the witness's remark must
be interpreted in context. The same witness testified earlier that
other companies had testing programs which met his standards (3 Tr.
179-181). His later comment thus plainly meant only that no test
which is used by a company without studying its racial impact and
job relatedness meets his standards. That is hardly an extraordinary
thought in light of the Griggs decision.
29
importance in United States v. Jacksonville Terminal Company, F . 2d
___ , 3 EPD 18324 (5th Cir. 1971). In holding that the defendant
employer had not adequately validated these tests, the Court there
observed:
Certainly the safest validation method is that
which conforms with the EEOC Guidelines
'expressing the will of Congress.'
3 EPD 18324 at 6993-149. At least two district courts have re
cognized the reasonableness and persuasive authority of the EEOC
Guidelines and have formally adopted them as the appropriate standard
for validation. In Hicks v. Crown-Zellerbach Corp.. 321 F.Supp.
1241 (E.D. La. 1971), Judge Heebe enjoined the use of any test un
less it had been
validated and proven valid in accordance with
the requirements of the 'Guidelines of Employee
Selection Procedures' published by the United
States Equal Employment Opportunity Commission.
321 F.Supp. at 1244. The court imposed a similar requirement in its
decree in Carter v. Gallagher, ___ F.Supp. ___ , 3 EPD 18205 at
6682-83 (D- Minn, 1971), aff'd in pertinent part, F.2d , 3 EPD2 9/
18 3 3 5 (8th Cir. 1971) .
2_9/ Other governmental agencies have also approved the Guidelines as standards for employment testing. The Office of Federal Contract
Compliance (OFCC) of the United States Department of Labor - the
agency charged with supervision of the fair employment program for
government contractors emanating from Executive Order 11246 - after
extensive study and consultation with a board of expert advisors including representatives of major businesses, promulgated its order
on employee testing and other selection procedures, which is virtually
identical to the EEOC Guidelines. See 36 Fed. Reg. 19307 (Oct. 2,
1971), in CCH Emp. Guide f 17,589 revising 41 C.F.R. §60-3. The
Pennsylvania Human Relations Commission has also recently published
new testing guidelines which are virtually identical to the EEOC
Guidelines. See Pennsylvania Guidelines on Employee Selection
Procedures, 1 Pa. Bull. 2005 (Oct. 16, 1971), in CCH Emp. Prac.Guide f5194.
30
The conflict between the adoption of the Guidelines by the
district courts in Hicks and Carter and the rejection of the Guide
lines by the court below lies at the heart of this appeal. If this
Court resolves the conflict as we urge and adopts the EEOC Guidelines
as the appropriate standard of test use, that resolution will dis
pose of the testing issue in this case, since there is no dispute
about the fact that Georgia Power's validation has not met the
standards of the EEOC Guidelines. The district court found
It is true that the validation does not meet
the rigid requirements recently set out by
the Equal Employment Opportunity Commission,
29 C.F.R. Section 1607. . . .
(Opinion at 51 n.8). Moreover, the expert witness for the Company,
Dr. Hite, in effect conceded in this lengthy testimony that his
30/
study and results did not comport with EEOC Guidelines standards.
This Court need not make any detailed assessment of the Hite study
if it adopts the EEOC Guidelines as its standard, because the study's
failure to meet those Guidelines is conceded.
We suggest at least four reasons why this Court should adopt
the approach of the Hicks and Carter courts and establish the EEOC
Guidelines as the applicable standard for test review.
1) The Guidelines are the expert interpretation of the agency
charged with carrying out the statutory purpose. The EEOC is, it
self, a creature of Title VII. It was created by Title VII for the
express purposes of settling charges of discrimination (42 U.S.C.
§§2000e-5, -8) and assisting employers and others in complying with
30/ 6 Tr. 89, 94-95, 97, 130.
31
the statutory requirements (§2000e-4 (f)). The EEOC is given
specific authority to issue written interpretations of the statute
and good faith reliance on such interpretations is made a defense
to any charge of discrimination §2 000e-12 (b)) . Thus Congress gave
the EEOC a central role in developing and elaborating the bare
bones of the statute.
This fact does not make EEOC Guidelines binding, per se,
on the courts. But it clearly entitles the EEOC Guidelines to
judicial deference, under the "venerable principle that the con
struction of a statute by those charged with its execution should
be followed unless there are compelling indications that it is
wrong." Red Lion Broadcasting Co. v. F.C.C.. 395 U.S. 367, 381
(1969); Udall v. Tallman, 380 U.S. 1, 16 (1965). And the Supreme
Court has clearly indicated that the EEOC Guidelines present a
situation, where "the administrative interpretation of the Act
by the enforcing agency is entitled to great deference," Griggs
v. Duke Power Co., supra at 433-434.
2) The EEOC Guidelines involve an interpretation
of complex and technical matters. The "venerable principle" of
deference to administrative agency interpretations has particular
application when the matter involved is highly complex and tech
nical. The EEOC Guidelines represent the end product of a major
technical effort involving industrial and educational psycho
logists and personnel experts, as well as lawyers. The Guidelines
deal with technical matters such as the difference between
"criterion-related validity" and "content validity", and when
32
each is appropriate, and the statistical standards appropriate
for determinining when data from a sample are sufficiently reliable.
In such matters as these a court's "instant expertise" is
unlikely to be as discerning as the well developed expertise of an
administrative agency. The Supreme Court found this reason per
suasive in S.E.C. v. New England Electric System. 384 U.S. 176
(1966), where it upheld the S.E.C. interpretation of a provision
in the Public Utility Holding Company Act of 1935 as "well within
the permissible range given those who are charged with the task
of giving an intricate statutory scheme practical sense and appli
cation, " I_d. at 185. Similarly, in Hardin v. Kentucky Utilities
Co., 390 U.S. 1 (1968), the Court upheld an administrative inter
pretation of the Tennessee Valley Authority Act, in part because
of "the complexity of the factors relevant to decision in this
matter, " l_d. at 9. See also Power Reactor Co. v . Electricians.
367 U.S. 396 (1961) (AEC regulations).
3) The Guidelines provide a reasonable standard. Deference
to administrative interpretation does not require a court to accept
unreasonable interpretations; but there is nothing unreasonable
about the Guidelines. The basic approach of the Guidelines is to
adopt recognized professional standards set out by the American
Psychological Association,29 C.F.R. §1607.5 (a). The Guidelines
permit validity to be established through a wide variety of means
including citerion-related validity studies, content and con
struct validity studies (§1607.5), and use of validity studies
conducted by groups of employers in the same industry
33
(§1607.4 (c) (2)). The Guidelines repeatedly provide that com
pliance with detailed rules is required only to the extent "tech
nically feasible" (§§1607.5 (b) (1) . 1607.5(b)(5)). The decisions
of the courts and agencies which have endorsed the Guidelines
necessarily imply a finding of reasonableness.
4) The EEOC Guidelines are supported by the legis
lative history of Title VII. A court should not defer to an
administrative interpretation which conflicts with the legisla
tive history of a statute; but there is no such conflict here.
While Title VII does not, of course, speak of such detailed matters
as validity studies, it does suggest the basic tone and purpose
of the law. And as Griggs affirms, the Guidelines are fully
supported by this history.
Since the Act and its legislative history
support the Commission's construction, this
affords good reason to treat the guidelines
as expressing the will of Congress
401 U.S. at 434.
* * *
From the sum of the legislative history relevant in the case, the conclusion is
inescapable that the EEOC's construction of §703 (h) to require that employment
tests be job-related comports with con
gressional intent. Id.* at 436.
For these reasons, the Court should adopt the EEOC Guidelines
on testing as standards of law under Title VII. As judged by those
standards, the Georgia Power Company's tests are unlawful.
B. Even If This Court Chooses To Develop Its Own
Standard For Review Of Validity Data, It Should
Enjoin The Georgia Power Testing Program. The
Hite Study Fails To Demonstrate Job Relatedness
Or Business Necessity As Measured By Any
Reasonable Standard.
34
We have argued above that the Hite study's failure to meet
the standards of the EEOC Guidelines is not because the Guidelines
impose any extraordinary requirements, but rather reflects the
study's own deficiencies. We now urge that, apart from the EEOC
standards, the Hite study completely fails, under any meaningful
standard, to demonstrate that Georgia Power's tests are job
related or required by business necessity. Therefore, even if this
Court chooses not to adopt the Guidelines standards formally it
should nonetheless enjoin Georgia Power's testing program.
The insufficiency of the Georgia Power validity study does not
stem solely or even principally from its technical defects -- al
though in fact these numerous important defects (summarized below
and described more particularly in Appendix B to this brief) cast
serious doubt on the study's conclusions. Rather, the study's
failure lies primarily in the fact that, even if it were assumed
to be technically perfect, its results do not offer a shred of
rational justification for Georgia Power's testing program. In
the first place, the Hite study attempted to validate a totally
different test battery from that actually in use at Georgia Power.
Moreover, the validity results produced by the Hite study are not
statistically significant. Both of these reasons, as well as its
technical defects, compel us to conclude that the Hite study, as
a validation of Georgia Power's existing (not hypothetical) testing
program, is valueless.
1) The Hite study is an attempted validation of a
totally different test battery from that actually used by Georgia
Powe r.
35
Dr. Hite studies a hypothetical testing program which differed
from Georgia Power's program in a number of significant ways. Since
Dr. Hite's purported validation was not addressed to the Company's
actual testing practices, its result can logically offer no justi
fication for those practices.
A brief description of Dr. Hite's methodology will make our
point clear. With one exception (for auxiliary equipment operator).
Dr. Hite relied on a method he termed "discriminant analysis" to de
termine test validity. Under this method he examined test scores of
samples of employees presently working in various jobs at Georgia
Power, and also had the same employees rated by their supervisors
(Co. Ex. 75, pp. 1-2). Although the tests involved were the same
ones used by Georgia Power to select persons for the jobs in ques
tion, Dr. Hite did not report how the tests worked when used in the
manner that the Company used them. The Company requires that a
person obtain a minimum passing score on each test: failure on a
single test blocks access to the job even if the person passes the
other tests with flying colors. If Dr. Hite had been validating
the Company's test system, therefore, he would have assessed the
job performance of persons who failed a single test in comparison
to that of persons who passed all the tests. Dr. Hite did not do
this. He testified that he did no computations based on Georgia
Power's method of using test scores (6 Tr. 130, 165-167), and in
deed that he made "no study of the Company's method whatsoever"
<6 Tr. 167) .
Instead of studying the Company's method Dr. Hite developed
from his data a new, independent weighting formula for each job, whic
combined all the test scores into a single score referred to as
36
the person's "total points" (6 Tr. 130, 165-168). Dr. Hite math
ematically derived these formulae in such a way as to produce the
weighting which would maximize the tests' predictive value (Co. Ex.
75, p. 1; 6 Tr. 168). Dr. Hite then ranked employees in each job
according to their "total points" and compared the job performance
of the high scoring group or groups with that of the low scorers.
The differences between the weighted formulae which Dr. Hite
studied and the separate cut-off scores which the Company uses
are like day and night:
(i) First, the mere step of shifting from separate cut-offs on
each test to combined scores,even if unweighted, may significantly
change results as to a person who scored very well on some tests
and marginally failed one. There are many such persons, of whom
3J_/plaintiff King is typical.
(ii) Next, the weighting used by Dr. Hite exaggerates the im
portance of some tests and de-emphasizes others. In one formula
(linemen). Dr. Hite actually determined that one of tests (the PTI-V)
had so little value that it should be entirely dropped from use (6
Tr. 148). In every formula one or more of the tests was given a
relatively insignificant weighting. For example, for winch
truck operators the PTI-N and the Bennett are each weighted
31/ For example, for the mechanics' category jobs, Dr. Hite derived
the following formula: "Total Points = .96 (PTI-V Raw Score) +2.3
(PTI-N Raw Score) +2.09 (Bennett Mechanical Raw Score). "(Co. Ex. 75, p. 16) .
The other formulae are set out in Dr. Hite's validity study,
Co. Ex. 75: p. 5 (linemen), p. 11 (winch truck and truck operators),
p. 23 (clerical), p. 32 (appliance servicemen), p. 36 (storekeepers),
p. 40 (metermen), p. 45 (helpers), p. 51 (coal equipment operators),
p. 56 (meter readers), unnumbered appendices (switchboard operators garage mechanics).
32/ King was ineligible for promotion because he failed the Bennett
by a single point (29, passing score 30). He passed the other two
tests in that battery, the PTI-V (31, passing score 16) and the
PTI-N (13, passing score 12.) Because of his high PTI-V score,King's combined score (73) was well in excess of the combined passing scores (53) (Gov. Ex. 12) .
37
.218 while the PTI-V is weighted 3.008 -- almost fifteen times as
heavy. For clerical employees the GCT is weighted .00014 while
the SET is weighted .01051 -- seventy-five times as heavy. The
results of such drastically weighted test batteries must vary
significantly from those produced by a simple, separate use of each
test on equal footing. The individuals who score high or low on
a combined and weighted battery may be almost totally different
from those who simply pass or fail each test.
(iii) Finally, and most strikingly, in seven of the twelve
weighted formulae there is a negative weight assigned to one or
33/more tests. This means simply that the higher a person scored
on that particular test or tests, the lower his "total point" score
would be. Since in every case the Company requires a person to
"pass" such a test by exceeding a certain score, the Company is
usina the tests in a manner exactly opposite to that called for347
by Dr. Hite's weighted formulae.
Dr. Hite clearly recognized that his weighting formulae
produced a system quite different from that in use at Georgia Power,
33/ See n.31, supra, for formulae citations.
34/ The district judge stated this point well in relation to the
switchboard operator job, for which the formula assigns negative
weights to two of the three tests involved (-2.5070 to the PTI-V
and -6.5607 to the Bennett, along with 4-8.9266 for the PTI-N) .In such a case, the judge observed, it becomes questionable". . .
Whether you have a ‘must pass'- ...You might have to have a 'must
flunk' in order to be a switchboard operator."(Hite testimony, 6 Tr.
147-148). The same observation would apply equally well to one or more tests given for the following seven of the twelve jobs analyzed
switchboard operators, appliance servicemen, meter readers, store
keepers, helpers, coal equipment operators, and garage mechanics
(See formulae citations in n.31, supra.)
38
and he recommended that the company use the weights he developed
(6 Tr. 150-151). The Company's present system is therefore in
consistent with its own expert's recommendations. Moreover, the
discriminatory impact of the Company's actual system on blacks far
exceeds that which use of Dr. Hite's formulae would produce. since
blacks generally score lower on Georgia Power's tests, the negative
weights assigned for some tests in seven of twelve formulae would
give an advantage to low-scoring blacks which would partially
offset the severe disadvantage which tests presently impose on them.
2 ̂ The Hite study fails to produce statistically sig
nificant data. While the foregoing section indicates that Georgia
Power can find no support for its present test practices in the
Hite study, it leaves open the possibility that the Company might
wish to modify its practices and make use of Dr. Hite's weighted
formulae. This too would be impossible, however. The data pro
duced by the Hite study are not statistically significant enough
to justify even use of the weighted formulae. And because dis
criminant analysis maximizes the predictive value of the tests,
it should be expected that the Company's actual test program has
even less validity than it would have under this discriminant
analysis technique. ( 6 Tr. 167-168).
25/ In some instances, use of the combined weighted scores
ven produce a hiring preference in favor of blacks For pvamnio
Dr. Hite s formula for the meter reader jcS wSSd p i o d ^ e an
average "total points" of 70.6 for blacks and 65.6 for^hi?es (based
on average scores of the limited sample of persons hi£ed and t2 ?ed
betweewanuary 1 . 1968 and March 20? I O V O W gov eS L ; So."ex
39
The concept of "statistical significance" is critical to
any assessment of data from a validity study. The data from any
experimental study will indicate that certain results were obtained,
but there is always a risk those results are coincidental, since
they are based on only a small experimental sample. The statis
tician, therefore, analyzes his data to determine how confident he
can be that the results are more than coincidental. This is ex
pressed in terms of levels of confidence. The .01 level means that
there is only 1 chance out of 100 that the results were coincidental;
the .05 level means that there is 1 chance out of 20. Unless a
sufficient level of confidence is reached, the results are too
unreliable to serve as the basis for any meaningful conclusions.
(3 Tr. 26-27, 6 Tr. 94-95). According to the accepted pro
fessional convention adopted by textbook writers and professional
statisticians, the level of confidence should reach at least the36/
.05 level (and preferably the higher .01 level).
Dr. Hite prepared a chart, introduced at trial as Gov.
Ex. 41, on which he computed the statistical significance of his
37/data. This chart shows, in one series of columns, statistical
measures of the results which Dr. Hite obtained (Columns (3), (7)
36/ A.L. Edwards, Statistical Methods (2nd ed, 1967), p. 184;
W. Hays, Statistics for Psychologists (1963), p. 269; M.H. Walker, Elementary Statistical Methods (1943), pp. 290-292; C.C. Peters
and W.B. Van Voorhis, Statistical Procedures And Their Math
ematical Bases (1940). pp. 175-176; I.D.J. Bross, Design "for
Decision (1953), pp. 2-21-222.
The EEOC Guidelines also adopt the .05 level as the minimum
standard. 29C.F.R. §1607.5 (c) (1) .
37/ This chart is reproduced in Appendix B to this brief, p. B - 2 .
40
and (1 0), and in another series of columns, the results which
would have been needed to reach various levels of confidence.
38/
(Columns (4), (5), (8), (9), (12)) In preparing this chart,
Dr. Hite in effect acknowledged that the .05 and .01 levels of
confidence are critical, because he selected these levels for
reporting on the chart.
Dr. Hite's own data contained in Gov. Ex. 41 demonstrate
that his results are not statistically significant under the appro
priate standards. The most important data on this chart were
derived by the discriminant analysis method, which Dr. Hite him-
39/self chose as his preferred method (6 Tr. 52-75, Co. Ex. 75, p. 1).
We reprint on the following page portions of the chart including,
in the last column, the approximate levels of confidence which
can be ascribed to Dr. Hite's discriminant analysis results.
38/ In his testimony, Dr. Hite sometimes referred to the 95%
confidence level and sometimes to the 5% or .05 level. These all
refer to the same thing. If a statistic is valid at the .05
level, we are 95% certain that it did not occur by coincidence
or chance, and there is only a 5% likelihood that the statistic
occurred because of chance.
39/ The chart in Gov. Ex. 41 also shows computations of statis
tical significance under two forms of correlation or regression
analysis. Since Dr. Hite rejected correlation analysis as being
inappropriate in this case (6 Tr. 52-75 ), those computations are
not discussed here. In any event, as is more fully explained
in Appendix B to this brief, at B-4 , there is no showing of
statistical significance in any of these other computations.
41
Excerpt from Gov. Ex. 41, Chart prepared by Dr. Hite
expert witness for Georgia Power Company
Job Discriminant Analysis
(3) (4) (5) Actual level ofStatistical
results from
Hite Study
Result needed Result needed
to reach .05 con- to reach .01
fidence level confidence
level
confidence obtained
Lineman 1.41 not reported not reported .25
Truck Operators
Mechanics, Electri-
4.47 3.13 4.92 .02 1
CM
cians, etc. 1.51 2.85 4.34 .25Clerical 1 .56 3.09 4.82 .25 '
Apprentice Mechanics 1.09 3.24 5.29 .40
Appliance Servicemen 0.87 2.83 4.29 .48
Storekeepers 0.77 2.95 4.57 . 50
Metermen 0.63 3.29 5 .42 .60
Helpers
Coal Equipment
1.02 2.82 4.26 .40
Operators 0 .18 3.29 5.42 .90Meter Readers 0.97 2.87 4.40 -.40Garage Mechanics
Switchboard Op-
3.46 2.98 4.64 .04
erators 1.93 2 .90 4.46 . 15
40/ The figures in this last column are derived from W .J. Dixon and F.J. Massey, J x .Introduction to Statistical Analysis, Table A-7c (2d ed. 1957),
Dr. Hite's explanation of this exhibit appears at 6 Tr. 95-97, 125-129.
The figures in the last column of this chart show that the
Hite study falls far short of accepted standards of statistical
reliability. Only for two jobs (truck operators and garage mechanics
of the thirteen listed did the discriminant analysis produce re-
41/
suits even approaching the .05 level of confidence. For the other
eleven jobs, the results relied on by Georgia Power are not even
42/
close to having statistical significance.
3) The serious technical defects of the Hite study
undermine the value of even the limited results obtained. We have
chosen to discuss the serious technical shortcomings of Georgia
Power's validity study in considerable detail in an Appendix to
this brief (Appendix B , infra), rather than in the text. We do
so because, as previously emphasized, it is our position that the
Hite study proves nothing, even if its results are taken at face
value as the product of a technically legitimate validity study.
For the sake of completeness, we briefly summarize these tech
nical defects here.
(i) The Hite study persistently shifts from one validation
43/technique to another in a futile quest for meaningful results.
41/ As to one of the two jobs for which the .05 level was reached
(garage mechanic) Dr. Hite himself testified that he would not re
commend use of the tests (6 Tr. 87-88). (Dr. Hite may have been
disturbed by the very heavy negative weighting of the PTI-N test
in his formula for this job).
42/ For example, as to the meterman job Dr. Hite's study produced
a statistical result of only 0.63, whereas a result of at least 3.29
was necessary to achieve significance at the .05 level, and 5.42 for
significance at the .01 level. Dr. Hite's result reaches only the
.60 level of confidence, meaning that there was a 60% probability -
it was more likely than not - that his results were based on
coincidence or chance.
43/ See Appendix B, pp.B-5- B-7 , infra.
- 43 -
The accuracy and credibility of data produced by this artificial
and self-serving process of technique-shopping is highly dubious.
(ii) A job performance rating used by Dr. Hite was based
on supervisory ratings containing a large element of built-in
bias. This "contamination" of raw data, coupled with the absence
of cross-validation of the "contaminated" results thereby ob
tained, and in particular the failure to derive separate validation
data for any distinct sample of black persons, further deprives
44/the study results of real meaning.
(iii) The record is replete with proof of inconsistencies
and racial disparities in the test administration process. The
study takes no account of any such possibility, and thereby fails
to assess the situation actually prevailing in the Georqia Power45/
Company.
The defects and inadequacies of Georgia Power's validity
study result from the Company's desperate attempt to prove its
tests’ worth by any means possible. Yet that effort falls short,
for the study itself utterly fails to evidence the validity re
quired to justify the continuation of this testing program with
its racially discriminatory impact. Therefore, following the rule
of Griggs v. Duke Power Company, supra, this Court should enjoin
usage of the tests and order relief for the victims of Georgia
Power's discriminatory testing.
44/ See Appendix B. pp. B-7-B-11, infra.
45/ See Appendix B, p. B-12, infra.
44
II. THE COURT BELOW ERRED IN FAILING TO FIND
THAT GEORGIA POWER COMPANY HAS CONTINUED
TO ASSIGN BLACK HIREES TO MENIAL JOBS ON
THE BASIS OF RACE, DESPITE PLAINTIFFS'
UNREBUTTED STATISTICAL EVIDENCE SHOWING
THESE PRACTICES
Georgia Power Company's record of post-July 29,
1963 assignment practices clearly shows that, whatever may have
been the consequences of the Company's change in policy allowing
laborers to seek transfer or promotion to better jobs, that policy
change had no effect at all upon the existing practices of assign
ing all black hirees to menial positions. The district court
erred in failing to recognize these assignment practices as dis
criminatory and in failing to grant their black victims appro
priate relief.
In its finding of no discrimination with respect to job
assignments, the district court ignored this Court's clear directive
that an inference of discrimination should be drawn under the cir
cumstances prevailing here.
When there is a finding of a pre-Act pattern
of discrimination, and little or no evidence
indicates a post-Act change in such pattern
or practice up to the time suit is filed, a
strong inference that the pre-Act pattern
or practice continued after the effectiveness
of the Act arises.
United States v. West Peachtree Tenth Corp., 437 F.2d 221, 227
(5th Cir. 1971).
The district court's denial of relief as to this issue was
based on legal standards which, we submit, were erroneously con
ceived and applied. The court found no discrimination in assign
ment because plaintiffs did not show the availability of specific
45
high level job openings for the 28 admittedly qualified blacks who
were, after July 29, 1963, assigned as laborers (Opinion at 44). In
adopting this reasoning, the court unjustifiably refused to draw
the plain inference of the statistics showing that all fully
qualified blacks hired before June 9, 1969 were assigned to laborer
jobs whereas over 80% of all the whites (including many of the un
qualified whites) were made helpers. That inference is that the
Company continued its pre-1963 policy of job assignments on a
racial basis. See e.g., Bing v. Roadway Express, Inc., 444 F.2d
687, 689 (5th Cir. 1971) (where this Court held that statistical
evidence made out a prima-facie case of discrimination); United
States V. Hayes International Corp., 415 F.2d 1038, 1043 (5th Cir.
1969); United States v. Sheet Metal Workers, Local 36, 416 F.2d 123,
127 n.7 (8th Cir. 1969); Jones v. Lee Way Motor Freight Co.. 431
F . 2d 245, 247 (10th Cir. 1970), cert, denied, 4 m U.S. 954 (1971).
The court found no specific evidence to contradict this inference,
and there is no such evidence in the record. Where the strong
inference of statistics showing a pattern suggesting discrimination
is not rebutted by specific facts to the contrary, the court was
not warranted in disregarding the inference of discrimination.
Bing v. Roadway Express, Inc., supra. United States v. Jacksonville
Terminal Company, F.2d ,3 EPD f8324 (5th Cir. 1971) at pp.
46/
137-142.
46/ This Court in Bing wrote, "Once Binq, by establishing a prima ■Facie case of discrimination, had carried his burden of proof, it was
incumbent upon Roadway to come forward and refute his case with
something more than mere conclusional statements that it never dis
criminated against Negroes in hiring road drivers," 444 F.2d at 689.
And in finding no assignment discrimination in the Jacksonville
Terminal case, this Court clearly relied on the Company's specific
affirmative proof of non-discrimination, which is totally lacking here.
46
Instead of requiring affirmative proof by the Company, the
court below held plaintiffs to the burden of specific proof des
pite their unrebutted statistical evidence. It found decisive
the fact that there was "no showing that there were 28 openings
at the time or that there were any openings at all into which they
could have immediately been hired, at any other job" (Opinion at 44).
We submit that the absence of proof of specific vacancies is not
fatal to plaintiffs' case here. Indeed, the record shows that at
least 600 non-laborer vacancies were filled by whites in five years
from January 2, 1965 to January 16, 1971 (cf. Opinion at 15) --
an average of approximately two per week. In requiring more in
these circumstances, the district court "saddled the United States
[and private plaintiffs] with an impossible burden of proof,"
just as did the district court which was reversed in United States47/
v. Jacksonville Terminal Co.., supra at 138-139. The decision of
the court below implies that the assignment of all blacks hired for
nearly six years after July 29, 1963 to laborer jobs is not proof
enough, because some of them might have been assigned as laborers
even without discrimination and because the plaintiffs had not
demonstrated specifically which ones were so assigned as a result
of discrimination. The district court's standard is erroneous.
As the Second Circuit has held in an analogous situation,
47/ That district court had required a comparative evaluation
of blacks and whites seeking each job - a task comparable we
submit, to cataloguing each of the hundreds of vacancies and
analyzing various applications for them.
47
It is true that some of the black employees
might have been assigned there even under the
best of systems. But there is no apparent
way of knowing that, or determining now who they would be, and appellees offer none.
The discrimination found illegal here was to
a group; group remedy is therefore appropriate .
United States v. Bethlehem Steel Corp.. 446 F.2d 652, 660 (1971).
Here too, the burden of specific proof should have been shifted
to defendants by the plaintiffs' prima-facie statistical case;
and in the absence of any such specific defense the finding of
no discrimination was error.
III. GEORGIA POWER COMPANY VIOLATES TITLE VII
BY PLACING PRIMARY RELIANCE ON WORD-OF-
MOUTH RECRUITMENT AND WALK-IN APPLICATIONS
TO FILL JOB VACANCIES, IN LIGHT OF ITS SUBSTANTIALLY SEGREGATED WORK FORCE.
The present, continuing racial stratification of its
entire workforce along pre-1963 lines clearly demonstrates that
the effect of Georgia Power's word-of-mouth recruitment practices
has been to perpetuate the racial status quo. Nevertheless, the
district court blithely concluded,
However, the court sees nothing sinister in this respect [recruitment]. Advertisement
of existing vacancies by word-of-mouth on
the part of company employees has been the
best means for recruitment, and often for
promotion, from time immemorial. It has been
done by both blacks and whites in this com
pany. Even if it were harmful, there is no
known way to halt it. The mere fact that a
company's personnel has a majority of one race
or another does not make such practice charge
able to the employer as a discriminatory act.Only if some preference were given the white
48
referrals could the question rise to the
point of decision. Nor is there any signi
ficance in the places of conducting inter
views for possible management personnel. In
short, the court finds no evidence substantia-
ing a racially motivated policy of recruiting on the part of the Company.
(Opinion at 46). We submit that as the foregoing passage shows,
the district court applied an erroneous standard and consequently
arrived at a clearly erroneous result.
Because its work force remains segregated, Georgia Power's
reliance on word-of-mouth advertising and exclusively walk-in
hiring leads inevitably to the perpetuation of existing racial
segregation. White employees tell their friends and relatives
of openings in the higher-level jobs which they hold in over
whelming proportions; black employees are able to pass on word
only of openings in the lower-level jobs to which they have been
relegated. As one commentator has explained this phenomenon,
"[t]he segregated social and residential patterns in the nation
make it unlikely that minorities will be included in the 'web of
information' which flows around opportunities in the white society."
Blumrosen, The Duty of Fair Recruitment Under the Civil Rights Act
of 1964, 22 Rutgers L. Rev. 465, 477 (1968) (hereafter cited48/as Blumrosen).
48/ Reliance on word-of-mouth recruitment implies acceptanceof walk-in application as the means of securing applicants for
employment. If an employer does not engage in affirmative recruit
ment efforts, potential applicants can learn about employment
opportunities only from incumbent employees or from the employer1s
general reputation in the community. Where the existing work force
is substantially segregated, these forms of communications can have
49
We submit that Title VII's proscription of a "failure" to
hire individuals because of their race or color, 42 U.S.C. §2000e-2
(a)(1 ), must be here construed as covering the failure to en
gage in recruitment activities which will assure potential black
employees equal employment opportunities. Blumrosen at 472-473;
Note, Employment Discrimination and Title VII of the Civil Rights
Act of 1964, supra, at 1153. We further submit that recruitment
practices which result in the maintenance of a substantially
all-white work force, in all or a range of an employer's jobs --
particularly where the better jobs are largely held by whites
and the lower-paying jobs are occupied by a disproportionately
F.N. 48 (Cont'd)
a devastating impact on racial minorities. Cf. Blumrosen at 478; Note, Employment Discrimination and Title VII of the Civil Rights
Act of 1964, 84 Harv. L. Rev. 1109, 1153-54 (1971),
See also the following OFCC regulation implementing Executive Order No. 11,246 (3 C.F.R. 339 (1967));
Contractors should employ procedures for
handling of applicants that are compatible with equal employment opportunity. In this regard,
employee referrals or "walk-ins" require par
ticular attention, since these recruitment
procedures are most susceptible to dis
crimination. Employees tend to refer friends
from their own racial and ethnic backgrounds
and, therefore, if few minority group workers
are already employed, few minority group
workers are likely to be referred. With re
gard to "walk-in", the entire recruitment
program is strongly dependent upon the attitudes of guards and receptionists and can thus be
easily frustrated. 41 C.F.R. Section
5-12.805-51 (b) (5) (1970) .
At least one Court of Appeals has explicitly adopted the rationale
of this regulation. Parham v. Southwestern Bell Telephone Co.,433 F . 2d 421, 429 (8th Cir. 1970).
50
large percentage of blacks — violate Title VII's prohibition of
practices that "limit, segregate, or classify employees" in a
way which deprives or tends to deprive potential black employees
of employment opportunities because of their race or color,
42 U.S.C. §2000e-2 (a) (2) . Cf. Blumrosen at 474-476.
Our position is in perfect accord with the Supreme Court's
holding with respect to employee selection procedures handed down
in Griggs v. Duke Power Company, supra, that a selection practice
which is "neutral on its face" may nevertheless contravene
§2000e-2 (a) if it has a discriminatory impact on minority group
employment opportunities. See also, Local 53, International
Ass'n. of Heat & Frost Insulators & Asbestos Workers v. Vogler.
407 F.2d 1047, 1054 (5th Cir. 1969). Under Griggs, once such an
adverse impact has been established, the crucial inquiry is whether
the employer can prove that the practice is required by business
necessity. 401 U.S. at 432. We seriously doubt that Georgia Power
could ever show that its exclusion, through recruitment practices,
of a large class of potential applicants was necessary to its
selection of a superior work force; at any rate, the Company made
no attempt to show any business necessity for these practices on
this record. Since these procedures have perpetuated the exclusion
of blacks from the better jobs, they thereby violate the Act.
This position received the approval of several careful
courts, on closely similar facts, even before Griggs. In Clark
v. American Marine Corp., 304 F.Supp. 603 (E.D.La. 1969), the
court found that, since the employer did not advertise job vacancies
51
or make any other public announcements of openings, u[t]he word-
of-mouth message that vacancies exist in better paying iobs
therefore usually goes only to whites." 304 F.Supp. at 606
(emphasis added). In effect, the court held that, although the
word-of-mouth system apprised blacks of openings in lower level
positions, its failure to inform them of vacancies in better jobs
constituted a violation of §2000e-2 (a) (1) and (2). In Parham v.
Southwestern Bell Telephone Co., 433 F.2d 421 (8th Cir. 1970),
the company employed relatively few black employees, who were
largely concentrated in the lower job categories. 433 F.2d at
426. The court held the employer's primary reliance on word-
49/
of-mouth recruitment an unlawful employment practice. Similarly,
the courts have recognized that a reputation for discriminatory
employment practices, spread by informal communication among the
members of the black community, has a discriminatory impact on
potential minority applicants which renders the failure to dispel
that reputation an unlawful employment practice. Lea v. Cone
Mills Corp., 301 F.Supp. 97, 102 (M.D.N.C. 1969), aff'd in per
tinent part, 438 F.2d 86 (4th Cir. 1971); United States v. Plumbers
Local 73, 314 F.Supp. 160 (S.D. Ind. 1969); United States v. Sheet
Metal Workers Local 36, 416 F.2d 123, 139 (8th Cir. 1969).
49/ In Parham, the Eighth Circuit, quoting the observation of
the trial judge that "employment in the ‘all white' categories,
would continue to be all white, and employment in the predomin
antly Negro lower categories would continue to be predominantly
Negro, held "as a matter of law that the Company's system of recruiting new workers . . . operated to discriminate against
blacks." 433 F>2d at 427.
52
The district court erred in adopting, for this issue, an
apparently subjective standard which was successfully met when the
court found no evidence of "a racially motivated policy of recruit
ing on the part of the Company " (Opinion at 46). These con
siderations do not dispose of a charge of racial discrimination,
since "good intent or absence of discriminatory intent does not
redeem employment procedures . . . that operate as 'built-in head
winds' for minority groups and are unrelated to measuring job
capacity," Griggs v. Duke Power Company, supra at 431. Accord.
Local 189, United Papermakers and Paperworkers v. United States.
416 F .2d 980, 996 (5th Cir. 1969), cert. denied 397 U.S. 919 (1970).
The crux of the district court's legal misconception is its belief
that Georgia Power had no affirmative duty to regulate or design
its racially neutral employee selection procedures in such a way
as to assure blacks equal access to job opportunities, as a
50/
practical matter and in fact. To read Title VII as so limited
would be to strip its broad provisions of real effect, and for a
court to sustain such a reading would be tantamount to merely
parroting the Act's prohibitions," which this Court disapproved
in Local 53, etc, v. Voqler, supra, at 1052.
Fortunately, the courts need not be content with such asserted!
neutral practices as Georgia Power's. On the contrary, this Court
50/ The court stated that "[o]nly if some preference were given
the white referrals could the question rise to the point of de
cision " (Opinion at 46). The decision in such a hypothetical
case would presumably be obvious. Yet in practical terms, the
system approved below has virtually the same result as a system of preferences for white referrals.
53
has recognized an affirmative judicial duty to devise solutions
that are effective as a practical matter, which may include affirm
ative steps. See, e.g., Local 53, of Int11. Ass'n of Heat &
Frost Insulators & Asbestos workers v. Vogler, 294 F.Supp. 368
(E.D. La. 1968), aff'd 407 F.2d 1047 (5th Cir. 1969); Bing v.
Roadway Express, Inc., supra, at 690; see also, United States v.
Sheet Metal Workers, Local 36, supra,at 139-140. These courts
have recognized that affirmative steps toward non-discriminatory
recruitment practices can and must be taken pursuant to 42 U.S.C.
§2000e-5 (g), providing for "such affirmative action as may be
appropriate" to remedy discrimination, so long as such steps
51/
are feasible.
This record and the annals of employment law are utterly
devoid of support for the district court's flippant conclusion
that "[a]dvertisement of existing vacancies by word-of-mouth on
the part of the Company employees has been the best means for
recruitment . . . from time immemorial " (emphasis supplied)
(Opinion at 46). Indeed, we would assert that, far from being * 3
51/ Although, as the court below feared, it may be impossible
(or undesirable) to stop present employees from telling their friends
and relatives about job openings, this does not dispos e of the issue.
Affirmative actions to counterbalance unpreventable employee
behavior are appropriate, where necessary. A number of courts
have ordered employers to adopt such affirmative recruitment programs to encourage minority hiring. See, e.g., Morrcur v. Crisler,
___ F.Supp. ___, 4 EPD 17541 (S.D. Miss. 1971); United States v .
Plumbers Local 73, supra; United States v. Ironworkers Local 392,
3 EPD 18063 (E.D. 111. 1970) (consent decree); United States v. Virginia Electric & Power Co., 4 EPD 17502 (E.D. Va. 1971)
(consent decree).
54
the best possible method of recruitment in an overwhelmingly-
segregated Company, word-of-mouth and walk-in methods fall below
the minimum standard of lawfulness under Title VII.
For the same reasons, Georgia Power's practice of re
cruiting only at white schools for management and skilled tech
nical and clerical positions violates the Act. By recruiting
at white institutions, while ignoring predominantly black insti
tutions, Georgia Power "fail[s] . . . to hire . . . individual[s],
or otherwise discriminate[s] against . . . individual[s] with
respect to [their] privileges of employment, because of [their]
race [or] color." Cf. 42 U.S.C. §2000e-2 (a) (1) . The discrim
inatory effect of such a practice is clear, and there can be no
showing of such "business necessity" as would justify it: since
the company is capable of maintaining recruitment programs at
white institutions throughout the South, there is no reason to
believe that it cannot maintain such programs at black schools
in the same area. Therefore, applying the test established by
the Supreme Court in Griggs v. Duke Power Co., supra, the practice
violates Title VII as a matter of law. Cf. United States v.
Ironworkers Local 8 6, 315 F.Supp. 1201 (W.D. Wash. 1970), aff'd
52/433 F .2d 544 (9th Cir. 1970), cert, denied 4 EPD 17526 (1971).
52/ The Ironworkers case held that "[i]t is unlawful for a
union or apprenticeship committee actively to attempt to recruit
whites while making no effort to recruit blacks."315 F.Supp. at 1236. See also, Local 53, etc, v. Vogler, supra; Lea v. Cone
Mills, supra.
55
IV. THE COURT BELOW ERRED IN FAILING TO GRANT
FULL AND EFFECTIVE SENIORITY RELIEF, OR IN
SOME INSTANCES ANY RELIEF, TO BLACKS WHOSE
SENIORITY RIGHTS AND STATUS WERE ADVERSELY
AFFECTED BY DEFENDANTS' DISCRIMINATION.
A. The Relief Granted By The District Court
Pertaining To Seniority Was Inadequate
And Ineffectual In Many Respects.
1) As to pre-July 1963 hirees.
Almost as soon as Georgia Power purportedly took the lid off
the trap of menial labor where its black employees were ensnared,
it put the lid back on, in the form of new testing and educational
requirements. No such requirements were or had been imposed on in
cumbent whites in non-menial jobs (Opinion at 13). Largely because
of these obstacles, blacks hired before July 29, 1963 remained in
menial positions, while contemporary whites of no greater quali
fications continued to advance into positions of increasing
responsibility and pay.
The Court below recognized under these circumstances that
pre-July 1963 black employees were thus suffering substantial
seniority disadvantages, the consequences of which have continued
right down to the present time. The Court recognized the pertinence
of the long line of precedents holding the continuation into the
present of the consequences of such past discrimination to be a
violation of Title VII (Opinion at 42—43), and ordered certain
measures that it felt would neutralize such discrimination (US decree
at 3a-4). As a practical matter and as a matter of law, this re
lief was inadequate. In certain of its provisions, the court's
remedy fell far short of according pre-1963 black employees effective
relief, and in some ways even tends to perpetuate discrimination.
56
The decree of the court below designates three "affected
classes", two of which are pertinent here: Class I, including all
incumbent black employees hired before August 19, 1963; and Class
II, including all incumbent black employees hired prior to July 29,
1963, who had a high school education and who had passed the
Company's aptitude tests (US decree at 2). Paragraph III of the
decree enjoins defendants, with one unimportant exception, from
maintaining a promotional seniority system based upon the amount
of time spent in the jobs immediately below the job bid for in a
line of progression. Instead the court ordered that promotions of
employees in these classes should be awarded based on the entire
time spent with the Company - in other words, required the sub
stitution of "company" or "employment" seniority for "job" seniority
(US decree at 3a).
Before proceeding to specify those ways in which we feel
that the remedy provided by the Court below with respect to
seniority was inadequate, we would note the need for clarification
of certain aspects of the district court's remedy. In some re
spects, the decree is not absolutely clear, and in others it appears
somewhat inconsistent with the conclusions of law in the court's
opinion of June 30, 1971. We assume that the decree governs,
rather than the opinion, but it would seem appropriate that this
Court clarify those points that remain uncertain.
For example, we assume that the provisions as to "company"
seniority for pre-July 29, 1963 employees, as set forth in Para
graph III of the decree, are determinative, and that the reference
57
in the opinion (at 43) that only time spent prior to July 29, 1963
5_3/must be credited toward a future promotion, has been superseded.
We also assume that the provisions of Paragraph III of the
decree apply with respect to each bid for promotion which may be
made by a member of one of the affected classes — not only to his
first advancement. Otherwise, for example, laborers who are now
in the bottom rung of various lines of progression could use their
newly acquired company seniority only to compete with other laborers
(virtually all of them also black) to move up one notch — as con
trasted with white employees originally hired at the same time or
later who hold positions many levels higher. The decree must,
therefore, mean that as openings develop from time to time that
black employees may use their company seniority each time they bid for
a promotion. In this way, they will gradually be able to rise to
whatever level their talents entitle them — rather than merely
to the second rung from the bottom of the ladder - and thus grad
ually reduce the gap between their status and that of those white
employees who hold higher positions only because of the earlier
54/
discrimination.
53/ Alternatively, it may conceivably have been intended to apply
only to the narrow situation of employees who had a high school
education and had successfully completed the aptitude tests. In
either case, clarification would seem desirable.
54/ Efficiency is protected by the requirement that the employee
be qualified to perform the job to which he is promoted (Gov.
decree at 9-10). Moreover, these procedures do not call for
removal of incumbent whites from jobs they now hold. We have no
quarrel with either of these aspects of the decree.
58
The decree in the private class actions incorporates the
company seniority provisions of the decree in the government case
(Private decree at 2). Again, we are assuming that this must be
taken literally despite the inconsistent reference in the earlier
conclusions of law of the court below (Opinion at 60-61), that the
five named plaintiffs hired prior to 1963 should have added to their
"job seniority" only the time going back to the latest of three
dates: that of the first bid for transfer made by each plaintiff;
the effective date of the Act; and 90 days before the filing of
the charge by that plaintiff with the Equal Employment Oppor
tunity Commission. Since the company seniority provided by the
decree will apparently be determinative in competition for vacancies
at higher levels, we shall assume that, if there is continued force
to the change in job seniority set forth in the conclusions of
law, they shall be in addition to (rather than instead of) the
rights to company seniority provided by the decree.
The relief granted by the Court below to black employees
hired before July 29, 1963 is deficient, even as we have con
strued th&t relief. The district court's injunction, as a practi
cal matter, provides effective relief only for those black laborers
who hold jobs outside the coal handling section in the Production
Department. The reason is found in Paragraph III(C) of the decree,
which states: "The job classification to or from which affected
employees are promoted, demoted, transferred or laid off shall be
in accordance with the lines of progression as set forth in the
Defendants' collective bargaining agreement." (US decree at 5).
59
This provision apparently means that company seniority may be ex
ercised only in the section in which the member of the affected
class was already working at the time the decree was entered.
1. The majority of the Company's black employees are
assigned to the coal handling section (Gov. Ex. 7), the top job
of which pays less than the top job of almost any other line of
progression (Gov. Ex. 7, Co. Ex. 7). Therefore, most of the
Company's black employees will under the terms of the decree be
competing against other blacks for the few jobs within that sec
tion, and will continue to be locked out, as they always have
been, from the more lucrative and desirable jobs inside the pro
duction plants, which are in different lines of progression. The
decree should be amended to allow blacks in the coal handling
section to exercise their company seniority at least once in trans
ferring to another section.
2. Furthermore, Paragraph III (C) completely cuts off
any relief to janitors, porters, and maids, all of whom are black,
even though their grievances would seem to be identical to those
of black laborers. This anomaly results from the fact that there
are no other jobs in the lines of progression of janitors, porters,
and maids into which these employees could progress with their
newly awarded company seniority. The decree should be clarified
to allow janitors, porters, and maids to transfer to other lines
of progression, taking their company seniority with them.
3. Class II employees (incumbent black employees hired
before July 29, 1963, who had passed the company's aptitude tests
and were high school graduates) who had been relegated to laborers'
60
jobs were not accorded the same relief granted other pre-1963
victims of racial discrimination. Instead of according them company
seniority, to enable them to bid for promotion against white
employees on the basis of total time spent with the company, the
court below refused to relieve them from the burden of competing
on the basis of job seniority, with the slight benefit of being
allowed to add to the time spent in their present jobs a period
equal to that between their first employment and July 29, 1963.
(Gov. decree at 3a) The theory of the court below was apparently
that from July 29, 1963 onward, these black employees have been
free from discrimination and could have bid for promotion or
transfer on equal terms fconcededly "onerous", however) with white
laborers. As pointed out above, however, discrimination did not
end on that date, and black employees were, until very recently,
and even now with very few exceptions, left in menial jobs, regard
less of whether or not they were qualified to do more responsible
work, even on the basis of the yardsticks applied by the company.
To assume that they were free to move up is to disregard the
entire pattern of company practice.
2) As to post-July, 1963 hirees.
The court below found only a limited degree of discrimination
in Georgia Power's operations after July 29, 1963. It granted
correspondingly little relief to post-July, 1963 hirees. The
court held only the requirement of high school graduation to be
61
unlawful, and in its decree purported to afford some relief to
only those post-1963 hirees who lacked a high school diploma or
its equivalent and "who have been denied job promotions or trans
fers because of this fact." (Affected Class III, US decree at 2).
The court accorded employees (white or black) in this class the
right to use "company seniority" instead of "job seniority",
relief similar to that granted to pre-1963 black employees.5-5̂
All other blacks hired after July 29, 1963 received little or
nothing from the district court. in denying seniority relief to
this class, the court erred in a number of areas.
C) . As demonstrated above, the testing requirements imposed
by the company are unlawful under Griggs v. Duke Power Co., supra.
If, as we feel confident, this court agrees, black employees
whose opportunities for advancement were blocked by failure to
pass the tests should be allowed to compete for promotions on
the basis of their company seniority.
To a great extent, what the district court did in the instant
Even this relief — the only adjustment of seniority ordered by the court below for blacks hired after July 29, 1963 -- was a
hollow remedy. Those lacking high school diplomas were in most
instances deterred by that fact from engaging in the futile ges
ture of applying for jobs which the company had said they could
not get. Moreover, most of them were also barred by the com
pany's testing requirements. And since the tests would not
normally be given to those who, even if they passed, would still
be disqualified by the educational requirement, there were
probably no black employees at all who fall within the court’s definition of Class III.
62
case was similar to what the Court of Appeals for the Fourth
Circuit had done in Griggs v. Duke Power Company, 420 F.2d
1225 (4th Cir. 1970). Both courts invalidated the test re
quirement for black employees who had been hired in menial jobs
at a time when white employees were employed in higher positions
without meeting test requirements. Both courts allowed such
blacks to compete for promotions on the basis of company seniority,
but refused to h dLd the tests invalid or to allow company seniority
as to black employees hired at a time when the tests were imposed
upon all employees. What we seek in this Court is the counterpart
of the Supreme Court's action in Griggs — a decision holding the
tests invalid and (although this is not specifically adverted to
in the Supreme Court's opinion) in effect to provide that the
later group of black employees who had been denied equal oppor
tunities because of the tests be also allowed to compete for pro
motions on the basis of their time with the Company. Any lesser
seniority relief would result in the continuation, for the rest
of their lives, of the competitive disadvantage suffered by blacks
held in the towest jobs through application of the unlawful test
ing requirements. If this remedy is not provided, although blacks
might no longer be blocked by the tests they would be forced to
compete for every promotion on the basis of "job seniority", and
would therefore continue forever to suffer a competitive disad
vantage vis-a-vis white employees junior to them in time with
the Company but who had had more time in the job immediately
below the one for which they were competing. CJf. Local 189,
63
United Paperworkers v. United States, 416 F. 2d 980, 988-991
(5th Cir. 1969), United States v. Bethlehem Steel Corp., 446
F.2d 652, 658 (2nd Cir. 1971).
2. Even where the court below did hold the teste unlawful
and enjoined their use for the jobs of switchboard operator,
garage mechanic, and coal equipment operator (Opinion at 34,
U.S. decree at 3), it failed to provide affirmative relief for
those victimized by their use in these respects. All such black
persons should receive the company seniority necessary to enable
them to compete for promotion on equal terms.
3. The district court also found that many black applicants
had, after July 2, 1965, been discriminatorily refused employment
when jobs existed (Opinion at 25). It identified specifically
at least 20 such discriminatees (U.S. decree at 5). While pro
viding for a procedure inviting these discriminatees to re-apply,
the court made no provision for them to assume their rightful
seniority standing upon hiring (US decree at 7-8). This Court
should order that these and other identifiable discriminatees
receive company seniority from the date of their first (post-
July 2, 1965) discriminatorily rejected application.
The relief we seek for this group is proper. This Court
has recently reaffirmed the broad discretion accorded the federal
judiciary under Title VII to fashion remedies and adjust relative
seniority rights of various groups of employees. Vogler v.
McCarty, Inc., __ F. 2d __, 4 EPD f7581 (5th Cir. 1971). Nor is
64
there any barrier to relief for those who had applied but never
become employees. See United States v. Bethlehem Steel Corp.,
supra, at 659 n. 11; Local 53, International Ass'n of Heat &
Frost Insulators v. Vogler, supra; United States v. Sheet Metal
Workers Internat'l Ass'n, supra.
B. This Court Should Correct the Deficiencies
of the District Court's Seniority Remedy in
Keeping With Its Duty to Provide Full and Effective Affirmative Relief From Racial
Discrimination.
This Court has repeatedly voiced its determination to pro
vide a remedy for the continuing effects of discriminatory systems
of seniority. Local 189, united Papermakers and Paperworkers v.
United States, supra, united States v. Hayes International Corp.,
415 F. 2d 1038 (5th Cir. 1969), Local 53, International Ass'n of
Heat & Frost Insulators & Asbestos Workers v. Vogler, supra, Bing
v. Roadway Express, Inc., supra, united States v. Jacksonville
Terminal Co., supra. The same principles which these cases apply
to companies' seniority systems to determine their unlawfulness
under Title VII are also applicable to review of remedial seniority
arrangements ordered by a district court. United States v.
Bethlehem Steel Corp., supra. In both cases, the seniority system
must be changed insofar as it carries forward the effects of
past discrimination, to the extent consistent with considerations
of safety and efficiency in operations and the limitations of
the "rightful place" theory. Local 189, etc, v. united States,
65
supra, at 988, 994.
The seniority arrangements provided in the district court's
decree fall far short of meeting these standards, in the many
ways specified above. Rather than presenting a separate legal
argument in support of each aspect of seniority relief which we
have sought, we are content to let the facts as set forth above
speak eloquently for themselves, and we here conclude more
generally.
To summarize the shortcomings of the decision of the court
below as to seniority rights, part, but only a small part, of the
relief necessary to correct illegal practices was provided. The
pattern of discrimination at Georgia Power -- after 1963 as well
as before -- is so pervasive, so widespread, that it would be
virtually impossible to isolate any appreciable number of black
employees and say that in their cases they have not suffered from
racial discrimination. The only remedy that can possibly be
adequate will be to allow all black employees to compete for
promotion on the basis of company seniority. As stated in the
assignment practices context in United States v. Bethlehem Steel
Corp., supra, at 660:
it is true that some of the black employees might have
been assigned there even under the best of systems.
But there is no apparent way of knowing that, or
determining now who they would be, and appellees
offer none. The discrimination found illegal here
was to a group; group remedy is therefore appropriate.
Cf. United States v. Board of Education, 372 F. 2d
836, 866 (5th Cir. 1966), aff'd on rehearing en banc,
380 F. 2d 285, cert, denied, 389 U. S. 840 (1967).
Just as in Bethlehem Steel, it is nearly impossible to determine
66
just which blacks missed out for which reason, and just as in that
case, relief should be accorded here to blacks as a class.
This Court has emphasized that courts of equity should be
creative in drafting decrees adequate to provide the remedies
needed. Local 53, supra, at 1052-53. See also Louisiana v.
United States, 380 U. S. 145, 154 (1965):
The court has not merely the power but the duty to
render a decree which will as far as possible
eliminate the discriminatory effects of the past
as well as bar like discrimination in the future.
We submit that complete relief of the nature described in
Louisiana v. United States should correct each of the deficiencies
in the seniority aspects of the district court's decrees, as set
forth above. As stated by this Court: "The ethic which permeates
the American dream is that a person may advance as far as his
talents and his merits will carry him." Miller v. International
Paper Co., 408 F. 2d 283, 294 (5th Cir. 1969). The common thread
in our seniority argument may be found in the failure of the
court below, in so many different respects, to approach that
guideline.
67
V. THE COURT BELCW ERRED IN FAILING TO
GRANT ADEQUATE TITLE VII RELIEF IN
THE NATURE OF BACK PAY
A. The District Court Improperly Limited
The Amount Of Back Pay It Awarded To
The Named Plaintiffs In Allowing Them
Far Less Than Their Actual Loss Due To
Defendants' Discrimination.
Ifi/The court below found five of the individual plaintiffs
entitled to affirmative remedial relief under its decision. It
57/directed that these plaintiffs be awarded job seniority and
back pay from an "effective date" defined as the later of two
dates: the date of the plaintiff's first bid for a job other than
laborer, or the date 90 days prior to the filing of plaintiff's
EEOC charge (Opinion at 60-61). In the decree implementing its
decision, the court found the various "effective dates" determined
in this manner to range from January, 1966 to February, 1969, and
awarded job seniority and back pay accordingly (Private decree at
3-4) .
56/ Paul Brown, Ed Dulaney, Sammie L. Davenport, Charles King,
and Willie C. Moreman. All were pre-1963 hirees, and thus eligible
for non-laborer jobs under the court's decree despite not having passed the tests.
The other named plaintiff, Rufus Mitchell, was a 1965 hiree
who was not test-qualified. He therefore received no relief from
the court's decree. If this Court finds the Company's testing re
quirement unlawful, Rufus Mitchell would be entitled to back pay
and seniority relief.
57/ The "job seniority" in question appears to be of no impor
tance in this appeal, for the reasons set forth at p. 5 9,
supra.
68
This back pay award to the five plaintiffs was legally
5E/inadequate. The court below erred, as a matter of law, in applica
tion of the "effective date" (as defined therein) for affirmative
relief from effects of past discrimination, and in limiting the
applicable rate of pay used for computing the back pay award to
that of the lowest-level promoted job.
All these plaintiffs should be awarded back pay from July 2,
5271965, regardless of whether they had bid for promotion or filed
EEOC charges as of that date. The court below found that these
plaintiffs and all other blacks hired before July 29, 1963 were
assigned to laborer jobs solely because of their race (Opinion at
12, 42). At the same time, the great majority of white employees
were hired into jobs higher than laborer, and those few whites who
were hired as laborers quickly achieved promotion (Opinion at 12,
42-43).
Thus, it is clear that if these five plaintiffs had not been
black, or if Georgia Power had not discriminated against blacks,
they would have held non-laborer jobs at all times since July 2,
6 0/1965. In these circumstances plaintiffs must receive back pay
58/ The job-seniority awarded to these plaintiffs was also in
adequate, for the same reasons. However, the issue is moot here
since the court's decree grants these plaintiffs adequate company
seniority. See pp. 5 7t 59 supra.
59/ The effective date of Title VII is the only appropriate cut
off date for back pay awards, despite dicta in the district court's
decision regarding a shorter "statute of limitations," cf. Opinion
at 58-59. See discussion at pp. 83-91, infra.
60/ Any lack of educational or test qualifications could not
consistent with Title VII and the judgment of the district court, have prevented them from obtaining a higher job, since all five were pre-1960 hirees.
69
for the jobs that they were prevented from holding on the basis of
race over this entire period, to make them whole from the direct
injury inflicted on them by defendants' discrimination. See, e.g.,
Johnson v. Georgia Highway Express Co., 417 F.2d 1122,1125 (5th
Cir. 1969); Bowe v. Colgate-Palmolive Company, 416 F.2d 711, 720
(7th Cir. 1969). The court erred as a matter of law in allowing
plaintiffs back pay only from the date of their first bid or EEOC
filing. White employees hired before July 29, 1963 had neither to
bid nor to file EEOC charges, in order to receive the superior pay
of non-laborer jobs.
The court below also erred in the amount it allowed, even
with respect to the limited periods for which it granted monetary
relief. The court allowed as back pay only the additional amount
plaintiffs would have earned in the job bid for — in every case,
the lowest jobs in a non-laborer line of progression (Opinion at
61, Private decree at 3-4). Yet in fact, absent discrimination,
these plaintiffs would not have remained in the entry-level job
61/of the line of progression. With their considerable seniority,
they could have successfully competed for higher-paying jobs.
Their back pay award should reflect this deprivation of advance
ment opportunities, as well as the denial of the first promotion
on which each had bid. This Court should reverse and remand for
determination of the correct amount of back pay, measured by what
plaintiffs would have earned since July 2, 1965, including second
or subsequent promotions based on their company seniority, absent
61/ For example, plaintiff Dulaney now has 31 years seniority,
plaintiff Davenport 16 years, and plaintiff Moreman 23 years.
70
defendants' discrimination.
B. The District Court Erred In Denying Back
Pay To Members Of Plaintiffs' Class, Even
Though The Court Found That Class Members
Had Suffered Severe Economic Loss Result
ing From Defendants' Discrimination.
The district court specifically found that members of plain
tiffs' class (including each of the designated subclasses) have
suffered economic injury as a consequence of defendants' policies
and practices. Nevertheless, the court ruled out an award of
back pay to all such non-party class members, including those it
found to have been detrimentally affected by race discrimination,
as a matter of law. We submit that the district court's limita
tion of back pay awards to named plaintiffs who filed EEOC charges
cuts against the entire thrust of Title VII remedial law, and
should be reversed by this Court.
The court below explicitly recognized the classwide economic
consequences of defendants' discrimination. It found that
The job classification of Laborer, Janitor,
Porter, Maid, Coal Sampler, and Switchman,
which blacks have traditionally held, pay
lower than the lowest paying white job. (Gov.
Exs. 1, 2). Since the majority of blacks
have held only laborer's jobs, and since
virtually all whites have held higher paying
jobs, it follows that earnings of black em
ployees , as â class, have been considerably
lower than earnings of their white contem
poraries , and lower than some later hired
whites.1/
71
(Opinion at 41) (emphasis added).
This devastating loss of opportunity for equal paying jobs
has injured class members hired both before and since July 29,
1963. With respect to the former group, the court properly con
cluded that
. . . all white employees hired prior
to July 29, 1963, solely because of their race, have freely worked in and accumulated
seniority in job classifications necessary
for promotion to higher-paying jobs while
prior to July 29, 1963, all black employees
were prevented solely because of race, from
accumulating such seniority. The present effect of that practice made those black em
ployees unsuccessful in bidding against white
employees for higher paying jobs.
(Opinion at 43) (emphasis added). With respect to individuals
hired after July 29, 1963, the court dramatically graphed the pay
disparities between whites and blacks hired after July 2, 1965 in
its Finding #24, set forth at p. 12 , supra. That finding shows
that as of September 6, 1968, 99% of the blacks in this category
earned under $400 per month, and not a single black earned over
$500 per month; whereas less than 3% of all whites in this cate
gory earned less than $400 per month, and over 65% earned over
$500 per month.Id.
Despite its unequivocal finding that in fact all members
of plaintiffs' class had been intentionally deprived of income
62/
62/ in footnote 3, the court cited and found as vivid factual
illustrations of this finding of pay-disparities three instances
in which later-hired whites earned up to $16,500 more than earlier-
hired blacks in a period of four and one half years (Opinion at
41 n.3).
Thus, it is clear that black class members' income losses
have been severe.
72
opportunities by defendants' practices, the court below refused
to award back pay to any black workers other than the named plain
tiffs who had filed EEOC charges. In summary, it held
In the court's view, such massive pay
ments and specific retroactive relief is
not contemplated in a suit by the Attorney
General under Section 707(a) of the Act
(42 U.S.C. §2000e-6), nor in a class action
generally. The scheme envisioned by the
Congress in Title VII relief was of two
types: individual and class relief; the
former to be brought by each aggrieved person after exhaustion of his administrative
procedures, the latter by the Attorney
General. . . . In individual cases, after
proper compliance with the administrative
prerequisites presumably fashioned to en
courage circulation [sic] and to insulate
the court from inundation by a mass of de
tail in one suit, the court is authorized
to give specific relief.
(Opinion at 55-56). The court below then carefully noted the dif
ference between the statutory remedies in private actions, includ
ing back pay, 42 U.S.C. §2000e-5(g), and the less specific remedies
provided by the statute for pattern - and - practice actions, 42
U.S.C. §2000e-6(a)(3) (Opinion at 57). Holding that, "primarily
for this reason, the court concludes that such relief is not author
ized and, in the discretion of the court, certainly not demanded,"
Id., and expressing its preference for "prospective remedies for
63/class-type relief," the court denied the class back pay. Id.
63/ The court below did not specify how this reasoning can sup
port a denial of back pay in the private class actions as distinct
from the consolidated §707 action. If anything, the court's reason ing would seem to support appellants' position here that such class
back pay is available in the present actions.
Nevertheless, the court's language, Opinion at 55 n.ll and
accompanying text, and its final decree, make it clear that the
court has finally denied private appellants' claim for back pay
on behalf of their class.
73
We submit that the district court's denial of back pay to
class members as a matter of law is contrary to the statute and to
applicable decisional law. Section 706(g) of Title VII, 42 U.S.C.
§2000e-5(g), specifically authorizes the district court to award
64/back pay as part of the remedy for a Title VII violation.
Although the language of Title VII makes the back pay award dis
cretionary, we submit that its denial in the circumstances of
this case was an improper exercise of that discretion, founded
on a plain misreading of the statutory powers granted to the
district court (see n .6 3, supra and accompanying text).
The narrow question of whether a district court erred in
denying class-wide back pay, where severe economic loss due to
discrimination has been proved, is of first impression in this
Court. The denial of back pay was erroneous, in our view, for
two principal reasons which have been most clearly enunciated by
the Seventh Circuit.
The first reason is that the statute contemplates full and
effective compensatory relief to victims of Title VII discrimina
tion, and that such relief necessarily includes back pay where
64 / Section 706(g) provides, in pertinent part:
If the court finds that the respondent
has intentionally engaged in or is enten-
tionally engaging in an unlawful employment
practice charged in the complaint, the court
may enjoin the respondent from engaging in
such unlawful employment practice, and order
such affirmative relief as may be appropriate,
which may include reinstatement or hiring of
employees, with or without back pay.
74
income loss has been shown. This Court indicated as much in
Johnson v. Georgia Highway Express, 417 F.2d 1122 (1969) where
it characterized the claim for back pay as "an integral part of
the statutory remedy." 417 F.2d at 1125. The same concept was
more fully developed by the Seventh Circuit in Bowe v. Colgate-
Palmolive Company, 416 F.2d 711 (1969), in reversing a district
court's refusal to award back pay to a class of persons who had
not filed individual EEOC changes.
The clear purpose of Title VII is to
bring an end to proscribed discrimina
tory practices and to make whole, in a
pecuniary fashion, those who have suffered
by it. To permit only injunctive relief
in the class action would frustrate the
implementation of the strong Congressional
purpose expressed in the Civil Rights Act
of 1964. To require that each employee
file a charge with the EEOC and then join
in the suit would have a deleterious effect
on the purpose of the Act and impose an
unnecessary hurdle to recovery for the
wrong inflicted.
416 F.2d at 720. Of the statutory authority to award back pay,
42 U.S.C. §2000e-5(g), that court said:
This grant of authority should be broadly
read and applied so as to effectively terminate the practice and make its
victims whole . . . The full remedial
powers of the court must be brought to bear and all appropriate relief given.
416 F.2d at 721.
The Fourth Circuit has endorsed the principles expressed
in Bowe in its recent landmark decision affirming a district
court's award of class-wide back pay, Robinson v. Lorillard Corp.,
444 F.2d 791 (1971). in that case, liability for class-wide back
75
pay resulted from a racially discriminatory seniority/transfer
system similar in form to that of Georgia Power Company, but far
less virulent in its details. The court noted that
This is a case in which final injunc
tive relief is appropriate and the defen
dants' liability for back pay is rooted
in grounds applicable to all members of
the defined class. Under these circum
stances the award of back pay, as one
element of the statutory remedy, con
flicts in no way with the limitations of Rule 23(b)(2).
444 F.2d at 802. The same analysis applies with equal force here,
and should inevitably require the back pay award. The Robinson
court also took the same view of the character of the back pay
award as that expressed in Bowe and Johnson: "The back pay award
is not punitive in nature, but equitable — intended to restore
the recipients to their rightful economic status absent the effects
of unlawful discrimination." Robinson v. Lorillard Corp., supra
at 804.
The second compelling reason to award back pay is that the
class back pay remedy is crucial to the effective implementation
of Title VII as a public policy against job discrimination. This
Court has forcefully declared its recognition of that policy:
Title VII of the 1964 Civil Rights Act
provides us with a clear mandate from
Congress that no longer will the United
States tolerate this form of discrimina
tion. It is therefore, the duty of the
courts to make sure that the Act works . . .
Culpepper v. Reynolds Metals Co., 421 F.2d 888, 891 (1970). The
Seventh Circuit has brought into focus the importance of Title VII
remedies, including back pay, as a means of enforcing the strong
76
national policy against discrimination recognized in Culpepper:
The vindication of the public interest
expressed by the Civil Rights Act con
stitutes an important facet of private
litigation under Title VII . . . and the
court has a special responsibility in the
public interest to devise remedies which
effectuate the policies of the Act as well
as afford private relief to the individual
employees instituting the complaint.
Sprogis v. United Air Lines, Inc., 444 F.2d 1194, 1201 (1971),
cert, denied ___ U.S. ___, 4 EPD 17588 (1971) . We submit that
the availability of class-wide back pay as part of a remedy for
across-the-board racial discrimination is crucial to the meaning
ful enforcement of Title VII. Back pay is therefore commanded by
65/adherence to national policy.
The district court's denial of class back pay was evidently
founded on its rejection of this Court's considered interpretation
of the statutory scheme with respect to Title VII class actions.
This Court has consistently favored class actions as an appropriate
eL/vehicle for presentation of Title VII claims for relief. The
court succinctly stated the heart of the matter in its land
mark holding that persons need not have filed EEOC charges in order
65/ Back pay has also been awarded to class members who did not
file EEOC charges in United States v. Wood, Wire, and Metal Lathers
Int. U.. Local 46. ___F.Supp. ___ , 3 EPD f8204 at 6664 (S.D. N.Y.
1971); and class back pay was approved in Local 186 v. Minnesota
Mining and Manufacturing Co., 304 F.Supp 1284, (N.D. Ind. 1969) and
Baxter v. Savannah Sugar Refining Corp., 46 F.R.D. 56 (S.D. Ga. 1969).
§6/ Jenkins v. United Gas Corp., 400 F.2d 28 (5th Cir. 1968);
Johnson v, Georgia Highway Express, 417 F.2d 1122 (5th Cir. 1969);
see also Griggs v. Duke Power Co., 401 U.S. 424 (1971).
77
to join as Title VII co-plaintiffs with other persons who have
filed such charges, where all co-plaintiffs are members of a pro
per class:
Racial discrimination is by definition
class discrimination, and to require
a multiplicity of separate, identical charges before the EEOC, filed against
the same employer, as a prerequisite to
relief through resort to the court would
tend to frustrate our system of justice and order.
Oatis v. Crown-Zellerbach Corporation, 398 F.2d 496, 499 (1968).
See also, Miller v. International Paper Co., supra, at 284-285.
In short, this Court has consistently ruled that each class mem
ber need not individually exhaust EEOC remedies and file suit, in
order to qualify for remedial relief under Title VII.
It is logically indefensible to rule, as the court below
did, that a class action is maintainable, and class relief is
available, for purposes of partial injunctive relief only, but not
for purposes of affirmative relief as part of the equitable remedy.
There appears to be a total lack of persuasive authority for
this proposition of the district court. On the contrary, the Bowe
court flatly commented, "We are also unable to perceive any justifi
cation for treating such a suit as a class action for injunctive
purposes, but not treat it so for purposes of other relief." 416
F.2d at 720. Apparently the court below was also unable to arti
culate any justification for its limiting ruling, since it relied
only on its own assertion of what Title VII remedies the Congress
"envisioned" for § 707 actions (Opinion at 56), and sidestepped
entirely the language of § 706(g).
78
Denial of affirmative relief is not only without rational
basis, but it squarely contradicts another basic principle govern
ing Title VII relief, best expressed by the Supreme Court in
Louisiana v. United States, supra, at 154:
The court has not merely the power but
the duty to render a decree which will
so far as possible eliminate the dis
criminatory effects of the past as well
as bar like discrimination in the future.
Acting pursuant to this principle, this Court has held in a Title
VII race discrimination case,
In formulating relief from such practices,
the courts are not limited to simply
parroting the Act's prohibitions but are
permitted, if not required, to "order such
affirmative action as may be appropriate."
Local 53, etc, v. Vogler; supra. at 1052. Any less-than-complete
relief from admitted race discrimination, such as that granted by
the district court, is therefore a breach of the equity court's
duty to eliminate, insofar as possible, all the effects of invidi
ous discrimination. Cf. Potts v. Flax, 313 F.2d 284, 289 (5th
Cir. 1963) .
The award of compensatory back pay to victims of unfair
labor practices under the National Labor Relations Act, 29 U.S.C.
67/§§ 151 et seq., a statute whose remedial provisions served as
67/ Section 10(c) of the NLRA, 29 U.S.C. § 160(c) provides the
statutory basis for such back pay awards in conjunction with rein
statement orders. That section directs the Board, upon finding
an unfair labor practice, to issue cease-and-desist orders "and
to take such affirmative action including reinstatement of employees
with or without back pay, as will effectuate the policies of this subchapter."
79
the model for those of Title VII provides an instructive analogy.
Such back pay awards have long been accepted as valid remedies, and
federal appellate courts have routinely enforced National Labor Re
lations Board back pay awards for reasons that are equally applic
able to Title VII back pay claims.
The Supreme Court in a recent opinion upholding the full
scope of an NLRB back pay order concisely summarized the applicable
NLRA principles:
The legitimacy of back pay as a remedy
for unlawful discharge or unlawful failure
to reinstate is beyond dispute, Mastro
Plastics Corp. v. N.L.R.B., 350 U.S. 270
(1956), and the purpose of the remedy is
clear. 'A back pay order is a reparation
order designed to vindicate the public
policy of the statute by makeing the
employees whole for losses suffered on
account of an unfair labor practice.'
Nathanson v. NLRB, 344 U.S. 25 (1952).
NLRB v. Rutter - Rex Manufacturing Co., 396 U.S. 258, 263 (1969)
(ordering up to 7^ years of back pay for 172 workers).
As the quotation from the Nathanson case indicates, NLRB back
68/
58 / The Clark-Case Interpretative Memorandum on Title VII, Con
gressional Record (Senate) April 8, 1964, pp. 7212-7215 clearly
states that Title VII remedies are to include familiar NLRA remedies,
including back pay: "The relief sought in such a suit would be an
injunction against future acts or practices of discrimination, but
the court could order appropriate relief, such as hiring or rein
statement of employees and the payment of back pay. This [Title
VII] relief is similar to that available under the National Labor
Relations Act in connection with unfair labor practices; 29 United
States Code 160(b)." Id. at p. 7214. This memorandum by Title VII's Senate floor managers has commanded the attention of numerous
courts seeking to clarify the legislative intent of Title VII. See,
e.g., Griggs v. Duke Power Company, 401 U.S. 424 (1971), Local 189,
United Papermakers and Paperworkers v. United States, supra, at 987
n.7; Quarles v. Philip Morris .Inc. , 279 F. Supp. 505, 515 (E.D. Va.
1968) .
80
pay awards are designed to give effect to the public policy of the
Act, and not merely to provide compensatory relief to wronged in
dividuals. This Court has pioneered in stating the public purposes
of such awards. "A backpay proceeding is designed to vindicate a
public, not a private, right as to deter unfair labor practices; the
employee is but a beneficiary." Trinity Valiev Iron & Steel Co. v.
NLRB, 410 F.2d 1161, 1168 (5th Cir. 1969). Accord; NLRB v. Mooney
Aircraft, 366 F.2d 809, 811 (5th Cir. 1966); see also NLRB v. United
Marine Division. Local 333. 417 F.2d 865, 868 (2nd Cir. 1969).
Thus, back pay in NLRA cases is both an element of adequate in
dividual relief and a means of deterring unlawful conduct.
Title VII of course embodies a national policy and commitment
as strong as that of the National Labor Relations Act. Surely we
cannot allow this national goal to be attended with any less vigor
and seriousness than the NLRA's goals of ensuring employee rights
and preserving industrial peace. The back pay award is therefore
as appropriate here as in the well-established line of NLRB pro
ceedings .
As a matter of sound judicial interpretation and administra
tion of the Act, the availability of class-wide back pay should not
be left to the unguided discretion of district judges. The incon
sistency of results thereby invited would seriously undermine the
statutory purpose of a uniform national commitment to end job dis
crimination. As shown above, the purposes of Title VII require
adoption of a policy allowing class-wide back pay in appropriate
circumstances. We propose for back pay the standard adopted by
81
the Supreme Court for attorneys' fees in Title II actions, under
Newman v. Piggie Park Enterprises, Inc., 390 U.S. 400 (1968): such
relief should be awarded unless special circumstances would render
such an award unjust. See also. Miller v. Amusement Enterprises,
Inc., 426 F.2d 534 (5th Cir. 1970). A growing number of decisions
are adopting the same standard for attorneys' fees under Title VII.
See, e.g., Lea v. Cone Mills, supra, at 88; Robinson v. Lorillard
Corp., supra, at 804; Clark v. American Marine Corp.. 320 F. Supp.
709 (E.D. La. 1970), aff'd per curiam 437 F.2d 959 (5th Cir. 1971).
Under both Titles, attorneys' fees have been awarded as an incen
tive to compliance with and enforcement of the statutory mandate
to end discrimination. The same standard should apply for back
pay awards, for the same reasons.
Under the standard we have proposed, the circumstances of
this case, including the severity of the economic injury inflicted
on black class members by defendants and the pervasiveness of
their discrimination, clearly warrant the back pay award we seek.
The district court's exercise of discretion in denying the award
69/
(if it was truly such) was erroneous. This Court should direct
the district court to conduct proceedings to determine the amount
of back pay due to each member of the class who suffered financial
69/ Where the district courts have abused their discretion or
where it is clear that they have not exercised discretion in light
of the objective of the Act, this and other Courts have not hesitated
to reverse. United States v. Hayes International Corp., supra,
at 1044-1045; Miller v. International Paper Co., supra; Pettway v.
American Cast Iron Pipe Co., 411 F.2d 998, 1003 (5th Cir. 1969);Lea v. Cone Mills Corp., supra.
82
injury as a result of defendants' discriminatory practices and
70/policies, as defined by this Court.
VI. THE COURT BELOW ERRED IN LIMITING THE
PERIOD FOR ANY CLASS BACK PAY AWARD
BY APPLICATION OF AN INAPPROPRIATE
AND UNDULY RESTRICTIVE STATUTE OF
LIMITATIONS
In considered dictum, the district court expressed the view
that the applicable statute of limitations would restrict the
amount of back pay claims by unnamed members of the class, even
if class-wide back pay were found proper (Opinion at 58-59).
While the district court did not explicitly state what limita
tions period it would apply, it apparently referred to statutes
21/specifying 2 or 4 years. (_Id. at 59) Appellants contend that
70/ All black individuals in the subclasses of pre-July 29, 1963
hirees denominated Class I, Class 2, and Class 3 by the court be
low (see US decree at 2) were found to have been victims of
defendants' discrimination; as to these individuals, therefore,
only the court's holding of law with regard to class-wide back pay
stands in the way of their recovery of financial relief. Of the
189 blacks hired after July 29, 1963, 28 were high school graduates
who passed the employment tests whose legality is disputed here,
but 23 were assigned despite these qualifications to laborer jobs.
These 23 would qualify for back pay relief to the same extent as
the pre-July 29, 1963 hirees, if this Court reverses the district
court's holding on defendants' assignment practices. The remain
ing 161 post-1963 hirees who did not pass the tests would qualify
for back pay to the same extent as pre-July 29, 1963 hirees if this Court holds Georgia Power's tests unlawful.
71/ Ga. Code Ann. §3-704, §3-706.
The district court also hinted that Title VII's 90-day
statute of limitations for filing the charge of discrimination with
EOOC, 42 U.S.C. §2000e-5(d), might apply to limit individual relief
to class members. We submit that any such application of §2000e-5(d),
which is a purely procedural section, to remedial issues would be
plain misreading of the Act. The legislative history makes clear
83
the correct limitations period in this case is the time since
July 2, 1965, the effective date of the Act. We reach this result
72/by application of the traditional doctrine of 42 U.S.C §1988 to
Title VII relief. Title VII does not specifically provide for any
limitations period with respect to relief. The applicable period
must therefore be imputed to the statute. Appellants know of no
authorities which have dealt directly with this issue under Title
73/VII; the narrow question appears to be of first impression.
We propose that the method of reference over to the analogous state
statute, embodied in 42 U.S.C. §1988, should be adopted for
Title VII purposes.
Section 1988 applies, by its terms, to actions under Chapter
__/f.n. 71 Cont'd
what the federal courts have ruled time and again: the 90-day limi
tations period for filing is designed only to relieve EEOC from the
burden of investigating stale or ancient claims. See, e.g.,
Boudreaux v. Baton Rouge Marine Contracting Corp.. 437 F.2d 1011,
1014 n .6 (5th Cir. 1971). The period is not to be used to limit
the full and effective relief which must be granted upon proof of
timely claims. Cf. Local 53, Int11 Ass'n. of Heat & Frost Insulators & Asbestos Workers v. Vogler, supra.
72/ 42 U.S.C. §1988 directs the federal courts, in civil rights
cases under Chapter 21 of Title 42, to apply the analogous remedial
statutes (including, as federal courts have construed §1988,
statutes of limitations) of the state where the court is located,
in a manner consistent with the liberal purpose of the federal Acts.
73/ This Court has apparently approved the method proposed by appellants, Boudreaux v. Baton Rouge Marine Contracting Corp.,
supra, at 1017 n.16 [dictum], although we would take issue with its choice of the analogous state law in that instance. Accord,
Clark v. American Marine Corp.. 304 F.Supp. 603, 611 n.12 (E.D. La. 1970) [dictum].
84
21 ("Civil Rights") of Title 42, United States Code. The instant
suit is brought under subchapter VI of Chapter 21 (Title VII),
and therefore §1988 may be applied here. The provisions of 42
U.S.C. §1988 direct that federal courts in certain civil rights
actions where applicable federal statutes"are deficient in the
provisions necessary to furnish suitable remedies" shall follow
74/the host state's own law (insofar as consistent with federal law).
Numerous federal courts have applied §1988 to supply a statute of
limitations in actions under the Civil Rights Acts of 1866 (42
U.S.C. §1982), 1870 (42 U.S.C. §1981), and 1871 (42 U.S.C.
§1983, 1985), which like Title VII have no express statute of
75/
limitations. In the case of §1981 several courts have used §1988
in the context of employment discrimination actions. See, e.g.,
Lazard v. Boeing Co.,___F.Supp.___ , 3 FEP Cases 643, 645 (E.D.
La. 1971); Waters v. Wisconsin Steel Works of International
Harvester Co., 427 F.2d 476, 488 (7th Cir. 1970).
The court below appeared to assume the applicability of
§1988 (or its method) to the present Title VII action, in invoking
Ga. Code Ann. §§3-704, 3-706. But the court below analogized this
employment discrimination claim to the wrong state statutes.
The method to be followed in applying §1988's provisions
to the selection of a statute of limitations in civil rights cases
74/ The same rule of reference is provided for federal actions
generally by 28 U.S.C. §1652.
75/ See, e.g., n.78 infra.
85
was succinctly stated by this Court in a series of recent cases.
In Beard v. Stephens. 372 F.2d 685 (1967), this Court summarized
the correct method as follows:
We look first to federal law to determine
the nature of the claim and then to state
court interpretations of the state's
"statutory catalogue" to see where the
claim fits into the state's scheme.
[citation omitted] The problem is a two
step one — first, what under federal law
is the "essential nature" of the claim,
and second, what statute of limitations would the state courts hold applicable to
this type or class claim.
372 F.2d at 688. See also, Franklin v. City of Marks. 439 F.2d
665, 670 (5th Cir. 1971). In the second step of the Beard pro
cedure, the court should select the analogous state statute with
a view to effectuating the liberal purposes of the federal act.
Thus, in general,
In civil rights cases, federal courts
should use that combination of federal law,
common law and state law as will be best
adapted to the object of the civil rights
laws and must use common law powers to
facilitate and not to hinder proceedings in vindication of civil rights. 42 U.S.C.§1988.
Brown v, City of Meridian. 356 F.2d 602, 605, (5th Cir. 1966). Accord.
Lefton v. City of Hattiesburg, 333 F.2d 280 (5th Cir. 1964).
In a case involving the availability of remedies under the Civil
Rights Act of 1871, this Court clearly outlined the liberal and
effect-oriented policy that should be implemented through §1988:
Indeed, §1988 uses sweeping language.
It reflects a purpose on the part of Congress
that the redress available will effect
uate the broad policies of the civil rights statutes .
* * * *
86
interest expressed in Title VII would be ill served by reference
to the inapt analogies of the shorter Georgia statutes mentioned
by the district court. The proper analogy is to the 20-year clause
76/
of §3-704. That clause imposes a 20 year limitations period
for "all suits for the enforcement of rights accruing to indi
viduals under statutes, acts of incorporation, or by operation
of law." A Title VII action seeking relief from employment dis
crimination is assuredly such a suit enforcing rights "under
77/statutes" and 'Accruing by operation of law," specifically in
this instance the Civil Rights Act of 1964, 42 U.S.C. §2000e-5(g).
Since there were no Georgia civil rights statutes and no common-
law right of action for racial discrimination in employment, back
76/ Section 3-704 provides: "All suits for the enforcement ofrights accruing to individuals under statutes, acts of incorpora
tion, or by operation of law, shall be brought within 20 years
after the right of action shall have accrued; provided, however,
that all suits for recovery of wages, overtime, or damages and
penalties accruing under laws respecting the payment of wages
and overtime . . . shall be brought within two years after the
right of action shall have accrued. (Acts of 1955-56, p. 234, p. 333) . "
For practical purposes here, the Title VII back pay claim
would date only from July 2, 1965, the effective date of Title VII, rather than from 20 years ago.
77/ The right so enforced is not simply a right to back pay.
It is also the right of effective injunctive relief; as this Court
has recognized, the back pay claim is merely incidental to the
equitable remedy, Johnson v. Georgia Highway Express, supra, at
1125. Beyond its purely compensatory aspect, the availability of
class-wide back pay serves as a deterrent to continuing discrimina
tion in violation of Title VII and will unquestionably foster
voluntary compliance of an injunctive nature with the Act's provisions.
8,
Thus §1988 declares a simple, direct
test: what is needed in the particular
case under scrutiny to make the civil
rights statutes fully effective?
Brazier v. Cherry, 293 F.2d 401, 408, 409 (5th Cir. 1961).
Because the policy underlying Title VII is a serious and
comprehensive national commitment to eradicate job discrimi
nation, Culpepper v. Reynolds Metals Company, supra, at 891,
federal courts in proceedings to enforce Title VII should assure
the selection of a statute of limitations appropriate to accom
plishment of Title VII's broad remedial aims. The Supreme Court
has recently directed that such choices with respect to remedial
damages be guided and informed by the liberal aims underlying
the civil rights acts.
The existence of a statutory right implies
the existence of all necessary and appro
priate remedies . . . This means, as we read§1981, that both federal and state rules on
damages may be utilized, whichever better
serves the policies expressed in the federal statutes.
Sullivan v. Little Hunting Park, 396 U.S. 429, 439-440 (1969)
(emphasis added). This Court has even held that the most closely
analogous state limitations statutes should be rejected where a
"significant federal interest makes them inappropriate." Huson
v. Oatis Engineering Corp., 430 F.2d 37, 42 (1970). Clearly, the
policy of remedial liberality relevant to the choice of the ana
logous state statute is a strong and comprehensive one.
We turn now to the choice of the applicable Georgia statutes,
guided by the principles discussed above. The substantial federal
pay under Title VII must be a "statutory liability". Georgia
state law, moreover, leaves no question that "the 20-year statute
of limitations is clearly applicable to a statutory liability
or penalty." Bankers Fidelity Life Insurance Co. v. Oliver. 106
Ga. App. 305, 126 S.E. 2d 887, 892 (Ga. 1962).
The other analogies followed by the district court, re
sulting in adoption of a two or four year limitations period,
fall far wide of the mark. The exception clause of §3-704, pre
scribing a two-year statute, applies only to claims for lost wages,
overtime, and damages and penalties under overtime-wages laws.
Obviously, appellants' claim here is not in the nature of a simple
claim for lost wages or overtime. Section 3-706, prescribing a
four year period, applies only to "Open accounts, breach of con
tract not under hand of party, [and] implied assumpsit." But
the present action is not for breach of contract or related
actions; it attacks the legality of the contract itself and
practices under the contract.
Our reference to the Georgia statute governing actions "for
the enforcement of rights accruing to individuals under statutes .
or by operation of law" is supported by numerous analogous civil
rights claims in which the federal court carefully selected the
applicable statutory analogue from among competing possibilities.
In each of these cases, courts considered and rejected application
of the profferred common-law analogue [e.g., tort or contract] and
chose instead to analogize to the statutes governing liability
created by statute or civil actions not otherwise provided for
89
under state law.
Our position if surther supported by federal court rulings
under the Equal Pay Act, 29 U.S.C. §206, et seq_. , a remedial statute
in many respects similar to Title VII. The Equal Pay Act provides
for actions to recover back pay lost as a result of sex discrim
ination, but contains no statute of limitations. In Shultz v.
Wheaton Glass Co., F.Supp. , 3 EPD ^8270 (D.N.J. 1970), aff'd
in pertinent part, F.2d___, 3 EPD 18296 (3rd Cir. 1971), the
court ordered restitution of back pay from the effective date of
the Act, March 1, 1965. In so holding, the court pointed out that
the availability of full back pay was an adjunct to the remedial
purpose of the Act, "not merely to enforce an employee's private,
individual right; rather, it was more expansive and expressly
designed to 'correct a continuing offense against the public
interest'" 3 EPD 18270 at p. 6910. The court also noted that
78/
78/ In Baker v. F & F Investment. 420 F.2d 1191 (7th Cir.
1970), the Court of Appeals for the Seventh Circuit held that a 42 U.S.C. §1982 action was "a cause of action created by
statute," and therefore fell within the realm of the Illinois
limitations statute governing "actions not otherwise provided
for." Earlier the Seventh Circuit had applied the identical
analysis to an action under 42 U.S.C. §§1983, 1985, Wakat v.
Harlib, 253 F.2d 29 (7th Cir. 1958). In both cases the court rejected the suggestion of a tort statute analogy.
A number of Circuit Courts have invoked the same
reasoning in rejecting the tort statute analogy for 42 U.S.C.
§1983 actions. See, e.g., Smith v. Cremins. 308 F.2d 187 (9th Cir. 1962); Glascoe v. Howell. 431 F.2d 863 (8th Cir. 1970);
Franklin v. City of Marks. 439 F.2d 665 (5th Cir. 1971);
see also Lazard v. Boeing Co.. __F.Supp._, 3 FEP Cases643 (E.D. La. 1971); Mclver v. Russell, 264 F.Supp. 22 (D. Md. 1967).
90
the discriminatory employer had extensive notice that its practices
would incur liability. Both reasons are equally applicable to
Title VII and the instant case. See also, Hodgson v. American
Can Co.. Dixie Products.___ F.2d __, 3 EPD J8171 (8th Cir. 1971),
adopting Wheaton Glass.
For these reasons the court should look to the 20-year
statute as both the most similar and the most liberal state
statutory analogue. Since the adoption in principle of the
suggested twenty year period would impose pre-Act Title VII lia
bility, however, the plaintiffs recognize that defendants' lia
bility should run only from July 2, 1965 to present.
VII. THE COURT BELOW ERRED IN LIMITING ITS
AWARD OF ATTORNEYS' FEES TO PLAINTIFFS'
COUNSEL BY APPARENTLY FAILING TO COM
PENSATE THEM FOR TIME SPENT AT AND AFTER
TRIAL AND BY BASING THE AWARD ON A SUG
GESTED MINIMUM FEE SCHEDULE THAT WAS
INAPPROPRIATE FOR THIS CASE.
The court below recognized that Title VII, 42 U.S.C.
§2000e-5(k), and Clark v. American Marine Corp.. 437 F.2d 959
(5th Cir. 1971) required an award of counsel fees to plaintiffs'
counsel. Nevertheless, the court below awarded plainly inadequate
fees to counsel for plaintiffs, Peter E. Rindskopf and Isabel
Gates Webster. We reach this conclusion by direct application of
the standard which this Court held to be appropriate in Clark v.
American Marine Corp., supra. That ruling affirmed the decision
91
of Judge Rubin, 320 F. Supp. 709 (E.D. La. 1970), which had defined
the proper counsel fees standard to include consideration of the
factors listed in the A.B.A Code of Professional Responsibility,
Disciplinary Rule 2-106. See also Annotation, 56 ALR 2d 13 (1957).
The record clearly shows that the court below ignored or
misapplied this clear standard in determining the amount of counsel
fees awarded. The court did not even compensate Mr. Rindskopf and
Mrs. Webster at the rate suggested under the pertinent minimum fee
schedule, and wholly failed to take account of factors which indicate
that a rate much higher than the minimum rate was appropriate here.
Accepting the district judge's findings with respect to time
spent by plaintiffs' counsel prior to trial and to the applicable
minimum suggested rate schedule of the Georgia Bar Association79/
(Opinion at 63-64), it is clear that Mrs. Webster received $360
less than the minimum for her pre-trial time, and Mr. Rindskopf
only $1,150 more than the minimum. This award clearly fails to com
pensate both counsel adequately (or at all) for their attendance
and participation at the seven-day trial, or Mr. Rindskopf for his
post-trial work including the post-trial brief, settling individual
client claims, and discharging plaintiffs' part of drawing the
decree. Thus the district court denied plaintiffs their due under
the Title VII counsel fee provision, 42 U.S.C. § 2000e-5(k), even
79/ The court found that Mr. Rindskopf had logged 214 hours on
the case, and Mrs. Webster 96 hours, prior to trial. It found
the applicable suggested minimum fee schedules to be $25 per hour
for Mr. Rindskopf, $35 for Mrs. Webster. The court awarded Mr.
Rindskopf $6,500, and Mrs. Webster $3,000. (Opinion at 63-64).
92
on a simple hourly, minimum rate basis. An examination of the
factors which the district judge relied on, and those he neglected,
demonstrates that his counsel fee award flouts both the standards
enunciated by Judge Rubin and the applicable principles of Title
VII law generally.
(1) Judge Smith suggested in his opinion that he had re
duced the award because of the participation of the U.S. Justice
Department through the consolidated action of United States v.
a _ i y
Georgia Power Co. (Opinion at 64). This was error. The court in
Dobbins v. Local 212, IBEW, 292 F. Supp. 413, 450 (S.D. Ohio, 1968)
originally made similar error but on reconsideration held that
80 /
80/ The trial transcript raises doubts that the district judge
even took proper account of the time plaintiffs' counsel testified
to spending prior to trial. During plaintiffs' submission of
evidence regarding their time records on the case, when Mrs. Webster
was asked on cross-examination to explain why she had attended the
entire trial, the court stated: "It doesn't make a lot of differenceI'm going to try to do something practical when it comes to attorneys
fees when we reach that point anyway." 4 Tr. 134. After the evi
dence was in, the district judge elaborated on his proposed "practical" solution:
One reason I like to handle this matter of
attorneys' fees in this way, I have a wage
hour case and one of my best friends got up
and he was the only evidence on the value of
the services to the employee who had a recovery in the case, and I awarded 60% of what he
testified. He has always taken unbrage at me
on account of that, and I told him that was
about right, I believe about three-fifths of
everything he said. If I don't have any evi
dence, I don't run into that problem. 4 Tr. 135.
81/ We would note that the government's case was filed several
months after both of the private cases, and that consolidation was
ordered 18 months after that. Plaintiffs engaged in substantial
litigation prior to consolidation, see 295 F. Supp. 943 (1968),
310 F. Supp. 327 (1969), and 61 LC H 9389 (1970), and continued to
litigate diligently as parties to the consolidated action.
93
although private counsel should not be compensated for work done by
the government, neither should their fees for work actually done be
reduced merely because the government has taken some part in the
litigation, 61 LC 19327 at p. 6860. Plaintiffs here seek compen
sation only for their efforts, not for the government's.
(2) The district court failed to take account of the truly
extraordinary complexity and novelty of the issues presented by
this case. As the lengthy opinion of the court below and foregoing
portions of this brief amply demonstrate, this case raises a number
of the most difficult and fundamental issues in the developing law
of employment discrimination. The nature of these issues, and
counsel's skill in presenting them, should have been a factor in
setting the fee award. Here these facts demanded more, not less,
than a minimum schedule award. As Judge Rubin pointed out (and
the district judge well knew), "in difficult matters of this kind
an hourly fee substantially in excess of the minimum is frequently
charged." Clark v. American Marine Corp., supra, at 712.
(3) The court failed to accord proper consideration to the
experience, reputation, and ability of plaintiffs' counsel - another
factor recognized by Clark. See also, ABA Code of Professional
Responsibility, Disciplinary Rule 2-106(B)(7). Plaintiffs' counsel
had appeared and represented numerous clients before the court
below and other district courts, in this and other Circuit Courts,
and in the Supreme Court.-&2l
82/ To list only partially counsel's reported participation in
Title VII cases in local courts:
Rindskopf: Sanders v. Dobbs Houses, Inc., 431 F.2d 1097 (5th
Cir. 1970, cert. den. 3 EPD f8127 (1971); Johnson v. Georgia Highway
94
In addition to the facts in their testimony relating to their ex-
83/
perience, the court should have taken notice of their experience,
reputation, and ability as shown in their reported decisions.
(4) Finally, the court below failed to consider that the
allowance of any fee for plaintiffs' counsel was contingent on 84/
their success. Cf. ABA Code of Professional Responsibility,
Disciplinary Rule 2-106(B)(8). Courts traditionally have awarded
greater fees to attorneys working on a contingent fee basis. See,
e.g., Freeman v. Ryan, 408 F.2d 1204 (D.C. Cir. 1968); Simler va
Conner, 352 F.21138 (10th Cir. 1965). This judicial inclination
is bottomed on the realization that the fair and equitable func
tioning of our judicial system requires adequate reward to risk-
82\J cont'd.
Express, 417 F.2d 1122 (5th Cir. 1969); Culpepper v. Reynolds
Metals Co., 421 F.2d 888(5th Cir. 1970); Long v. Georgia Kraft Co..
4 EPD 17556 (5th Cir. 1971); Colbert v. H-K Corp., 3 EPD 18248 (5th Cir. 1971); Evans v. I-T-E Corp.. 313 F. Supp. 1354 (N.D.
Ga. 1970); Long v. International Brotherhood, 60 LC 19306 (N.D.
Ga. 1969); Holliday v. REA Express, 306 F. Supp. 898 (N.D. Ga.
1969), etc.
Webster: Sanders v. Dobbs Houses, Inc., supra; Jones v.
Montag, 3 EPD 18243 (N.D. Ga. 1969); Banks v. Lockheed-Georgia
Corp., 46 FRD 442 (N.D. Ga. 1968); Allen v. Lockheed-Georgia Corp.,
61 LC 19333 (N.D. Ga. 1968).
83y Mrs. Webster had 12 years of litigation experience. Although
Mr. Rindskopf had been in practice only a few years, his extensive
experience in civil rights and Title VII cases and his manifest
ability required a more substantial award. See Clark v. American
Marine Corp., supra, at 712; Cape Cod Food Products, Inc, v.
National Cranberry Ass'n., 119 F. Supp. 242 (D. Mass. 1952).
84 / By way of contrast, counsel for defendant companies in Title
VII suit may be presumed to receive compensation win, lose, or draw.
95
taking by attorneys on behalf of clients who would otherwise be
unable to pay the price of litigation. As was recognized long
ago in a bankruptcy case, the success of such counsel should be well
rewarded as
an incentive to attorneys to put forth
their best efforts in cases which appear
unpromising . . . . There must always be
a flavor of generosity in the awards to
success in order that an appetite for
effort may be stimulated.
In re Osofsky, 50 F.2d 925, 927 (S.D. N.Y. 1931).
The contingency factor should be given special consideration
in setting counsel fees in civil rights cases. As the Supreme
Court has observed,
Lawsuits attacking racial discrimination
. . . are neither very profitable nor very
popular. They are not an object of general
competition among . . . lawyers: the pro
blem is one of an apparent dearth of lawyers
who are willing to undertake such litigation,
(footnote omitted)
NAACP v. Button, 371 U.S. 415, 443 (1963).
A plaintiffs' attorney in a Title VII case acts "as a
'private attorney general,' vindicating a policy that Congress con
sidered of the highest priority," and this was the primary reason
for the enactment of the counsel fee provisions in the Civil Rights
Act of 1964. Newman v. Piggie Park Enterprises, Inc., supra, at
402. See Note, Employment Discrimination and Title VII of the
Civil Rights Act of 1964, 84 Harv. L. Rev. 1109, 1255 (1971). The
court below disregarded this policy as well as the variety of
specific factors we have discussed above. This Court should cor
rect the district court's errors by ordering that the principles
96
stated by Judge Rubin are to govern and that these principles
require a substantially more generous award of attorneys' fees
here.
CONCLUSION
We respectfully urge this Court to hold that the decision
below was in error in each of the respects set forth herein,
and in reversing to enter an appropriate order correcting each
of the district court's enumerated errors. This Court's order
should hold that (1) Georgia Power's use of its employment tests
for promotion and hiring must be enjoined; (2) Georgia Power
has discriminated against black employees in its assignment
practices; and (3) Georgia Power's word-of-mouth recruitment
system violates Title VII and must be replaced by appropriate
affirmative measures. The Court should also direct the award
of further relief to plaintiffs and class members by (1 ) award
ing all identified victims of discriminatory practices and all
black employees company-wide seniority to be applicable in
a practically effective manner; (2 ) awarding named plaintiffs
back pay to compensate them for the full amount of their loss
due to post-Act discrimination; (3) directing the award of
back pay to members of plaintiffs' class; (4) directing that
back pay awards be computed for the entire period since July 2,
97
1965; and (5) ordering the award of fully adequate attorneys'
fees (as defined herein) to plaintiffs' counsel.
Respectfully submitted,
75 Piedmont Avenue, N.E. Suite 1154
Atlanta, Georgia 30303
ISABEL GATES WEBSTER75 Piedmont Avenue , N.E. Suite 1170
Atlanta, Georgia 30303
JACK GREENBERG
WILLIAM L. ROBINSON
MORRIS J. BALLER
10 Columbus Circle Suite 2030
New York, New York 10019
Attorneys for plaintiffs-Appellants
ALBERT J. ROSENTHAL GEORGE COOPER
HARRIETT RABB
435 West 116th Street
New York, New York 10027
Of Counsel
98
APPENDIX A
Note on Form of Citations
In accordance with its usual procedure in cases involving
a record of this size, the district court clerk's office did
not individually number the pages of the record. Instead, it
assigned "document numbers" to each pleading in the record.
Because of the interlocking and overlapping of both these
documents and the numbers assigned to them in these three con
solidated cases, appellants believe that citation to page numbers
of documents identified only by number would prove unnecessarily
confusing. Instead, appellants have chosen to cite to the
record by brief, descriptive document titles as follows:
(1) Each citation to the record indicates in which
civil action the pertinent was filed. Record
documents in No. 11723 are identified "King" documents;
those in No. 12185 as "Moreman" documents; and those
in No. 12355 as "US" documents.
(2) Where no indication appears as to which action's
record is involved, the document involved is common
to all three actions. For example, the transcript,
trial exhibits, and opinion citations are so treated.
(3) Citations to the trial transcript indicate the
volume and page of the pertinent citation as follows:
"2 Tr. 100" means Volume 2, page 100 of the transcript of trial held January 18-26, 1971.
(4) Citations to trial exhibits indicate the official
number of the exhibit, its source, and the page number,
if any, within the exhibit. For example, "Gov. Ex. 15 at 100"; "Co. Ex. 75 at 10".
(5) Citations to specific pages of the Court's opinionof June 30, 1971 are made as follows: "Opinion at _____."
(6 ) Citations to specific pages of the decrees are
made as follows: "US decree at ____" for the decree
entered September 27, 1971 in No. 12355, and "Private
decree at ____" for the decree entered October 26, 1971in Nos. 11723 and 12185.
A-1
For the Court's convenience of reference, appellants list
on the following page the brief citation title, full title, date,
and document number (as indicated by the district court clerk
frequentlyon the docket sheets in each action) of every item/cited in this
brief, except trial exhibits (which did not receive any document
number).
[ See p. A-3, Table of Documents ]
Also, on second following page appellants list by
subject matter the trial exhibits which are most frequently
cited in their brief.
[See p. A-4, Table of Exhibits ]
A-2
Citation Title
NO. 11723
Document NO
NO. 12185
Document No
>iu>
King Complaint 2
Moreman Complaint
US Complaint
King Answers 14, 16
Moreman Answers
Pre-trial Order 28
Motion for Preliminary
Injunction
Preliminary Injunction
Amended Preliminary
Injunction
Preliminary Injunction
Decree
1-7 Transcript
Opinion 52
2
12, 11
24
36
US Decree
Private Decree 55 39
Notice of Appeal 58 42
NO. 12355
Document No Full Title & Date Filed
Complaint, King et al v. Georgia Power
Company et al.. filed April 12, 1968.
Complaint, Moreman et al v. Georgia Power
Company et al., filed October 28, 1968
2 Complaint, United States y. Georgia Power
Company et al., filed January 10, 1969
Answers of Georgia Power Company and
Local No. 84, filed August 16, & 22, 1969
Answers of Georgia Power Company and
Local No. 84, filed March 19 & 21, 1969
Pre-trial Order filed August 19, 1969
68 Plaintiffs' Motion for Preliminary
Injunction, filed July 13, 1970.
88 Order of Court, filed September 22, 1970.
92 Order of Court, filed October 6, 1970.
93 Order of Court, filed October 6, 1970.
109-115 Transcript of proceedings held January
18-26, 1971, N.D. Ga., Vols. 1-7
127 Memorandum including Findings of Fact,
Conclusions of Law, and Order, filed
June 30, 1971.
131 Final Decree in No. 12355, filed September
27, 1971
Final Decree in Nos. 11723 and 12185,
filed October 26, 1971.
Notice of Appeal by Appellants Charles King,
et al., filed November 9, 1971.
Brief Citation
Gov. Ex. 1
Gov. Ex. 2
Gov. Ex. 6
Gov. Ex. 7
Gov. Ex. 12
Gov. Ex. 13
Gov. Ex. 14
Gov. Ex. 15
Gov. Ex. 17
Gov. Ex. 41
O 0 • Ex. E
Co. Ex. 1-B
t
0u Ex. 3
Co. Ex. 7
Co. Ex. 9
o 0 • Ex. 75
Subject Matter
Memorandum of Agreement between Georgia Power
and defendant Unions, dated July 1, 1965
Memorandum of Agreement between Georgia Power
and defendant Unions, dated July 1, 1968.
Employment application forms (microfilm)
Employee history cards (microfilm)
Georgia Power's answers to plaintiffs' second interrogatories
Georgia Power seniority lists, various
divisions, 1969-70.
Number of employees by race and job as of
January 16, 1970 (supplemented by Co. Ex. 9).
Employment history of all Black employees as
of January 16, 1970 (supplemented by Co. Ex. 9)
Employment history and educational background
of all empl yees in Steam Plants Atlanta and
Macon Operating Divisions (bargaining unit
jobs) (supplemented by Co. Ex. 9).
Summary data sheet from Georgia Power
validation study.
Lines of progression
Manual-Testing Program Procedures, Georgia
Power Company, August 1, 1964
Manuals and Tests Utilized by Georgia
Power Company
Payroll print-out for all employees as of
January 16, 1970
Chart correcting Gov. Exhibits 14, 15, & 17,
with supporting documents
Validation Data and Procedures for the
Employment Testing Program of Georgia Power
Company, January , 1971.
A-4
References to Gov. Exhibits 14, 15 or 17 which carry the
notation "(Supplemented)* indicate that the data reported appears
in the indicated exhibits as corrected by Co. Ex. 9. The latter
exhibit was a final and official adjustment of the figures and
data in the Government exhibits, to reflect the truth as it
could be best determined by the parties. It is these appellants'
understanding that the figures shown in these exhibits, as
amended by Co. Ex. 9, were not in dispute at trial and are not
open to purely factual dispute here.
A-5
APPENDIX B
Materials on the Technical Defects of the Georgia Power Testing Program and Validity Study
We discuss in this Appendix several of the various technical
flaws in the Georgia Power Company's defense of its testing program.
These flaws are centered on the defects in Dr. Hite's validity study,
but extend to the nature of the testing program itself.
(1) We first set forth a list of all tests and the jobs they
apply to, and reproduce Government Exhibit 41 as the basis for the following discussion.
[ See Use of Tests on p. B-3 ]
[ See Gov. Ex. 41 on p. B-2 ]
Explanation of Column Headings on Gov. Ex. 41
(1) N = Number of Observations
(2) df - degrees of freedom for evaluation of Discriminant Analysis
(3) F = obtained F ratio for Discriminant Analysis
(4) F ratio required for .05 level of confidence
(5) F ratio required for .01 level of confidence
(6) df = degrees of freedom for regression analysis
(7) R = Multiple regression coefficient
(8) Multiple R needed to obtain .05 level of confidence
(9) Multiple R needed to obtain .01 level of confidence
(10) Standard Error of Multiple R obtained by
d R where m = no. of variables
(11) RS*r
(12) Probability of obtaining R (column 7) by chance
B-l
Government Exhibit 41
Tests of Significance for Georgia Power Statistics"
(1 ) |(2) (3) (4)
Job N df F .05
Linemen 24 2 , 2 1 1.41
Linemen 50
Truck Operators 76 2,72 4.47 3.13
Mechanics, Electricians, 42 3,38 1.51 2.85
etc.
Clerical L01 2,98 1.56 3.09
Apprentice Mechanics 20 3,16 1.09 3.24
Appliance Servicemen 46 3,42 0.87 2.83
Storekeepers 32 3,28 0.77 2.95
Metermen 19 3,15 0.63 3.29
Helpers 48 3,44 1.02 2.82
Coal Equipment Operators 19 3,15 0.18 3.29
Meter Readers 39 3,35 0.97 2.87
Garage Mechanics 37 3,26 3.46 2.98
Switchboard Operators 30 3,33 1.93 2.90
* [sic]. These two figures are in error; see
** Level of confidence reached
DISCRIMINANT REGRESSION ANALYSIS
(5)
.01
(6)3f_
(7)
R
(8)
.05
(9)
.01
(in
- 5 4
(1 2)**
46 .31 .38 .45 .13 2.32 .02
4.92
4.34
4.82
5.29 16 .51 .62 .71 .18 2.83 .01
4.29 42 .20 .41 .48 .15 1.33 .18
4.57 28 .44 .49 . 57 .15 2.93 .01
5.42 15 .39 .57 . 66 .22 1.77 .03
4.26 44 .17 .40 .47 . 1 1 1.55 .12
5.42 15 .46 .40* ** .47* .20 2.30 .02
4.40 35 .14 .44 .52 .17 0.82 .41
4.64 33 .36 .46 .54 .15 2.40 .02
4.46 26 .50 .51 .59 .14 3.57 .01
p. b -4, infra.
Use of Tests by Georgia Power Company
The following positions require passing only of the Personnel
Tests for Industry: apprentice trouble dispatcher, beginner; automotive pool attendant; dispatcher - appliance repair; duplicating
machine operator; forest inspector aid [sic]; janitor; maid; porter; watchman; rodman and chainmen.
The Bennett Mechanical Comprehension Test as well as the PTI tests
are given for the following positions: auxiliary equipment operator;
blue print operator; coal equipment operator; display helper; display- men; helper; helper-garage; operator D;
servicemen C; truck operator; communication technician; inspector
(electrical, mechanical or civil); instrument technician; laboratory assistant; laboratory technician; test helper.
For draftsmen, field estimator, and instrumentmen the Revised
Minnesota Paper Form Board Test (Minnesota) is given in addition to the PTI tests.
The Short Employment Tests are given for the following classifications: Collector C; I.B.M. operator C; key punch operator B; mail clerk;
meter reader, PBX operator; pool car clerk; trouble clerk.
For the following positions the General Clerical Tests are given to
gether with the SET: clerk, clerk C (acc.); district cashier; district
clerk B; estimate clerk; field office clerk; file clerk; insurance
clerk; local cashier; local clerk, receptionists; steno-clerk; stenographer, teller; typist.
For home economist and merchandise salesmen, the Wonderlic Personnel Test (Wonderlic) is given.
The operative norms or cut-off scores, since August 1, 1964,have been as follows: for the PTI-V below 22 use caution, below 16
do not employ; for the PTI-N below 14 use caution, below 12 do not
employ; for the Bennett below 36 use caution, below 30 do not employ; for the SET below 78 use caution, below 70 do not employ; for GOT
below 130 use caution, below 115 do not employ, for the Minnesota
below 45 use caution, below 40 do not employ; and for the Wonderlic
below 25 use caution, below 18 do not employ. [Co. Ex. 1-B]
B-3
(2) Technical information regarding significance of resultsshown in Gov. Ex. 41
The computations in Government Exhibit 41 fail to show
statistical significance under either of the correlation (or regres
sion) analysis computations.
First, Gov. Ex. 41 indicates on its face that the multiple
regression technique covered in columns (6 ) through (9) produces no
statistically significant results for the tests being employed by
Georgia Power Company, except possibly for the coal equipment
operator job. Only for coal equipment operators does the coefficient
of column (7) appear to reach and exceed the required level; but for
this job the reported confidence levels are in error. With a sample
size of nineteen (N-19) and fifteen degrees of freedom, the required
coefficients, as indicated for the "metermen" job, are . 57 at the .05
level of confidence and .66 at the .01 level of confidence.^/
Using the proper required coefficients, it is clear that the coal
equipment operator results fail to reach the .05 level necessary for
statistical significance.
Using the second multiple regression technique covered in
columns (10) - (12), Dr. Hite might seem to have derived a substantial
number of significant results at last. Column (12), which reports
the levels of confidence reached, indicates that six of the ten
Bl/— ' J.4th ed. P. Guilford,
McGraw-Hill, Fundamental Statistics in Psychology and Education, 1965, App. B, p. 508, Table D.
B-4
classifications surpass the .05 level of confidence. However, these
results prove upon examination to be specious. Dr. Hite used an out
dated statistical formula to effect these reported confidence levels.
Dr. Hite s reported formula is very nearly the same as the classical
formula for the standard error of a multiple correlation for n vari
ables, about which a very reputable statistical text has stated,
. . . it is not safe to use (11.13) [this classical formula] unless
N is indeed very large. Even then it is questionable."”^ in the
Georgia Power Company study there is not even a large sample size, N.
When the proper modern regression formula is applied?-^/the derived
results are those found in column (7) of Government Ex. 41, wherein
no statistical significance was evidenced.
(3) Technical information regarding Company Exhibit 75,
"Validation Data and Procedures for the Employment TestingProgram of Georgia Power Company."
The Hite study was introduced in evidence as Co. Ex. 75. We
here examine the most serious technical and procedural defects of that
study.
a) Dr. Hite failed consistently to follow any single recognized validation method.
The major technical defect in the Hite study is its persistent
shifting from technique to technique in a quest for some meaningful
B 2/ McNemar, Quinn, Psychological Statistics, 4th ed. John Wiley and Sons, Inc. (New York, 1969), p. 203.
Ĵ/ Ibid, p. 320.
B-5
results. A psychologist who tries enough different approaches in
enough different situations might be expected to come up with something.
This explanation appears to characterize the few statistically sig
nificant results which Dr. Hite did manage to produce.
Dr. Hite first attempted to validate the Georgia Power tests
by a correlation (or regression) analysis (Co. Ex. 75, p.l). He
apparently continued to pursue this analysis for most jobs, because
correlation analysis data is reported for 10 of 13 jobs in Gov. Ex.
41. However, Dr. Hite evidently stopped reporting such data in the
course of his validity study because it showed no significant results
for any job. Instead, he shifted to the discriminant analysis tech
nique .
Dr. Hite's basic approach in using the discriminant analysis
technique was to divide high scores and low scores into two groups by
drawing a cut-point line at the mean score level. However, as to
three job classifications (truck operators, mechanics, and clerical
employees), Dr. Hite abandoned his stated approach (Hite testimony,
6 Tr. 155) and shifted to a three group approach by using two cut-
point lines artfully drawn so as to include marginal scorers in groups
that would increase the apparent predictive value of the tests.
There appears to be no objective basis for treating these three job
classifications differently. It is evidently not because the jobs
involved larger sample groups since one job which was not divided into
three groups (appliance serviceman) involved a larger sample group
B-6
than one (mechanics) which was so divided.
In addition, although it is nowhere specifically reported, Dr.
Hite apparently tried other arrangements of his data to see if better
results could be generated. Thus, in Gov. Ex. 41 we note that Dr.
Hite did computations for mechanics and apprentice mechanics as
separate classifications, yet in his validity study he reports
these as a single classification (Co. Ex. 75, pp. 14-18).
Finally, in one other case, Dr. Hite shifted to yet a fourth
technique. He compared the performance of tested persons with un
tested persons for the auxiliary equipment operator job and concluded
that the tests were valuable because tested persons were doing sig
nificantly better. It is difficult to say what merit that con
clusion can possibly have, because no attempt was made to determine
whether the tested and untested groups were otherwise comparable,
e.g., whether or not one group was older, less experienced, better
trained, etc. Dr. Hite indicates nothing as to this critical ques
tion. Notwithstanding that uncertainty, he elected to use the tested-
vs.-untested comparison as his technique in just one instance, al
though he conceded that it was available for other jobs as well
(Hite testimony, 6 Tr. 158).
b) One job performance criterion used made the study's
predictive analysis tend to be self-fulfilling, and not probative.
One crucial aspect of any validity study is its choice of job
performance criterion -- the measure of job performance against which
B-7
test scores will be matched. A properly selected job performance
criterion must distinguish among levels of job performance. in
choosing a criterion, one of the most frequent and serious dangers
is that of "criterion contamination" — that is, including within
the job performance rating system some element which detracts from
its ability to appraise performance fairly.
A job performance criterion used by Dr. Hite in preparing his
validity study was highly suspect and possibly biased, in that it
contained such a "contaminating" element. This performance criterion
was a system of ratings by supervisors. Dr. Hite instructed super
visors to rate workers in terms of
[advancement] potential reflected by arithmetic skills, verbal skills, and an understanding of mechanical and
physical relationships. Try to focalize on the under
lined characteristics above. [emphasis in original.]
Co. Ex. 75, Appendix entitled "Georgia Power Company Personnel
Evaluation Instructions, Part I." The skills thus isolated for the
supervisor's consideration tended to be biased toward those
4/specific skills the tests purportedly measure. under these circum
stances an artificial correlation between test scores and job per
formance as reflected by the same criteria would naturally tend to
develop. The chosen job performance criterion, by building in this
self-serving element, has potentially contaminated the results, par
ticularly if these skills are not demonstrated to be the only proper
PTI-N, PTI-V, and Bennett Mechanical.4/
B-8
standards for assessing 30b performance. See EEOC Guidelines, 29
C.F.R. § 1607.5 (b) (3)- (4).
It is true that this contaminated "Part I" criterion described
above was used for only some of the classifications. Unfortunately,
the study does not indicate which ones, a rather careless oversight,
and we can assume that it was used for a substantial number of
classifications.
The EEOC Guidelines suggest a most reasonable approach to
the question of criterion contamination. Section 1607.5(b)(4)
provides:
. . . supervisory rating techniques should becarefully developed, and the ratings should be closely examined for evidence of bias. In
addition, minorities might obtain unfairly low
performance criterion scores for reasons other
than supervisors' prejudice, as, when, as new
employees, they have had less opportunity to
learn job skills. The general point is that
all criteria need be examined to insure free
dom from factors which would unfairly de
press the scores of minority groups.
By this reasonable approach, the job criteria Georgia Power used to
measure performance are clearly insufficient to support an accurate
predictive analysis.
c) The Hite study is further defective because of its failure to perform a cross-validation.
Cross-validation is, in effect, a statistical method of
double-checking validation results. In cross-validation, the
weighted formula derived from the initial sample group would be
applied to a new group, to see how accurately the new group's test
scores predict its actual job performance. (Kirkpatrick testimony,
3Tr. 86-89). At least one leading authority on the use of
B-9
personnel tests insists upon cross-validation:
"Before any test battery is put to use as a selection
system, therefore, cross-validation is essential.
Cross-validation means that the battery is administered
to a totally new sample, expected or predicted criterion
levels are determined for each person, and expected
performance is correlated with actual performance."
Robert M. Guion, Personnel Testing (1965), p. 165.
With one exception in the lineman classification, Dr. Hite
did not cross-validate his derived results. He attempted to justify
his failure to do so by claiming that the sample he had to work with
was too small to generate a separate group on which to cross-validate
and that it would take too long to produce a new sample of tested
personnel on whom to cross-validate. (6 Tr. 153-154). If this
were true, the EEOC Guidelines would surely permit cross-validation
to be deferred, because it would not be "technically feasible,"
and the Guidelines do not generally insist on unfeasible action.
However, it should be noted that Dr. Hite studied groups as small
as nineteen (coal equipment operators) in his reported data. Where
such a critical issue is being studied, and where the results reported
by Dr. Hite are at best questionable and tenuous, cross-validation
is imperative. It should have been possible for Dr. Hite to cross-
validate in several instances - for example, in the clerical job
where the sample included 100 employees. Instead, Dr. Hite termi
nated his study without any significant effort at cross-validation.
d) The Hite study made no attempt to obtain separate
data regarding test results for blacks and whites.
B-10
We have shown above that Georgia Power's tests produce a
racially disparate scoring pattern (See p. 17 supra). In all cases
where this phenomenon is present, accepted standards require
separate validation for whites and blacks, insofar as feasible.
EEOC Guidelines §1607.5 (b) (5) , Kirkpatrick testimony, 3 Tr. .16,
31-35, 53-54, 55. Georgia Power has argued that the number of
black workers was insufficient to form a meaningful sample group,
and therefore a separate study of blacks was not feasible, But,
as Dr. Hite admitted at trial, there were black samples for both
coal equipment operators and helpers which were as large as the
sample group for one job he included in the study. (Hite testimony,
6 Tr. 92 -- sample size of 11 for cross-validation.) Dr.
Hite’s weak excuse for not differentially validating for these two
job classifications was that the test scores for those black
employees could not be found in the Company's records. But, the
necessary test scores could easily have been obtained by readminister
ing the tests to those blacks. The EEOC Guidelines would have re
quired at least this effort as part of a proper validity study.
And what the Guidelines call for in this situation - the compilation
of differential test data only "whenever technically feasible",
§1607.5(b)(5) - is eminently reasonable.
e) The Hite study totally fails to take into account
the manifest evidence of test administration dispari
ties.
B-ll
The record is replete with evidence of racially tainted
irregularities surrounding notice of testing, how and when tests
were administered, notice of results, use of the "waiver" formB5/
and retesting. The district court found that there were indeed
instances of administrative deficiencies in Georgia Power's test
program (Opinion at 35, Finding #71). The court however dis
counted this problem by finding that there was no differential
racial impact of these administrative problems, ignoring sig-
B6/nificant testimony to the contrary. In any case, it is import
ant to consider that to the extent that these administrative
deficiencies did indeed exist, the data on which Dr. Hite's
study is built, the test scores, are not accurate measures of
even the person's ability to perform well on the employment
tests, but rather reflect in part the conditions in which the
Company tested him. The data then becomes doubly self-ful
filling, since it measures in part the effects of the Company's
testing practices, rather than the aptitude of the person taking
the test.
B5/ 1 Tr. 91-92, 93, 108, 109, 121, 124, 131-132, 150, 153,
164, 174-181, 192-193, 200-202, 225-227; 2 Tr. 15, 27, 44.125; 4 Tr. 74.
B6/ See note b 5 above, and compare with normal company (white) procedures, 5 Tr. 30, 52, 53-54, 57, 63, 65, 68, 87-88, 91;
6 Tr. 268; 7 Tr. 26-27, 80, 82; and specific white retesting 6 Tr. 6 Tr. 266-267.
B-12
CERTIFICATE OF SERVICE
The undersigned counsel for plaintiffs-appellants
King, Moreman, et al. hereby certifies that on the //
day of January, 1972, he served copies of the foregoing
Brief for Plaintiffs-Appellants upon counsel of record
the other parties as listed below, by placing said
copies in the United States mail, airmail postage pre
paid ,
J. Lewis Sapp, Esq.
1900 Peachtree Center Building
230 Peachtree Street, N.W.
Atlanta, Georgia 30303
Robert L. Mitchell, Esq.
1841 First Nat'l Bank Building Atlanta, Georgia 30303
Stephen Glassman, Esq.
United States Department of Justice Civil Rights Division
Washington, D. C. 20530