Dollie W. HESTER, Plaintiff-Appellee, v. SOUTHERN RAILWAY COMPANY, Defendant-Appellant

SIMPSON, Circuit Judge: Under this appeal we review a district court order based upon a finding that the appellant Southern Railway (“Southern” or “Railway”) discriminated against blacks in hiring for the Data Typist/1050 position at Southern’s Atlanta general office. The suit was brought as a class action under Title YII of the Civil Rights Act of 1964, 42 U.S.C. Sec. 2000e et seq. (1970). The district court order enjoined the use of certain tests and interviews employed by Southern in selecting its Data Typists. The court also awarded the individual plaintiff differential back pay and attorneys’ fees and ruled that the plaintiff should be awarded seniority from the date of her original application should she subsequently apply for and be hired by Southern for the Data Typist position. D.C., 349 F.Supp. 812. The origin of the law suit was a classified advertisement placed by Southern in the Atlanta Constitution, an Atlanta, Georgia daily newspaper. The want ad sought applicants for the Data Typist position who “should be able to type 60 wpm [words per minute] and be willing to work night shift and weekends.” It continued that Southern was an “Equal Opportunity Employer.” Appellee resonded to an advertisement which ran from June 30 to July 12, 1969. Ms. Hester was one of forty individuals who responded. She went to the Southern Railway office and completed an application form, and was then required to take an SRA (Science Research Associates) typing test in order to demonstrate her ability to type a minimum of sixty words per minute. In addition, she was required to take SRA verbal and non-verbal tests, and the J. P. Cleaver self-description test. She passed the typing test on her second attempt. According to testimony, applicants were not given pass/fail grades on the three other tests. Rather, these tests were used as informational inputs in evaluation of the applicants. After finishing the testing phase of the application process, Ms. Hester was interviewed by Southern Personnel Officer James Melton. She was later notified by mail that Southern had no position for her at that time. About two weeks later, Ms. Hester noticed the same want ad for typists in the Atlanta newspaper. She called Mr. Melton to inquire about the availability of a position at that time. Mr. Melton responded that she did “not fit into the picture.” Ms. Hester and her husband tried to see Melton at his office, but he did not respond to messages to call her. Ms. Hester at about this time filed a complaint with the Equal Employment Opportunity Commission (EEOC) alleging discrimination against her by Southern on the basis of race. An EEOC investigator examined Southern’s record of those hired for the Data Typist/1050 position 1 between June and October 1969 and also of those “in training” for the position as of October 2, 1969. The investigator concluded that Southern was discriminating, not against blacks generally, but against black women with children. 2 Southern declined an EEOC offer to conciliate on Ms. Hester’s charge of discrimination, asserting *1376 that it was not practicing any discrimination. The EEOC notified Ms. Hester by letter of her right to sue in federal court within 30 days, and she timely filed her suit in the court below in August, 1970. I The complaint stated that Southern had violated the provisions of Title VII of the Civil Rights Act of 1964, Title 42 U.S.C. Sec. 2000e et seq., by refusing, on grounds of race and sex, to hire the plaintiff for the position of Data Typist/1050 operator. Plaintiff was alleged to represent a class composed of black women with children who had not been hired by the defendant because of sex and race. Southern denied engaging in racially discriminatory hiring and submitted affidavits of four black women with children at the time that they were hired by Southern and who were working as Data typist/1050 operators at the time of Ms. Hester’s application. Southern also presented employment figures showing that, for the period January 1, 1969, through June 1, 1971, 20.5% of all the data typists hired were non-white and of these, 50% had children at the time they were employed. These affidavits and figures were the basis for Southern’s Motion for Summary Judgment, which the district court denied. The case was set for non-jury trial. Evidence was taken on April 10-11, 1972. The proof submitted was primarily: (i) testimony of Southern Personnel Officer Melton as to his reasons for not offering Ms. Hester employment at the time of her application; 3 (ii) statis-' *1377 tics, introduced in the main by Southern, showing that its hirees for the Data Typist position for the years 1968-69 were 24-25% non-white. Melton also testified without producing actual figures that about 35% to 40% of the persons responding to the newspaper ad for Data Typists were black. The court ruled the plaintiff had failed to make an affirmative showing during trial that the requirements of Rule 23, F.R.Civ.P. were met and held that the case was not a valid class action, citing Oatis v. Crown-Zellerbach Corporation, 5 Cir. 1968, 398 F.2d 496. The judge then evaluated the evidence in the light of the plaintiff’s claim of discrimination in hiring. Respecting the plaintiff’s claims of (discrimination based on sex or on race éombined with status as parent, the court failed to find any discrimination demonstrated: It is clear from the evidence that defendant has not discriminated against black females for the Data Typist position on the basis of. sex. Practically all, if not all' Data Typists hired by defendant have been females. It is also clear that defendant has not discriminated against black females who have children as against black females without children or white females with or without children. The evidence shows that for the years 1968-1969 12 of the 23 (52%) female blacks hired as Data Typists had children while 31 of the 70 female whites hired (44%) had children. This evidence utterly refutes plaintiff’s claim that defendant does not hire black females with children for the position of Data Typist. 349 F.Supp. at 816-817. Proceeding then to the question of discrimination based on race, the court found: (i) that the testing procedure employed by Southern was not validated (i. e., proven to be job-related and non-racially biased) at the time of Ms. Hester’s application; and (ii) the interviewing procedure employed by Mr. Melton was “purely subjective” and based upon “no formal guidelines, standards, or instructions . . . .” Id. at 817. The question at this stage was whether over all the hiring procedure employed by the Railway was discriminatory in its operation. The court in its summary of the evidence stated: While it is apparent that defendant employed Data Typists in Atlanta from the end of 1964 through 1967, no evidence was introduced as to their exact racial or sexual composition. The only remotely relevant evidence introduced by plaintiff concerning this period establishes that as of February 1966 defendant reported to the Equal Employment Opportunity Commission that it employed as “office and clerical” help 248 white males, 60 white females, 11 black males, and 1 black female. Although it was not shown whether Data Typists were considered “office and clerical” employees by defendant at that time, defendant reported no other non-laborer female employees at that time, black or white, and only 27 other non-laborer and non-managerial male employees, all white. The court deduces from this evidence that as of February 1966 defendant employed quite a few white Data Typists but few, if any, black Data Typists at the Atlanta office. During the year 1968 defendant hired 51 females as Data Typists and Data Typists Trainees in the Atlanta office. Of this number 38 were white and 13 (25%) were black. Eighteen of the whites and nine of the blacks hired had children at the time of their employment. From January 1, 1969 through August 31, 1969, defendant hired 22 Data Typists and Data Typist Trainees in the Atlanta office, 21 of whom were females. Of the females, 17 were white and 4 (19%) were black. There was no evidence as to how fnany of these females had children. During the entire year of 1969 defendant hired 42 females as Data Typists and Data Typist Trainees in the Atlanta office. Of this number 32 were white and 10 (24%) were black. *1378 Thirteen of the whites and three of the blacks hired had children at the time of their employment. Plaintiff introduced no evidence as to the total number of people who applied for the position of Data Typist at the Atlanta office in 1968 or 1969, nor did she introduce any evidence as to the total male: female or black: white ratio of the applicants. There was some testimony, however, that in 1969 approximately 35%-40% of those who responded to defendant’s advertisements for the Data Typist position were black. As of October 2, 1969, defendant employed a total of 110 Data Typists and Data Typist Trainees. Of this number, 92 were white and 18 (17%) were black. According to the 1970 Census of Population, blacks constitute over 50% of the population of Atlanta, and black females constitute over 50% of the female population of Atlanta. 349 F.Supp. at 813-814. Upon this showing, the court found the following facts and reached its conclusion of law: To be sure, the statistical evidence adduced by plaintiff is weak and in some respects inconclusive. Nevertheless, the court found that in the early years of the Data Typist position defendant employed few, if any, blacks. As of October, 1969, around the time plaintiff applied, only 17% of the Data Typists were black. During 1968 and 1969, when it appears whites were applying in only slightly larger numbers than blacks for the Data Typist position, three times as many whites as blacks made it through the testing and interviewing process and were hired. In view of this disparity and the holdings in Griggs [401 U.S. 424, 91 S.Ct. 849, 28 L.Ed.2d 158], Jacksonville Terminal [5 Cir., 451 F.2d 418], and Rowe [5 Cir., 457 F.2d 348], the court concludes that the procedure used by defendant to hire Data Typists at its Atlanta office constitutes an unlawful employment practice within the meaning of Title VII. Id. at 817-818. The order entered below, based upon the foregoing: (i) enjoined the further use by Southern of the non-validated tests and the single discretionary interview procedure; (ii) awarded differential back pay to Ms. Hester from the date of the alleged discriminatory refusal to hire until the date of the filing of the complaint in the court below; (iii) awarded costs and attorneys’ fees to the plaintiff; 4 and (iv) required the defendant to accord Ms. Hester seniority from the date of the rejected application in the event she applied for and was hired for the Data Typist position. II On appeal, Southern challenges the heart of the district court opinion, contending that the court’s finding that the Railway’s hiring procedure — both the testing and the interviewing stages— discriminated against blacks is clearly erroneous for want of an evidentiary basis. Rule 52(a) F.R.Civ.P. Without an underlying finding of discrimination, Southern says that the court lacked power to enjoin the further use of this hiring procedure. Southern argues further that whether or not the tests were discriminatory was not before the court below since the evidence showed the plaintiff passed the typing and other tests and therefore had no standing to challenge the tests. Both parties challenge the district court’s award of differential back pay and attorneys' fees. 5 *1379 We deal first with the sufficiency of the evidence to support the finding below of discrimination in Southern’s hiring procedure, treating the finding as presumptively correct and subject to overturn only upon a showing that it is “clearly erroneous”. Rule 52(a), F.R.Civ.P.; McAllister v. United States, 1954, 348 U.S. 19, 75 S.Ct. 6, 99 L.Ed. 20; Baxter v. Savannah Sugar Refining Co., 5 Cir. 1974, 495 F.2d 437 [decided June 6, 1974]; Pettway v. American Cast Iron Pipe Co., 5 Cir. 1974, 494 F.2d 211, 234 n. 56. The paths to establishment of a prima facie case of racial discrimination are several. In some instances, a showing of disparity between the percentage of minority group members in the local population and the percentage in a particular job may be of significance. Cf. Chance v. Board of Examiners, S.D.N.Y.1971, 330 F.Supp. 203, 214, aff’d 2 Cir. 1972, 458 F.2d 1167; Penn v. Stumph, N.D.Cal.1970, 308 F.Supp. 1238, 1243 nn. 7 & 8. There is no competent evidence of such a disparity in this ease. 6 In other situations, a showing of a total absence of blacks from a particular job position, both currently and as a matter of historical record, may be significant as an indication of racial discrimination. Franks v. Bowman Transportation Co., 5 Cir. 1974, 495 F.2d 398, at p. 419 [1974], This again is not the case before us. 7 The most direct route to proof of racial discrimination in hiring is proof of disparity between the percentage of blacks among those applying for a particular position and the percentage of blacks among those hired for the position. From such statistics, the district court could have proceeded to consider evidence as to the numbers of whites and blacks taking and passing the typing and other SRA tests administered here. Griggs v. Duke Power Co., 1971, 401 U.S. 424, 91 S.Ct. 849, 28 L.Ed.2d 158 suggests such a test of employment testing. The initial inquiry that a court must make in evaluating employment testing is whether the tests “operate to disqualify Negroes at a substantially higher rate than white[s] . . . ” *1380 Id. at 426, 91 S.Ct. at 851, 28 L.Ed.2d at 161. Accord: Johnson v. Goodyear Tire & Rubber Co., 5 Cir. 1974, 491 F.2d 1364, 1372 n. 22; United States v. Georgia Power Co., 5 Cir. 1973, 474 F.2d 906, 911. Upon a presentation of additional evidence a similar inquiry could have been made regarding the percentage by race of those passing the typing test but being denied employment on the basis of the personal interview. The district court in fact sought the additional evidence making such a determination possible. Less than two months after the hearing in this case, and several months before the opinion-order came down, the court notified counsel of a post-trial hearing through which he sought answers to the following questions, inter alia: (1) With respect to the years 1966-1972: (a) how many blacks and how many whites applied for the position in question? * ' * -X- -X- -X- * (4) With respect to the years 1966-1972: (a) how many whites and how many blacks who applied for the position in question took the “J. P. Cleaver” (sic) typing test? * -x- * * * * (d) how many whites and how many blacks who took the typing test achieved a score of “60 corrected” (words per minute) or better? * -X- -X- * * * (g) how many whites and how many blacks who scored “60 corrected” or better were interviewed for the position; . . . ? (h) how many whites and how many blacks who scored “60 corrected” or better and were interviewed were hired for the position; . . . ? (i) how many whites and how many blacks who scored below “60 corrected” (even after retesting) were interviewed for the position; . . . ? (j) how many whites and how many blacks who scored below “60 corrected” were hired for the position ; . . . ? The record does not show whether this hearing ever took place. If it did, the information sought by the interrogatories is not in the record before us. The evidence in the record as to the percentage of blacks in the applicant pool was: (i) the estimate of Southern Personnel Officer Melton that blacks constituted probably 35%-40% of the respondents to Southern’s want ad; (ii) records indicating that, of the 40 persons answering the ad to which Ms. Hester responded, 21 were black and 19 were white. This evidence fails to support the district court’s finding that “[d]uring 1968 and 1969, whites were applying in only slightly larger numbers than blacks for the Data Typist position.” 349 F.Supp. at 817. Melton’s testimony is defective both as to relevance and as to probative value. Regarding relevance, Melton testified that blacks constituted approximately 35%-40% of those responding to the newspaper ads placed by Southern in the Atlanta newspaper. Other testimony indicated, however, that a substantial number of those responding to the newspaper ads failed to advance to the testing and interviewing stages. The testimony was that some applicants “deselected” themselves before submitting to the selection process. This evidence is too conjectural and inconclusive to permit any conclusions as to the incidence or rate of deselection. But this testimony does bring into question the relevance of the percentage of blacks responding to the newspaper ads, since the percentage actually applying for the Data Typist position might have been considerably different. The value of Melton’s testimony is seriously undermined by its subjectiveness and .lack of precision illustrated by the exchange quoted in Note 3, supra. Such “gut feel” testimony is not competent to support any finding respecting the racial composition of the applicant pool. *1381 The insufficiency of Melton’s testimony to sustain the district court’s finding is underscored, we think, by the avenues open to the plaintiff of utilization of the liberal discovery techniques under Rules 26-37 of the F.R.Civ.P. Assuming the existence of the evidence required to establish a prima facie case, means of access to it and its perpetuation were readily available. 8 In these circumstances, lack of diligence in assembling statistical evidence in support of her case presents us with serious problems. We recognize that statistics are a powerful tool in the hands of a Title VII plaintiff, but we are also aware that undue emphasis on their use may obscure 9 rather than advance the judicial process. Likewise, the evidence concerning the racial composition of the pool of applicants responding along with the plaintiff to the ad of June 30 — July 12, 1969, does not meet the standards of Rule 52(a). Evidence that blacks constituted 50% of one group of respondents to an ad run periodically throughout the year is no logical support for a finding that blacks constituted a like percentage over an entire one or two year period. Such extravagent extrapolation is not permissible. The district judge’s finding that blacks and whites applied in equal numbers for the Data Typist position during 1968-69, is rejected for lack of a sufficient evidentiary foundation. It is “clearly erroneous”, Rule 52(a), F.R.Civ.P. Ill Without factual underpinning, the lower court’s finding of discrimination in hiring is erroneous as a matter of law. The conclusory finding of discrimination is among the class of ultimate facts dealt with as conclusions of law and subject to review outside the constrictions of Rule 52(a). United States v. Winthrop, 5 Cir. 1969, 417 F.2d 905, 910, and cases cited therein. We come now to the question of whether the relief ordered by the district court was appropriate without competent proof of discrimination by the Railway in its hiring of Data Typists. At the time that the district court held its hearing the SRA and J. P. Cleaver tests utilized by Southern had not been validated for job-relatedness. Likewise, the interview procedure comes uncomfortably close to the promotion procedure proscribed by Rowe v. General Motors Corp., 5 Cir. 1972, 457 F.2d 348, 358-359. Nonetheless, nonvalidated tests and subjective hiring procedures are not violative of Title VII per se. Title VII comes into play only when such practices result in discrimination. At that point, the burden of producing evidence shifts to the employer, who must offer satisfactory justification for his procedures. As stated in United States v. H. K. Porter Co., N.D.Ala.1968, 296 F.Supp. 40, 76-77: The courts must decide the cases which come before them on the evidence and not on abstract propositions. For a court to find racial dis *1382 crimination in the use of aptitude tests which have not been validated, there should be- at least some evidence that the use of an aptitude test which has not been validated has resulted in discrimination and not merely the abstract proposition that test validation is desirable._ The missing ingredient in the proof here was the necessary showing of discrimination. Without such proof the district court lacked authority to enjoin the further use of the testing and interviewing procedures by Southern for selection of Data Typists. That injunction is vacated because of this deficiency. The court below is free to consider re-imposing it if the plaintiff by additional proof establishes a prima facie case of racial discrmination in the use of either the tests or the interviews, and Southern fails to meet its burden of justification. IV The lower court awarded the plaintiff-appellee differential back pay and attorneys’ fees. Successful litigants in Title VII suits are eligible for back pay awards under Title 42 U.S.C., Sec. 2000e-5(g). Franks v. Bowman Transp. Co., supra, 495 F.2d at 421; Johnson v. Georgia Highway Express, Inc., 5 Cir. 1969, 417 F.2d 1122, 1125. Similarly, the “prevailing party” in Title VII litigation is eligible for an award of reasonable attorneys’ fees under Title 42 U.S.C. Sec. 2000e-5(k). But in this case the award of back pay and attorneys’ fees falls with the failure to prove discrimination. See Banks v. Seaboard Coastline R. R. Co., N.D.Ga.1973, 360 F.Supp. 1372, 1375. The district court’s order and supplemental order awarding attorneys’ fees and back pay is vacated, against subject to re-imposition if the appellee succeeds after remand in proving discrimination in Southern’s hiring policies for the Data Typist/1050 position. Because discrimination was not proved, we also set aside the district court’s order that the plaintiff be awarded seniority from the date of her original application should she re-apply for and be hired for the Data Typist position. It is not clear to us on this record whether the plaintiff, armed with the necessary statistics, would or would not have succeeded in proving the discrimination claimed. It is therefore appropriate to remand for further proceedings in the district court. Noted swpra, is the attempt by the district court to obtain the statistics necessary to determine whether Southern has discriminated in the hiring of its Data Typists. This inquiry, apparently unsuccessful as it was, may have sufficed to end the matter. This silent record does not reveal why the additional information was not forthcoming. The district judge, in the exercise of his sound discretion, should decide what further course, if any, it is necessary for this litigation to take along lines not inconsistent with this opinion. Reversed in part, vacated in part, and remanded. Footnotes [1] The job description of the position applied for by Ms. Hester. [2] The sample showed that Southern had hired or had in training four married white women; all had children. Of the three'married black women hired or in training none had children. The plaintiff had listed one child on her application, and given her age as twenty-one. [3] Mr. Melton’s interview notes, entered on Ms. Hester’s application, read as follows: Twenty-one years old, married two years, one child, hometown Atlanta, also husband from Atlanta, works at Lockheed, has been there four years, high school education, very prim and proper air, parents here, mother seamstress, father works at Scripto, types all day now but didn’t seem to like it, has never worked nights. Husband — looks like worked days, might be a problem with that. Like her except for no night work experience. A big if. Reject for this reason. At the trial, Melton stated that these notes were not intended as a permanent record, but simply something to jog bis memory. Proceeding, he further explained his reasons for not offering Ms. Hester a job: “Judging from this I would say I didn’t feel that her overall background and work experience and general personality fit what I felt would be an acceptable pattern for a data typist — and I’m just guessing, but looking at her application, her work experience, though it has been in repetitive type, I have a statement here she didn’t seem to like typing all day, she lias never done any night work, never done anything in a production-type work environment, never worked week-ends, her husband would have worked the day shift and she would have worked nights, they’re young people, they have been married, well maybe they hadn’t been married a great while. There are a lot of things that I could conjectorize (sic) here. Her work experience seems to be a little bit more professional, a little bit more towards the secretary, the steno, a little, maybe a cut or two above in work responsibility than what the work of a data typist would be. Now, I’m just guessing but that would be, probably be, from looking at the application — I have no recollection of our conversation at all. Q And you took into consideration everything she put on the application? A Hopefully. Yes ma’am. Q And you also considered that she had no night work experiences? A Apparently. I’ve made a note with specific reference to that. I may have asked questions about that. I don’t know. Q Then, how many persons whom you interviewed had night work experience? A I’m not sure. Q Do you recall that many of them had night work experience? A No ma’am. I don’t recall that specifically. Q Sir, was the fact that Mrs. Hester didn’t indicate on the application that, in these blocks, where you asked whether or not she was willing to work any day of the week, overtime or any shift, did that have anything to do with your not hiring her? A It could have. It might not have. I don’t recall.” In view of this testimony the district court declined to “place much reliance on Melton’s handwritten notes”. [4] By a supplemental order entered December 12, 1972, the court below awarded the plaintiff Ms. Hester differential back pay in the amount of $2,500 and reasonable attorneys’ fees in the amount of $750. [5] The plaintiff filed a notice of appeal from the district court’s initial opinion and order, filed October 3, 1972, apparently intending to challenge the court’s denial of class action status for her suit. This cross appeal was later dismissed by this court for want of prosecution under 5th Circuit Local Rule 9(b). [6] Plaintiff was allowed to introduce as an exhibit a pamphlet published by the U.S. Department of Commerce entitled “1970 Census of Population — Georgia (No PC(V2)-12), which indicated that the population of the City of Atlanta in 1970 was approximately 51% black. Testimony by Southern officials, however, indicated that the applicant pool for the Data Typist position was composed of individuals from the greater Atlanta metropolitan area. The census report shows the population of the Atlanta Standard Metropolitan Statistical Area (SMSA) as approximately 22% black. Choice between these disparate figures would lead to widely differing conclusions in comparison with figures showing the percentage of blacks employed by Southern as Data Typists in 1970. Unfortunately the record contains only figures for 1969, when the Data Typist pool was 16.4% black and 1972 when the pool was 29.6% black. Meaningful comparison with the statistics for the general population is not possible on this record. Moreover, comparison with general population statistics is of questionable value when we are considering positions for which, as here, the general population is not presumptively qualified. Data Typist applicants were required to prove their ability to type at a minimum speed of sixty corrected words per minute as a prerequisite to consideration by Southern for employment. (The record indicates deviation by a word or two from this requirement in isolated instances). A more significant comparison might perhaps be between the percentage of blacks in the population consisting of those able to type 60 wpm or better and the percentage hired into the Data Typist position by Southern. Even this, however, would constitute no more than an indication that discriminatory practices may be in operation. As the text, infra, indicates, recourse would still have to be had to the statistics concerning the applicant pool and its racial composition before meaningful comparison with the percentage of blacks actually employed could be made. [7] The Data Typist position was created in 1964. Statistical evidence for years prior to 1969 was inconclusive as to the number of blacks or other minorities working as Data Typists. Blacks constituted 16.4% of the Data Typist force in 1969, according to the record. [8] Louis Sak, an officer of Southern testified that the Railway maintains its applicant records by the month for a period of two years from the date the application is filed. Thus, had plaintiff sought by use of deposition or interrogatories information from Southern’s records as to the percentages of blacks and whites in the applicant pool for the Data Typist position, it would have been feasible for her to construct a statistical portrait dating back to August of 1968 and covering the period of claimed discrimination. The continuing availability of this information is doubtful. [9] For an example of the manner in which the interpretation of statistics may mislead a court’s attempt to investigate the existence of discrimination, see Johnson v. Goodyear Tire & Rubber Co., 5 Cir. 1974, 491 F.2d 1364, 1372. An excellent discussion of the legitimate and logical force which proper statistics properly interpreted can have in an employment discrimination context appears in Chance v. Board of Examiners, 2 Cir. 1972, 458 F.2d 1167, 1171-1173.