INGRAHAM, Circuit Judge:
Nine blacks (appellants) filed a civil rights action against Union Carbide Corporation, charging it with various discriminatory employment practices perpetrated at the Chickasaw, Alabama plant. On appellants’ motion the district court certified the case as a class action filed in behalf of:
Those black job applicants who had an outstanding job application at the time of the commencement of this cause and those black job applicants who have filed job applications since the commencement of this cause and all future black job applicants, along with all current black employees o'f defendant Union Carbide Corporation, Materials System Division.
Additionally, the district court included a provision in its Notice of Pendency of Class Action which required potential class members to “opt-in” to preserve their back pay claims.
After considering all the evidence, the district court concluded that Union Carbide’s hiring and promotional practices were not discriminatory and did not violate Title VII of the Civil Rights Act.
On appeal appellants claim that the district court’s ruling was clearly erroneous and that the order requiring class members to opt-in to obtain back pay is contrary to Rule 23 of the Federal Rules of Civil Procedure and Title VII of the Civil Rights Act.
TITLE VII CLAIMS
1.
Hiring Practices
In 1965 Union Carbide opened a plant in Mobile County, Alabama for the manufacture of “molecular sieves,” a product that strains and filters liquid or gaseous molecules. The work force at the plant is composed primarily of residents in the vicinity of Mobile, Alabama, an area which is approximately 26% black. When the plant began operations, the work force consisted of 35 whites and 1 black. Nevertheless, within nine years the plant hired 84 black
and 259 white employees. Furthermore, blacks serve in both salaried and wage-earning positions at the Union Carbide plant.
Union Carbide continually accepts employment applications irrespective of the existence of current vacancies. The personnel files keep a record of applicants; whenever a position becomes available, the applications from the “active” file are reviewed for employment consideration.
If the files contain an insufficient number of applications, Union Carbide places advertisements in the local newspaper. Then the plant conducts short screening interviews to determine the applicants’ basic vocational interests and to answer any questions relating to the job openings. Following the initial interview, selected applicants are requested to return for an in-depth interview with three or four other plant employees— usually two engineers and one foreman. The applicant also tours the plant to view the working conditions. According to Mr. James W. Garner, manager of employee relations, the plant prefers applicants who are “more likely to be long term Carbide employees [and] who want the kind of work and opportunities that [the plant has] to offer.” Three members of the personnel department discuss the qualifications of each applicant and decide which of the ap
plicants are most qualified for the particular job opening. An offer is then extended to the candidates deemed to be most qualified.
Appellants claim that the Union Carbide hiring practices were discriminatory and violative of the Civil Rights Act, Title VII. Nevertheless, the employment practices of Union Carbide appear to be fair in form — that is, reasonably directed to secure the best qualified candidate for the position available. Although perhaps fair in form, the hiring and promotional practices must also be fair in operation. In
Rowe v. General Motors Corp.,
457 F.2d 348, 355 (5th Cir. 1972), Chief Judge Brown stated:
“It is clearly not enough under Title VII that the procedures utilized by employers are fair in form. These procedures must in fact be fair in operation. Likewise, the intent of employers who utilize such discriminatory procedures is not controlling since ‘Congress directed the thrust of the Act to the
consequences
of employment practices, not simply the motivation.’
Griggs v. Duke Power Co., supra,
401 U.S. [424] at 432, 91 S.Ct. [849] at 854, 28 L.Ed.2d [158] at 165.
“It is therefore clear that employment practices which operate to discriminate against people because of their race, col- or, religion, sex or national origin, violate Title VII, even though the practices are fair on their face and even though the employer had no subjective intention to discriminate.”
An examination of the consequences of Union Carbide’s employment practices is thus critical to a disposition of this case.
The litigants proposed separate methods to analyze the hiring data. The appellants suggested that the total number of applications filed by black persons is the critical factor to be considered. Their statistical analysis showed that blacks filed approximately 50% of all applications but secured only 26% of new hourly positions:
At least one witness, however, cast suspicion on the reliability of appellants’ approach, comparing it to “trying to measure jellyfish with a rubber band [because] you don’t know where all the applicants come from . . . [and] how many are duplicated. [S]ome applicants fill six applications a year out. We have even had three or four in one month from a particular applicant.”
Union Carbide, offering an alternative method of analysis, urged that the employment record should be proportionate to the composition of the local work force. Since 1970 approximately 33% of all employees hired by Union Carbide have been black. Operating its plant in an area where the labor force is approximately 26% black, Union Carbide presents convincing statistics tending to negate the existence of a discriminatory effect of its hiring practices and procedures:
Adopting Union Carbide’s method of analysis, the district court concluded that the employer “ . . . has truly equal access to jobs at its Chickasaw plant by blacks and whites and the Court finds that this has in fact produced an employee population with substantially the same proportion of blacks represented as are in the Mobile area work force.” The district court’s recognition of the employer’s method of analysis as more reliable and indicative of racial discrimination is supported by testimony that appellants’ alternative analysis is subject to suspect data, such as the duplication of employment applications. Additionally, the trial court’s finding that Union Carbide’s method is more reliable is supported by the practice of other federal courts. On many occasions federal courts have compared the composition of the company’s work force to the composition of the labor force in the surrounding area.
E. g., Jones v. Tri-County Elec. Cooperative, Inc.,
512 F.2d 1 (5th Cir. 1975);
Morrow v. Crisler,
479 F.2d 960 (5th Cir. 1973),
rehearing,
491 F.2d 1053,
cert. denied,
419 U.S. 895, 95 S.Ct. 173, 42 L.Ed.2d 139 (1974);
NAACP v. Beecher,
371 F.Supp. 507, 515 (D.Mass.Mod.)
aff’d
504 F.2d 1017 (1st Cir. 1974),
cert. denied,
421 U.S. 910, 95 S.Ct. 1561, 43 L.Ed.2d 775;
Crockett v. Green,
388 F.Supp. 912, 917 (E.D.Wis.1975);
Fowler v. Schwartzwalder,
351 F.Supp. 721 (D.Minn. 1972).
Under the standard of review in this circuit there must be “requisite subsidiary facts to undergird the ultimate facts.”
Causey v. Ford Motor Co.,
516 F.2d 416, 420-21 (5th Cir. 1975). We conclude that the record presents sufficient underpinnings to support the ultimate finding that Union Carbide did not engage in discriminatory hiring practices.
2.
Promotional Practices
Even an employer who exhibits nondiscriminatory hiring practices can fall short in its system of promotion,
Rowe v. General Motors Corp.,
457 F.2d 348, 355-59 (5th Cir. 1972) (Brown, C. J.), and that appears to be the situation in the instant case. Although the employer’s affirmative action hiring program has resulted in the employment of blacks in close proportion to the composition of the area labor force, the system of promotion has failed to provide equal access to the most preferred positions at the Chickasaw plant. For this reason we REVERSE and REMAND on the issue of Union Carbide’s promotional practices.
(a)
System of promotion.
The plant uses a seniority system to promote employees within the rank of “hourly personnel.” When there is an hourly position available, current employees can bid for the job. The senior employee satisfying the minimum qualifications for the position will secure the job. According to Mr. Garner, “[the plant] can’t hire anybody outside the plant who makes a high test score or promote a person to that job who makes a higher test score . . . . [W]e promote the most senior qualified employees.” Job candidates demonstrate that they are
qualified
by passing the examinations prepared and administered by the Employee Relations Manager and a plant foreman or supervisor.
The plant provides training programs designed to prepare employees for the exams, the passing of which is essential to secure a better hourly wage position. Thus, assuming the requisite seniority, completion of a training program and passing the company’s exam are the keys to advancement within the hourly ranks.
In the case of “non-exempt salaried personnel”
the system of promotion is somewhat different. The personnel department posts a notice of the particular job opening and accepts applications for the position. After the initial screening process, applicants are referred to the department supervisor under whose direction the selected applicant would work. The ultimate employment decision is made by the responsible department supervisor, with personnel’s concurrence.
A third category, “exempt salaried employees,”
is not derived from job posting or seniority bidding, but is the product of Carbide recruitment. Chemical engineers and other technical and supervisory personnel, for example, are exempt salaried employees. To fill the engineering positions Carbide representatives approach the colleges and universities to conduct job interviews. When the plant hires new foremen, the superintendents and supervisors evaluate all employees in the particular division and select the most qualified individuals for advancement from the hourly wage to salaried positions. The ultimate decision, however, is not based on uniform procedures or objective standards. Instead, the selection of one candidate over another depends on highly subjective criteria which shift in importance from case to case.
(b)
Statistical evidence of racial discrimination.
As mentioned earlier, the original work force at the Chickasaw plant was composed of 35 whites and 1 black. Six of the 35 whites are presently employed by Union Carbide: one is plant manager, one is an area supervisor, two are first line supervisors, and two are group leaders. Only the plant manager began his employment in that capacity; the other five white Carbide employees began in low priority wage-earning positions but were promoted through successive promotions. In contract, the first black employee began as an oiler and is currently a warehouseman; both jobs are low priority wage-earning positions.
Blacks continue to dominate the ranks of menial employment at Union Carbide’s
plant in grossly disproportionate numbers. For example, 40% of the oiler and helper-trade employees are black; blacks constitute only 19.2% of the total work force. Only 7.1% of the plant’s maintenance department is black, exclusive of members of the oiler/helper trades.
The following charts demonstrate that blacks occupy a severely disproportionate number of the low-priority, non-salaried positions:
A substantial disparity between the proportion of blacks in a specific job classification is sufficient to establish a prima facie case of employment discrimination.
E. g., Sagers v. Yellow Freight System, Inc.,
529 F.2d 721, 729-30 (5th Cir. 1976);
Wade v. Missis
sippi Cooperative Extension Service,
528 F.2d 508, 516-18 (5th Cir. 1976);
United States v. T. I. M. E.
— D.
C., Inc.,
517 F.2d 299, 311-14 (5th Cir. 1975);
cert. granted,
- U.S. -, 96 S.Ct. 2200, 48 L.Ed.2d 814 (1976);
Rowe v. General Motors Corp.,
457 F.2d 348, 358 (5th Cir. 1972). The contrast between the black population of the Mobile area (approximately 28%) and the percentage of blacks in supervisory or salaried positions (below 10%) constitutes a substantial statistical discrepancy and, thus, establishes a prima facie case of unlawful racial discrimination. After the prima facie case is established, the burden of persuasion shifts to the corporation whereby it must show the statistical discrepancy results from causes other than racial discrimination.
Sagers v. Yellow Freight System, Inc., supra; Wade v. Mississippi Cooperative Extension Service, supra.
(c)
Response to a prima facie case of discrimination.
Union Carbide contends that the employment statistics are the result of a nondiscriminatory system of promotion. For promotion within the hourly wage category, the employer relies on ostensibly nondiscriminatory data — the candidates’ seniority and the successful completion of comprehensive tests designed and administered by the employer. In this employment category, we conclude that the district court’s no-discrimination finding has inadequate factual underpinning in the record and should be REVERSED.
Union Carbide failed to establish that the qualification exams were “job related;” that is, there must be a correlation between test scores and actual job performance.
Crockett v. Green,
388 F.Supp. 912, 919-20 (E.D.Wis.1975). Additionally, the employer failed to show that the admission standards for its training program were objective and nondiscriminatory. Because the training programs are a springboard to occupational advancement, the admission standards are clearly relevant to an analysis of the plant’s promotional practices.
Promotion within the ranks of salaried personnel is based upon the supervisors’ subjective evaluation of a job candidate, without definite qualification factors being considered in a fixed ratio. Mr. Garner described the basic process for promoting an hourly wage earner to a salaried supervisory position:
Q Mr. Garner, would you please describe the process of selecting first line maintenance and production supervisors?
A When there is a vacancy, an increased need for a first line supervisor in production, the positions are first approved by top management, who in this case would be the Plant Manager . . . . [Then the responsible Production Superintendent requests] his Area Supervisor to evaluate our employees for this position.
And the Area Foremen send out memos, or by any means, ask Foremen and Employee Relations Departments, Superintendents and general management to recommend anyone they feel would be qualified for the position. Evaluations are made . . . [by] anyone who had the opportunity to evaluate their performance in a Temporary Foreman’s job or in a work capacity that they were being considered for.
These evaluations and reports are turned in to the Production Superintendent and in group meetings, we discuss applicants — applicants meaning from within the company at this time, we used to go outside — and a group decision is reached on who would be the best candidate for the position. At times, depending on the circumstances, we might choose to have a period of time of several months to try out Temporary Foremen who have not tried out before; or it might be the case that we have already have done this as the year went along and that might not be necessary.
But, after the
compilation of all of the results of the questionnaires or comments and evaluations,
then a group decision is made of who is going to be offered that job. That person is then interviewed, it could be by more than one person, by four managers — myself, the Plant Manager, the Engineering and Construction Manager and Mr. Portzer, our Production Manager. And if all agree that that is
the right person for the right job at that time, an offer is made to that individual.
Q Then the final decision is made by no one particular person?
A That is correct.
The questionnaires and evaluation forms used by Union Carbide require the interviewer’s subjective opinion concerning the candidates’ “adaptability,” “bearing, demeanor, manner,” “verbal expression,” “appearance,” “maturity,” “drive,” and “social behavior.” Such high-level subjectivity subjects the ultimate promotion decision to the intolerable occurrence of conscious or unconscious prejudice.
Rowe v. General Motors Corp., supra,
at 358-59;
Wade v. Mississippi Cooperative Extension Service, supra,
at 517-18.
(d)
Conclusion.
In
Wade,
this court noted that the employer’s employee evaluation form was constitutionally defective under
Griggs
v.
Duke Power Co.,
401 U.S. 424, 91 S.Ct. 849, 28 L.Ed.2d 158 (1971) in three respects:
(1) “the questions on the evaluation form were in part subjective and vulnerable to either conscious or unconscious discrimination by the evaluating supervisors;”
(2) “the evaluation scores themselves were not consistently used as a basis for promotion;” and,
(3) “the defendants wholly failed to make a showing that the test was substantially related to the particular job of the individual being evaluated.” 528 F.2d 508, 518.
Similarly, in'
Rowe
this court suggested several other conditions reflecting Title VII violations:
(1) “[t]he foreman’s recommendation is the indispensable single most important factor in the promotion process, [but he is] given no written instructions pertaining to the qualifications necessary for promotion;”
(2) “standards which were determined to be controlling are vague and subjective;”
(3) “[h]ourly employees are not notified of promotion opportunities nor are they notified of the qualifications necessary to get the job;” and,
(4) “there are no safeguards in the procedure designed to avert discriminatory practices.” 457 F.2d 348, 358-59.
In regard to Union Carbide’s promotion practices in the ranks of salaried personnel, we also conclude that the trial court’s no discrimination finding lacked adequate factual support in the record. Therefore, under our standard of review, we REVERSE and REMAND the district court’s findings and conclusions regarding Union Carbide’s system of promoting employees in the ranks of the hourly wage and salaried personnel.
Causey v. Ford Motor Co.,
516 F.2d 416, 420-21 (5th Cir. 1975). The employer clearly failed to rebut the statistics or to explain the disparity in promotion.
CLASS ACTION CLAIM
Appellants argue that the trial court’s “opt-in” notice violates the “opt-out” notice provisions of Rule 23(c)(2).
Indeed, Rule 23(c)(2) provides for mandatory notice to (b)(3)-type
class members whereby each has a right to exclude himself, or opt out of
the class. Concerning this opt out option under Rule 23(c)(2), one commentator pointed out that the provision was patterned after the highly successful procedure of the Book-of-the-Month Club. Frankel, “Some Preliminary Observations Concerning Civil Rule 23,” 43 F.R.D. 39, 44 (1967). The rule was designed to prevent “sideline sitting” and subsequent “one-way intervention” by the eligible class member.
Compare, Escott v. Barchris Const. Corp.,
340 F.2d 731, 735-36 (2nd Cir. 1965) (Friendly, J., concurring).
Appellants’ claim essentially contests the second portion of the district court’s “Notice of Pendency of Class Action” requiring class members to opt in to obtain back pay.
The supplemental notice to class members that do not opt out under the first portion of the Notice is permissible as requiring “some affirmative action as a condition of ultimate recovery.” 3B Moore’s Federal Practice 123.55 at 23-1161 (1975);
Iowa v. Union Asphalt & Roadoils, Inc.,
281 F.Supp. 391, 404 (S.D.Iowa 1968);
Philadelphia Elec. Co. v. Anaconda Am. Brass Co.,
43 F.R.D. 452, 459 (E.D.Pa.1968);
Harris v. Jones,
41 F.R.D. 70, 74-75 (D.Utah 1966).
Thus, assuming the sufficiency of the first section of the Notice, the district court did not err by requiring the affirmative action in the second section as a condition of ultimate recovery.
We AFFIRM this part of the district court’s judgment.
CONCLUSION
We AFFIRM the district court on the class action and hiring practices issues; REVERSE and REMAND on the no discrimination findings as to Union Carbide’s system of promotion.