Watson v. Fort Worth Bank and Trust Brief for Respondent

    ■4 >

No. 86-6139

IN THE

Supreme Court of the United States
OCTOBER TERM, 1987

CLARA B. WATSON, Petitioner

v.

FORT WORTH BANK & TRUST, Respondent

On Writ of Certiorari to the United States 
Court of Appeals for the Fifth Circuit

Brief for Respondent

Bruce W. McGee*
Joseph W. Spence
GANDY MICHENER SWINDLE

WHITAKER & PRATT 
2501 Parkview Drive, Suite 600 
Fort Worth, Texas 76102 
817/878-0565

Attorneys for Respondent 
*Counsel of Record

< >
Fort Worth Brief Printing Co., 610 South Jennings Avenue, Fort W orth, Texas 76104 Phone 817-332*4070



1

TABLE OF CONTENTS
Page

TABLE OF AUTHORITIES........................................................ii

STATEMENT OF FACTS.........................................................  1

SUMMARY OF ARGUMENT...............    5

ARGUMENT.............................................................................  7
I. The Decisions of This Court Restrict Disparate Impact 

Analysis to Objective Selection Procedures.....................  7
II. The Legislative History of the 1972 Amendments to 

Title VII Does Not Support Application of Disparate 
Impact Analysis to Subjective Selection Criteria..............18

III. Administrative Interpretation Has Not Been Consistent
Regarding Application of Subjective Employee Selection 
Procedures to Disparate Impact Analysis.........................27

IV. The Disparate Impact Theory Should Not Be Applied
to Claims Involving Subjective Employment Practices 
or Criteria Because the Disparate Impact Theory Is 
Inappropriate for Analysis of Such Claims............   33

V. The Better Reasoned Decisions of the Courts of 
Appeals Recognize That the Disparate Impact Model 
Is Not Appropriate for Analyzing Subjective 
Employment Judgments................................................... 42

CONCLUSION 48



Cases
TABLE OF AUTHORITIES

Page
Albemarle Paper Co. v. Moody,

422 U.S. 405 (1975)............. !................................... 12, 13, 14, 30, 34
Allen v. City of Mobile, 466 F.2d 122 (5th Cir. 1972).......................25
Atonio v. Wards Cove Packing Co., 810 F.2d 1477 

(9th Cir. 1987) (en banc)..................................................... 43, 45
Bauer v. Bailar, 647 F 2d  1037 (10th Cir. 1981)............................... 35
Brewster v. Barnes, 788 F.2d 985 (4th Cir. 1986)..............................44
City of Los Angeles Dept, of Water and Power v. Manhart,

435 U.S. 702(1978).............................................................................. 40
Coates v. Johnson & Johnson Co., 756 F.2d 524

(7th Cir. 1985)...................................................................................... 44
Coe v. Yellow Freight Sys., Inc., 646 F.2d 444

(10th Cir. 1981).....................................................................................38
Connecticut v. Teal, 457 U.S. 440 (1982)...........  15, 17, 26, 34, 40, 41

Davis v. Califano, 613 F.2d 957 (D.C. Cir. 1980)............................. 41
Dothardv. Rawlinson, 433 U.S. 321 (1977)...........  12, 13, 14, 23, 34
Douglas v. Hampton, 512 F.2d 976 (D.C. Cir. 1975)........................ 30

EEOC v. Greyhound Lines, Inc., 635 F.2d 188 
(3d Cir. 1980)........................................................................................39

Firefighters Local Union No. 178U v. Stotts,
467 U.S. 561 (1984).............................................................................25

Fumco Construction Corp. v. Waters,
438 U.S. 567(1978).........................................................  13, 14, 36, 40

General Telephone v. EEOC, 446 U.S. 318 (1980)..................... . 20
Goodman v. Lukens Steel Co., 482 U .S____ , 96 L.Ed 2d

572, 584, 107 S. C t____ (1987)......................................................... 42
Griffin v. Carlin, 755 F.2d 1516 (11th Cir. 1985).........................43, 45
Griggs v. Duke Power Co., 401 U.S. 424 (1971).......................passim

Harris v. Ford Motor Co., 651 F.2d 609 (8th Cir. 1981).................. 44
Hazelwood School District v. United States,

433 U.S. 299 (1977)...................................... 11, 12, 13, 34, 37, 41, 42
Heagney v. University of Washington, 642 F.2d 1157

(9th Cir. 1981).....................................................................................39
Hill v. Seaboard Coast Line RR. Co., 767 F.2d 771 

(11th Cir. 1985) 35



Cases
TABLE OF AUTHORITIES (Continued)

iii

International Brotherhood of Teamsters v. United States,
431 U.S. 324(1977)......................................................

Johnson v. Transp. Agency, Santa Clara County, Califor 
nia, 480 U.S____, 94 L.Ed.2d 615 (1987)..............

Lewis v. Chicago State College, 299 F.Supp. 1357
(N.D. 111. 1969)....................................................................... 26

Lilly v. City ofBeckley, 797 F.2d 191 (4th Cir. 1986)....................39
McDonnell Douglas Corp. v. Green, 411 U.S. 792 (1973)......passim
Moody v. Albemarle Paper Co., 474 F.2d 134 (4th Cir. 1973).......  13
Mortenson v. Callaway, 672 F.2d 822 (10th Cir. 1982).................43
New York City Transit Auth. v. Beazer, 440 U.S. 568 (1979).......34
NLRB v. Gullett Gin Co., 340 U.S. 361 (1951).................. .........  18
Paige v. Marsh, No. 86-1282 (slip op; available on Lexis)

(8th Cir. Feb. 3, 1987)............................................................. 44
Pattern Makers League v. NLRB, 473 U.S. 95 (1985).................33
Payne v. Travenol Laboratories, Inc., 673 F.2d 798

(5th Cir. 1982)...................................................................41, 42
Pouncy v. Prudential Insurance Co. of America, 668 F.2d 

795 (5th Cir. 1982)............................................. 41, 42, 43, 44, 45
Reed v. Lockheed Aircraft Corp., 613 F.2d 757

(9th Cir. 1980).........................................................................41
Robinson v. Union Carbide Corp., 538 F.2d 652

(5th Cir. 1976), cert, denied, 434 U.S. 822 (1977).....................37
Rossini v. Ogilvy & Mather, Inc., 798 F.2d 590

(2d Cir. 1986).................................................................... 44, 45
Rowe v. Cleveland Pneumatic Co. Numerical Control,

690 F.2d 88 (6th Cir. 1982)...........................................  37, 38, 43
Rowe v. General Motors Corp., 457 F.2d 348 (5th Cir. 1972).......37
Segarv. Smith, 738 F.2d 1249 (D.C. Cir. 1984), cert, denied

sub nom. Meese v. Segar, 471 U.S. 1115 (1985).............43, 46, 47
Smith v. Board of Education of Morrilton School District,

365 F.2d 770 (8th Cir. 1966)............................................... 25, 26
Stastny v. Southern Bell Telephone & Telegraph Co.,

628 F.2d 267 (4th Cir. 1980)

Page 

passim 

. 21, 23

44



IV

Texas Department of Community Affairs v. Burdine,
450 U.S. 248 (1981)............................................................passim

United States v. Georgia Power Co., 3 FEP Cas. 767 (N.D.
Ga. 1971), affd in part and vacated in part, 474 F.2d 906
(5th Cir. 1973)..........................................................................30

United States Postal Service Board of Governor v. Aikens,
460 U.S. 711 (1983)...................... '.....................................16, 36

United Steelworkers of Am. v. Weber, 443 U.S. 193 (1979)..........39
Wang v. Hoffman, 694 F.2d 1146 (9th Cir. 1982)............................35
Waters v. Fumco Construction Co., 551 F.2d 1085 

(7th Cir. 1977)......................................................................... 13
Zahorik v. Cornell University, 729 F.2d 85 (2d Cir. 1984)..............35
Ziman v. USPS, No. 01842595 (July 23, 1986)............................. 33

Statutes and Regulations
Civil Rights Act of 1964, Tit. VII, 42 U.S.C. 2000e, et seq. .. passim

Section 703(a)(2), 42 U.S.C. 2000e-(2)(a)(2).................. 18, 26, 41
Section 703(h), 42 U.S.C. §2000e-(2)(h).................. 10, 11, 20, 28
Section 703(j), 42 U.S.C. §2000e-2(j)........................................ 40
Section 706(f), 42 U.S.C. §2000e-f(f)(l)..................................... 20
Section 715............................................................................... 31
42 U.S.C. 1981......................................................................  3
42 U.S.C. 1983....................................................................25, 26

Decision of the EEOC, CCH Empl. Prac. Guide H17304.53
(Dec. 2, 1966)........................................................................... 28

Equal Employment Opportunity Act of 1972,
Pub. L. 92-261, 86 Stat. 103.................................................18, 31

Guidelines on Employee Selection Procedures, 35 Fed.
Reg. 12333 (Aug. 1, 1970)................................................... 29, 33
1607.2 .................................................................................... 29
1607.3 ................................................................................ 29, 31
1607.4 - 1607.8....................................................................29, 30
1607.13 ............................................................................... 29, 32

Guidelines on Employment Testing Procedures,
August 24, 1966 ................................................................. 28, 29

TABLE OF AUTHORITIES (Continued)
Cases Page



V

Uniform Guidelines On Employee Selection Procedures,
29 C.F.R. 1607 (1978).................................................... 31, 32, 33
1607.3A ....................................................................................31
1607.6B..................................................................................... 32
1607.6B(1).................................................................................32
1607.16Q................................................................................... 31

41 Fed. Reg. 51734 (1976)............................................................. 31
43 Fed. Reg. 38291 (1978)............................................................32
44 Fed. Reg. 12002 (1979)............................................................. 32

Legislative History
H.R. 1746...................................................................    19, 20
H.R. 9247 ................................................................   19
H. Rep. No. 238, 92d Cong., 1st Sess. (1972)........................22, 23
H.R. Conf. Rep., 92d Cong., 2d Sess. (1972)................................24
S. 878.......................................................................................... 21
S. 884.......................................................................................... 21
S. 2515.............................................................................20, 21,23
S. Rep. No. 92-415, 92d Cong., 1st Sess. (1971).....................20, 22
110 Cong. Rec. 7218 (1964).........................................................  40
117 Cong. Rec.

31960 ................................................................................20, 22
31961 .................................................................................... 20, 24
32094 ..................................................................................... 24
32095 .....................................................................................20
32096 .....................................................................................20
32102........................................................................................20
32104........................................................................................20
32106........................................................................................20
32111........................................................................................19

118 Cong. Rec.
294 .......................................................................................... 21
296.......................................................................................... 24
580.......................................................................................... 21
590 .......................................................................................... 21
944-945.....................................................................................21
1815-1816..................................................................................24
3962......................................................................................... 20

TABLE OF AUTHORITIES (Continued)
Statutes and Regulations Page



VI

3965-3979...............................................................................................21
4080 ......................................................................................................20
4081 ......................................................................................................27
4082 ..............................................................................................20, 27
4944........................................................................................................ 23
6647........................................................................................................ 26
7166........................................................................................................ 24
7170...........................   . . . . . 23
7573........................................................................................................ 23

Other Authorities
Blumrosen, Strangers In Paradise: Griggs v. Duke Power 

Co. And The Concept of Employment Discrimination,
71 Mich. L. Rev. 59 (1972)........'........................................................ 28

Booth & MacKay, Legal Constraints on Employment Test­
ing and Evolving Trends in the Law, 29 Emory L.J. 121 
(1980)...................................................................................................... 30

Campbell, Regression Analysis In Title VII Cases: Mini­
mum Standards, Co?nparable Worth, and Other Issues 
Where Law And Statistics Meet, 36 Stan. L. Rev. 1299
(1984)..................................................................................................... 46

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)................................................................30, 31

Comptroller General of the United States, Report to the 
Congress: Problems With Federal Equal Employment 
Opportunity Guidelines On Selection Procedures Need 
to Be Resolved (1978)..........................................................................30

Gold, Griggs’ Folly: An Essay on the Theory, Problems 
and Origin of the Adverse Impact Definition of Employ­
ment Discrimination and a Recommendation for
Reform, 7 Indus. Rel. L.J. 429 (1985)................................. 8, 30, 31

Gold, Reply to Thomson, 8 Indus. Rel. L.J. 117 (1986)...........  18, 26

TABLE OF AUTHORITIES (Continued)
Legislative History Page



3A A. Larson & L. Larson, Employment Discrimination
(1987)...............................................................................................25, 35

Lerner, “Washington v. Davis”: Quantity, Quality and.
Equality in Employment Testing, 1976 Sup. Ct. Rev. 263........ 31

Meltzer, The 'Weber Case: The Judicial Abrogation of the 
Antidiscrimination Standard in Employment, 47 U.
Chi. L. Rev. 423 (1980)...................................................................... 26

Note, Application of the EEOC Guidelines to Employ­
ment Test Validation: A Uniform Standard for Both 
Public and Private Employers, 41 Geo. Wash. L. Rev.
505(1973)............................ '................................................................30

Note, The Uniform Guidelines on Employee Selection 
Procedures: Compromises and Controversies, 28 Cath.
U. L. Rev. 605(1979)...........................................................................31

Opinion Letter, dated October 2, 1965, GC 296-65, cited at 
Pet. App. 147b, in Griggs v. Duke Power Co., 401 U.S.
424 (1971).............................................................................................. 28

Sape and Hart, Title VII Reconsidered: The Equal 
Employment Opportunity Act of 1972, 40 Geo. Wash. L.
Rev. 824 (1972)............................. '...................................................... 26

Thomson, The Disparate Impact Theory: Congressional 
Intent in 1972-A Response to Gold, 8 Indus. Rel. L.J.
105 (1986).............................................................................................. 21

vii

TABLE OF AUTHORITIES (Continued)
Other Authorities Page



1

STATEMENT

In August of 1973, Clara Watson applied for employment at 
Fort Worth Bank & Trust. (J. A. 5). Although she had no bank­
ing experience when she applied for employment, Watson was 
offered employment by the bank and accepted a position as a 
proof operator. (J.A. 5, 190). Approximately one year after 
starting with the bank, Watson was promoted to the position 
of teller trainee. (J.A. 6 and I Tr. 21). Upon completion of her 
training, Watson became a teller in the motor bank or drive-in 
bank. (J.A. 7). She remained in the position of motor bank 
teller until 1980 when she was promoted to the position of 
commercial teller and informal assistant to Richard Burt. 
(I Tr. 47).

Up to 1978, the bank conducted oral evaluations of its 
employees. The evaluations considered various factors which 
the bank considered important to its business. (J.A. 8). In 
1979, the bank instituted a written evaluation system to eval­
uate employees. (II Tr. 80). The written evaluation forms 
were obtained from an outside consultant who performs simi­
lar services for other banks.

Beginning in 1979, the bank also adopted a posting proce­
dure whereby job openings in the bank were posted to inform 
current bank employees of the positions available. (II Tr. 36). 
An employee who wished to apply for the position would fill 
out a form or application for promotion to the position. (J.A. 
121). The supervisor over the open job position was in charge 
and had responsibility for interviewing each of the applicants, 
as well as evaluating the relative qualifications of the appli­
cants. The supervisor would then make his or her best judg­
ment as to which applicant would be given the job. (J.A. 131).

In February 1980, two supervisory positions became open 
at the bank. Watson was considered for both positions, how­
ever, Richard Burt was promoted to supervisor of lobby tell­



2

ers and Pat Cullar was promoted to supervisor of motor bank 
tellers. (I Tr. 40).

At the time of his promotion to supervisor of lobby tellers, 
Burt had experience in the general ledger department, as a 
credit analyst, as an assistant to Shipp, and as supervisor of 
the bookkeeping department. (Ill Tr. 28-29). He received his 
bachelor of arts degree in banking and finance shortly after his 
promotion to supervisor of lobby tellers. (Ill Tr. 28-29).

At the time of Cullar’s selection as supervisor of motor bank 
tellers, she was working as the assistant to the lobby teller 
supervisor whose place Burt had taken and had obtained valu­
able supervisory experience. (Ill Tr. 87). In addition, Cullar 
had approximately eighteen years experience in banking. 
(Ill Tr. 202-204).

In addition to Watson, a white employee who had experi­
ence as supervisor of the motor bank tellers was also not 
selected. Watson acknowledged that she was not as qualified 
as Gail Leavitt, the white applicant. (I Tr. 147).

Immediately after these supervisor positions were filled, 
Watson was placed in the informal position as assistant to 
Burt. (I Tr. 157-158). Watson remained in the position of com­
mercial teller and assistant to Burt until January of 1981. At 
that time, Watson was required to have foot surgery and took 
a leave of absence from the bank. (J.A. 18). While Watson was 
on leave, the positions of motor bank supervisor and lobby 
teller supervisor both became open again. (J.A. 18). Cullar, 
the motor bank teller supervisor, who had reported directly to 
Burt as lobby supervisor, was chosen by Burt to become the 
new lobby teller supervisor. (Ill Tr. 31). In choosing Cullar, 
Burt testified that there were many factors considered in 
arriving at his decision. Included in these factors were previ­
ous supervisory experience, leadership ability and the ability 
to get along with others. At the time Burt was considering the 
applicants, he formed an opinion that Cullar was the best qual­



3

ified candidate to assume the position of supervisor of lobby 
tellers. Further, Burt testified that he did not consider Wat­
son to be as qualified as Cullar because there had been resent­
ment from other tellers who had been under Watson’s direc­
tion during the time she acted as his assistant. (Ill Tr. 33).

Burt testified that in making his decision to select Kevin 
Brown for the other position he again considered the fact that 
Watson had dem onstrated an inability to get along with 
others. Burt was of the opinion that Watson did not have the 
respect nor the leadership ability that Brown possessed. (Ill 
Tr. 76). Burt considered the fact that Brown was a good teller, 
knew the people in the drive-in very well, and was of the opin­
ion that they had respect for him. Burt also considered the 
fact that Brown had worked as the “lead person” in the after­
noon shift in the motor bank. (Ill Tr. 76). As was the situation 
with the 1980 promotions, Burt considered a number of appli­
cants before selecting Cullar and Brown. (J.A. 131).

Watson filed suit in the District Court alleging that the bank 
discriminated against her and other persons similarly situated 
on the basis of race in violation of 42 U.S.C. §1981 and Title 
VII of the Civil Rights Act of 1964, 42 U.S.C. §2000e, et seq. 
In support of her individual claim, Watson attempted to show 
that she had been the victim of intentional discrimination in 
the bank’s decisions regarding the four promotions she 
sought. She also sought to show that the bank’s use of a sub­
jective, discretionary procedure for evaluating applicants for 
promotion resulted in a disparate impact upon her. Although 
she alleged a claim under both theories, Watson’s proof was 
clearly patterned after the disparate treatm ent model as 
described in McDonnell Douglas Corp. v. Green, 411 U.S. 792 
(1973). The thrust of Watson’s case (as supported by her own 
testimony) was that she had been intentionally denied a pro­
motion on four separate occasions by white supervisors on the 
basis of her race. (J.A. 45-47).



4

Although Watson offered certain statistical data concerning 
blacks and whites with regard to the hiring, promotion, and 
salary increase practices of the bank, Watson offered no evi­
dence which identified a specific, facially neutral practice of 
the bank which created an adverse impact upon blacks.

The bank challenged the statistical evidence offered by 
Watson, pointing out that the hiring data was flawed because 
Waton’s expert had failed to include at least 28 applicants in 
the data analyzed (II Tr. 183) and that her analyses failed to 
control for the relative qualifications of the applicants. (II Tr. 
186). Likewise, the bank established that Watson’s analyses 
were further flawed because her expert had based all of his 
analysis upon the dubious assumption that there were always 
job openings when black applicants sought employment at the 
bank. (II Tr. 185). Dr. Marshall also admitted that he did not 
consider the frequency which blacks and non-blacks actually 
sought promotions. (II Tr. 204). The District Court apparently 
determined that the statistical evidence offered by Watson 
was not probative because it found that her statistical evi­
dence did not present a prima facie case of race discrimination. 
(J.A. 202).

However, with regard to Watson’s individual claim, the trial 
court concluded that Watson had established a prima facie 
case of individual employment discrimination in that she was a 
member of a protected class, she applied for and was qualified 
for the position of lobby supervisor and motor bank supervisor 
in 1980 and 1981 and that she was not selected. The Court then 
concluded that the bank had met its burden of proof by artic­
ulating legitim ate non-discrim inatory reasons for each 
employment decision. The Court finally concluded that Wat­
son failed to establish that the reasons articulated by the bank 
for choosing Burt, Cullar, and Brown were pretexts for racial 
discrimination. Accordingly, judgment was entered that Wat­



5

son take nothing and th a t her action be dismissed on the 
merits.

On appeal to the Fifth Circuit, the District Court’s judg­
ment rejecting Watson’s individual claim was affirmed.

SUMMARY OF ARGUMENT

Petitioner complains that the Court of Appeals refused to 
examine her employment discrimination claims under the dis­
parate impact model announced by this Court in Griggs v. 
Duke Power Co., 401 U.S. 424 (1971). The Griggs decision 
permits utilization of the disparate impact theory only when 
objective, facially neutral tests or requirements are at issue. 
Petitioner seeks to expand the Griggs holding to permit the 
challenge of not only subjective employment criteria, but also 
the “bottom line” effect of an employer’s subjective employ­
ment process.

Eradication of intentional discrimination was the primary 
goal of Congress when Title VII was adopted. The disparate 
impact model of proof was crafted by this Court after Title 
VII was enacted to address employment practices which were 
neutral in intent but had a disparate impact upon protected 
groups. Subsequent opinions of this Court applying the dis­
parate impact model have been careful to follow the distinc­
tion between “objective” and “subjective” employment crite­
ria. While this Court has perm itted objective employment 
practices to be tested under both the treatm ent and impact 
theories, the Court has refused to extend impact analysis to 
subjective criteria such as those which are present in this 
case.

Contrary to the assertions of petitioner, Congress did not 
mandate that disparate impact analysis should apply to all 
employment practices when it passed the 1972 amendments to 
Title VII. Accordingly, the legislative history of the 1972



6

amendments have little or no relevance to this Court’s inquiry 
in this case.

Petitioner’s reliance upon the allegedly consistent view of 
the government with regard to analysis of subjective employ­
ment criteria by application of the disparate impact model is 
w ithout basis. Contrary to petitioner’s assertions to this 
Court, the position of the EEOC on the issue has changed sev­
eral times since Title VII was enacted. The current guidelines 
recognize that certain employment criteria cannot or need not 
be validated and provide that the employer must only “justify 
continued use of the procedurejs] in accord with Federal law.”

The expansion of Griggs sought by petitioner is unwar­
ranted and inappropriate. The disparate treatm ent model is 
the only appropriate method for analyzing subjective employ­
ment practices of the type utilized by respondent. Disparate 
impact analysis is only appropriate for analyzing employment 
criteria which are applied to all applicants mechanically and 
without exception.

Although as this Court observed in International Brother­
hood of Teamsters v. United States, 431 U.S. 324 (1977), both 
the disparate treatment and disparate impact theories may be 
applied to a particular set of facts, the theories are not inter­
changeably applied to all sets of facts. When the employment 
practices challenged are subjective, the basis of the complaint 
is that the subjective criteria have been applied in a discrimi­
natory fashion. The disparate treatment model provides the 
only proper framework for analyzing such complaints.

If the position advocated by the petitioner is adopted 
employers will be faced with the impossible burden of validat­
ing all subjective criteria involved in normal personnel deci­
sions. While perhaps it is theoretically possible to validate 
subjective criteria, practically speaking the economic cost to 
small employers makes the task impossible. As is demon­
stra ted  by the b e tte r reasoned opinions of the Court of



7

Appeals, the disparate treatment model provides an appro­
priate method for analyzing the application of subjective 
employment criteria. There is no basis or need to extend dis­
parate impact analysis to subjective criteria.

ARGUMENT

P etitioner presented a case alleging th a t she was not 
selected for promotion to four supervisory positions because 
of her race and further that the bank’s use of subjective selec­
tion procedures had a disparate impact upon her. The District 
Court considered fully the statistical and anecdotal evidence 
presented by both parties and determined that Watson had 
not met her burden of proof under the disparate treatm ent 
model. The Court of Appeals affirmed the trial court’s judg­
ment regarding Watson’s individual claim. Watson now asks 
this Court to ignore its previous decisions and to permit her to 
employ the disparate impact model to attack the bank’s use of 
a selection process which utilizes subjective evaluation of mul­
tiple factors for the purpose of selecting the best qualified 
applicant. I. * * * * * VII,

I. The Decisions of This Court Restrict Disparate Impact Anal­
ysis to Objective Selection Procedures

The basis for the disparate impact theory of proof is, of
course, this Court’s opinion in Griggs v. Duke Power Co., 401
U.S. 424 (1971). In Griggs, this Court announced that Title
VII applied not only to intentional discrimination, but also to
an employment practice or selection criterion which is “fair in 
form, but discriminatory in operation.” Griggs at 431. At issue 
in Griggs was the employer’s requirement that employees 
have a high school degree and that they pass an aptitude test 
before they were qualified for promotion. In analyzing Title
VII, this Court emphasized that the Act does not command 
that any person be hired simply because he was formerly the



8

subject of discrimination or because he is a member of a 
minority group, rather, what is required by Congress is the 
removal of artificial, arbitrary, and unnecessary barriers to 
employment when the barriers operate invidiously to discri­
minate on the basis of racial or other impermissible classifica­
tion. Griggs, 401 U.S. at 431. Petitioner has construed the 
holding in Griggs as a proclamation that disparate impact 
analysis is the rule rather than the exception in Title VII anal­
ysis. Petitioner emphasizes the Court’s reference to “prac­
tices, procedures, or tests neutral on their face” as referring 
to any and all employment practices or procedures. However, 
in Griggs, the Court expressly defined the “p ractices” 
referred to as the objective, facially neutral requirements of 
high school degrees and the passing of a general intelligence 
test.

Petitioner has attempted to expand the holding in Griggs by 
arguing that the Griggs’ opinion does not expressly limit dis­
parate impact analysis to objective criteria in that there was 
no distinction made between subjective and objective employ­
ment practices in this Court’s opinion. While it is true that the 
terms “subjective” or “objective” were not used to label the 
specific employer practices being challenged, it is clear that 
the holding in Griggs focused only on the narrow issue of the 
employer’s requirement of a high school diploma and stan­
dardized testing requirements.

In enacting Title VII, there is little question that Congress 
was principally concerned with intentional discrimination.1 
However, this Court construed Title VII to encompass not 
only intentional discrimination, but also that which is “neutral 
in terms of intent”. Griggs at 430. Although this Court did hold 
that employment practices which were neutral in intent were 
prohibited if they had a disparate effect upon minorities, noth- *
'See Gold, Griggs' Folly: An Essay on the Theory, Problems, and Origin of 
the Adverse Impact Definition of Employment Discrimination and a Rec­
ommendation for Reform, 7 Indus.Rel. L.J. 429, 497-503 (1985).



9

ing in the opinion would suggest that this Court considered 
disparate impact analysis appropriate for testing discretion­
ary employment practices.2

In fact, the Court’s narrow analysis of Title VII and its ulti­
mate holding is reflected in the final paragraph of the opinion 
wherein the Court states:

Nothing in the Act precludes the use of testing or 'mea­
suring procedures-, obviously they are useful. What Con­
gress has forbidden is giving these devices and mecha­
nisms controlling force unless they are demonstrably a 
reasonable measure of job performance. Congress has 
not commanded that the less qualified be preferred over 
the better qualified simply because of minority origins. 
F ar from disparaging job qualifications as such, Con­
gress has made such qualifications the controlling factor, 
so that race, religion, nationality, and sex become irrele­
vant. What Congress has commanded is that any tests 
used must measure the person for the job and not the 
person in the abstract. Griggs, 401 U.S. at 436. (Empha­
sis added).

The distinction between disparate treatment and disparate 
impact analysis was addressed by this Court in International 
Brotherhood o f Teamsters v. United States, 431 U.S. 324
(1977). The governm ent alleged th a t the em ployer had 
engaged in a pattern  or practice of discriminating against 
minorities in hiring “line drivers”. In addition, the government 
challenged the seniority system established by the collective 
bargaining agreements between the employer and the union 
on the basis that the seniority system contained therein had 
an adverse impact on minorities. In analyzing the alleged dis­
2Contrary to the implication from Petitioner’s Brief at 17, the Solicitor Gen­
eral did not advocate application of the disparate impact model to discre­
tionary employment practices. The Solicitor General is misquoted. The 
final sentence of the quoted portion of the Solicitor General’s Amicus Brief 
in Griggs states that “Discriminatory ‘employment practices’ -  not atti- 
tudes-are declared unlawful.” Brief of the United States as Amicus Curiae 
at 16, Griggs v. Duke Power Co., October Term, 1970. (Emphasis added).



10

parity in treatm ent involving the refusal to recruit, hire, 
transfer or promote minority group members on an equal 
basis with whites, this Court applied the disparate treatment 
analysis set forth in McDonnell Douglas Corp. v. Green; see, 
Teamsters, 431 U.S. at 335. In holding that the disparate 
treatment analysis applied to the recruiting, hiring, transfer­
ring and/or promoting of minority group members, this Court, 
by way of footnote, specifically distinguishes disparate treat­
ment from disparate impact:

Claims of disparate treatment may be distinguished from 
claims that stress “disparate impact”. The latter involved 
employment practices that are facially neutral in their 
treatment of different groups but that in fact fall more 
harshly on one group than another and cannot be justi­
fied by business necessity. Teamsters, 431 U.S. at 335, 
see n.15.

While the Court noted that either theory may be applied to 
a particular set of facts, it is significant that Teamsters specif­
ically distinguished claims which stress disparate impact from 
claims involving the employer’s practices of recruitment, hir­
ing, transferring, and/or promotion of minority group mem­
bers. The Court made clear that such claims were to be ana­
lyzed as under the disparate treatment model.

In considering the government’s contentions that the collec­
tive bargaining agreement and its seniority system had an 
adverse impact on minorities, the Court in Teamsters had an 
opportunity to further explore the application of the disparate 
impact analysis. After reviewing the application of the senior­
ity system, the Court determined that the ultimate effect was 
that it “locked” minority workers into inferior jobs and per­
petuated prior discrimination by discouraging transfers to 
jobs as line drivers. While the Court determined tha t the 
seniority system was protected by Section 7.03(h) as a “bona 
fide” seniority system , the Court noted th a t otherwise it



11

would have fallen within the Griggs rationale. The Court 
stated:

One kind of practice “fair in form, but discriminatory in 
operation” is that which perpetuates the effects of prior 
discrimination. As the Court held in Griggs'. “Under the 
Act, practices, procedures, or tests neutral on their face, 
and even neutral in terms of intent, cannot be maintained 
if they operate to ’freeze’ the status quo of prior discrim­
inatory employment practices. (Citations omitted).
Were it not for Section 7.03(h) the seniority system in 
this case would seem to fall under the Griggs rationale. 
Teamsters, 431 U.S. at 349.

It was the seniority systems’ effect of segregating minority 
employees in the less desirable jobs, and otherwise adversely 
impacting minorities, which caused the Court to conclude it 
would be subject to disparate impact analysis. As in Griggs, 
the seniority system created an artificial barrier to advance­
ment by members of a protected class. As the Court in Team­
sters observed, with regard to claims of discriminatory hiring 
practices, the “controlling legal principals are relatively clear.” 
Teamsters at 334-35. The disparate treatm ent model was held 
to be appropriate for analyzing such practices, not impact 
analysis. Contrary to petitioner’s view of Teamsters, the 
seniority system was not the only employment practice at 
issue. This Court was also faced with a direct challenge of a 
largely subjective employment system which resulted in a 
racial imbalance, just as were Watson’s allegations here. This 
Court held that it was clear that such claims should be ana­
lyzed under the disparate treatment model. That same “rela­
tively clear” legal principal was properly used to analyze Wat­
son’s claims.

Petitioner relies on Hazelwood School District v. United 
States, 433 U.S. 299 (1977). However, petitioner’s suggestion 
that Hazelwood is authority for the application of disparate



12

impact analysis to subjective criteria is incorrect. Hazelwood 
is authority for application of a “pattern or practice” test to 
Watson’s claim, which is of course an analysis for disparate 
trea tm en t, not d isparate impact. As petitioner correctly 
observes, the facts of Hazelwood are very similar to those at 
hand, and the claim asserted by Watson was properly tested 
under the disparate treatm ent model.

That Hazelwood is authority for application of disparate 
treatm ent analysis to employment practices which grant to 
supervisory personnel “virtually unlimited discretion”3 to 
evaluate applicants for promotions based upon numerous 
intangible criteria, is further borne out by this Court’s opinion 
in Dothard v. Rawlinson, 433 U.S. 321 (1977). Decided the 
same day as Hazelwood, Dothard applied disparate impact 
analysis to certain minimum height and weight requirements. 
In Dothard, the Court again took pains to distinguish the 
facially neutral employment standards at issue from the cases 
where proof of intent was required. The Court stated:

The gist of the claim that the statutory height and weight 
requirem ents discriminate against women does not 
involve an assertion of purposeful discriminatory motive. 
It is asserted, rather, that these facially neutral qualifi­
cation standards work in fact disproportionately to 
exclude women from eligibility for employment with the 
Alabama Board of Corrections. We dealt in Griggs v. 
Duke Power Co. and Albem arle Paper Co. v. Moody 
(citations omitted), with similar allegations that facially 
neu tra l employment standards disproportionately  
excluded Negroes from employment, and those cases 
guide our approach here. Dothard, 433 U.S. at 328-29.

It is significant that Hazelwood was not one of those cases 
which the Court mentioned as guiding its consideration of the 
height and weight requirements at issue in Dothard. Decided 
by this Court on the same day, Dothard and Hazelwood distin­
iHazelwood School District v. United States, 433 U.S. 299 at 302 (1979).



13

guish this Court’s application of disparate impact analysis to 
the objective employment standards present in Dothard from 
disparate treatm ent analysis of the essentially subjective 
employment practices utilized in Hazelwood. Although this 
Court did not engage in characterizing the practice as “objec­
tive” versus “subjective”, it seems clear from the facts in 
Dothard th a t the employer practice in question is clearly 
objective.4

In Furnco Construction Corp. v. Waters, 438 U.S. 567
(1978), this Court once again distinguished between disparate 
treatment and disparate impact analysis. Petitioner has rep­
resented to the Court on page 23 of her Brief that the Court of 
Appeals concluded that Furnco’s policy of not hiring at the 
gate was racially neutral on its face, but that there was no 
showing that the policy had a disparate impact or effect. How­
ever, petitioner’s assertion is incorrect. Rather, in applying 
the narrow holding of Griggs, the Court of Appeals stated:

As we understand the Griggs' principle, it provides a 
method for establishing the presence of discrimination 
where application of a test of a group of employees or 
applicants results in a proportionately higher rate of 
rejection of members of a minority group. Here the reli­
ance upon the superintendent’s recollection of brick lay­
ers he considered competent was not a test or device for 
comparing the members of a group of prospective 
employees, where there was no defined group. . . .
This was not a case, like Griggs, of comparing white or 
black applicants by objective, but irrelevant standards, in 
rejecting more blacks, proportionately, than white. 
(Em phasis added). Waters v. Furnco C onstruction  
Corp., 551 F.2d 1085 at 1089-90 (7th Cir. 1977).

4Dothard further relies on Albemarle Paper Co. v. Moody, 422 U.S. 405 
(1975) in support of its disparate impact analysis. Albemarle involved a 
class action brought by black employees who challenged a job seniority sys­
tem and the requirement of employment testing conducted by standardized 
written test. (Objective criteria). See Moody v. Albemarle Paper Co 474 
F.2d 134 (4th Cir. 1973).



14

This Court agreed with the Court of Appeals th a t the 
proper approach was the analysis contained in McDonnell 
Douglas (disparate treatment). Fumco, 438 U.S. at 575. In a 
supporting footnote, this Court noted that:

This case did not involve employment tests, which we 
dealt with in Griggs v. Duke Power Co. And in Albemarle 
Paper Co. v. Moody, for particularized requirements 
such as the height and weight specifications considered in 
Dothard v. Rawlinson, and it was not a “pattern  and 
practice” case like Teamsters v. United States. (All cita­
tions omitted). Fumco, 438 U.S. at 575 n. 7.

Petitioner tries to down play the significance of footnote 7 of 
Fumco  by claiming that the footnote is simply a reiteration of 
the disparate impact cases decided by the Court as of that 
date. However, the footnote supports the single sentence 
holding of the Court that the proper approach in Fumco  is the 
disparate treatment analysis of McDonnell Douglas v. Green. 
It seems highly improbable that the Court, after holding that 
the disparate treatment analysis applies, would support that 
holding with a footnote re ite ra ting  the various disparate 
impact cases which it had decided. It seems obvious that foot­
note 7 was added to distinguish between the factual situations 
which require disparate treatment analysis and those where 
disparate impact analysis is appropriate.

A few years later, in Texas D epartm ent o f C om m unity  
Affairs v. Burdine, 450 U.S. 248 (1981), this Court explained 
the basic allocation of burdens and order of presentation 
approved in a disparate treatm ent case under McDonnell 
Douglas Corp. v. Green. Burdine, 450 U.S. at 252. Nowhere 
in Burdine does the Court focus on disparate impact analysis, 
rather, the Court clearly indicates that the respondent’s Title 
VII claim was properly analyzed under disparate treatment 
theory. Burdine alleged that the failure to promote her and



15

The significance of this Court’s application of the disparate 
treatm ent analysis to the facts in Burdine becomes more sig­
nificant in light of this Court’s application of the disparate 
impact analysis to the facts in Connecticut v. Teal, 457 U.S. 
440 (1982). The plaintiffs alleged that the employer violated 
Title VII by imposing, as an absolute condition for considera­
tion for promotion, that applicants pass a 'written test that 
excluded blacks in disproportionate numbers and that it was 
not job related. Teal, 457 U.S. at 444. The employer asserted 
a “bottom line” result defense in that the overall result of the 
selection process showed that a greater percentage of blacks 
were promoted.

This Court, in its m ajority opinion, concluded tha t the 
employees’ claim should be analyzed under the disparate 
impact analysis of Griggs. In so holding, the majority consis­
tently focuses on the fact that the case concerns a specific, 
facially neutral employment test that allegedly had a dispar­
ate impact on minorities.

Petitioner misconstrues Teal by asserting that the majority 
and the dissent agree that the disparate impact approach 
includes examination of the entire range of employee selection 
devices. The references of the dissent to the “total selection 
process” clearly relate to the majority’s rejection of the “bot­
tom line” defense asserted by the employer. The dissent is not 
suggesting that disparate impact analysis can be applied to a 
multiple component employment selection process. Teal dem­
onstrates this Court’s consistent limitation of the Griggs test 
to objective requirements and tests and that an employer will 
not be permitted to raise a “bottom line” defense to defeat a 
prima facie case of disparate impact. Teal is not to be read as 
an expansion of Griggs. As the m ajority holds, disparate 
impact analysis is not applied to the employer’s entire selec­

the subsequent decision to term inate her had been predicated
on gender discrimination in violation of Title VII.



16

tion process, but is only applied to distinct requirements and 
tests within the selection process.

Perhaps this Court’s decision which is closest to the facts of 
the case at hand is United States Postal Service Board of Gov­
ernor v. Aikens, 460 U.S. 711 (1983). Aikens involved an indi­
vidual claim that the employer had discriminatorily refused to 
promote. In Aikens, the plaintiff showed that white employ­
ees consistently received promotions over him and other black 
employees, and that he was be tter qualified than the white 
who was selected for promotion in all but one instance. The 
plaintiff also introduced evidence showing that the person 
who was responsible for the promotion decisions had made 
numerous derogatory comments about blacks in general and 
the plaintiff in particular. Aikens at 713. These facts are quite 
similar to those which Watson claims to have established in 
support of her individual claim.

The Court left no doubt tha t the M cDonnell Douglas/ 
Burdine model should be applied to the Plaintiffs claim:

We have consistently distinguished disparate-treatment 
cases from cases involving facially neutral employment 
standards that have disparate impact on minority appli­
cants. See, e.g. Texas Department o f Community Affairs 
v. Burdine, 450 US 248, 252, n5, 67 L Ed 2d 207, 101 S Ct 
1089 (1981); McDonnell Douglas Corp. v. Green, 411 US 
792, 802, n 14, 36 L Ed 2d 668, 93 S Ct 1817 (1973). Aikens 
at 713.

Although the Court does not expressly so sta te , it seems 
obvious that disparate treatm ent analysis was appropriate 
because of the subjective criteria involved in the employment 
selections challenged. The exercise of discretion in choosing 
one applicant over another is not a facially neutral employ­
ment standard which may be tested by impact analysis. On 
the contrary, a claim that one was preferred over a minority 
employee as a result of a subjective interview process, where



17

the employer exercises discretionary judgment to chose the 
applicant best qualified, will always be a claim that the minority 
applicant was the victim of intentional racial discrimination.

An examination of this Court’s decisions following Griggs, 
and those which distinguish Griggs, clearly demonstrate that 
subjective employment practices are not appropriate for dis­
parate impact analysis. As Justice Brennan acknowledged in 
Teal, disparate impact analysis tests for discrimination which 
arises from facially neutral policies, while disparate treatment 
involves facially discriminatory policies. Teal at 455.

Watson complains of separate decisions made by different 
members of the bank’s supervisory staff. She says that her 
nonselection was due to the bank’s consideration of her race. 
There is little question that her position before the trial court 
was that she was the victim of intentional discrimination. 
After she could not satisfy her burden of proof to make a case 
of disparate treatment, she now attempts to characterize the 
bank’s selection procedures as the same type involved in 
Griggs and its progeny. But as Griggs and those decisions of 
this Court following it have shown, the bank’s evaluation pro­
cess for determining which of several applicants is to receive a 
particular promotion is not the kind of facially neutral require­
ment or test which is appropriate for disparate impact analy­
sis. Watson cannot point to any specific criteria which causes 
the disparate impact, but acknowledges that the bank relied 
upon numerous factors in its decisions. Watson’s claim was 
properly tested  under the M cDonnell Douglas!Burdine 
model in that she was afforded a full and fair opportunity to 
convince the trier of fact that the bank’s actions were moti­
vated by intentional discrimination because of her race. The 
decisions of this Court offer no basis for now analyzing her 
claim under the disparate impact model. While the decisions 
of the Court do not refer by name to the distinction between 
claims which involve subjective selection criteria and those



18

which involve objective criteria, a review of the Court’s deci­
sions clearly demonstrates tha t the Court has consistently 
limited disparate impact analysis to those claims where objec­
tive, facially neutral selection criteria were present.

II. The Legislative History of the 1972 Amendments to Title
VII Does Not Support Application of Disparate Impact
Analysis to Subjective Selection Criteria

Petitioner and her amici contend that the legislative his­
tory of the 1972 amendments to Title VII shows that Congress 
mandated that the Griggs rule should apply to all employment 
practices, including subjective ones.5 Specifically, they point to 
(1) the references in the legislative history to “systemic” or 
“institution” discrimination, (2) the concern with discrimina­
tory practices by governmental employers, particularly those 
of the Civil Service Commission, and (3) Congress’s asserted 
ratification of the result of Griggs v. Duke Power Co., 401 U.S. 
424 (1971), and of the pre-1972 case law on employment dis­
crimination. These contentions are demonstrably incorrect.

Before examining the legislative history in detail, a basic 
preliminary point must be made. In 1972, Congress simply did 
not amend any of the provisions relevant to this case. See the 
Equal Employment Opportunity Act of 1972, Pub. L. 92-261, 
86 Stat. 103. The specific provision at issue in this case, Sec­
tion 703(a)(2), was amended only to add the words “or appli­
cants for employment.” This change does not suggest an intent 
to codify Griggs', still less does it suggest an intent to extend 
the coverage of Griggs to subjective practices. See Gold, 
Reply to Thomson, 8 Indus. Rel. L. J. 117, n.4 (1986). If Con­
gress has re-enacted Title VII in 1972, such a decision might 
have signified an endorsement of existing judicial and admin­
istrative glosses on the statute. Cf. N LRB v. Gullett Gin Co.,
5See Brief for Petitioner 35-38; Brief for the NAACP Legal Defense and 
Educational Fund, Inc. et al., 13-20; Brief for the Lawyers’ Committee for 
Civil Rights Under Law 9-10.



19

340 U.S. 361, 365-366 (1951). But Congress did not reenact the 
statute. Thus, since the statutory language governing this 
case “was enacted in 1964, not 1972,” and since “[t]he views of 
members of a later Congress, concerning different sections of 
Title VII, are entitled to little if any weight,” it is the intent of 
the Congress that enacted Section 703(a) in 1964 that controls. 
Teamsters v. United States, 432 U.S. 324, 354 n. 39 (1977). 
Any debate about the intent of the 1972 Congress is therefore 
really academic.

Assuming, arguendo, that the 1972 legislative history has 
some relevance to this case, petitioner and her amici in any 
event misinterpret it. That history reveals no intent whatever 
to extend Griggs to subjective practices. If anything, it shows 
that Congress looked with favor on administrative enforce­
ment of the aims of Title VII through disparate treatment lit­
igation, i.e., pattern-or-practice suits.

The genesis of the 1972 amendments, as relevant here, was 
as follows. A bill designated H.R. 1746 was introduced by 
Cong. Augustus Hawkins and reported out unchanged by the 
House Committee on Labor and Education. The Hawkins bill 
was turned down in a floor vote in favor of the “Erlenborn sub­
s titu te” H.R. 9247, a measure introduced by Congressman 
John N. Erlenborn as an alternative to the Hawkins bill. See 
117 Cong. Rec. 32111.6

During the House floor debate on September 14-15, 1971, 
several supporters of the Hawkins bill referred  in their 
remarks to “systemic” or “institutional” discrimination. Con­
gressman Perkins, who chaired the House committee, spoke 
to that issue in his introductory remarks on September 14; 
however, he clearly linked the existence of such discrimina­
tion to the need to give broader enforcement and litigation 
authority to the Equal Employment Opportunity Commission
Petitioner seems to imply, mistakenly, that the Hawkins bill passed the 
House. See Brief for Petitioner 36.



20

(EEOC), which under the 1964 Act had only conciliatory and 
persuasive powers.7 Congressman Perkins did not suggest 
that Griggs should be extended to cover these employment 
practices. See 117 Cong. Rec. 31960. Nor did other supporters 
of H.R. 1746 who raised the issue of institutional discrimina­
tion; they advocated the same solution as Congressman Per­
kins’s, i.e., they sought to empower the EEOC to issue cease 
and desist orders, and to initiate litigation. See 117 Cong. Rec. 
32095 (remarks of Del. Fauntroy); id. at 32096 (remarks of 
Cong. Eckhardt); id. at 32102 (remarks of Cong. Badillo); id. 
at 32104 (remarks of Cong. Fraser); id. at 32106 (remarks of 
Cong. Stokes).8

7Prior to the 1972 amendments, the only civil actions authorized by the stat­
ute were private lawsuits or “pattern-or-practice” actions brought by the 
Attorney General. The EEOC’s role in eliminating employment discrimi­
nation was limited chiefly to conciliation and persuasion. As a result of the 
1972 amendments, authority to bring “pattern or practice” cases was trans­
ferred to the EEOC, except in cases where the defendant was a govern­
mental entity. See Section 706(f)(1), 42 U.S.C. 2000e-f(f)(l). As the Court 
has said, “Congress became convinced * * * that the ‘failure to grant the 
EEOC meaningful enforcement powers has proven to be a major flaw in the 
operation of Title VII.’ []S. Rep. No. 92-415 p. 4 (1971). The 1972 amend­
ments to Section 706 accordingly expanded the EEOC’s enforcement pow­
ers by authorizing the EEOC to bring a civil action in federal district court 
against private employers reasonably suspected of violating Title VII.” 
General Telephone v. EEOC, 446 U.S. 318, 325 (1980) (footnote omitted). 
See, also, e.q., 118 Cong. Rec. 3962 (remarks of Sen. Javits) (“the price of 
the Civil Rights Act of 1964 was a very weak job discrimination title”); id. 
at 4080 (remarks of Sen. Williams) (to like effect); id. at 4082 (remarks of 
Sen. Javits).

8One provision of the rejected Hawkins bill would have amended Section 
703(h) of Title VII insofar as that section provides that “it shall not be an 
unlawful employment practice * * * for an employer to give and to act upon 
the results of any professionally developed ability test provided that such 
test, its administration or action upon the results is not designed, intended 
or used to discriminate * * *.” Congressman Perkins explained that the pro­
posed amendment to Section 703(h) was designed to “alter[] the language of 
Title VII better to reflect the congressional intent as interpreted by the 
Court in the Griggs case.” 117 Cong. Rec. 31961. The proposed amendment 
was defeated together with the rest of the Hawkins bill; there were no com­
ments on it other than Congressman Perkins’s. No parallel provision 
appeared in S. 2515, the Senate bill sponsored by Senator Williams. While



21

The substantive provision of the Senate bill introduced by 
Senator Williams, S. 2515, were similar in relevant respects to 
those of the Hawkins bill;9 in particular, S. 2515 would have 
conferred cease and desist powers on the E E O C -a  measure 
both houses ultimately rejected.10 * In the Senate debate, as in 
the House debate, those who spoke of institutional discrimi­
nation also urged expanded EEOC authority, not the exten­
sion of Griggs, as an antidote to the problem. Senator Wil­
liam’s introductory remarks on his bill were similar in tenor to 
Congressman Perkins’s introductory remarks on the Hawkins 
bill in the House. See 118 Cong. Rec. 294 (remarks of Sen. Wil­
liams). Other Senators who spoke of institutional discrimina­
tion likewise envisaged EEOC enforcement authority as the 
cure. See id. a t 580 (rem arks of Sen. Jav its); id. a t 590 
(remarks of Sen. Humphrey); id. at 944-945 (remarks of Sen. 
Spong) (opposing Williams bill, but urging EEOC be given 
authority to litigate); id. at 3978 (remarks of Sen. Moss) (sup­
porting Williams-Javits position).11
it may be hazardous to draw any inferences from such Congressional inac­
tion, see Johnson v. Transp. Agency, Santa Clara Cty., 480 U .S .______ ,
94 L.Ed.2d 615 (1987) (Scalia, J., dissenting), certainly the result does 
nothing to support the view that the 1972 Congress codified or ratified 
Griggs.

9'But see n.8, supra.
'“Petitioner states, not entirely accurately, that the Williams bill was “ulti­

mately enacted.” See Brief for Petitioner 35. In this crucial respect, how­
ever, the original Williams bill survived only as amended. Both supporters 
and critics of the Williams bill regarded the conferral of cease and desist 
powers on the EEOC as one of its key provisions. See, e.g., 118 Cong. Rec. 
3965-3979; Thomson, The Disparate Impact Theory: Congressional Intent 
in 1972-A Response to Gold, 8 Indus. Rel. L. J. 105, 109 (1986) (“[t]he 
main struggle in both houses of Congress concerned whether the EEOC 
would have power to issue cease and desist orders”). Nevertheless, the 
Senate ultimately passed an amendment by Senator Dominick, S. 884, that 
denied such enforcement power to the EEOC, and required it to initiate 
litigation in a federal district court. See 118 Cong, Rec. 3979.

nSee 118 Cong. Rec. 3965 (colloquy between Sens. Dominick and Javits) 
(describing differences between S. 884, the Dominick amendment, and S. 
878, the Williams-Javits proposal to which Sen. Moss spoke).



22

The same perceived deficiency in Title VII, and the same 
proposed cure, were discussed at length in the Report of the 
Senate Committee on Labor and Public Welfare, S. Rep. 92- 
415, 92nd Cong., 1st Sess. 4-5 (1971). (The R eport was 
attached to the original, unamended Williams bill). In a sec­
tion entitled “Need For The Bill” the Report states (id. at 12) 
that:

[i]n 1964, em ploym ent discrim ination tended to be 
viewed as a series of isolated and distinguishable events, 
for the most part due to ill-will on the part of some iden­
tifiable individual or organization. * * * Employment dis­
crimination as viewed today is a far more complex and 
pervasive phenomenon. Experts familiar with the sub­
ject now generally describe the problem in terms o f‘sys­
tems’ and ‘effects’ rather than simply intentional wrongs, 
and the literature on the subject is replete with discus­
sions of, for example, the mechanics of seniority and lines 
of progression, perpetuation of the present effect of pre­
act discriminatory practices through various institutional 
devices, and testing and validation requirements.

See also H. Rep. No. 238, 92d Cong., 1st Sess. 21-22 (1972) 
(attached to the Hawkins bill). The statements immediately 
preceding this reference to “systems” and “effects,” however, 
make clear th a t the Senate Com m ittee’s solution was to 
change the role of the EEOC from an agency devoted to vol­
untary conciliation efforts, to an agency that, like the National 
Labor Relations Board or the Securities and Exchange Com­
mission,12 could issue cease and desist orders (id. at 11):

The most striking deficiency of the 1964 Act is that the 
EEOC does not have the authority to issue judicially 
enforceable orders to back up its findings of discrimina­
tion. In prohibiting discrimination in employment based 
on race, religion,, color, sex or national origin, the 1964 
Act limited the Commission’s enforcement authority to

>2See, e.q., 117 Cong. Rec. 31960 (remarks of Cong. Perkins) (drawing com­
parison to cease and desist powers of other agencies).



23

‘informal methods-of conference, conciliation and persua­
sion.’ * * * This failure to grant the EEOC meaningful 
enforcement powers has proven to be major flaw in the 
operation of Title VII. While the statutes dealing with 
discrimination in housing and in education provide appro­
priate enforcement powers for the agencies responsible 
for the elimination of discrimination in those areas of the 
law, Title VII, as it now stands, is little more than a dec­
laration of national policy. Regretably [sic], the practices 
and policies of discrimination are so deeply ingrained that 
the voluntary conciliation approach has not succeeded in 
adequately combating the existence of such practices.

On February 22, 1972, the Senate passed the amended S. 
2515. 118 Cong. Rec. 4944. The differences between the House 
and Senate bills were resolved in conference, and the final ver­
sion, as reported from the conference, was ratified by both 
houses. 118 Cong. Rec. 7573 (House); id. at 7170 (Senate).

Petitioner’s amici further argue that the expansion of Title 
VIPs coverage in 1972 to include governmental employers also 
evidences an intent to extend Griggs to subjective decision- 
making.13 But this Court pointed out in Johnson v. Transp.
Agency, Santa Clara Cty., 480 U.S--------, 94 L.Ed.2d at 628
“[wjhile public employers were not added to the definition of 
‘employer’ in Title VII until 1972, there is no evidence that 
this mere addition to the definitional section of the statute was 
intended to transform the substantive standard governing 
employer conduct. Indeed, ‘Congress expressly indicated the 
intent that the same Title VII principles be applied to govern­
mental and private employers alike’” (citing Dothard v. Raw- 
linson, 433 U.S. 321, 323 n. 14 (1977)). Consequently, the 
broader post-1972 coverage of Title VII does not suggest that 
Congress altered the pre-1972 substantive standards for 
determining the lawfulness of subjective practices.
nSee Brief for the NAACP Legal Defense and Educational Fund, Inc., et al 

15-17.



24

The legislative record bears out this conclusion. The refer­
ences in the floor debates to practices by governm ental 
employers th a t had racially disparate effects were in te r­
twined with the argument that extending Title V II’s protec­
tions to governmental employees would bring the necessary 
relief. See, e.g., 118 Cong. Rec. 296 (remarks of Sen. Wil­
liams); id. at 1815 (remarks of Sen. Williams) (introducing into 
the record a 1969 repo rt of the U.S. Commission on Civil 
Rights on which petitioner relies (Brief for petitioner 35-36)). 
Senator Williams did not argue that impact theory had to be 
applied across the entire range of governmental -  and hence 
private-employment procedures. He said instead that “[t]he 
inclusion of State and local government employees within the 
jurisdiction of Title VII will fulfill the congressional duty to 
enact the ’appropriate legislation’ to ensure that all citizens 
are treated equally in this country” (118 Cong. Rec. 1816). In 
other words, Senator Williams apparently believed that the 
extension of Title VII’s protections to government workers 
would in  itself he an adequate remedy for the abuses pointed 
out in the Report of the Commission on Civil Rights.14

Nor do the statements in the 1972 legislative reports, to the 
effect that then-developed case law would continue to govern 
Title VII, prove that Congress intended Griggs to apply to 
subjective practices. See H.R. Conf. Rep., 92nd Cong., 2d 
Sess., reprinted in 118 Cong. Rec. 7166 (1972).

First, this Court has repudiated excessive reliance on pre- 
1972 lower Court decisions as a basis for interpreting Title
14Other speakers emphasized that the problems in the federal civil service 

would be corrected by empowering the EEOC, rather than the Civil Ser­
vice Commission, to take antidiscriminatory measures. See 117 Cong. Rec. 
32094 (remarks of Del. Fauntroy); id. at 31961 (remarks of Cong. Perkins). 
Congressman Perkins added that it was also expected that the Civil Ser­
vice Commission would continue its own “affirmative measures” against 
federal employment discrimination. Id. These remarks (which in any case 
were addressed to the unsuccessful Hawkins bill) also fall far short of the 
claim that Griggs was to be extended to subjective practices in both gov­
ernmental and private employment.



25

VII. Thus, in Teamsters v. United States, 431 U.S. 324, the 
Court considered and rejected the argument that the 1972 
amendments had codified all prior Title VII cases that had 
struck down good faith seniority systems. And, in Firefighters 
Local Union No. 1784. v. Stotts, 467 U.S. 561, 582 n. 15 (1984), 
the Court again rejected the “contention th a t Congress 
in te n d e d  [in 1972] to  codify  all e x is t in g  T itle  V II 
decisions * *

Furthermore, none of the pre-1972 cases, including Griggs, 
had settled  the issue w hether subjective practices with 
adverse impact on protected groups had to have a “manifest 
relation” to the job in order to avoid breaching Title VII. “[I]t 
may be said with confidence that there is nothing either in the 
facts of [Griggs] * * * or in the language of the opinionf] that 
could serve as authority for extending [its] requirements 
beyond objective tests and other objective measures such as 
educational requirements . . .” 3A A. Larson & L. Larson, 
Employment Discrimination para. 76.30 at 15-81 (1987).

Finally, there was pre-1972 case law (in the form of suits 
against public employers under the Fourteenth Amendment 
or 42 U.S.C. 1983) holding that some subjective decisionmak­
ing was shielded from such testing.15 In Smith v. Bd. ofEduc. 
of Morrilton Sch. Dist., 365 F.2d 770, 781-782 (8th Cir. 1966), 
then Judge Blackmun wrote that “teaching is an art and that 
excellence does not depend upon knowledge, experience, for­
mal training and classroom conduct alone. Fitness for teach­
ing rests  upon a broad range of factors and encompasses 
numerous personality and character traits. * * * Nothing con­
tained in this opinion is intended to be restrictive of a school 
board’s freedom to make a full inquiry and to give due consid­
15Although these cases were not decided under Title VII, their relevance to 

the 1972 amendments is clear. See Allen v. City of Mobile, 466 F.2d 122, 
125-126 (5th Cir. 1972) (Goldberg, J . , dissenting) (noting that prior to 1972 
amendments, Section 1983, as applied to governmental employees, had 
“strong analogy” to Title VII).



26

eration to an applicant’s qualifications so long as the board 
does not act unreasonably, arbitrarily, capriciously or unlaw­
fully?’ Id. at 632-633. See also Lewis v. Chicago State College, 
299 F. Supp. 1357 (N.D. 111. 1969) (to like effect in Section 1983 
case). If Congress codified (or even approved) existing case 
law, it presumably also codified (or approved) Smith, Lewis 
and similar employment discrimination decisions, and so has 
arguably ratified a “reasonableness” test for at least some 
subjective decisionmaking.

Similar arguments undercut petitioner’s reliance on the 
C ourt’s statem ent in Connecticut v. Teal, 457 U.S. 440, 
447 n. 8 (1982), that the legislative history of the 1972 amend­
ments “demonstrates that Congress recognized and endorsed 
the disparate-im pact analysis employed by the Court in 
Griggs;”16 The Teal footnote does not suggest that Congress 
actually codified Griggs (as opposed to just endorsing it). Fur­
thermore, Congress’s approval of Griggs would not assist in 
the resolution of this case, which unlike Griggs concerns sub­
jective, not objective, processes. And in any event, as amici 
concede,17 the Court has placed little or no weight on the 1972 
legislative history where the relevant statutory language-in 
this case, that of Section 703(a)(2) -  was unaffected by the
K'See Brief for Petitioner 35; see also Brief for the NAACP Legal Defense 

and Educational Fund, Inc. et al. 14.
We note, however, that some commentators consider Congress’s 

response to Griggs to have been more ambiguous. See Gold, Reply to 
Thomson, 8 Indus. Rel. L. J. 117-119; Meltzer, The Weber Case: The Judi­
cial Abrogation of the Antidiscrimination Standard in Employment, 47 
U. Chi. L. Rev. 423, 436-437 (1980). Thus, e.q., a Senate proposal to strike 
the word “intentionally” from the remedial provision of Title VII, Section 
706(g), was deleted at the insistence of the House, so that the 1964 wording 
was retained. See 118 Cong. Rec. 6647 (Joint Explanatory Statement of 
Managers at the House-Senate Conference); Sape and Hart, Title VII 
Reconsidered: The Equal Employment Opportunity Act of 1972, 40 Geo. 
Wash. L. Rev. 824, 883-884 (1972). See also n.8, supra.

llSee Brief of the NAACP Legal Defense and Educational Fund, Inc., et al., 
14 n. 12.



27

amendments. Cf. Teamsters v. United States, 481 U.S. at 354 
n. 39.

In sum, petitioner and her amici err in reliance on the 1972 
legislative record. Congress simply did not deal with the issue 
of subjective employment practices in any of its actual enact­
ments. Indeed, the fact that the 1972 amendments confer 
authority on the EEOC to bring pattern-or-practice suits sug­
gests that, contrary to petitioner’s thesis, Congress envisaged 
that disparate treatment litigation of that form would be a 
large part of the answer to the problem of “institutional” dis­
crimination.18 If the 1972 legislative history is relevant at all, 
then, it tends to demonstrate exactly the opposite of petition­
er’s view.

III. Administrative Interpretation Has Not Been Consistent 
Regarding Application of Subjective Employee Selection 
Procedures to Disparate Impact Analyses

Petitioner urges (Pet. Br. 20-21, 24, 29-34) that the Court 
defer to what she claims has been the consistent view of the 
government -  advocated since Title VII’s enactment in 1964 -  
that the disparate impact analysis and its corollary manifest 
relation/business necessity requirement apply to subjective 
selection processes. But petitioner overstates the consistency 
with which the government has adhered to this view and, per­

I8Senator Javits stressed the importance of the statutory change transfer­
ring to the EEOC the power to initiate pattern-or-practice suits against 
private employers. “By our decision yesterday we gave the EEOC the 
power to bring suit in big as well as small cases. * * * I have said on a num­
ber of occasions in this debate that [the Civil Rights Act of 1964] was a 
compromise and that much was given up, especially in respect of seeking a 
remedy against discrimination in employment. * * * That is why we felt 
we had to give the Attorney General the power to sue in big cases, in class 
action cases, and in cases where there was a constant pattern of discrimi­
nation directed at individuals with limited resources whom we were rele­
gating to the courts and who could hardly be expected to carry such a 
broad and deep case. We are now changing that and giving it back to the 
Commission.” 118 Cong. Rec. 4081-4082. See also id. at 4081 (remarks of 
Sen. Williams).



28

haps more importantly, mischaracterizes the views of the gov­
ernment, including the EEOC, today.

Contrary to petitioner’s suggestion, the EEOC has not con­
sistently interpreted Title VII as prohibiting use of subjective 
selection processes that have an adverse impact and that have 
not been proven manifestly necessary to the job or other busi­
ness need. In fact, it appears that the EEOC initially took the 
position that “discrimination” based on educational qualifica­
tions (of the kind later at issue in Griggs) did not violate Title 
VII. See Opinion Letter, dated October 2, 1965, GC 296-65, 
cited at Pet. App. 147b, in Griggs v. Duke Power Co., 401 U. S. 
424 (1971). Thereafter, the EEOC took notice of the fact that 
objective tests might to used to exclude qualified minority 
applicants. See Guidelines on Em ploym ent Testing Proce­
dures, August 24, 1966. Accordingly, it issued guidelines 
advocating that employers refrain from using such tests wher­
ever they would have the effect of perpetuating past discrim­
ination and were not job-related. See id.; see also Decision of 
the EEOC, CCH Empl. Prac. Guide 1117304.53 (Dec. 2, 1966). 
But the guidelines were directed exclusively to the meaning of 
the phrase “professionally developed ability test[s]” that 
appears in Section 703(h) of the statute (42 U.S.C. 2000e- 
7(h)).19 They were not addressed to other kinds of selection 
devices or employment practices, objective or subjective.

To be sure, the 1966 Guidelines did advocate the use of non- 
discriminatory “total personnel systems.” But they did not, as 
petitioner suggests (Pet. Br. 20), state or imply that such per­
sonnel systems in their entirety-or, for that matter, any sin­
gle component of them -had to be formally validated. Rather, 
the Guidelines advocated that test results not alone deter­
mine an employer’s personnel choices, at least not where other 
factors suggested that minority group applicants would be
I9The origin of the 1966 Guidelines is discussed (by an EEOC official who 

was instrumental in creating them) in Blumrosen, Strangers In Paradise: 
Griggs v. Duke Power Co. And The Concept of Employment Discrim ina­
tion, 71 Mich. L. Rev. 59, 60-61 nn.5-7 (1972).'



29

successful employees. Indeed, where the 1966 Guidelines did 
mention a practice tha t is ordinarily considered subjective 
(i.e., interviews), they did not propose that that practice be 
formally validated. Instead, they stressed the desirability of 
finding interviewers whose racial attitudes were likely to be 
free from subjective biases.

It was only in 1970 that the EEOC extended its guidelines 
specifically to cover subjective selection processes and to 
require that those processes, like objective selection prac­
tices, be validated or otherwise formally substantiated. See 
Guidelines on Employee Selection Procedures, 35 Fed. Reg. 
12333 (Aug. 1,1970). Thus, in Section 1607.3 of the 1970 Guide­
lines, the EEOC defined “discrimination” to include “any test 
which adversely affects hiring, promotion, transfer or any 
employment or membership opportunity of classes protected 
by Title VII * * * unless * * * the test has been validated and 
evidences a high degree of utility * * *” (id. at 12334). The 
concept of a “te s t” was then explained in Section 1607.2 to 
encompass “any paper-and-pencil or performance measure 
used as a basis for any employment decision,” including 
“scored interviews, biographical information blanks, in ter­
viewers’ rating scales, [and] scored application forms” (ibid.), 
and Sections 1607.4 though 1607.8 described a single standard 
for demonstrating the validity of all of these selection devices 
(see id. at 12334-12336). Finally, Section 1607.13 noted that 
“[selection techniques other than tests, as defined in §1607.2, 
may be improperly used so as to have the effect of discrimi­
nating against minority groups” and that these “techniques 
include, but are not restricted to, unscored or casual in ter­
views and unscored application forms” (id. at 12336). In such 
circumstances, Section 1607.13 provided (ibid.) th a t the 
employer

* * * may be called upon to present evidence concerning 
the validity of his unscored procedures as well as of any 
tests which may be used, the evidence of validity being of



30

the same types referred to in §§ 1607.4 and 1607.5. Data 
suggesting the possibility of discrimination exist, for 
example, when there are differential rates of applicant 
rejection from various minority and nonminority or sex 
groups for the same job or group of jobs or when there 
are disproportionate representation of minority and non­
minority or sex groups among present employees in dif­
ferent types of jobs. If the person is unable or unwilling 
to perform such validation studies, he has the option of 
adjusting employment procedures so as to eliminate the 
conditions suggestive of employment discrimination.

The EEOC’s Guidelines were in effect from 1970 to 1978. 
During that period, however, the Department of Labor, the 
Department of Justice, and the Civil Service Commission 
were operating under anti-discrimination and employment 
testing regulations that differed in many critical respects.20 
Moreover, courts and commentators became increasingly crit­
ical of the EEOC’s Guidelines, charging that they imposed 
professionally unrealistic  and effectively unattainable 
requirements on employers and that they vested government 
prosecutors with excessive and uncontrolled authority to 
restrict private and public employment practices.21 Accord­
2aSee Douglas v. Hampton, 512 F.2d 976, 961-962 & n.37 (D.C. Cir. 1975) 

(noting apparent inconsistencies); see also Note, Application of the EEOC 
Guidelines to Employment Test Validation: A Uniform. Standard for Both 
Public and Private Employers, 41 Geo. Wash. L. Rev. 505, 527 & nn. 138- 
139 (1973). The various problems that arose as a result of these inconsis­
tent regulatory frameworks are discussed in Comptroller General of the 
United States, Report to the Congress: Problems With Federal Equal 
Employment Opportunity Guidelines On Employee Selection Procedures 
Need to Be Resolved 12-17 (1978); Comment, Courts, Psychologists, And 
The EEOC’s Uniform. Guidelines: An Analysis Of Recent Trends Affect­
ing Testing As A  Means Of Employee Selection, 36 Emory L. J. 203, 219 
(1987).

21See, e.g., Albermarle Paper Co. v. Moody, 422 U.S. 405, 449 (1975) (Black- 
mun, J ., dissenting); United States v. Georgia Power Co., 3 FEP Cas. 767, 
787 n.8 (N.D. Ga. 1971), affd inpart and vacated in part,, 474 F.2d 906 (5th 
Cir. 1973); Booth & MacKay Legal Constraints on Employment Testing 
and Evolving Trends in the Law, 29 Emory L. J. 121, 125 (1980); see also 
Gold, “Griggs’” Folly: An Essay on the Theory, Problems, and Origin of



31

ingly, the EEOC joined with these other agencies in rethink­
ing the principles that underlay the extant standards govern­
ing employee selection procedures.22

The upshot of these inter-agency deliberations was the pro­
mulgation of the Uniform Guidelines On Employee Selection 
Procedures, 29 C.F.R. 1607, to which all four agencies, includ­
ing the EEOC, are signatories. These Guidelines provide that 
the “use of any selection procedure which has an adverse 
impact on * * * employment * * * opportunities of members of 
any race, sex, or ethnic group will be considered to be discrim­
inatory * * * unless the procedure has been validated * * * or 
the provisions of Section 6 [of the Guidelines] * * * are satis­
fied” (29 C.F.R. 1607.3A (emphasis added). Moreover, they 
state that a “selection procedure” includes [a]ny measure, 
combination of measures, or procedure used as a basis for any 
employment decisions] * * * including] the full range of 
assessment techniques from traditional paper and pencil tests 
* * * through informal or casual interviews and unscored 
application forms” (29 C.F.R. 1607.16Q). Section 6 of the Uni­

the Adverse Impact Definition of Employment Discrimination and a Rec­
ommendation for Reform., 7 Indus. Rel. L. J. 429, 459 (1985); Lerner, 
Washington v. Davis : Quantity, Quality and Equality in Employment 

Testing, 1976 Sup. Ct. Rev. 263, 304, 306. Oddly enough, as these commen­
tators note, the American Psychological Association (APA) was one of the 
foremost critics of the EEOC’s Guidelines. See, e.g., Brief Amicus Curiae 
of the APA in Washington v. Davis, No. 74-1492, at 18-24, 28.

-"In adopting the Equal Employment Opportunity Act of 1972, Congress 
created the EEO Coordinating Council (EEOCC) in order to eliminate 
inter-agency conflicts and regulatory inconsistencies. See Sec. 715 of Title 
VII, as amended in 1972, 42 U.S.C. 2000e-14. In November 1972, the 
EEOCC decided to develop a set of uniform guidelines reconciling inter­
agency differences. The Department of Justice, the Department of Labor, 
and the Civil Service Commission joined in 1976 in adopting the resulting 
Federal Executive Agency (FEA) guidelines, which those agencies found 
better reflected professionally accepted standards than did the 1970 EEOC 
Guidelines. See 41 Fed. Reg. 51734 (1976); Comment, supra, 36 Emory 
L. J. at 220; Note, The Uniform Guidelines on Employee Selection Pro­
cedures: Compromises and Controversies, 28 Cath. IL L . Rev. 605, 608- 
610 (1979). The EEOC joined the other enforcement agencies in this effort 
in 1978.



32

form Guidelines, however, states that “[t]here are circum­
stances in which a user cannot or need not utilize the valida­
tion techniques contemplated by these guidelines” (29 C.F.R. 
1607.6B (emphasis added)), that such “circumstance” include 
the use of “informal or unscored selection procedurefs]” (29 
C .R .R . 1607.6B(1)), and th a t, w here such informal or 
unscored selection procedures are involved, the user need 
only “justify continued use of the procedurefs] in accord with 
Federal law” (ibid.). Thus, unlike Section 1607.13 of the 1970 
Guidelines, which imposed a formal substantiation require­
ment on employers using subjective selection processes asso­
ciated with numerical disparaties in the workforce, the 1978 
Guidelines only require the employer to “justify continued use 
of the procedurefs] in accord with Federal law.”

The Uniform Guidelines do not clearly define how an 
employer “justifies]; continued use of the procedure in accord 
with Federal law.” In subsequent questions and answers, the 
agencies stated that “ft]he method of justification of a proce­
dure by means other than validity is one to which the Guide­
lines are not addressed. See Section 6B.” 44 Fed. Reg. 12002
(1979). But, in the prefatory comments to the Guidelines 
themselves, the agencies did state that “fa] few practices may 
be used w ithout validation even if they have an adverse 
impact. See e.g., McDonnell Douglas v. Green, 411 U.S. 792 
(1973) and section 6B.” 43 Fed. Reg. 38291 n. 17 (1978). In 
short, the Uniform Guidelines not only provide no support for 
petitioner’s argument that employers must validate or other­
wise formally substantiate their subjective selection pro­
cesses, the agencies that promulgated them appear to believe 
that an employer satisfies the “in accord with Federal law” 
standard by articulating a legitimate, non-discriminatory rea­
son for its procedures (as in McDonnell Douglas v. Green)?

“ Petitioner errs in suggesting (Pet. Br. 33) that administrative decisions of 
the EEOC establish that the Uniform Guidelines require that disparate 
impact analysis be applied to subjective selection processes. The decisions



33

This interpretation of the Uniform Guidelines is confirmed, 
of course, by the brief amicus curiae of the United States and 
the EEOC in this case. In that brief, the government and the 
EEOC argue that, where a subjective selection process is in 
issue, an employer justifies “continued use of the procedure in 
accord with Federal law” by setting forth legitimate, nondis- 
criminatory reasons for maintaining the procedure and for 
applying it in particular cases, as is required in a disparate 
treatment case. The current position of the government and 
the EEOC represents a “sensible construction” of both the 
Guidelines and the s ta tu te , a construction which is “not 
negated by legislative history” (Pattern Makers League v. 
NLRB, 473 U.S. 95, 116-117 (1985) (White, J., concurring)). 
Thus, contrary to petitioner’s suggestion, if the Court is to 
defer in this case to the Uniform Guidelines, it should find that 
disparate impact theory does not apply to subjective selection 
processes.

IV. The Disparate Impact Theory Should Not Be Applied to 
Claims Involving Subjective Employment Practices or Cri­
teria Because the Disparate Impact Theory is Inappro­
priate for Analysis of Such Claims

In the presentation of her case to the District Court, peti­
tioner alleged that she was the victim of intentional discrimi­
nation, as well as discrimination resulting from the impact of 
respondent’s subjective employment practices. However, it is 
clear from the record that the proof offered by petitioner was 
intended to present her case under the disparate treatment 
theory. Quite probably petitioner recognized that even if the 
disparate impact theory were available to her, the facts of her

that petitioner cites arose under the 1970 Guidelines. Administrative deci­
sions of the EEOC under the 1978 Guidelines in fact indicate that “a sub­
jective selection process cannot be facially neutral and, thus, cannot be 
analyzed under the disparate impact doctrine” (Ziman v. USPS, No 
01842595 (July 23, 1986), at 14).



34

case would prevent its employment.24 Because petitioner was 
unable to satisfy her burden of proof on her disparate treat­
ment case, and because she did not identify any aspect of 
respondent’s employment process which was suitable for test­
ing under the Griggs model, the District Court found that 
petitioner had failed to prove her case 25 The District Court’s 
judgment, as it relates to the issue before this Court, was 
affirmed by the Court of Appeals for the Fifth Circuit.

In the decisions following Griggs, this Court has applied dis­
parate impact analysis only to criteria which, like the high 
school diploma and objective te s t score requirem ents in 
Griggs, are specific, defined, identifiable, and objective26

P etitioner’s assertion th a t her claim should have been 
tested under the impact theory is erroneous because she chal­
lenges hiring criteria which she acknowledges are purely sub­
jective. When subjective hiring practices or criteria are chal­
lenged, it is clear that the disparate treatment model is the 
only appropriate method for analyzing Watson’s claim.27

"4As has been pointed out by one of the am ici briefs supporting petitioner’s 
position, the evidence presented by petitioner demonstrated that she 
viewed her case as appropriate for disparate treatment analysis, and fur­
ther that the number of promotions sought by blacks were not of sufficient 
number to establish any pattern. Brief for the NAACP Legal and Educa­
tion Fund, Inc.; et al, at 2 n.l.

25The District Court concluded that petitioner had failed to show that the 
reasons articulated by respondent for promoting others, rather than peti­
tioner, were pretext for racial discrimination. (J. A. 197).

MSee Connecticut v. Teal, 457 U.S. 440, 446 (1982) (standardized test); New 
York City Transit Auth. v. Beazer, 440 U.S. 568, 589-91 (1979) (policy 
against employing persons who use narcotic drugs); Dothard v. Rawlin- 
son, 433 U.S. 321, 329 (1977) (height and weight requirements); Albemarle 
Paper Co. v. Moody, 422 U.S. 405, 412-13 (1975) (standardized tests). Cf. 
International Bhd. of Teamsters v. United States, 431 U.S. 324 (1977) 
(applying the disparate treatment model to a Title VII action challenging 
subjective hiring and promotion practices); Hazelwood School Dist. v. 
United States, 433 U.S. 299 (1977) (applying the disparate treatment 
model to a Title VII action challenging standardless, largely subjective 
hiring procedures).

“'Title VII does not preclude use of discretion or subjectivity in decisions by 
employers. If the subjective criteria are applied without consideration of



Objective criteria are neutral in design and application. By 
definition, objective criteria are applied mechanically and 
without exception to all applicants. It is for this reason that 
disparate impact analysis is appropriate where objective cri­
teria are involved. The disparate impact model looks at a 
facially neutral requirement which has a disparate effect upon 
members of a protected class, despite fairness and uniformity 
in treatment and application. Once such a requirement is iden­
tified by plaintiff, the employer must demonstrate that there 
is a valid business necessity for utilization of the requirement 
Griggs, 401 U.S. at 432.

Conversely, subjective criteria are determined by factors 
which are personal and individualized to the one making the 
selection decision. They are determined, whether consciously 
or not, by that person’s own perspective, beliefs, background, 
circumstances, judgments, and ultimately intent. Subjective 
criteria lack the uniformity associated with objective criteria. 
When selection decisions are made by many different people, 
as was the bank’s practice, the lack of uniformity is magnified 
even greater. When objective criteria are utilized there is a 
hazard that a facially neutral policy or requirement may result 
in discrimination, regardless of lack of discriminatory intent. 
However, with use of subjective criteria unintentional dis­
crimination is not the concern. To the contrary, the principal 
worry is that subjective criteria will be influenced by the 
biases, especially racial biases, of the one making the selecting 
decisions. Objective criteria are applied in a mechanical fash­
ion, but subjective criteria are employed on an ad hoc basis.28 
The disparate treatment model was created so that discrimi­

race or other irrelevant factors, there is no violation of Title VII. See Hill 
v. Seaboard Coast Line RRCo., 767 F.2d 771, 775 (11th Cir. 1985); Wang v. 
Hoffman, 694 F.2d 1146, 1148 (9th Cir. 1982); Bauer v. Bailor, 647 F.2d 
1037, 1046 (10th Cir. 1981).

28See Zahorik v. Cornell University, 729 F.2d 85, 95 (2d Cir. 1984); 3 A. Lar­
son & L. Larson, Employment Discrimination §76.34 at 15-87 to 15-88.



36

nation rising from improper use of subjective criteria could be 
uncovered.

Although the disparate treatm ent and disparate impact the­
ories are each designed to unearth discrimination based upon 
consideration of impermissible factors, the two are not appli­
cable interchangeably. Each theory has its own unique func­
tion in bringing about the goals Congress envisioned when 
Title VII was enacted. As this Court observed in Teamsters, 
“[ejither theory may, of course, be applied to a particular set 
of facts.”"9 However, that is not to say that both theories are 
applicable to any and all sets of facts.29 30 Even when applied to 
the same set of facts, the two theories serve completely differ­
ent roles and functions in accomplishing Title VIPs goals.31

If Title VII plaintiffs are permitted to apply the disparate 
impact model to all sets of facts, obviating the need to provide 
discriminatory intent, there would be no role left for the dis­
parate treatment model in any case involving statistical proof 
of discrimination. By merely demonstrating an imbalance, a 
plaintiff could shift the heavy burden to the employer to prove 
that any employment practice, whether objective or subjec­
tive, is justified by business necessity.

Elimination of the role of the disparate treatm ent model 
would be contrary  to both Congressional in ten t and this 
Court’s previous interpretation of Title VII. The Court has
29International Board of Teamsters v. United States, 431 U.S. 324 336 

n.15 (1977). ’
Where it is alleged that an objective selection criterion results in a statis­
tically significant racial disparity, a plaintiff could proceed under either or 
both theories. But one should not infer from the use of both theories to a 
particular set of facts that either theory can be applied to all sets of facts.

,!1This Court continues to stress that disparate treatment and disparate 
impact are conceptually and fundamentally different theories of discrimi­
nation. See, e.g., Fumco 438 U.S. at 581-82 (Marshall, J., concurring in 
part and dissenting in part) (“It is well established under Title VII that 
claims of employment discrimination because of race may arise in two dif­
ferent ways”) See also United States Postal Serv. Bd. of Governors v Ath­
ens, 460 U.S. 711, 713 n.l (1983); Burdins, 450 U.S. at 252 n.5.



37

recognized that “[undoubtedly disparate treatm ent was the 
most obvious evil Congress had in mind when it enacted Title 
VII.” Teamsters, 431 U.S. at 335 n.15 (citing legislative history 
of Title VII). Moreover, the Court’s approval of the use of sta­
tistics as evidence of intentional discrimination in Teamsters, 
431 U.S. at 339, and Hazelwood School Dist. v. United States, 
433 U.S. 299, 307-08 (1977), undeniably demonstrates that a 
plaintiff cannot be relieved of the burden of showing discrim­
inatory intent simply because the case involves statistical 
proof. It also demonstrates that the disparate impact model 
should not be permitted to inundate and replace the disparate 
treatment model in discrimination actions.

Moreover, if disparate impact analysis applies to subjective 
criteria, then the Griggs requirement that the employment 
practice at issue be proved to be job related will also apply to 
subjective criteria. Validating subjective criteria  as job 
related, however, is impossible. It is impossible to devise 
objective measures of subjective qualifications. It would be 
futile to attem pt to score criteria which involve subjective 
judgment. One cannot convert the subjective elements of a 
selection process to objective by merely ascribing numbers to 
them. For the same reason, it is impossible to define and 
devise reliable measures of job performance or employee ade­
quacy in terms of subjective criteria, as is necessary in vali­
dation studies.32

More importantly, validating subjective criteria such as 
those used by the bank is unnecessary. The purpose of vali-

’“Even courts which mistakenly apply disparate impact analysis to subjec­
tive systems implicitly assume that those systems inherently are incapable 
of being validated. They analyze the lawfulness of those systems through 
consideration of factors other than technical validation studies, such as the 
court’s perception of the overall fairness to employees of the subjective 
procedures. See, e.g., Rowe v. Cleveland Pneumatic Co., 690 F.2d 88, 90, 
95 (6th Cir. 1982); Robinson v. Union Carbide Corp., 538 F.2d 652, 662 (5th 
Cir. 1976), cert, denied, 434 U.S. 822 (1977); Rowe v. General Motors 
Corp., 457 F.2d 348 (5th Cir. 1972). Indeed, often the factors considered 
have nothing to do with job-relatedness. See, e.g., Id ., 457 F.2d at 358-59.



38

dation is to determine the job-relatedness of the criterion 
itse lf w ithout regard to the m anner through which it is 
applied. Common sense dictates that attitude, reliability, hon­
esty and loyalty are related to any job. The real issue’ with 
subjective criteria is whether they are fairly applied in a color­
blind manner, not whether they are job related.

No employer could defend itse lf successfully against a 
charge that its subjective selection procedure was discrimi- 
natorily applied by relying upon the job-relatedness of the cri­
teria. Thus the petitioner’s assertion that the bank should be 
lequired to validate its promotion process is misplaced and 
unpersuasive.

Because the employer cannot defend itself against a dispar­
ate impact challenge to its subjective selection system on the 
grounds of business necessity, disparate impact actions chal­
lenging subjective systems are in reality merely statistical 
battles. The employer’s only defense is to refute the plaintiffs’ 
statistical evidence.33 The burden which the employer must 
bear under disparate impact analysis to justify subjective sys­
tems which result in a significant racial disparity, therefore, is 
insurmountable.

The employer is then left with two choices. F irs t, the 
employer can install a totally objective system. It would, for 
example, refrain entirely from interviewing applicants or hav­
ing any other face-to-face contact with applicants (since that 
inevitably results in the consideration of subjective factors) 
even though it is well settled that subjective interviews are

See, e.g., Rowe v. Cleveland Pneumatic Co., 690 F.2d at 95 (“If the 
employer m the present case had been confronted with a prima facie case 
ot disparate impact as a result of its [subjective] hiring practice, the unre­
stricted discretion and unexplained subjective decisions would have failed 
to rebut that prima facie case.”) C oe v. Yellow Freight Sys., Inc. 646 F 2d 
444, 453 n.2 (10th Cir. 1981) (“[W]here the policy objected to is vague and 
general, once plaintiff shows discrimination by the use of statistics the 
employer may only rebut the claim of discrimination by the use of 
statistics”).



39

not per se illegal and that employers are permitted to use sub­
jective selection systems.34

The other more likely consequence of exposing subjective 
interviews to disparate impact analysis is employer reliance 
upon surreptitious, inflexible quota systems outside the per­
missible parameters of voluntary affirmative action plans.35 
The practical effect of applying disparate impact analysis to 
subjective criteria is simple, pure quota-based employment 
decisions because reliance upon a quota is the only means by 
which an employer can both use subjective criteria and pro­
tect itself from Title VII liability36 Race, then, rather than 
being irrelevant, becomes the most relevant factor in hiring 
decisions.

Quota systems, however, are blatantly contrary to the basic 
intent and purpose of Title VII.37 From Griggs onward, the
uSee, e.g., Heagney v. University of Washington, 642 F.2d 1157 at 1163 (9th 

Cir. 1981); EEOC v. Greyhound Lines, Inc., 635 F.2d 188 at 193 (3d Cir. 
1980). See also McDonnell Douglas Corp., 411 U.S. at 803 (rejecting a 
lower court holding that the defendant’s proferred reasons for refusing to 
hire the plaintiff were entitled to little weight because they were 
subjective).

!>Under Supreme Court precedent, while certain forms of race-conscious 
relief to remedy past discrimination are lawful, the type of informal, sur­
reptitious, inflexible quota system that would inevitably result from apply­
ing disparate impact analysis to subjective criteria is’unlawful. The test 
established in United Steelworkers of Am., v. Weber, 443 U.S. 193 (1979), 
for voluntary affirmative action programs is whether the program (1) con­
tains safeguards necessary to avoid trammelling the rights of non-minori­
ties, (2) is designed to remedy past discrimination and (3) is temporary. An 
informal quota system that is covert, unstructured, open-ended and/or has 
only the general goal of avoiding liability under Title VII cannot meet the 
criteria set forth in Weber. Lilly v. City ofBeckley, 797 F.2d 191, 195 (4th 
Cir. 1986) (holding that a white police officer who was not hired for a job on 
the Beckley, W. Va. police force because of the city’s informal affirmative 
action program was a victim of racial discrimination under Title VII). 

'^Hiring by quota to protect oneself from liability under Title VII, however, 
is prohibited. The Supreme Court, in Weber, 443 U.S. 193, refused to adopt 
the “arguable violation” theory as its justification for upholding an affir­
mative action plan. The “arguable violation” theory contends that volun­
tary affirmative action can stem merely from an employer’s wish to insu­
late itself from Title VII liability.



40

Supreme Court has consistently held that Title VII was not 
designed to force an equality of group representation in the 
workplace.38 As the Court first articulated in Griggs, “the Act 
does not command th a t any person be hired simply . . . 
because he is a member of a minority group. Discriminatory 
preference fo r any group, minority or majority, is precisely 
and only what Congress has proscribed.”39 I t would be ironic if 
Title VII and the Griggs decision, which were intended to 
require employers to hire applicants without regard to race 
according to their qualifications, should have the ultimate 
practical effect of discouraging color blind merit hiring.

The purported rationale of Watson’s position before this 
Court also is prohibited by the Supreme Court’s holding in 
Connecticut v. Teal, 457 U.S. 440. In Teal, the Court rejected 
the “bottom line” defense.40 The Court explained that it “has
,!'In response to an objection posed by opponents of the bill that Title VII 

would require employers to establish quotas for nonwhites in proportion to 
the percentage of nonwhites in the labor market area, Senator Clark 
remarked that “[qjuotas themselves are discriminatory.” 110 Cong. Ree. 
7218 (1964). Both proponents and opponents of the Act agreed that Title 
VII did not mandate a rigid quota system in which employment opportu­
nities would be allocated without regard to merit and soleiy in proportion 
to the representation of protected groups in the general population or in 
any particular pool. Thus, Section 703(j) of Title VII, 42 U.S.C. §2000e- 
2(j), provides in pertinent part:

Nothing 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 num­
ber 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 community, State, section, or other area, or in the 
available work force in any community, State, section, or other area. 

:*See, e.g., Burdine, 450 U.S. at 259, Fumco, 438 U.S. at 579; City of Los 
Angeles Dept, of Water and Power v. Manhart, 435 U.S. 702 at 709 (i978); 
McDonnell Douglas, 411 U.S. at 800-04; Griggs, 401 U.S. at 430-31.

*’401 U.S. at 430-31 (emphasis added). See also Teamsters, 431 U.S. at 340, 
n.20 (Title VII does not require an employer’s work force to be racially 
balanced).

“The plaintiffs in Teal challenged Connecticut’s selection process for welfare 
supervisors. The state’s promotion scheme required, first, passing a writ-



41

never read §703(a)(2) as requiring the focus to be placed . . . 
on the overall num ber of m inority or female applicants 
actually hired or promoted,” id. at 450, and ruled that Title 
VII requires that the focus be upon discrimination against 
individuals, not groups. Id. at 451, 453-54. Just as the “bottom 
line” cannot provide a defense to a charge of discrimination, it 
cannot provide the basis for finding liability.

Watson and her am ici argue th a t the d isparate impact 
model must be made available to analyze subjective selection 
criteria or such employment practices will be insulated from 
challenge under Title VII. The only basis for her assertion is 
the unsubstantiated assumption that the disparate treatment 
model is not suited to prove unlawful discrimination.* 41 Watson 
would have this Court turn all Title VII cases which contain 
evidence of a statistical imbalance into a disparate impact 
case.

As an illustration of the problem she perceives, Watson 
cites the Fifth Circuit’s decision in Payne v. Travenol Labo­
ratories, Inc., 673 F.2d 798 (5th Cir. 1982).42 Watson implies 
that because of the Fifth Circuit’s decision in Potency v. Pru­
dential Insurance Company of America, 668 F.2d 795 (5th 
Cir. 1982) the plaintiff was denied an opportunity to challenge

ten test and, second, consideration of work experience, recommendations 
and seniority. Evidence showed that disproportionately more blacks than 
whites failed the test but that disproportionately more blacks than whites 
were promoted. The plaintiffs argued that the scheme violated Title VII 
under the disparate impact theory because the state had not validated the 
written test as job-related. The state argued that the result of the entire 
process, reflecting no adverse impact on minority group members, pre­
cluded the finding of a Title VII violation.

41Merely because petitioner’s case was largely based upon statistical proof 
does not mean that the disparate treatment model was unavailable to her. 
In light of Teamsters, 431 U.S. 324 (1977), and Hazelwood, 433 U.S. 299 
(1977), courts have held that statistical proof is relevant and important in 
an individual claim of disparate treatment. See, e.g., Davis v. Calif ano, 613 
F.2d 957, 962 (D.C. Cir. 1980); Reed v. Lockheed Aircraft Corn., 613 F.2d 
757, 762 (9th Cir. 1980).

^Petitioner’s Brief at 38-9.



42

Tj'avenol’s subjective interview system because he could not 
use the disparate impact model. However, such was not the 
result in Payne. The Fifth Circuit analyzed proof which the 
plaintiff had tailored to meet the Griggs disparate impact ele­
ments under the d isparate trea tm en t model required by 
Pouncy. Relying upon this Court’s decisions in Teamsters and 
Hazelwood, the Fifth Circuit affirmed the trial court’s conclu­
sion that because of its use of a subjective interview process, 
Travenol had engaged in intentional discrimination.43

Thus, it seems clear that Title VII plaintiffs do not need an 
additional method for challenging subjective employment 
practices. The disparate trea tm en t model of M cDonnell 
Douglas IBurdine and its application in Teamsters and Hazel­
wood provide the proper theory for rooting out any discrimi­
natory intent or treatment contained in an employer’s subjec­
tive employment practices. Petitioner’s claims were properly 
considered by the trial court under the disparate treatment 
model. There is no necessity or basis for creating a different 
model to review them again. Petitioner was unable to con­
vince the trier of fact that she had been the victim of discrim­
inatory treatm ent and the Court of Appeals has affirmed. 
Those factual findings are not challenged before this Court, 
therefore this Court should affirm the Court of Appeals 
judgment.44

V. The Better Reasoned Decisions of the Courts of Appeals 
Recognized that Disparate Impact Model is not Appropriate 
for Analyzing Subjective Employment Judgments

Petitioner has incorrectly represented to the Court that 
nine of the twelve Courts of Appeals have adopted the dispar- * 4 * * *
iAPayne v. Travenol Laboratories, Inc., 673 F.2d 798, at 815-20.
i4See, Goodman v. Lukens Steel Co., 482 U .S .____ , 96 L.Ed.2d 572, 584,

107 S .C t.------- (1987) (In a Title VII case, where the Court of Appeals did
not set aside the trial court’s findings of fact, on certiorari this Court will
not examine the record “absent some extraordinary reason.”).



43

ate impact model for review of subjective selection criteria. 
(Pet. Br. 35, 36, 47). While there are indeed some of the Cir­
cuits which have adopted the position advocated by petitioner, 
impact analysis of subjective criteria has only been consis­
tently applied in four Circuits.40 * * * * * The better reasoned and more 
recent decisions of the remaining Circuits either directly sup­
port respondent or indicate approval of the authority followed 
in the case at bar.

In Pouncy v. Prudential Ins. Co. of America, 668 F.2d 795 
(5th Cir. 1982) the Fifth Circuit first considered the suitability 
of applying the disparate impact model to analyze a claim that 
certain employment practices fell more harshly on blacks 
when no facially neutral employment practices were identified 
by the plaintiff. Among the challenged employment practices 
was the claim that Prudential used subjective criteria to eval­
uate its employees. The Fifth Circuit held that such practices 
were not the type of facially neutral employment practices to 
which the disparate impact model applies.

The Tenth Circuit considered this issue in a factual context 
which is quite similar to that presented by Watson. In Morten- 
sen v. Callaway, 672 F.2d 822, 824 (10th Cir. 1982), the plain­
tiff sought to apply impact analysis to a subjective evaluation 
system which determined who was selected for promotion. 
The Tenth Circuit rejected use of disparate impact analysis, 
holding that the subjective selection criteria “were not crite­
ria an employee had to meet or test an employee had to pass in 
order to be considered” for the promotion.46

4oSixth Circuit, see, e.g., Rowe v. Cleveland Pneumatic Co., Numerical
Control, 690 F.2d 88, 93 (6th Cir. 1982); Ninth Circuit, see, e.g., Atonio v.
Wards Cove Packing Co., 810 F.2d 1477,1480-1486 (9th Cir. 1987) (en banc);
Eleventh Circuit, see, e.g., Griffin v. Carlin, 755 F.2d 1516,1524-1525 (11th
Cir. 1985); District of Columbia Circuit, see, e.g., Segarv. Smith, 738 F.2d 
1249, 1270-1272, 1288 n.34 (D.C. Cir. 1984) cert, denied, 471 U.S. 1115
(1985).

ibMortensen v. Callaway, 676 F.2d at 824 n.l.



44

This same result was reached by the Eighth Circuit in Har­
ris v. Ford Motor Co., 651 F.2d 609, 611 (8th Cir. 1981). The 
Court stated: “A subjective decision-making system, how­
ever, is not the type of practice outlawed in Griggs”. Harris v. 
Ford Motor Co. at 611. The Eighth Circuit’s rejection of use of 
impact analysis of subjective criteria was recently reaffirmed 
in Paige v. Marsh, No. 86-1282 (slip op; available on Lexis) 
(8th Cir. Feb. 3, 1987).

The position of the Fourth Circuit is in line with the Fifth 
Circuit’s decision in Pouncy. Subjective employment prac­
tices are not appropria te  for d isparate  im pact analysis 
because of the lack of any “objective standard, applied evenly 
and automatically”. Stastny v. Southern Bell Telepho?ie & Tele­
graph Co., 628 F.2d 267, 274 (4th Cir. 1980). See also, Brewster 
v. Barnes, 788 F.2d 985, 993 (4th Cir. 1986). The Seventh Cir­
cuit also recognizes that claims which challenge the manner 
which employment practices are applied  are inherently 
intent-based and must be tested under the treatment theory. 
In refusing to apply impact analysis to subjective criteria, the 
Court in Coates v. Johnson & Johnson Co., 756 F.2d 524 (7th 
Cir. 1985) said that the essence of the plaintiffs claim did not 
involve rules which were facially neutral, but instead the 
plaintiffs complaint was that the rules were “more severely 
applied, through supervisory discretion”, to blacks than to 
whites.47 Where the issue is the manner a rule or requirement 
is applied, rather than the effect of the rule or requirement, 
the proper framework for analyzing the evidence presented is 
the disparate treatment model.

Perhaps the Second Circuit’s opinion in Rossini v. Ogilvy & 
Mather, Inc., 798 F.2d 590 (2d Cir. 1986) best characterized 
the paradoxical nature of Watson’s position. On the one hand 
she attacks the bank’s subjective evaluation system as being 
inherently suspect (hinting strongly of a claim of intentional
47Coates v. Johnson & Johnson Co., 756 F.2d at 530-531 n.4.



45

discrimination), but on the other she claims that the bank’s 
subjective evaluation system constitutes a facially neutral 
standard, appropriate for d isparate impact analysis. The 
Court points out that when subjective criteria are at issue, it 
is “logically impossible to prove both propositions.”48 Judging 
from the manner Watson presented her case in the District 
Court, it appears that she recognized that impact analysis was 
not available to her and, accordingly, concentrated her efforts 
in an attempt to prove a disparate treatment case.

Although the recent decision of the Ninth Circuit, Atonio v. 
Wards Cove Packing Co., 810 F.2d 1477 (9th Cir. 1987) (en 
banc) differs with respondent on use of impact analysis to test 
subjective employment practices, Atonio is consistent with 
Pouncy in prohibiting use of the disparate impact model to 
attack  an employer’s overall selection process. Both the 
majority and concurring opinions make clear that impact anal­
ysis would be available only where the plaintiff (1) identifies a 
specific selection criterion, and (2) establishes a casual connec­
tion between the identified practice and the impact.

The D.C. and Eleventh Circuits have also held that the dis­
parate impact model can be applied to subjective, multiple cri­
teria. In Griffin v. Carlin, 755 F.2d 1516 (11th Cir. 1985), the 
Court, completely ignoring the availability of the disparate 
treatm ent model to analyze adverse impact against a pro­
tected group, reasoned that “limiting the disparate impact 
model to situations in which a single component of the process 
results in an adverse impact completely exempts the situation 
in which an adverse impact is caused by the interaction of two 
or more components.” 755 F.2d at 1525. Further, the Court 
errs in suggesting that employers are better situated to iden­
tify the variable of their multi-component or multi-step selec­
tion process that caused the adverse impact. There is no merit 
to the Eleventh Circuit’s suggestion. A two-equation regres­
48Rossini v. Ogilvy & Mather, Inc., 798 F.2d at 605.



46

sion analysis will identify which variable, if any, is responsible 
for the challenged disparity.49 The plaintiff can conduct that 
regression analysis as easily as can the employer. As this 
Court has observed, the liberal rules of discovery and plain­
tiffs access to the EEOC’s investigatory files makes the plain­
tiff sufficiently well situated to explain the challenged employ­
ment decisions.50 The Eleventh Circuit’s holding on this issue 
was unnecessary  and not consistent w ith th is C ourt’s 
decisions.

In Segarv. Smith, 738 F.2d 1249, 1271-72 (D.C. Cir. (1984), 
cert, denied sub nom. Meese v. Segar, 471 U.S. 1115 (1985), the 
Court created its own hybrid d isparate im pact/disparate 
treatm ent model in which disparate treatm ent analysis is 
completely subsumed within disparate impact analysis. In 
Segar, the Court attempted to justify requiring the employer 
to pinpoint the employment practice in the subjective system 
at issue causing the disparity. The Court reasoned that after 
the employer effectively rebuts the disparate treatm ent claim 
by articulating a legitimate business reason for the disparity, 
the employer already will have identified which employment 
practices caused the observed disparity and, therefore, which 
employment practices it will have to validate as job-related 
under the disparate impact theory. 738 F.2d at 1271-72, 1288. 
The Court, therefore, not only required the defendant to 
prove the causation element of the plaintiffs’ prima facie dis­
parate impact case, but also rendered the disparate treatment 
model totally unnecessary. Under Segar, when the employer 
avoids liability under the disparate treatm ent theory by artic­
ulating a legitimate business reason, the plaintiff can, without 
any further proof, merely switch over to the disparate impact 
theory and require the employer to prove that the legitimate
49See, Campbell, Regression Analysis In Title VII Cases: Minimum. Stan­

dards, Comparable Worth, and Other Issues Where Law And Statistics 
Meet, 36 Stan. L. Rev. 1299, 1312-1319 (1984).

mTexas Dept, of Community Affairs v. Burdine, 450 U.S. 248, 258 (1981).



47

business reason is also a business necessity. Disparate treat­
ment thereby would become completely subsumed within the 
disparate impact.

Segar further errs in suggesting that discriminatory intent 
is difficult to prove, necessitating an alternative method for 
plaintiffs to prove their cases. Segar incorrectly proposes that 
the M cDonnell Douglas and Burdine  framework may be 
abandoned, and the Griggs proof substituted, whenever a 
defendant articulates a nondiscriminatory reason for the chal­
lenged action. The Segar approach is further flawed in that it 
created a procedural scheme which allows the plaintiff to pur­
sue a Title VII claim without giving the defendant notice of 
the claim to be advanced. Since the true focus of inquiry will 
not be identified until the defendant articulates it nondiscri­
minatory reason for the challenged employment action, the 
defendant is required to show the business necessity of 
employment practices which were not identified by the plain­
tiff. The hybrid procedural approach suggested by Segar is 
totally contrary to the carefully crafted analytical framework 
created by this Court and has no application to the case at 
hand.

As has been demonstrated, the better reasoned decisions of 
the Court of Appeals support respondent’s position tha t 
impact theory should not be extended to analyze subjective 
employment criteria. The decision of the Fifth Circuit below is 
consistent with this Court’s prior decisions and represents the 
most logical approach for the analysis of such subjective 
criteria.



48

CONCLUSION

In the above reasons, the judgment of the Court of Appeals 
for the Fifth Circuit should be affirmed.

Respectfully submitted,

Bruce W. McGee*
Joseph W. Spence 
GANDY MICHENER SWINDLE 

WHITAKER & PRATT 
2501 Parkview Drive, Suite 600 
Fort Worth, Texas 76102 
(817) 878-0565 

Attorneys for Respondent 

*Counsel of Record



SHS®
’ ■ * ' .

i  > ■