Williams v. New Orleans Steamship Ass'n (673 F.2d 742, 1982)

JERRE S. WILLIAMS, Circuit Judge:

Plaintiffs George Williams, Duralph Hayes, and Ernest Turner, Jr. and intervenors Matthew Richard and John Aaron sued New Orleans Steamship Association (NOSA), sixteen of its member stevedoring companies, Locals 1418 and 1419 General Longshore Workers, International Longshoremen Association (ILA), and Locals 1802 and 1863 Sacksewers, Sweepers, Waterboys, and Coopers, ILA, alleging individual and class-wide employment discrimination in the Port of New Orleans in violation of Title VII of the Civil Rights Act of 1964, 42 U.S.C. § 2000e et seq., and 42 U.S.C. § 1981. After an eighteen day trial, the court ordered the merger of the previously segregated Locals 1418 and 1419 and Locals 1802 and 1863 but dismissed the remainder of plaintiffs’ claims and refused to certify a class pursuant to Fed.R.Civ.P. 23. In this appeal, plaintiffs contest the court’s dismissal of their claims of discrimination in the allocation of grain work and in the assignment of preferred positions in general cargo gangs as well as the court’s refusal to certify a plaintiff class. We hold that the court erred in denying plaintiffs’ claim concerning the allocation of grain work and in refusing to certify a class with respect to that claim. We affirm the dismissal of the discriminatory assignment claim and the court’s refusal to certify a class in that instance.

I. FACTS

The district court’s opinion amply describes the factual background of this eleven year old case. See Williams v. New Orleans Steamship Association, 466 F.Supp. 662 (E.D.La.1979). Because this appeal concerns the employment practices in particular areas of the stevedoring industry, it is necessary to describe briefly the operation of the industry in the Port of New Orleans.

Originally, employment on the waterfront was casual and unsystematic. After a 1974 Department of Labor study, a registration system was instituted. Today, waterfront workers are categorized by craft and registered accordingly. Each craft is under the jurisdiction of a different ILA local.

This appeal involves alleged discrimination against Craft I workers. Craft I is comprised of general longshoremen previously under the jurisdiction of Locals 1418 (white) and 1419 (black), now under the jurisdiction of the merged Local 3000. Within each craft, workers are classified according to their priority for employment opportunities based upon their length of service with a particular company. The highest ranking categories are considered “registered” (as opposed to “casual”) and comprise the majority of longshoremen. Throughout the relevant time period, approximately 75% of all registered longshoremen were black. Thus membership in Local 1419 was three times as great as that of Local 1418.

Longshoremen load and unload ships. The work is performed on a day-to-day basis, seven days a week. Stevedoring companies hire longshoremen daily through the Waterfront Employment Center owned and operated by NOS A. The workers are organized into “gangs” of varying size, depending on the type of cargo to be loaded or unloaded. Typically, grain gangs employ eight men and general cargo gangs sixteen. Pay rates for the various jobs are set out in the Deep Sea Agreement negotiated between NOSA and the Locals.

Hiring is done twice daily at “shape-ups.” Each stevedore chooses a foreman for each gang needed that day and the foreman then hires the necessary gang; members. Many companies have “regular” gangs comprised of registered longshoremen who frequently work for that company and are given preference on available work. If a company has more work than can be handled by its regular gangs, it employs “non-regulars” who are either regulars with other companies or “casuals” not associated with any particular company. This results in all longshoremen working for virtually all the stevedores at one time or another. Although many workers become associated with a particular company from time to time, none works exclusively for one company.

Despite the registration system, employment is still largely casual. The work performed by a longshoreman varies from day to day depending on a variety of factors including the work available as well as personal choice. Longshoremen are not required to work any particular days or hours. If they want to work, they simply come to a shape-up. The rates of pay vary for different jobs and different shifts. For example, in 1973, grain work paid a 20$ per hour premium. Work performed during mealtimes, weekends, and holidays also pays a premium. A longshoreman may choose to shape-up only when there is a certain type of work available or only during certain hours. Thus the longshoreman’s job and wages are determined in part by personal factors.

The employment relationship ultimately is controlled by the Deep Sea Agreement which specifies the terms and conditions of employment and the varying wages for each job. In this appeal, plaintiffs contest the employment practices concerning the allocation of grain work and the job assignments in general cargo gangs. Before 1974, the Agreement required that so far as practicable, grain work should be divided evenly between Locals 1418 and 1419, notwithstanding their disproportionate memberships. There was no corresponding provision respecting assignments in general cargo gangs; plaintiffs contest an alleged practice by white foremen of assigning white longshoremen to the more preferable jobs.

II. ALLOCATION OF GRAIN WORK

A. Separate Claim of Racial Discrimination in Grain Work.

Grain work comprises approximately 8% of all longshore work and involves the loading and unloading of ships carrying grain. It is physically less demanding than other types of work because the cargo is pumped rather than carried onto and off of vessels. Grain workers often become covered from head to toe in. grain and inhale particles, however, and because of this unpleasantness, a premium is paid for grain work.

Pursuant to the Deep Sea Agreement, grain work was allocated equally between the members of the black and white locals. Plaintiffs contend that this practice violated Title VII and § 1981.

The district court’s original order in this case, dated February 14, 1979, 466 F.Supp. 662, was five years after the trial. In that order, the court found that the 50-50 allocation of grain work was discriminatory, yet it did not find a violation of Title VII. Upon plaintiffs’ motion for reconsideration of the issue, the court clarified by saying its earlier holding meant that the contract clause requiring 50-50 allocation was not discriminatory but that it was evidence of the discriminatory potential inherent in a system that maintained segregated local unions.

The court then reconsidered plaintiffs’ allegation that the practice of 50-50 allocation was discriminatory in and of itself. Again, the court did not find a violation of Title VII. This time its conclusion was based upon its finding that, because longshoremen perform more than one type of work, evidence focusing on only one of these jobs, Le., grain work, would distort the analysis. “The important thing is how [the black longshoreman] fares overall.” The court then considered the overall welfare of black longshoremen and concluded that in the Port of New Orleans, they received their proportionate share of work and pay. Thus the court found no discrimination.

An appellate court must accept a trial court’s findings of subsidiary facts unless clearly erroneous; however, the court’s ultimate finding on the issue of discrimination is subject to review free of the clearly erroneous standard. Wright v. Western Electric Co., 664 F.2d 959, 963 (5th Cir. 1981). Furthermore, this Court is not bound to any degree by the district court’s conclusions as to the law, Parson v. Kaiser Aluminum & Chemical Corp., 575 F.2d 1374, 1382 (5th Cir. 1978), cert. denied, 441 U.S. 968, 99 S.Ct. 2417, 60 L.Ed.2d 1073 (1979), nor are we bound by findings of fact based upon erroneous applications of law, Johnson v. Uncle Ben's, Inc., 628 F.2d 419, 422 (5th Cir. 1980), vacated on other grounds, 451 U.S. 902, 101 S.Ct. 1967, 68 L.Ed.2d 290 (1981). We find that the district court’s opinion was based upon an erroneous legal conclusion as to the viability of plaintiffs’ grain claim. Thus our review of the court’s findings of fact is not bound by the clearly erroneous standard of Fed.R.Civ.P. 52(a).

In concluding that the plaintiffs failed to prove discrimination, the district court misconstrued plaintiffs’ allegation. By comparing the overall economic picture of blacks and whites, the court transformed plaintiffs’ claim of discrimination in the allocation of grain work into a claim of discrimination in the entire industry. The court erroneously concluded that plaintiffs’ segmented claim was not cognizable under Title VII and § 1981.

Claims alleging discrimination in only one segment of an employer’s workforce are cognizable under Title VII and § 1981. The Supreme Court recognized this principle in one of the landmark Title VII cases, International Brotherhood of Teamsters v. United States, 431 U.S. 324, 97 S.Ct. 1843, 52 L.Ed.2d 396 (1977). There, the United States brought suit on behalf of blacks and Spanish-surnamed persons against a large trucking company and the union which represented its employees claiming that the company engaged in a pattern or practice of discrimination in the hiring of line drivers. Minorities who were hired allegedly were relegated to the lower paying, less desirable positions and met wjth insurmountable difficulties in trying to gain promotions and transfers.

The government’s case consisted of powerful statistical evidence showing a gross disparity between the percentage of minorities in the company’s workforce as a whole and those in line driver positions. Those statistics, coupled with the testimony of individuals who recounted over forty specific instances of discrimination, established the government’s prima facie case. The company was unable to rebut this evidence, and the government prevailed on the issue.

Line drivers only comprised approximately one-fourth of the truck driver positions, yet the Court recognized the validity of the departmentalized claim. The Court did not look to the overall picture of minorities within the company’s employ to determine the validity of the claim. Minorities may have fared overall as well as their white counterparts, yet the Court did not consider this factor in evaluating the evidence. The government was permitted to charge discrimination in one job classification only and to prove its case by offering statistical evidence relevant only to that job.

In determining whether a particular claim of discrimination in only a segment of an employer’s work force is cognizable the operative factor must be the distinctiveness of the segment. In Teamsters, supra, the duties and compensation of line drivers differed from those of other driver positions and thus line drivers constituted a distinct job classification.

A similar distinction exists between grain work and other types of longshoring jobs in New Orleans. The tasks performed by grain workers differ from those performed by other longshoremen because of the nature of the cargo involved. Furthermore, the Locals and NOSA traditionally have treated grain work separately from other types of longshore work, as evidenced by the Deep Sea Agreement. The hourly wage paid for grain work is negotiated separately from other types of work. The mode of allocating jobs evenly between the black and white Locals was also unique to grain work. Thus, while longshoremen can and do perform a variety of jobs, the distinctiveness of grain work makes plaintiffs’ claim of discrimination in this one area of employment viable under Title VII and § 1981.

In fact, the court below recognized the validity of segmented claims within the industry in New Orleans. Although it refused to consider plaintiffs’ claims of discrimination in the allocation of grain work, it permitted similar segmented claims with respect to other sectors of the industry. A clear example is the court’s consideration of plaintiffs’ allegation of discrimination in job assignments in general cargo gangs. Although the court ultimately rejected plaintiffs’ claim on the merits, it recognized the segmented claim and considered statistics relevant only to general cargo gangs, not the entire industry.

Furthermore, both logic and the policies underlying Title VII dictate the conclusion that a segmented claim is cognizable. Title VII proscribes racial discrimination of any kind in virtually all aspects of employment. There is nothing in the Act to indicate that a- claim for discrimination in one distinct category of employment within a plant, industry or office, or even in one job, is not cognizable.

Were we to accept defendants’ contentions, we would be ruling in effect that an employer may discriminate with impunity as long as it confines such action to some units or departments within the workforce while maintaining an overall facade of equality. An employer guilty of plant or industry-wide discrimination may be a worse offender of Title VII than one who discriminates in only a single department. The remedies awarded to the former’s employees as well as the penalties assessed against it may be greater than those imposed upon the latter. As to liability, however, both are equally subject to the prohibitions of Title VII.

The cases relied upon by defendants in support of the judgment below are unpersuasive. None of these cases involved the question of whether a segmented claim may be brought against an employer. In Swint v. Pullman-Standard, 539 F.2d 77 (5th Cir. 1976), cert. granted (after an earlier remand), 451 U.S. 906, 101 S.Ct. 1972, 68 L.Ed.2d 293 (1981), plaintiffs alleged that the defendant engaged in a pattern or practice of racial discrimination in departmental assignments within its production plant. To prove their claim, plaintiffs offered statistical evidence of racial disparity in .nine of the twenty-eight departments. We discarded the proffered analysis because it focused on only nine of twenty-eight depártments. 539 F.2d at 94-95. Instead, we held the proper statistical focus should have encompassed the entire plant and thus plaintiffs’ evidence of racial imbalances in a segment of the plant was insufficient to establish a prima facie case of plant-wide discrimination.

In E.E.O.C. v. Datapoint Corp., 570 F.2d 1264 (5th Cir. 1978), plaintiffs raised a claim of racial discrimination in hiring and job assignments. They introduced statistical evidence showing gross imbalances in some job categories, but no evidence as to the employer’s workforce as a whole. The district court rejected this as proof of a prima facie case. We upheld its decision stating that while “plaintiffs’ statistical evidence alone might be sufficient to infer discrimination in these job categories ... [to prove] ■ a prima facie case of racial discrimination, such statistics must be relevant, material and meaningful. .. . ” 570 F.2d at 1269. Because plaintiffs’ claim was one of employer-wide discrimination, the segmented, isolated statistics offered were misleading and therefore meaningless.

Contrary to defendants’ contention, these two cases do not stand for the proposition that plaintiffs’ claim of discrimination in grain work is not cognizable. In both cases, plaintiffs charged the employer with plant-wide discrimination but offered statistical evidence which was limited to small segments of the plant. In denying plaintiffs’ claims, we held that the proffered limited statistics were insufficient to prove the broad claims asserted.

Here, plaintiffs alleged discrimination in one separate job category within the industry. Naturally, then, their statistics focused on the subject area. Unlike the Swint and Datapoint plaintiffs, they did not offer a segmented statistical analysis as a means of trying to prove industry-wide discrimination. Indeed, that was not their claim. The district court, nevertheless, treated their offer of proof as one for industry-wide discrimination. As such, plaintiffs’ statistics certainly would not prove a prima facie case, for the reasons cited in Swint and Datapoint. The court refused to consider plaintiffs’ proof in the context in which it was offered, and we find this to be reversible error.

Lee v. City of Richmond, 456 F.Supp. 756 (E.D.Va.1978) and Croker v. Boeing Co., 437 F.Supp. 1138 (E.D.Pa.1977), aff’d, 662 F.2d 975 (3d Cir. 1981), also relied upon by defendants, are equally unsupportive. . Both cases, as do our decisions set out above, stand only for the proposition that segmented statistical evidence of racial imbalances within a few departments cannot prove a prima facie case of employer-wide discrimination. Neither of these cases addresses the validity of a claim of segmented, departmentalized discrimination such as the one posed by plaintiffs herein. Therefore, they do not support defendants’ position in this' appeal.

We hold that the district court erred in refusing to recognize plaintiffs’ claim and in transforming it into one of industry-wide discrimination. Plaintiffs presented a cognizable claim under Title VII and § 1981. We now turn to the merits of that claim.

B. The Prima Facie Case

Having found plaintiffs’ claim for discrimination in the allocation of grain work to be cognizable under Title VII and § 1981, we turn now to plaintiffs’ evidence to determine whether a prima facie case was made. We note at the outset that the contract clause requiring the 50-50 allocation of grain work was eliminated in 1974. Because the clause formed a significant part of plaintiffs’ evidence, our discussion in this section pertains only to the time period in which the objectionable clause was in effect.

Plaintiffs’ claim is one of class-wide disparate treatment. Thus, the inquiry must be whether defendant treated members of the plaintiff class differently from their white counterparts, and if so, whether that difference was the result of discriminatory intent. In some cases, proof of the latter can be inferred from strong statistical evidence of the former. Teamsters, 431 U.S. at 335 n.15, 97 S.Ct. at 1854 n.15; Pouncy v. Prudential Insurance Company of America, 668 F.2d 795 at 802 (5th Cir. 1982); Wilkins v. University of Houston, 654 F.2d 388, 395 (5th Cir. 1981). Plaintiffs introduced statistical evidence which they contend proved a significant disparity between blacks and whites in the allocation of grain work.

Statistics have become an important and useful tool in civil rights litigation. In many cases, they are the only evidence available to plaintiffs. Pragmatically, too, statistics assist in the evaluation of employer practices because “absent explanation, it is ordinarily to be expected that nondiscriminatory . . . practices will in time result in a work force more or less representative of the racial and ethnic composition of the population ... from which employees are hired.” Teamsters, 431 U.S. at 339 n.20, 97 S.Ct. at 1856 n.20. Thus, “[wjhere gross statistical disparities can be shown, they alone may in a proper case constitute prima facie proof. . . . ” Hazelwood, 433 U.S. at 307-08, 97 S.Ct. at 2741. Wilkins, 654 F.2d at 395.

Plaintiffs introduced statistical evidence showing the allocation of grain work in ten of the NOS A companies for the years 1971-73. In 1971, there were six grain gangs composed of 31 whites and 32 blacks; in 1972, there were 19 gangs composed of 72 whites and 76 blacks; and in 1973 there were 18 gangs composed of 67 whites and 71 blacks. Thus while blacks comprised 75% of the registered workforce, they held only 51.3% of the grain positions. A calculation of the binomial distribution, see Castaneda v. Partida, 430 U.S. 482, 97 S.Ct. 1272, 51 L.Ed.2d 498 (1977), reveals that the actual number of black grain workers deviated from the expected number by 10.24 standard deviations. This is well sufficient to prove a prima facie case of discrimination. Id. at 496 n.17, 97 S.Ct. at 1281 n.17.

In addition to this strong statistical data, plaintiffs introduced into evidence the contract between the Craft I unions and NOSA which specifically compelled the allocation of grain work according to race. It provided that “so far as is practical, work is to be divided between members of ILA Locals 1418 and 1419 in grain trimming machine gangs, and/or hand trimming gangs.” Williams v. NOSA, 466 F.Supp. at 673. In practice, this meant that grain gangs were half black and half white. Id.

The contract clause and the statistics are powerful evidence of purposeful discrimination. The statistics alone show that it is quite unlikely that random, impartial hiring practices would have produced such disparities. And, in light of the contract’s compelled discrimination, any doubts as to purposefulness must be resolved in plaintiffs’ favor. We hold that plaintiffs established a prima facie case of purposeful discrimination in the allocation of grain work.

C. Defendants’ Rebuttal

Once plaintiffs have proved a prima facie case of purposeful discrimination, defendants must rebut plaintiffs’ case by discrediting plaintiffs’ evidence or providing a “nondiscriminatory explanation for the apparently discriminatory result.” Teamsters, 431 U.S. at 360 n.46, 97 S.Ct. at 1876 n.46. If the defendants fail to rebut the prima facie case, plaintiffs prevail on the issue.

At no point in this eleven year proceeding have defendants disputed the validity of plaintiffs’ statistics relating to the allocation of grain work. In fact, the statistical data was derived from the defendant companies’ Answers to plaintiffs’ Interrogatories. Defendants challenged the designation of black grain workers as the sample and argued that the analysis should have focused upon a 11 black longshoremen. We responded to this contention in Part I, supra and need not reiterate our holding.

Instead, defendants offered explanations for the statistical disparity which they contend satisfied their rebuttal burden. First, they contend that because the offensive clause was included at the insistence of black Local 1419, in following its mandate they were only following the wishes of those who now claim to have been discriminated against. This appears to be factually correct; nevertheless, it does not excuse discrimination on the part of the union or the employer. “The rights assured by Title VII cannot be bargained away — either by a union, by an employer, or by both acting in concert.” United States v. St. Louis-San Francisco Railway Co., 464 F.2d 301, 309 (8th Cir. 1972) (en banc), cert. denied, 409 U.S. 1116, 93 S.Ct. 913, 34 L.Ed.2d 700 (1973). We have held “[t]he fact that a contract required this limitation [of employment opportunities] is not a justification if it is discriminating.” United States v. Hayes International Corp., 456 F.2d 112, 117 (5th Cir. 1972). Clearly, then, neither the existence of the contract nor its origin can exonerate defendants’ discriminatory practices.

Second, defendants claimed that because blacks in the longshore industry fared overall as well as or better than whites, any misallocation of grain work is irrelevant. As we stated above, this type of analysis transforms plaintiffs’ claim into one of industry-wide discrimination. Whether blacks as a whole suffered economically has no bearing on the issue of discrimination in the allocation of grain work. Swint, 539 F.2d at 92. Plaintiffs readily admit that industry-wide earnings of blacks may be relevant on the issue of backpay. This evidence has no place in the liability phase, however. Id. Defendants also argue that grain work was no more desirable than any other type of work. Although it paid a 20