Detroit Police Officers Association v. Young Brief Amicus Curiae
Public Court Documents
July 31, 1978
Cite this item
-
Brief Collection, LDF Court Filings. Detroit Police Officers Association v. Young Brief Amicus Curiae, 1978. 494c60ba-af9a-ee11-be37-00224827e97b. LDF Archives, Thurgood Marshall Institute. https://ldfrecollection.org/archives/archives-search/archives-item/b1cedc62-893c-4f0f-9e3f-28eb993f3875/detroit-police-officers-association-v-young-brief-amicus-curiae. Accessed July 17, 2025.
Copied!
*
IN THE UNITED STATES COURT OF APPEALS
FOR THE SIXTH CIRCUIT
receive:
Ho. 78-1163 JUL 3 1197B
_________________ CITY OF DETROIT
ON APPEAL FROM THE UNITED STATES
DISTRICT COURT FCR THE EASTERN DISTRICT OF MICHIGAN
BRIEF FOR THE UNITED STATES AND EQUAL
OPPORTUNITY COMMISSION AS AMICUS
EMPLOYMENT
CURIAE
ABNER W. SIBAL,
General Counsel.
Equal Employment
Opportunity Ccrrjr.ission.
Washington. D.C. 20506.
DREW S. DAYS, III,
Assistant Attorney General.
BRIAN K. LANDSBERG,
ROBERT J. REINSTEIN,
JOHN C. HAMMOCK,
Attorneys.
Department of Justic
Washington, D.C. 20 '-n|
fC
ij. j
*•»
TAELE OF CONTENTS
Interest of the United States and the
Eaual Employment Opportunity Commission ......... 1
Statement ....................................... 2
I. 1943-1967: The Exclusion of Blacks
from the Detroit Police Department
and the Effects of that Exclusion ...... 3
II. 1968-1974: The Police Department’s
Efforts to End Discrimination Against
Blacks in Hiring and Promotions and
to Remedy the Effects of Past Dis
crimination ........................... ^
A. The City's Self Analysis and
Affirmative Action in Hiring,
1968-1974 ......................... 11
3. The City's Self-Analysis and
Affirmative Action in Promo
tions, 1968-1974 .................. 17
1. Dipping ........................ 18
2. Service Rating ................ 19
3. Seniority ..................... 21
4. Written Promotion
Examination ................. 21
5. Oral Boards ................... 27
6. College education and veteran's
preference .................. 28
7. Rank Order .................... 29
8. Summary and results, 1968-
1974 ........................ 32
i
C. The Affirmative Action Pro
motions, 1974-1977 ....... ........ 33
Argument ............................ ............ 38
The Detroit Police Department's Voluntary
Affirmative Action Promotions Are Lawful
Under The Constitution And The Federal
Civil Rights Acts ........................ 38
I. The Federal Civil Rights Acts And
The Constitution Do Not Prohibit,
And Sometimes Require, The Use Of
Race-Conscious Practices To Eliminate
The Effects Of An Employer's Unlawful
Discrimination ....................... 38
II. The Detroit Police Department Practiced
Systematic and Unlawful Past Discrimina
tion Against Blacks In Hiring And
Promotions ...................... , 4 8
A. Past Discrimination in Hiring .... 49
B. Past Discrimination In
Promotions ....................... 56
C. The Prior Discrimination In
Hiring And Promotions Was
Unlawful ......................... 60
III. The Detroit Police Department May
Voluntarily Institute Affirmative
Action To Remedy The Effects Of
Its Past Exclusionary Policy ......... 64
IV. The Particular Race-Conscious Remedy
Adopted By the Detroit Police Depart
ment Did Not Violate The Legal Rights
Of Plaintiffs ..................... 73
A. A Promotion Ratio Is An Appropriate
Method of Remedying Past Discrimina
tion ............................. 73
3. A Promotion Ratio Is Appropriate
To Eliminate The Effects of Past
Discrimination Even If The Depart
ment's Recent Promotional Models
Have Been Validated .............. 75
ii
C. The Particular Promotion Ratio
(50 Percent) Is Appropriate,
But An Ultimate Goal Must Be
Set on Remand ....................
Conclusion ...................................... 90
CITATIONS
Cases:
Afro-American Patrolmen's League v. Duck,
503 F .2d 294 (6th Cir. 1974), affirming
366 F. Supp. 1095 (N.D. Ohio 1973) ........... . 53,57,60
Albemarle Paper Co. v. Moody, 422 U.S. 405
(1975) ..................... 42,66
Alexander v. Gardner-Denver Co., 415 U.S.
36 (1974) ..................................... 65
Arnold v. Ballard, 12 FEP Cases 1613
(6th Cir. 1976), vacated, 16 FEP Cases
396 (6th Cir. 1976) ....................... ••• 44,76
Associated GeneralContractors of Mass., Inc,
v. Altshuler, 490 F.2d 9 (1st Cir. 1973),
certiorari denied, 416 U.S. 957 (1974) ........ 69,88
Barnett v. International Harvester, 12 FEP
Cases 786 (W.D. Tenn. 1976) ............ -...... 65
Baumgartner v. United States, 322 U.S. 665
(1944) ......................................... 48
Bolden v. Pennsylvania State Police, 73 F.R.D.
370 (E.D. Pa. 1976) ........................... 87
Boston Chapter, NAACP v. Beecher, 504 F.2d
1017 (1st Cir. 1974), certiorari denied,
421 U.S. 910 (1975) ................. .......... 44
3rown v. Board of Education, 349 U.S. 294
(1955) ................................. ....... 68
iii
Cases (continued):
Carter v. Gallagher, 452 F.2d 315
(8th Cir. 1971), certiorari
denied, 406 U.S. 950 (1972) ................... 42,45,61,75
Castaneda v. Partida, 430 U.S. 482
(1977) .................................... 48,49,52
Castro v. Beecher, 459 F.2d 725
(1st Cir. 1972), reversing 334
F. Supp. 930 (D. Mass. 1971) .................. 54
Causey v. Ford Motor Co., 516 F.2d 416
(5th Cir. 1975) ............................ . 48
Chmill v. City of Pittsburgh, 31 Pa. Cmwlth.
98, 375 A.2d 841 (1977) ....................... 66
Civil Rights Cases, 109 U.S. 3 (1883) ........... 62
Crockett v. Green, 534 F.2d 715 (7th
Cir. 1976) .................................... 45
Davis v. County of Los Angeles, 566 F.2d
1334 (9th Cir. 1977), certiorari granted,
No. 77-1553, June 19, 1978 .................... 42,45,62
Dayton Board of Educ. v. Brinkman,
433 U.S. 406 (1977) ........................... 69
Detroit Firefighters Assfn v. City of
Detroit, 17 FEP Cases 186 (E.D.
Mich. 1976), vacated, 17 FEP Cases
190 (6th Cir. 1978) ...................... . 66
Detroit Police Officers Ass'n v. Young,
446 F. Supp. 979 (E.D. Mich. 1978) ............ Passim
Dobbins v. Local 212, I.B.E.W.,
292 F » Supp. 413 (S.D. Ohio 1978) ............ 77
Dothard v. Raw1inson, 433 U.S. 321 (1977) ...... . 54
iv
Cases (continued):
EEOC v. A.T. & T. Co., 556 F.2d 167
(3d Cir. 1977), certiorari denied,
Nos. 77-241-243, July 3, 1978 ....... .......... 44,74,75,87
EEOC v. Contour Chair Lounge, 17 FEP Cases
309 (E.D. Mo. 1978) .......................... . 65
Erie Human Relations Comm, v. Tullio,
493 F .2d 371 (3d Cir. 1974), affirming
357 F. Supp. 422 (W.D. Pa. 1973) .............. 54
Firefighters Institute for Racial Equality
v. City of St. Louis, 549 F.2d 506
(8th Cir. 1977), certiorari denied,
434 U.S. 819 (1977) ........................... 85,86
Franks v. 3owman Transp. Co., 424 U.S.
747 (1976) .................................... 42,46,69,75
Furnco Const. Co. v. Waters, No. 77-369
(U.S. June 29, 1978) ............... ........... 41,70
Germann v. Kipp, 429 F. Supp. 1323
(W.D. Mo. 1977), vacated, 17 FEP Cases
72 (8th Cir. 1978) ............................ 65
Green v. Countv School Board, 391 U.S. 430
(1968) ...................................... .. 41,67,71
Griggs v. Duke Power Co., 401 U.S. 424
(1971) ....................................... . 39,62
Hazelwood School Dist. v. United States,
433 U.S. 299 (1977) .......................... . 49,51,52,61
Hills v. Gautreaux, 425 U.S. 284 (1976) ......... 42
Hutchinson Commission v. Midland Credit
Management, Inc., 213 Kan. 399, 517 P.2d
158 (1973) ..... .............................. . 39,65
v
Cases (continued):
Int'l Brotherhood of Teamsters v. United
States, 431 U.S. 324 (1977) ................... 49,51,56,69
Jackson v. Dukakis, 526 F.2d 64
(1st Cir. 1975), affirming 394
F. Supp. 162 (D. Mass. 1975) ......... ......... 53
Johnson v. Railway Express Agency, Inc.,
421 U.S. 454 (1975) ........................... 61
Johnson v. Ryder Truck Lines, 17 FEP Cases
571 (4th Cir. 1978) ........................... 62
Jones v. Alfred Mayer Co., 392 U.S. 409
(1968) ........................................ 61,62
Kirkland v. Dept, of Correctional Services,
520 F .2d 420 (2d Cir. 1975), certiorari
denied, 429 U.S. 823 (1976) ................... 44,74,75,85
Lindsay v. City of Seattle, 86 Wash. 2d 698,
548 P.2d 320 (1976), certiorari denied,
429 U.S. 886 (1976) ......................... . 39,65
Long v. Ford Motor Co., 496 F.2d 500
(6th Cir. 1974) ...................... .. ....... 61
Louisiana v. United States, 380 U.S. 145
(1965) ......................................... 41,76
McDaniel v. 5arresi, 402 U.S. 39 (1971) ......... 41,69,71
McDonald v. Santa Fe Trail Trans. Co.,
427 U.S. 273 (1976) ........................... 40
McDonnell-Douglas Corp. v. Green, 411 U.S.
792 (1973) .................................... 41,70
Milliken v. Bradley, 433 U.S. 267 (1977) ....... 68
Morgan v. Kerrigan, 509 F„2d 599 (1st
Ciro 1975) ........................... ......... 44
vi
Cases (continued):
Morrow v. Crisler, 491 F.2d 1053 (5th Cir.
1974), certiorari denied, 419 U.S. 895
(1974) ....................................... 44,45
NAACP v. Allen. 493 F.2d 614 (5th Cir.
1974) ......................................... 44
Norris v. Alabama, 294 U.S. 587 (1935) ......... 48
North Carolina Bd. of Educ. v. Swann,
402 U.S. 43 (1971) ............................ 46
Officers for Justice v. Civil Service Comm.,
371 F. Supp. 1328 (N.D. Cal. 1973) ............ 87
Fapermakers and Paperworkers, Local 189 v.
United States, 416 F.2d 980 (5th Cir.
1969) , certiorari denied, 397 U.S. 919
(1970) ........................................ 74
Patterson v. American Tobacco Co., 535 F.2d
257 (4th Cir. 1976), certiorari denied,
429 U.S. 920 (1976) ............... ............ 44
Patterson v. Newspaper Deliverers* Union,
514 F .2d 767 (2d Cir. 1975), certiorari
denied, 427 U.S. 911 (1976) .................. 44,75,87
Pomoey v. General Motors Corp. 385 Mich.
537, 189 N.W. 2d 243 (1971) ................ .... 61
Porcelli v. Titus, 431 F.2d 1254 (3d Cir.
1970) , certiorari denied, 402 U.S.
944 (1971) .................................... 65,69
R,eeves v. Eaves, 411 F. Supp. 531 (N.D. Ga.
1976) 66
Reed v. Lucas, 11 FEP Cases 153 (E.D.
Mich. 1975) ........................ ........... 87
vii
Cases (continued):
Regents of the Univ. of Calif, v. Sakke,
No. 76-811 (U.S. June 28, 1978) ............... 45,46,47,74
Rios v. Steamfitters, Local 638. 501 F.2d
622 (2d Cir. 1974) .......................... . 44,88
Robinson v. Union Carbide Corn., 538 F.2d
652 (5th Cir. 1976) ........................... 58
Schaeffer v. Tannian, 394 F. Supp. 1128
(E.D. Mich. 1974) ............................. 29,34
Senter v . General Motors Corn., 532 F.2d
511 (6th Cir. 1976), certiorari denied,
429 U.S. 870 (1976) ... ......................... 49
Sims v. Sheet Metal Workers Local 65.
489 F.2d 1023 (6th Cir. 1973) ................. 44
Stamps V. Detroit Edison Co.. 365 F. Supp.
87 (E.D. Mich. 1973), affirmed sub nom.
ELCC v. Detroit Edison Co.. 515 F.2d 301
(6th Cir. 1975), vacated, 431 U.S. 951 (1977) .. Passim
Stewart v. General Motors Corp., 542 F.2d
445 (7th Cir. 1976), certiorari denied,
433 U.S. 919 (1977) ........................... 58
Swann v. Charlotte-Mecklenburg Bd. of
Edue.. 402 U.S. 1 (1971) ...................... 41,42,46,68,69,71
United Jewish Organizations of Williamsburg
v. Carev. 430 U.S. 144 (1977) ................. 69,71
United States v. Alleghenv-Ludlum Industries.
517 F .2d 826 (5th Cir. 1975), certiorari
denied, 425 U.S. 944 (1976) ................... 65
United States v. City of Chicago. 549 F.2d
415 (7th Cir. 1977), certiorari denied,
434 U.S. 875 (1977) ........................... 44,74
United States v. City of Chicago, 573 F.2d
416 (7th Cir. 1978) ........................... 85,86,87
vii i
Cases (continued):
United States v. Duke, 332 F.2d 759 (5th
Cir. 1964) .................................... 76
United States v. Elevator Constructors,
Local 5, 538 F.2d 1012 (3d Cir. 1976) ......... 44
United States v. I.B.E.W.. Local 38,
428 F .2d 144 (6th Cir. 1970),
certiorari denied, 400 U.S. 943 (1970) ........ 43,44,45
United States v. Ironworkers, Local 86,
443 F.2d 544 (9th Cir. 1971), certiorari
denied, 404 U.S. 984 (1971) .................. . 45
United States v. Jefferson County Bd.
of Educ., 372 F .2d 836 (5th Cir.
1966) , opinion adopted by court,
380 F.2d 385 (5th Cir.
1967) , certiorari denied, 389 U.S.
840 (1967) ..................................... 42
United States v. Lathers, Local 46,
471 F .2d 408 (2d Cir. 1973),
certiorari denied, 412 U.S. 939
(1973) ......................................... 73
United States v. Local 1, Bricklayers,
5 E.P.D. Para. 8480 (W.D. Tenn. 1972-73),
affirmed, 497 F.2d 871 (6th Cir. 1974) ........ 77
United States v. Local 169, Carpenters,
457 F.2d 210 (7th Cir. 1972),
certiorari denied, 409 U.S. 851 (1972) ........ 45
United States v. Local 212, I.B.E.W.,
5 E.P.D. para. 8428 (S.D. Ohio 1972),
affirmed, 472 F.2d 634 (6th Cir. 1973) ........ 77
United States v. Local 212. I.B.E.W.,
472 F.2d 634 (6th Cir. 1973) .................. 43,77
ix
Cases (continued):
United States v. Masonry Contractors Ass'n,
Inc., 497 F.2d 871 (6th Cir. 1974) ............. 44,77 '
United States v. N.L. Industries. Inc.. 479 F.2d
354 (8th Cir. 1973) ............................ 45,66,74,87
United States v. Solomon, 563 F.2d 1121
(4th Cir. 1977) ................................ 61
United States v. United States Gypsum Co.,
333 U.S. 364 (1948) ............................ 49
United States v. United States Steel Corn.,
5 FEP Cases 1253 (N.D. Ala. 1973), affirmed;
520 F .2d 1043 (5th Cir. 1975), certiorari
denied, 429 U.S. 817 (1976) .................... 87
Vulcan Society v. New York Civil Service
Comm,, 360 F. Supp. 1265 (S.D. N.Y. 1973),
affirmed 490 F.2d 387 (2d Cir. 1973) ........... 85,86
Washington v. Davis, 426 U.S. 229 (1976) ......... 40,61
Waters v . Wisconsin Steel Works Co.,
427 F .2d 476 (7th Cir. 1970),
certiorari denied, 400 U.S. 911
(1970) .......................................... 61
Watkins v. Scott Paper Co., 530 F.2d
1159 (5th Cir. 1976), certiorari denied,
429 U.S. 861 (1976) ............................ 44,45,58,74
Weber v. Kaiser Aluminum & Chemical
Corp., 563 F.2d 216 (5th Cir.
1977) . 47,65,67
Young v. Int’l Telephone & Telegraph Co.,
438 F .2d 757 (3d Cir. 1971) .....................61
x
Constitution, statutes and regulations:
Constitution of the United States of America,
Fourteenth Amendment ..................... 39,42,61
Civil Rights Act of 1964, Title VI, 78 Stat.
252 et sec., 42 U.S.C. 20Q0d et seq. ..... 39,42,61
Section 602, 42 U.S.C. 2000d-l ........ 65
Section 604, 42 U.S.C. 2000d-3 ........ 61
Civil Rights Act of 1964, Title VII, 78 Stat
253 et sec., 42 U.S.C. 2000e et seq. , ....Passim
Section 706(b), 42 U.S.C. 2000e-5(b) ... 65
Section 706(f)(1), 42 U.S.C. 2000e-5
(f)(1) .............................. 1»65
42 U.S.C. 1931 ...............................Passim
42 U.S.C. 1983 ............................. 70,71
Michigan Civil Rights Act (1977),
Section 37.2101 et seq. ................63
Michigan Fair Employment Practices Act,
M.C.L.A. (1972):
Section 423.301 et: seq................. 61
Section 423.302(b) ................... 61
29 C.F.R. 1607.5(a) ........................85,86
45 C.F.R. 80.3(b)(6)(i) ....................46
45 C.F.R. 80.3(b)(6)(ii) ...................46
45 C.F.R. 80.5(i) ..........................46
Miscellaneous:
American Psychological Association,
STANDARDS FOR EDUCATIONAL AND
PSYCHOLOGICAL TESTS (1964) ............... 85,86
ENA Labor Relations Reporter, Fair
Employment Practices Maunel 455:
1091-1095 ................................ 63
xi
Miscellaneous (continued):
Edwards, THE POLICE ON THE URBAN FRONTIER
(1968) ........................ ........... 8,9,17
Edwards. Order & Civil Liberties: A
Complex Role for the Police, 64 Mich.
L. Rev. 47 (1965) ........................ 3,18
W. Harrison, THE DETROIT RACE RIOTS (1943) .. 4
Policy Statement on Affirmative Action
Programs for State and Local Govern
ment Agencies, 41 Fed. Reg. 38814
(1976) ................................... 66
President's Commission on Law Enforcement
and Administration of Justice, THE
CHALLENGE OF CRIME IN A FREE SOCIETY
(1967) ................................... 7,8
President's Commission on Law Enforcement
and Administration of Justice, TASK
FORCE REPORT: THE POLICE (1967) ......... 4,5,6,8
Proposed EEOC Guidelines on Remedial
And/or Affirmative Action, 42 Fed.
Reg. 64826 (1977) .......... •............. 47,66
REPORT OF THE NATIONAL ADVISORY COMMISSION
ON CIVIL DISORDERS (1968) ................ 4,7,8,9
R. Shogan & T. Craig, THE DETROIT RACE
RIOT (1964) .............................. 3,4
U.S. Bureau of the Census, 1940 CENSUS
OF POPULATION, Vol. II, Part 3 ........... 4
U.S. Civil Rights Commission, ADMINISTRATION
OF JUSTICE STAFF REPORT (1963) ........... 6,11
U.S. Civil RightsCommission, HEARINGS
BEFORE THE UNITED STATES COMMISSION ON
CIVIL RIGHTS, HEARINGS HELD IN DETROIT,
MIGHICAN DECEMBER 14, 15, 1960 ........... 18
xii
Miscellaneous (continued):
U.S. Dept, of Commerce, Bureau of the
Census, 1950 CENSUS OF POPULATION,
Vol. II, Part 22 ....................... .. 4
U.S. Dept, of Commerce, Bureau of the
Census, 1960 CENSUS OF POPULATION,
Vol. I, Part 24 .......................... 5
U.S. Dept, of Commerce, Bureau of the
Census, 1970 CENSUS OF POPULATION ........ 6
W. White 6c T. Marshall, WHAT CAUSED THE
DETROIT RIOT? (1943) .................... 4
xiii
IN THE UNITED STATES COURT OF APPEALS
FOR THE SIXTH CIRCUIT
No. 78-1163
DETROIT POLICE OFFICERS ASSOCIATION, et al.,
Plaintiffs-Appellees
v.
COLEMAN A. YOUNG, et al. ,
Defendants-Appellants
ON APPEAL FROM THE UNITED STATES
DISTRICT COURT FOR THE EASTERN DISTRICT OF MICHIGAN
BRIEF FOR THE UNITED STATES AND EQUAL EMPLOYMENT
OPPORTUNITY COMMISSION AS AMICUS CURIAE
-1-
INTEREST OF THE UNITED STATES
AND THE EQUAL EMPLOYMENT OPPORTUNITY COMMISSION
Federal enforcement of Title VII of the Civil Rights Act
of 1964 has been vested by Congress in the Equal Employment
Opportunity Commission (EEOC) and the Department of Justice.
The EEOC has the responsibility to investigate charges of
employment discrimination, to attempt voluntary conciliation
and, if necessary, to bring civil actions against private
employers under 42 U.S.C. (Supp. V) 2000e—5(f)(1). The
Attorney General has enforcement responsibility when the
employer is a state government, governmental agency, or
political subdivision. 42 U.S.C. (and Supp. V) 2000e-6. In
discharging these federal enforcement responsibilities, the
Government relies, in substantial part, on encouraging
employers to self-evaluate their employment practices and to
voluntarily adopt affirmative, corrective action to increase
the participation of minorities in jobs that were formerly
closed to them.
This case raises important questions concerning the
legality of voluntary affirmative action by a public employer
that sought to rectify the exclusionary effects of its past
racial discrimination. The resolution of these questions
may significantly impact on federal enforcement of Title VII.
The United States and Equal Employment Opportunity Commission
file this brief as amicus curiae in support of reversal of
the district court's decision.
-2-
S TATEM ENT_1 /
This lawsuit challenges a voluntary affirmative action
program instituted by the Detroit Police Department in 1974.
In order to remedy the effects of its own past discrimination
against blacks, the department decided to accelerate the pro
motions of qualified black police officers to the rank of
sergeant. When this affirmative action plan began in 1974,
only 61 of the department's 1185 sergeants (5.1 percent) were
black, although the City of Detroit was approximately 50 per
cent black and the department as a whole was 17.2 percent
black. 2/ Between 1974 and 1977, 160 whites and 152 blacks
were promoted to sergeant; this result was achieved by
departing from strict rank order on the eligibility lists so
that promotions of police officers would be roughly in a
ratio of one qualified black for each qualified white. 3/ By
the end of 1977, the sergeants' rank was 15.1 percent
black. 4/
The district court's opinion accurately describes the
procedures which were followed in promoting sergeants between
1/ This appeal is proceeding on a deferred appendix.
References are to the transcript ("Tr.", followed by the date
and page of testimony and identification of witness), Plain
tiffs' Exhibits ("PX"), Defendants' Exhibits ("DX"), and the
district court's opinion, reported at 446 F. Supp. 979 (E.D.
Mich. 1978).
2/ 446 F. Supp. at 994 n. 29, 995-996.
_3/ Id., 987-989.
4/ DX 264.
-3-
1974 and 1977. 5/ However, the opinion does not relate
accurately the historical facts and reasons which impelled
the police department to begin promoting on a race-conscious
basis. The district court disregarded almost entirely the
massive and largely undisputed evidence of past, systematic
discrimination against blacks in hiring and promotions; the
extensive self-analysis undertaken by the department over a
period of years; and the failure of less drastic remedies to
eliminate the continuing effects of the department's past
exclusionary policies. Because of the district court's
failure to address fully the evidence, and in light of the
size of this record, it is necessary to set forth the facts
at unusual length.
I. 1943-1967: The Exclusion Of Blacks From The
Detroit Police Department And The Effects of
_________ That Exclusion_____________________
In 1943, when the first Detroit race riot occurred, the
city's police department was virtually all white and
segregated. 6/ There was "open warfare between the Detroit
5/ 446 F. Supp. at 986-988.
6/ See generally Robert Shogan & Tom Craig, THE DETROIT
RACE RIOT (1964); Walter White & Thurgood Marshall, WHAT
CAUSED THE DETROIT RIOT? (1943). Writing in 1965, Judge
George Edwards reviewed the 1943 Detroit riot and subsequent
riots in other cities, and concluded that "hostility between
the Negro communities in our large cities and the police
departments is the major problem in law enforcement in this
decade. It has been a major cause of all recent race riots."
Edwards, Order and Civil Liberties: A Complex Role for the
Police, 64 Mich. L. Rev. 47, 54-55 (1965).
(continued)
-4-
Negroes and the Detroit Police Department." 7/ In the wake
of the 1943 riot, responsible observers emphasized the
necessity of integrating the Police Department. 8/ But this
did not occur. Over the next decade, the black population of
Detroit rose beyond 16 percent; 9/ between 1944 and 1953, the
Department hired 3,122 police officers— 3,005 whites and 117
blacks (3.7 percent). 10/ In 1953, the Department was only
2.4 percent black. 11/ The small number of black officers in
6/ (Continued)
According to former Commissioner Nichols, there was a
"minimal number" of blacks in the Detroit Police Department
in 1942. Tr. 8/10/77, 41. At the time of the 1943 riot, the
Department had only 4 3 black officers out of- a total complement of
3400, R. Shogan & T. Craig, supra at 42, 70. To control the riot,
the Police Department activated a "Negro auxiliary", a volunteer
group of some 200 untrained and unarmed black citizens. Id., at
72, 109.
7/ REPORT OF THE NATIONAL ADVISORY COMMISSION ON CIVIL
DISORDERS 48 (1968) (hereinafter cited as "KERNER COMMISSION
REPORT").
8/ See, e.g., Walter White & Thurgood Marshall, supra note
6 at 17"(We recommend that the number of Negro officers be
increased from 43 to 350; that there be immediate promotions
of Negro officers in uniform to positions of responsibility
* * * *"); Robert Shogan & Tom Craig, supra note 6 at 116; Walter R.
Harrison, THE DETROIT RACE RIOTS 11-12, 16 (1943).
9/ U. S. Department of Commerce, Bureau of the Census, 1950
CENSUS OF POPULATION, Vol. II, part 22, Table 33 (reporting
that minorities constituted 16.4 percent of Detroit's
population in 1950). The Detroit SMSA (Wayne, Oakland and
Macomb Counties) was 12.0 percent minority in 1950. Ibid.
Ten years earlier, the City of Detroit was 9.2 percent black;
and the Detroit Metropolitan District was 7.4 percent black.
U.S. Bureau of the Census, 1940 CENSUS OF POPULATION, Vol.
II, Part 3 (Michigan), Tables 34 & A-45.
10/ DX 208, p. 5.
11/ The President's Commission on Law Enforcement and
Administration of Justice, TASK FORCE REPORT: THE POLICE 172
(1967) .
-5-
the Department were given segregated job assignments. 12/ And
the exclusion of blacks from the supervisory ranks was even
more severe; 0.7 percent of all officers at the rank of
sergeant or above were black. 13/
The same pattern continued for another decade. Between
1954 and 1962, the Department hired 1,663 white and 68 black
(3.9 percent) police officers. 14/ By the end of 1962, the
Department was only 3.5 percent black; the supervisory ranks
(sergeant and above) were only 1.1 percent black; 15/ while
the City of Detroit was over 29 percent black. 16/ A 1962
United States Civil Rights Commission study found that blacks
12/ As late as 1962, the Police Department admitted to the
U. S. Civil Rights Commission that assignments were made on a
racial basis. Id., at 174. Numerous witnesses who were on
the police force in the 1940's, 1950's and 1960's testified
that racial segregation in job assignments was a reality until
the early 1960's. See, e.g., Tr. 8/10/77, 55-58 (former Commis
sioner Nichols); 11/4/77, 23-29, 50-51 (Executive Deputy Chief
Bannon); 11/2/77, 9-14 (Chief Hart); 11/15/77, 7-19 (Officer
Baldwin; id., 40-46 (former Sgt. Stewart).
13/ There were 344 white and 3 black sergeants, 167 white
and 1 black lieutenant, and 42 whites and no black above the
rank of lieutenant in 1953. TASK FORCE REPORT: THE POLICE,
supra note 11 at 172.
14/ DX 208, p. 5.
15/ TASK FORCE REPORT: THE POLICE, supra note 11 at 172.
(There were 340 white and 5 black sergeants, 152 white and 1
black lieutenants, and 56 whites and no black above the rank
of lieutenant) .
16/ U. S. Dept, of Commerce, Bureau of the Census, 1960
CENSUS OF POPULATION, Vol. I, Part 24, Table 21. The Detroit
SMSA was 14.9 percent black by 1960. Ibid.
-6-
were excluded from the Detroit Police Department because of
both intentional discrimination and the effect of the written
entrance examination.17/ Formal racial segregation in job
assignments continued until the early 1960's.18/
In February of 1967, the President's Commission on Law
Enforcement and Administration of Justice concluded that
minorities were grossly under-represented in the police
departments of most large cities and that discrimination
against minority applicants was prevalent, singling out
Detroit as a case in point.19/ The population of Detroit was
approaching 44 percent black,20/ but there were only 214
blacks on the 4,356 member police department (4.9 percent
black).21/ The President's Commission also found that
"[t]here is an even more marked disproportion of minority
group supervisory personnel than of minority group officers
generally throughout the police service" of major cities,_22/
and that "there is evidence that discrimination is practiced
against minority group officers, perhaps more in promotion
17/ U.S. Civil Rights Commission, ADMINISTRATION CF JUSTICE
STAFF REPORT ch. 11, 8-16 (1963), summarized in TASK FORCE
REPORT: THE POLICE, supra note 11 at 169.
18/ See note 12, supra.
19/ TASK FORCE REPORT: THE POLICE, supra note 11 at
167-171.
20/ U.S. Dept, of Commerce, Bureau of Census, 1970 CENSUS OF
POPULATION, Vol. I, Part 24, Table 23. The Detroit SMSA was
18.0 percent black in 1970. Ibid.
21/ DX 208, pp. 3, 5 (data for end of 1967).
22/ TASK FORCE REPORT: THE POLICE supra note 11 at 171.
-7-
than in recruitment."23/ The supervisory ranks of the
Detroit Police Department were then approximately 2.1 percent
black.24/
The President's Commission concluded that the exclusion
of minorities from police departments had created a crisis in
law enforcement. The relations between the police and
minority communities "is as serious as any problem the police
have today."25/ These relations could not be improved until
there is ”a sufficient number of minority-group officers at
all levels of activity and authority."26/
"Police departments in all communities
with a substantial minority population
must vigorously recruit minority group
officers. The very presence of a pre
dominantly white police force in a
Negro community can serve as a
dangerous irritant * * *. In
neighborhoods filled with people
suffering from a sense of injustice
and exclusion, many residents will
reach the conclusion that the
neighborhood is being policed not for
the purpose of maintaining law and
order but for the purpose of
maintaining the status quo.
23/ Id., at 172.
24/ KERNER COMMISSION REPORT, supra note 7 at 169 Table A
(data for October, 1967).
25/ The President's Commission on Law Enforcement and
Administration of Justice, THE CHALLENGE OF CRIME IN A FREE
SOCIETY 99 (February, 1967).
26/ Id., at 101.
I
"In order to gain the general
confidence and acceptance of a
community, personnel within a police
department should be representative of
the community as a whole. * * *"27/
The President's Commission set as a "high priority objective"
for all police departments "in communities with a substantial
minority population to recruit minority-group officers, and
to deploy and promote them fairly."28/ And in the area of
promotions, "more than nondiscrimination" was necessary "to
overcome the legacy of the past."29/ The Commission
advocated a number of race-conscious remedies to meet the
"urgent need for high-ranking minority officers," including
the deliberate "exercise of * * * discretion" to promote
minorities.30/ And finally, emphasizing the immediacy of the
problem, the Commission warned that "everyday police
encounters in [minority] neighborhoods can set off riots."31/
On the evening of July 22, 1967, "everyday police
encounters" in black areas of Detroit triggered the second
Detroit race riot.32/ During the next five days, 43 people
-8-
27/ TASK FORCE REPORT, supra note 11 at 167. See also G.
Edwards, THE POLICE ON THE URBAN FRONTIER 86 (1968).
28/ THE CHALLENGE OF CRIME IN A FREE SOCIETY, supra note 25
at 102. See also G. Edwards, supra note 27 at 40, 86.
29/ TASK FORCE REPORT, supra note 11 at 173.
30/ Ibid.
31/ THE CHALLENGE OF CRIME IN A FREE SOCIETY, supra note 25
at 99.
32/ KERNER COMMISSION REPORT, supra note 7 at 47-59.
-9-
were killed, hundreds were injured, thousands were arrested,
large portions of Detroit were devastated, and the army
occupied the city.33/ A principal cause of this and the
other riots during that summer was the hostility between the
police and the black community— a hostility caused in large
measure by the exclusion of blacks from all ranks of the
police.34/
II. 1968-1974: The Police Department's Efforts
to End Discrimination Against Blacks in Hiring
and Promotions and to Remedy the Effects of
____________Past Discrimination________________
The second Detroit race riot had a traumatic effect on
the City and its police force. There was extreme distrust
between the black community and an "essentially * * * white"
police department.35/ By the end of 1967, the department was
only 4.9 percent black.36/ At the supervisory ranks, blacks
accounted for only 2.6 percent of the sergeants (9 of 348),
1.3 percent of the lieutenants (2 of 158), and 1.6 percent of
the officers above the rank of lieutenant (1 of 63).31/ The
Mayor stated that it had finally become "obvious to me and to
33/ _Id_!_ at 60-61.
34/ Id. at 165-169. See also note 6 supra, and G. Edwards,
supra note 27 at 35 ("[T]he conflict between Negroes and the
police actually is a conflict between two of the most
segregated groups in American society.").
35/ Tr. 8/8/77, 68, 87 (Johannes Spreen, Police Commissioner
between 1968 and 1970).
36/ DX 208, pp. 3, 5 (214 blacks of 4,356).
37/ KERNER COMMISSION REPORT, supra note 7 at 169 Table A.
-10-
this entire community, that this proportion of Negro
policemen was clearly unacceptable."38/ The Department
understood that its "biggest problem to try to overcome" was
the polarization between itself and the black community39/
and that it was essential to rectify the under-representation
of blacks at all ranks.40/ The Police Department knew that
blacks had been "excluded" during all of the years until 1968
and that "there was some form of discrimination effectively
working during those years."41/The City then began to
undertake actions in both the hiring and promotion areas,
which span a six year period (1968-1974) and are recounted
below. The City's initial objective was to determine the
reasons why blacks were so grossly under-represented in the
department. It then implemented a series of steps to
eliminate existing racially discriminatory barriers and to
rectify the effects of past discrimination.
38/ PX 106, p. 17 (statement of Mayor Cavanagh).
39/ Tr. 8/8/77, 87 (former Commissioner Spreen).
40/ Id., 76, 89-91. Commissioner Spreen's objective was
that the Department should be racially reflective of the
City. ("I think any enlightened Chief of Police or
Commissioner would prefer that, if he could have that." Id.,
89-90). See also Tr. 8/9/77, 21-22 (former Commissioner
Nichols).
41/ Tr. 8/24/77, 22; 8/25/77, 63 (Commander Caretti,
director of recruiting). Johannes Spreen, who was Police
Commissioner in 1968, testified that "I made inquiry as to
what possibly could have been the cause [of so few blacks in
the department]. One of the causes was an exclusion policy
and I made certain that there was none when I was the
Commissioner." Tr. 8/8/77, 77. But cf. _id., 79.
-11-
A. The City's Self Analysis and Affirmative
Action in Hiring, 1968-1974__________
In May, 1968, Mayor Cavanagh announced the beginning of
an intensive campaign to recruit black applicants for the
Police Department.42/ The Mayor also appointed a Special
Task Force to determine why blacks were under-represented in
police hiring. The Task Force found initially that 47
percent of police applicants in 1967 were black, a figure
greater than the black proportion of Detroit's population, a
finding which "exploded” the "myth" that blacks "do not want,
nor are they willing, to join the Police Department."43/ The
Task Force also found that black applicants were disporpor-
tionately rejected: (1) at the preliminary application stage
for "'miscellaneous' reasons"; (2) at the written examination
stage; and (3) at the background investigation and oral inter
view stages "where subjective opinions are critical."44/
Mayor Cavanagh also appointed a committee of seven
prominent industrial psychologists and personnel selection
experts (the "Vickery Committee") to examine the present
selection standards and to make recommendations for the
development of non-discriminatory and valid standards.45/
42/ PX 106, p. 20. The recruitment efforts are fully
described in the district court opinion, 446 F. Supp. at
997-998.
43/ PX 106, p. 45.
44/ Ibid. Compare the 1962 U.S. Civil Rights Commission
staff study, supra note 17.
45/ PX 106, p. 3; Tr. 8/11/77 24, 36 (Commander Caretti, who
worked with the Vickery Committee).
-12-
The Vickery Committee found that many of the qualifications
at the preliminary screening stage were not relevant and
recommended changes.46/ It also found that the entry-level
test used up to that time (a three-hour IQ test) was "very
bias[ed]" against blacks,47/ and a validation study showed
that there was no relationship between scores on the test and
job performance.48/ The Vickery Committee recommended that
the 12-minute Wonderlic IQ test should be used as an interim
measure until a validated test could be obtained.49/ The
Committee knew that the Wonderlic test discriminated against
blacks and that there was a body of knowledge repudiating
this test as a predictor of job performance,50/ but it hoped
that as an interim measure the Wonderlic would have a lesser
adverse impact against blacks.51/ The changes recommended by
the Vickery Committee were instituted in 1968,52/ and the
46/ PX 106, pp. 3-6.
47/ Tr. 8/11/77, 24 (Caretti). Three times as many blacks
failed this test as whites (DX 294, pp. 14-15).
48/ PX 106, pp. 6-7. This result was not surprising. As
Caretti stated, "no one has been able to establish a relation
ship between IQ and [job] performance" (Tr. 8/11/77,46).
49/ Tr. 8/11/77, 24-25 (Caretti).
50/ Tr. 8/24/77, 25-28 (Caretti).
51/ Id., 28. Caretti explained that "it's important to
understand that the Vickery Committee did not want to come in
and move with a hatchet, so to speak, to try and correct and
change the system. They were cautious. * * * [T]here was no
opinion in the Vickery Committee that would support Wunder
lich [sic]. It was strictly an interim measure to try and
improve the process." Ibid.
52/ Tr. 8/18/77, 15 (Caretti); Tr. 8/8/77, 83 (Spreen).
-13-
applications of blacks were processed on an expedited
basis.53/ Between 1968 and 1970, the hiring rate of blacks
increased to 26% (408 out of 1575),54/ and the Police Depart
ment was 11.0% black by the end of 1970.55/ However, black
applicants were still being rejected at a highly dispropor
tionate rate, because about 50% of the applicants between
1968 and 1970 were black.56/ Blacks failed the entrance-
level IQ tests at a much greater rate than whites.57/ This
was also true of the medical examination and background
investigation.58/
In 1971, as a result of another recommendation from the
Vickery Committee, the "Chicago battery" test was
53/ Tr. 8/8/77, 125 (Spreen).
.54/ DX 208, p. 4.
55/ DX 208, pp. 3, 5 (568 blacks of 5158).
56/ Tr. 8/8/77, 81-84 (Spreen); 10/31/77, 27 (Sgt.
Broadnax). The Police Community Relations Project Committee
reported in 1970 that 47 percent of the applicants in 1969
were black (DX 294, p. 11). This committee included civic
and business leaders, a federal and a state judge, high
ranking police officers, and the president and vice-president
of Plaintiff-Appellee Detroit Police Officers Association
(^d., pp. 9-10). The committee stated: "The fact that 47 percent
of all police applicants in 1969 were black shatters the myth
that the black community is less interested in law
enforcement and destroys a conclusion, held by some, that
blacks are less interested in a police career." (Id., p.
11) .
57/ On the Wonderlic IQ test, used in 1968, only 30% of
black applicants passed, compared to 70-80% of white appli
cants. Tr. 8/24/77, 24-27 (Caretti); in 1969 and 1970,
approximately 40% of black applicants and approximately 80%
of white applicants passed the IQ tests (PX 190, p. 42).
58/ Tr. 8/24/77, 35-38 (Caretti).
-14-
ins tituted. 59/ This battery included both job-related and IQ
components, 60/ and was used until 1973. 61/ This reduced but
did not eliminate the disparity between the passing rates of
whites and blacks. 62/ The Police Department also critically
examined the background investigation and medical procedures
because blacks continued to be rejected at higher rates than
whites. 63/ The background investigators could choose whom
they wanted to investigate, 64/ sergeants could reject appli
cants on the recommendation of investigators even before the
investigation was completed, 65/ and the time for investigat
ing blacks was substantially longer than whites with similar
backgrounds. 66/ These practices were ended, and the racial
disparity in rejection rates was substantially reduced.67/
The major problem at the medical stage was that 80
59/ PX 190, p. 42; Tr. 8/11/77, 45 (Caretti).
60/ Tr. 10/27/77, 15-16 (Commander Ferrebee); Tr. 8/19/77, 4
Tcaretti).
61/ Tr. 8/19/77, 5 (Caretti).
62/ PX 190, p. 42 (between 1971 and 1973, passing rates on
"Chicago battery" were 59.2% for blacks and 82.5% for whites).
6_3/ Tr. 8/24/77, 35-38, 42-43 (Caretti); Tr. 10/13/77, 26-29 ,
6 8 (Broadnax) .
64/ Tr. 10/31/77, 35 (Broadnax); Tr. 10/26/77, 33 (Ferrebee).
65/ Tr. 10/26/77, 44-46 (Ferrebee).
66/ Id., 31-32, 98-100.
67/ Investigations were assigned on a "blind draw" basis_(Tr.
To/ 31/77, 35 (Broadnax); Tr. 10/26/77, 33 (Ferrebee)); rejections
could be made only by a lieutenant or commander (Tr. 10/26/77,
45-46 (Ferrebee)); irresponsible investigators were transferred
(id., 41); and the background investigation unit was integrated
(Id., 42) .
-15-
percent of psychiatric rejectees were black, and the
psychiatrist refused to give reasons or document the
rejections.68/ A new psychiatrist was appointed who docu
mented all rejections; the result was to reduce significantly
the rejection rate of blacks, and many black applicants who
had been rejected were reevaluated and accepted.69/
As a result of the above changes instituted successively
in 1971-1973, the hiring rate of blacks increased to about 30
percent.70/ In late 1973, the Police Department began using
an entrance test developed by John Furcon of the University
of Chicago, who had been retained by the Vickery Committee to
construct a validated test.71/ The City was satisfied that
this test was appropriately validated,72/ and black and white
68/ Id., 28-30.
69/ Id., 30-31; Tr. 10/31/77, 54 (Broadnax). Another
problem at the medical stage was that blacks who barely
passed the height and weight standards were reexamined, but
this was not required of whites (Tr. 10/26/77, 25
(Ferrebee)).
70/ DX 208, p. 4.
71/ Tr. 8/11/77, 43-45; 8/18/77, 28 (Caretti). The goal of
the Vickery Committee was "to develop a fair, equitable test
that would measure or predict job performance, and would also
reduce and, if at all possible, eliminate the cultural biasis
[sic] that were part of previous practices." Tr. 8/11/77, 36
(Caretti).
72/ Tr. 8/16/77, 12 (Caretti). The City's expert consult
ants on the Vickery Committee examined the empirical
validation data and concluded that the test was validated.
Id., 13-14. Furcon's test was differentially validated—
different passing scores are used for black and white appli
cants in order to predict the same probability of job per
formance by both groups. Tr. 8/11/77, 45; 8/17/77, 31-32;
8/24/77, 43-47 (Caretti). The minimum passing level was set
higher than the reading and writing abilities of incumbent
police officers. Tr. 8/25/77, 49-51 (Caretti).
-16-
applicants passed at about the same rate.73/ The result of
using this test, as well as the other changes in selection
described above, was that blacks and whites were hired in
equal numbers in 1974.74/
As a result of the City's affirmative action efforts in
hiring, black representation on the Police Department
increased from 4.9 percent at the end of 1967 to 17.2 percent
in 1974.75/ The City's ultimate goal since 1968 has been
that the department should ultimately be 50 percent black and
i
thereby reflect the City's population.76/ By the end of
1977, the department was 32 percent black.77/
73/ PX 190, p. 42.
74/ Tr. 10/26/77, 67-71 (Ferrebee); Tr. 8/25/77, 54
(Caretti).
75/ DX 208, p. 3 (data for June 13, 1974).
76/ This goal was held by Commissioner Spreen in 1968 (Tr.
878/77, 90-91) and was formally adopted as the City's policy
by Mayor Gribbs in 1971. Tr. 10/27/77, 25 (Ferrebee). Mayor
Young endorsed this policy both before and after his election
in 1973. Tr. 8/26/77, 17-19 (Chief Tannian).
77/ In 1974, the City also imposed a pre-application
residency requirement because there were enough qualified
applicants living in Detroit. PX 190, pp. 31-32. The effect
of this policy has been to reduce the percentage of white
applicants. Also, since 1968, black applicants have been
processed on an expedited basis "in response to the
conditions to try and correct the problems of the past." Tr.
8/25/77, 29 (Caretti); see also pp. 12-13, supra. Beginning
in late 1973, the City created two hiring lists by race and
exhausted the list of blacks before hiring whites, Tr.
10/27/77, 19-25 (Ferrebee), although there is no evidence of
the number of whites (if any) who were not hired because of
this procedure. In 1977, 60-70 percent of the applicants
were black, and 80 percent of those hired were black. Tr.
10/28/77, 37 (Ferrebee).
-17-
B. The City's Self-Analysis and Affirmative
Action in Promotions, 1968-1974______
As previously noted, the under-representation of blacks
historically has been even more severe at the supervisory
ranks (sergeant and above) than in the Police Department as a
whole. Thus, in 1953, when the Department was 2.4% black,
the supervisory ranks were 0.7% black (see, pp. 4-5, supra).
By the end of 1967, when the Department was 4.9% black, the
supervisory ranks were 2.1% black (see pp. 6-7, supra).
These disparities were caused by a policy of deliberate
exclusion. There was a widespread belief in the Department
that black officers would not enforce the law as diligently
against blacks as would white officers.78/ This* belief
resulted in the deliberate exclusion of blacks from the
investigative and specialized units.79/ The upward mobility
of blacks was limited to the detective operations, where they
would not be "leading white police officers in a command
function."80/
78/ Tr. 11/4/77, 26, 61-62 (Executive Deputy Chief Bannon).
Bannon, who is white, was first hired as a police officer in
1949 and at one time subscribed to that belief "[bjecause I
was socialized as every other police officer was." _Id_., 4-5,
62.
79/ Id., 61-62.
80/ Id., 27. Judge Edwards, who was Police Commissioner of
Detroit in 1962-1963, has stated: "The unwritten color lines
in police administration die hard. For example, as of 1962
no Negro police officer in Detroit had ever advanced to the
rank of uniformed lieutenant; there were many units in the
department where not a single Negro officer had ever served."
G.*Edwards, THE POLICE ON THE URBAN FRONTIER 87 (1968).
Judge Edwards promulgated a formal policy against
discrimination in promotions and assignments, see
(continued)
-18-
Following the 1967 race riot, the Department began to
examine and reform its promotional practices. As in the
hiring area, the Department's dual goals were to eliminate
existing discrimination in promotions and to remedy the
effects of past discrimination; and reforms were implemented
haltingly and in stages over a period of years.81/
1. "Dipping" Until 1968, promotions were made at the
discretion of the Commissioner. Rank-order eligibility lists
of candidates who passed the test were developed, with the
written examination score weighted 50 percent, service
ratings weighted 35 percent, seniority weighted 15 percent,
and a veterans' preference of 2 percent.82/ But Police
Commissioners followed a practice of "dipping"— i.e.,
promoting out of order for favoritism purposes, a
80/ (continued) Edwards, Order and Civil Liberties: A
Complex Role for the Police, 64 Mich. L. Rev. 47, 59-60
(1965). But vestiges of the exclusionary practice persisted
until the late 1960's because of resistance in some of the
bureaus and precincts. Tr. 11/4/77, 25-26, 50-51 (Bannon).
81/ However, unlike the hiring area, the Vickery Committee
did not examine the promotional process; and the Department's
corrective actions resulted from internal processes. Tr.
8/24/77, 57 (Caretti).
82/ DX 263 (weights for 1965). These weights varied over
the years. In 1960, the written examination counted 50
percent, service ratings 40 percent and seniority 10 percent.
HEARINGS BEFORE THE UNITED STATES COMMISSION ON CIVIL RIGHTS,
HEARINGS HELD IN DETROIT, MICHIGAN DECEMBER 14, 15, 1960, p.
394 (statement of Police Commissioner Herbert Hart).
-19-
practice which was not ended until 1968.83/
2. Service Ratings. Until 1973, service ratings were
widely abused.84/ The ratings were dependent upon job
assignments,85/ and police officers in the specialized units
were given improperly high ratings.86/ As noted above (p.
17), there was a historical practice.of racial segregation in
job assignments and of excluding blacks from specialized
units.
Johannes Spreen, who was Police Commissioner between
1968 and 1970, wanted to eliminate service ratings as a
factor in promotions.87/ His successor, John Nichols, also
felt that the service rating system "was not being properly
administered."88/ In November, 1972, Nichols made "drastic
changes" in the service rating system by issuing detailed
guidelines "to insure that the program would be administered
83/ Tr. 8/8/77, 49-50, 110 (former Commissioner Spreen); Tr.
8/9/77, 14-15, 26 (former Commissioner Nichols). The record
does not show the number of people promoted out of order
because of this practice. According to Nichols, who had been
on the Police Department since 1942, the practice was
prevalent enough to be very bad for morale (Tr. 8/9/77, 26).
Caretti,who was assigned to the personnel section in
September, 1968 (Tr. 8/11/77, 6), testified that "dipping"
affected less than 2 percent of promotions (Tr. 8/19/77, 35).
But Spreen, who was Commissioner in 1968, denied that he ever
dipped on promotions (Tr. 8/8/77, 49-50), while acknowledging
that his predecessors had (_id. 110).
84/ Tr. 8/10/77, 59-60 (former Commssioner Nichols).
85/ Id., 63; Tr. 8/9/77, 28 (Nichols), Tr. 8/8/77, 100-101
(former Commissioner Spreen).
86/ Tr. 8/8/77, 100 (Spreen).
87/ Id., 100-102.
88/ Tr. 8/9/77, 28 (Nichols).
-20-
fair ly and equitably across the Department."89/ These
guidelines provided, for the first time, objective standards
for service ratings and warned against the improper influence
of bias.90/ An appeals panel was also established to review
complaints of allegedly improper service ratings.91/ The
appeals panel found that a number of black police officers
were given low service ratings because of racial
discrimination.92/
Following these reforms, the average service ratings of
black and white police officers of comparable seniority were
the same.93/ However, the weight of service ratings in the
promotion model has been halved— from 30 percent in 1970 to
15 percent in 1975.94/ There is nothing in the record to
explain why this reduction occurred after the service rating
system was made job related and nondiscriminatory.95/
89/ Id., 28, 30.
90/ PX 51; PX 43.
91/ Tr. 8/9/77, 38-39 (Nichols).
92/ Tr. 8/24/77, 15 (Commander Caretti, who was Chairman of
the Appeals Board). There was no case of a white police
officer who got a low rating because of racial
discrimination. Ibid.
93/ PX 191; Tr. 8/16/77, 50-57 (Caretti).
94/ DX 263.
95/ Plaintiffs' expert testified, and the district court
found, that the existing (i.e., post-1972) service rating
system is job related and non-discriminatory. Tr. 10/13/77,
17-19 (Dr. Wollack); 446 F. Supp. at 991-992.
-21-
3. Seniority. In 1965, seniority was weighted 15
percent on the promotional model.96/ The Police Department
determined that there is no value in seniority as a deter
minant of promotability and that seniority adversely affected
black police officers because of past hiring practices.97/
For these reasons, the weight of seniority in the promotion
model was reduced to 8 percent in 1970 and 6 percent in
1974.98/
4. Written Promotion Examination. Until 1969, the
written examinations for promotion to sergeant were
essentially IQ tests.99/ These tests were not related to a
sergeant's job performance.100/ This heavy emphasis on IQ
was a "barrier” to the promotional opportunity of black
police officers— "a cultural bias * * * that did effectively
impede the success of minorities in the system.”101/
In September, 1968 Richard Caretti, then a lieutenant,
was assigned to the personnel section to develop promotion
examinations.102/ Caretti had first become involved with the
96/ DX 263.
97/ Tr. 8/24/77, 86-87 (Caretti).
98/ Id., 88-89; DX 263.
99/ Tr. 8/16/77, 67-68; 8/24/77, 61-64, 81 (Caretti).
100/ Tr. 8/24/77, 63; 8/25/77, 75 (Caretti).
101/ Tr. 8/24/77, 81 (Caretti). See also id., 61-62.
Comparative pass/fail statistics by race on these tests were
not presented because such statistics were not recorded until
1973 and could not be developed retrospectively. Id., 71.
102/ Tr. 8/11/77, 6 (Caretti).
-22-
subject of personnel selection a year earlier when he
attended a General Motors institute.103/ Caretti is not a
psychologist, and his efforts in test construction "were the
results of counselling and guidance from psychologists."104/
Based upon the experience he had obtained working with
the Vickery Committee, Caretti realized that the promotion
examinations had to be changed in order that black officers
would have equal opportunity.105/ His goal was to eliminate
all IQ questions and to develop a test that would closely
relate to the content of the sergeant's job.106/ Caretti's
initial efforts were not fruitful. He first worked on the
1969 sergeant promotion test,'but he was allowed to change
only one part of the test (the job-related section) because
"I had just arrived on the scene and possibly the supervisor
had not gained the confidence in my ability to do the
job * * *."107/ In 1970, Caretti prepared a qualifying
examination for promotion from the rank of detective to
103/ Id., 5.
104/ Id., 9.
105/ Tr. 8/24/77, 61 (Caretti). The expert witnesses who
testified for the plaintiffs agreed that promotion tests
which measure IQ are inappropriate, Tr. 10/14/77, 51, 53-54
(Dr. Wollack), and that such tests present a substantial
possibility of racial bias, Tr. 10/12/77, 62-64 (Dr. Ebel).
106/ Id., 62-63; Tr. 8/16/77, 67-68 (Caretti).
107/ Tr. 8/24/77, 64; id., 56 (Caretti).
-23-
sergeant, but the detectives' union exerted sufficient
political power to induce the Police Department to promote
all detectives who took the examination.108/ The result was
that the detectives did not study for the examination but
"just simply went through the motions."109/ No one failed
this examination, and 158 detectives were automatically
promoted, in mass, regardless of qualifications.110/ When
John Nichols became Commissioner in 1970, the promotional
examination was in disrepute. "I had a concern for the
examination per se * * * because a great many people felt
that the entire examination process had been contaminated and
what [sic] direction I gave then [to] Lieutenant Caretti was
that I wanted an examination put together so nobody could say
the fix was on, that they were discriminated against and that
somebody had an advantage that somebody did not have * * *.
What we tried to do was to re-establish the integrity of the
system itself * * *."111/
Caretti attempted to achieve content validity on the
1973, 1974 and 1976 sergeant promotion examinations.112/
108/ Id., 66-67. See also Tr. 8/10/77, 97 (former
Commissioner Nichols).
109/ Tr. 8/24/77, 68 (Caretti).
110/ Id., 69-70; Tr. 8/10/77, 98-102 (Nichols).
111/ Tr. 8/10/77, 80-81 (Nichols).
112/ Tr. 8/16/77, 20, 40 (Caretti).
-24-
He defined content validity as follows:113/
"Certainly the knowledge, skills and
abilities to do the job should be
identified through a job analysis, and
then the content of the examination
should relate to that job analysis.
Essentially that's the way you go
about to obtain content validity is
[sic] in an examina
tion process."
There were two job analyses conducted, by John Furcon in 1973 and
by Andres Inn in 1975.114/ However, Caretti "didn't understand"
Furcon's job analysis "that well"; so for the 1973 and 1974 tests
he "didn't have the benefit of a professional job analysis" and
relied instead on his own experience.115/
Caretti was also restrained in using the job analyses because
of labor negotiations concerning the promotional model.116/ The
examinations are based upon a complex bibliography of reading
materials.117/ Caretti inherited this bibliography when he began
working on promotion examinations in 1969, and it remained
unchanged because of directions by his superiors.118/ Caretti's
discretion in constructing the examinations was limited to
selecting questions from the bibliography.119/ For this purpose,
113/ Id., 19.
114/ Id., 20-21.
115/ Id., 21-22. See also Tr. 8/17/77, 10; 8/18/77, 45-47
(Caretti).
116/ Tr. 3/16/77, 30 (Caretti).
117/ Tr. 8/11/77, 48 (Caretti).
118/ Tr. 8/18/77, 41-42, 47 (Caretti).
119/ Id., 47.
-25-
Caretti retained qualified outside consultants,120/ gave each a
portion of the bibliography, and instructed each to develop a
representative cross-section of multiple-choice questions, which
Caretti then reviewed.121/ Thus, the questions on the examination
tested knowledge of "book content" and, according to Caretti,
"there were sections that dealt with knowledge of the book that
would in many cases surface on the job but many in some other
cases wouldn't surface on the job."122/
As noted above, Caretti's objective was to make the 1973-1976
examinations content valid. However, Caretti refused to express
an opinion as to whether these examinations were in fact content
valid. "I can't say * * * the test is content valid."123/ "Well,
we never found them [the promotion examinations] to be job
related. We worked towards that goal."124/ "I cannot sit here
any say that we have a content valid examination process. I can
say that we worked in that direction and we tried to do the things
that would lead to that goal."125/ "How well we succeeded would be
120/ Id., 42-45, 49-50.
121/ Tr. 8/17/77, 14, 22; 8/24/77, 105-106 (Caretti). One
outside consultant, Dr. Reginald Wilson, described the process
which he followed in developing questions in his assigned area.
He read two textbooks and two study guides from the bibliography.
Then he selected a cross-section of questions that would test,
knowledge of the books and guides. He tried to make the questions
as fair and clear as possible. Tr. 11/17/77, 95-96 (Wilson).
122/ Tr. 8/17/77, 22 (Caretti).
123/ Tr. 8/16/77, 81 (Caretti).
124/ Id., 76.
125/ Id., 82.
-26-
extremely speculative."126/ Indeed, on eight occasions during his
testimony, 'Caretti was asked whether the examinations were content
valid; each time he candidly refused to express that opinion.127/
However, two expert witnesses (Drs. Wollack and Ebel) testified
for the plaintiffs that, in their opinion, the 1973 and 1974 promo
tion tests were content valid.128/ They did not perform a
validation study but relied primarily upon their review of the
tests and of Caretti's testimony on how the tests were
developed.129/ They did not review the 1976 promotion test and
refused to express an opinion as to whether that test was content
valid.130/ The district court found that the 1973, 1974 and 1976
promotion tests were content valid on the basis of "the testimony
of Caretti, Wollack and Ebel."131/
Despite the elimination of IQ questions 132/ and other
126/ Id., 29.
127/ See notes 123-126, supra, and Tr. 8/16/77, 73, 75, 86;
8/17/77, 18.
128/ Tr. 10/12/77, 16, 35, 63-67 (Dr. Ebel); Tr. 10/13/77, 22,
32-34; 10/14/77, 35 (Dr. Wollack).
129/ Tr. 10/12/77, 65-67 (Dr. Ebel); Tr. 10/13/77, 21-22 (Dr.
Wollack).
130/ Tr. 10/14/77, 35-37 (Dr. Wollack); Tr. 10/12/77, 36, 63 (Dr.
Ebel).
131/ 466 F. Supp. at 991.
132/ The 1973 promotion examination contained a 100 question sec
tion, counting for 12.5% of the total score, consisting of the
"Watson-Glaser Critical Thinking Appraisal," which purports to mea
sure such abstract psychological traits as "inference," "recogni
tion of assumptions," "deduction," "interpretation," and "evalua
tion of arguments"; this section is an IQ test. PX 15. It was
eliminated from the 1974 and 1976 tests. PX 16, 17.
-27-
substantial improvements in the promotion examinations, blacks
failed the 1973, 1974 and 1976 tests at greater rates than
whites.133/ The weight of the written test in the promotional
model was increased from 60 percent in 1973 to 65 percent in 1974
and 1976.134/
5. Oral Boards. The department recognized that many capable
applicants for promotion were not good test takers but neverthe
less had good judgment and communications skills and would perform
excellently as sergeants.135/ To tap these "abilities and know
ledges that were not measured in any other aspect of the pro
motional model," the department instituted oral boards in
1974 136/ which were constructed by two outside experts.137/ The
members of each oral board were police command officers from out
side of Detroit,138/ who were trained in an orientation process so
that their ratings would be reliable and objective.139/ The
133/ The pass rates were 43% for whites and 28% for blacks on the
1973 test; 53% for whites and 39% for blacks on the 1974 test; 51%
for whites and 42% for blacks on the 1976 test. DX 198, 199, 200.
134/ PX 21, 22, 23.
135/ Tr. 8/16/77, 61-62 (Caretti); Tr. 9/6/77, 45; 9/19/77, 26-27
(former Commissioner Tannian); PX 36.
136/ Tr. 8/16/77, 61-62 (Caretti).
137/ Tr. 8/17/77, 54-59 (Caretti); 11/17/77, 97 (Dr. Wilson, one
of these experts).
138/ Id., 98 (this was done to reduce the potential for favori
tism or bias).
139/ Id., 98-99. Each oral board had three members, one of whom
was black. Id., 109; Tr. 8/17/77, 54 (Caretti). Each board mem
ber rated each applicant independently on the basis of a set
rating scale; and the ratings were then averaged. The interviews
were taped, and applicants could appeal to the experts; but there
was a minimal number of appeals. Tr. 11/17/77, 107-110 (Dr. Wilson).
-28-
applicants were tested on a cross-section of typical command
situations, and rated for their knowledge, judgment, logic,
intelligence, quickness of thought, ability to express themselves,
and sensitivity to human relations.140/ The oral boards are job
related and non-discriminatory.141/ On the average, black appli
cants were rated higher than whites on the oral boards.142/
However, the oral board rating was weighted only 10 percent in the
promotional model,143/ although the department believed that this
weight was too low.144/
6. College education and veterans1 preference. To be pro
moted to sergeant from the 1973 eligibility list, a police officer
had to have accumulated at least 15 quarter or 10 semester hours
of college credit; this requirement was raised to 30 quarter or 20
semester hours in 1974, and to 45 quarter or 30 semester hours in
1976.145/ The composite score was increased by one-half point for
each year of college up to two points, and a veterans' preference
of up to two points was also added.146/
140/ Id., 97-110.
141/ Ibid.; 446 F. Supp. at 992.
142/ Tr. 8/17/77, 72 (Caretti); 446 F. Supp. at 992.
143/ PX 22, 23.
144/ Tr. 8/17/77, 58 (Caretti). See also note 154, infra.
145/ PX 21, 22, 23. However, this requirement was waived for all
police officers with 12 1/2 years of seniority as of December 31,
1973. Ibid.
146/ Ibid.
-29-
7. Rank Order. Rank order on the promotion lists is deter
mined by weighting each applicant's scores on the various
components of the promotion model. The rank order scores
adversely affected blacks.147/ In 1973, 19 percent of the
applicants for promotion to sergeant were black;148/ if
rank order had been followed, only 10.6 percent of those
promoted would have been black.149/ In 1974, 27.7 percent
of the applicants were black;150/ rank order promotions
would have been 11.2 percent black.151/ In 1976, 30.9
percent of the applicants were black;152/ rank order pro
motions would have been 26.9 percent black.153/
The reason that rank order adversely affected blacks
is that the component of the promotion model on which
whites scored higher (the written test) was given a
147/ Tr. 8/17/77, 80 (Caretti).
148/ DX 200 (226 of 1191); see Tr. 8/24/77, 82 (Caretti).
The figures reported in this paragraph are for males only.
A number of women were promoted out of order to comply
with the decree in Schaeffer v. Tannian, infra note 168.
We limit these figures to men only because plaintiffs have
never challenged the out of order promotions of women.
150/ DX 199 (318 of 830); see Tr. 8/24/77, 83 (Caretti).
151/ DX 274 (14 of 125).
152/ DX 198 (300 of 971); see Tr. 8/24/77, 83 (Caretti).
153/ DX 274 (18 of 67).
-30-
maximum weight (60-65 percent), while the component on
which blacks scored higher (the oral boards) was given
a minimum weight (10 percent). There is no evidence in
the record which explains how the weights on the promo
tions model were chosen. Caretti inherited the promotion
model in 1969; he did not attempt to justify the weights
on the model but instead explained that his directions
from the City were to make no major changes in the model
because of union negotiations.154/ Nor. did any of the
plaintiffs' experts attempt to justify the component
weights and resulting composite rank-ordered scores.
Dr. Ebel did not even review these matters;155/ Dr. Wollack
had no opinion as to whether the weighted model was job-
related;!^/ and Mr. Guenther, a labor market expert, did
154/ Tr. 8/18/77, 46 (Caretti). Subsequently, in 1975
negotiations, the City proposed to eliminate the seniority
component and to weight the written examination 32 percent,
the oral board 32 percent, service ratings 20 percent,
college credit 14 percent, and veterans' preference 2 per
cent. Tr. 10/20/77, 12-14 (David Watroba, Vice-President
of D.P.O.A.). The union in turn proposed to increase the
seniority component, eliminate the oral board, and elimi
nate the college credit component. Id., 8. The negotia
tions reached an impasse. Id., 9.
155/ Tr. 19/12/77, 54, 71-72 (Dr. Ebel).
156/ "I have made some testimony with regard to the indi
vidual components of the model, but I really have no
opinion as to the method by which the various components
were put together so I really couldn't say whether it's
job related." Tr. 10/14/77, 60-61 (Dr. Wollack).
-31-
not testify on this issue at all.157/ However, plaintiff
D.P.O.A. did retain Dr. Andres Inn to examine the 1974:
promotion model.158/ Dr. Inn's validation study concludes
that the composite scores are based upon arbitrary weights,
that the department's scoring procedure was inappropriate,
that rank order did not measure relative qualifications
and that M[t]he rank order scores convey a false illusion
of overall superiority or inferiority."159/
Nevertheless, the district court stated that there
is "no written validation report regarding the 1973-1976
examination models. However, equally important to note
is the fact that Caretti, Wollack, Guenther and Ebel con
sistently testified that the promotional models for 1973-
1976, including each component part, were job related and
content valid"160/ Based upon this "testimony," the
district court found that rank order scores measured rela
tive qualifications for promotion.161/
157/ See Tr. 9/27/77; 9/28/77; 9/29/77 (Guenther).
158/ Tr. 10/20/77, 44 (Watroba).
159/ PX 298 (a).
160/ 446 F. Supp. at 993-994.
161/ Id., at 994.
-32-
8. Summary and results, 1968-1974. Between 1968 and
June, 1974, the.Police Department took substantial steps
which were designed to reduce or eliminate continuing
discrimination in the promotion process. The practice of
"dipping" was ended, the service ratings were made job-
related and non-discriminatory, the written test was re
formed, the weight of seniority was reduced, and the oral
boards were added. However, the department had taken no
race-conscious steps to remedy the effects of past discri
mination. Commissioners Spreen (1968-1970) and Nichols
(1970-1973) opposed promoting out of order from the eligi
bility lists, although they recognized the necessity of
having more blacks in the supervisory ranks.162/ There
was much less progress in integrating the supervisory ranks
than the entry-level police officer rank. By June of 1974,
the department as a whole was 17.2 percent black, but only
5.1 percent of the sergeants (61 of 1185) were black.163/
And the promotional model continued to have an adverse impact
against minorities. On May 9, 1974, thirty police officers
162/ Tr. 8/8/77, 76, 87-91, 111-113 (Spreen); Tr. 8/9/77,
18-22; 8/11/77, 48-50 (Nichols). Commissioner Spreen
expressed the dilemma thusly: "I think it's a hell of a
mess to try to rectify what ought to be done." Tr. 8/8/77,
96.
163/ DX 208, p. 3. The lieutenants' rank was 4.7 percent
black (11 of 230). Ibid.
-33-
were promoted to sergeant according to rank order from the
top of the eligibility list; 29 of those promoted were
white.164/ The department then determined that affirma
tive race-conscious steps were necessary to remedy the
exclusion of blacks at the supervisory ranks caused by
past discrimination.
D. The Affirmative Action Promotions, 1974-1977
On July 22, 1974, the Chief of Police, Phillip
Tannian, requested permission from the Board of Police
Commissioners to make promotions on an affirmative action
basis.165/ The Board was created by the Detroit City
Charter of July 1, 1974; and its five members had sole
authority to authorize promotions out of rank order from
the eligibility lists.166/ Tannian believed that he had
a legal and moral duty to take affirmative action to rec
tify the effects of past discrimination.167/ He believed
164/ 446 F. Supp. at 987.
165/ PX 240 (minutes of 7/22/74 meeting).
166/ PX 276, Section 7-1114 (City Charter).
167/ Tr. 8/26/77, 81; 8/30/77, 6; 9/12/77, 11-13 (Tannian).
The Corporation Counsel's Office advised the department
and the Board of Police Commissioners "that we had a duty
to move." Tr. 9/16/77, 20, 23, 33; 9/19/77, 28 (Tannian);
DX 44.
-34-
that the department could voluntarily take corrective action
without having to be sued and ordered to act by a court:168/
"[If] the [judicial branch] upon * * *
appropriate finding[s] * * * can order
certain actions, and those actions are
entirely proper and lawful, then it should
not be a requirement upon the executive
branch of government, where they analyze
the facts and find things to be dispro
portionate and not in compliance, [to]
stand around and wait for somebody to sue
them to have a court order to deal with."
Chief Tannian also was aware of the extreme hostility between
the black community and the depart
ment resulting from the department's exclusionary practices.169/
He believed that the law enforcement capabilities of the
department had been impaired by the exclusionary effects of
past discrimination, and that "a department which more accurately
reflects the pluralistic characteristics of our City will
be best equipped to carry out the primary responsibility of
168/ Tr. 9/16/77, 18 (Tannian). The department had in
fact been sued for sex discrimination. The district court
found that the department had historically discriminated
against women in hiring and promotion. Schaeffer v. Tannian,
394 F. Supp. 1128 (E.D. Mich. 1974). The court ordered the
department to hire one woman for each man, ibid., and to
promote 19 women preferentially to sergeant from the existing
eligibility list and to thereafter promote without regard
to sex. 10 FEP Cases 896 (E.D. Mich., order of June 7, 1974).
169/ Tr. 9/7/77, 86 (Tannian).
35
the Department: effective, fair, equal and professional
law enforcement." 170/
At the July 22, 1974, meeting of the Board of Police
Commissioners, Tannian admitted the past practices of dis
crimination in hiring and promotion, produced evidence to
support that admission and emphasized the continuing effects
of those practices at the supervisory ranks, which were
still virtually all white. 171/ The Board took his request
for affirmative action under advisement and reconvened on
July 26, 1974. The Board again convened on July 31, 1974,
at which representatives of D.P.O.A. spoke. 172/ The Board
then announced that affirmative action was necessary because
of past discrimination in hiring and promotions. 173/ The
170/ Tr. 9/7/77, 9 (Tannian). See also _id. , 85-86, 88.
171/ PX 240 (minutes of 7/22/74 meeting).
172/ Id. (minutes of 7/31/74 meeting).
173/ Affirmative Action Resolution of Board of Police
Commissioners (PX 32); PX 240 (minutes of 7/31/74 meeting,
p. 2). In concluding that the department had engaged in
past hiring and promotion discrimination, the Board members
relied not only on Tannian's presentation but also their
own knowledge, as life-long residents of Detroit, that
historical discrimination by the department was "notorious,"
Tr. 11/9/77, 67 (Chairman Littlejohn), and that blacks
"were not welcome." Id., 93.
-36-
Board agreed with Tannian that the past discrimination had
a two-fold effect--it created a present racial balance at
the supervisory ranks, and it diminished the department’s
law enforcement abilities:
M[T]he discrimination existed, the
impact of that discrimination was a
dramatic imbalance of Police personnel
within the Department that affected
the Department's ability to control
crime in the City of Detroit. The two
are inter-related, you cannot separate
one from the other because one exists
because the other existed previously."174/
The Board unanimously authorized the department to make
promotions on a basis of roughly one black for each white.
The Board chose this ratio after considering other alterna
tives; it decided on this ratio because anything greater
would be too disruptive and anything less would be too slow
in remedying past discrimination.175/ The affirmative
action program had two goals: to remedy the condition
caused by past discrimination and to develop racially neutral
hiring and selection procedures.176/ Thus, the Board consi
dered the accelerated promotion of blacks as a temporary
174/ Tr. 11/9/77, 10 (Littlejohn).
175/ Tr. 11/9/77, 150-152 (Littlejohn). This ratio was
not chosen by the Board because it reflected the City’s
population, ibid., although Tannian proposed a 1:1 ratio
for that reason. Tr. 8/30/77, 7 (Tannian).
176/ Tr. 11/9/77, 104-105 (Littlejohn).
-37-
measure which was ultimately consistent with merit selec
tion:177/
"We were not challenging the [promotion]
examination at that time as to whether it
was either biased or unbiased, that was
not the question before us. What we were
trying to get to was a model, that once
we got into an equal position that that
model could be applied equally and then,
as I said before, there would be no
further need for an Affirmative Action
Program.
"We were talking about remedying a condi
tion, not the validation of an examination,
that was new and was then being implemented,
that hopefully would be our model for the
future."
The Board therefore reviewed the affirmative action program
each time the Chief requested to be allowed to promote out
of order, to determine whether the condition caused by past
discrimination was remedied.178/
Between 1974 and 1977, 159 whites and 152 blacks were
promoted to sergeant pursuant to the affirmative action
program.179/ All of the persons promoted, both white and
177/ Id., 80.
178/ Id. , 85-86, 118-120. See also PX 240 (minutes of
4/28/77 meeting, pp. 9-12, statement of Commissioner Cohn).
179/ DX 274 (figures for males only). See note 148, supra.
-38-
black, had passed the written examination, were on the
eligibility list, and were fully qualified;180/ and they
were substantially more qualified than persons promoted to
sergeant a decade earlier.181/ By the end of 1977, the
sergeant's rank was 15.1 percent black.182/ If rank order
had been followed, 287 whites and 47 blacks would have been
promoted,183/ and the sergeants' rank would have been only
7.1 percent black.184/
ARGUMENT
THE DETROIT POLICE DEPARTMENT'S VOLUNTARY
AFFIRMATIVE ACTION PROMOTIONS ARE LAWFUL
UNDER THE CONSTITUTION AND THE FEDERAL
CIVIL RIGHTS ACTS
I. The Federal Civil Rights Acts and the
Constitution Do Not Prohibit, and Some
times Require, the Use of Race-Conscious
Practices To Eliminate The Effects of an
Employer's Unlawful Discrimination______
The district court held that Detroit's affirmative
action promotions violated Titles VI and VII of the 1964
180/ Tr. 8/24/77, 106, 111 (Caretti).
'181/ Tr. 8/25/77, 74-76 (Caretti).
182/ DX 264.
183/ DX 274 (figures for males only).
184/ Before the disputed promotions, there were 61 black
and 1124 white sergeants (see p. 32, supra). If rank order
had been followed, and discounting turnovers, there would
have been 108 black (61 + 47) and 1411 white (1124 +287)
sergeants by the end of 1977.
-39-
Civil Rights Act, 42 U.S.C. 2000d, 2000e; the Civil Rights
Act of 1866, 42 U.S.C. 1981;185/ and the Equal Protection
Clause of the Fourteenth Amendment.186/ Although these
statutory and constitutional provisions have different
perimeters in their substantive prohibitions against racial
discrimination in employment,187/ the core of each is the
same: individuals are protected against intentional racial
185/ For convenience, we refer hereinafter to Titles VI
and VII and Section 1981 collectively as "the Civil Rights
Acts."
186/ The district court also struck down the promotions as
violative of Michigan law. This conclusion is unsupported
by Michigan precedent. None of the Michigan cases cited
by the district court (446 F. Supp. at 1012) deals with
the issue presented in this case--the validity of voluntary
race-conscious relief by an employer which had practiced
unlawful discrimination--and we have found no such reported
decision in our research. The district court also cited
two decisions of courts of other states (id.) which invali
dated such relief under the laws of those states. However,
the highest courts of another two states, in decisions not
cited by the district court, have upheld such voluntary
affirmative action. Lindsay v. City of Seattle, 86 Wash.2d
698, 548 P.2d 320 (1976), certiorari denied, 429 U.S. 886 (1976)
Hutchinson Commission v. Midland Credit Management, Inc.,
213 Kan. 399, 517 P.2d 158 (1973). There is no basis for
asstuning that the Michigan Supreme Court would decide this
difficult issue one way or the other. When the substantive
reach of state law is so clearly unsettled, a federal court
should not, in the exercise of pendent jurisdiction, hold
that the actions of public officials violated state law.
See, United Mine Workers v. Gibbs, 383 U.S. 715, 726 (1966);
Moor v. County of Alameda, 411 U.S. 693, 715-716 (1973).
187/ For example, Title VII applies to private as well as
public employers and prohibits all forms of employment
discrimination, whether intentional or unintentional. See,
e.g., Griggs v. Duke Power Co., 401 U.S. 424 (1971). The
(continued)
-40-
discrimination; 188/ and the State, as an employer, ordinarily
may not base its decisions upon a racial classification,
whether the group burdened by the classification is in the
minority or majority.189/
The plaintiffs have proven (and defendants have never
denied) that since 1974, promotions to the rank of sergeant
in the Detroit Police Department have been made on the basis
of an overt racial classification. This is a prima facie
violation of the Civil Rights Acts and the Equal Protection
Clause.190/
The police department's principal justification for
using a racial classification in promotions is that this
was necessary and proper to eliminate the effects of its
own past unlawful discrimination against blacks. Its purpose
187/ (continued) Equal Protection Clause is, of course,
limited to situations of "state action" and prohibits only
intentional discrimination. See Washington v. Davis,
426 U.S. 229 (1976).
188/ See, e.g. , Washington v. Davis, supra, 426 U.S. at 239;
City of Los Angeles Department of Water& Power v. Manhart,
46 U.S.L.W. 4347, 4349 (U.S.#April 25, 1978); McDonald v.
Santa Fe Trail Trans. Co.. 427 U.S. 273, 283, 295 (1976).
189/ See, e■g., Regents of the University of California v.
Bakke, No. 76-811 (U.S., June 28, 1978); McDonald v. Santa
Fe Trail Trans. Co., supra.
190/ The reason for this is straight-forward: plaintiffs
proved that the police department intentionally used race
as a factor in promotions. The district court's adaptation
(continued)
-41-
is to remedy voluntarily past violations of the law and
thereby bring about a condition of true racial neutrality.
We agree that this purpose--and the use of an appropriate
numerical racial classification to achieve it— does not
violate the Civil Rights Acts or the Constitution.
When unlawful racial discrimination has been practiced
by the State, the Constitution and the Civil Rights Acts
do not command that the consequences of that discrimination
should go uncorrected. On the contrary, they impose upon
the State the affirmative duty to eliminate the present
effects of past discrimination. Green v. County School
Board, 391 U.S. 430 (1968). This duty governs all areas
of racial discrimination, whether involving schools, e.g.,
ibid; Swann v. Charlotte-Mecklenburg Bd. of Educ., 402 U.S.
1 (1971); McDaniel v. Barresi, 402 U.S. 39 (1971); or voting,
e.g., Louisiana v. United States, 380 U.S. 145, 154 (1965);
190/ (continued) of the standards set out in McDonnell
Douglas Coro, v. Green, 411 U.S. 792, 802 (1973) (see
446 F. Supp. at 1002-1003) is unnecessary and imposes too
high a burden of proof on plaintiffs. McDonnell Douglas
articulates evidentiary standards for inferring racial
intent from circumstantial evidence when the employer denies
such intent. See also Furnco Const. Co. v. Waters, No. 77-
369 (U.S., June 29, 1978). In this case, there is no need
to infer racial intent--it has been conceded from the outset.
-42-
or housing, e.g., Hills v. Gautreaux, 425 U.S. 284, 297
(1976); or employment, e.g., Albemarle Paper Co. v. Moody,
422 U.S. 405, 418 (1975). Moreover, this affirmative duty
is imposed by each of the federal laws involved in this
case: by the Constitution, e.g., Swann, supra; by Title VII,
e.g., Albemarle Paper Co., supra; Franks v. Bowman Transp.
Co., 424 U.S. 747 (1976); by Title VI, e.g., United States
v. Jefferson County Bd. of Educ., 372 F.2d 836, 852-878
(5th Cir. 1966), opinion adopted by court en banc, 380 F.2d
385 (5th Cir. 1967), certiorari denied, 389 U.S. 840 (1967);
and by Section 1981, e.g., Carter v. Gallagher, 452 F.2d
315, 327 (8th Cir. 1971) (en banc), certiorari denied,
406 U.S. 950 (1972); Davis v. County of Los Angeles, 566 F.2d
1334 (9th Cir. 1977), certiorari granted, 46 U.S.L.W. 1553
(June 19, 1978).
The use of a racial classification is therefore appro-
priate--and sometimes required--to remedy past employment
discrimination. Thus, when there has been unlawful discri
mination, back pay and seniority credits may be awarded on
a racial basis, Albemarle Paper Co., supra; Franks, supra,
even though the latter form of relief upsets the expecta
tions of incumbent white employees, see id., 424 U.S. at
775. And employers may also be required to institute racial
-43-
classifications to cure the basic conditions which result
from their past discrimination. This Court has consistently
upheld district court orders requiring employers or labor
unions to adopt fixed numerical racial goals in hiring,
promotion, union membership and referrals to rectify gross
racial imbalances caused by their discrimination. For
example, in the Detroit Edison case (where the labor force
and evidence of past discrimination are comparable to that
presented here, see p. 53 , infra), the district court ordered
the company to hire at a ratio of three blacks for every two
whites and to promote at a ratio of one black for every
white, Stamps v. Detroit Edison, 365 F. Supp. 87, 122-123
(paragraphs 9, 11, 12) (E.D. Mich. 1973); and this Court
affirmed those orders, 515 F.2d 301, 317 (6th Cir. 1975).191/
See also United States v. I.B.E.W., Local 212, 472 F.2d 634,
636 (6th Cir. 1973) (upholding union membership goals);
191/ The Supreme Court vacated and remanded for reconsidera
tion in light of Teamsters, 431 U.S. 951 (1977). On remand,
this Court modified certain provisions of its earlier judg
ment (relating to seniority and the burden of proof at Stage
II proceedings) and otherwise reaffirmed its decision. Nos.
74-1007, 74-1008, 74-1009, 74-1675 (6th Cir., August 12, 1977).
The company petitioned for rehearing and challenged specifi
cally the affirmance of the district court's 1:1 promotion
goals. This Court denied rehearing. Nos. 74-1007, 74-1008,
74-1009, 74-1675 (6th Cir., November 8, 1977).
-44-
Sims v. Sheet Metal Workers, Local 65, 489 F.2d 1023, 1027
(6th Cir. 1973) (upholding union referral ratio); United
States v. Masonry Contractors Ass'n of Memphis, Inc., 497 F.2d
871, 877 (6th Cir. 1974) (upholding hiring goals); Arnold v.
Ballard, 12 FEP Cases 1613 (6th Cir. 1976) (upholding hiring
goals), vacated on other grounds, 16 FEP Cases 396 (6th Cir.
1976); see United States v. I.B.E.W., Local 38, 428 F.2d
144, 149-151 (6th Cir. 1970) (instructing district court
to order affirmative action), certiorari denied, 400 U.S.
943 (1970). Eight other circuits have also upheld the
authority of the district courts to order numerical relief
to rectify the effects of past discrimination in hiring and
\
promotions.192/
192/ See, e.g,, Morgan v. Kerrigan, 509 F.2d 599 (1st Cir.
1975); Boston Chapter, NAACP v. Beecher, 504 F.2d 1017,
1026-1028 (1st Cir. 1974), certiorari denied, 421 U.S. 910
(1975) ; Patterson v. Newspaper Deliverers' Union, 514 F.2d
767, 773-775 (2d Cir. 1975), certiorari denied, 427 U.S.
911 (1976); Rios v. Steamfitters, Local 638, 501 F.2d 622,
628-633 (2d Cir. 1974); but cf. Kirkland v. Department of
Correctional Services, 520 F.2d 420, 427-430 (2d Cir. 1975),
reh. denied, 531 F.2d 5 (2d Cir. 1975), certiorari denied,
429 U.S. 823 (1976); EEOC v. A.T. & T. Co., 556 F.2d 167
(3d Cir. 1977), certiorari denied, Nos. 77-241-243 (July 3,
1978); United States v. Elevator Constructors, Local 5,
538 F.2d 1012 (3d Cir. 1976); Patterson v. American Tobacco
Co., 535 F .2d 257, 273-275 (4th Cir. 1976), certiorari denied,
429 U.S. 920 (1976); Watkins v. Scott Paper Co., 530 F.2d
1159, 1194 (5th Cir. 1976), certiorari denied, 429 U.S. 861
(1976) ; NAACP v. Allen, 493 F.2d 614 (5th Cir. 1974); Morrow
v. Crisler, 491 F.2d 1093 (1974), certiorari denied, 419 U.S.
895 (1975); United States v. City of Chicago, 549 F.2d 415
(7th Cir. 1977), certiorari denied, 434 U.S. 875 (1977);
(continued)
-45-
Moreover, this Court has held that affirmative race-conscious
relief to correct the effects of past discrimination is
mandatory,193/ and three other Circuits have held that numeri
cal goals for hiring and promotion must be used when less
drastic relief fails to achieve that end.194/
The validity of race-conscious remedies for past unlaw
ful discrimination is also supported by the Supreme Court's
recent decision in Regents of the University of California v.
Bakke, No. 76-811 (decided June 28, 1978). Of course, Bakke
involved a very different and more complex problem than the
case at bar--the Davis Medical School did not attempt to
justify its special admissions program as a remedy for past
192/ (continued) Crockett v. Green, 534 F.2d 715, 718-719
(7th Cir. 1976); United States v. N.L. Industries, 479 F.2d
354, 377 (3th Cir. 1973); Carter v. Gallagher, 452 F .2d 315,
327 (8th Cir. 1971), certiorari denied, 406 U.S. 950 (1972);;
Davis v. County of Los Angeles, 566 F.2d 1334 (9th Cir. 1977).
certiorari granted, 46 U.S.L.W. 1553 (June 19, 1978); United
States v. Ironworkers, Local 86, 443 F.2d 544, 552-544, (9th
Cir. 1971), certiorari denied, 404 U.S. 984 (1971).
193/ United States v. I.B.E.W., Local 38, 428 F.2d 144,
159-161 (6th Cir. 1970), certiorari denied, 400 U.S. 943
(1970).
194/ Watkins v. Scott Paper Co., 530 F.2d 1159, 1194 (5th
Cir. 1976) (ordering numerical goals for promotions); Morrow
v. Crisler, 491 F.2d 1053 (5th Cir. 1974) (en banc) (same for
hiring), certiorari denied, 419 U.S. 895 (1974); United States
v. N.L. Industries, 479 F.2d 354, (8th Cir. 1973) (same for
promotions); United States v. Local 169, Carpenters, 457 F.2d
210, 220-221 (7th Cir. 1972) (same for union work permits),
certiorari denied, 409 U.S. 851 (1972).
-46-
unlawful discrimination, and there was no claim that Davis had
ever practiced such discrimination. Nevertheless, the Court
held that, even in the absence of past discrimination, the
Constitution and the Civil Rights Acts do not require "color
blind" behavior by public institutions. The opinions affirmed
that race may be a factor in admissions, either to achieve broader
diversity in a student body,195/ or to ameliorate the effects of
societal discrimination.196/ The Court also held that, in the
absence of past unlawful discrimination, Title VI prohibited a
university from setting aside a fixed number of admissions on a
racial basis.197/
Bakke did not present the situation of an institution which
adopted a numerical admissions goal to remedy its own past discri
mination, and four justices (in the opinion authored by Justice
Stevens) expressed no view on the legality of such a remedy under
the Civil Rights Acts or the Constitution.198/ Four other justices
195/ Opinion of Powell, J., slip op. at 42-45.
196/ Opinion of Brennan, White, Marshall & Blackmun, JJ., slip
op. at 38-45.
197/ Opinion of Stevens, J. (joined by Burger, C.J., Stewart &
Rehnquist, J.J.); Opinion of Powell, J., slip op. at 50-51.
198/ Justice Stevens made clear at the outset of his opinion that
he was addressing only those questions he considered squarely
before the Court (slip op. 1-4). Justice Stevens did not assert
that the use of racial criteria for admissions would violate Title
VI if this was designed to remedy a past exclusionary policy.
Indeed, Justice Stevens stated that HEW regulations requiring the
use of racial criteria were inapposite in Bakke because those
regulations did not apply " [w]here no discriminatory policy was
in effect." (Slip oo. 11, n . 22) . See 45 C.F.R. 80.3(b)(6)(i),(ii),
80.5(i). And to'hold that Title VI precludes the use of racial
criteria to overcome the effects of past discrimination would ce
directly contrary to Swann, supra, and Franks, supra, and could
render Title VI unconstitutional. See North Carolina 3d. or Ecuc.
v. Swann, 402 U.S. 43 (1971). There is no reason to believe that
Justice Stevens was advocating such a position.
-47-
(in the joint opinion of Justices 3rannan, White, Marshall, and
Blackmun), specifically approved the decisions of this and other
federal courts (see pp. 43-45, supra) which upheld the use of numeri
cal goals as a remedy for unlawful employment discrimination,
whether that remedy was imposed by a court or adopted voluntarily
by an employer (slip op. 29-31 & n. 28, 39-41). Finally, Justice
Powell, writing for himself, also endorsed those opinions upholding
numerical goals in remedial contexts, at least when a court or
other appropriate governmental body had made a finding of unlawful
employment discrimination (slip op. at 32-33, 38). Thus, a majority
of the court in Bakke approved race-conscious numerical relief for
past unlawful discrimination, and four of five justices who dealt
with this question also held that such relief could be adopted
voluntarily.
Thus the central issue in this case— which was not decided
by a majority of the Court in 3akke and is a matter of first
impression in this Court— is the extent to which an employer may
voluntarily adopt race—conscious numerical goals to remedy its
past unlawful discrimination. The initial question in this case
is whether the Detroit Police Department had unlawfully discriminated
against blacks.199/ We think that the evidence clearly establishes
199/ It is unnecessary in this case to determine whether affirma
tive action can be undertaken voluntarily only when an employer
admits to and proves past discrimination. Compare the^majoritv
opinion in Weber v. Kaiser Aluminum & Chemical Corp. , _56 3 F.̂ 2d̂ 216
(5th Cir. 1977) with Judge Wisdom's dissent, id. at 227, and the
proposed EEOC guidelines on affirmative action, 42 Fed. Reg. 64826
(1977). Ordinarily, an employer would be extremely reluctant to^
admit, let alone prove, past discrimination because that would^risx
a suit for class back pay. Requiring an admission and/or prcoi.
of past discrimination would tend to chill voluntary compliance;
it might also undermine consent decrees and conciliation agreements.
Thus, in appropriate cases, an adequate predicate for voluntary
(continued)
-48-
that it had, and that the contrary conclusions of the district
court must be set aside. We address the department's past dis
crimination in Part II (pp. 48-63), which follows. We then
address in Part III (pp. 64-72) whether Detroit could voluntarily
remedy that discrimination by adopting race-conscious numerical
goals. In Part IV (pp. 73-90), we examine whether the particular
remedy adopted by Detroit is appropriate.
II. THE DETROIT POLICE DEPARTMENT PRACTICED
SYSTEMATIC AND UNLAWFUL PAST DISCRIMINA-
TION AGAINST BLACKS IN HIRING AND PROMOTIONS
The district court held unlawful the Detroit Police Depart
ment's affirmative action plan principally because the court con
cluded that the justification for such a remedy— prior discrimina
tion against blacks in hiring and promotion— was factually absent.200/
Whether there was past discrimination is a question of ultimate
fact to be determined by an evaluation of evidentiary facts accord
ing to governing legal standards. See, e .g., Castaneda v. Partida,
430 U.S. 482 (1977); Norris v. Alabama, 294 U.S. 587 (1935);
Causev v. Ford Motor Co., 516 F.2d 416 (5th Cir. 1975). Cf.
Baumgartner v. United States, 322 U.S. 665, 670-671 (1944). And,
199/ (continued) affirmative action can be a reasonable factual
basis for an employer's belief that its employment practices were
not in conformity with the law. Since, however, the Detroit Police
Department has admitted to and proven past discrimination, this
question is not properly before the Court.
200/ 446 F. Supp. at 1005-1010. See, e.g., id.at 1009 ("In the
absence of proof of prior discrimination as claimed by the defen
dants the preferential promotion plan of the derendants clearly
violates Title VTI as an impermissible racial quota.").
-49-
in reviewing the district court's evidentiary findings concerning
past discrimination, the "clearly erroneous" rule is not applicable
because, as we show below, these findings were induced by incorrect
legal principles. See, e.g., Senter v. General Motors Corp.,
532 F.2d 511, 525 (6th Cir. 1976), certiorari denied, 429 U.S. 870
(1976). However, even applying the "clearly erroneous" rule, these
evidentiary findings must be set aside because a review of the
entire evidence necessitates the firm conviction that the district
court was wrong. United States v. United States Gypsum Co., 333 U.S
364, 394-395 (1948). When the evidence relating to the Police
Department's historical hiring and promotion practices— evidence
which is basically undisputed but was largely disregarded by the
district court— is evaluated under established legal principles,
the conclusion of past intentional discrimination against blacks
in hiring and promotions is required as a matter of law. See, e.g. ,
Castaneda v. Partida, supra; International Brotherhood of Teamsters
v. United States, 431 U.S. 324 (1977); Hazelwood School Dist. v.
United States, 433 U.S. 299 (1977).
A. Past Discrimination in Hiring
The district court acknowledged that between 1944 and
1968, the number of whites hired by the Detroit Police Department
"far outdistanced" the number of blacks hired (446 F. Supp. at 998).
The court also found that (id. at 1000):
"Until 1973 entry level (hiring)
written examinations * * * may
have constituted a source of dis
crimination against blacks seeking
entry into the department because
-50-
these examinations were heavily
weighted on I.Q. type questions,
were not job-related and tended
to fail large numbers of blacks
vis-a-vis whites."
The district court concluded, however, that there was no past
discrimination in hiring. The court reached this conclusion by
accepting plaintiffs' argument that the appropriate labor force
was the unweighted Detroit SMSA (18.6 percent black in 1970) and
by comparing this with the composition of the police department
in 1974 (17.2 percent black). See 446 F. Supp. at 995-996, 1006.
This analysis is fundamentally incorrect.
Although the finding that the available labor force is as
low as 18.6 percent black is open to the most serious doubt,201/
the district court's basic error inheres in the assumption that the
201/ Plaintiffs' expert conceded that his use of the unweighted
SMSA as the relevant labor force was "tentative." Tr. 9/29/77, 85
(Mr. Guenther). He reiterated that conclusions based upon the
geographical area of recruitment are "far less precise" than an
analvsis using the actual residence of applicants. Tr. 9/27/77,
28-30. Tr. 9/29/77, 86. Generally, people tend to apply to jobs
closest to where they live; thus one would expect that a dispro
portionate number of applicants would come from Detroit rather than
from the surrounding counties. Tr. 9/28/77, 21; Tr. 9/29/77,
52-54. The correct approach would be to weight the counties in the
SMSA according to the*proportions of applicants. Tr. 9/29/77,
52-54, 79-80. Since almost all blacks who live in the Detroit
SMSA reside in the City, and since it is expected that most^
applicants come from the City, a weighted approach should yield a
higher percentage black in the labor force than the unweighted
figure of 18.6 percent. Indeed, the fact that 47 percent of the
applicants in 1967 were black (see p. 11, supra)— before the City
engaged in affirmative recruitment— strongly suggests that the
18.6 percent figure is a gross underestimate.
-51-
police department's racial composition in 1974 (17.2 percent
black) refutes the existence of past hiring and promotion discri
mination. By 1974, the City had engaged in affirmative action in
hiring for six years.202/ In 1968, the department was only 4.9
percent black203/ which, compared to a labor force of 18.6 percent,
is prima facie evidence of systematic intentional discrimination
(see infra). The department's composition in 1974 means only that
its affirmative action efforts had ameliorated the effects of
past discrimination at the hiring level.204/ But these efforts
cannot erase the reality that prior discrimination existed,
Teamsters v. United States, supra, 431 U.S. at 341-342. Nor did
these efforts remedy the effects of past discrimination at the
sergeants' level, which was only 5.1 percent black in 1974.205/
As noted above, the department was only 4.9 percent black
when its affirmative action efforts began in 1968, while the labor
force was, according to the district court, 18.6 percent black.
This disparity is substantially greater than disparities which
have constituted, as a matter of law, prima facie evidence of
purposeful discrimination. The Supreme Court in Hazelwood held
that a prima facie case of intentional discrimination was esta-
202/ See pp. 11-16, supra. The 1968-1974 affirmative action
efforts included a preference to blacks in processing applications.
See id. at 12-13.
203/ See p. 6, supra.
204/ This assumes arguendo that the 13.6 percent labor force
figure is correct. See note 200, supra.
205/ See p. 32, supra.
-52-
blished by statistics showing a work force of 1.8 percent black
drawn from a labor force of 5.7 percent black. 433 U.S. at 307-
309.206/ Using the statistical methodology set forth in the
Castaneda case, the Court found in Hazelwood that the difference
between the actual number of black employees and the expected
number (based on the labor force percentage) was between five and
six standard deviations, while any difference greater than two or
three standard deviations creates an inference of purposeful dis
crimination. 433 U.S. at 308-309 n. 14. See Castaneda, supra, 430
U.S. at 496-497 n. 17. Applying that statistical methodology
here, the actual number of blacks in the police department in
1968 is twenty-three (23) standard deviations less than the expected
number based on a labor force which is 18.6 percent black.207/
And a more precise statistical calculation reveals that the likeli
hood that this result occurred by chance is less than 1 in 10^. 208/
206/ Hazelwood was, of course, a case where the government sought
to prove intentional discrimination. Id. at 306 n. 12.
207/ At the end of 1967, there were 214 blacks in the department
of a total complement of 4356. DX 208, pp. 3, 5. Given a labor
force that is 18.6 percent black, the expected number of blacks
would be 808 (i.e ., 4356 x .186).
Using the Hazelwood and Castaneda methodology, the standard
deviation is the square root of the product of the total number in
the department (4356) times the probability of selecting a black
from the labor force (.186) times the probability of selecting a
white (.814). Thus, the standard deviation is V(4356)(.186)(.814),
which equals 25.7. The difference between the expected and actual
numbers of blacks on the department (808-214 = 694) is thus about
23 standard deviations (694 -r 25.7).
208/ This calculation is based on the statistical formulae in the
authorities cited in Castaneda, supra, 430 U.S. at 496-497 n. 17.
It can be replicated simply by using the pre-packaged program in
the Texas Instruments 59 pocket calculator.
-5 3-
That is, the probability that racial discrimination was not a
factor in the department's hiring until 1968 is equivalent to the
odds of being dealt eight consecutive royal flushes by an honest
dealer, or flipping a fair coin and getting 170 straight heads.209/
The lower federal courts have also held consistently that
disparities less than that presented here (4.9% black employment
in an 18.6% black labor force) are, standing alone, proof of dis
crimination. For example, in the Detroit Edison case,210/ the
company hired from the identical labor force as in this case (the
Detroit SMSA),211/ and its work force was 7.8% black.212/ This
Court held that this "statistical evidence alone— showing a dispro
portionately low number of black employees— would be sufficient to
support a finding of discrimination in hiring.n213/ And numerous
other cases have rested findings of discrimination on statistical
disparities which are less than those in the case at bar.214/
209/ The odds of being dealt one royal flush is approximately one
in one million.
210/ Stamps v. Detroit Edison Co., 365 F. Supp. 87 (E.D. Mich.
1973), aff'd in rel. part sub nom. EEOC v. Detroit Edison Co.,
515 F.2d 301 (6th Cir. 1975), vacated on other grounds, 431 U.S.
951 (1977), reaffirmed in rel. part on remand, Nos. 74-1007,
74-1008, 74-1009, 74-1675 (6th Cir., August 12, 1977), rehearing
denied, Nos. 74-1007, etc. (6th Cir., Nov. 18, 1977).
211/ See 365 F. Supp. at 98.
212/ See 515 F.2d at 307 (832 blacks of 10,630 employees).
213/ Id. at 313 (emphasis added). The Court also noted that addi
tional evidence had been produced, ibid. Similar evidence is also
present in this case. See pp. 55, 58-59, infra.
214/ See, e.g., Afro-American Patrolmen's League v. Duck, 503 F.2d
294, 299 (6th"cir. 1974), aff'g, 366 F. Supp. 1095 (N.D. Ohio 1973)
(police force 8.2% black and City population 16% black); Jackson
v. Dukakis, 526 F.2d 64 (1st Cir. 1975), aff'g, 394 F. Supp. 162,
171 (D. Mass. 1975) (work force 9% black and City population 23%
(continued)
Between 1944 and 1967, the police department hired 5,537
persons, of whom only 325 (or 5.5%) were black (DX 208). The
district court discounted this because, the court found, there was
no evidence "showing what the relevant labor market was during
this period" (446 F. Supp. at 998)♦215/ This finding is clearly
erroneous. Plaintiffs' expert calculated the relevant labor
market during this period, using the number of persons in the
Detroit SMSA who were under thirty-four years of age and were high
school graduates. (Tr. 9/27/77, 45-47; 9/28/77, 14-16; PX 242,
Table 2A). Although accepting this calcuation for the 1970-1974
period (see 446 F. Supp. at 995-996 & nn. 38-42), the district
court ignored it for the fearlier period. According to the calcu
lations of plaintiffs' expert, 645 blacks would have been hired by
the department between 1944 and 1967 if selections were in propor
tion to the relevant labor market (PX 2 42, Table 2A) ; this is
twice the actual number of blacks hired.216/
214/ (continued) black); Erie Human Relations Comm, v. Tullio,
493 F.2d 371 (3d Cir. 1974), aff'g, 357 F. Supp. 422_ (W.D. Pa.
1973) (police department 1.4% black and City population 6.8%
black); Castro v. Beecher, 459 F.2d 725, (1st Cir. 1972),
rev'g, 334 F. Supp. 930, 936 (D. Mass. 1971) (police department
2% black and City popuplation 9% black).
215/ The district court also stated (ibid.) that there was no
applicant flow data for the 1944-1967 period. The presence of
such data is not necessary to prove discrimination. Dothard v.
Rawlinson, 433 U.S. 321, 330 (1977).
216/ The calculations of plaintiffs' experts also show that
blacks were grossly under-represented in the department during
this entire period. The following table compares the black compo
sition of the department and the relevant labor market (using
(continued)
-55-
The district court also stated that the evidence of past
discrimination in hiring "was largely, if not totally, statisti
cal." (446 F. Supp. at 1005). This is also clearly erroneous.
The evidence reveals with clarity the specific discriminatory
practices which caused the gross under-representation of blacks in
the department: (a) the use, until 1973, of written IQ tests for
hiring which failed blacks at much greater rates than whites and
which were not job-related;217/ (b) the discriminatory rejection
of blacks for "'miscellaneous' reasons" at the preliminary appli
cation stage;218/ (c) the disproportionate failure rate of blacks
at a subjective oral interview;219/ and (d) overt discrimination
in the conduct and evaluation of the background investigation
and medical examination.220/ This evidence was not discredited by
the district court; it was largely ignored.221/
216/ (continued) plaintiffs' definition) for 1943, 1953, 1962
and 1967:
Year Department (% Black)* Labor Force (% Black)**
1943 1. 3% 7.4%
1952 2.4% 10.2%
1962 3.5% 13.9%
1967 4.9% 15.2%
"*Source: Supra, pp. 4-6.
**Source: PX 242, Tab le 2A.
217/ see pp. 11-14, supra? 446 F. Supp. at 1000.
218/ See p. 11, supra.
219/ See ibid.
220/ See pp. 14-15, supra.
221/ Thus, except for the written entrance tests, which the court
found "many have * * * [been] a source of discrimination" (446
F. Supp. at 1000) , the district court did not address the evidence
of discrimination at the preliminary application, background invest
gation, medical examination and oral interview stages.
(continued)
-56-
B . Past Discrimination in Promotions
By excluding blacks from entry-level positions, the police
department clearly limited the number of blacks who could be
promoted to the supervisory ranks. The systematic nature of the
department's intentional discrimination in hiring (and assignments)
is also relevant in ascertaining the motives underlying the depart
ment's past promotion practices. The hiring and promotion decisions
of the police department cannot be neatly compartmentalized. When,
as here, there is proof of illicit intent in hiring and assignments,
that proof has "substantial probative value" in determining the
department's intent in making promotions. See Keyes v. School Dist.
No. 1, Denver, Colorado, 413 U.S. 189, 207-211 (1973). Moreover,
even if the department's employment decisions were compartmentalized,
there is undisputed evidence of overt discrimination in promotions;
but this evidence was almostly entirely disregarded by the district
court. The conclusory finding that there was no past discrimina
tion against blacks in promotions (446 F. Supp. at 1000) should
be set aside.
221/ (continued) It is unnecessary to determine whether an
employer, "in adoption of a so-called 'affirmative action' plan,
can rely solely upon statistical comparisons in an effort to prove
prior discrimination" (446 F. Supp. at 1Q05). Statistical dispari
ties are relevant because "absent explanation, it is ordinarily to
be expected that nondiscriminatory hiring practices will in time
result in a work force more or less representative of the racial
and ethnic composition of the population in the community from
which employees are hired." Teamsters, supra, 431 U.S. at 339—
340 n. 20. Here, the Detroit Police Department had no "explana
tion" for the underrepresentation of blacks other than racial
discrimination, and the department identified the discriminatory
practices.
-57-
When the police department adopted its affirmative action
plan for promotions in July, 1974, only 61 of 1185 sergeants (or
5.1 percent) were black.222/ A part of this under-representation
of blacks at the sergeants' rank223/ is attributable to the effects
of racial discrimination in hiring. See Afro-American Patrolmen's
League v. Duck, 503 F.2d 294 (6th Cir. 1974). But the promotion
process is also partly responsible, for the under-representation
of blacks at the sergeants' rank has been consistently more
severe than in the department as a whole, as summarized in the
following table.
222/ DX 208, p. 3; see 446 F. Supp. at 994 n. 29.
223/ If blacks had been represented at the sergeants| rank in
proportion to the labor force as defined by the district court
(18.6 percent black), there would have been 220 black sergeants
in 1974 (i.e., 1185 x .186).
-58-
Year Department: % Black 224/ Sergeants: % Black 225/
1953 2.4% 0.9%
1962 3. 5% 1. 4%
1967 4.9% 2.6%
1974 17.2% 5.1%
Thus , even though by July, 1974, the police department's
affirmative action efforts may have largely rectified the past
exclusion of blacks at the hiring level (see p. 51 / supra) , the
continuing statistical disparity at the sergeants' rank estab
lishes a prima facie case of racial discrimination in promotions.
See, e .g., Robinson v. Union Carbide Corp., 538 F.2d 652, 659-661
(5th Cir. 1976), and cases cited therein; Watkins v. Scott Paper Co.
530 F.2d 1159, 1191-1192 (5th Cir. 1976), certiorari denied, 427 U.S
861 (1976); Stewart v.' General Motors Corp., 542 F.2d 445 (7th Cir.
1976), certiorari denied, 433 U.S. 919 (1977).
The evidence in the record also explains why blacks were so
grossly under-represented at the supervisory ranks: (a) Until the
early 1960’s, there was a racially exclusionary policy in
existence, and the upward mobility of black police officers was
224/ See pp. 4-6, 16, supra.
225/ See notes 13, 15 & pp. 9, 32, supra. These disparities cannot
be explained by the effect of the time-in-grade requirement in per
petuating hiring discrimination. As this table shows, black repre
sentation at the sergeants' rank was less in 1962 than in the
department as a whole in 1953; the same is true comparing the
sargeants' rank in 1967 with the department in 1962. And in 1970-
1974, the time-in-grade requirement was three years. Yet only 5.1
percent of the sergeants were black in 1974, while the department
was 13.1 percent black in 1971 (711 of 5437). DX 208, pp. 3, 5.
-59-
deliberately and severely restricted; vestiges of this policy
persisted into the late 1960's (see p. 17 & n. 80, supra);
(b) Until 1968, promotions were made at the discretion of the police
commissioner, who selected out of order from the eligibility lists
for favoritism purposes (see pp. 18-19, supra) ; (c) Until 1973,
service ratings, which counted 30-40 percent on the promotion
models, were unstructured, totally subjective and widely abused.
Arbitrarily high ratings were given to police officers in the
specialized units, from which blacks were excluded. And black
officers were given low ratings because of racial discrimination
(see pp. 19-20, supra); (d) Until the early 1970's, the written
promotion examinations were non-job-related IQ tests which were
ba'rriers to the promotion of blacks (see pp. 21-23, supra) ; And
(e) the use of seniority in the promotion model perpetuated the
effects of hiring discrimination (see p. 21, supra). As in the
hiring area, the district court simply ignored the undisputed evi
dence that these practices occurred. 226/ And much less evidence
than this has been considered sufficient to establish racial
226/ The district court made no finding concerning the
policy of exclusion, even though this evidence was not contra
dicted. Similarly, the district court ignored undisputed
evidence that "dipping" was practiced until 1968 and found,
instead, that "[p]rior to the August 1, 1974, promotionspromo
tions to the rank of Sergeant had always been made in strict
numerical rank according to the current eligibility register."
446 F. Supp. at 989. This finding is clearly erroneous. The
district court also discussed in detail the service rating system
only since 1973, 446 F. Supp. at 991-992. The conclusory finding
that there is "no evidence" of discrimination in service ratings
"prior to 1973" (ibid.) ignores undisputed evidence and is clearly
erroneous. The district court found that the promotion tests
historically placed a "heavy emphasis" on IQ-type questions, 446
F. Supp. at"989-990, but ignored Caretti’s undisputed testimony
that these tests were barriers to the promotion of blacks (even
though the district court specifically credited and relied upon
Caretti's testimony as a whole,see id. at 990 & n. 11, 991-994).
-60-
discrimination. See, e.g., Afro-American Patrolmen's League v.
Duck, supra. 227/
C. The Prior Discrimination in Hiring and
Promotions Was Unlawful________________
Finally, the district court held that the evidence concerning
past hiring practices was legally irrelevant because it concerned,
at most, unintentional discrimination which was not illegal before
1972, when Title VII became applicable to the police department
(446 F. Supp. at 1006-1007, 1013-1014). This holding is wrong.
First, as we have shown, when judged by proper legal stan
dards, the conclusion is inescapable that blacks were excluded
intentionally from hiring and promotions. This is not a simple
testing case. The department has admitted— and the evidence
supports that admission— that there was an exclusionary policy
which prevented blacks from being hired and promoted. Although
intentional discrimination by a public employer was not illegal
under Title VII before 1972, it was illegal under the Equal
227/ See especially Duck, supra, 503 F.2d at 300:
"We believe that the fact that the
entrance level examination employed by the
[police department] prior to November 1972
was a general intelligence test which
rewarded verbal skills and that in
addition to this test 30 percent of an
applicant's total score was derived from
a completely unstructured and totally
subjective interview, when considered with
the statistical imbalance of minority
representation on the police force, fully
supports the district court's determina
tion that racial discrimination existed
in the police hiring practices."
If the word "promotion" is substituted for "entrance level" and̂
"hiring," the above is almost an exact description of part of the
undisputed evidence of discrimination which the district court
disregarded in this case.
-61-
Protection Clause, 228/ the Civil Rights Act of 1866 (42 U.S.C.
1981), 229/ the Michigan fair employment laws, 230/ and Title VI
of the 1964 Civil Rights Act. 231/ Thus, the police department's
discrimination against blacks in hiring and promotions was clearly
unlawful. 232/
228/ See, e.g., Hazelwood, supra, 433 U.S. at 309-310 n. 15.
229/ See, Johnson v. Railway Express Agency, Inc., 421 U.S. 454
719 75); Jones- v! Alfred H. Maver Co., 392 U.S. 409 (1968); Carter
v. Gallaaher, 452 F.2d 315, 327 (8th Cir. 1971), certiorari denied,
406 U.S."950 (1972); Young v. Int'1 Telephone & Telegraph Co., 438
F.2d 757 (3d Cir. 1971); Waters v. Wisconsin Steel Works, 427 F.2d
476 (7th Cir. 1970), certiorari denied, 400 U.S. 911 (1970).
230/ Michigan's Fair Employment Practices Act, M.C.L.A. 423.301
et seq., was enacted in 1955 and specifically applied to public as
well as private employers. Id., Sec. 423.302(b).
231/ The district court found that the police department has "both
before and after adoption of the affirmative action plan" received
federal funds which were used in part to pay salaries of police
officers and sergeants. 446 F. Supp. at 1000, n. 61. See Section
604 of Title VI, 42 U.S.C. 2000d-3.
232/ This case does not present the "post-Act" problem of
Hazelwood . In that case, the United States was the sole plaintiff
and asserted only that the school district's hiring policies
violated Title VII; the government did not assert any other statu
tory basis for standing to sue. Cf. United States v. Solomon, 563
F .2d 1121 (4th Cir. 197 7) (holding that, absent statutory authoriza
tion, the United States generally lacks standing to assert Equal
Protection Clause violations). Accordingly, the Supreme Court in
Hazelwood , after finding a prima facie case of purposeful discrimi
nation, remanded for a determination of whether any discrimination
occurred which was subject to correction under Title VII the only
asserted basis for jurisdiction. 433 U.S. at 309-313.
In the case at bar, the Government is not suing the Detroit
Police Department under Title VII. The question is whether the
department may voluntarily remedy the effects of past intentional
discrimination, which was unlawful apart from Title VII. Indeed,
before 1972, private black plaintiffs could have sued the depart-_
ment under the Equal Protection Clause, Section 1981, Title VI and
the Michigan acts. See, e.g., Washington v. Davis, 426 U.S. 229
(1976); Pompev v. General Motors Corp., 385 Mich. 537, 189 N.W. 2d
243 (1971) .
-62-
Second, given the existence of prior intentional discrimina
tion against blacks in hiring and promotions, it is unnecessary to
determine whether unintentional discrimination against blacks was
also unlawful before Title VII became applicable to government
employment practices in 1972. However, in Long v. Ford Motor Co.,
496 F.2d 500 (6th Cir. 1974), this Court held that 42 U.S.C. 1981,
a statute which has independently prohibited employment discrimina
tion since 1866, 233/ reaches any employment practice which provides
"treatment similar on its face but dissimilar in its effects upon
racial minorities and unfounded on business necessity." Id., at
505, adopting the standard of Griggs v. Duke Power Co., 401 U.S.
424, 430-431 (1971). Accord, e.g., Johnson v. Ryder Truck Lines,
17 FEP Cases 571 (4th Cir. 1978); Davis v. Ccuntv of Los Angeles,
566 F.2d 1334 (9th Cir. 1977), certiorari granted, 46 U.S.L.W. 1553
(June 19, 1978). Furthermore, the Michigan Fair Employment Practices
Act prohibited racial discrimination in employment by public agencies
since 1955 (see note 230, supra). The Michigan Civil Rights
Commission, which has statutory authority to issue rules and
regulations to implement that Act, has published interpretative
guidelines which incorporate into state law the definition of
discrimination set forth in Title VII and Griggs v. Duke Power Co.,
233/ The Supreme Court first held in 1968 that the 1866 Civil
Rights Act reached private acts of racial discrimination. Jones
v. Alfred H. Mayer Co., supra. But the Court always recognized
that the 1866 Act prohibited racial discrimination by the State,
see Civil Rights Cases, 109 U.S. 3, 16-17 (1883).
-63-
sup r a . 234/ Thus, even if all of the evidence of overt discrimi
nation is put to one side, and only the use of discriminatory
written examinations is considered, it would still b;e the case
that the police department unlawfully discriminated against blacks
in hiring and promotions under Section 1981 and Michigan law.
For the above reasons, the evidence in this case, when
considered according to governing legal standards, establishes
that the Detroit Police Department intentionally and unlawfully
discriminated against blacks in hiring and promotion. Thus, the
district court erred in holding that the factual predicate for
race-conscious affirmative action was absent. We now turn to the
remaining issues in the case— whether an employer may voluntarily
adopt race-conscious remedies to rectify the continuing effects of
its past illegal discrimination (see Part III, infra ) and whether
the particular remedy adopted by Detroit is appropriate (see Part IV
infra) .
234/ These guidelines were originally published in 1972 and appear
in BNA Labor Relations Reporter, Fair Employment Practices Manual
455: 1091-1095. These guidelines prohibit the use of employment
tests which adversely affect minorities and are not shown to be
job-related, id., at 1091-1092, and adopt the EEOC testing guide
lines, id. at 1094. There appear to be no reported Michigan cases
dealing with this question under state law. However, in 1977, the
Fair Employment Practices Act was repealed and the Michigan Civil
Rights Act, M.C.L.A. 37.2101 et seq., was enacted. The new
Act's substantive provisions follow Title VII verbatim, which
evidently confirms the position of the Michigan Civil Rights
Commission that the definition of discrimination under state law
is the same as under Title VII.
-64-
III. THE DETROIT POLICE DEPARTMENY MAY
VOLUNTARILY INSTITUTE AFFIRMATIVE
ACTION TO REMEDY THE EFFECTS OF
ITS PAST EXCLUSIONARY POLICY______
As we have noted earlier (p. 53), the evidence of
past discrimination in this case is greater than that
presented ’in the Detroit Edison litigation. To correct
the effects of past discrimination by Detroit Edison,
the district court ordered the company, inter alia,
to promote at the ratio of one black for each white;
and this order was affirmed by this Court (see pp. 43-44,
235/
supra). The Detroit Police Department did not wait
to be sued; instead, the department recognized its legal
and moral duty to rectify its past wrongs and voluntarily
instituted the same remedy which was imposed on Detroit
Edison by court decree. Yet in this case, the district
court held that only the judiciary has the power to
235/ In affirming the 1:1 promotion ratio, this Court
held that it must be modified in two respects: only
qualified persons could be promoted and an ultimate
goal or time limit must be provided. We think that
a similar limitation is necessary in this case. See
PP. 88-90, infra.
-65-
remedy the effects of past discrimination and that an
employer may not do so voluntarily (446 F. Supp. at 1010).
236/
This holding is wrong for several fundamental reasons.
1. In combatting the evil of employment discrimination,
both Congress and the courts have emphasized that voluntary
compliance is the most effective and preferred method of
enforcement. See Section 706(b), (f)(1) of Title VII,
42 U.S.C. 2000e-5(b), (f)(1); Section 602 of Title VI,
42 U.S.C. 2000d-l; Alexander v. Gardner-Denver Co.,
415 U.S. 36, 44, 59 (1974); United States v. Allegheny-
Ludlum Industries, 517 F 2d 826, 846-848 (5th Cir. 1975),
certiorari denied, 425 U.S. 944 (1976). And the federal
236/ As noted earlier, the four justices who expressed
a clear view on this matter in Bakke held that employers
could voluntarily adopt numerical relief for past unlawful
discrimination (see pp. 46-47, supra).
Most of the lower courts faced with this question
have concluded that employers may voluntarily remedy
past unlawful discrimination with numerical goals.
Weber v. Kaiser Aluminum & Chemical Corp., 563 F.2d 216,
223 (5th Cir. 1977) (dictum); id. at 227, 229 (Wisdom, J.,
dissenting); Porcelli v. Titus, 431 F.2d 1254 (3d Cir.
1970), certiorari denied, 402 U.S. 944 (1971); Lindsay v.
City of Seattle, 86 Wash. 2d 698, 548 P.2d 320, certiorari
denied, 429 U.S. 886 (1976); Hutchinson Comm, v. Midland
Credit Management, 213 Kan. 399, 517 P.2d 158 (1973);
Germann v. Kipp, 429 F. Supp. 1323 (W.D. Mo. 1977),
vacated as moot, 17 FEP Cases 72 (8th Cir. 1978);
EEOC v. Contour Chair Lounge, 17 FEP Cases 309 (E.D.
Mo. 1978); Barnett v. International Harvester, 12 FEP
Cases 786 (W.D. Tenn. 1976).
(continued)
-66-
departments and agencies which have statutory responsibility
for enforcing the Civil Rights Acts have consistently
encouraged employers to voluntarily remedy past discrim
ination, through the use of appropriate numerical goals
when necessary. See, e_.g_. , Policy Statement on
Affirmative Action Programs for State and Local Govern
ment Agencies, 41 Fed. Reg. 38814' (1976); Proposed EEOC
Guidelines on Remedial and/or Affirmative Action, 42
Fed. Reg. 64826 (1977).
To be sure, Congress has equipped the district
courts with the authority to impose potent remedies
when voluntary compliance fails, but the very purpose
of these remedies is to "provide the spur or catalyst
which causes employers and unions to self-examine and
to self-evaluate their employment practices and to
endeavor to eliminate, so far as possible, the last
vestiges" of discrimination. Albemarle Paper Co. v.
Moody, supra, 422 U.S. at 417-418, quoting United
States v. N. L. Industries, 479 F.2d 354, 379
(continued)
Contra: Chmill v. City of Pittsburgh, 31 Pa.
Cmwlth. 98, 375 A.2d 841 (1977); Reeves v. Eaves,
411 F. Supp. 531 (N.D. Ga. 1976); Detroit Firefighters
Ass'n v. Citv of Detroit. 17 FEP Cases 186 (E.D. Mich.
1976), vacated as moot, 17 FEP Cases 190 (6th Cir. 1978).
-67-
(8th Cir. 1973). Yet the district court's holding in
this case precludes voluntary compliance and converts
court-ordered relief - a last-resort remedy under the
law - into the sole remedy. It is not only pointless
to require expensive and lengthy litigation before an
employer may correct its past wrongs; this approach
would inevitably delay the realization of equal
employment opportunity, and it is at odds with "the
central theme" of the employment discrimination laws.
Weber v. Kaiser Aluminum & Chemical Corp., 563 F .2d 216,
223 (5th Cir. 1977).
2. There is no merit in the district court's
suggestion that the equitable remedial discretion of a
federal court is broader than the ability of an employer
voluntarily to correct its own wrongs. When a public
agency has engaged in past racial discrimination, it is
"clearly charged with the affirmative duty" to eliminate
that discrimination "root and branch." Green v. County
School Board, 391 U.S. 430, 437-438 (1968) (emphasis added).
-68-
The public agency has the "primary responsibility" for
remedying its past discrimination, Milliken v. Bradley,
433 U.S. 267, 281 (1977) (emphasis in original); Brown v.
Board of Education, 349 U.S. 294, 299 (1955); and
judicial authority may be invoked only when that
affirmative obligation is not discharged voluntarily.
Milliken, supra, 433 U.S. at 281; Swann v. Charlotte-
Mecklenburg Board of Educ., supra, 402 U.S. at 15.
The district court's holding turns these basic
237/
principles upside down. Instead of placing the
primary responsibility for remedying discrimination on
the state employer which committed the wrong, the district
court would prevent the state from voluntarily discharging
its affirmative duty. A district court is empowered to
order an employer to rectify the effects of past dis
crimination only when this is necessary to bring the
employer into compliance with the law. There is no reason
that the employer may not voluntarily bring itself into
legal compliance without being ordered to do so by a court.
237/ The district court's holding could also lead to totally
illogical results. The lengthy and costly litigation in the
Detroit Edison case (which is still not entirely completed)
probably would have been avoided if the company had.volun
tarily adopted race-conscious remedies for its past dis
crimination. But under the district court's rationale,
such voluntary compliance would be illegal; the company
would have to be sued by the government or private
plaintiffs; and then the court could impose such relief.
-69-
3. We deal here with the executive branch of government
as a public employer. In many respects, "the discretionary power
of public authorities to remedy past discrimination is even broader
than that of the judicial branch," Associated General Contractors
of Massachusetts, Inc. v. Altshuler, 490 F.2d 9, 17 (1st Cir. 1973),
certiorari denied, 416 U.S. 957 (1974). A federal court's remedy
must be preceded by a finding of unlawful discrimination based on
facts elicited at an adversary hearing, Swann v. Charlotte-
Mecklenburg Bd. of Educ., supra, 402 U.S. at 16; and a judicial
remedy is limited by the extent of the finding of discrimination.
Dayton Bd. of Educ. v. Brinkman, 433 U.S. 406 (1977). But public
authorities may adopt race-conscious remedies without holding hear
ings or making findings of a judicial character, McDaniel v. Barresi,
402 U.S. 39, 41 (1971); Porcelli v. Titus, 431 F.2a 1254 (3d Cir. 1970),
certiorari denied, 402 U.S. 944 (1971); see United Jewish Organization
of Willi*™*burg v. Carey, 430 U.S. 144, 157 (1977) (plurality opinion);
Swann, supra, 402 U.S. at 16. Moreover, public and private employers
mav rectify the effects of their own past discrimination in some
situations which could not be reached by a federal court's decree.
For example, an employer may voluntarily reorder a bona fide seniority
system which perpetuates past discrimination, Franks v. 3owman Trans._
Co., supra, 424 U.S. at 778-779, even though that system might pass
judicial scrutiny. Teamsters v. United States, supra, 431 U.S. at
348-355.
4. That Detroit's affirmative action plan affects the
interests and expectations of some incumbent employees does not
disqualify it from appropriate voluntary compliance. All race-
conscious efforts by an employer, whether adopted voluntarily or
affect such interests and expectations. Thisby compulsion, can
-70-
is true even in the simplest case when an employer determines
that an applicant was rejected because of racial discrimination.
A convenient example is Rodney Brown, discussed in the district
court's opinion (446 F. Supp. at 999-1000). Mr. Brown, who is
black, applied to the Detroit Police Department in 1963 and was
rejected for being too short. The minimum height requirement was
5' 8 1/2" and Brown was 5'11". The department had no non-racial
explanation for this rejection; and courts, faced with this
evidence, would find purposeful racial discrimination. McDonnell
Douglas Corp. v. Green, 411 U.S. 792 (1973).238/ The department
surely could have decided to voluntarily remedy this wrong by
hiring Brown with retroactive seniority. Such action would, we
think, be considered enlightened voluntarily compliance with the
law, notwithstanding the effect such action would have on others
(e.g., the white applicant who was not hired because Brown took
the last place at the Police Academy, or the white incumbents whose
seniority positions were diluted.)239/
238/ Although the district court in this case refused to find that
Brown was rejected because of his race, 446 F. Supp. at 999-1000,
it entirely disregarded the decision in McDonnell Douglas. And i
strains credulity to hypothesize that some reason other than
racial discrimination caused Brown’s rejection. See also Furnco
Const. Co. v. Waters, No. 77-369, slip op. 9 (U.S., June 29,
1978) .
239/ This situation also presents an example where an employer's
efforts at voluntary compliance may exceed the remedial authority
of district courts. If Brown had sued the department^in 1974, it
could have successfully argued that his Title VII claim was fore-
closed because the Act did not apply to municipal employers until
1972; and it could have defeated his claims under 42 U.S.C. 19ol,
(continued)
-71-
5. In this case, Detroit concluded it had discriminated
against scores of blacks on a class basis, and it voluntarily
adopted numerical relief as a class remedy. There is no merit in
the district court's reasoning (446 F. Supp. at 1010) that an
employer who admits to past discrimination cannot be trusted to
voluntarily adopt numerical relief. If this reasoning were valid,
school boards which had practiced segregation could not voluntarily
rectify the effects of their past wrongs; nor could governmental
entities which had discriminated in voting adopt remedial measures.240/
And in stressing voluntary compliance as the central theme of the
Civil Rights Acts, Congress could not have viewed employers who had
practiced discrimination akin to criminal recidivists. The anti-
discrimination laws are designed to change ways of thinking and to
persuade employers, who had been hostile or indifferent to the
rights of minorities, to correct their past wrongs.
6. An employer that critically examines its employment
practices, and concludes that there has been unlawful discrimina
tion, may allocate its resources in two different ways. It may
acknowledge its duty to eliminate the present effects of
discrimination and institute necessary affirmative action to that
239/ (continued) 1983 by pleading the statute of limitations as
!n^affirmative defense/ But the?e is no rule of law which requires
employers to raise such defenses; and certainly an employer is no
precluded from voluntarily correcting a past wrong because
defenses might be available in a lawsuit.
240/ But see McDaniel v. Barresi, supra;̂
County School Board, supra; United Jewish
Swann, supra;
Organizations
Green v.
, supra.
-72-
end. Or it may resist change and instead retain a battery of
lawyers, statisticians and labor market and testing experts to
construct sophisticated defenses against meritorious claims. The
former course is consistent with the underlying purpose of the
Constitution and the Civil Rights Act--the elimination of unlaw
ful discrimination. The latter course will invite unnecessary and
costly litigation and will perpetuate, for an indeterminate
period, the status quo of unlawful discrimination which the Consti
tution and Civil Rights Act command must be eradicated "root and
branch."
In this case, the Detroit Police Department voluntarily
adopted race-conscious numerical goals to rectify the effects of
past discrimination. For the reasons stated above, such
affirmative action is a lawful form of voluntary compliance and
should be encouraged by the courts. This does not mean, of
course, that an employer has carte blanche authority to do
anything it wishes in the name of affirmative action. Just as
analogous remedies imposed by the courts must be within the scope
of sound judicial discretion, affirmative action undertaken volun
tarily must not be clearly inappropriate under all of the circum
stances. We examine, therefore, the appropriateness of the
particular remedial steps undertaken by the department.
-73-
IV. THE PARTICULAR RACE-CONSCIOUS REMEDY ADOPTED
BY THE DETROIT POLICE DEPARTMENT DID NOT VTO-
LATE THE LEGAL RIGHTS OF PLAINTIFFS___________
A. A Promotion Ratio is an Appropriate Method
of Remedying Past Discrimination____________
To remedy the residual effects of past discrimination, the
police department adopted a racial classification whereby pro
motions to the rank of sergeant were made at a ratio of approxi
mately one qualified black for each qualified white. The use of
numerical race-conscious relief is appropriate in this situation.241/
Because of discrimination in hiring and promotion, the existing
work force of the police department is not composed as it
would have been had constitutional and statutory standards of non
discrimination been maintained. In such a discriminatorily created
work force, only a limited number of vacancies open up at the sergeant's
rank periodically. A policy of promoting on a racially "neutral"
basis, when superimposed on the existing racially shaped pattern,
would not effectively remedy the discriminatory system. Instead,
it would perpetuate the effects of past discrimination. While it
has been suggested that one "pure" and certain method of restoring
eauality of opportunity would be to displace all incumbents who
were the beneficiaries of the past discriminatory system and require
them to compete with all applicants for each position, this solution
241/ " [W]hile quotas merely to attain racial balance * * * are
forbidden, * * * quotas to correct past discriminatory practices
are not * * *." United States v. Lathers, Local 46, 471 F.2d 40 8
(2d Cir. 1973), certiorari denied, 412 U.S. 939 (1973).
-74
would have the harshest and most dramatic possible impact on
the interests of incumbents. As an alternative to this "freedom
now" approach, 242/ the federal courts have approved the
accelerated selection of qualified blacks on a numerical basis
(see pp. 43-45, supra). 243/ For the reasons stated in Part III
of this Brief (pp. 64-73, supra), an employer may voluntarily adopt
such a remedy for discrimination. And such relief is not limited
to the hiring level. Where, as here, there was past discrimination
in promotions, this Court has upheld numerical promotion relief, 244/
as have the Third, Fifth, Seventh and Eighth Circuits. 245/ More
over, it is well-settled that such relief need not be limited to
242/ See Papermakers & Paperworkers, Local 189 v. United States,
TT6 F .2d 980, 988 (5th Cir. 1969), certiorari denied, 397 U.S. 919
(1970).
243/ The Supreme Court's decision in Bakke is fully consistent
with these holdings (see pp. 45-47, supra).
244/ The Detroit Edison case, pp. 43-44, supra.
245/ EEOC v. A.T.&T. Co., 556 F.2d 167 (3d Cir. 1977),
certiorari denied, 46 U.S.L.W. — (July 3, 1978); Watkins v.
Scott Paper Co., 530 F.2d 1159 (5th Cir. 1976), certiorari
denied, 429 U.S. 861 (1976); United States v. City of Chicago,
549 F.2d 415 (7th Cir. 1977), certiorari denied, 434 U.S. 875
(1977); United States v. N. L. Industries, Inc., 479 F.2d 354,
377 (8th Cir. 1973) .
In Kirkland v. New York State Dept, of Correctional Services,
520 F.2d 420 (2d Cir.), reh. denied, 531 F.2d 5 (1975), certiorari
denied, 429 U.S. 823 (1976), the Second Circuit upheld an interim
promotion quota but struck down its use after the development of
a new testing system. It is unnecessary for this Court to become
engaged in the conflict of legal theory which has embroiled
different panels of the Second Circuit. Compare Kirkland with
(continued)
-75-
identifiable victims of discrimination. 246/
B. A Promotion Ratio Is Appropriate to Eliminate
the Effects of Past Discrimination Even if the
Department's Recent Promotional Models Have
Been Validated________________ __________________
All of the persons promoted to sergeant under the affirmative
action plan, both black and white, were fully qualified, and were,
in fact, more qualified than those promoted a decade earlier (see
p. 38, supra). However, the department passed over a number of
whites on the eligibility list to meet the promotion ratio, and
the district court found that these whites were "more qualified
than the blacks who were lower on the list because the promotion
models in effect between 19 73 .and 19 77 were valid (446 F. Supp. at
993-994). This .finding does not detract from the legality of the
department's affirmative action to remedy the present effects of
past discrimination.
Patterson v. Newspaper Deliverers' Union, 514 F.2d 767, 773-775
(2d Cir. 1975). The apparent premise of Kirkland— that race
conscious relief which affects incumbents must be more circumspect
than relief which affects applicants for initial employment— was
clearly rejected by the Supreme Court in Franks v. Bowman Trans. Co.,
supra, 424 U.S. at 775, and has not been followed by this or any
other Circuit. Moreover, Kirkland is factually distinguishable
because of the absence of proof in that case of "egregious" past
discrimination. See 520 F.2d at 427.
246/ See e.g., the line of cases decided by this Court upholding
numerical relief, pp. 43-44, supra; EEOC v. A.T. &T. Co. , supra,
556 F.2d at 175-177, 180; Carter v. Gallagher, supra, 452 F.2d
at 330-331. Moreover, "incumbent employees will be affected
identically by a remedy in favor of identifiable victims of
specific discrimination as by a remedy which includes employee
members not so identifiable. The impact on incumbent employees
goes to the scope rather than the availability of class relief.
EEOC v. A.T.&T. Co., supra, 556 F.2d at 177 (footnote omitted).
-76-
1. Assuming arguendo that the district court's validation
finding is correct (but see point 2, infra), it would follow that
the police department had ceased overt discrimination against blacks
in 1973. But the development of non-discriminatory selection proce
dures did not relieve the department of its affirmative duty to
eliminate the present effects of past discrimination. Non-discrimi
nation and affirmative action are complementary, and not mutually
exclusive, remedies. The department's affirmative duty was to
"eliminate the discriminatory effects of the past as well as bar
like discrimination in the future." Louisiana v. United States, 380
U.S. 145 , 154 (1965). As we have stated above (p. 74, supra) ,
racially "neutral" promotions superimposed on a work force distorted
by past discrimination would perpetuate that discrimination and would,
indeed, give a continuing and indefinite advantage to whites who have
already benefitted from that discrimination. If promotions had been
made in rank order since 1974, the sergeants' rank would have been
only 7.1 per cent black by the end of 1977 (see p. 38, supra). This
would have preserved the discriminatory status quo, and past injustices
would be perpetuated anew in the course of rejecting black applicants
for promotion who met standards much more rigorous than those required
of whites when overt discrimination and non-job related tests were
the rule. Cf. United States v. Duke, 332 F.2d 759, 768 (5th Cir.
1964) .
The law in this Circuit is settled that numerical relief is
appropriate to eliminate the effects of past discrimination, even
when that relief requires an employer to select out of order from
a validated hiring or promotion list. This Court so held in Arnold
v. Ballard, 12 FEP Cases 1613 (6th Cir. 1976), vacated on other
-77-
grounds, 16 FEP Cases 396 (6th Cir. 1976). In United States v.
I.B.E.W., Local 212, 472 F.2d 634 (6th Cir. 1973), this Court
upheld numerical relief to remedy past discrimination, even though
non-discriminatory and valid selection procedures were in place. 247/
And in United States v. Masonry Contractors Ass'n of Memphis, Inc._,
497 F.2d 871 (6th Cir. 1974) and EEOC v. Detroit Edison Co., supra,
this Court affirmed district court orders requiring the development
of nondiscriminatory selection procedures and the institution of
numerical relief for past discrimination. 248/
The Detroit Police Department undertook voluntarily what the
federal courts ordered employers and unions to do in the above cases:
to eliminate future discrimination by developing validated selection
procedures, and to eradicate the effects of past discrimination by
using a racial classification for promotions. These complementary
efforts are fully consistent with the department's affirmative duty
under the Constitution and the Civil Rights Acts.
2. In any event, the district court's finding that the
promotion ratio resulted in "better qualified" whites being passed
over should be set aside.
(a) Rank order on the eligibility list is determined by a
composite weighting of scores on the written test, service ratings,
247/ In the Local 212 case, the district court initially entered
detailed orders to end continuing discrimination but refused affirma
tive relief. Dobbins & United States v. I.B.E.W., Local 212, 292 F.
Supp. 413 (S.D. Ohio 1968). After some four years of operating under
court-approved and supervised standards, the union was ordered to
adopt numerical relief to overcome the effects of its pre-1968 dis
crimination, United States v. I.B.E.W.,_Local 212, 5 E.P.D. Para.
8428 (S.D. Ohio 1972), and this order was affirmed, 472 F.2d at 636.
248/ See United States v. Bricklayers, Local 1, 5 E.P.D. Para. 8480
Tw 7d . Tenn. 1972-1973), aff’d, 497 F.2d at 878; Stamps v. Detroit
Edison Co., supra, 365 F. Supp. at 123-124 paragraphs 9, 11, 12,
aff'd, 515 F.2d at 317.
-78-
oral board, seniority, college credits and veterans' preference.
There is no evidence that the various weights assigned to each
component, and the resulting rank order scores, are justified. On
the contrary, Caretti explained that due to labor negotiations he
was able to make only minor adjustments in the model which he
inherited in 1969; plaintiffs' experts, Drs. Wollack and Ebel, did
not testify that the weighted model was job-related; and Dr. Inn, who
had been retained by D.P.O.A., concluded that the weightsvere arbi
trary and that rank order did not measure relative qualifications 249/
(see pp. 30-31, supra). The district court's finding that the promo
tion models for 1973-1976 were job-related is clearly erroneous;
indeed, it is based on non-existent testimony.
In his written presentation to the Court (PX 256) , Dr. Wollack,
plaintiffs' expert, emphasized that the various components of a
selection model must be appropriately weighted according to their
relative importance as identified in a job analysis, in order for
the model to be job-related (id. at 258-259);
"An employment system which is broadly-based
must provide for a methodology to determine
the candidate's eligibility in a manner
which is reflective of this broadly-based
249/ Dr. Inn concluded specifically (PX 298(a), p. 22):
Consider the white officers who feel
that they have been unduly discriminated
against because they were skipped over
in the promotional process despite their
higher rank order scores. I have tried
to suggest that in terms of composite
scores these persons are indistinguishable
from the promoted black applicants who
had lower rank order scores. The rank
order scores convey a false illusion of
overall superiority or inferiority.
-79-
assessment. Each required capability or
attribute which is measured in the employ
ment process must be factored into the
determination of the candidate's eligibility,
and this must be done by a method which is
consistent with the relative importance of
each characteristic measured. Weights must
be assigned to the assessable performance
dimensions or personal characteristics which
have been identified in the job analysis and
in a manner which is proportionate to their
importance. The proposed weighting rationale
differs from traditional civil service
procedures which are technique oriented.
For example, often written examinations
are given some weight, say 507o in the
selection process, while an oral board
might also be weighted 507o. Under the
proposed employment system, it is the
performance dimensions or job analysis
factors which are weighted; so the
proposed system is job analysis oriented
rather than technique "oriented. The
proposed approach, which Consultants
call "Total Personnel Assessment"
(TPA) requires that many selection
tools be combined in a systematic
fashion to get the most realistic
estimate of an applicant's quali
fications on each required job factor.
An applicant's overall standing on a
final eligibility list is determined
by summing the ratings which he obtains
on each factor. Some of these factors
may be measured by written examinations
While the results of written examinations
may contribute to the applicant's over
all evaluation, the TPA process utilizes
such information in a way as to give
those test scores an appropriate degree
of weighting in the total assessment
of each applicant's capabilities. In
this fashion, an applicant's position
on the eligibility list reflects a
valid, comprehensive assessment of
all relevant measurable job analysis
factors. The weighting process assures
broad-based assessment and a balanced
and appropriate use of the selection
tools. "
-80-
The police department's 1973-1976 promotion models were not
developed by corresponding weights in the model to a job analysis.
Indeed, as Caretti admitted, he could not make major changes in
the model which he had inherited and did not even have a professional
job analysis with which to work (see pp. 24, 29, supra). For this
reason alone, the promotion models are not job-related.
The absence of an appropriate weighting system for determining
rank order on the promotion lists is particularly important in the
context of this case. As Dr. Wollack also emphasized in his written
presentation, proper weighting is essential "to reduce inappropriate
adverse effect against protected classes which acrues from the
inordinate weighting of certain select components in the process"
(PX 256, p. 259) (emphasis in original). For example, Dr. Wollack
compared the 1974 written examination results of the whites who
were passed over with the blacks who were promoted out of order for
affirmative action purposes. He found that, as a group, the whites
scored better on 70 per cent of the written examination questions
(446 F. Supp. at 991 n. 15). This represents the sum total of the
empirical evidence in the record that the whites passed over were
"better qualified" than the blacks promoted out of order. 3ut
Dr. Wollack admitted that he did not examine the comparative perfor
mance of these two groups on the other components of the promotion
model. 250/ Such an examination paints a far different picture.
250/ Tr. 10/14/77, 65-68 (Dr. Wollack). Indeed Dr. Wollack testi
fied that he was not claiming that this group of blacks were less
qualified than the whites who were passed over - only that they
performed poorer on the writted test. Id., 67.
-81-
In Appendices A and B to this Brief, we list the names of
the whites who were passed over, and the blacks promoted out of
order, from the eligibility list derived from the 1974 examination.
*
We also list each person's score on each component of the promotion
model - i.e., written examination, oral board, service ratings,
seniority, college credits and veterans' preference. In Appendix C,
we present the total and average scores for the two groups on each
component. The following table presents the difference in average
scores (based on a uniform scale of 0-100) of the two groups on each
component of the 1974 selection model. (A "+" in the last column
indicates that the average score of the blacks promoted out of order
is greater than the average score of the whites passed over; a
indicates the reverse.)
Component
Average Score
of Blacks
Average Score
of Whites Difference
Written Test 79.43 86.24 - 6.81
Oral Board 75.06 70.10 + 4.96
Service Rating 80.13 82.74 - 2.61
Seniority 84.50 89.00 - 4.50
• College Credit 43.00 34.00 + 9.00
Military Service 40.50 22.50 + 18.00
A determination of whether the blacks promoted out
order were more, less, or equally qualified as the whites
were bypassed clearly turns on which component of the promotion
model one examines. Of the three components which the district court
found to be job-related (see 446 F. Supp. at 989-992), blacks did sig
nificantly worse on the written test, significantly better on the
-8 2-
oral board and slightly worse on the service ratings.
0
In addition, this group of blacks had less seniority
but much more college education and military experience.
Thus, the blacks promoted out of order scored
better, on the average, on three of the six components
than the whites who were passed over. Nevertheless,
when rank order scores were computed on the basis of
the arbitrary weights used on the model, the average
rank order scores of this group of blacks was 4.1
points less than this group of whites (see Appendix C).
The reason for this is that the one job-related component
on the model on which whites scored substantially better
(the written test) was weighted 65 per cent, while the
one job-related component on which blacks scored sub
stantially better (the oral board) was weighted only
10 per cent. And of the remaining four components,
seniority and service ratings were weighted 6 and 15
per cent respectively, and college credits and military
service were each weighted 2 per cent (446 F. Supp. at
989).
251/ The slightly higher average service ratings of the whites
is due to the facts that they have greater seniority and that
there is a tendency of raters to give more senior police officers
higher ratings. See 446 F. Supp. at 992; PX 240 (minutes of
7/24/74 meeting, p. 12). When blacks and whites of the same
sniority are compared, their service ratings are the same on
the average. See 446 F. Supp. at 992; PX 191.
251/
-8 3-
It is evident that the comparative rank order
scores of these two groups are highly sensitive to
the weights assigned to each component on the model.
If, for example, the weights on the written test and
oral boards were reversed, with the former counting
10 per cent and the latter 65 per cent, with the other
weights unchanged, this group of blacks would have
had an average rank order score which was 2.4 points
■ 252/
greater than this group of whites. If the written
test were weighted 30 per cent and the oral boards
253/
45 per cent, with no other change in the model,
the average rank order scores of the two groups would
have been identical. And, as a final example, if the
department had used the weights which it proposed in
254/
labor negotiations, the average rank order scores
of the blacks would be slightly higher (0 .6 points)
than the whites.
252/ Each of the calculations in this paragraph is presented
in Appendix D to this Brief.
253/ In his written presentation, Dr. Wollack proposed
through a job analysis study that, for the police selection
model which he prepared, the written and physical performance
tests should have a combined weight of 35 per cent, while the
oral interview should be weighted 65 per cent. PX 256, p. 261.
254/ See note 154, supra (written test 32 per cent, oral
board 32 per cent, service rating 20 per cent, college
credit 14 per cent, veterans' preference 2 per cent, and
no weight for seniority).
-84-
Of course, we do not suggest that any of the hypothetical
weights used for the computations in the above paragraph are any
less arbitrary than the weights actually used by the department.
Without using a professional job analysis, it is impossible to
validly weigh the various components according to their actual
relative importance in measuring job-related skills and abilities.
Nor do we suggest that the blacks who were promoted because of
the affirmative action plan were more, less or equally qualified
than the whites who were passed over. On the basis of the present
record, there is simply no way of validly comparing the relative
qualifications of these two groups. Since the composite promotion
model has not been shown to be job-related, and since rank order
is a function of arbitrarily fixed weights, the district court's
finding concerning relative qualifications must be set aside.
(b) On the basis of the present record, the district
court's findings concerning the content validity of the 1973,
1974 and 1976 tests are also clearly erroneous. We fully agree
with the district court that Commander Caretti attempted, to the
best of his ability, to make these tests as content valid as he
could. But the evidence does not support the district court's
conclusion that Caretti succeeded.
First, no one testified that the 1976 test is content
valid. Caretti refused to express an opinion on the content
validity of this test (as well as of the 1973 and 1974 tests),
and plaintiffs' experts neither examined nor testified about the
1976 test (see p. 26, supra).
-85-
Second, Caretti acknowledged the difficulties which
he faced in attempting to develop content valid tests in 1973
and 1974. In particular, Caretti recognized that the only appro
priate way to develop a content valid test is to first do a
professional job analysis and then relate the test to the job
analysis (see p. 24, supra). And both professional testing
standards and the case law uniformly hold than an adequate job
analysis is an indispensable prerequisite for developing a
content valid test. 255/ It is, indeed, the "cornerstone" for
developing a content valid test, without which "a test construc
tor is aiming in the dark and can only hope to achieve job
relatedness by blind luck." 256/
The district court found that the 1973 and 1974 tests were
constructed on the basis of a job analysis prepared by Caretti and
John Furcon in 1973 (446 F. Supp. at 990) . This crucial find
ing is clearly erroneous. The 1973 job analysis was prepared
solely by Furcon, and Caretti admitted that he did not understand
it and therefore prepared the 1973 and 1974 tests without "the
benefit of a professional job analysis" (see p. 24, supra). 257/
255/ See, e.g., American Psychological Association, STANDARDS FOR
EDUCATIONAL AND PSYCHOLOGICAL TESTS 28, 45-46 (1974); 29 C.F.R.
1607.5(a) (EEOC testing guidelines); United States v. City of
Chicago, 573 F. 2d 416, 425 (7th Cir. 19 78); Firefighters Institute
for Racial Equality v. City of St. Louis, 5T9 F .2d 506, 511 (8th
Cir. 1977), certiorari denied, 434 U.S. 819 (1977); Vulcan Society
v. New York Civil Service Comm., 360 F. Supp. 1265, 1274 (S.D.N.Y.
1973), aff'd in rel. part, 490 F.2d 387 (2d Cir. 1973).
256/ Kirkland v. New York State Dept, of Corrections, 374 F. Supp.
1361, 1373 (S.D. N.Y. 1974), aff'd in rel. part, 520 F.2d 420 (2d
Cir. 1975), certiorari denied, 429 U.S. 823 (1976).
257/ Caretti's reliance on his own prior experience as a sergeant is
clearly an inadequate substitute for a professional job analysis.
See e.g., Kirkland, supra, 374 F. Supp. at 1375-1376.
-86-
Caretti also admitted that he was constrained to use a pre
selected bibliography from which he could choose questions and
that some sections of the test dealt with knowledge which would
surface on the job but some others would not (see pp. 24-25, supra).
We are unaware of any decision which has found a test constructed
under these circumstances to be content valid.
Third, " [f]or a test to be content valid, its content must
closely approximate tasks to be performed on the job. * * * [T]here
must be a correlation between the importance of a job function as
determined by the job analysis and the weight given to this func
tion on the examination." 258/ It would surely have been "blind
luck" for Caretti to have accomplished this without a job analysis
and given the constraints under which he was operating. And there
is no empirical evidence in the record that Caretti succeeded.
The district court's conclusions concerning validation
and relative qualifications cannot be supported on this record.
However, for the reasons stated in point 1, supra (pp. 76-77 ) r
even if these findings are sustained, the police department's
accelerated promotion of blacks to remedy past discrimination
would be lawful.
258/ United States v. City of Chicago,- supra, 573 F.2d at
425-426. Accord, e.g., Firefighters Institute, supra, 549 F.2d
at 511-512. Kirkland, supra, 374 F. Supp. at 1378; Vulcan
Society, supra, 360 F. Supp. at 1274; APA Standards, supra at
29; EEOC Guidelines, supra, 29 C.F.R. 1607.5(a).
As plaintiffs' expert, Dr. Wollack, has written: "The
degree of similarity or correspondence between the content
domain of the test and that of the job is, of course, the
most essential aspect of content validity." PX 256, p. 146.
-87-
C. The Particular Promotion Ratio (50
Percent) Was Appropriate, But An
Ultimate Goal Must 3e Set on Remand
The plaintiffs' position throughout this litigation has been
that any raciallv-based promotion ratio was unlawful. The district
court accepted this argument. Neither the plaintiffs nor the dis
trict court have suggested that the particular numerical ratio used
by the department (one qualified black for each qualified white) was
excessive. And we think that such an argument would be without merit.
The Board of Police Commissioners considered alternative ratios and
selected 50 per cent because any greater ratio would be too disruptive
while any lesser ratio would be too slow in remedying the effects of
past discrimination (see p. 36, supra).. The ratio chosen does not
invidiously exclude whites as a class, and it is comparable to pro
motion ratios which federal courts have ordered employers to utilize
in similar factual circumstances. 259/ Thus, we think that a 50 per
cent ratio was within a zone of reasonableness which Detroit could
voluntarily adopt to remedy past discrimination.
259/ See, e.g., the Detroit Edison case, suora, pp. 39-40
(50 per cent promotion ratio); EEOC v. A.T. & T. Co., supra
(absolute preference for promotions when necessary to meet
under-utilization goals); United States v. City of Chicago,
supra (40 per cent promotion ratio); United States v. N._L.
Industries, supra (50 per cent promotion ratio);
Patterson v. Newspapers Deliverers Union, supra (50 per
cent promotion ratio); Reed v. Lucas, 11 FEP Cases 153
(E.D. Mich. 1975) (ordering the Wayne County Sheriff's
Department to promote on a 1:1 basis); United States v.
United States Steel Corp., 5 FEP Cases 1253, 1259-1260
(N.D. Ala. 1973) (33 per cent promotion ratio), aff'd
in rel. part, 520 F.2d 1043 (5th Cir. 1975), certiorari
denied, 429 U.S. 817 (1976): Officers for Justice v.
Civil Service Commission, 371 F. Supp. 1328 (N.D. Cal.
1973) (50 per cent promotion ratio); Bolden v. Pennsylvania
State Police. 73 F.R.D. 370 (E.D. Pa. 1976) (33 per cent
promotion ratio).
-83-
However, a much more serious problem, in our view, is that
the police department has never placed a fixed time limit or ulti
mate goal on the promotion ratio. See EEOC v. Detroit Edison Co.,
supra, 515 F.2d at 317 (holding that a time limit or ultimate goal
must be placed on the district court's 50 per cent promotion ratio).
The use of a racially-based promotion ratio is appropriate only until
the effects of past discrimination have been eliminated; after that
has been accomplished, the use of race as a criterion in promotions
must cease.
The Board of Police Commissioners has properly viewed the
affirmative action program as a temporary measure which will end
when past discrimination has been remedied (see pp. 36-37, supra).
However, the 3oard has never declared when this result will occur.
The setting of an ultimate goal is crucial because this will deter
mine the number of white police officers who will be adversely
affected by the remedy. 260/ Since the object of the remedy is to
eliminate the racial imbalance at the sergeants' rank caused by past
discrimination, it is therefore essential to determine, as accurately
as possible, what proportion of the sergeants' rank would have been
black absent discrimination. See Rios v. Steamfitters, Local 638,
501 F .2d 622 (2d Cir. 1974).
The appropriate ultimate goal should be set on remand because
this issue was not litigated. The district court did conclude that
the appropriate labor force was 13.6 per cent black, but this rigure
was used only as a benchmark for determining whether past discrimination
260/ On the other hand, the ratio on promotions only determines how
fast the ultimate goal will be reached.
occurred and is probably a serious underestimate. 261/ And even
if the 18.6 per cent figure were found to be the appropriate goal,
plaintiffs would be entitled to prospective relief only because the
sergeants' rank is now only 15.1 per cent black (seep. 38, supra).
Finally, on remand both the police department and the dis
trict court should consider the changed circumstances which have
occurred as a result of the City's affirmative action efforts. In
1974, the sergeants' rank was only 5.1 per cent black; and promotions
made in rank order between 1974 and 1977 would have corrected this
imbalance only marginally (see p. 38, supra). Consequently, there
was a compelling need for accelerating the promotions of qualified
blacks. Now, however, there has been considerable progress; and the
police officer rank is on the order of 40 per cent black. By further
improving the written examination and/or by modifying the weights on
the oromotion model, the department may now be able to promote almost
as many black sergeants with a racially-neutral promotion process as
by continuing its 50 per cent ratio. In light of these changed cir
cumstances, there may no longer be a remedial need to use a racial
criterion for promotion to sergeant. Cf. Patterson v. American
261/ See note 200, supra. Because the police department
hired applicants who lived in the Detroit SMSA until
1974 (_i. e., during the period of discrimination) we
think it is untenable for the City to maintain that
the ultimate goal should be based on the population
of Detroit alone (50 per cent black). However, the
district court's use of the unweighted SMSA (18.6 per
cent black) is incorrect because the job is located in
Detroit and police officers were required to live in
the City after they were hired. See note 200, supra.
Faced with a similar question in Detroit Edison, the
district court set an ultimate hiring goal at 30 per
cent black, which was between the Detroit City and
Detroit SMSA proportions. Stamps v. Detroit Edison,
supra, 365 F. Supp. at 122 n. 4. See Associated
General Contractors v. Altshuler, supra, 490 F.2d at
14 n. 11, 18-19. The district court should consider
this solution on remand.
-90-
Tobacco Co., 535 F.2d 257, 274 (4th Cir. 1976), certiorari denied,
429 U.S. 920 (1976).
For the above reasons, the police department's affirmative
action promotions did not violate the rights of plaintiffs, but
future promotions to sergeant may threaten legal violations, depend
ing on what time limit or ultimate goal is appropriate. Since this
question was not decided by the district court, it should be addressed
by that court in the first instance. And, in light of the resolution
of that question, the district court should also decide whether plain
tiffs are entitled to a prospective injunction against future race
conscious promotions.
CONCLUSION
The judgment and orders of the district court should be
reversed and the case remanded for further proceedings.
Respectfully submitted
DREW S. DAYS, III
Assistant Attomev General
ABNER W. SIBAL,
General Counsel,
Equal Employment
ROBERT J. REINSTEIN,
Opportunity Commission
Washington, D. C. 20506
JOHN C. HAMMOCK,
Attorneys,
Department of Justice,
Washington, D. C. 20530
APPENDIX A
Following is a list of all white male police officers who
were on the eligibility register for sergeant derived from the
1974 examination and who were passed over as a result of the
affirmative action plan. Their names and ranks on the eligi
bility register are taken from Plaintiffs' Proposed Findings
of Fact and Conclusions of Law, p. 96 (Appendix III) (Dist.
Ct. Docket #162, filed December 22, 1977). We list each
person's written examination score, service rating (performance
evaluation), oral board score, seniority credits, college
credit, veterans' preference and rank order score. This infor
mation is taken from PX 195.
A - 1
1974 WHITES
RANK NAME WRIT
PERF
EVAL
71 Buterakos, G. 89 75.25
72 Hunn, T. 84 89
73 Lindstrom, P. 92 76
75 Connell, R. 89 80.5
77 McGuire, H. 85 85.25
78 Barba, G. 89 86.5
79 Gratopp, W. 89 87.5
80 Saad, M. 83 86.5
81 Kowalczyk, B. 87 83
82 Sauvage, C. 87 81.5
83 McLean, R. 84 80.5
84 Moore, A. 89 85
85 Hall, J. 84 87.5
86 Purrlngton, J. 86 88.25
87 Bajerclus, A. 86 79.5
88 Balmas, B. 93 77.25
89 Laube, P. 88 80.25
91 Frelgruber, E. 89 83
92 Rachas, R. 87 80
93 Htx, C. 86 85
94 McKinnon, D. 88 82.5
95 Pierce, R. 90 77.75
96 Morris, R. 87 82.5
97 Sleloff, N. 86 80.25
98 Schlhl, K. 86 79.5
99 Petrimoulx, K. 85 89.5
100 Thompson, J. 86 84.75
1 0 1 Howell, L. 91 78.75
105 Lashbrook, J. 85 83.5
106 Rlzk, R. 85 83.0
PASSED OVER
ORAL MIL. RANK ORDER
BOARD SEN. COL. SERV. SCORE
42.67 5.17 .5 _ — 83.34
38.33 5.63 — 2 .0 83.24
29.67 4.08 2 . 0 * - - 83.21
35.67 6 . 0 - - — 83.05
34.33 6 . 0 2 .0 - - 82.90
30.33 6 . 0 — — 82.89
28.67 5.17 1 . 0 — 82.87
40.67 5.63 2 .0 — 82.68
34.67 4.67 .5 1.58 82.68
32 5.5 2 .0 — 82.67
41.33 5.71 - - 2 .0 82.65
30 6 . 0 - - - - 82.6
38.67 6 . 0 1 . 0 - - 82.45
31.67 5.46 1.5 — 82.43
35.33 5.54 .5 1.5 82.43
35.67 2.17 1.08 - - 82.42
37.33 5.17 .5 - - 82.37
34 4.25 1 . 0 — 82.35
36.33 6 . 0 .5 - - ' 82.31
38.33 6 . 0 .5 - - 82.31
33.67 6 . 0 — — 82.30
40.67 3.5 .5 - - 82.29
31.67 6 . 0 1 . 0 - - 82.25
39.33 5.67 .5 .17 82.14
41.33 5.38 .5 — 81.97
37 5.88 — — 81.95
33.67 6 . 0 .5 — 81.84
36 3.58 - - — 81.74
31.33 5.71 — 1.92 81.67
34.67 6 . 0 1.0 - - 81.63
RANK NAME WRIT
PERF
EVAL
107 Keck, J. 83 79.5
108 Joseph, M. 86 74.25
109 Parlow, W. 87 84.75
110 Coracl, V. 88 80.75
1 1 1 Kopec, D. 90 75.5
1 1 2 Martin, K. 84 79.5
114 Collier, T. 87 78.5
115 Eveleth, C. 84 85.5
116 VanDerMeulen 83 87
118 Obidzinski, R. 84 82
> 119 Mackenzie, D. 82 88.5
• 120 Birdseye, W. 86 85.5
w 1 2 1 Cockell, J. 78 89.25
122 Graves, R. 85 89
123 Craddock, J. 86 81
124 Yoakum, H. 79 87
125 Hudson, R. 85 80
126 Marvin, R. 88 81.25
127 Mestdagh, R. 86 86.5
ORAL
BOARD SEN. . COL.
40.67 5.29 .5
34 5.29 .5
31.33 4.08 .5
29.33 4.33 .5
26.67 4.25 2.
39 6 . 0 1.0
38.33 4.83 .5
31.67 6 . 0 1.5
35.67 6 . 0 1.0
32.67 5.0 1.0
40 6 . 0 .5
26.67 6 . 0 1.0
42.33 6 . 0 .5
35.67 5.29 mm mm
33.33 5.79 .5
42.67 6 . 0 --
39 3.92 1.96
23.33 6 . 0 —
30.33 5.92 —
MIL RANK ORDER
SERV. SCORE
1.83 81.62
1.83 81.45
1.33 81.43
1.42 81.42
- - 81.40
- - 81.32
— 81.32
- - 81.25
- - 81.13
1.67 81.10
- - 81.07
— 81.05
2 .0 81.05
— 81.02
- - 81.00
2 .0 80.93
mm mm 80.93
.83 80.88
— 80.86
APPENDIX B
Following is a list of all black male police officers who
were promoted out of order, as a result of the affirmative
action plan, from the eligibility list derived from the 1974
examination. The last white male promoted from this list was
ranked #70, while the last black male promoted was #342 (446
F. Supp. at 988). The list which follows is taken from PX 195.
Listed is the name of each black male between #70 and #342 on
the eligibility register who was promoted and his rank on the
register, written examination score, service rating (performance
evaluation), oral board score, seniority credits, college credits,
veterans' preference and rank order score.
B - 1
1974 BLACKS
RANK NAME WRIT
PERF
EVAL
74 Green, R. 85 85.5
76 Rucker, N. 87 81.75
90 WllSon, D. 84 80.25
102 Ma rch, J. 81 86.75
103 Stewart, R. 85 83.25
104 McClendon, S. 87 73.75
113 Jones, L. 85 81.75
143 Johnson, J. 81 80.75
151 Grant, J. 81 80.5
154 McBride, P. 84 78
158 Packnet, M. 88 76
159 Smith, H. 89 74.75
167 llyman, W. 84 84.75
169 Bullock, K. 81 85.25
189 Donaldson, C. 76 84.5
191 Cook, R. 80 82.5
192 Baker, R. 79 78
194 Sanders, C. 76 86.5
195 Anderson, N. 84 82
199 Williams, B. 84 74
201 Bell, W. 84 74.75
202 Scott, H. 81 78
213 Williams, A. 79 87.5
214 Mays, D.
O'Bannon, C.
77 82.75
229 77 83.5
234 Bush, H. 78 80.5
237 Simmons, H. 75 82.25
242 Hubbard, W. 77 82.5
PROMOTED OUT OF ORDER
ORAL
BOARD SEN. COL.
MIL
SERV.
RANK ORDER SCORE
43.67 5.79 .5 83.09
38.67 5.25 1.0 .25 83.04
39 5.17 1.0 1.75 82.35
40.33 6 . 0 2 .0 — 81.72
38.67 5.75 .5 — 81.72
40.33 6 . 0 - - — 81.67
42 3.33 1.0 1.08 81.32
40.33 5.5 1.0 1.08 80.40
38 5.75 2 .0 - - 80.07
45 3.67 1.0 — 79.97
37.67 3.17 .5 .08 79.88
33.67 3.08 1.0 — 79.87
38.33 3.08 1.54 — 79.59
26.67 5.71 1.0 2 .0 79.48
40.67 5.75 2 .0 .83 78.78
34.33 6 . 0 1.5 — 78.74
42 5.71 1.5 -- 78.66
38 6 . 0 1.5 1.17 78.64
30.67 5.08 .5 — 78.61
40.33 2.83 1.0 .92 78.51
38 3.42 .5 1.17 78.50
35.33 5.25 — 1.83 78.49
29.67 5.83 2 .0 - - 78.23
36.33 5.46 1.0 2 .0 78.18
33.33 6 . 0 .5 2 .0 77.74
41 5.71 - - .92 77.60
37.67 5.5 1.5 1.92 77.54
37.67 6 . 0 1.5 - - 77.45
RANK Name WRIT
PERF
EVAL
243 Beard, K. 81 83.25
259 Sneed, L. 79 77
262 Clowney, J. 76 76.5
267 Bivens, J. 84 74.25
270 Staples, C. 77 84
273 Miller, S. 77 76.75
283 George, J. 78 78.75
285 McNeil, R. 84 71
296 Stricklen, R. 73 79
300 Hollingsworth 73 80.25
302 Oliver, N. 77 75.5
304 Jackson, A. 75 81.25
305 Hall, S. 77 75.75
315 Newkirk, G. 73 89
316 Barnes, E, 74 73
319 Harper, R. 76 72.5
321 Tomlinson, C, 77 78.25
325 Ford, W. 77 82
329 Ferguson, R. 73 86.5
333 Rice, W. 76 82
334 Neal, J. 70 87.75
335 Powell, G. 82 73.5
342 Thomas, M. 73 76.5
ORAL
BOARD SEN. COL.
MIL
SERV.
RANK ORDER SCORE
33.67 3.83 .5 1.25 77.45
40 6 . 0 — — 76.9
36.67 5.29 1.5 1.83 76.82
37.33 3.42 -- — 76.62
39.67 5.96 -- - - 76.54
29 5.54 1.5 2 .0 76.40
33 5.08 2 .0 — 76.19
32.67 4.33 - - — 76.11
40.33 6 . 0 .5 1.92 75.78
40 5.71 1.0 1.5 75.69
42 5.88 - - - - 75.65
31 6 . 0 .5 2 .0 75.63
32.33 5.04 1.0 1.67 75.58
43 6 . 0 - - — 75.4
42.33 5.88 — 2 .0 75.39
38.67 4.75 1.0 1.58 75.33
38.33 5.79 - - - - 75.24
36.67 3.75 .5 1.25 75.18
36.67 4.75 1.0 1.58 75.08
39 4.08 1.5 — 75.08
43 5.04 1.0 1.67 74.97
32.67 3.58 .5 - - 74.93
38.67 5.29 1.0 1.83 74.77
APPENDIX C
There were 49 white males passed over (Appendix A) and 51
black males promoted out of order (Appendix B) from the eligi-
* /
bility register derived from the 1974 written examination.
1. The total rank order scores of the white males listed
in Appendix A (by addition) is 4016.1. The average rank order
score for this group (divided by 49) is 82.0. The total rank
order scores of the black males listed in Appendix B (by addi
tion) is 3972.9. The average rank order score for this group
(divided by 51) is 77.9. Thus, the average rank order score
for the whites passed over is 4.1 points more than for the
blacks promoted out of order.
2. Following are the total and average scores for the
white and black groups on each component of the promotion model
(again by adding and dividing the data in Appendices A and B): *
* / The numbers do not match exactly because the white and black
women promoted are not relevant to this lawsuit (446 F. Supp. at
987); a number of women were promoted out of order as a result
of the order in Shaefer v. Tannian, sunra, p.33 n. 168.
C - 1
Blacks Promoted Whites Passed Over
Component
Total
Score
Average
(Tota 14-51)
Total
Score
Average
(Tota1449)
Written
(scale 0-100) 4051 79.43 4226 86.24
Oral Board
(scale 0-50) 1914.02 37.53 1717.68 35.05
Service Rating
(scale 0-10 0) 4086.50 80.13 4054.50 82.74
Seniority
(scale 0-6) 258.78 5.07 261.86 5.34
College
(scale 0-2) 44.04 0 .8 6 33.54 0 .6 8
Military
(scale 0-2) 41.08 0.81 22.08 0.45
uConverting the above to uniform scales of 0-100 produces the
following average scores for each group, and the following
differences in average scores:
Component
Avg. Score
Blacks
Avg. Score
Whites Difference
Written Test 79.43 86.24 -6.81
Oral Board 75.06 70.10 +4.96
Service Rating 80.13 82.74 -2.61
Seniority 84.50 89.00 -4.50
College Credit 43.00 34.00 +9.00
Military Service 40.50 22.50 +18.00
* / The relative weights used by the department were: written
test 65%, oral board 10%, service rating 15%, seniority 6%,
college credits 2% and veterans' preference 2%. Although oral
board scores are reported on a scale of 0-50 on PX 195 (and
therefore on Appendices A and B), they are' doubled (to a scale
of 0-100) in calculating rank order scores. The seniority,
college, and military credits are already adjusted in the raw
scores shown. Thus, for example, using the raw scores as shown, B.
Kowalcyk (#81 on Appendix A and the first white listed to have
points on each component) has a rank order score calculated as:
Rank order score s (.65 x Writ) + (.15 x Perf.)
+ (2 x.,10 x Oral) + Sen. + Coll. + Mil.
- (.65 x 87) + (.15 x 83) + (2 x .10 x 34.67)
+ 4.67 + .5 + 1.58
= 56.55 + 12.45 + 6.93 + 4.67 + .5 + 1.58
“ 82.68 (as reported on PX 195 and Appendix A)
C - 3
w
APPENDIX D
The calculations presented on pp. 82-83, of the Brief are
based on a simple formula. For convenience, we will use the
following notations:
Let A R
wAw
bA B
P
A p
sA s
difference in average rank order score
(between blacks and whites)
weight of written test in model
difference in average written test scores
(between blacks and whites)
weight of oral board in model
difference in average oral board scores
(between blacks and whites)
weight of service {performance) ratings
in model
difference in average service ratings
(between blacks and whites)
weight of seniority in model
difference in average seniority (between
blacks and whites)
c » weight of college credit in model
A C « difference in average college credits
(between blacks and whites)
m - weight of military credit
A m * difference in average military credits
(between blacks and whites)
It can be shown by simple algebra that the following formula
gives the difference in average rank order scores (between blacks
and whites):
A R - W • A W + b * A B + p • A P
+ s * A S + c * A C + m • A M
(The proof of this formula is presented in Appendix E).
D - 1
The average actual differences for each component between
blacks and whites for the eligibility register derived from the
1974 test are presented in Appendix C, p. C-3. Substituting
those values, the formula becomes:
=• -6.81 w + 4.96 b - 2.61 p
L________ ~ 4,50 s + 9.00 c + 18.00 m
1. The actual weights used by the department were: w = .65;
b » .10; p « .15; s ■ .06; c * * .02; and m * .02.
Substituting these figures produces:
& R - -6.81 (.65) + 4.96 (.10) - 2.61 (.15)
- 4.50 (.06) + 9.00 (.02) + 18.00 (.02)
- -4.43 + .50 - .39 - :27 + .18 ■+ .36
* -4.1 (the actual difference, see Appendix
C, p * C—1).
2. If the weights of the written test and oral boards were
reversed, with all other weight unchanged, then
& R * -6.81 (.10) + 4.96 (.65) - 2.61 (.15)
- 4.50 (.06) + 9.00 (.02) + 18.00 (.02)
- - .68 + 3.22 - .39 - .27 + .18 + .36
* + 2.4 (see p. 33 of Brief).
3. If the written test were weighted 30% and the oral
boards weighted 45%, with all other weights unchanged, then
A R - - 6.81 (.30) + 4.96 (.45) - 2.61 (.15)
- 4.50 (.06) + 9.00 (.02) + 18.00 (.02)
* - 2.04 + 2.23 - .39 - .27 + .18 + .36
* + .07 (see p. 83 of Brief)
D - 2
4. If we use the weights proposed by the City in negotia
tions (w * 32%; b * 32%; p * 20%; s * 0; c = 14%; and m
2%), then:
R - - 6.81 (.32) + 4.96 (.32) - 2.61 (.15)
- 4.50 (0) + 9.00 (.14) + 18.00 (.02)
- - 2.18 + 1.59 - .39 - 0 + 1.26 + .36
* + .64 (see p. 83 of Brief)
D - 3
y
APPENDIX E
Proof
Let Rl, R2
A R
Wl, W2
AW
Bl, B2
AB
PI, P2
A ?
of formula used in Appendix D:
=■ Average rank order scores for blacks and for whites
- Rl - R2
* Average written test scores for blacks and whites
* Wl - W2
* Average oral board scores for blacks and whites
* Bl - B2
■ Average service (performance) ratings for blacks
and for whites
- PI - P2
SI, S2
AS
Cl, C2
AC
Average seniority credits for blacks and for whites
SI - S2
Average college credits for blacks and for whites
Cl - C2
Ml, M2 - Average military credits for blacks and for whites
A M - Ml - M2
Then Rl « W - Wl + b * Bl + P ' PI
+ s • SI + c ♦ Cl + m - Ml
and R2 * w • W2 + b • B2 + P * P2
+ s • S2 + c • C2 + m • ]
where w, b, p, s, c, and m are the weights of the
components (see Appendix D).
Then A R - Rl - R2
■ w *Wl + b • Bl + P PI
+ s • SI + c • Cl + m • Ml
- w . W2 - b • B2 - p • P2
- s • S2 - c • C2 mm m • M2
* w (Wl - W2) + b (Bl -
+ p (pi - P2) + s (SI -
+ c (Cl - C2) + m (Mlr
i A * = w •A w + b •AB + p • AP! + s • As + c • A C + m-AM
B2)
S2)
- M2)
DOJ-1978-07
E - 1