Robinson v. Polaroid Corp., 567 F. Supp. 192 (D. Mass. 1983)

*193 Jonathan Shapiro, Boston, Mass., for plaintiffs.

Sandra L. Lynch, J. Harold Flannery, Foley, Hoag & Eliot, Boston, Mass., for defendant.

SKINNER, District Judge.

This is an action under Title VII of the Civil Rights Act of 1964 seeking injunctive relief and damages for alleged racial discrimination in the course of a massive layoff of Polaroid employees in 1974. The named plaintiffs are black persons who were laid off from salaried positions at Polaroid. I earlier certified a class consisting of both the salaried and hourly employees and ruled that the named plaintiffs were appropriate representatives of both types of employees. The evidence adduced at trial cast considerable doubt on the propriety of that decision. The layoff method for salaried and hourly employees was significantly different; no hourly employees have ever complained to Polaroid, filed EEOC claims or come forward in this action; the number of salaried employees laid off was 33, insufficiently numerous to justify a class action; and it became clear that the request for class certification was a stratagem to circumvent the failure of all but the named plaintiffs to file administrative claims, a jurisdictional prerequisite to joining as named plaintiffs in this action. I will not decertify the class, however, since the case has been fully tried as a class action. Decertification, furthermore, would not change the ultimate result which I reach in this case.

Polaroid is a manufacturer of cameras, film and related optical and light sensitive products. It operates a number of different facilities, mostly in eastern Massachusetts, with distribution centers in other parts of the country. It is principally noted for cameras which have a self-contained capacity for producing a completely developed photographic print within a few seconds after the picture is taken. This process was developed by Dr. Edwin Land, the founder of Polaroid and Chairman of the Board of Directors until 1982. The development and production of these cameras require Polaroid to maintain a number of different laboratories and a staff of engineers qualified in optics, electrical systems, *194 electronics, light-sensitive materials and quality control and production. It is organized by divisions, which are subdivided into departments, which in turn may be subdivided into groups or teams for specific projects. In addition to technical specialties, these divisions and departments include sales, marketing, accounting, inventory control and all the administrative superstructure of a large corporation.

Some of the departments, however, are larger than some of the divisions. While the organizational nomenclature is traceable in Polaroid administrative documents, most of the employees who testified either used the terms division and department interchangeably or were not clear whether their own organization unit was a division, a department or a group.

In 1969, following the assassination of Dr. Martin Luther King, Dr. Land determined to institute a vigorous affirmative action program at Polaroid, aimed at increasing the minority, and specifically black representation at all levels of the company, including the professional engineering and executive positions. Polaroid recruiters regularly visited black colleges and advertised for black professionals who were seeking advancement. The management encouraged a committee of black workers to express their views of the company's personnel policies with respect to minority employees, and when a group of black professionals resigned, instituted a study to find out the reason. It has maintained a technical training program for black people in Boston called Inner City. This program continues to train black people and place them in local industries, not necessarily Polaroid. It is, however, entirely financed by Polaroid.

In June, 1974, the proportion of black hourly employees was 18.1%, of black salaried employees was 7.1%, and the overall percentage was 14.7%.

During 1974, the company management realized that the lack of market acceptance of some of Polaroid's products required drastic retrenchment and the layoff of many of its employees. The layoff was accomplished between July and December, 1974. Seniority was a determinative factor in the layoff of hourly employees, and had some impact on the layoff of salaried employees. After the layoff, the proportion of black hourly employees was 15.6%, of black salaried employees was 6%, and blacks in the entire work force was 12.5%. Since 1975 the company has increased its work force again, and the proportions by the end of 1978 were 17.2%, 7% and 14.4% for the hourly, salaried and total work force, respectively. These proportions have remained relatively stable to the time of trial.

Layoff of hourly employees was based entirely on seniority, under a company policy which had been in effect ever since the company was organized. Any employee selected for layoff had the right to bump any employee anywhere in the company who had less seniority if he was qualified for the job held by the less senior employee. An employee committee maintained layoff centers at all of the principal Polaroid facilities. Complete records of the seniority status of all employees were available at these centers, and laid-off employees received assistance in bumping. An extensive grievance and appeal procedure was available. Some employees did not exercise bumping rights because opportunities outside the company were superior to the bumped positions. An employee post-audit showed that there were ten employees out of approximately 1,000 who were erroneously laid off out of turn, all of whom were rehired.

No hourly employee filed a grievance with Polaroid on the basis of racial discrimination. No hourly employee filed a complaint with the EEOC. No hourly employee testified in this case in support of the plaintiffs' position.

Plaintiffs' statistical expert testified that the percentage of black employees laid off was significantly greater than would have been expected in a random selection. For *195 purposes of establishing a prima facie case, however, he failed to take into account the factor of seniority, which had an unfavorable impact on blacks. While plaintiffs concede that the statistical significance disappears if seniority is taken into account,^[1] they assert that seniority was not applied neutrally and in good faith. They base this assertion on the fact that there remained in the work force white employees with less seniority than blacks who were laid off. Plaintiffs say that 23% of the blacks who were laid off were laid off out of turn. No comparable figure was given as to whites who were laid off.

In my opinion, the specific evidence as to the layoff machinery and the open employee status information rebuts any inference of racial bias in the operation of the seniority system. That 23% of the black employees who were laid off out of turn is more readily explainable by the likelihood that some did not have the requisite qualifications to bump (the range in skills of hourly employees were from warehouseman to laboratory technician, neither of whom would have been likely to master the job of the other) and that some could find opportunities outside the company which were superior to jobs which they were entitled to bump.

I find that the defendant's system for the layoff of hourly employees was executed without racial bias in accordance with a pre-existing and long standing seniority system. I find that the seniority system was operated with exemplary openness and good faith and that its adverse impact on black employees, who were by and large the most recently hired under the company's affirmative action program, was unintentional and inevitable given the history of the company.

The salaried employees created a more complex problem because their skills were more specialized and diversified, and thus viewed as less interchangeable than the skills of the hourly workers. The top managers assigned layoff quotas to the various divisions, the managers of which assigned quotas to the departments. In some cases, departmental quotas were parcelled out among groups or teams. The assignments of quotas was based on the expected future operations of the company and management's assessment of what functions would be required to keep the company going. Initial layoff nominations were made by the manager of each operational unit after consultation with his senior supervisors. The operational unit was sometimes a department, sometimes a group or team, occasionally a small division. If the operational unit was a group or department, the layoff decisions were reviewed by the manager and senior supervisors of the next higher organizational level.

Finally, each divisional manager was required to justify layoff decisions to the Operational Committee, the principal management group of the company. Layoff of each member of a racial minority was specifically reviewed on an individual basis. Changes were ordered within the divisions in twenty or thirty cases, all of which were in favor of blacks.

The company established criteria for the layoff of salaried employees in a document called Personnel Policy No. pp-250A, the key paragraph of which is as follows:

By necessity the process for an exempt layoff must be less structured. Members in exempt classifications are not necessarily interchangeable. Each situation is different. Prior to a layoff decision, both job requirements and personal factors which should be considered include the particular knowledge and qualifications of the affected members and their seniority, past performance, future potential and ability to locate elsewhere.

*196 Other parts of pp-250A provided in effect that anyone with ten years of service as a salaried employee was exempt from layoff.

The quoted criteria were not uniformly applied throughout the company. For instance, some managers considered an employee's seniority as a salaried employee while others took into account prior service as an hourly employee. The several criteria were weighted differently by different managers. In general, managers of departments and divisions rated their employees comparatively and made pragmatic decisions on the basis of which employee they could lose with the least loss of function. Efficiency and the possession of particular skills thought to be required in the immediate future weighed heavily in these decisions.

I agree with plaintiffs' personnel expert that the criteria called for subjective evaluation of salaried employees by management. Application of these criteria could potentially mask racial bias. I find, however, that collective decision-making and the elaborate review system established by the company provided some control over misuse of the 250A criteria. How effective these controls were is difficult if not impossible to determine on a case by case basis, however. If they were not effective in a general way, the existence of racial bias in the layoff process should show up in statistical analysis, which thus becomes crucial, not only for the class but also for the named plaintiff.

I further find that the ten-year rule was basically arbitrary, was without historic basis and was not justified by any compelling interest of the company that couldn't be served by making seniority one of the general criteria in layoff decisions.

All of the named plaintiffs and plaintiff class established a prima facie case by showing that they were black and laid off while white employees were not.

The defendant introduced evidence that each named defendant and class member had been carefully evaluated in comparison with other employees in the division or department. In each case there was credible evidence, most of which was corroborated by personnel records, that other employees retained had superior performance records or more needed skills than the plaintiffs. The plaintiffs' response was to dispute these evaluations. Some of the decisions were extremely close judgment calls, as in the case of Charles Roane, a member of the class but not a named plaintiff. Neither individually nor as a class, however, did the plaintiffs establish by a preponderance of evidence that the justification for layoff was pretextual or a sham or other than a good faith exercise of professional judgment.

There was indeed some evidence that some people at Polaroid were biased against blacks, but the evidence did not establish a probability that there was a causal relationship between the instances of bias and the decisions to lay off the plaintiffs.

For instance, Mr. Curington testified to a racial slur written on the blackboard, but there is no evidence as to who wrote it. He also testified that a department manager named George Howard told a joke which was insulting to blacks. George Howard was not Curington's supervisor and had nothing to do with Curington's layoff. Howard testified that he did not even know Curington.

Howard is a significant figure, however, with respect to the plaintiff Gerald Smith. Howard hired Smith initially and assigned him to various supervisors within the marketing services divisions. Each of these supervisors found Smith's work unsatisfactory. Eventually Howard gave him a special assignment as manager of performance analysis reporting directly to him. It was the testimony of Smith's supervisors that they viewed Howard as, in fact, giving *197 Smith special consideration. Smith's view was, however, that he was unable to do better work only because Howard failed to provide him with necessary staff assistance and that this was the cause of his eventual layoff from his final position. The layoff decision was not made by Howard, but Smith alleges that Howard's bias prevented him from doing his job properly and this was the indirect cause of his layoff.

All of his supervisors were unanimous in their testimony that Smith could not even get started on the projects assigned to him. He never took notes when his assignments were explained to him. The evidence, furthermore, does not show any bias against him by Howard, who apparently went out of his way to give Smith one chance after another. There was some suggestion that Howard was a demanding manager and insensitive in his language, but no evidence of any bias, racial or otherwise, in his treatment of Gerald Smith.

Robert L. Davison, a class member but not a named plaintiff, testified as to three purported examples of racial bias. A supervisor of Chinese descent invited Davison to join a group at lunch and said, "We are both colored, but the trouble is you're the wrong color." The manager of Davison's department made a reference to "incompetent engineers". Davison assumed that the reference was to himself and that the manager never made such comments about white employees. I find it extremely doubtful that this remark was motivated by racial bias, even if it was intended to apply to Davison. A supervisor named Harrison plagued Davison with constant criticism. Harrison testified that he liked Davison. I do not believe this testimony. Harrison disliked Davison, but I find that this was because of specific conduct by Davison and not because of race. I do not find any support for Davison's assertion that he was the victim of racial bias.

One further incident perhaps deserves comment. Richard DeLima, the secretary and general counsel, laid off a black lawyer named William Hill even though he had no layoff quota to meet. In fact, it is obvious that DeLima could hardly wait after layoffs were announced to send Hill packing. He says that he was motivated solely by a desire to aid the company in its time of need. He may actually believe this, but I do not. Hill's prior experience was principally in constitutional litigation and criminal defense. Before coming to Polaroid, Hill had no experience in corporate work. Both men testified. I have no trouble envisaging mutual personal and professional antipathy. While the precipitate removal of Mr. Hill certainly raises a suspicion of racial bias, I cannot say on this evidence that racial bias is the probable cause of Mr. Hill's discharge.

Of course, if there were evidence of discriminatory treatment of either Mr. Davison or Mr. Hill which did not affect the whole class, no consequences would flow, since neither satisfied the jurisdictional prerequisites for individual relief.

In any case, there is no evidence of any identifiably discriminatory treatment of any of the named plaintiffs or of the salaried employees in the certified class.

The ten-year rule, however, although facially neutral, had a potentially discriminatory impact on the plaintiff class because of the historical seniority patterns at Polaroid. I find and rule that there was no business justification for the rule that was not satisfied by the general inclusion of seniority as one of the criteria to be used in layoff decisions in pp-250A. In fact, however, the ten-year rule did not have any impact on any of the named plaintiffs or the plaintiff class. There is no evidence that any of the plaintiffs were laid off because of the rule or would have escaped layoff if the rule did not exist.

Even though there is no evidence of specific discriminatory intent and the evidence on discriminatory impact revealed potential *198 discrimination but no actual impact, Polaroid's layoff criteria were so subjective in their application that they could mask tacit racial discrimination. This is true even with Polaroid's elaborate system of review. The management personnel doing the reviewing were, after all, still predominantly white, despite Polaroid's affirmative action program.

The plaintiffs' statistician, Dr. Goldstein, testified that the statistics revealed a greater number of blacks discharged than would have been the case if the layoff system had been racially neutral. The defendant's statistician, Dr. Michelson, testified that there was no statistically significant bias against blacks, and in some respects, blacks were statistically favored.

Testing of statistical evidence is difficult at best and has been rendered nearly impossible by virtue of the use of computer programs, the critical characteristics of which are unknown. Dr. Goldstein used a commercially available statistical program called BMDP. Dr. Michelson used a proprietary program called MULQUALS, devised by his own consulting firm.

Dr. Goldstein testified first in the plaintiffs' case-in-chief, a second time on rebuttal, and a third time as the court's witness. Dr. Michelson testified in the defendant's case-in-chief and in sur-rebuttal.

In their first go-round, the different opinions of the two experts resulted from the fact that Dr. Goldstein performed an unconditional logit, that is, a comparison based on significant factors between the persons laid off and other employees on a company-wide basis. Dr. Michelson performed a conditional logit in which the comparison was made between persons laid off and other employees in the same department or division. In my view, Dr. Michelson's system most nearly reflects the actual facts as revealed by the evidence.

There were also differences in the data base resulting from the fact that each expert counted the population of Polaroid employees over a slightly different time period. I am unable to determine which data base was more accurate. Since the plaintiffs have the burden of proof, to the extent that their case depends on the superiority of Dr. Goldstein's data base, they lose.

In the second round, Dr. Goldstein testified that he had redone his study using a conditional logit like Dr. Michelson's and had run it on BMDP. He still came to the conclusion that the statistics revealed a bias against blacks. He described his procedure in some detail.

In his second appearance, Dr. Michelson testified that he had also rerun the conditional logit using the BMDP program, which he had obtained. He came up with virtually the same result that he had reached with MULQUALS. He testified that Goldstein had reached a different result because he had given the BMDP program the wrong instruction. According to Dr. Michelson, BMDP is a program designed to test the effect of drugs on the survival rates of the people who take them. The program makes certain assumptions about death and survival which must be taken into account in feeding in the data. Failure to take these programmed assumptions into account result in a serious distortion of the result. Dr. Michelson testified very positively that Dr. Goldstein had made this error in running the program.

I then recalled Dr. Goldstein as the court's witness and asked him if he had made the error which Dr. Michelson described. In reply, he said among other things, that it was not completely clear that BMDP did what he thought it did. With this statement on the record, there is no way that I could find that the plaintiffs had sustained their burden of proof with respect to the statistical evidence.

In fact, I find that Dr. Michelson's analysis is probably correct. Since subjective evaluation of performance was not included as a variable, I conclude that subjective evaluation of performance did not skew the statistically predictable result. I infer, *199 therefore, that the performance evaluation was conducted in a racially neutral manner.

Accordingly, I find no evidence of discriminatory treatment or discriminatory impact in Polaroid's layoff program, either as applied in particular cases or viewed overall.

Plaintiffs allege that defendant treated black salaried employees less favorably than white salaried employees in the layoff process. The following analysis governs this "discriminatory treatment" claim:

First, the plaintiff has the burden of proving by the preponderance of evidence a prima facie case of discrimination. Second, if the plaintiff succeeds in proving the prima facie case, the burden shifts to the defendant to articulate some legitimate, nondiscriminatory reason for the employee's [layoff] ... Third, should the defendant carry this burden, the plaintiff must then have an opportunity to prove by a preponderance of the evidence that the legitimate reasons offered by the defendant were not its true reasons, but were a pretext for discrimination...

Texas Department of Community Affairs v. Burdine, 450 U.S. 248, 252-253, 101 S. Ct. 1089, 1093, 67 L. Ed. 2d 207 (1981); McDonnell Douglas Corp. v. Green, 411 U.S. 792, 93 S. Ct. 1817, 36 L. Ed. 2d 668 (1973). Plaintiffs retain the ultimate burden of persuading the trier of fact that defendant intentionally discriminated against them. Texas Department of Community Affairs v. Burdine, supra at 253, 101 S. Ct. at 1093.

The "ten year rule" is analyzed under a different legal rubric. This rule is a facially neutral policy that allegedly had an adverse impact on blacks. It is appropriately analyzed under the "disparate impact model" articulated by the Supreme Court in Griggs v. Duke Power Co., 401 U.S. 424, 91 S. Ct. 849, 28 L. Ed. 2d 158 (1971). See also Dothard v. Rawlinson, 433 U.S. 321, 328-329, 97 S. Ct. 2720, 2726-27, 53 L. Ed. 2d 786 (1977). Under Griggs, plaintiffs need not prove that defendant intended to discriminate against them when promulgating the policy. Plaintiffs' prima facie case under this theory consists of establishing that the ten year rule had a discriminatory impact upon blacks.

Plaintiffs cite Patterson v. Greenwood School District 50, 696 F.2d 293 (4th Cir. 1982) for the proposition that black salaried members were discriminated against by virtue of the ten year rule even if they would still have been laid off without the rule. Patterson, however, was a discriminatory treatment case where plaintiff proved that defendant discriminated against female applicants in promotion decisions. It is inapposite to a discriminatory impact situation where a critical element of plaintiff's case is a showing that the challenged rule actually had an adverse impact upon the plaintiff class.

Applying the foregoing rules of law to the facts which I have found, I conclude that while the plaintiffs proved a prima facie case, they were unable on any theory to satisfy their overall burden of proof. Accordingly, I find for the defendant and order judgment to be entered for the defendant with costs.

[1] In the initial selection by the company of employees for layoff, blacks were actually favored. The operation of the bumping system wiped out their statistical advantage, but the result was statistically consistent with racial neutrality given the seniority factor.