540 F.2d 744
12 Empl. Prac. Dec. P 11,243
David RICHARDSON et al., Appellants,
v.
J. Means McFADDEN et al., Appellees.
David RICHARDSON et al., Appellees,
v.
J. Means McFADDEN et al., Appellants.
Nos. 73-2512, 73-2513.
United States Court of Appeals,
Fourth Circuit.
Argued April 8, 1976.
Decided Aug. 30, 1976.
Ray P. McClain, Charleston, S. C. (Laughlin McDonald, Neil Bradley, Atlanta, Ga., F. Henderson Moore, Charleston, S. C., Melvin L. Wulf and E. Richard Larson, New York City, on brief), for David Richardson et al.
Randall T. Bell, Columbia, S. C. (Daniel R. McLeod, Atty. Gen., and A. Camden Lewis, Asst. Atty. Gen., Columbia, S. C., on brief), for J. Means McFadden et al.
Before BOREMAN, Senior Circuit Judge, CRAVEN, Circuit Judge, and HADEN, District Judge.*
CRAVEN, Circuit Judge:
This action for declaratory and injunctive relief was brought by four black law school graduates1 who had satisfied all requirements for admission to the South Carolina Bar except that they received failing scores on the bar examination. They challenge the constitutionality of the South Carolina Bar Exam as applied generally to black applicants. Appellants Spain and Kelly also attack its validity as applied to them personally.
In the district court appellants alleged and undertook to prove: (1) direct and purposeful discrimination against blacks by the State Board of Law Examiners in assigning grades to their papers; (2) failure of the State to demonstrate that the bar examination is job related as opposed to simply a measurement of general educational preparation; (3) denial of due process of law in the failure of the Bar Examiners to provide an established procedure whereby review of, and challenge to, the assignment of a failing score could be had; and (4) arbitrary and capricious application of the Examiners' own standards and criteria to the examinations of Spain and Kelly, with the result that they were denied passing scores.
The district court, after trial, rejected appellants' arguments, except as to the claim that due process was denied unsuccessful applicants by the failure of the Bar Examiners to provide a system for review of failing papers. Judge Blatt abstained as to that issue until appellants had presented it to the South Carolina Supreme Court. On appeal the plaintiffs below assign error as to all adverse rulings except their contention that tests were graded in an intentionally racially discriminatory manner. The Bar Examiners cross-appealed the court's decision to abstain from decision rather than dismissing outright the due process challenge to the absence of a failing test challenge procedure.2
We affirm the district court, except as to the individual claims of Spain and Kelly.
I.
Appellants' main challenge to the South Carolina Bar Examination is that it is not job related. They argue that the State Bar's past history of racial discrimination in admitting blacks to the practice of law and the disproportionate impact of the examination on blacks places on the Examiners the burden of showing that the exam is job related, and that in this context the standards for judging job relatedness should be those of Title VII of the 1964 Civil Rights Act3 rather than traditional Fourteenth Amendment tests. Appellants contend that to be constitutional the test in general must be shown to measure skills relevant to the practice of law, and the passing score must be selected so as to draw the line at minimal professional competency. While the main thrust of their argument goes to establishing Title VII standards as applicable to the bar examination, they contend that, regardless of the standard employed, the Examiners have failed to demonstrate the requisite job relatedness.
The district court, examining the question solely under the criteria of the Fourteenth Amendment, held that "the South Carolina bar examination as presently administered has a rational connection with any applicant's fitness or capacity to competently practice law in this state." App. 44. He made no explicit finding concerning the passing score, probably because the question was not presented to him in those terms, but we believe that, in ruling that the bar examination "as presently administered" was "job related," he meant to indicate approval of both the test in general and the selection of a passing score.
But if Title VII's standards for job relatedness are applied, the test used must be "shown, by professionally acceptable methods, to be 'predictive of or significantly correlated with important elements of work behavior which comprise or are relevant to the job or jobs for which the candidates are being evaluated.' " Albemarle Paper Co. v. Moody, 422 U.S. 405, 431, 95 S.Ct. 2362, 2378, 45 L.Ed.2d 280 (1975). While the Bar Examiners do not concede that they would lose under this requirement, we believe the record is inadequate to demonstrate either "criterion" ("predictive"), "content," or "construct" validity under professionally acceptable methods.4 Thus, if we were to determine that Title VII standards were applicable, it would be necessary to reverse and declare the South Carolina Bar Examination constitutionally invalid.
Appellants agree that Title VII does not apply to the bar exam by its own terms. But they point out that in Walston v. County School Board, 492 F.2d 919 (4th Cir. 1974), we incorporated Title VII standards into the Fourteenth Amendment equal protection guarantee in the context of past, state-sanctioned racial discrimination and the disparately adverse impact of the test upon blacks. So we did. See also United States v. Chesterfield County School District, 484 F.2d 70 (4th Cir. 1973).
However, in Washington v. Davis, --- U.S. ----, 96 S.Ct. 2040, 48 L.Ed.2d 597 (1976), decided after oral argument in this case, the Supreme Court limited our Walston approach,5 holding that, where discriminatory purpose by the state is not proven, it is inappropriate to "adopt this more rigorous standard (Title VII) for the purposes of applying the Fifth and the Fourteenth Amendments . . . ." Id. at ----, 96 S.Ct. at 2051.
To prove discriminatory purpose, appellants rely on circumstantial evidence generated by the chronological juxtaposition of three changes in admission practices to the State's Bar. The first was the elimination of "diploma privilege," which allowed graduates of the State's accredited law school to gain automatic admission to the Bar. This path was eliminated in 1950, three years (the normal law school term) after a "separate but equal" law school was started at South Carolina State College, a black school. Appellants argue the same pattern was followed as to the practice of "reading law." This avenue for admission to the Bar was eliminated in 1957, "coincidentally" shortly after a black applicant used this method. Finally, they contend that reciprocity was abolished in January 1972, not long after a black member of the Oklahoma Bar applied under the reciprocity rule for admission to the Bar.
The Bar Examiners respond with three points. First, they note, and appellants do not contend otherwise, that there have never been laws or rules of court prohibiting blacks from practicing law in the State or imposing different standards based on race. This is important perhaps of controlling importance. Second, it is statistically clear that admission to the State's Bar has been relatively open to blacks according to the 1970 Census, South Carolina has the highest proportion of black lawyers in its Bar of any state in the nation.6 Finally, they cite neutral reasons for all three changes in admission practices.7
We think it clear that the evidence falls short of proving (unlike the school cases) a deliberate state scheme of de jure discrimination. Thus, despite proof of disparate impact, the application of Title VII standards is not appropriate. Davis, supra.
Job-Relatedness Under the Equal Protection Clause
We agree with the Fifth Circuit in Tyler v. Vickery, 517 F.2d 1089 (5th Cir. 1975), cert. denied, --- U.S. ----, 96 S.Ct. 2660, 49 L.Ed.2d 393 (1976)8 that under the Equal Protection Clause of the Fourteenth Amendment the issue is still whether the examination is job related, albeit a less demanding inquiry. "The hallmark of a rational classification is not merely that it differentiates, but that it does so on a basis having a fair and substantial relationship to the purpose of the classification." Id. at 1099. And here the purpose of the classification is to distinguish between persons demonstrating minimal competence to practice law and those lacking such knowledge and skill.
The testimony introduced at trial on the validity of the test questions consisted primarily of testimony by the individual Bar Examiners. Each testified to basically the same facts: (1) that he is a successful practicing lawyer in the State and, from observation and experience, understands the skills necessary to practice competently; (2) that he has examined sample questions prepared by the National Council of Bar Examiners and/or discussed question formulation with others; and (3) that he designed the questions he placed on the Bar so as best to determine whether the applicant possesses the minimal level of competence necessary to practice law in the State. In addition, the Examiners' experts, Statler and Bernreuter, testified that they had performed correlation studies between performance on the bar examination and performance in law school as measured by the law school grades. Both experts testified that, in their opinion, the results demonstrated the "content validity" of the bar examination, where "content" is defined as law school performance.9 Appellants did not challenge the existence of such correlation; they only attacked its significance and argued that it was irrelevant to any then established procedure for showing content validity. They contend that, absent a job analysis, which was not performed here, there can be no demonstration of content validity.10
In Washington v. Davis, supra, the Supreme Court held that, at least in validating admission to a police training program, a "positive relationship between the (admission) test and training course performance (is) sufficient to validate the former, wholly aside from its possible relationship to actual performance as a police officer." --- U.S. at ----, 96 S.Ct. at 2053. In fact, the Court found this relationship to be "the much more sensible construction of the job relatedness requirement." Id.
While the Court's treatment of the job relatedness requirement in Washington v. Davis has no direct application to professional licensing examinations because of the differing state interests involved, we believe the Court's general reasoning gives substantial support to the Examiners' argument that under the Fourteenth Amendment the relationship between law school performance and the bar examination is significant in establishing job relatedness.11
We believe that this statistical evidence plus the other efforts of the Examiners to intelligently relate the examination questions to the skills involved in the practice of law are sufficient to satisfy the demands of the Fourteenth Amendment.
Whether the passing score selected by the Bar Examiners bears "a fair and substantial relationship" to the determination of minimal competency presents a much more difficult question. In United States v. North Carolina, 400 F.Supp. 343 (E.D.N.C.1975) (three-judge court), the district court, in a similar context, held that to satisfy the demands of the Equal Protection Clause the cut-off score selected must be reasonably related to minimal competency, and it held invalid the minimal passing score selected by North Carolina for a National Teacher Examination because of a total absence of evidence showing it to be related to minimal teaching competency.
In this case, some evidence was introduced to validate the cut-off score of 70 used by the Examiners, but it was very subjective and general in nature and hardly acclaimed by the educational testing experts who testified.12 Perhaps the testimony most supportive of the validity of the cut-off score was provided by Examiner David Freeman, who described his grading procedures as follows:
My own approach is that, preliminary to the grading process, to go back to the exam question and in studying through them very carefully, I make a mental assessment of the importance to be attached to each one. I do not go through, for my own purposes: I feel that it's a mechanical process of assigning a point value to each question. I then read the examinations. I treat them not as questions to which so may points were assigned to this or to that issue of this question, but as a totality and assign to that paper a grade which I think is reflective of the student's evidence of ability in answering the whole.
Q. What form would that grade take, a letter grade?
A. It would be a numerical grade. And I think, for my own testing purposes, the magic passing point is 70, and I range upward or downward through that.
Q. All right sir, as I understand it, you read the entire paper and then assign a single numerical grade, with 70 as passing?
A. That is right.
Q. As I understand it, you don't attempt to assign points to any particular portion of the test?
A. Not in a numerical fashion. It is a matter, in the preliminary process, of giving a mental assessment of importance to this question, or lesser importance to this question, or lesser importance to that question. When I have finished grading a paper, what I would have is one grade that I put on there.
App. Exhibits at 325-26.
Other Examiners employed a very mechanical system, assigning points to particular parts of questions, summing those points, and then in some cases obtaining the 70 cut-off line simply by raising the highest score to an "A" or perfect score.13 We tend to agree with appellants' expert that, if this second system is utilized in the precise manner described by the Bar Examiners, it would be almost a matter of pure luck if the "70" thereby derived corresponded with anybody's judgment of minimal competency.
But absent professionally validated, administered, and evaluated examinations, it is not clear that to require grading along the lines discussed by Freeman rather than the more mechanical and arbitrary method used by McFadden is anything more or less than to demand greater subjectivity. It is not at all certain which of these two, both of whom are competent lawyers but laymen at question design and evaluation, generates numerical scores which more accurately reflect their "true" evaluation of competency.
In view of the fact that all Examiners both designed their exams and assigned scores so as to indicate their judgment as to minimal competency, we cannot find the results obtained so unrelated to the State's objectives as to violate the Equal Protection Clause.14
II.
Appellants Spain and Kelly argue that individually their rights under due process and equal protection were denied in that, through arbitrary and capricious application of the Examiners' own standards, their "borderline" papers were certified as "failing." It is beyond question that the Bar Examiners are subject to the requirements of the Due Process and Equal Protection Clauses in the conduct of their duties. As the Supreme Court stated in Schware v. Board of Examiners, 353 U.S. 232, 238-39, 77 S.Ct. 752, 756, 1 L.Ed.2d 796 (1957):
A State cannot exclude a person from the practice of law or from any other occupation in a manner or for reasons that contravene the Due Process or Equal Protection Clause of the Fourteenth Amendment. . . .
There is no more fundamental proposition in our law than that a state may not act arbitrarily and capriciously to deprive a citizen of "liberty" or "property" interests. Such is the basic thrust of Spain and Kelly's argument, and it is clear that if their allegations are accurate they are entitled to relief.
Kelly and Spain base their case on the following table of examination scores:
Applicant Scores for each 6 examiners Average Pass/Fail
--------- --------------------------- ------- ---------
# 160 June, 1971 66 67 68 71 78 81 71.8 Fail
# 128 June, 1970 66 67 67 72 75 79 71.0 Pass
Spain June, 1971 66 66 68 71 72 80 70.5 Fail
# 121 June, 1969 66 69 F(69) 71 73 73.8 70.3 Pass
Kelly Feb., 1971 63 1/2 66 69 k 70 71 78 69.6 Fail
# 17 Feb., 1970 60 67 71 72 73 74 69.5 Fail
# 10 Feb., 1971 63 66 70 71 73 73 69.3 Pass
Appellants' Reply Brief at 22.
Their basic argument is that under standard Bar Examiner practice of "rounding up" grades of .5 or better to the next whole number both had passing cumulative scores Kelly 70 and Spain 71. The Examiners' response is that under their practices a person might "fail" the Bar even though he had a cumulative passing score if he had failed three or more of the individual Examiners. In those cases, the decision would be based, the Examiners testified, upon the "configuration of scores" and Examiners' notes containing remarks on the general quality of papers. We understand that during these proceedings the actual exam papers were not before the Examiners, and there was no review of the applicant's performance other than examination of the pattern of the grades and whatever notes the Examiners may have made concerning each paper.
But, appellants point out, individuals who had lower cumulative totals and not obviously different "configuration of scores" were passed. As to Spain, the Examiners' basic response is that one cannot expect perfection in the difficult borderline cases. As to Kelly, they argue that no one who failed three Examiners had a lower score and was judged to have passed the exam. They contend that individual # 10 "passed because he passed four examiners," indicating that passing would be automatic in such situations. Brief for Appellee at 63 n.37. That, at least, was their response before applicant # 17 was brought to their attention. He passed four Examiners, had a higher cumulative score than # 10, and still failed the examination. We therefore think the scores are comparable since there is no consistently applied distinction between them.
At oral argument, the Examiners offered a further explanation of how an individual may receive a passing grade when his cumulative total was lower than another applicant who failed. They tell us that written comments often accompany borderline scores and are employed to make these difficult decisions. In this fashion, they explain their different treatment of two applicants in 1973 (one passed and the other failed) who both had cumulative averages of 692/3. The one who passed the exam failed three Examiners and, in addition, failed the Multistate portion15 of the exam. The one who failed it also failed three Examiners but passed the Multistate portion.
The Examiners told us that this was perfectly reasonable because comments on the grading sheets corresponded with the ultimate results. The one who failed was noted to be "poor in expressing himself" and "didn't seem to have an understanding of legal principles." The one who passed was "marginal plus" and "had some good answers." We find reliance on these comments irreconcilable with the Board's contention that numerical scores are used to capture precise gradations in performance. It is not possible to pursue the goal of objectivity and also put ultimate reliance on subjective notes as general and vague as those cited above.
We would be concerned about this general inconsistency of treatment of scores in arriving at an overall result even if the scores from each Examiner were absolutely precise. However, as noted in the previous section, they are far from precise as to any Examiner, and for three Examiners, who raised each score by an arbitrary number of points, the correspondence of a score of 70 with even their own judgment of minimal competency was little more than fortuitous. On these facts, we hold that, as to Spain and Kelly, the Bar Examiners acted arbitrarily and capriciously in violation of both the Due Process and Equal Protection Clauses of the Fourteenth Amendment.
Accordingly, on remand, the district court will order that they be certified as having passed the South Carolina Bar.16
III.
The Bar Examiners appeal Judge Blatt's refusal to dismiss appellants' claim that lack of an established process for challenging exam results violated due process. The Examiners argue, first, that while there was at the time the suit was begun no such "express provision for review" or "ascertainable basis upon which to challenge bar examination results," there was no denial of due process because the State's Supreme Court, under its exclusive jurisdiction of admission to the Bar and under its inherent powers, could "review actions of the law examiners and . . . make final determination of who will and will not be admitted to" the Bar. Brief for Appellee at 35. They contend that no other rule or court action was required to guarantee an opportunity for review, and, as a result, due process is satisfied.
We cannot determine on this record whether adequate review was then available. We are not told whether such a system has ever operated. Under such circumstances, we find nothing improper in the district court's order which deferred action on this claim until the question was presented to the State Supreme Court. It is true that Judge Blatt's order may be read to indicate his tentative belief that the establishment of new review procedures may be necessary. But cautiously he deferred decision in favor of application to the State Supreme Court, which was the one body which could definitely determine what, if any, review was available and appropriate under state law.17
As a second argument, the Examiners contend that no review procedure was necessitated because the right of reexamination satisfied the requirements of due process. See Tyler, supra, at 1103-05; Whitfield v. Illinois Board of Law Examiners, 504 F.2d 474 (7th Cir. 1974).18 They contend that, if a person actually possesses the requisite skills, it will certainly be demonstrated by at least one of a series of reexaminations, and that the chance " 'the same individual would be the victim of error after two re-examinations is literally one in a million.' " Brief for Appellees at 38, citing Tyler, supra, at 1104.
To our knowledge, a person is not required by any state to repeatedly demonstrate his competence to practice law. The rule is: once is enough. And the reason for the rule is that it takes work, effort, and, nowadays, money to prepare for a bar examination. Moreover, the license is deemed of sufficient value that delay in getting it is an injury.
It is true that some courts have held that reexamination is a more effective remedy than review because the administrative burden of allowing challenges was perceived to be too great.19 We are not persuaded.20
AFFIRMED IN PART; REVERSED IN PART; AND REMANDED.
Sitting by designation
David Richardson was graduated from Washington College of Law of American University in 1970. Patrick Kelly and Hiram Spain received law degrees from Howard University in 1966 and 1971, respectively. Patricia King was graduated from North Carolina Central University Law School in 1969
On appeal, the Examiners argue that this court has no jurisdiction since generally only the Supreme Court can review decisions of a state court, and contend that the individual Examiners are immune from suits challenging the exercise of their judicial function. While both propositions are correct in principle, it is crystal clear that in giving and grading examinations and certifying passing scores the Board of Bar Examiners neither renders judicial decisions nor exercises judicial functions
42 U.S.C. § 2000e et seq
For a discussion of the meaning of these terms see, e. g., Douglas v. Hampton, 168 U.S.App.D.C. 62, 512 F.2d 976, 984 (1975); Bridgeport Guard, Inc. v. Bridgeport Civil Service Commission, 482 F.2d 1333, 1337-38 (2d Cir. 1973), cert. denied, 421 U.S. 991, 95 S.Ct. 1997, 44 L.Ed.2d 481 (1975)
In disapproving numerous cases from the lower federal courts that reached the contrary conclusion, the Court failed to cite either of our cases upon which appellants chiefly rely. Id. at ----, 96 S.Ct. at 2050 n. 12. This omission might be interpreted to mean that disproportionate impact upon blacks in the contest of historically stated-sanctioned discrimination establishes "discriminatory purpose." But at least two other explanations may be given for the Court's failure to cite either Walston or Chesterfield County
Inadvertent omission is, of course, possible. Also, the Court may have considered these "school cases" part of the "special context of school desegregation cases" upon which Keyes v. School District No. 1, 413 U.S. 189, 208, 93 S.Ct. 2686, 2697, 37 L.Ed.2d 548 (1973), relied. In that case, the Supreme Court cited with approval our holding in Chambers v. Hendersonville City Bd. of Educ., 364 F.2d 189, 192 (4th Cir. 1966) (en banc), that "in a school system with a history of segregation, the discharge of a disproportionately large number of Negro teachers incident to desegregation 'thrust(s) upon the School Board the burden of justifying its conduct by clear and convincing evidence.' " Id. at 189. Thus, Walston, which relied heavily upon Chambers, might be read as simply a part of that "special context," as the Examiners contend, or much more broadly, as appellants urge.
However, we need not in this case decide either the scope of Walston or the degree to which it survived Davis, since there is lacking here a history of invidious discrimination sufficient (with disparate impact) to trigger its application.
The District of Columbia has a higher percentage. Of D. C. lawyers, 9.4% are black, as compared with 3.4% in South Carolina. North Carolina is next with 3.3%, followed by Mississippi with 2.8%. Defendants' Exhibit A. App. Exhibits at 720
They contend that as early as 1940 the South Carolina Bar Association was on record supporting the American Bar Association effort to abolish both the "diploma privilege" and "reading law." Each change, they contend, was simply part of the effort to upgrade the educational qualifications of those admitted to the practice of law in the State, and each predated any possible racial motivations for the changes. The Examiners further note that both changes were in line with national trends. As to "reciprocity," the Examiners contend that the law was changed "in order to curb an increasing flow of semi-retired lawyers into the state not sufficiently skilled in the principles and practice of law in South Carolina."
Cited with approval in Washington v. Davis, supra, --- U.S. at ----, 96 S.Ct. at 2050 n. 12
See App. Exhibits at 618-21; App. 851-54, 871-72, 972-77
See App. 865-66, 1026-27. See also Equal Protection Challenge to the Bar Examination, 1975 Ariz.St.L.J. 531 (1976)
If it were carried to its logical extreme, seldom the path of the law, the Court's opinion on this point surprisingly might invalidate almost all state professional examinations. If the only demonstration of job relationship required is that it has a positive relationship to training course performance e. g., law school then why does not training school performance itself demonstrate that the applicant is fit to practice his profession? It is certainly clear that nothing correlates better with training school performance than training school performance itself. An applicant for the Bar who has graduated from an accredited law school arguably may be said to stand before the Examiners armed with law school grades demonstrating that he possesses sufficient job-related skills. Why, then, any bar examination at all?
See App. 715-16
See, e. g., Deposition of J. Means McFadden, App. Exhibits at 266:
I took the top man and I figured what an A paper would be and how many points I'd have to add to his score to get him an A paper, and I gave him that number of points to get him to an A. And I gave everybody else the same number of points.
That is not to say that such an unprofessional approach leaves us with much confidence in the precise numerical results obtained. It is unfortunate the State has not combined the advantage of utilizing Bar Examiners who are practicing professionals with consultation and guidance from experts in question evaluation
The Multistate Bar Examination (MBE) is an "objective" multiple-choice examination constructed by the National Conference of Bar Examiners. It was first given in South Carolina in February 1972
Appellants argue on appeal that the district court erred in refusing to certify this as a class action. While we believe that Judge Blatt might have properly done so (see, e. g., Doe v. Charleston Area Medical Center, Inc., 529 F.2d 638, 645 (4th Cir. 1975)), given our disposition of this case, we find the error, if any, to be harmless
Because of the pendency of this appeal, the South Carolina Supreme Court has stayed action on appellants' petition for admission to the Bar and review of their examination scores
See also Comment, Review of Failing Bar Examinations: Does Reexamination Satisfy Due Process?, 52 Bos.U.L.Rev. 286, 300-01 & n.114-15 (1972)
It is significant, we think, that after initiation of this suit the State voluntarily established procedures for review of failing papers. Furthermore, the administrative burden argument has not been made by the State in its brief
In both Tyler and Whitfield, the state bar under challenge allowed the applicant unlimited opportunity to retake the bar. In South Carolina at the time of the district court's opinion in this case, an applicant was allowed only three opportunities to take the examination