Markey v. Tenneco Oil Co., 439 F. Supp. 219 (E.D. La. 1977)

U.S. District Court for the Eastern District of Louisiana - 439 F. Supp. 219 (E.D. La. 1977)
October 21, 1977

439 F. Supp. 219 (1977)

Dempsey J. MARKEY et al.

Civ. A. No. 75-1583.

United States District Court, E. D. Louisiana.

October 21, 1977.

*220 *221 Joseph W. Thomas, Okla Jones, II, New Orleans, La., for plaintiffs.

David E. Walker, Alvin J. Bordelon, Jr., Kenneth P. Carter, Monroe & Lemann, New Orleans, La., for defendant.


SEAR, District Judge:

This is an individual and class action brought under the Civil Rights Act of 1964, as amended, 42 U.S.C. § 2000e et seq. (better known and hereinafter referred to as Title VII) and 42 U.S.C. § 1981 for damages and injunctive relief. Plaintiffs allege that the Tenneco Oil Company refinery located in Chalmette, Louisiana discriminates *222 against blacks in its employment practices, more particularly in the areas of hiring, promotion, pay, and discharge.

On April 12, 1976, Edward J. Boyle, Sr., United States District Judge who was temporarily handling the case during the vacancy in this section of Court certified the case as a class action, the class being composed of:

"I. All black applicants for employment at the Tenneco Oil Company Refinery in Chalmette, Louisiana, from July 2, 1965 to the date of the final order of this Court;
II. All blacks who were, are and will be employees of Tenneco Refinery in Chalmette, Louisiana, from July 2, 1965 to the date of the final order of this Court."[1]

The trial of the case was conducted over a four day period beginning on March 7, 1977. At the conclusion of all the evidence, the matter was submitted, and both sides were given 30 days to submit suggested findings of fact and conclusions of law and other legal memoranda. This period was subsequently extended for an additional two weeks upon a joint motion of counsel.


In 1955 Tenneco Oil Company purchased an oil refinery located in the city of Chalmette in St. Bernard Parish, Louisiana. The refinery operates on a continuous basis, processing an average of 90,000 barrels of crude oil a day, and manufacturing a variety of petroleum products.

The refinery is organized into four separate departments. The Operations Department, the largest and most essential department, is charged with the operation of eight interrelated complexes, each composed of one or more process and refining units. The Maintenance Department provides labor, mechanical services and warehousing for the refinery at large and the Operations Department in particular. The Administrative Department provides personnel, accounting, fiscal, and clerical services. Lastly, the Technical Department furnishes engineering and other technical support to the Operations Department.

All nonsupervisory personnel in the Maintenance and Operations Departments, with the exception of nine Operations employees who work on the Tenneco wharf,[2] are represented by the Oil, Chemical and Atomic Workers International Union. Also included are ten nonsupervisory employees in the Laboratory, a division of the Technical Department. Since 1964 a series of collective bargaining agreements between the O.C.A.W. and Tenneco have governed all terms and conditions of employment of union members.[3]


The named plaintiff in this case, Dempsey J. Markey, alleges that he was discharged from his employment at the refinery because of his race. Markey was hired by Tenneco into an entry-level position as a laborer in the Maintenance Department on May 23, 1974. Pursuant to the collective bargaining agreement between the O.C.A.W and Tenneco, all new employees must successfully complete a 120 day probationary period before acquiring plant seniority with its concomitant rights and benefits. During this period the employee is "on trial". Towards the end of the period, the employee is rated by his supervisor, and may be discharged if his performance has been unsatisfactory. Markey admitted that he was aware that his status with Tenneco was probationary and that his job performance was being closely scrutinized for the purpose of the ultimate evaluation.

By Markey's own account he had worked in the Maintenance Department for two weeks to a month when he was temporarily transferred to the Laboratory to work vacation relief as a "lab sampler". On or about July 10, 1974 Markey drove a Cushman motor cart to collect samples from the tanks. He parked the cart next to the railroad tracks and climbed up to the tanks. Before he returned, a train came down the tracks, catching and dragging the Cushman cart. The cart was badly damaged. Markey contended that he had often parked the cart there without incident, and that he had observed other employees park the cart in the same spot. Nonetheless, he made no protest when a supervisory panel determined he had been negligent and suspended him from work for three days.

A second incident occurred on August 20, 1974 when Markey was asked to work overtime in order to help clean up a crude oil spill. Markey reported to the site of the spill at the end of his regular shift, but instead of beginning work, told the supervisor he was sick and went home. Markey saw no doctor, however, and reported for work on time the next day.

On August 21, 1974, supervisor Hudson filled out a rating sheet on Markey, evaluating his performance on thirteen different job functions on a scale from one to five. Markey scored no higher than a three on any of the job functions. Additionally, the supervisor placed Markey in the lower third of company employees overall, and did not recommend that Markey's employment with Tenneco be continued.

On September 8, 1974 Markey was summoned to the office of W. G. Pollard, where, in a meeting with Pollard and Paul Floth, the administrative manager of the Tenneco plant, Markey was told he was fired. Floth testified at trial that the decision to fire Markey was based on the rating sheet submitted by supervisor Hudson, and the two incidents which had occurred during Markey's probationary period.

On September 11, 1974 Markey filed a charge with the Equal Employment Opportunity Commission. The charge stated that Markey "believed" he had been discharged from Tenneco because of his race. Plaintiff made no other charge of discrimination, and indeed at trial testified that he had not been discriminated against in any other way in connection with his employment at Tenneco. The EEOC found in a Determination dated February 25, 1975, that "the evidence does not support the charge" and that there was "not reasonable cause to believe that Title VII of the Civil Rights Act ... has been violated in the manner alleged."

The Fifth Circuit has recently elaborated upon the respective burdens to be borne by both plaintiff and defendant when trying an individual claim under Title VII in Turner v. Texas Instruments, Inc.[4] There, the Court stated:

"[T]he plaintiff must present a prima facie case of racial discrimination; the employer then has the burden of proving, by a preponderance of the evidence, that legitimate, nondiscriminatory reasons existed to support his action; and the plaintiff *224 then has the burden of proving by a preponderance that the employer's articulated reason was a pretext for discrimination."[5]

It is questionable in this case whether Markey has even gone so far as to present a prima facie case of discrimination. A plaintiff's mere suspicion that he has been discriminated against, is not sufficient evidence to support an individual charge under Title VII.[6] Markey did not demonstrate that the position he vacated was filled by a white person. In fact, Paul Floth testified that Markey had been replaced by a black. Markey simply argued that he disagreed with the supervisor's ratings, and that he felt he had not been at fault in the two incidents which occurred during his tenure with Tenneco.

In regard to the first incident, Markey testified that in parking the Cushman cart close to the railroad tracks, he was doing as he had seen his fellow employees ("lab samplers") do many times. However, Markey did not demonstrate that this practice, later labelled by a supervisory panel as negligent, was officially condoned by Tenneco, or even that supervisory personnel were aware of the practice. Markey could not point to a single instance in which any other employee, black or white, had caused similar damage to company property. Moreover, at the time of the incident, although Markey was disciplined with a three day suspension, Markey made no protest that his treatment was unfair. Under these circumstances, and particularly because plaintiff could not isolate any instance in which a white employee was treated differently for a similar offense, I find that Tenneco had good cause to consider this incident against the plaintiff in determining not to retain him at the end of the probationary period. That such an incident would reflect poorly upon an employee's capabilities is a facially neutral result, and plaintiff has not shown by any evidence that the incident furnished a mere pretext for the exercise of race discrimination against him.[7]

As to the second incident when Markey refused to work overtime claiming that he was ill, I find that this also was properly considered against the plaintiff in evaluating his job performance. Markey had been told at the outset of his employment that he would be required on occasion to work overtime. On August 20, 1974 he was asked to do so. The task was to be an unpleasant, dirty onethe cleaning of an oil spill. Markey claimed that he was ill and could not do the work. It was the opinion of his supervisor that Markey was not ill, but rather that he merely wished to avoid the disagreeable assignment. This opinion was supported by certain objective facts: Markey worked his entire normal shift on the day he was asked to work overtime and reported on time for work the following day also working that entire day; Markey did not see a doctor for his illness. Under these circumstances I find that the assumption of the supervisor was a reasonable one, and was not influenced in any way by the race of the plaintiff.

The final element in the decision to discharge Markey was the poor rating he received from supervisor Hudson. Markey was rated on a scale of one to five on each of thirteen job functions, and scored four "ones" (unsatisfactory), six "twos" (passable), and three "threes" (usual). At trial Markey took issue with each of his ratings, arguing that he should have rated higher in *225 each category. Markey contends that the rating system is entirely subjective and that in his case it was used to discharge him because of his race.

While one of the categories on the rating form was rather amorphous[8], on the whole the functions listed were susceptible to fairly objective ascertainment by one who had observed the employee's job performance.[9] Given that the rating form is an objective one and is facially neutral, the issue remaining is whether it was discriminatorily applied in plaintiff's case, i. e., was plaintiff rated lower than he objectively should have been because of his race. While I have no way of knowing the accuracy of most of the evaluations, other than Markey's self-serving statement that he felt he should have rated higher in each one, in several categories I am able to assess the ratings based on testimony at trial. For example, Markey testified that during his brief tenure with the company he was late for work three or four times, and that he missed entirely approximately two days of work. Given this record, I find that a rating of "three" (usual) in the category of "punctuality and attendance record" was reasonable, indeed it may have been generous. Likewise, considering the two incidents in which Markey was involved, ratings of "one" (unsatisfactory) in the categories of "willingness to work on unpleasant jobs" and "observation of Company rules including Safety rules" were also reasonable. I find that plaintiff has not demonstrated that he was unfairly rated because of his race.

In sum, I find that the factors considered in the decision to terminate Markey's employment at Tenneco were legitimate, neutral criteria, and that plaintiff has failed to prove by a preponderance of the evidence, or even to make out a prima facie case, that he was discharged because of his race.

A. Effect of dismissal of named plaintiff on class claims.

Until very recently, the jurisprudence was well established that a plaintiff who fails on his individual claim may still pursue the class claims as class representative.[10] However, more recently in Satterwhite v. City of Greenville, Texas,[11] the Fifth Circuit decided that assuming that a class has been properly certified in the first place, if the claim of the named plaintiff is dismissed the class claims can be further pursued only if two requirements are met.

First, after dismissal of the individual claim, the class members must "retain sufficient interest in the outcome of [the] litigation to serve as Article III plaintiffs".[12] In other words, absent the claim of the named plaintiff, does the suit present a justiciable dispute or "case or controversy" for decision? This question can be answered by a demonstration that the class, or at least certain of its members, continue to have a personal stake in the outcome of the litigation. Satterwhite suggests that one method of proving this is to show that on the record there are identifiable individuals who nurse a grievance against the defendant or who have in fact sought relief against the defendant's practices in the past. The Supreme Court's decision in *226 Franks v. Bowman Transportation Co.,[13] one of the cases relied upon in Satterwhite,[14] indicated that the court must consider whether an adversary relationship between the class and the defendant survives once the named plaintiff is dismissed. Is there any question "concerning the continuing desire of any class members for the ... relief presently in issue?"[15]

Turning to the facts of this case, I find that a "case or controversy" does survive the dismissal of plaintiff Markey's claim. The exhibits in this case are filled with names of specific class members who, should the plaintiffs prevail on the statistical evidence, would be entitled to relief. Moreover, one unnamed class member, Lynwood Epps, was called to the stand at trial, and testified to events which he believed constituted discrimination. There has been no indication that unnamed class members would be content to have their claims simply passed over without decision.

The second issue to face upon dismissal of the named plaintiff's claim is whether that named plaintiff can justify his continuing representation of the class. Is he an adequate class representative? Of course, this question has already been answered by implication at the outset of this litigation when Judge Boyle certified the case as a class action.[16] Rule 23(a) (4) sets forth the existence of an adequate class representative as a prerequisite to class certification. However, Satterwhite seems to contemplate a second examination of this issue after the named plaintiff's claim is dismissed.

Adequacy of representation encompasses both the competence of plaintiff's counsel, and the character and interest of the plaintiff himself.[17] In retrospect, the performance of plaintiff's counsel is easily evaluated. Both of Markey's attorneys are thoroughly experienced in Title VII litigation. Both prosecuted their clients' case diligently and thoroughly. As to Markey himself, his moral character has not been questioned. Although his particular claim has been dismissed, it does not appear that his interests are antagonistic to class interests.[18] I conclude that plaintiff Markey remains an adequate class representative although I have denied him the relief he personally sought.

B. Narrowing the class

The scope of the class originally certified in this case in the spring of 1976 was quite broad.[19] At the trial at the close of *227 plaintiffs' case, defendant moved for an order redefining the class[20] in the following manner:

"I. All black applicants for employment in jobs which are part of the bargaining unit represented by the O.C.A.W. Union at the Tenneco Oil Company Refinery in Chalmette, Louisiana, from March 15, 1974, to the date of the final order of this Court;
II. All blacks who were, are or will be employees in jobs which are part of the bargaining unit represented by the O.C. A.W. Union at the Tenneco Oil Company Refinery in Chalmette, Louisiana, from March 15, 1974, to the date of the final order of this Court."[21]

I took the motion under submission, and I believe that this opinion is the appropriate vehicle for my ruling on it.

The motion presents a dual argument for limitation of the class. First, defendant contends the plaintiffs may not raise any issue based upon events occurring earlier than 180 days prior to Markey's EEOC charge, and that the class should therefore be limited to persons employed and events occurring after March 15, 1974. Second, defendant contends that Markey lacks the "nexus" required to represent any employees who are not members of the bargaining unit, and that the class should be limited to persons represented by the O.C.A.W. Union.

1. Statute of limitations

Under Title VII, a person seeking relief from employment discrimination must file a charge with the EEOC within 180 days of the occurrence of the allegedly discriminatory act.[22] Failure to file within the prescribed time bars not only any EEOC action, but also bars the ultimate filing of suit in federal court.[23]Oatis v. Crown Zellerbach Corp.[24] established, however, the proposition that if the named plaintiff has properly proceeded with his charge before the EEOC he may nevertheless represent unnamed class members who have not filed charges with the EEOC prior to the institution of suit. The reasoning is that the filing requirement is primarily procedural rather than substantive, and

"... no practical purpose could be served by requiring scores of substantially identical grievances to be processed through the EEOC when a single charge would be sufficient to effectuate both the letter and the spirit of Title VII."[25]

This analysis raises the question of what if any of the other limitations contained in Title VII apply to unnamed class plaintiffs. Specifically in this case, given that unnamed class members need not have actually filed EEOC charges, does the time limitation of 180 days which is imposed on the *228 named plaintiff also apply to the class claims? It would seem unfair if the individual plaintiff who had gone to the trouble and effort of retaining a lawyer and actively prosecuting his case was more restricted in his claim than the unnamed class members who take little or no part in the litigation and rely upon the labors of the named plaintiff.

The Fifth Circuit has not confronted this issue. However, the Third Circuit in Wetzel v. Liberty Mutual Ins. Co.[26] held that a named plaintiff could not represent persons who could not have filed a charge with the EEOC at the time he filed his charge. This meant that anyone who left the employ of the defendant more than 180 days[27] before the named plaintiff filed his charge was time barred from asserting a claim as a member of the class. It did not mean that any discriminatory act that occurred prior to that time was not actionable, because certain policies are considered to be continuing violations of Title VII and would allow the filing of a charge at any time by a present employee. Therefore, for example, an employee who was denied a promotion prior to the named plaintiff's 180 day period, but who was still employed by the defendant during the 180 period would have an actionable claim since failure to promote is a continuing violation of Title VII[28] and the injury is ongoing until it is remedied or the employment relationship is severed. On the other hand, since failure to hire is not a continuing violation[29] a person who had applied for employment and been rejected prior to the named plaintiff's 180 day period would not be eligible for class membership.[30]

Although the Fifth Circuit has not considered Wetzel's holding, the case has been followed by a number of district courts[31] and has been recently reaffirmed as the position of the Third Circuit.[32] There has been no substantial disagreement with Wetzel in the jurisprudence, and the Supreme Court has denied certiorari in the Wetzel case itself. Therefore, I believe that Wetzel should be adopted in this Circuit, and since Wetzel imposes a time limitation on all class members I find defendant's motion to be meritorious on the statute of limitations issue.

2. "Nexus" of named plaintiff with class members

In order to properly represent a class, the named plaintiff must establish his "nexus with the class and its interests and claims which is embraced in the various requirements of 23(a) and (b)."[33] This requirement is commonly paraphrased by saying simply that membership in the class is a prerequisite to pursuance of a class action.[34] Defendant contends in its motion to clarify the class that plaintiff Markey, a probationary *229 employee and member of the O.C.A.W. union, cannot properly represent non-union employees at Tenneco.

The primary authority for this position is the 1975 Fifth Circuit case of Wells v. Ramsay, Scarlett & Co., Inc.[35]Wells was the inverse of the present case. In Wells a black non-union supervisor sought to represent black union employees of the defendant company. The court stated:

"We conclude after studying the record that there was no nexus between plaintiff, a foreman, and the longshoremen named in the class. Plaintiff does not seek to represent a class consisting of black foremen. He was a salaried employee, employed on a monthly basis, who supervised longshoremen. The longshoremen in the asserted class were members of Local 1830, one of the defendant unions. They were assigned on a daily basis through the unions, under collective bargaining agreements, to stevedoring companies for the available longshoremen jobs on the docks in the Baton Rouge area. Plaintiff was not in the longshoremen group for job assignment purposes; indeed, while at one time a member of Local Union 1830, he had severed his affiliation with the union long before his complaint was filed. Moreover, he testified that he was physically unable to perform the services of a longshoreman.
We conclude that the nexus between plaintiff and the class was insufficient and thus the district court did not err in declining to allow the cause to proceed as a class action. . . . One may not represent a class of which he is not a part."[36]

Wells appears to be somewhat of a maverick case in this Circuit. Except for one district court case[37] this aspect of Wells has not been further explained in any manner by any court in this circuit. It remains, nevertheless, the controlling jurisprudence on the subject.[38]

In this case plaintiff Markey was hired as an hourly-paid bargaining unit employee. Markey was given a copy of the collective bargaining agreement and understood that this document governed the terms of his employment. Markey admitted that he never sought or aspired to any job beyond the bargaining unit. Under these circumstances, I find no nexus between Markey and the non-union employees he seeks to represent.

In sum, I find that defendant's motion to clarify class has merit, both as to the request for a time limitation, and as to the request to limit the class to union employees. The motion is therefore GRANTED.


In order to prove that defendant engaged in a "pattern and practice"[39] of employment *230 discrimination against its black employees plaintiffs offered both individual testimony of two former black employees (one of whom was the named plaintiff Dempsey J. Markey), and the statistical analysis of their expert Dr. Arnold Levine. Before addressing the various areas in which plaintiffs claim discrimination, it may be helpful to set forth in summary the principles which govern the shifting burden of proof and the use of statistical evidence in Title VII cases.

Because the objective of Title VII is to achieve equality of employment opportunity for all citizens, the Act requires "the removal of artificial, arbitrary, and unnecessary barriers to employment when the barriers operate invidiously to discriminate on the basis of racial or other impermissible classification."[40] The Act prohibits not only overt discrimination, but also practices which, although facially neutral, operate to segregate and classify along racial lines.[41]

As in any other case, a Title VII plaintiff bears the initial burden of proof. He must offer evidence sufficient to create an inference of the existence of discrimination. The burden then shifts to the employer to defeat the inference which the plaintiff has raised by exposing plaintiff's proof as either inaccurate or insignificant.[42] Whether the plaintiff chooses to approach his case by demonstrating disparate treatment or disparate impact[43] he will most likely employ a statistical analysis as a method of proof. The Fifth Circuit has often accorded statistical evidence critical weight in Title VII cases,[44] and the Supreme *231 Court has recently held that "gross statistical disparities" alone may constitute a prima facie case of discrimination in certain circumstances.[45]

With these general principles in mind, I turn now to the specific allegations of discrimination.


The defendant Tenneco accepts applications on a continuing basis from persons who hear about the refinery by word of mouth. When a vacancy occurs, that position is first posted for bid by current employees in accordance with the terms of the collective bargaining agreement.[46] If the position is not filled in this manner, Tenneco then reviews the most current applications in its file and selects a certain number of applicants to call in for a screening interview. If an applicant survives the screening, he is then scheduled for a pre-employment test. Both Charles H. Kilgore the refinery general manager, and Paul Floth the administrative manager testified that the test results did not influence hiring decisions, but rather that the test was given for the sole purpose of compiling data for a longitudinal validation study[47] of the test. Generally, Tenneco receives a sufficient number of walk-in applications to fill vacancies. On the occasions when it does not, it advertises the vacancies in the "Help Wanted" section of several newspapers such as the New Orleans Times Picayune. Tenneco recruits from St. Bernard, Orleans, Plaquemines, Jefferson, and St. Tammany Parishes.

There are two "entry level" positions at the Tenneco refinery: laborer in the Maintenance Department, and operator trainee in the Operations Department. Since these are the lowest paying jobs at the plant and require no particular skills, these positions are usually filled from the applications file. The more advanced positions in the bargaining unit requiring special skills, if not filled by a qualified bidder from within, are more difficult to fill, and Tenneco more often takes an active role in recruitment of individuals for these positions than it does for the entry level jobs.

In recent years Tenneco has followed an explicit policy of hiring an equal number of blacks and whites.

Plaintiffs' statistical evidence was structured around five hypotheses. Using statistical data, plaintiffs' expert Dr. Arnold Levine performed various tests and either rejected or accepted each hypothesis.[48] Hypotheses 1 and 2 relate to hiring practices.

Hypothesis 1 reads, "The proportion of Black personnel [at Tenneco] is greater than or equal to the proportion of Blacks in the 1970 Orleans SMSA and the 1975 populations. The hypothesis is tested for each of the following EEOC categories: Managers, professionals and technicians; office and clerical; craftsmen; operatives; laborers." By comparing the percentage of blacks in each employment category in the general population as reflected in the 1970 New Orleans Standard Metropolitan Statistical Areas (SMSA) report[49] with the percentage *232 of blacks in each employment category at Tenneco Dr. Levine rejected hypothesis 1 for several categories in several different years. In light of the narrowed class,[50] the only relevant rejection of the hypothesis is that for the category of craftsmen in the years 1974 and 1975.

On the subject of such statistical analyses, the Supreme Court has recently stated:

"[A]bsent explanation, it is ordinarily to be expected that nondiscriminatory hiring practices will in time result in a work force more or less representative of the racial and ethnic composition of the population in the community from which employees are hired. Evidence of longlasting and gross disparity between the composition of a work force and that of the general population thus may be significant even though § 703(j) makes clear that Title VII imposes no requirement that a work force mirror the general population."[51]

Therefore ordinarily, assuming for the moment the validity of the SMSA figures as a comparative data base[52], the disparity between the percentage of black craftsmen in the general population and the percentage of black craftsmen at Tenneco for the years 1974 and 1975 establishes a prima facie showing of race discrimination in hiring in that job category for those years. However, given Tenneco's policy of filling job vacancies by promoting from within ranks whenever possible[53] and resorting to hiring from the outside only in the absence of qualified current employees, the test of hypothesis 1 and its results lose meaning in relation to hiring discrimination.[54]

Hypothesis 2 presented the framework for a more exacting analysis. It reads: "The proportion of Black Laborers hired by year [at Tenneco] (from 1970-1975) is greater than or equal to their proportionate representation in the SMSA workforce." Dr. Levine performed two tests to ascertain the validity of this hypothesis. The first, the "sign" test[55] revealed that for each year from 1970 to 1975[56] the percentage of black laborers hired by Tenneco fell below the percentage of black laborers in the SMSA workforce.[57] According to Dr. Levine, *233 the probability of this occurring by chance was approximately 2/100, well below the 5 per cent significance level[58] he employed as a standard against which to measure his results. The second test, the "test of proportions"[59] revealed that the percentage of total black laborers hired by Tenneco over the years 1970 through 1975 (.483[60]) varied significantly from the percentage of black laborers in the SMSA (.597). In each case hypothesis 2 which assumed nondiscriminatory hiring of black laborers was rejected and a prima facie case of discrimination was made out.

Defendant's primary attack upon Dr. Levine's analysis is that the SMSA does not reflect the relevant labor force from which Tenneco draws its employees. The SMSA includes the parishes of Orleans, Jefferson, St. Bernard, and St. Tammany. Dr. Levine did not ascertain the actual residences of Tenneco employees, and consequently made no attempt to weight his analysis in accordance with the distribution of Tenneco employees among the surrounding parishes.

The importance of accurately selecting the relevant labor market is illustrated by the recent Supreme Court case of Hazelwood School District v. United States.[61] In that case the plaintiff submitted statistical proof to show that the percentage of black teachers hired by the Hazelwood School District varied significantly from the percentage of black teachers available in the relevant labor market. The plaintiff used a labor market consisting of St. Louis County (the county in which the school district was located) and the city of St. Louis, a city enclosed by but not included in St. Louis County. The defendant argued that the inclusion of the city of St. Louis in the relevant labor market distorted the analysis because of the special attempts by the city to maintain a 50 per cent black teaching staff. Overturning the Court of Appeals' unquestioning acceptance of plaintiff's statistics, the Supreme Court commented, "What the hiring figures prove obviously depends upon the figures to which they are compared."[62] The Court then remanded the case to the district court for a detailed consideration of whether the city of St. Louis should have been excluded from the relevant labor force. In a footnote the Court remarked:

"These observations are not intended to suggest that precise calculations of statistical significance are necessary in employing statistical proof, but merely to highlight *234 the importance of the choice of the relevant labor market area."[63]

The Tenneco refinery is located in St. Bernard parish and is not accessible via public transportation. Quite naturally, the majority of Tenneco employees reside also in St. Bernard parish (63.5 per cent[64]). The remaining employees reside mainly in Orleans (28 per cent), Plaquemines (5 per cent), and Jefferson (2.5 per cent).[65] As shown in published census data, the racial proportions of the civilian labor force in these parishes is:

                        White         Black
St. Bernard              96%           4%
Orleans                  61%          39%
Jefferson                89%          11%
Plaquemines              84%          16%[66]

The fact that a substantial majority of Tenneco's employees comes from a parish with only a 4 per cent black work force emphasizes the basic flaw in Dr. Levine's analysis. And plaintiffs have not contended that Tenneco practiced discrimination by drawing the majority of its work force from the parish in which the refinery is located, nor could they logically do so. By weighting the black labor force in each parish to reflect its actual contribution to Tenneco's work force a relevant labor market of 14.53 per cent blacks results.[67] Compared with this figure, Tenneco's hiring practices are clearly nondiscriminatory.

Refining this data further in terms of laborers only, Tenneco should have a relevant labor market in the laborer category of 32.61 per cent blacks.[68] Since plaintiffs' *235 own evidence shows that 48.3 per cent[69] of the laborers hired over the years 1970 to 1975 are black, it is obvious that Tenneco has hired black laborers in proportions beyond their representation in the workforce, again negating any prima facie case of discrimination. In light of this conclusion it is not necessary to address Tenneco's contentions that Tenneco operatives should have been included in the analysis of laborers at Tenneco; inclusion of operatives would only have strengthened Tenneco's position further.

Before leaving the subject of hiring, two more topics should be briefly addressed: method of recruitment and pre-employment testing.

As discussed earlier, Tenneco rarely advertises job openings. Ordinarily news of employment vacancies is spread simply by word of mouth. Plaintiffs attack the word of mouth system as discriminatory.

"Under word-of-mouth hiring practices, friends of current employees admittedly [receive] the first word about job openings. [If] most current employees are white, word-of-mouth hiring alone would tend to isolate blacks from the `web of information' which flows around opportunities at the company. See Blumrosen, The Duty of Fair Recruitment Under the Civil Rights Act of 1964, 22 Rutgers L.Rev. 465, 477 (1968). No business necessity compels the company to continue to rely so heavily on this hiring technique."[70]

However, plaintiffs fail to recognize that word of mouth recruitment is not per se discriminatory; the method itself is neutral only when the company's work force does not reflect the racial balance in the general working population does word of mouth recruitment serve to perpetuate discrimination.[71] Here, plaintiffs have not carried their initial burden of proving a substantial disparity between Tenneco's work force and the relevant labor market. Therefore, assuming Tenneco's work force is racially balanced, word of mouth recruitment can only serve to reinforce that balance.

Plaintiffs further contend that the use of a pre-employment test at Tenneco violates Title VII because the test used has not been validated in accordance with EEOC guidelines and because it has an adverse impact on the hiring of blacks.

The Supreme Court held in Griggs v. Duke Power Co. that Title VII forbids the use of employment tests that effect discrimination unless the employer can show that the test has a manifest relationship to the job in question.[72] The EEOC has promulgated guidelines[73] by which employers may validate employment tests. The Fifth Circuit has found that:

"[T]hese guidelines undeniably provide a valid framework for determining whether a validation study manifests that a particular test predicts reasonable job suitability. Their guidance value is such that we hold they should be followed absent a showing that some cogent reason exists for noncompliance."[74]

The Fifth Circuit has further cited longitudinal validation as preferable over concurrent validation.[75] However, as the Supreme Court has recently remarked, the *236 burden of showing that an employment test is job related "arises . . . only after the complaining party . . . has shown that the tests in question select applicants for hire . . . in a racial pattern significantly different from that of the pool of applicants."[76]

In this case plaintiffs failed to demonstrate statistically or otherwise that test results were in any manner correlated with hiring selections.[77] In fact, the testimony of Tenneco officials was that the test results were disregarded entirely, and that the test was administered to applicants solely for the very purpose of conducting a longitudinal validation study of the test, i.e., records of the applicants' test results were kept for comparison with future job performance in order to determine the accuracy with which the test could predict job performance.

In sum, plaintiffs have not proved that hiring discrimination exists at Tenneco in any form.


Promotions within the bargaining unit are governed by the terms of the collective bargaining agreement[78] which provides for a bidding and seniority system. Pursuant to this procedure vacant positions are posted for bid. If there are no bidders within the department in which the vacancy occurs, the bid is posted plant wide. If no current employee bids on the position the job is filled by hiring a new employee. With the exception of craftsmen positions, promotions are awarded on a strict seniority basis. In the case of craftsmen, a further demonstration of journeyman skills is required. Under the Tenneco system it is possible for an employee to "bid down" if he so wishes, and he may have to if he desires to switch from one department to another.

In order to prove that Tenneco discriminates in its promotion practices, Dr. Levine first referred to the data compiled for the test of proportions on hypothesis 1[79] to show that for the relevant years 1974 and 1975[80] the proportion of black craftsmen at Tenneco varied significantly from the proportion of black craftsmen in the SMSA.

Contrasting the proportion of blacks in a specific job classification with the proportion of blacks in the general population is a proper method of establishing a prima facie case of promotion discrimination.[81] However, the SMSA again does not reflect the relevant labor force for Tenneco. Census data indicate that in the craftsman category, the relevant labor force is 11 per cent black.[82] Representation of blacks in *237 craftsman positions at Tenneco also figures at 11 per cent.[83] Clearly, there is no disparity, and hence no proof of discrimination.

However, Dr. Levine did not stop at the population comparison, but went on to try a second approach to proving promotion discrimination. The analysis began with the following hypothesis (hypothesis 4 in Dr. Levine's report): "Blacks hired as laborers [at Tenneco] are promoted in the `same manner' as Whites hired as laborers [at Tenneco]." For each year from 1970 through 1976 and using only employees who had not been terminated as of 1976 Dr. Levine compared the job status of blacks to whites. By analyzing for each black the number of whites hired at the same time ahead, equal, or behind that black in job position, Dr. Levine was able to compute the average proportion of whites ahead of blacks for each year.[84] Dr. Levine urged that in the absence of discrimination this proportion would amount to 33.3 per cent whites would be evenly distributed above, equal to, and behind blacks. He then compared the average proportion of whites *238 ahead for each year[85] to the 33.3 per cent figure. Dr. Levine concluded that the variation between the two figures steadily increased the longer an employee remained at Tenneco, and reached a level of statistical significance at five years and after.[86]

Plaintiffs' statistical evidence was rebutted by defendant through both expert testimony and specific evidence relating to promotions. Dr. Richard Olson, defendant's expert statistician, testified convincingly that the radical fluctuations in the data[87] rendered Dr. Levine's approach inappropriate. Applying the "sign test"[88] to the first and second years of employment[89] Dr. Olson concluded that in those cases there was no evidence of discrimination. In other words, for the various test groups Dr. Levine's "proportion of whites ahead" figure fluctuated above and below the expected average of 33 per cent approximately an equal number of times.[90] I find that Dr. Olson discredited Dr. Levine's analysis on this point.

Defendant further produced evidence concerning every opening in certain job categories from January 1970 to January 1977. In the categories of operator trainee and stillman[91] the seniority system was equally applied, that is to say the senior bidder whether black or white was awarded the job, in every instance except one.[92] In the case of every opening there were blacks within the Maintenance Department[93] with *239 higher seniority than the successful bidder who chose not to bid for reasons which are not explained.[94]

The Supreme Court addressed the issue of seniority relief for nonapplicants in International Brotherhood of Teamsters v. United States.[95] The Court initially determined that nonapplicants are not "inexorably barred" from an award of seniority relief commenting:

"The effects of and the injuries suffered from discriminatory employment practices are not always confined to those who were expressly denied a requested employment opportunity. A consistently enforced discriminatory policy can surely deter job applications from those who were aware of it and are unwilling to subject themselves to the humiliation of explicit and certain rejection."[96]

The Court agreed in that case that the plaintiff had proved that the defendant employer had engaged in "post-Act"[97] discriminatory policies entitling discriminatees to relief. However, the Court held that this proof created a rebuttable presumption in favor of individual relief only for those persons who had applied for positions and been turned down.[98] Nonapplicants bore a further burden:

"To conclude that a person's failure to submit an application for a job does not inevitably and forever foreclose his entitlement to seniority relief under Title VII is a far cry, however, from holding that nonapplicants are always entitled to such relief. A nonapplicant must show that he was a potential victim of unlawful discrimination. Because he is necessarily claiming that he was deterred from applying for the job by the employer's discriminatory practices, his is the not always easy burden of proving that he would have applied for the job had it not been for those practices. [Citation omitted.] When this burden is met, the nonapplicant is in a position analogous to that of an applicant and is entitled to the presumption [of discrimination accorded to an applicant].
. . . . .
"While the scope and duration of the company's discriminatory policy can leave little doubt that the futility of seeking [certain positions] was communicated to the company's minority employees, that in itself is insufficient. The known prospect of discriminatory rejection shows only that employees who wanted [those certain positions] may have been deterred from applying for them. It does not show which of the nonapplicants actually wanted such jobs, or which possessed the requisite qualifications."[99]

*240 And further in a footnote:

"Inasmuch as the purpose of the nonapplicant's burden of proof will be to establish that his status is similar to that of the applicant, he must bear the burden of coming forward with the basic information about his qualifications that he would have presented in an application. . . . [T]he burden then will be on the employer to show that the nonapplicant was nevertheless not a victim of discrimination."[100]

The case was remanded with instructions that the plaintiff must carry its burden of proof as to each individual nonapplicant.[101]

In this case seniority is the only factor in promotions into the categories of operator trainee and stillman. In every case[102] seniority rank has been strictly followed. Therefore there has been no discouraging pattern of discrimination which might deter blacks from bidding on vacancies. Moreover, there is no past discrimination in hiring or otherwise which might be perpetuated by this neutral seniority system.[103] I conclude that promotions into these job categories have been free of discrimination.

Defendants also produced evidence concerning every job opening in the following crafts positions from January 1970 to January 1977:[104] instrument electrician, carpenter, painter/insulator, first class pipefitter, machinist and welder. Openings for these positions are awarded to the senior qualified bidder. Since in this case another determining factor is added to seniority, a different approach is necessary. However, it should be noted initially that without regard to qualifications, the record shows that in every case where an opening has been filled by a current employee[105] the *241 award might as well have been made on the basis of seniority alone except in two instances. Therefore, it appears that skill requirements have not operated to exclude blacks in statistically significant proportions.

In enacting Title VII Congress

"did not desire to melt job qualifications having no racially discriminatory ingredient or controlling pre-Act antecedent. In light of Title VII's legislative history, ascribing such an altruistic yet impractical purpose to that legislative body would surely be erroneous`reverse discrimination' of the most blatant sort."[106]

Job requirements which serve to perpetuate discrimination, however, even though they may be facially neutral must qualify as "business necessities".

"The business necessity test essentially involves balancing the need for the challenged practice or policy against its discriminatory impact. The business purpose must be `sufficiently compelling to override any racial impact.'; it must `effectively and efficiently' carry out a business purpose; and there must be no acceptable alternative practice."[107]

However, the burden of proving business necessity "arises, of course, only after the complaining . . . class has made out a prima facie case of discrimination, i. e. has shown that the [requirements] in question select applicants for hire or promotion in a racial pattern significantly different from that of the pool of applicants."[108] Here plaintiffs have not demonstrated that the skill requirements of the crafts positions at Tenneco operate to disproportionately exclude black bidders. In only two instances have skill requirements meant that a white bidder with less seniority than a black bidder has been awarded a vacant position. In eleven instances, skill requirements have excluded all bidders, both black and white.

Since skill requirements do not disproportionately exclude blacks from crafts positions, Tenneco is not required to come forth with proof of the "business necessity" or "job relatedness" of these requirements. However, even had a prima facie case been established, I believe that business necessity exists in this case. The Fifth Circuit has recently held that

". . . a facially neutral job criterion can be so manifestly job-related so as not to be `the kind of "artificial, arbitrary, and unnecessary barriers"' Title VII prohibits. In such a case, the employer need not make an evidentiary showing of business necessity, even though the criterion may have a discriminatory impact."[109]

Several years ago Judge Wisdom remarked in dicta that skill in typing was manifestly a business necessity for a secretarial position.[110] It seems to me that it is also manifestly necessary that a welder know how to weld, an electrician know how to repair electrical equipment, and so forth.

In short, plaintiffs have not shown that promotion discrimination exists at Tenneco.

*242 VII. PAY

Rates of pay at Tenneco for union employees are governed by the collective bargaining agreement. For each position, a specific hourly rate is set forth which applies to every employee in that position. Since I have already determined that there is no promotion discrimination at Tenneco and that there are no segregated departments there, it is difficult to imagine how plaintiffs could prove discrimination in rates of pay to blacks and whites. In fact, they have not.

Hypothesis 5 of Dr. Levine's report reads as follows: "There is no pay differential between Blacks and Whites [at Tenneco]." Dr. Levine averaged salaries of black and white employees at Tenneco and concluded that there was a pay disparity of $215.00 per year between blacks and whites, or 1.6 per cent of total average earnings. However, Dr. Levine admitted that his data was flawed, and therefore refused to test the statistical significance of the disparity arguing that to do so would be to "dignify" incompetent data. Therefore, Dr. Levine neither accepted nor rejected hypothesis 5. Nevertheless, plaintiffs continue to argue that Tenneco does discriminate against blacks in pay.

There is no basis for this contention. Dr. Levine's figures are indeed inaccurate. Since he did not have precise numbers of what each employee earned, Dr. Levine extracted what data he could from tables which showed by year which employees occupied which jobs. He assumed that each man in question earned only his straight-time rate at his job (ignoring overtime), and since he did not know when within the year a man acceded to a higher-rated position, he assumed an identical promotion date for all men in the analysis. Similarly when a man resigned or was discharged he did not know how much of the year had been worked, and so made an arbitrary assumption on that point too. Despite the flaws in the data, defendant's expert statistician Dr. Olson ran a statistical test on the disparity to determine whether it was statistically significant. He found that it was not.

Therefore, by no stretch of the imagination have plaintiffs shown discrimination in pay on the basis of race at Tenneco.


Employees within the bargaining unit at Tenneco are initially hired for a 120 day probationary period during which they have no effective plant seniority.[111] During that period each employee is rated by his supervisor. If he receives a satisfactory rating he is hired permanently; if not, he is discharged. The vast majority[112] of employees who have been discharged in recent years have been discharged during the probationary period.

Plaintiffs contend that the rating form[113] for probationary employees is too subjective and leaves room for race discrimination. They claim that Tenneco has indeed practiced race discrimination over recent years and offer statistical evidence to support this contention. Defendant argues that Dr. Levine's data is inaccurate and that his statistical approach is incorrect.

Hypothesis 3 of Dr. Levine's report reads: "The proportion of Blacks terminated [at Tenneco] is less than or equal to the proportionate termination of Whites." Using the sign test[114] and the test of difference in proportions[115] Dr. Levine rejected the hypothesis concluding that Tenneco discriminates against blacks in "terminations". Dr. Levine's data consisted of figures on all laborers in the Maintenance Department terminated for any reason from 1970 through 1975. This included those who voluntarily resigned, retired, died, or quit for medical reasons.

*243 I agree with defendant that Dr. Levine's data is flawed and that Tenneco can only be held responsible for the employees that it discharged, not for those who left Tenneco of their own accord. Of all employees fired during the period 1970 through 1975, fifteen were discharged for failure to satisfactorily complete their probationary period.[116] Seven of these were black, or 47 per cent.[117] The proportion of blacks hired into entry level jobs during this same period is 40 per cent.[118] Considering the small size of the group that was fired, these percentages compare favorably and do not evidence discrimination.

While I have previously questioned to some extent the objectivity of the rating form for probationary employees,[119] since plaintiffs have not demonstrated any discriminatory impact of the form it is not necessary for defendant to justify the form as a business necessity.[120]

Of the permanent nonprobationary employees discharged during the 1970 through 1975 period, three were white and one was black.[121] Again, the small sample size undermines the effectiveness of a statistical analysis. However, the proportion of blacks fired (25 per cent) does in fact compare favorably with the proportion of permanent black employees at Tenneco.

Finally, even if Dr. Levine's data had been accurate, defendant's statistical expert Dr. Olson testified that by performing a statistical test based on frequency of termination rather than proportions hypothesis 4 which assumed nondiscrimination would be accepted rather than rejected. Since I find that Dr. Levine's data was seriously flawed, it is unnecessary to resolve the question of which statistical test was proper.

I conclude that Tenneco does not discriminate in the discharge of probationary or permanent employees.

Since plaintiffs have not proved discrimination at Tenneco either against the named plaintiff Dempsey J. Markey or against blacks as a class, judgment shall be entered in favor of defendant Tenneco Oil Company dismissing plaintiffs' suit at plaintiffs' cost.


[1] Court's Order, Record Doc. No. 15. In light of evidence adduced at trial and additional legal arguments brought to the attention of the Court, I must modify the initial certification pursuant to Rule 23(c) (1), Fed.R.Civ.P. See discussion limiting class, infra at pp. 226-227 et seq.

[2] The wharfmen are not represented by a union. In view of the difference in job requirements and union representation, there is no flow of personnel between the wharf and other positions in the refinery, as there is among the other departments and divisions of the refinery.

[3] Labor organizations, as well as employers, have a duty to prevent the perpetuation of discriminatory employment practices. Myers v. Gilman Paper Corp., 5 Cir. 1977, 544 F.2d 837, 850, appeal pending. Quite often, therefore, Title VII suits are brought against the employer and the union in tandem. See, e. g., Carey v. Greyhound Bus Co., 5 Cir. 1974, 500 F.2d 1372. In this case, the union has not beer, joined as a defendant.

In regard to the place of collective bargaining agreements in a scheme of employment practices attacked as discriminatory, the Fifth Circuit has recently cautioned that "when a union has chosen from among arguably acceptable alternatives, a minority of employees should not be given free reign to upset the union's bargain simply because the faction or a court might have subjectively deemed other alternatives better." Myers, supra at 858. Nonetheless, the law is clearly established that the employer may not use a collective bargaining agreement in any way as a shield when discriminatory practices are shown to exist. Carey v. Greyhound Bus Co., Inc., 5 Cir. 1974, 500 F.2d 1372, 1377; Peters v. Missouri-Pacific Railroad Co., 5 Cir. 1973, 483 F.2d 490, 497, cert. denied, 414 U.S. 1002, 94 S. Ct. 356, 38 L. Ed. 2d 238; United States v. Jacksonville Terminal Co., 5 Cir. 1971, 451 F.2d 418, 454, cert. denied 1972, 406 U.S. 906, 92 S. Ct. 1607, 31 L. Ed. 2d 815.

[4] 5 Cir. 1977, 555 F.2d 1251.

[5] Id. at 1256.

[6] Long v. Sapp, 5 Cir. 1974, 502 F.2d 34, 37.

[7] Statistics may be used to show pretext in some individual Title VII cases. Turner v. Texas Instruments, Inc., 5 Cir. 1977, 555 F.2d 1251. However, in this case, the only offer of statistical proof was a general one on the number of blacks and whites discharged at Tenneco over the years in question. It is clear from the Turner case, that what is required is a statistical analysis tailored to the individual facts at issue. In this case, that would mean a showing that under similar circumstances whites received lesser punishments than blacks, or that whites were not disciplined at all whereas blacks were. In order to be meaningful, the statistical sample would have to be of a fair size. As mentioned earlier, plaintiff could not point to a single similar incident involving any employee, white or black.

[8] Category 13: character and moral standards. In this category Markey rated a "three".

[9] The other categories were: thoroughness in completing job assignments, understanding of job assignments, ability to work without constant supervision, willingness to accept responsibility for own mistakes, willingness to work on unpleasant jobs, ability to learn the job, ability to accept orders, ability to get along with fellow employees, observation of Company rules including Safety rules, willingness to look for additional work when assignments are completed, ability to work efficiently and quickly, and punctuality and attendance record.

[10] Huff v. N. D. Cass Co., 5 Cir. 1973, 485 F.2d 710; Long v. Sapp, 5 Cir. 1974, 502 F.2d 34; Smalls, Class Actions under Title VII: Some Current Procedural Problems, 25 Am.U.L.Rev. 821, 840-41 (1976).

[11] 5 Cir. 1977, 557 F.2d 414.

[12] Id. at 423.

[13] 424 U.S. 747, 96 S. Ct. 1251, 47 L. Ed. 2d 444.

[14] The other Supreme Court case cited in Satterwhite was East Texas Motor Freight System, Inc. v. Rodriguez, 1977, 431 U.S. 395, 97 S. Ct. 1891, 52 L. Ed. 2d 453 in which the Court approved the District Court's decision dismissing the plaintiffs' class allegations upon a finding that the named plaintiffs were not proper class representatives. In a footnote the Court speculated that:

". . . a different case would be presented if the District Court had certified a class and only later had it appeared that the named plaintiffs were not class members or were otherwise inappropriate class representatives. In such a case, the class claims would have already been tried, and, provided the initial certification was proper and decertification not appropriate, the claims of the class members would not need to be mooted or destroyed because subsequent events or the proof at trial had undermined the named plaintiffs' individual claims."

Id. at 1898.

[15] 96 S. Ct. at 1260.

[16] See discussion at note 1, supra and accompanying text.

[17] Wright & Miller, Federal Practice and Procedure § 1766.

[18] In Eisen v. Carlisle & Jacquelin, 2 Cir. 1968, 391 F.2d 555, the court commented that the proper measure of the named plaintiff's interest was whether it was antagonistic to class interests, rather than whether it was coextensive with them.

[19] The definition of the class adopted by Judge Boyle is set forth supra at text accompanying note 1. At the time plaintiffs first moved to certify the class, defendant opposed primarily on the ground that the class claims were not reasonably related to Markey's EEOC charge. Defendant argued that if a class were certified it should be limited to "discharged Black probationary employees within the bargaining unit." Record Doc. No. 13. Judge Boyle granted plaintiffs' motion without assigning written reasons.

The leading case on the relation of the EEOC charge to the judicial complaint is Sanchez v. Standard Brands, Inc., 5 Cir. 1970, 431 F.2d 455. Quoting from a Northern District of Georgia case, King v. Georgia Power Co., 1968, 295 F. Supp. 943, the Fifth Circuit held that the judicial complaint "may encompass any kind of discrimination like or related to allegations contained in the charges and growing out of such allegations during the pendency of the case before the Commission." The Fifth Circuit then continued to state an even broader test: "the scope of the judicial complaint is limited to the scope of the EEOC investigation which can reasonably be expected to grow out of the charge of discrimination." Sanchez, 431 F.2d at 467. The Court then affirmed the purpose of the EEOC to encourage voluntary compliance by employers, and stated that an overly restrictive relation requirement would thwart that purpose since employers would have no incentive to conciliate on issues which would never go to court. Given this liberal standard, I believe Judge Boyle was correct in rejecting defendant's argument that the class claims were not "like or related to" Markey's EEOC charge. However, the question of the named plaintiff's "nexus" with the proposed class is, I believe, an entirely separate issue, see discussion, infra at pp. 228-229, which was not fully argued at the time of the original motion to certify class.

[20] Under Rule 23(c) (1), Fed.R.Civ.P., class certification is not immutable, but may be altered or amended at any time before the decision on the merits.

[21] Record Doc. No. 73.

[22] 42 U.S.C. § 2000e-5(e).

[23] Macklin v. Spector Freight Systems, Inc., 1973, 156 U.S.App.D.C. 69, 478 F.2d 979.

[24] 5 Cir. 1968, 398 F.2d 496.

[25] Miller v. International Paper Co., 5 Cir. 1969, 408 F.2d 283.

[26] 3 Cir. 1975, 508 F.2d 239, cert. denied 1975, 421 U.S. 1011, 95 S. Ct. 2415, 44 L. Ed. 2d 679.

[27] Actually, in Wetzel the time limitation was 210 days because a charging party had previously instituted proceedings with a state agency which dealt with discriminatory employment practices. 42 U.S.C. § 2000e-5(d) (1970).

[28] Clark v. Olinkraft, Inc., 5 Cir. 1977, 556 F.2d 1219.

[29] East v. Romine, Inc., 5 Cir. 1975, 518 F.2d 332.

[30] Although a person whose application for employment was rejected during or after the named plaintiff's 180 day period would be.

[31] Grogg v. General Motors Corp., S.D.N.Y. 1976, 72 F.R.D. 523, 534; Lamphere v. Brown University, D.R.I. 1976, 71 F.R.D. 641, 648-49; Presseisen v. Swarthmore College, E.D.Pa. 1976, 71 F.R.D. 34, 47-48; Ungar v. Dunkin' Donuts of America, E.D.Pa. 1975, 68 F.R.D. 65, 140, reversed on other grounds 3 Cir. 1975, 531 F.2d 1211, cert. denied 1977, 429 U.S. 823, 97 S. Ct. 74, 50 L. Ed. 2d 84; Jones v. United Gas Improvement Corp., E.D.Pa. 1975, 68 F.R.D. 1, 15.

[32] Ostapowicz v. Johnson Bronze Co., 3 Cir. 1976, 541 F.2d 394, 397, cert. denied 1977, 429 U.S. 1041, 97 S. Ct. 741, 50 L. Ed. 2d 753.

[33] Huff v. N.D. Cass Co., 5 Cir. 1973, 485 F.2d 710, 714.

[34] Long v. Sapp, 5 Cir. 1974, 502 F.2d 34, 42.

[35] 506 F.2d 436.

[36] Id. at 437-38.

[37] McClinton v. Turbine Support, Division of Chromalloy American Corp., W.D.Tex. 1975, 68 F.R.D. 236, relying on Wells held that the plaintiff, a black foreman, could not represent black union members.

[38] A recent Supreme Court case although not precisely on point indicates a retreat from prior jurisprudence which had loosely construed class action requirements in civil rights cases, and so offers some support for the restrictive view taken in Wells. In East Texas Motor Freight System, Inc. v. Rodriguez, 1977, 431 U.S. 395, 97 S. Ct. 1891, 52 L. Ed. 2d 453, three individual plaintiffs sought to represent a class composed in part of all applicants for the position of "line driver" with the defendant. The Court found, inter alia, that since the individual plaintiffs were clearly unqualified to become "line drivers", they were not members of a class of potential "line drivers", and thus the class claims on that score were dismissed.

[39] This language which has become a catch phrase in Title VII litigation is contained in 42 U.S.C. § 2000e-6(a) and was intended not as a term of art, but only in its usual meaning. See International Brotherhood of Teamsters v. United States, 1977, 431 U.S. 324, 97 S. Ct. 1843, 1855 n. 16, 52 L. Ed. 2d 396 for a discussion of the legislative history behind the phrase.

[40] Griggs v. Duke Power Co., 1971, 401 U.S. 424, 91 S. Ct. 849, 853, 28 L. Ed. 2d 158.

[41] Id.

[42] International Brotherhood of Teamsters v. United States, 1977, 431 U.S. 324, 97 S. Ct. 1843, 1866-67, 52 L. Ed. 2d 396, discussing McDonnell Douglas Corp. v. Green, 1973, 411 U.S. 792, 93 S. Ct. 1817, 36 L. Ed. 2d 668 and Franks v. Bowman Transportation Co., 1976, 424 U.S. 747, 96 S. Ct. 1251, 47 L. Ed. 2d 444.

[43] The Supreme Court has recently defined these terms in International Brotherhood of Teamsters v. United States, 1977, 431 U.S. 324, 97 S. Ct. 1843, 1854 at n. 15, 52 L.Ed.2d 396:

"`Disparate treatment' . . . is the most easily understood type of discrimination. The employer simply treats some people less favorably than others because of their race, color, religion, sex, or national origin. Proof of discriminatory motive is critical, although it can in some situations be inferred from the mere fact of differences in treatment. [Citation omitted.] Undoubtedly disparate treatment was the most obvious evil Congress had in mind when it enacted Title VII. . . .

Claims of disparate treatment may be distinguished from claims that stress `disparate impact.' The latter involve employment practices that are facially neutral in their treatment of different groups but that in fact fall more harshly on one group than another and cannot be justified by business necessity. . . . Proof of discriminatory motive, we have held, is not required under a disparate impact theory."

The Fifth Circuit has recently carved an exception to the general "disparate impact" rule as stated in International Brotherhood of Teamsters. In Smith v. Olin Chemical Corp., 5 Cir. 1977, 555 F.2d 1283 the Court stated:

". . . a facially neutral job criterion can be so manifestly job-related so as not to be `the kind of "artificial, arbitrary, and unnecessary barriers"' Title VII prohibits. In such a case, the employer need not make an evidentiary showing of business necessity, even though the criterion may have a discriminatory impact."

[44] James v. Stockham Valves & Fittings Co., 5 Cir. 1977, 559 F.2d 310; Sagers v. Yellow Freight System, Inc., 5 Cir. 1976, 529 F.2d 721; United States v. T.I.M.E.-D.C., 5 Cir. 1975, 517 F.2d 299, vacated and remanded on other grounds, 1977, 431 U.S. 324, 97 S. Ct. 1843, 52 L. Ed. 2d 396; Rodriguez v. East Texas Motor Freight, 5 Cir. 1974, 505 F.2d 40, vacated and remanded on other grounds, 1977, 431 U.S. 324, 97 S. Ct. 1891, 52 L. Ed. 2d 453; Johnson v. Goodyear Tire & Rubber Co., 5 Cir. 1974, 491 F.2d 1364; Ochoa v. Monsanto Co., 5 Cir. 1973, 473 F.2d 318; Rowe v. General Motors Corp., 5 Cir. 1972, 457 F.2d 348; United States v. Hayes International Corp., 5 Cir. 1972, 456 F.2d 112; United States v. Jacksonville Terminal Co., 5 Cir. 1971, 451 F.2d 418, cert. denied, 1972, 406 U.S. 906, 92 S. Ct. 1607, 31 L. Ed. 2d 815.

[45] Hazelwood School District v. United States, 1977, ___ U.S. ___, 97 S. Ct. 2736, 2741, 53 L. Ed. 2d 768.

[46] Defendant's Exhibit 4, Art. 7, Sect. 7.

[47] Dr. James W. Herrine, an industrial psychologist testified as to the difference between a longitudinal validation study and a concurrent validation study. Under the longitudinal method, the test is administered only to job applicants. By tracking the individual's subsequent job performance and correlating this with the test results, it is possible to ascertain to what degree of accuracy the test predicts job performance. Under the concurrent method the test is administered to current employees, and the results are correlated with their job performance to determine the test's accuracy. The obvious advantage of the concurrent method is that such a study can be quickly performed, whereas a longitudinal study necessarily takes place over an extended period of time.

[48] All of plaintiffs' statistical evidence is contained in Dr. Levine's report. Plaintiffs' Exhibit 4.

[49] The 1970 SMSA report contains precise census data. "Updates" conducted in 1975 and 1976 do not reflect precise figures, but are merely estimates. In order to use the 1970 data in years other than 1970, Dr. Levine had to demonstrate that there had been no significant variation in population distribution since 1970that the 1970 proportions had remained fairly constant. The comparison is set forth in Table 1 of Dr. Levine's report, and does indeed show that only minor variations have occurred since 1970, thus justifying the use of the 1970 data for the years following 1970.

[50] The class is limited to persons employed after March 15, 1974. See discussion pp. 227-228, supra.

[51] International Brotherhood of Teamsters v. United States, 1977, 431 U.S. 324, 97 S. Ct. 1843, 1856-57 at n. 20, 52 L. Ed. 2d 396.

[52] Defendant's contention that the SMSA does not accurately represent the relevant work force in this case is addressed, infra, at pp. 233-235.

[53] The positions at Tenneco equated with the category of "craftsman" as used in the SMSA are not entry level jobs.

[54] Plaintiffs did not present or analyze this data by breaking down the numbers hired and the numbers promoted to craftsmen positions. Promotion discrimination was presented separately and is discussed, infra, at p. 236 et seq.

[55] The sign test is explained in W. Dixon & F. Massey, Introduction to Statistical Analysis 335 (3d ed. 1969). Assuming no discrimination, it would be expected that the percentage of blacks at Tenneco would fluctuate around the percentage of blacks in the workforce, so that approximately half of the time it would fall above that number and the other half it would fall below.

[56] Although the class claims as I have redefined them cannot include acts of discrimination in hiring which occurred prior to March 15, 1974, statistics on the prior years are relevant to show a pattern which would not be visible if only the years 1974 and 1975 were used. See United States v. Jacksonville Terminal Co., 5 Cir. 1971, 451 F.2d 418, 441, cert. denied 1972, 406 U.S. 906, 92 S. Ct. 1607, 31 L. Ed. 2d 815, on the importance of a historical approach to statistical analysis in Title VII cases.

[57] Defendant argued that the proper comparison was not between blacks at Tenneco and blacks in the general population, but rather between blacks actually hired by Tenneco and black applicants to Tenneco. This contention is clearly erroneous. "The statistics which the Courts have always considered is [sic] the racial composition of the employer's work force as compared to the percentage of the minority population in the employer's service area." Jones v. Tri-County Electric Cooperative, Inc., 5 Cir. 1975, 512 F.2d 1, 2.

[58] A significance level is a percentage below which test results become significant, i.e., it becomes highly likely that the results are not due to mere chance, but rather reflect the presence of some selective factor, in this case race discrimination. Dr. Levine used a 5 per cent significance level throughout his report which he testified he chose because of the size of his sample. Defendant strenuously contended that a 1 per cent significance level should have been used, and through the testimony of its expert Dr. Richard Olson demonstrated that with a 1 per cent significance level many of plaintiffs' conclusions of discrimination were reversed. However, in light of the sample size and the use of a 5 per cent level in other Title VII cases, I believe the 5 per cent level is the more accurate one in this case. W. Dixon & F. Massey, Introduction to Statistical Analysis 100 (3d ed. 1969). Note, Beyond the Prima Facie Case in Employment Discrimination Law: Statistical Proof and Rebuttal, 89 Harv.L.Rev. 387, 400-01 at n. 58 (1975) and cases cited therein.

[59] The test of proportions is illustrated in W. Dixon & F. Massey, Introduction to Statistical Analysis 101 (3d ed. 1969).

[60] The figure used in Dr. Levine's report at page 3 is .477. However, upon cross-examination Dr. Levine admitted a minor error in his data and revised this figure to the figure referred to in the text. This revision did not change the significance of the results.

[61] 1977, ___ U.S. ___, 97 S. Ct. 2736, 53 L. Ed. 2d 768.

[62] Id. at 2743.

[63] Id. at 2743-44, n. 17 (emphasis added).

[64] Defendant's Exhibit 13.

[65] Id.

[66] This table is compiled from data found in U.S. Bureau of the Census, "Census of Population: 1970 General Social and Economic Characteristics Final Report PC(1)-C20 Louisiana," Government Printing Office (1972), pp. 325-30; 355-60.

[67] The following table shows how this result was reached:

                  Residence of Employees         Black Population            Contribution to
                  When Hired, Defendant          in Parish Civilian          Tenneco Relevant
Parish            Exhibit 13                     Labor Force                 Labor Market    
St. Bernard         63.5%                   ×       4%                 =         2.54%
Orleans               28%                   ×      39%                 =        10.92%
Plaquemines            5%                   ×      16%                 =          .80%
Jefferson            2.5%                   ×      11%                 =          .27%
                                    Total Theoretical Black
                                    Population in Relevant
                                    Labor Market                                      14.53%

[68] Data found in U.S. Bureau of the Census, The following table applies the same analysis used in footnote 67, supra for laborers only:

                      Residence of Employees         Black Population         Contribution to
                      When Hired, Defendant          in Total Parish          Tenneco Relevant
Parish                Exhibit 13                     Laborers                 Labor Market    
St. Bernard             63.5%                  ×        15%             =        9.53%
Orleans                   28%                  ×        71%             =       19.88%
Plaquemines                5%                  ×        44%             =        2.20%
Jefferson                2.5%                  ×        40%             =        1.00%
                                        Total Theoretical Black
                                        Population in Relevant
                                        Labor Market                                  32.61%

[69] See note 60, supra, and accompanying text.

[70] United States v. Georgia Power Co., 5 Cir. 1973, 474 F.2d 906, 925.

[71] Id.; Sagers v. Yellow Freight System, Inc., 5 Cir. 1976, 529 F.2d 721, 729; Long v. Sapp, 5 Cir. 1974, 502 F.2d 34, 41; Franks v. Bowman Transportation Co., 5 Cir. 1974, 495 F.2d 398, reversed on other grounds 1976, 424 U.S. 747, 96 S. Ct. 1251, 47 L. Ed. 2d 444.

[72] 1971, 401 U.S. 424, 91 S. Ct. 849, 854, 28 L. Ed. 2d 158.

[73] 29 C.F.R. Part 1607.

[74] United States v. Georgia Power Co., 5 Cir. 1973, 474 F.2d 906, 913.

[75] Id. at 912. See note 47, supra, for a discussion of these two methods.

[76] Albemarle Paper Co. v. Moody, 1975, 422 U.S. 405, 95 S. Ct. 2362, 2375, 45 L. Ed. 2d 280.

[77] See Defendant's Exhibit 18.

[78] Defendant's Exhibit 4, Art. 8, Sect. 7.

[79] This analysis was rejected as it related to hiring precisely because Tenneco attempts if at all possible to promote from within into vacancies in upper level jobs. See note 54, supra, and accompanying text.

[80] Since failure to promote is a continuing violation, promotions in years prior to 1974 would have been admissible had plaintiffs structured this analysis in such a manner as to indicate which employees from those prior years were still employed in 1974. However, they did not. Clark v. Olinkraft, Inc., 5 Cir. 1977, 556 F.2d 1219. The analysis of hypothesis 4, also on the subject of promotion discrimination, used only currently employed employees. In that instance, proof of discriminatory failures to promote prior to 1974 is indeed relevant and admissible.

[81] Robinson v. Union Carbide Corp., 5 Cir. 1976, 538 F.2d 652, 660-61, modified on other grounds 1977, 544 F.2d 1258.

[82] Data found in U.S. Bureau of the Census, "Census of Population: 1970General Social and Economic CharacteristicsFinal Report PC(1)-C20 Louisiana," Government Printing Office (1972), pp. 331-36, 361-67 indicates that among craftsmen blacks comprise the following percentages in each of the following parishes: St. Bernard2 per cent; Orleans32 per cent; Plaquemines13 per cent; Jefferson7 per cent.

The following table applies the same analysis used in note 67, supra, for craftsmen only:

                  Residence of Employees         Black Population        Contribution to
                  When Hired, Defendant          in Total Parish         Tenneco Relevant
   Parish         Exhibit 13               Craftsmen         Labor Market    
   St. Bernard      63.5%                   ×        2%             =     1.27%
   Orleans            28%                   ×       32%             =     8.96%
   Plaquemines         5%                   ×       13%             =      .65%
   Jefferson         2.5%                   ×        7%             =      .17%
                                     Total Theoretical Black
                                     Population in Relevant
                                     Labor Market                          11.05%

In James v. Stockham Valves & Fittings Co., 5 Cir. 1977, 559 F.2d 310, the Fifth Circuit held that the relevant work force for comparison to crafts positions was not the local labor market, but rather was the employer's own force of maintenance and production workers, in that case 66 per cent black. I believe that the James case is distinguishable, however, since there the employer maintained substantial training and apprenticeship programs for craft positions. It was therefore to be expected that given equal access to those programs the crafts positions would be filled in the same race proportions as the entry level jobs. No such widescale programs were maintained by Tenneco in this case. The only similar program was one instituted approximately one year before the trial for the position of instrument electrician. Openings in that program were posted for bid and were awarded on a strict seniority basis. The program was instituted when Tenneco failed to fill an instrument electrician vacancy after posting the position for bid and advertising in local papers. There is no such program for any of the other crafts positions at Tenneco. Therefore, it is logical to expect that the crafts positions in this case would reflect the racial balance of qualified craftsmen in the surrounding community, rather than the racial balance in entry level jobs at Tenneco.

[83] There was dispute at trial as to the proper correlation between categories in census data and the job categories at Tenneco. The census data used in the preceding paragraph, as well as that used by Dr. Levine gives figures for a category termed "Craftsmen, foremen, and kindred workers". Proceeding from job descriptions supplied by Tenneco, I include the following Tenneco positions within the census category: Controlman, Stillman, Second Class Pipefitter, First Class Pipefitter, First Class Carpenter, First Class Painter/Insulator, First Class Machinist, First Class Welder, and Electrician Instrumentman. The total number of employees in these positions is 111, including 12 blacks. Defendant's Exhibit 3. This amounts to a black representation of 10.81 per cent, very close to the 11.05 per cent relevant labor market.

[84] See Plaintiffs' Exhibit 4, Table 6.

[85] By "year" I refer to the employee's tenure at Tenneco rather than to a specific calendar year. For example, year one of an employee hired in 1970 is 1970, of an employee hired in 1971 it is 1971, and so forth.

[86] See Plaintiffs' Exhibit 4, Tables 6B and 6C. It may be of passing moment that with the use of a calculator I myself discovered numerous minor mathematical errors in these tables, although none were of such significance to change the ultimate outcome.

[87] The "proportion of whites ahead" varied from 10 per cent to 72 per cent, and in not one single test group did that percentage consistently rise or consistently fall as the years of employment progressed. See Plaintiffs' Exhibit 4, Table 6.

[88] See note 55, supra.

[89] The third, fourth, fifth and sixth years could not be tested in this manner since each of these years contained only four or fewer sample groups to compare.

[90] In rebuttal Dr. Levine attempted to argue that the sign test could not be used on this data because the figures being compared were not independent of one another, i.e. what happened in each successive year depended on what had occurred in the previous years. However, I believe Dr. Levine misconstrued Dr. Olson's analysis. Dr. Olson did not track each group's performance over the years. Rather, he compared each year of tenure among the different groups. To illustrate by example, Dr. Olson did not take the employees hired in 1971 and show that the "proportion of whites ahead" rose and fell for that group over the years in the following pattern: + - = + +. To have done this would indeed be an attempt to administer the sign test to data which were not independent. What Dr. Olson did was to take the first year "proportion of whites ahead" figure for the group hired in each year (six groups over the 1970 to 1976 period) and by comparing that figure to the expected average of 33 per cent derived the following sequence: - + - - = +. In other words, for the group hired in 1970, the figure fell below 33 per cent, in 1971 above, in 1972 below, in 1973 below, in 1974 it equaled 33 per cent, and in 1975 it again fell above. The groups compared are not in any manner related, and therefore the sign test was properly administered by Dr. Olson. See W. Dixon & F. Massey, Introduction to Statistical Analysis 336 (3d ed.1969).

[91] Defendant's Exhibits 6 and 7.

[92] For an opening as an operator trainee on September 19, 1973, a white employee, E. Calagesi, who had worked at Tenneco since June 7, 1973 was awarded the position over a black bidder D. Harris who had worked at Tenneco since April 12, 1973. Although this discrepancy was not explained and might form the basis of an individual claim of discrimination it can hardly be deemed to show any widespread pattern or practice of discrimination.

[93] Although the position of operator trainee is in the Operations Department, since it is an entry level position only someone in the Maintenance Department wishing to switch departments would ordinarily bid on an opening.

[94] In the case of operator trainee openings, I suspect that this is probably because the blacks with higher seniority occupied better paying positions in the Maintenance Department and were not willing to sacrifice this advantage in order to switch departments. However, since defendant did not supply more specific information on the non-bidders I cannot be certain that this was the case. The Supreme Court has found that a similar seniority system was not discriminatory because "[t]o the extent that it `locks' employees into [certain departments], it does so for all." International Brotherhood of Teamsters v. United States, 1977, 431 U.S. 324, 97 S. Ct. 1843, 1865, 52 L. Ed. 2d 396. It may be of note that the evidence reveals 26 instances in the operator trainee category in which no white employee submitted a bid.

[95] 1977, 431 U.S. 324, 97 S. Ct. 1843, 52 L. Ed. 2d 396.

[96] Id. 97 S. Ct. at 1869.

[97] I. e., after Title VII's effective enactment date. The major holding of this landmark case is that employees who suffered only pre-Act discrimination are not entitled to seniority or back pay relief because § 703(h) of Title VII protects bona fide seniority systems even though they perpetuate pre-Act discrimination. However, in this case no showing has been made of either pre- or post-Act discrimination, so International Brotherhood of Teamsters v. United States is not relevant on that point.

[98] 97 S. Ct. at 1866-67, n. 45, and 1871.

[99] Id. at 1871.

[100] Id. at 1871-72, n.53 (emphasis added).

[101] The Fifth Circuit recently held in James v. Stockham Valves & Fittings Co., 5 Cir. 1977, 559 F.2d 310, that "percentage of requests for specific jobs" by black employees may not be determinative of a claim of discrimination. In that case the district court found that because 61 per cent of white employees and 53 per cent of black employees received the job assignments they requested that there was no discrimination. The Fifth Circuit overturned that finding as "factually insufficient and legally irrelevant" citing four reasons for its conclusion: 1) over two-thirds of the work force was assigned to jobs without request; 2) the statistics did not reflect the number of blacks who were not hired in the first place; 3) the statistics did not reflect requests by blacks for transfers to traditionally white departments; 4) since the employer practiced blatant discrimination, many blacks were no doubt discouraged from making a request in the first place.

I believe that this case is factually distinguishable from the James case on each of these four points, and for that reason absence of bids by blacks is relevant here. First, Tenneco does not seek out current employees for openings. The system operates entirely by voluntary bid on the part of the individual. The only exception to this practice of which there was any evidence has been the occasional efforts of Tenneco to seek out and encourage black employees to bid for promotions. Second, I have already determined that in this case Tenneco's hiring policies and practices are not discriminatory. Third, the statistics here do not reflect any segregation of departments by race. And fourth, plaintiffs have not demonstrated blatant or even subtle discrimination which would discourage blacks from bidding on vacancies and therefore have not carried their initial burden as required by International Brotherhood of Teamsters v. United States, 1977, 431 U.S. 324, 97 S. Ct. 1843, 52 L. Ed. 2d 396. See note 100, supra, and accompanying text.

[102] With the one exception explained in note 92, supra.

[103] James v. Stockham Valves & Fittings Co., 5 Cir. 1977, 559 F.2d 310, suggests that the issue of whether a seniority system is "bona fide" within the meaning of § 703(h) of Title VII is relevant only when pre-Act discrimination is shown. In this case no evidence whatsoever has been adduced of pre-Act practices. However, to the extent that it may be relevant, I believe the seniority system in this case is a bona fide one under the standards set forth in International Brotherhood of Teamsters v. United States, 431 U.S. 324, 97 S. Ct. 1843, 52 L. Ed. 2d 396, since 1) transferees between departments do not lose seniority; 2) the seniority system did not have its genesis in racial discrimination; and 3) the system was negotiated and has been maintained free from any illegal purpose.

[104] Defendant's Exhibits 8 through 12.

[105] The exhibits show that in eleven instances openings could not be filled by current employees because no employee, white or black, possessed the requisite qualifications. The total number of bids for these jobs was 48, of which 13 or 27 per cent were made by blacks.

[106] United States v. Jacksonville Terminal Co., 5 Cir. 1971, 451 F.2d 418, 445, cert. denied 1972, 406 U.S. 906, 92 S. Ct. 1607, 31 L. Ed. 2d 815.

[107] Rodriguez v. East Texas Motor Freight, 5 Cir. 1974, 505 F.2d 40, 56-7, vacated and remanded on other grounds 1977, 431 U.S. 395, 97 S. Ct. 1891, 52 L. Ed. 2d 453. See also, e. g., Myers v. Gilman Paper Corp., 5 Cir. 1977, 544 F.2d 837, 849, appeal pending.

[108] Albemarle Paper Co. v. Moody, 1975, 422 U.S. 405, 95 S. Ct. 2362, 2375, 45 L. Ed. 2d 280 (emphasis supplied).

[109] Smith v. Olin Chemical Corp., 5 Cir. 1977, 555 F.2d 1283, 1287.

[110] Local 189, United Papermakers & Paperworkers, AFL-CIO v. United States, 5 Cir. 1969, 416 F.2d 980, 989, cert. denied 1970, 397 U.S. 919, 90 S. Ct. 926, 25 L. Ed. 2d 100. See also Cotton v. Hinton, 5 Cir. 1977, 559 F.2d 1326.

[111] Defendant's Exhibit 4, Art. 8, Sect. 1.

[112] 15 out of 19. Defendant's Exhibit 17.

[113] See notes 8 and 9, supra, and accompanying text.

[114] See note 55, supra.

[115] W. Dixon & F. Massey, Introduction to Statistical Analysis 249-51 (3d ed. 1969) explains this test.

[116] Defendant's Exhibit 17. There is some dispute as to whether one black employee, E. Labat, should be included in this total since he simply left work one day and never returned. After two days absence he was discharged. If E. Labat is considered as having voluntarily resigned, the total fired during probation is lowered to fourteen.

[117] This figure would be lowered to 43 per cent if E. Labat is not counted as a "discharge".

[118] Defendant's Exhibit 15.

[119] See notes 8 and 9, supra, and accompanying text.

[120] See discussion at p. 241 supra on business necessity and the burden of proof.

[121] Defendant's Exhibit 17.