July 2 0 , 1978
78-41 MEMORANDUM OPINION FOR THE COUNSEL
TO THE ASSOCIATE ATTORNEY GENERAL,
ATTORNEY PERSONNEL
Employee Selection Procedures— Use of LSAT Scores
in the Department’s Honor and Summer Intern
Programs
This responds to your predecessor’s request for our opinion whether the
Department may consider Law School Admission Test (LSAT) scores of
applicants for the Honor and Summer Intern Programs. For the reasons that
follow we recommend against such use.
In a memorandum from your predecessor to the Civil Rights Division, he
explained how the Department uses and considers LSAT scores in these
programs. The score, he maintains, is only a minor factor in the evaluation of
program applicants. He also states that the score is considered a “ rough
indication of intellectual ability.”
The use of tests and test scores for employment purposes is a major subject in
employment-discrimination law. The Supreme Court in Griggs v. Duke Power
C o., 401 U.S. 424 (1971), established the basic standards by which employee
selection devices, including tests, were to be judged to determine whether they
illegally furthered discrimination in violation of Title VII of the Civil Rights
Act of 1964, 42 U.S.C. § 2000e et seq. The Court held that employment
practices that operate to exclude protected class members under Title VII1 and
that cannot be shown to be related to job performance are prohibited. Id., at
431. If a practice operates disproportionately to exclude minorities, the
employer must meet the heavy burden of proving that the practice “ bear[s] a
demonstrable relationship to successful performance of the jobs for which it is
used.” Id. The Equal Employment Opportunity Commission (EEOC) has
published “ Guidelines on Employee Selection Procedures.” 29 CFR § 1607.1
'T itle VII prohibits em ploym ent discrim ination on the basis o f race, color, religion, sex, or
national origin. 42 U .S .C . § 2000e-2(a). Because our discussion relates to tests as em ploym ent
devices and because racial m inorities generally do not perform as well as the rest o f the population
on written tests our focus will be on how the D epartm ent’s use o f the LSAT affects racial
m inorities.
166
et seq.2 One of the concerns which led to the publication of these guidelines is
the common practice of “ using tests as the basis for employment decisions
without evidence that they are valid predictors of employee job performance.”
29 CFR § 1607.1(b). Where such evidence is lacking “ the possibility of
discrimination in the application of test results must be recognized.” Section
1607.3 of the guidelines, in defining discrimination, essentially restates the
Griggs standard. It provides, in pertinent part, that the use of a test that
disproportionately rejects minorities in the hiring process constitutes discrimi
nation unless the test is predictive of, or significantly correlated with, actual job
requirements. Even where the test is reasonably related to job requirements, if
it disproportionately rejects minorities the employer must show that there is no
suitable alternative hiring procedure that would impact less heavily on
minorities. Id.
We now tum to the Department’s use of LSAT scores to see whether its
procedure comports with the above rules. The memorandum from your Office
explains the Department’s use of these scores as follows. The Department
operates on the premise that the LSAT score is a rough indicator of intellectual
ability. Proceeding from this premise it explains the significance that the
Department attaches to these scores:
In order to evaluate the non-intellectual abilities of the candidate the
LSAT score is compared to the applicant’s academic record. If a
person has a high LSAT score, but only average grades then it
suggests that the person is an underachiever and we are therefore not
interested in him. By the same token, a mediocre or low LSAT score
coupled with high academic performance suggests that the candidate
is a hard worker and self disciplined. The person did not achieve his
excellent grades by intellectual ability alone. This weighs very
heavily in the candidate’s favor. Finally, a high LSAT score and high
academic performance suggests that not only is the person very bright
but he or she is also a hard worker.
This explanation may be illustrated by the following categorization of appli
cants:
(1) high grades— high LSAT
(2) high grades— average LSAT
(3) average grades— average LSAT3
(4) average grades— high LSAT
2These guidelines are entitled to great deference and have been follow ed by virtually every court
dealing with these issues. See. Douglas v. Hampton, 512 F. (2d) 976, 986 (D .C . Cir. 1975), and
cases cited. See also. Washington v. Davis, 426 U .S . 229, 247, n. 13 (1976).
■
’ Although the m em orandum does not state how this com bination o f grades and LSAT scores
bears upon the em ploym ent decision, this category o f applicants seem s logically to fall betw een
classes 2 and 4. Class 4 m em bers are unfavorably viewed as “ underachievers.” That label would
not fit class 3 m em bers since their grades are com m ensurate with their LSAT scores. T hus, class 3
members would appear to be considered more desirable applicants than those in class 4. Class 3
m em bers, how ever, are not viewed as favorably as class 2 mem bers. Class 2 mem bers are seen as
hard working and self-disciplined. It would seem to follow that class 3 m em bers do not warrant
these labels because their grades were consistent with their LSAT scores.
167
We have listed the categories in the order of the most desirable applicants (class
1) to the least desirable applicants (class 4); desirability is based on the
reasoning of the above quoted statement.
At this point it is important to keep in mind that the use of test scores where
the test is not predictive of or correlated with job performance is a discrimina
tory practice only insofar as it operates to reject disproportionate numbers of
protected class members. Therefore, we must consider the adverse impact that
use of LSAT scores has on minority applicants. To proceed with our analysis
we make two basic assumptions to determine whether there is a possible
adverse effect on minority applicants. First, we assume that minority members
as a general rule do not perform as well as nonminority persons on the LSAT.
Second, we assume that minority members as a general rule receive lower
law school grades than nonminority persons.4
Accepting these assumptions as valid, we can now evaluate how the
Department’s use of LSAT scores may affect minority applicants in the Honor
and Summer Intern Programs.
I. Class 1 (high grades— high LSAT)
Class 1 would include very few minority members because the high grades
and high LSAT score are inconsistent with assumed minority performance.
Therefore, this class would, to a large degree, be composed of whites. A high
LSAT score adds to their desirability since the Department would view an
individual in this class as very bright and hard working. Thus, in this class the
LSAT is considered as a positive factor. The effect of this is to give these
predominantly white applicants an additional advantage based on their LSAT
scores. Members of the other classes are adversely affected by this because the
effect of increasing the ratings for class 1 members serves to make members of
the other classes less desirable comparatively.
II. Class 2 (high grades— average LSAT)
Because of their high grades, members of this class would also be
predominantly white. Interestingly, here the average LSAT would actually be
considered favorably. The theory is that the class member received grades
higher than expected. Thus, in this case, an average LSAT score is a positive
factor. Although it seems anomalous, this favorable consideration of average
LSAT scores adversely affects minority members. Most minorities would not
4U nfortunately, we do not have ready access to the actual statistics on this subject, if indeed any
exist. H ow ever, we believe that these assum ptions are fully warranted since there is a “ substantial
body o f evidence that black persons and other disadvantaged groups perform on the average far
below the norm for whites on generalized intelligence or aptitude tests. ’ ’ Douglas v. Hampton, 5 12
F. (2d) 976, 983, quoting from Arrington v. Massachusetts Bay Transp. Auth., 306 F. Supp. 1355,
1358 (D. M ass. 1969). The LSA T concededly is a general intelligence test. T herefore m inorities
would not be expected to score as high as w hites. See also. Racial Bias and the LSAT: A New
Approach to the Defense o f Preferential Admissions, 24 Buff. L. Rev. 439, 456 (1974), and Bell,
In Defense o f Minority Admissions Programs: A Response to Professor Graglia, 119 U. Penn. L.
Rev. 364, 367 (1970).
168
receive the benefits connected with average LSAT scores because, not having
received high grades, they would not be in this class. Thus, here again, it is fair
to assume that whites would, to a disproportionate degree, benefit from the use
of LSAT scores.
III. Class 3 (average grades— average LSAT)
Most black applicants would fall in this class. Consideration of the LSAT
scores would result in no discernible advantage to blacks based upon the
reasoning of the memorandum.
IV. Class 4 (average grades— high LSAT)
This is the one class where LSAT would adversely impact on predominantly
white class members. The high LSAT would result in fewer black class
members. The negative inferences drawn from the average grades— high LSAT
combination would diminish employment opportunities for these applicants.
As the foregoing illustrations demonstrate, it is quite possible that the
Department’s use of LSAT scores may work to the disadvantage of minorities.
Accurate data on the Department’s use of these scores would be required before
we could say, with any assurance, that this possible adverse impact is
consistent with what actually occurs. However, it is surely a possibility.
Proceeding under the premise that the Department’s use of the LSAT has a
possible discriminatory impact, the issue becomes whether such use is
reasonably predictive of job performance. The Educational Testing Service
(ETS) prepares the LSAT. This organization has consistently warned against
use of the LSAT in employment decisions since it is of doubtful validity as a
predictor of success in practice. Your memorandum states that the Depart
ment’s primary use of LSAT scores is to measure motivation. That is, the test
results are used to see how an individual’s grades stack up against his or her
LSAT rating. The validity of the Department’s assumption regarding appli
cant motivation has not been established. The EEOC guidelines state that:
Evidence of a test’s validity should consist of empirical data
demonstrating that the test is predictive of or significantly correlated
with important elements of work behavior which comprise or are
relevant to the job or jobs for which candidates are being evaluated.
28 CFR § 1607.4(c).
It thus seems that the Department is obliged to determine whether its
assumptions on applicant motivation are empirically supportable.
Congress, in passing the 1972 amendments to Title VII, Pub. L. 92-261, 86
Stat. 103, extended the protections of Title VII to Federal employees. Their
legislative history shows that the Federal Government’s use of unvalidated
hiring criteria was a major concern to the Congress.
Civil Service selection and promotion requirements are replete with
artificial selection and promotion requirements that place a premium
on “ paper” credentials which frequently prove of questionable value
as a means of predicting actual job performance. The problem is
169
further aggravated by the [Civil Service Commission’s] use o f
general ability tests which are not aimed at any direct relationship to
specific jo b s. The inevitable consequence of this, as demonstrated by
similar practices in the private sector, and found unlawful by the
Supreme Court, is that classes of persons who are culturally or
educationally disadvantaged are subjected to a heavier burden in
seeking employment. [Emphasis added.]5
Congress decried the use of such hiring criteria, stating that the “ inevitable
consequence” of this is to create an added and unwarranted burden on
disadvantaged classes. The D.C. Court of Appeals in Douglas v. Hampton,
512 F. (2d) 976 (D.C. Cir. 1976), likewise condemned the use of unvalidated
hiring criteria in holding that Federal employment tests must rationally measure
required job skills. Cf., Washington v. Davis, 426 U.S. 229, 247, n. 13 and
accompanying text (1976).
Private employers covered by title VII are required to use validated selection
criteria. It would be anomalous for the Federal Government not to meet this
same requirement. The United States Commission on Civil Rights in a July
1975 report, The Federal Civil Rights Enforcement Effort— 1974, Vol. V.—To
Eliminate Employment Discrimination, stated:
The Federal Government must not be permitted the continued use of
employment selection standards which close the doors to groups
victimized by years of discrimination without any empirical proof of
such standards’ relation to job performance; to do so, would permit
the Government to escape adherence to the requirements it, itself,
imposes on private employers. Such policy decisions within the
Government seriously erode the Government’s own credibility as an
enforcer of the law [footnotes omitted]. Id., at 42-43.
We are not unsympathetic with the unique problems involved in formulating
accurate predictors of attorney job performance. Nor do we fail to recognize
that other employment criteria for attorneys are not purely objective. Law
school grades6 and employment interviews are not validated as predictors, and
we do not here address the question whether they should be. We agree,
however, with the Civil Rights Division’s view insofar as it maintains that:
In light of the difficulty in evaluating job performance, we are forced
to use imprecise indicators of ability. Where, however, as here, we
have reason to question the usefulness of an indicator, we believe it
should be eliminated as a criterion of selection, particularly in light of
the appearance it creates of the application of a lesser standard of
compliance with Title VII in Department hiring than in hiring by
other employers.
5H. Rept. N o. 238, 92d C o n g ., 1st sess. 24 (1971), 1972 U .S. Code Cong. Adm in. N ew s, 2159.
6G rades in law schools are m ore reliable indicators than are LSAT scores, in that grades are
given on the basis o f legal work perform ed or questions answ ered. Therefore, a student is graded
for activities sim ilar in m any respects to w ork he or she will do in practicing law.
170
Your memorandum states that the Department is not insensitive to the
possibility of the LSAT being tainted by cultural biases. It is because of this
concern, it maintains, that the LSAT score plays a less significant role in
evaluating minority candidates. While this may be true, it fails to deal with the
possibility of nonminority candidates’ ratings being augmented because of
LSAT considerations. This would have the same effect as penalizing minority
candidates for their performance on the LSAT because even though they might
not be negatively considered as a result of their scores, their ratings would be
comparatively lower because of LSAT consideration.
A second and possibly more important issue is this. If a selection procedure
impacts adversely on minorities as a result of a discriminatory practice, the
degree of discrimination is irrelevant. See, Bolton v. Murray Envelope Corp.,
493 F. (2d) 191 (5th Cir. 1974); Rowe v. General M otors Corp., 457 F. (2d)
348 (5th Cir. 1972). While we concede it to be far from clear that the
Department’s use of LSAT scores to measure motivation results in a dispropor
tionate rejection of minority applicants, a plausible case can be made that this in
fact occurs. In sum, we recommend that the Department abandon the policy of
considering LSAT scores to determine employee motivation because of the
procedure’s (1) questionable reliability, (2) uncertain legality, and (3) apparent
conflict with requirements of private employers.
Jo h n M . H arm on
Assistant Attorney General
Office o f Legal Counsel
171