Amicus Curiae Brief — Watson v. Fort Worth Bank & Trust

~

ae \

A

No. 86-6139

In the Supreme Court of the United States

OCTOBER TERM, 1987

CLARA WATSON. PETITIONER

}

FORT WORTH BANK AND TRUS]

ON WRIT OF CERTIORARI TO THi

UNITED STATES COURT OF APPEAI

FOR THE Fifk ij tH CIRCUTI

BRIEF FOR THE UNITED STATES

AS AMICUS CURIAE SUPPORTING RESPONDELN!

ROMOER CLEC

Depu

OULSTION PRESENTED

ullineg

TABLE OF CONTENTS

Page

Interest of the United Siates |

Statement ....... |

Summary of argument 4

Argument ........ 6

The decisions resulting from a subjective promotion proc-

ess may be scrutinized effectively under the disparate

treatment analysis and should not also be subjected to

challenge under the disparate impact analysis 6

A. Disparate impact analysis has been applied only to

selection devices that do not assess each person’s

suitability for employment in an individualized way 6

B. Disparate treatment analysis provides ample protec

tion against possible discriminatory animus or

cultural bias arising from the use of subjective selec

tion devices. 12

C. Extending disparate impact analysis to the decisions

resulting from subjective selection processes would

upset Congress’s delicate balance of competing con

cerns in Title VII 17

D. The Uniform Guidelines on Employee Selection Pro

cedures do not provide that subjective selection proc

esses must be validated 26

Conclusion 28

TABLE OF AUTHORITIES

{ asses

4euilera v. Cook Cty. Police & Corrections Merit Bd

760 F.2d 844 (7th Cir.), cert. denied, 474 U.S. 907

(1985) 18

i/bemarle Paper Co. v. Moody, 422 U.S. 405 (1975) 8, 10

18, 19, 25

irlington Heights v. Metropolitan Housing Dev. Corp.,

429 U.S. 252 (1977) 14, 15

A4tonio v. Wards Cove Packing Co., 810 F.2d 1477 (9th

Cir. 1987) 14

Bauer v. Bailar, 647 &.2d 1037 (10th Cir. 1981) 14

Bazemore v. Friday, No. 85-93 (July 1, 1986) 9 12. 17

(111)

IV

Cases-Continued Page

Connecticut v. Teal, 457 U.S. 440 (1982) 8, 1

Davis v. City of Dallas, 777 F.2d 205 (Sth Cir. 1985), cert

denied, No. 85-1533 (May 19, 1986) 18

Dothard v. Rawlinson, 433 U.S. 321 (1977) 8, 10

Douglas v. Hampton, §12 F.2d 976(D.C. Cir. 1975) 23

Firefighters v. Cleveland, No. 84-1999 (July 2, 1986) 24

Furnco Constr. Corp. v. Waters, 438 U.S. 567 (1978) 5,

10, 11, 15

General Telephone Co. v. EEOC, 446 U.S. 318 (1980) 24

Goodman v. Lukens Steel Co., No. 85-1626 (June 19,

1987) 16

Gottlieb v. Tulane Univ. of Louisiana, 809 F.2d 278 (Sth

Cir. 1987) 25

Griffin v. Carlin, 755 F.2d 1516 (ith Cir. 1985) 15

Griges v. Duke Power Co., 401 U.S. 424 (1971) 4, 6,

7, 8, 10, 15, 17, 18, 24

Guardians Ass'n v. Civil Serv. Comm'n, 630 F.2d 79

(2d Cir. 1980), cert. denied, 452 U.S. 940 (1981) 21, 23

Harrison vy. Lewis, $59 F. Supp. 943 (D.D.C. 1983) 19

Hazelwood Schv i Dist. v. United States, 433 U.S. 299

(1977) 9,12, 14

Johnson v. Transp. Agency, No. 85-1129 (Mar. 25,

1987) 24, 25

Los Angeles Dep’t of Water & Power v. Manhart, 435

U.S. 702 (1978) y

\fcDonnell Douglas Corp. v. Green, 411 U.S. 792 (1973) > we

10, 11, 12, 21

Nanty v. Barrows Co., 660 F.2d 1327 (9th Cir. 1981) 13

Nashville Gas Co. v. Satty, 434 U.S. 136 (1977) 12

New York City Transit Auth. v. Beazer, 440 U.S. 568

(1979) 8, 18

Parson v. Kaiser Aluminum & Chem. Corp., 575 ¥.2d

1374 (Sth Cir. 1978), cert. denied, 441 U.S. 968 (1979) 13

Personnel Adm'r v. Feeney, 442 U.S. 256 (1979) 14-15

Rogers v. Int'l Paper Co., 510 F.2d 1340 (8th Cir.)

vacated, 423 U.S. 809 (1975) 25

Rowe v. General Motors Corp., 457 F.2d 348 (Sth Cu

1972) 13-14

Rovaly Viissouri Hiw) a Transp Communion 655 | 2d 159

(Sth Cir. 1981) 13

y

Cases —- Continued: Page

Sheet Metal Workers v. EEOC, No. 84-1656 (July 2,

1986) 24, 25

Smith v. Olin Chemical Corp., 555 F.2d 1283 (Sth Cir.

1977) 18

Steelworkers v. Weber, 443 U.S. 193 (1979) 24, 25

Stewart v. General Motors Corp., 542 F.2d 445 (7th Cir.

1976), cert. denied, 443 U.S. 919 (1977) 14

Teamsters v. United States, 431 U.S. 324 (1977) 6,9

11, 12, 14, 15, 24

Texas Dep't of Community Affairs v. Burdine, 450 U.S.

248 (1981) 13

United States v. South Carolina, 445 F. Supp. 1094

(D.S.C. 1977), aff'd mem. sub nom. National Educa

tion Ass'n v. South Carolina, 434 U.S. 1026 (1978) 18

U/.S. Postal Serv. Bd. of Govs. v. Aikens, 460 U.S. 711

(1983) 13, 17

Wade v. Mississippi Coop. Ext. Serv., 528 F.2d 508 (Sth

Cir. 1976) 13

Washington v. Davis, 426 U.S. 229 (1976) 18, 19, 22

Yuhas v. Libbey-Owens-Ford Co., 562 F.2d 496 (7th Cir

1977), cert. denied, 435 U.S. 934(1978) 18

Zahorik v. Cornell, 729 F.2d 85 (2d ¢ 1984) 20

Statutes, regulations and rule

Civil Rights Act of 1964, Tit. VII, 42 U.S.C. 2000e ef

seq 2,6

§ 703(aXil), 42 U.S.C. 2000e-2(a)(1) 12, 23

§ 703(aX2), 42 U.S.C. 2000e-2(a)(2) 7, 11, 23

§ 703(h), 42 U.S.C. 2000e-2(h) 26

§ 703(j), 42 U.S.C. 2000e-2()) 25

§ 706, 42 U.S.C. 2000e-5

Equal Employment Opportunity Act of 1972, Pub. L. No

92-261, 86 Stat. 103 23.24

29 C.F.R

Pt. 1607 1 5. %6

Section 1607.3A %

Section 1607.6B 27

Section 1607.16Q 26

Fed. R. Cw. P. 2a) ;

Vi

Miscellaneous Page

APA, Principles for the Validation and Use of Personne!

Selection Procedures (3d ed. 1987) 22

Bartholet, Application of Title V11 to Jobs in High Places,

95 Harv. L. Rev. 945 (1982) 21, 23, 27

Blumrosen, The Legacy of Griggs: Social Progress and

Subjective Judements, 63 Chi.-Kent Il Rev |

(1987) 19, 21, 23

Booth & Mackay, Lega/ Constraints on Employment Test

ing and Evolving Trends in the Law, 29 Emory L.J. 121

(1980) lv

W. Byham & M. Spitzer, The Law and Personnel Testing

(1971)

W. Cascio, Applied Psychology in Personnel Management

(2d ed. 1982) 21

Comment, Courts, Psychologists, and the EEOC’s Uni

form Guidelines: An Analysis of Recent Trends Affect

ing Testing as a Means of Employee Selection, 36

Emory L.J. 203 (1987) 19

100 Cong. Rec. (1964)

p. 6562 24

p. 8442 24

p. 8500 24

p. 9601 24

p. 13078 24

p. 14314 24

p. 15893 24

117 Cong. Rec. 31960 (1972) 24

118 Cong. Rec. 1815-1816 (1972) 24

Cooper & Sobol, Seniority and Jestinge Under Fai

Employment Laws: A General Approach to Objective

Criteria of Hirine and Promotion, 82 Harv. L. Rev

1598 (1969)

Denis, Subjective Decision Makine: Does ut Have a Place

mt tite Emplovyine nt Process’, 11 Empl Rel L.J. 270

(1985) 13, 15

EEOC, Guidelines on Employvinen ir ne Procedures

(Auge. 24, 1966) Z

38 Fed. Reg. 12333 (1970) 26

Vil

Miscellaneous — Continued: Page

43 Fed. Reg. 38291 (1978) 27

44 Fed. Reg. 12002 (1979) 27

Furnish, A Path Through the Maze: Disparate Impact

and Disparate Treatment Under Title VII of the Civil

Rights Act of 1964 After Beazer and Burdine, 23 B.

L. Rev. 419 (1982) 1]

E. Ghiselli, Theory of Psychological Measurement (1964) 20

Gwartney, Asher, Haworth & Haworth, Séfatistics, the

Law and Title Vil: An Economist’s View, 54 Notre

Dame Law. 633 (1979) 19, 21, 23

Haney, Employment Tests and Employment Discrimina

tion: A Dissenting Psychological Opinion, 5 Empl

Rel. L.J. 1 (1982) 16

H.R. Rep. 92-238, 92d Cong., Ist Sess. (1972) 24

Hunt & Pazuniak, Special Problems in Litigatine Upper

Level Employment Discrimination Cases, 4 Del. J

Corp. L. 114 (1978) 21, 23

3 A. Larson & L. Larson, Employment Discrimination

(1986) 10, 11, 1S, 21, 22, 27

Lerner, Employment Discrimination: Adverse Impact,

Validity, and Equality, 1979 Sup. Ct. Rev. 17 19, 23, 25

Lerner, Washington v. Davis: Quantity, Quelity and

Equality in Employment Testing, 1976 Sup. Ct. Rev

263 21, 23

Maliz, Jitle Wil and Upper Level Employment—A

Response to Professor Bartholet, 77 Nw. U. L. Rev. 776

(1983) 23, 24

M. Miner & J. Miner, Emplovee Selection Within the

Law (1979) 21, 23

Note, Developments in the Law—Employment Dis

crimination and Title Vil of the Civil Riehts Act of

1964, 84 Harv. L. Rev. 1109 (1971)

Note, Evaluation of Subjective Selection Systems in Tith

Vill Employment Discrimination Cases: A Misuse oj

Disparate Impact Analysis, 7 Cardozo L. Rev. $49

(1986) Z

Vill

Miscellaneous — C ontinued

Note. lL eval Implications of the Use of Standardized

Ability Tests in Employment and Education, 68 Colun

L. Rev. 691 (1968)

4. Outerbridge, Personnel Res. and Dev. Ctr., OPM

TM-79-18. A Survey of Test Validation Study Cosi!

(Aug. 1979)

Rachlin, Title Vil: Limitations and Qualifications, 7 B.«

indus & { om | Rev 473 (1966)

B. Schiei & P. Grossman, Employment Discriminatior

Law (1983) ;

“~*

S. Rep. 92-415, 92d Cong., Ist Sess. (1972)

Hn the Supreme Court of the United States

OcTOBER TERM, 1987

No 86-6139

CLARA WATSON. PETITIONER

’

FORT WORTH BANK AND TRUS!

ON URIT OF CERTIORARI TO THi

UNITED STATES COURT OF APPEAI!l

POR THE FiF lH CIR i]

Battt FOR THE UNITED STATES

AS AMICUS CURIAE SUPPORTING RESPONDENI

INTEREST OF THE UNITED STATES

Ihe Attorney General and the tr qual Employment Oj

imty Commussion (EEOC) have significant Title VII entorce

zz. ’ , , . ’ tf saec ; ’

nent responsibilities. See 42 U.S.C. 2000e Ihe United State

as the nation’s largest employer, tS also subDject to litle Vil

bligations. Finally, the proper interpretation of the Unit

suidelines on Employee Selection Procedures, 29 C.F.R. P

1607, promulgated | riOUS agencte t the Unned S

d78 i i < |

STATEMENT

i Pet (lara Watson s+ a2 DOCK ‘ na SI

/ ; (i \ ,. _

(jn () ' I51. DE eT | } j Vila ;

‘4 Lae @d.} i ving ina . :

it ' ; j ne [ 1. - : _ | '

| A *h JH) ©)

| i De :

| ‘ : na ai) j } Te

| ‘5 ’ P ,79y

; 1A yz

" , ,

: i Dia

; 4 4 . ~

. » “i Tf

go ti 7. jx 3 : :

>

;™ s™* \ | ;

.

: j _

\ ‘ ; >40 Ul

>

| .

~ : ’

lL.

. 7 4~ As ;

general review (and rare reversal) by respondent’s highest of-

ficials (J.A. 194-195; Il Tr. 80-92). Respondent also pointed out

that petitioner’s sample size was in many respects too small for

drawing legitimate statistical inferences and that her statistician

had not considered the relative qualifications of the persons in

the applicant pool, the distribution of applicants for specific

openings, or the influence of market factors (11 Tr. 181-196).

Finally, respondent demonstrated that blacks comprised 11.8%,

192%, and 22.8% of the population of the county,

metropolitan area, and city, respectively (J.A. 202).

Based on this evidence, the district court found (J.A.

199-201) that existing black employees (such as petitioner) did

not share a common question of law or fact with nonemployee

applicants and that existing employees formed too small a group

to satisfy the “numerosity” requirement of Fed. R. Civ. P

2¥a); accordingly, ut subdivided the class into applicants and

employees and held that petitioner could not advance claims on

behalf of exher class (J.A. 207). On the merits, the court held

that respondent's articulated reasons for failing to promote pet!

lioner were not pretextual and that, although petitioner “made

allegations of discrimination by defendant in hiring, compensa-

tion, initial placement, promotions, * * * and ot ier terms and

conditions of employment, there [was] little or no evidence of

discrimination in these categories” (J.A. 197, 201). Finally, after

finding that petitioner had “proceeded zealously on behalf of

the group of blacks who applied for employment [with] the

[bjank,” u ruled that “no discrimination against blacks as a class

hald] been proved on the basis of defendant's hiring practices”

(J.A. 202)

b The court of appeals affirmed in part and reversed in part

(J.A. 204-252). lt reyected petitioner’s argument that the district

court had abused us discretion in decertifying the class and m

holding that petitioner was not a proper class representative

(J.A. 208-212). On the merits, u held that “a Title Vil challenge

io an allegediy discretionary promotion system ts propertly

analyzed under the disparate treatment model rather than the

disparate impact model” (J.A. 213), that there was “no indica

ion m the record, nor [was] there any suggestion * * * Dy [peti

* * * that the district court neglected to consider the

poner}

uustical evidence presented Dy [petitioner]

redibility of the

pianations tor ws

in terms |

jluatineg the

lrespondent's} prottered

decisions” (J.A. 214 n.13), and

‘the district court's finding that the [Dlank’s explanation

is action were not pretexts for discrimination ts not clear!

rroneous (J \ 215)

promotvuon

hos

inal

4)

Finally, the court held that the hiring

rimination claims of the applicant class should have been

usmissed without prejudice (since there was no proper cia

presentative); wt remanded the case tor this purpose (J.A

1S)

SUMMARY OF ARGUMENT

\ Uisparate treatment was the principal evil ¢ T had

md when wt enacted T ithe Vil ihe title Vil ni eri

mMmMmunity soon observed however (nat en pioyel A

mn2g tO U landardized tests and

other obvpective credential

‘ a1

juirements which often had an adverse exclusionary ett

nbers of groups protected by Title Vil and which bor:

anil clationship to the employment in question. It wa

ponse to tn oncern that the Court in Griges v. Duke Power

401 U.S. 424 (1971)

rirst approved use Of dispar ;

mpa

iS a means of establishing unlawful employment discrin

ihe ( it held that a Title Vil violation may b :

; ul pt ; in : \T in opr i¢ i ] !

i Kcludes | 1 mdividuals tron

pportu ind i “manit lated

ft Db. The Court ha

’ : » 7 ’ ’ ' | ’ "

4 : ‘ na i] | Nill ) :

uf : nai a si vying (hal a pi

Md 1] nembet ia : is :

wmnsn at mata Vil ia

( : : ‘

\ , 4

\ if ;

jective selection processes, only disparate treatment, and not

disparate impact, theory applies. See AfcDonnell Douglas Corp

v. Green, 411 U.S. 792 (1973); Furnco Constr. Corp. v. Waters,

438 U.S. 567 (1978).

B. Disparate treatment theory provides the protection that

Congress intended against any discriminatory animus of

cultural bias that may arise from the use of subjective selection

processes. Selection processes are subjective when they are

discretionary and nonmechanical and depend upon the mental

operations of the decisionmaker. This Court has routinely re

jected arguments that disparate treatment analysis ts unsuited to

the task of examining the legality of subjective selection devices

Courts have often found disparate treatment to exist where a

subjective selection process with an aaverse impact has inex

plicably been used. That petitioner iailed to prevail on het

disparate treatment claim here does not show that disparate

treatment analysis ts an inadequate response to discriminatory

subjective decisionmaking; i shows rather that petitioner tailed

to convince the courts below that she was a victim of such

discriminatory decisionmaking

( Extending disparate impact theory to employee selec

tions resulting from subjective processes would undermine other

policies with which Congress was concerned in adopting Title

Vil. Disparate impact analysis is simply not suited to in

vestigating whether the decisionmaking processes that constitute

subjective selection procedures are discrimimatory or not. A

detendant could rarely, if ever, validate or otherwise formally

substantiate its subjective judgments. Subjective judgments

typically assess intangible qualities, which cannot be measured

in the manner that ts necessary for validation or other tormal

supstantiation. Accordingly, applying disparate impact theory

to the decisions resulting from subjective selection processes

woyld inevitably lead employers either to abaridon subjective

processes or, where such replacement is too burdensome,

restrictive, OF expensive, to superimpose numerical quotas on

employment systems using subjective devices. Neuher resull

would be consistent with the ment of the 1964 C onegress

) Ihe Uniform Guidelines on Employee Selection Pro

cedures, 29C FR. Pt. 1607 (1978), do not provid that subspes

h

ive selection processes must be validated or otherwise formally

upDstantiated. (in the

contrary, the Ceudelmes provide

ina

,here a subsective selection DrOCcess t5 ih ssuc

an employer may

justify continued use of the procedure in accord with Federal

law.” The United States and the EEOC beleve that an employ

an justify “contmmued use of the procedure m accord wit!

tederal law” by showing that its recourse to and application

part ular subpective

ne

ia

selection process is reasonably related

requirements Of the joD or other |

usiness need asm Al

:

isparate treatment case

ARGUMENT

THE DECTSIONS RESULTING FROM ASUBIECTIVE PRO

MOTION PROCESS MAY BE SCRUTINIZED EFFECTIVE

LY UNDER THE DISPARATE TREATMENT

AND SHOULD NOT ALSO &I

CHALLENGE UNDER THE

ANALYSIS

SUBJECTED T0

DISPARATE IMPACT

ANALYSIS

Petitioner contends that the outcon | subrecti /

on processes like that employed Dy pondent mu pa

ru under the disparate impact analysis of Grigg Duk

Power ( 401 U.S. 42411971). Based on our understanding ot

’ ‘ our ASE finciuding Crriges usell). our CA pPEe»rien MJ

rine tittle Vil of the Civel Riehi Ac he bala ‘

‘ ere 1964, and the Umtorm Crndel n Emmy

? ‘ ibe that ti ( }

’ ; it 74

4 Disparate Impact Analysis has Been Applied Only to Setex

tien Devices [hat Do Not Assess Each Person's Suitability

for Employment in an Individualized Way

VA Hewin f pia y ti ( ul ispara Da

i | / if ‘ na Hd ma tt

' 1Y64 : i* ;

: : ‘a | ‘

' Vil

, 1) ‘4 > ,iiy } : ‘

9

Contrary to petitioner’s argument (Br. 15-20), however, the

Court has not treated disparate impact analysis as the rule and

disparate treatment analysis as the exception. In fact, the Court

since Griggs has refused to accept the proposition that a prima

facie Title VII violation may be established without proof ol

discriminatory motive in every circumstance in which an

employment practice or combination of employment practices

has an adverse effect on a protected group. Thus, in Los

Angeles Dep’t of Water & Power v. Manhart, 435 U.S. 702, 711

n.20 (1978), the Court rejected the contention that the adoption

of a gender neutral pension plan “would itself violate Title VII

because of its disproportionately heavy impact on male

employees. Cf. Griggs v. Duke Power Co., 401 U.S. 424.” The

Court said (435 U.S. at 711 n.20):

* * *(Ujnder Title VII * * * the male employees would not

prevail. Even a completely neutral practice will inevitably

have some disproportionate impact On one group Of

another. Griggs does not imply, and this Court has never

held, that discrimination must always be inferred trom

such consequences.

Moreover, where non-mechanical selection processes have been

in issue, the Court has consistently inquired only whether Title

VII plaintiffs have successfully proven — by statistics or other

evidence —the presence of discriminatory motivation. See, é.g.,

Bazemore v. Friday, No 85-93 (July 1, 1986); Hazeiwood

School Dist. v. United States, 433 U.S. 299 (1977); Teamsters v.

United States, supra. Indeed, the Court has at least twice

strongly implied that disparate impact theory ts inapplicable in

challenges to subjective selection processes.

McDonnell Douglas Corp. v. Green, 441 U.S. 792, 805-807

(1973), concerned a discretionary decision not to rehire an in-

dividual who had participated in unlawful conduct directed

against an employer. Contrasting such a decision with the type

of mechanical decisionmaking processes that were in issue in

Grieves, the Court noted (411 U.S. at 806 (citation omitted)) that

Criees

dealt with standardized testing devices which, however

neutral on their face, operated to exclude many blacks who

were capable of performing effectively in the desired pos!

10

tions. Griggs was rightly concerned that childhood defi

ciencies in * * * education and background of minority

citizens, resulting from forces beyond their control, not be

allowed to work a cumulative and invidious burden on such

citizens for the remainder of their lives. Respondent,

however, appears in different clothing. *** [His

employer] does not seek his exclusion on the basis of a

testing device which overstates what is necessary for com-

petent performance, or through some sweeping disquali

fication * * *. [His employer] assertedly rejected [him] for

unlawful conduct * * * and * * * this cannot be thought

the kind of “artificial, arbiirary, and unnecessary barriers

to employment” which the Court found to be the intention

of Congress to remove.

Accordingly, the Court did not require the employer to

demonstrate that its subjective refusal to rehire had a manifest

relationship to the requirements of the job (id. at 805-806)

In Furnco Constr. Corp. v. Waters, 438 U.S. 567, 572, 575

(1978), the Court expressly refused to apply the disparate im

pact analysis to an employer’s practice of hiring only those per

sons whom his superintendent knew were experienced or had

been recommended. The Court contrasted the selection device

in issue there with “employment tests * * * or particularized re

quirements such as * * * height and weight specifications” (id

at 575 n.7), as were involved in Griggs, Albemarle Paper Co.,

and Dothard v. Rawlinson.*

As petitioner notes (Br. 22), there was no evidence of group-wide

disparate impact in McDonnell Douglas. But that is not the reason the Court

gave for rejecting application of the Griggs “manifest relationship” require

mem; rather, the Court in McDonnell Douglas said that the employer was not

subject to the more stringent Griggs-detense because the challenged practice

was a “predictive evaluation resistant to empirical proof” (411 U.S. at 806

n.21)

* Petitioner seeks to distinguish Furnco (Br. 23) on the ground that there

was no showing that the employer's hiring practices had a disparate impact

Bul this distinction does not explain why the Court “foreclose[d] on remand

further litigation on the Griggs question of whether [the employer's} hiring

practices had a disparate impact” (fFurnco Constr. Corp. v. Waters, 438 U.S

at 583 (Marshall, J., dissenting)). That is why various commentators have con

cluded that the Court in Furnco held disparate iapact theory inapplicable u

chalienges to subjective selection processes. See, ¢.2 1 A. Larson &

| Larson, Employment Discrimination § 76.42 a «615-83 to 15-85 (1986)

McDonnell Douglas and Furnco illustrate that the Court has

heretofore declined to extend the disparate impact analysis

beyond the particular form of employment discrimination

associated with the use of objective, mechanical selection

devices. It has said that an employer “limit[s], segregate[s], or

classif[ies] his employees or applicants” within the meaning of

Section 703(aX(2) when, by using such a device, the employer

separates individuals into uniformly defined groups from which

ultimate employment decisions will be made; and where the de-

vice is not job-related and has a statistically sign'ficant adverse

effect on members of a protected group, the Court has said that

the “limit{ation], segregat{ion], or classiffication]” “deprive|s]

or tend[s}] to deprive any individual of employment oppor-

tunities or otherwise adversely affect{[s] his status as an

employee, because of such individual’s race, color, religion, sex,

or national origin” (42 U.S.C. 2000e-2(a)(2)). But the Court has

also strongly implied that an employer does not “limit, segre-

gate, or classify his employees or applicants” within the meaning

of Section 703{a)(2) when it makes subjective decisions about

particular employees or applicants, even though a series of those

decisions may cumulatively result in a numerical or propor-

tional racial disparity in the employer’s workforce.’ Rather, the

Court has indicated that such cases are properly treated as

{hereinafter Larson}; Furnish, A Path Through the Maze: Disparate Impact

and Disparate Treatment Under Title VII of the Civil Rights Act of 1964 After

Beazer and Burdine, 23 B.C. L. Rev. 419, 438-439 (1982); AFL-CIO Br. at

10-11, in Connecticut v. Teal, supra

* See AfcDonnell Douglas Corp., 411 U.S. at 805-806; Furnco Constr

Corp. v. Waters, 438 U.S. at $75 n.7, 578; Teamsters, 431 U.S. at 334-336,

339-340 n.20

Where mixed-component or multi-step selection processes are involved, of

course, the Court has said that a plainuff may challenge any objective compo

nemt of that process for its own disparate impact. See Connecticut v. Teal,

supra. Bul petitioner has not attempted to do so in this case. Rather, she

challenges the selection process as a whole, which she characterizes as subjec

tive and discretionary. See Pet. Br. 4-6, 40, 47-48. Because the selection proc

ess includes both objective factors and subjective discretion, we agree that i

should be treated as subjective and turther submu that u should be

challengeable only under the disparate treatment theory. See Larson § 76.34

at 15-85 to 15-86, 15-91 to 15-98

| ,

arising under Section 7OMavl) of the statute and aS involving

only a possible “pattern or practice” of intentional discrimina

von

B. Disparate Treatment Analysis Provides Ample Protection

Against Possible Discriminatory Animus or Cultural Bias

Arising from the Use of Subjective Selection Devices

Petitioner’s principal reason for suggesting that disparate im

pact analysis should be applied to subjective selection devices 15S

(Br. 34-44) that, in her view, disparate treatment analysis Goes

not provide sufficient protection against the discrimimatory

ynimus or cultural bias that may arise from the application ol

[Nose devices We adisaegree

Subjective selection devices mvolve, in this Cour

vords, “predictive evaluation|s| that are “resistant (oO empiri 1

proof” (McDonnell Dougias Cor} 441 U.S. at 806 0.21). Such

include decisional components of unstructured perso

~ * F%

1€.4 c il ervicews | purl

an | ‘an ,}

decisions, OF supervisor ratings based on personal experience

with candidates for purposes of promotion decisions) and

assessments based on unstructured weighing of specified objec-

live considerations (such as grades, prior experience, and

references). The common attribute of subjective selection proc

esses 1s their reliance on judgment, intuition, and discretion

Mental operations such as these are, of course, the focus of

disparate treatment analysis and this Court has repeatedly re

jected arguments that this analysis is inadequate to its task. See,

e.g., U.S. Postal Serv. Bd. of Govs. v. Aikens, 460 U.S. 711.

716-717 (1983); Texas Dep't of Community Affairs v. Burdine,

450 U.S. 248, 256-260 (1981).

Ihe Court’s confidence in disparate treatment analysis is

equally appropriate here. Courts applying disparate treatment

analysis have consistently found decisions resulting from inex

plicable and unwarranted recourse to subjective selection prox

esses to be motivated by discriminatory intent. See generally B

Schler & P. Grossman, Employment Discrimination Law

191-205 (1983); Denis, Subjective Decision Makine: Does it

Have a Place in the Employment Process?, 11 Empl. Rel. L.J

270-276 (1985). Courts are well aware that subjective selectior

processes are “vulnerable to either conscious or unconscious dis

crimination by the evaluating supervisors” (Wade v. Mississippi

Coop. Ext. Serv., 528 F.2d 508, 518 (Sth Cir. 1976)), especially

where the evaluating supervisor is of a race or sex different from

that of th: applicant (see, e.g., Roval v. Mlissouri Hwy. &

Transp. Comm'n, 655 F.2d 159, 164 (8th Cir. 1981); Nanry »

Barrows Co., 660 F.2d 1327, 1334 (9th Cir. 1981)). Likewise.

courts have tound discriminatory treatment where the employe:

has no substantial job or business-related reason for employine

an entirely subjective selection process — as is often true when

ihe employer ts tilling jobs that require minimal or easily quan

fred skills. See. e.¢ Parson \. Aatser Aluwmnum & Chen

Corp., 575 F.2d 1374, 1385-1386 (Sth Cir. 1978). cert. denied

441 U.S. 968 (1979); Rowe v. General AMlotors Corp., 487 F.2d

348, 359 (Sth Cir. 1972). Moreover, this Cou Je.

pattern and practice disparate treatment cases make clear tha

statistically significant disparities Ma pe used | IpproT ii¢

rcumstances to miler discriminatory intent And. tollowine

hese “pattern and practice’ decisions, the courts of appea

ave consistently held that unwarranted recourse ID pec

ection processes can provide the missing tink f

atistical disparities and iiegal discrin

Bailar, 647 &.2d 1037, 1045 (10th Cir. 1981); S

General Motors Cor $42 F.2d 445. 450-451 (7th Cu 976)

" deemed. 433 U.S. 919 (1977). Thus. both theory and pra

contirm that disparate treatment theory an erie

rerre ‘ mul GISc ir y sSUDIEC . 1¢ Sid alk

~ 2. Pe er, however, suggests (Br. 24-25, 35, 41-44)

. 2

S10 F.2d 14 (1987) (en Danc), tha iD }< e sek

. i) ud i , Hit [

’ ’ ’ '? -

aca DA ~ | . . . ; . : |

vt t if | C i t .

“ea ‘a : . \ | : : :

. : .

. 3 . :

na , i ' : ih

.

' _ _

‘ ‘

a

15

Adm'r v. Feeney, 442 U.S. 256, 279 0.25 (1979): Arlington

Heights v. Metropolitan Housing Dev. Corp., 429 U.S. 252, 266

(1977); Teamsters, 431 U.S. at 336-339. Importantly, as noted

above, the fact that an employer cannot provide a job or

business-related reason for using a particular subjective selec-

tion process may also constitute evidence of discriminatory in-

tent. Thus, an employer plainly cannot expect to rebut a prima

facie statistical showing of discriminatory treatment with an

apology that such consequences are merely the result of a

longstanding practice with an unknowable purpose. Rather, as a

practical matter, the employer must show that its recourse to,

and its particular applications of, the selection process are

reasonably related to the requirements of the job or other

business need. Accord, B. Schiei & P. Grossman, supra, ai

1287-1288; Denis, supra, 11 Empl. Rel. L.J. at 270-276.

Nor is there merit to the concern voiced by petitioner (Br

36-41) and the Eleventh Circuit (see Griffin v. Carlin, 755 F.2d

1516, 1525 (1985)) that the disparate impact theory must be ex-

tended to subjective selection processes m order to prevent

employers from abandoning unvalidated objective selection

practices and substituting subjective selection processes in their

place. This extension simply turns Griggs on its head. The very

purpose of the Griggs decision was to require employers to

abandon unvalidated objective selection practices (where those

practices had a disparate impact on a protected group). More-

over, Griegs and its progeny in this Court make clear that Title

Vil leaves employers with the choice of adopting validated ob-

jective selection devices or considering each employee or appli-

cam individually.'* And, in any event, there are a host of

pressures that keep empivyers from unwarranted and pretextual

recourse to subjective selection processes: objective selection

devices In many circumstances are the only efficient means of

selection and the pressures of competitive markets will compel

. Sec Csriees, +4) 4 Ss al 43) 433. Furnco « overs { orp ‘ “ wieTs 43s

U.S. at 578. Accord, Larson § 76.33, at 15-86 to 15-87 (“The duty mmposed by

Tulc Vile a duty to retram trom racial discrimunation m employment. If an

employer can manage to comply with the duty wahout the and of formally

devised and mmplemented objective criteria, he Canmat De held have \nnated

AC)

t xtending Disparate Impact Analysis to the Decisions Result

ing from Su ective Selection Processes Would Upset Con

gress's Delicate Balance of Competing Concerns in Tithe Vil

ita mpa ' | we rm : led it

rve :

: | j ‘ ie _ : ' : _

nai * ik ’ ; ; | vil

; ¥ j i} Gispa :

"

sparta i ; ifa

' ' na 1 OeT la i*

i ita i i : ifa

" ;

; ; ; : :

1 i . : i :

" ;

: . _ ; :

: | : ; _

/ . Bb ; Da

. : 1. 7

19

business operation. See Washington v. Davis, 426 U.S. at

246-252; Albemarle Paper Co., 422 U.S. at 425-436. Though

this process of job definition, data collection, and formal

shstantiation can be quite costly and time-consuming, ?® certain

objective selection devices — such as reading ability tests and ver-

bal skill requirements — have been shown to have the necessary

“manifest relationship” to performance of particular jobs

(especially those involving a small number of simple and

repetitive tasks).?'

By stark contrast, we have been unable to find a single

reported case in which a subjective selection process was suc-

cessfully substantiated by formal means.?? The problem is that

subjective selection processes are typically used to assess

has a “manifest relationsmp” to the job by any means other than reasonable

inference. Therefore, inferential reasonableness ts all that can be required of

selection practices that rely on the employer's subjective judgment

** For discussions of the costs and complexities attendant to such efforts.

see A. Outerbridge, Personnel Res. and Dev. Cir., OPM, TM-7*./8, A Survey

of Test Validation Study Costs a | (Aug. 1979) (costs of stedies surveyed

ranged from $24,000 for measurement of a single characteristic to $1,245.00

for measurement of multiple characteristics); Lerner, Luplovment

Discrimination. Adverse Impact, Validity, and Equality, 1979 Sup. Ci. Rev

17, 18 n.6 (adequate validation study costs between $100,000 and $400,000).

(wartney, Asher, Haworth & Haworth, Si wisrics, the Law and Title WI, An

Economist’ View, $4 Nowre Dame Law. 633, 643 (1979) (cost of measuring

one sample characteristic ranges from $20,000 to $100,000)

See Comment, Cowrts, Pivchologists, and the EPOCS Uniform

Cuidetines’ An Analysts of Recent Trends Affecting Testing as a Means of

Empltovee Selection, % Emory LJ. 203 (1987) (summarizing developments),

Booth & Mackay, Lege! Constraints on Employment Testing and Evolving

Trends in the Law, 29 Emory LJ. 121 (1980) (same

Professor Blumroven surveys the cases and reache, the same Conclusion

See Blumrosen, The Legacy of Griggs: Social Proeres ond Subjective

Judgements, 63 Chi. Kem LL. Rev. 1, 22, 33-397 (1987). The NAACP Legal

Detense and Education Fund, et al, errs in suggesting (Amicus Br. 29-10) that

a subjective selection process was waccessfully validated in Merrion. 1 ewr

SF. Supp. 43 (D D.C. 1983) The employer in Muerrison did produce a

repixation study” of it) selection procedures, but the court held that the

vudy did “not m tact demonvurate the vahdity of y.ve} selection procedures

** *" td at O48)

(hvavonally, courts have upheld sulyectine decrmonmaking processes

agai divparate impact attacks, Dut only by pudeging the facil reasonablenes

SSSa_a___a_E

6

: s* ~

7

a. ; :

om : od ™ " 4 " .

. “ ‘ . _— -

~ ' . : Ae ,

« * ~ 7 | ~ ju & an .s "

: ' . : >

» ri sn , ,

n * av /_ “a ‘ -

s “ sme

:

\

>

such judgment therefore be rigidly eschewed. “In the aggregate,

heightened probabilities for excellence are not a function of the

validity of any particular selection device,” but “are the result of

a total situation in which vast numbers of employers are tree to

experiment with an unlimited number of possible qualitative

requirements, discarding those that seem unsatisfactory and

replacing them with others that seem more promising, and so on

and on, in a continuing, open-ended process” (Lerner, supra,

1976 Sup. Ct. Rev. at 304-305).?*

2. Ruling that the disparate impact theory ts applicable to

decisions resulting from subjective selection processes would,

therefore, create an irresistible incentive for employers to aban-

don subjective selection processes in favor of objective ones or,

where such replacement is too difficult or expensive, to

eliminate the statistical disparity by superimposing quotas upon

them.?* Neither result would be consistent with the intent of the

1964 Congress. *°

necessary to validate these objectified decision processes. But, as several

courts have noted, “construct validation” builds on content and criterion

validation and, accordingly, cannot piacticably be conducted where content

and criterion validation are infeasible. See, e.g., Guardians Ass'n v. Civil Sers

ice Comm'n, 690 F.2d at 92; Douglas v. Hampton, 512 F.2d 976, 985 (D.¢

Cir. 1975)

** Accord, M. Miner & J. Miner, supra, at 143; Blumrosen, supra, 63

Chi.-Kent L. Rev. at 18; Maltz, Title Vil and Upper Level Employment —A

Response to Professor Bartholet, 77 Nw. U. L. Rev. 776, 789 (1983), Hunt &

Pazuniak, supra, 4 Del. J. Corp. L. at 133; Lerner, supra, 1979 Sup. Ct. Rev

at 23; Gwartney, Asher, Haworth & Haworth, supra, 54 Notre Dame Law., at

642 0.20

2% Even petitioner’s most vocal supporters disagree with her assertion (Br

39-41, 44-49) that employers would not be forced down one of these two

roads. See, e.¢., Bartholet, supra, 95 Harv. L. Rev. at 1026-1027 (“Another

result might be increased pressure for quota or racially proportionate hiring

among those satisfying ceriain minimum qualifications{,| * * * [fjor u will

probably be difficult to demonstrate the job-relatedness and business necessity

of many of these selection systems, in large part because ut ts difficult to deter

mine what good performance is and how to measure ut. * * * Selection on a

racially proportionate basis seems an appropriate solution.”), Blumrosen,

supra, 63 Chi.Kem L. Rev. at 17, 35-36 (same)

” Petitioner errs in suggesting (Br. 26-29) that the intent of the 1972 Con

gress is relevant here. In 1972, Congress did not amend the language of Ses

tion 706¥¢Maxl) and (2) in any pertinent respect. See Equal Employment

24

As this Court has noted, “Title Vil would not have |t

enacted into law without substantial support from legisiator

both Houses who traditionally resisted tederal reguiatior

private business (S/fee/workers \ Weher. 444 US. 1934 ut”

(1979)). See also Firefighters v. Cleveland, No. 84-1999 (July 2

1986), slip op. 16-18; Sheet Metal Workers v. EEOC, N

84-1656 (July 2, 1986), sip op 27-39 [hese legisiat

resistance arose trom their Dele! that “jajny attem;

prescribe the qualifications that employers may or may

in 1b selection necessarily conthicts with a value thal g

held by members of a broad spectrum of American

ilue of employer autonomy” (Maltz, supra Nw. U. LR

it /3Y) [hese “legislators demanded as a { lor

port fa concession] that ‘management prerogat

doms |would] De ltelt undisturbed ne grea

[ Te) (WU eher 4()3 i iT OH | 1)) |

; \ 7 } : ; f a/

se 6 na

j ‘ 44 i ; 4 ;

i ' x ; ;

: ’ ;

W hile ' i i<Da ind

; i

EEO i ;

i K 4 ’

. H/ 2). TLR. OR ; . Y f

; A am ¢ be i* st. ) ) ‘ A

bh ‘ of ; ; . kK " .

» tl

:

: | vil

j ; { ii [ 4

, As } i

. ; \1 sm ;

a s'y ify

> . >

: ’ ™ :

.

25

even assuming that objective selection criteria exist or can

reasonably be derived for all job categories, '? employers cannot

be required to replace their subjective selection processes with

those objective criteria; to do so would create the very

bureaucratization and regimentation of the employee selection

process that the key legislators would not accept.

Moreover, given the generally prohibitive cost of converting

subjective selection processes to validated objective ones, the

more likely response of employers would be to superimpose

numerical quotas on their subjective selection processes. This

Court, of course, has held that Title VII pernurs employers to

engage in limited forms of voluntary affirmative action. See

Johnson v. Transp. Agency, supra. But the Court has also tirm-

ly asserted that Section 703(j)’’ was added to Title VII to assure

that the statute would not be interpreted to “require employers

or labor unions to use racial quotas or to grant preferential

trearment to *** avoid being charged with unlawful

discrimination” (Sheet Metal Workers, slip op. 27). The con-

gressional record is replete with comments “that employers

would not be required to institute preferential quotas to avoid

Title VU liability” (Weber, 443 U.S. at 207 n.7 (emphasis in

original)). Thus, a rule of law that would “leavie] * * *

employer|s} [with] little choice * * * but to engage in * * * sub-

jective quota systems] of employment” would be “far from the

intent of Tithe VII" (Albemarle Paper Co., 422 U.S. at 449

(Blackmun, J., concurring in the judgment)).

‘2 Of course, this ts an unrealistic assumption. For many employment dect-

sions, especially those involving professional or managerial positions, a com-

plete and sufficient range of such objective criteria does not exist. See Rogers

\. Int'l Paper Co., $10 F.2d 1340, 1345 (8th Cir.), vacated on other grounds,

423 U.S. 809 (1975); Gortleb v. Tulane Univ. of Louisiana, 809 F.2d 278, 283

& 0.6 (Sth Cir. 1987); Lerner, supra, 1979 Sup. Ct. Rev. at 23.

'* Section 70%j) provides that “[njothing contained in this subchapter shall

be interpreted to require any employer * * * to grant preferential treatment to

any individual or to any group because of the race * * * of such individual or

group on account of an imbalance which may exist with respect to the total

number or percentage of persons of any race * * * employed by any employer

* * * in comparison with the total number or percentage of persons of such

race * * * in any * * * area, or in the available work force in any * * * area”

(42 U.S.C. 2000e-2)))

i) lhe | : scle line tin tmopl ree Selection Procedures

’ : }’ + bale i hal subprective sSetection Processes Niust Hy

27

Guidelines goes on to state, however, that “[t}here are cir-

cumstances in which a user cannot or need not utilize the valida-

tion techniques contemplated by these guidelines,” that such

“circumstances” include the use of “informal or unscored selec-

tion procedure{s},” and that, where such informal or unscored

selection procedures are involved, the user need only “justify

continued use of the procedure[s] in accord with Federal law”

(29 C.F.R. 1607.6B). Unfortunately, the Guidelines do not

clarify how a user “justiffies}] continued use of the procedure(s]

in accord with Federal law.”*’ And to the extent that additional

explanation concerning what the Guidelines require in such

situations has been offered, the signals have been mixed. ’*

2. In the view of the United States and the EEOC, as a

practical matter, an employer “justiffies] continued use of the

procedure in accord with Federal law,” within the meaning of

Section 6B (29 C.F.R. 1607.6B), by showing that the selection

procedure is reasonably related to the requirements of the job or

other business needs and has been applied in a nonpretextual

fashion, as the employer would do in any disparate treatment

case (and as the courts below held that respondent did in this

case). Where an employer chooses to rely on the predictive value

of an objective selection device, it is reasonable to require that

employer to justify its reliance under the disparate impact

theory. But where an employer chooses to use subjective selec-

* Commentators on both sides of the debate have agreed with our

characterization of the Guidelines. See Bartholet, supra, 9S Harv. L. Rev. at

983. OBB 9.141; Larson, § 76.333, at 18-86 to 15-87.

'* When the Guidelines were first promulgated, the pretatory comments

stated that “[a] few practices may be used without validation even uf they have

an adverse impact. See, ¢.¢., McDonnell Douglas ¥. Green, 411 U.S ~%.)

(1973) and section 6B.” 43 Fed. Reg. 38291 9.17 (1978). Six months later, im

response lo a question Concerning how users can “pustity continued use of a

procedure on a basis other than validity,” the agencies said, “The method ot

justification of a procedure by means other than validity ts one to which the

CGundelines are not addressed. See Section 6B.” 44 Fed. Reg. 12002 (1979). The

response added that, “jijn Griges|,| * * * the Supreme Court ind ated that the

burden on the user was a heavy one,” that “the selection procedure could be

used if there was a “business necessity’ for its continued use,” and that “the

Federal agencies will consider evidence that a velection procedure ty Necessary

tor the safe and efficient operation of a Dusimess to pustity commued use of

selection procedure” (bed )

CONCLUSION