Wetzel v. Liberty Mut. Ins. Co., 449 F. Supp. 397 (W.D. Pa. 1978)
April 17, 1978
LIBERTY MUTUAL INSURANCE COMPANY, a corporation, Defendant.
United States District Court, W. D. Pennsylvania.
*398 Howard A. Specter, Litman, Litman, Harris & Specter, Pittsburgh, Pa., for plaintiffs.
Kalvin M. Grove, Lederer, Fox & Grove, Chicago, Ill., for defendant.
Robert A. Penney, Boston, Mass., for defendant (corporate counsel).
Clem R. Kyle, Pittsburgh, Pa., for defendant (local counsel).
Milton A. Smith, Gen. Counsel, Richard B. Berman, Chamber of Commerce of the United States of America, Washington, D. C., Gerard C. Smetana, Laurence D. Ehrlich, Julian D. Schrieber, Chicago, Ill., Walter P. DeForest, Pittsburgh, Pa., amicus curiae, defendants.
Robert F. Stone, Pittsburgh, Pa., for plaintiffs' co-counsel.
Equal Employment Opportunity Commission, Washington, D. C., James F. Bennett, *399 Equal Employment Opportunity Commission, Pittsburgh, Pa., amicus curiae, plaintiffs.
WEBER, Chief Judge.
In prior proceedings we have determined that the defendant discriminated against women technical employees in its claims department in hiring and promotional opportunities. We now consider whether the pay differential between the positions of Claims Adjuster (CA) and Claims Representative (CR) amounts to sex discrimination under Title VII of the Civil Rights Act of 1964, 42 U.S.C. § 2000e et seq.
Title VII became effective on July 2, 1965. In that same year the defendant insurance company created the position of Claims Representative in its claims department. From that date until sometime in 1969, these positions were filled exclusively by females. Prior to 1965 the functions which were performed by claims representatives in this new job classification had been performed by Claims Adjusters who were exclusively male. After 1965, the job title of Claims Adjuster continued to exist and was filled exclusively by males until early 1971, after the filing of administrative charges by the named plaintiffs herein, when the first female Adjusters were hired by Defendant.
There is no dispute that the male CAs were paid substantially more than were the female CRs through the 1965-1971 period. At a typical point in the period material to this question, the difference in the average annual pay was $2,700; the female CRs were being paid $5,200/year while the male CAs were receiving $7,900. The male adjusters also had the use of a company automobile and had the opportunity to earn extra pay for occasional weekend duty.
I. The Issues as Framed by the Parties
The complaint alleges that the Defendant is guilty of discrimination against women with respect to compensation in violation of Title VII of the Civil Rights Act of 1964. It makes no reference to the Equal Pay Act of 1963, 29 U.S.C. § 206(d) (1), but the allegation that the female CRs performed work equal to that done by the male CAs for less pay does state a cause of action thereunder. Both parties have referred to this question throughout the preparatory and trial phases of this case as the Equal Pay issue and have thoroughly briefed the question of whether Plaintiffs have established that these two job classifications actually involved "substantially equal" work. Plaintiffs, of course, contend that they have clearly established the following facts:
"26. Claims representatives exercise the same degree of skill, effort and responsibility as do claims adjusters.
27. Claims representatives perform their duties under similar or less favorable working conditions than do claims adjusters."
Furthermore, Plaintiffs contend that
"Where an employer segregates jobs by sex in violation of Title VII and assigns female employees to the lower paying jobs, a prima facie showing of discrimination in compensation is shown without regard to the relative differences or similarities between the higher and lower paying jobs,"
and that, without regard to whether these two jobs were "equal" Liberty Mutual violated Title VII by considering the sex of its *400 female employees in establishing the compensation of the Claims Representatives at a lower rate than that of the Claims Adjusters.
Liberty Mutual counters by arguing that the position of CA and CR involved substantially dissimilar work and that the difference in compensation for these groups of employees was based on factors other than sex. According to Liberty Mutual, Title VII affords Plaintiffs no relief with respect to discrimination in compensation other than that afforded by the Equal Pay Act. Therefore, if the jobs of CA and CR were not equal within the meaning of the Equal Pay Act, Plaintiffs may not recover even though they have established that one of Defendant's reasons for paying the CRs less than the CAs was the sex of the former group's members. In addition, Defendant argues that each of its branch offices constitutes a separate "establishment" within the meaning of the Equal Pay Act and that Plaintiffs have failed to show that these jobs were equal at each, or indeed at any single, branch office.
II. The Relationship Between Title VII and Equal Pay Act.
Since the hearing on this case, several courts have ruled on the relationship between Title VII's prohibition of discrimination "against any individual with respect to his compensation . . .", 42 U.S.C. § 2000e-2(a) (1), and the Equal Pay Act. The problem revolves around the interpretation of the following statutory provision:
"Notwithstanding any other provision of this title, it shall not be an unlawful employment practice for an employer to apply different standards of compensation . . . provided that such differences are not the result of an intention to discriminate because of . . . sex . . . It shall not be an unlawful employment practice under this title for any employer to differentiate upon the basis of sex in determining the amount of the wages or compensation paid or to be paid to employees of such employer if such differentiation is authorized by the provisions of Section 6(d) of the Fair Labor Standards Act of 1938, as amended (29 U.S.C. § 206(d))," 42 U.S.C. § 2000e-2(h). (emphasis added)
The emphasized portion of the section quoted above is commonly referred to as the Bennett Amendment; it is now the task of this Court to determine the effect of this sentence on Plaintiffs' Title VII claim. Those courts which have ruled on the effect of the Bennett amendment have uniformly rejected Plaintiffs' argument on this point, see Section I, ante. An employer who violates Title VII by segregating job classifications by sex is not liable for the entire difference between the wages for the higher paid male jobs and the lower paid female jobs to each female who was thus denied a job opportunity unless the two jobs being compared were substantially equal in skill, effort and responsibility. These courts have held that only those women who can show that they have been denied their "rightful place" in a higher-paying position or a promotion will be entitled to this amount of damages. As the Court in Orr v. Frank R. MacNeill & Son, Inc., 511 F.2d 166, 171 (5th Cir. 1975) held:
"[T]o establish a case under Title VII, it must be proved that a wage differential was based on sex and that there was the performance of equal work for unequal compensation."
See also, e. g., Christensen v. State of Iowa, 563 F.2d 353 (8th Cir. 1977) and cases cited therein; DiSalvo v. Chamber of Commerce of Greater Kansas City, 416 F. Supp. 844 (W.D.Mo.1976); Molthan v. Temple University, 442 F. Supp. 448 (E.D.Pa.1977).
We see no justification, either in the rules of statutory construction or in public policy, for rejecting the conclusion that this portion *401 of Title VII incorporates the requirement of the Equal Pay Act that a plaintiff who was hired for a lower-paying job must prove that she did work equal in skill, effort and responsibility to that performed by men in higher paying jobs or that, had it not been for discrimination based on sex, she would have been one of the women hired for an available high paying job.
III. The Equal Pay Case.
A. Substantial Equality of Jobs.
29 U.S.C. § 206(d) (1) provides as follows:
"No employer having employees subject to any provisions of this section shall discriminate, within any establishment in which such employees are employed, between employees on the basis of sex by paying wages to employees in such establishment at a rate less than the rate at which he pays wages to employees of the opposite sex in such establishment for equal work on jobs the performance of which requires equal skill, effort, and responsibility, and which are performed under similar working conditions, except where such payment is made pursuant to (i) a seniority system; (ii) a merit system; (iii) a system which measures earnings by quantity or quality of production; or (iv) a differential based on any other factor than sex: Provided, That an employer who is paying a wage rate differential in violation of this subsection shall not, in order to comply with the provisions of this subsection, reduce the wage rate of any employee."
As the Third Circuit Court of Appeals noted in Schultz v. Wheaton Glass Company, 421 F.2d 259, 264 (3d Cir.), cert. denied, 398 U.S. 905, 90 S. Ct. 1696, 26 L. Ed. 2d 64 (1970),
"[I]n its final form . . . [the Equal Pay Act] bears evidence of the competing tendencies which surrounded its birth. There are problems of construction which leap up from the reading of its language."
In Schultz v. Wheaton Glass Company, supra, the standards for application of the Act, recently reasserted by the Third Circuit Court of Appeals, are set forth:
". . . Congress in prescribing `equal' work did not require that the jobs be identical, but only that they must be substantially equal. Any other interpretation would destroy the remedial purposes of the Act.
The Act was intended as a broad charter of women's rights in the economic field. It sought to overcome the age-old belief in women's inferiority and to eliminate the depressing effects on living standards of reduced wages for female workers and the economic and social consequences which flow from it.
Differences in job classifications were in general expected to be beyond the coverage of the Equal Pay Act. This was because in the case of genuine job classifications the differences in work necessarily would be substantial and the differences in compensation therefore would be based on the differences in work which justified them. Congress never intended, however, that an artificially created job classification which did not substantially differ from the genuine one could provide an escape for an employer from the operation of the Equal Pay Act. This would be too wide a door through which the content of the Act would disappear." (footnotes omitted). 421 F.2d at 265-66.
Plaintiffs have made out a prima facie case of an Equal Pay violation by Defendant Liberty Mutual only if they have shown that the jobs of CA and CR involved "equal work on jobs the performance of which requires equal skill, effort and responsibility, and which are performed under similar working conditions." 29 U.S.C. § 206(d) (1). Furthermore, the Equal Pay Act requires that "the jobs to be equated be substantially the same," Angelo v. Bacharach Instrument Co., 555 F.2d 1164, 1173 (3d Cir. 1977) and the requirements of equal *402 skill, effort, responsibility, and similar working conditions cannot be aggregated to establish job equality. Id. at 1175.
From all the testimony produced in the weeks of hearings on this aspect of the case, we have reached the conclusion that these jobs were substantially equal except for a uniform difference in working conditions. It is clear that the purpose of the functions of both CAs and CRs were identical: to determine or evaluate the insurance company's potential liability for a given claim and to arrange settlements advantageous to the insurance company without unnecessarily annoying customers. In addition, a prime goal of all these employees was to discourage claimants from contacting lawyers or filing suits. To accomplish these ends, employees in both job classifications were required to have a familiarity with the terms of the basic kinds of policies and to exercise considerable tact in dealing with claimants themselves as well as with, for example, witnesses to an accident.
Plaintiffs presented expert testimony by Dr. Richard S. Barrett, whose doctorate is in industrial psychology and whose qualifications we find to be substantial. We note that although Dr. Barrett claimed no special familiarity with the insurance industry, he used a job evaluation system which was quite suitable for evaluation of these two positions and with which he was thoroughly familiar. In preparing his pretrial report and his trial testimony, this witness relied solely on documentary evidence such as depositions already filed and job descriptions prepared by the defendant. Because the jobs had been changed before trial, he did not personally interview any Liberty Mutual employees. However, we find that the assumptions on which he based his analysis were supported by credible testimony offered by the Plaintiffs at trial and therefore we are inclined to give substantial weight to his testimony.
In performing his analysis, Dr. Barrett analyzed each factor separately for both job classifications and then assigned points based on that analysis. He concluded that both CAs and CRs were required to assume responsibility at approximately the same level that both were presumed "to know the routine work details and performance standards of his [or her] job and to perform assigned tasks without appreciable advice or spur from the immediate superior," to exercise, within stated limits, discretion as to work schedule and to "refer questionable items to the immediate supervisors when in doubt as to procedure." Tr. 873-4. Neither the CAs nor the CRs were responsible for supervising other employees. Neither position required more than three months training. For both positions, Liberty Mutual required a college degree but no other work experience. Both CAs and CRs would be responsible at any one time for more than $1,000 worth of money. Both positions, as noted above, required a substantial ability to deal easily with others. Both involved regular work with confidential information. Mistakes by either a CA or a CR could cause considerable financial loss or loss of prestige to the insurance company.
It is also clear that the CRs generally had a much higher number of cases at any one time than did the CAs, and consequently could spend less time on each case. Travelling between appointments necessarily consumed large chunks of a typical CAs workday, so that time available for actual interviews was reduced. We have had some *403 difficulty in determining whether, after the Court of Appeals decision in Angelo v. Bacharach Instruments Corp. supra, a conclusion that the generally higher dollar amount of the cases on which the CAs were involved could be balanced out by the higher number of cases on which the CRs worked would be proper. However, because the total risk to the company from the CRs work was at least comparable to the total risk from the CAs work, we have concluded that, as to this specific aspect of the responsibility criterion, the CAs and the CRs did comparable work.
Different point values were assigned for the following aspects of the jobs: Because claims adjusters typically used company-owned automobiles, they were considered to have greater responsibility for machinery and equipment than did the CRs. The CRs work required "constant concentration on a very large volume of work which must be completed within a limited period of time," while the CAs work "involved uniform mental attention" for which they were awarded fewer points. In addition, because the CRs had more monotonous jobs, they were given more points than were the CAs. On a scale of 2 to 10, the CRs were given 2 points for working conditions ("not unusual," Tr. 877) while the CAs were given 6 points for working "regularly under poorer than average conditions; illumination and ventilation are considered only adequate . . ." Id.
It is noteworthy that on only two points of comparison were the CRs rated lower than the CAs: the first is the amount of responsibility for employer-owned machinery and the like, while the second is working conditions. Because responsibility for equipment is only one aspect of the generalized criterion of responsibility, as to all others of which the CRs were rated the same as or higher than the CAs, we attach no great weight to this and find the jobs to be substantially equal except for the working conditions under which they were performed.
B. Working Conditions.
The real difficulty lies in the question of what weight should be attached to the difference in working conditions between these job classifications, and whether this difference has been shown to be merely a pretext for a sex-motivated pay differential. We note that the terms of the statute itself apparently give to this factor a weight equal to that of any of the others listed skill, effort or responsibility. Furthermore, the applicable regulation provides as follows:
"Generally, employees performing jobs requiring equal skill, effort, and responsibility are likely to be performing them under similar working conditions. However, in situations where some employees performing work meeting these standards have working conditions substantially different from those required for the performance of other jobs the equal pay principle would not apply. For example, if some sales persons are engaged in selling a product exclusively inside a store and others employed by the same establishment spend a large part of their time selling the same product away from the establishment, the working conditions would be dissimilar. Also, where some employees do repair work exclusively inside a shop while others employed by the shop spend most of their time doing similar repair work in customers' homes, there would not be similarity in working conditions. On the other hand, slight or inconsequential differences in working conditions that are essentially similar would not justify a differential in pay. Such differences are not usually taken into consideration by employers or in collective bargaining in setting wage rates." (emphasis added) 29 C.F.R. § 800.132.
*404 The only significant decision to deal with this phrase is Corning Glass Works v. Brennan, 417 U.S. 188, 94 S. Ct. 2223, 41 L. Ed. 2d 1 (1974):
"[T]he legislative history reveals that Congress incorporated words having a special meaning within the field regulated by the statute so as to overcome objections by industry representatives that statutory definitions were vague and incomplete.
While a layman might well assume that time of day worked reflects one aspect of a job's `working conditions,' the term has a much different and much more specific meaning in the language of industrial relations. As Corning's representative testified at the hearings, the element of working conditions encompasses two subfactors; `surroundings' and `hazards.' `Surroundings' measures the elements, such as toxic chemicals or fumes, regularly encountered by a worker, their intensity, and their frequency. `Hazards' takes into account the physical hazards regularly encountered, their frequency and the severity of injury they can cause." 417 U.S. 202, 94 S. Ct. 2231.
Applying these definitions, the Court then concluded that
"the concept of `working conditions' as used in the specialized language of job evaluation systems, simply does not encompass shift differentials." Id.
However, the defendant, Corning Glass, had the opportunity to show as a defense that the shift differential was justified as being based on a "factor other than sex." On this point, the company's intent in establishing the differential was considered crucial.
In this case Dr. Barrett testified for the plaintiff that the phrase "working conditions,"
"refers generally to adverse working conditions, particularly those that are caused by an unpleasant environment, heat, noise, wetness and things like that, or hazards that may be involved on a job.
Q. In the field of job evaluation, what significance or weight is attached to the conditions under which jobs are performed?
A. Generally, its a very minor consideration . . .
Q. In the field of job evaluation, what significance or weight is attached to performing a job inside of an office as opposed to performing it outside of an office.
A. In my experience, I haven't seen any difference because that is not along the line of adverse working conditions that are generally referred to such as working by a heat treating furnace where the temperature is terribly high or in chemical plants where there are fumes and dust and things like that."
On cross examination, Dr. Barrett testified that he rated the CAs working conditions in the middle of three possible point levels because
"the working conditions factor for the top level says `works under unusual conditions, perhaps at the sacrifice of personal safety, comfort and convenience (as continuous travel out of the city or in the heat and noise of a boiler room).' So this would be a top level which would be worth 10 points.
* * * * * *
Q. Why didn't you assign the 10 points then?
A. I don't think that going out in a car is particularly hazardous. People do it every day to go to the movies or to the ball game. This is not a hazardous occupation in terms of say, working on high iron or working in the coal mine or being a house painter. These are much more hazardous. We mean by hazardous not the usual hazards as in going to and from work." Tr. 901.
*405 However, because the CA for a large portion of his average workweek works in less desirable conditions than does the average office-bound CR, he was awarded more points for working conditions.
Dr. Rigassio generally agreed with this assessment.
"[F]or example, just taking the Department of Labor. Cited in a number of their publications is the fact they point out that even a truck driver that rides in a truck you know, the modern type of truck where the atmosphere is relatively controlled, you know, warmed in the winter and cooled in the summer they do not consider this an outside job. So that you have to look at this thing. You have to see whether or not an outside job really is an outside job. Now, what is an outside job? Well, if you are delivering mail, this is an outside job. If you are a field worker, you know, working in the field picking crops, this is an outside job. If you happen to be a construction superintendent where you are out there standing in a big excavation, this is an outside job. See, these are the things I have to know." Tr. 990-91.
The record has established that the typical claims adjuster was only infrequently exposed to the elements and that the major hazard to which he was exposed was that of a traffic accident. The adjusters usually spend one day a week in the nearest branch office and some portion of the rest of the week making telephone calls from their homes to set up appointments. Furthermore, we agree with the Plaintiffs that the character of the work done by the claims adjusters was substantially unaffected by the fact that it was not performed in an office. Finally, we agree that credible expert evidence of the technical meaning of the phrase "working conditions," in the job evaluation field may, under Corning Glass, supra, outweigh the effect of the interpretation in the Regulations on which Liberty Mutual has so heavily relied. Neither CAs nor CRs were regularly exposed to adverse working conditions which could possibly justify the difference in wages paid them. We find that Dr. Barrett gave proper weight to this factor in his analysis and that the slight difference in working conditions here is legally insignificant.
C. What is an "establishment?"
Finally, the case law on the definition of the term "establishment" in the Equal Pay Act tends to support the Defendant's argument that equality of the CS and CA jobs must be determined separately at each branch office. The regulations on point provide as follows:
"Application to Establishments
It should be kept in mind in determining an employer's obligation under the equal provisions, that `employer' and `establishment' as used in these and other provisions of the Act are not synonymous terms. An employer may have more than one establishment in which he employs employees within the meaning of the Act. In such cases, the legislative history makes clear that there shall be no comparison between wages paid to employees in different establishments." 29 C.F.R. § 800.103
"Application to Employees
On the other hand, it is clear from the language of the Act that in each distinct physical place of business where employees of an employer work (including, but not limited to, the employer's own establishments), the obligation of the employer to comply with the equal pay requirements must be determined separately with reference to those of his employees who are employed in that particular establishment. Accordingly, where there are a number of distinct physical places of business in which an employer's employees are employed, compliance with the equal pay provisions must be tested within each establishment by comparing the jobs in which employees are employed in that establishment and the wages paid for work on such jobs when performed by employees of opposite sexes." 29 C.F.R. § 800.104.
Our research has disclosed few cases dealing specifically with the question of what is an "establishment" for Equal Pay purposes. *406 The only opinion which offers such guidance is Brennan v. Goose Creek Consolidated Independent School District, 519 F.2d 53 (5th Cir. 1975).
In Goose Creek, the Secretary of Labor alleged that the female light cleaners in a school district were being paid less than were the male heavy cleaners in violation of the Equal Pay Act. In affirming the trial court's holding that the physically separate schools in the district constituted a single establishment for purposes of the Act, the Court of Appeals construed the term in light of the analogous provision in 29 U.S.C. § 213(a) (2) and the case law developed thereunder.
"A narrow construction of the word `establishment,' as used in . . . [the Equal Pay Act] might make proof of discrimination more difficult, thus frustrating congressional intent.
* * * * * *
In the present case a central authority was responsible for the janitors' employment and wages often transferred them from one school to another. The school district failed to show any difference in the janitors' working conditions at the eleven schools. We, therefore, hold that the janitors were employed by one "establishment." 519 F.2d at 57-58.
In his concurring opinion in Goose Creek, Judge Gee set forth the factors he considered relevant to the conclusion that the school district was a single "establishment."
"First, the central administration acted as the hiring authority for all the schools in the system. Wage scales and job classifications were also set by the district administration office. The central administration assigned the individual janitors to schools and occasionally switched janitors from one school to another. . . A district official also testified that evaluation of janitors for so-called merit increases was not solely in the hands of individual principals. In fact the principals consulted and reached a decision with the Deputy Supervisor in charge of maintenance . . . Finally, on a broader scale, the schools in the district are under common control and a unified plan to provide educational services within the district. 519 F.2d at 58-9.
Plaintiffs have not addressed this problem in any depth. Their Proposed Finding of Fact No. 23 summarized their position as follows:
"Uniform starting salaries for claims adjusters and claims representatives and subsequent salary increases are determined by the defendant's home office personnel in Boston, Massachusetts. Claims Adjusters receive automatic salary increases of four hundred dollars ($400.00) a year at the end of their third month of employment. In the past, they received similar annual salary increases of $300.00. Claims representatives receive no similar increases. Starting salaries for claims adjusters at the defendant's approximately 130 claims office are and have been uniform with few exceptions. Starting salaries for claims representatives vary throughout the defendant's offices and from one office to another within each division. There is no rational correlation between the exceptions to the uniform national adjusters' starting salary and the claims representatives' starting salaries at the few offices where the exceptions apply. Nor does there appear to be any rational relationship between variations in claims representatives' starting salaries and the cost of living in the cities where defendant's offices are located."
Assuming that the foregoing quotation is an accurate summary of the evidence *407 of record, we must nevertheless hold that under the Equal Pay Act, Liberty Mutual's approximately 130 branch offices do not constitute a single establishment because of the potential and actual range of dutiesin terms of skill, effort, responsibility and working conditions, that isamong the physically separate branch offices. We can find no authority under the Equal Pay Act for focusing on the similarity of these jobs on a nationwide scale rather than on an office-by-office level. Applying Judge Gee's test in Goose Creek, and mindful of the emphasis on physical proximity as a determinative factor in the Regulations and in the cases decided under the Fair Labor Standards Act, we can reach no other conclusion.
The only remaining question, then, is whether Title VII also contains a requirement that plaintiffs prove their case "establishment" by "establishment." Although the cases discussed in part II, ante, have uniformly held that a Title VII plaintiff must, in this kind of case, prove that the two jobs in question are substantially equal in skill, effort and responsibility, we have found no precedent precisely on this point. In our view, however, the purpose of the Bennett amendment is to ensure that the same standard of proof is applied under Title VII as under the Equal Pay Act by eliminating the argument that Plaintiff has here made. In other words, the Bennett amendment serves only to eliminate the possibility that a defendant who has been guilty only of job segregation will be liable to every female in the lower paid jobs for the difference between her wages and the wages paid for the jobs from which she has been excluded. Where two jobs are segregated by sex but genuinely different, the employer should be liable only to an individual who can prove that, but for the discrimination, she personally would have been hired for a more lucrative position from which she was excluded merely because of her sex. Women who would not, for any reason, have qualified for or been hired to fill the jobs which were held only by men even were those positions open to women in general are entitled to no relief. Because Title VII nowhere contains a requirement similar to the separate "establishment" of the Equal Pay Act, because Title VII is such an appropriate vehicle for class-wide relief from discrimination, and because the Bennett amendment fully serves its function without incorporating the establishment requirement into Title VII, we conclude that there is no such limitation on Title VII relief with regard to discrimination in compensation. It follows, then, that Plaintiffs, having proved that the jobs of CA and CR are substantially the same, have proved their claim under Title VII and should be awarded damages accordingly.
An order will be entered setting a conference of counsel to discuss outstanding discovery matters and the evidence necessary to calculate damages in accordance with the foregoing opinion.NOTES
 Wetzel v. Liberty Mutual Insurance Company, 372 F. Supp. 1146 (W.D.Pa.1974), aff'd 508 F.2d 239 (3d Cir. 1975), cert. denied 421 U.S. 1011, 95 S. Ct. 2415, 44 L. Ed. 2d 679 (1975).
 "Between July 1, 1965 and March 17, 1972, 2,329 persons were hired as `Claims Representatives' of which 2,302 or 98.84% were women. The first male `Claims Representative' was hired in 1969, four in 1970, and twenty in 1971." 372 F. Supp. at 1151.
 The CAs' $7,900 annual salary included an increment of $400 paid after the first 90 days on the job, a benefit which was not awarded to the female CRs.
 Plaintiffs' Proposed Findings of Fact Nos. 26 and 27.
 Plaintiffs' Proposed Conclusion of Law No. 18.
 Even in a class action the total amount of damages recoverable by a class will be substantially less under this theory than under that advocated by plaintiffs because it is obviously unreasonable to assume that even a law-abiding employer would have placed or hired every woman who held a lower-paying job in a different and higher paying position instead of hiring only men.
 See Usery v. Allegheny County Institution District, 544 F.2d 148, 153, n. 4 (3d Cir. 1976).
 See also Corning Glass Works v. Brennan, 417 U.S. 188, 195, 94 S. Ct. 2223, 41 L. Ed. 2d 1 (1974).
 We find testimony by various Plaintiffs' witnesses to the effect that settlement of small claims generally involves at least as much persuasiveness as does settlement of more substantial claims to be eminently reasonable.
 Liberty Mutual has placed a great deal of emphasis on testimony tending to show that CRs received, usually, only about 1-2 weeks of formal training, while the CAs were trained in classes for as much as three months, during which time they were given little or no field experience. Clearly, though, the CRs were subject for far longer periods to such close supervision as might be considered on-the-job training. While in the field, the CAs could not be subject to any comparable supervision. In any case, we find the demarcation point adopted by the manual used by Dr. Barrett three months or less of training to be rationally based.
 This conclusion is bolstered by the testimony of the government expert, Dr. James L. Rigassio, Tr. 948 et seq. which is generally consistent with Dr. Barrett's evaluation.
 Liberty Mutual relies heavily on the emphasized portion of the regulation quoted above. We note that, while this court should defer to those regulations which can be considered consistent with the Equal Pay Act, the regulations are by no means conclusive in every case. Angelo v. Bacharach, supra, 555 F.2d at 1171-2 n. 10.
 Defendants have cited several cases dealing with what constitutes a single establishment in the context of an application for injunctive relief. Marshall v. Goodyear Tire & Rubber Co., 554 F.2d 730 (5th Cir. 1977) is one of the more recent cases in this line. See also Belt v. Johnson Motor Lines, Inc., 458 F.2d 443 (5th Cir. 1972), and Usery v. Sears, Roebuck & Co., 421 F. Supp. 411 (N.D.Iowa 1976). We have not found these helpful in this case where back pay is the only relief sought by the plaintiffs.
 We note that the Defendant has not seriously contested the fact that local or district officials had little discretion in setting salary scales for CAs or CRs or that in all branch offices the CAs were paid significantly more than were CRs of comparable seniority.