Equal Employment Opportunity Commission v. Local 638, 401 F. Supp. 467 (S.D.N.Y. 1975)

U.S. District Court for the Southern District of New York - 401 F. Supp. 467 (S.D.N.Y. 1975)
July 18, 1975

401 F. Supp. 467 (1975)

LOCAL 638 et al., Defendants.
LOCAL 28, Third-Party Plaintiff,

No. 71 Civ. 2877 (HFW).

United States District Court, S. D. New York.

July 18, 1975.

*468 *469 *470 Paul J. Curran, U. S. Atty., S. D. N. Y., New York City, for plaintiff United States Equal Employment Opportunity Commission; by Taggart D. Adams, Louis G. Corsi, New York City.

W. Bernard Richland, New York City Corp. Counsel, New York City, for plaintiff City of New York; by Beverly Gross, Thomas A. Trimboli, New York City.

Sol Bogen, New York City, for defendant Local 28.

Rosenthal & Goldhaber, Brooklyn, N. Y., for defendant Joint Apprenticeship Committee and Trust; by William Rothberg, Brooklyn, N. Y.

Louis J. Lefkowitz, Atty. Gen., New York City, for third and fourth-party defendant New York State Division of Human Rights; by Dominic J. Tuminaro, New York City.


WERKER, District Judge.

This is an action filed in 1971 by the United States pursuant to § 707(a) of Title VII of the 1964 Civil Rights Act, 42 U.S.C. § 2000e-6(a). It was originally part of a larger action against four unions in the building trades industry and their joint apprenticeship committee for engaging in a past and continuing pattern and practice of discrimination in admission and employment of non-whites.[1] Soon after issue was joined in that case separate trials were ordered for each union and its related co-defendants.[2] After severance of the four groups for purposes of trial, the City of New York (the City) was granted leave to intervene in that portion of the action relating to Local Union No. 28 of the Sheet Metal Workers' International Association (Local 28). United States v. Local 638, Enterprise Ass'n, etc., 347 F. Supp. 164 (S.D.N.Y.1972).[3]

The defendants who were on trial before this court from January 13 to February 3, 1975 are Local 28, Local 28's *471 Joint Apprenticeship Committee and Trust (JAC), and, for purposes of relief only, the New York City Chapter of the Sheet Metal and Air Conditioning Contractors' National Association (Contractors' Association). By virtue of third and fourth party complaints filed by Local 28 and JAC, the New York State Division of Human Rights (the Division) is a defendant in this action for purposes of relief. The third and fourth-party pleadings were predicated upon administrative and judicial proceedings instituted by the State Attorney General against Local 28 and JAC in which the defendants were directed to end racially discriminatory selection and admission practices under the supervision and direction of the Division.[4]See State Commission on Human Rights v. Farrell, 43 Misc.2d 958, 252 N.Y.S.2d 649 (Sup.Ct.N.Y.Cnty. 1964). Although a nominal defendant in this case, the Division has in its papers and at trial consistently aligned itself with the plaintiffs' cause.

The complaint filed by the United States (the government) alleges that Local 28 is engaged in a pattern and practice of resistance to the full enjoyment by non-whites of rights secured to them by Title VII of the 1964 Civil Rights Act, 42 U.S.C. §§ 2000e-2(c) and 2000e-2(d). The pattern and practice alleged includes, but is not limited to, the following:

(a) Failing and refusing to admit nonwhite workmen . . . as journeymen members on the same basis as whites are admitted;
(b) Failing and refusing to refer nonwhite workmen for employment ... (within its jurisdiction) ... on the same basis as whites are referred by applying standards for referral which have the purpose and effect of ensuring referral priority to ... (its) . . . members ...;
(c) Failing and refusing to recruit blacks for membership in and employment through ... (Local 28) . . . on the same basis as whites are recruited;
(d) Failing and refusing to permit contractors with whom ... (Local 28) ... has collective bargaining agreements to fulfill the affirmative action obligations imposed upon those contractors by Executive Order 11246 by refusing to refer out blacks whom such contractors wish to employ;
(e) Failing and refusing to take reasonable steps to make known to nonwhite workmen the oportunities for employment in the . . . (sheetmetal trade) ..., or otherwise to take affirmative action to overcome the effects of past racially discriminatory policies and practices.

The government's complaint does not allege specific acts of discrimination by defendant JAC. To the extent, however, that plaintiffs have succeeded in establishing such violations at trial, the government's complaint is deemed amended to conform to the proof. See Rule 15(b), Fed.R.Civ.P.

The City's complaint (¶ 10) alleges that Local 28 is engaged in a pattern and practice of resistance to the full enjoyment by non-whites of rights which are secured to them by § B1-7.0 of the New York City Administrative Code as well as by Title VII of the 1964 Civil Rights Act. It sets forth the same allegations as to Local 28 quoted above, and adds as an additional example of discriminatory practice:

(f) Adopting standards for admission to union membership which are not *472 job related and which operate to disqualify a disproportionate number of non-whites for membership.

The City's complaint, furthermore, alleges that defendant JAC is also engaged in a pattern and practice of discrimination, which includes, but is not limited to:

(a) Failing and refusing to make information concerning apprenticeship opportunities available to non-whites on the same basis as it is made available to whites;
(b) Failing and refusing to make apprenticeship opportunities available to non-whites on the same basis as they are made available to whites;
(c) Adopting standards for the selection of apprentices which are not job related and which operate to disqualify a disproportionate number of non-white applicants for apprenticeship.

The allegations of both complaints have been largely substantiated by the evidence produced at trial.

A. Local 28

1. Local 28 is an unincorporated labor union. It is the recognized bargaining agent for journeymen and apprentice sheet metal workers[5] hired by sheet metal contractors within its geographical jurisdiction.

2. The geographical jurisdiction of Local 28 includes the five boroughs of the City of New York.

3. A non-white has never been an officer of Local 28, or a member of the Executive Board of Local 28.

4. Since its inception in 1913 Local 28 has been governed by its own Constitution and By-Laws, and by the Constitution and Ritual of the Sheet Metal Workers' International Association. Prior to November 1946, the Constitution of the International Association contained a provision for the establishment of an "auxiliary" local union when there was a "sufficient number of eligible Negro applicants." As stated in this provision, the auxiliary was:

subordinate to the established and affiliated white local union and shall be represented by said white local union at all conferences and conventions, including International Conventions .... The same initiation, reinitiation and reinstatement fees shall apply to auxiliary members and the privilege of transfer shall be limited to transferring from one auxiliary to another auxiliary.

5. As of October 1, 1974 Local 28 had collective bargaining agreements with approximately 133 sheet metal contractors in New York City. Those contractors do not employ an individual to perform sheet metal work within the trade jurisdiction of Local 28[6] unless *473 the individual is a member or apprentice of Local 28, or has been given an "identification slip" (ID Slip) by Local 28 permitting him to temporarily work in the sheet metal industry within the geographic jurisdiction of Local 28. The reason for this is that members of Local 28 will not work with sheet metal workers who do not fit within either of the two categories above. Thus, despite a contract provision which grants Local 28 contractors autonomy in hiring, Local *474 28 has substantial, if not complete, control of job opportunities which arise with them.

6. In preventing its contractors from hiring non-union non-white sheet metal workers Local 28 has precluded them from fulfilling the affirmative action obligations imposed on them by Presidential Executive Order 11246, 3 C.F.R. Chapter IV § 202, and Mayoral Executive Order 71, April 2, 1968, 96 The City Record 2842 (April 10, 1968). Those orders require recipients of federal construction funds and city construction contracts to take affirmative action to ensure that all applicants for employment enjoy equal access to work opportunities without regard to race, color or national origin.

7. Since 1960 there have been four methods by which individuals have been admitted to membership in Local 28:

a) successful completion of a four-year apprentice program administered by JAC;
b) successful performance on a written and practical examination administered by the Examining Board of Local 28 (journeyman's test);
c) transfer from a sister local union, affiliated with the Sheet Metal Workers' International Association;
d) employment with a newly organized sheet metal contractor who will certify as to its need for the applicant and the applicant's ability to work in accordance with journeyman standards of performance.

The availability of the first method is determined by the collective bargaining agreement between Local 28 and the employers. The availability of the other three methods is determined by the Executive Board of Local 28, with the approval of the Union membership.

8. As of July 1, 1974, 3.19% of the union's total membership (including pensioners) was non-white.

9. Between January 1965 and July 1974 Local 28 admitted 1103 new members, 79.78% from the apprentice program, 9.07% through the use of written and practical examinations, 5.98% via transfer from sister unions, and 2.81% from newly organized sheet metal shops. Of the 1103 new members, 111 or 10.06% were non-white.

10. Local 28 does not maintain a hiring hall. Referral and hiring are done informally, through word of mouth and contacts with other members, apprentices and contractors. Sometimes business agents call Local 28 members and advise them of job opportunities; sometimes members call the agents seeking information on work openings. In good times, each business agent makes a job referral approximately five to ten times a week.

11. Local 28 refused to participate in the New York Plan when it was in effect in New York City. The Plan was a joint industry, City and State effort to increase participation of minority employees in the construction trades.[7]

B. The Contractors' Association

12. The Contractors' Association is an association of building contractors in New York City who are engaged in sheet metal construction work. It has a collective bargaining agreement with Local 28, and its members employ approximately 70-80% of the union's members and apprentices.

13. The manpower requirements of the Contractors' Association is a mandatory subject of collective bargaining by and between the Association and Local 28.

C. JAC and the Apprentice Program

14. JAC is a joint labor-management committee composed of representatives of Local 28 and of the Contractors' Association. It administers the Local 28 apprentice program.

15. A non-white individual has never been a member of JAC.

16. Since 1964 the operation and organization of the apprentice program *475 has been governed by the Standard Form Union Agreement (the Collective Bargaining Agreement), the order and opinion in State Commission on Human Rights v. Farrell, supra, which includes the Corrected Fifth Draft of Standards for the Admission of Apprentices for the Sheetmetal Industry of New York City, New York (Corrected Fifth Draft), JAC's Agreement and Declaration of Trust, and JAC's Rules and Regulations.

17. In 1965 non-white enrollment in the apprentice program was .37%. It increased to a high of 21.80% in July 1967, fell to 9.77% in July 1973 and returned to 13.99% in July 1974.

18. The Local 28 apprentice program presently consists of eight terms of six months each. Apprentices attend ten all-day class sessions per term, receiving eight hours of pay for each such session.[8] The other days they work for employers who have collective bargaining agreements with Local 28.

19. Under the most recent (1972) Collective Bargaining Agreement apprentice classes are to be appointed every six months, and the size of the entire program is to be stabilized at 568 apprentices. Between January 1, 1972 and July 1, 1974 the total number of apprentices enrolled in the program has decreased:

    January 1, 1972     - 568
    July 1, 1972        - 482
    November 11, 1972   - 517
    January 1, 1973     - 498
    July 1, 1973        - 399
    January 1, 1974     - 323
    July 1, 1974        - 286

20. An apprentice must pass a physical exam and be between the ages 18 and 25 at the time of admission, although exceptions up to age 30 are made for time spent in military service.

21. Since 1969, as a result of the Corrected Fifth Draft contained in the New York Supreme Court decision in State Commission on Human Rights v. Farrell, supra, apprentices are required to have high school diplomas or equivalency certificates at the time of admission. For the years 1967-1968 only three years of high school education were required. For the years 1965-1966, only two years of high school education were required. Prior to 1965 apprentices were appointed from an applicants list, with high school diplomas, veteran's status and recommendations of relatives by members of Local 28 receiving some weight in the appointment process.

22. Application forms utilized for the apprentice program require the applicant to list his police record, if any, and his citizenship.

23. Applicants satisfying the age, education and physical requirements are admitted to the apprentice program in accordance with the ranking obtained on an apprentice entrance exam. There is no cut-off pass/fail score for the entrance exam, which is an aptitude exam consisting of tests in five areas (the JAC battery):

a) mental alertness
b) mechanical reasoning
c) space relations
d) mathematical computations and concepts
e) mathematical analysis and problem solving

24. Since 1966 the choice of tests in the aforementioned areas, their administration, *476 and the ranking of those tested has been performed by Stevens Institute of Technology (Stevens Institute). In 1965 and 1966 this work was done by the New York Testing and Advisement Center.

25. None of the defendants have kept records of the race or the ethnic identification of persons who have applied or sought to apply to the apprentice program, of persons whose applications have been rejected prior to the aptitude examination, or of persons who have taken the aptitude examination and have been rejected or have themselves rejected admission to the apprentice program.

26. None of the defendants prior to 1973 kept records of the race or ethnic identification of persons taking the apprentice aptitude examination.

27. No one fails out of the apprentice program because of school performance, although he may be left back.

28. JAC assigns apprentices for employment. However, no apprentice can begin working for a Local 28 employer without first receiving a union "apprentice work card."

29. The current Collective Bargaining Agreement between Local 28 and the Contractors' Association provides:

§ 3(c). There shall be a moratorium on apprentices during the period of the six (6) hour day. Apprentices shall be appointed, but work assignments deferred until return to the seven (7) hour day and shall then be made by the Joint Apprenticeship Committee in accordance with the required needs of the program.

Since October 31, 1973, the industry has been on a six hour day.


Prior to the institution of a new selection procedure in accordance with the Corrected Fifth Draft, Local 28 alone controlled admission to the apprentice program. Union officers testified that the basic selection criterion applied by Local 28 was the applicant's relationship to, or friendship with, union members. Thus the union's selection procedure was largely nepotistic, with the result that a majority of the individuals enrolled in the apprentice program were related in some manner to members of Local 28.[9] This is further borne out by evidence that no Black was over enrolled in the apprentice program, and at least during the period from January 1, 1960 through March 15, 1965, only one other non-white individual was a Local 28 apprentice.

Congress's objective in enacting Title VII of the 1964 Civil Rights Act was to achieve equality of employment opportunity by removing those selection barriers which have historically operated to favor white employees; therefore under the Act,

practices, procedures, or tests neutral on their face, and even neutral in terms of intent, cannot be maintained if they operate to `freeze' the status quo of prior discriminatory employment practices.

Griggs v. Duke Power Co., 401 U.S. 424, 429-30, 91 S. Ct. 849, 853, 28 L. Ed. 2d 158 (1970).

Whether the new selection procedure provided for in the Corrected Fifth Draft discriminates against non-whites and/or operates to "freeze the status quo" of the defendants' prior discrimination would best be determined by a *477 thorough statistical analysis of the entire procedure as a whole. However, the fact that no records were kept of applicants' race and national origin has precluded this approach.[10] Plaintiffs at trial therefore sought to establish that each individual component of the selection procedure operates separately to discriminate against non-whites. For the reasons discussed below the court finds that the selection procedure does not fully comport with the mandate of Title VII.

A. The Apprentice Entrance Exam

Plaintiffs' argument as to the JAC battery of tests was based almost exclusively upon the testimony and statistical analysis of Dr. Raymond Katzell, Professor of Psychology at New York University, and an expert in industrial psychology and psychological testing. Dr. Katzell analyzed the percentage of identified non-whites tested, as opposed to that of whites tested, who appeared in the top 50, top 100, and top half of the aggregate rankings for all eight exams administered between April 1968 and March 1973.[11] He found that the percentage of non-whites at each level was smaller than that of whites to a statistically significant degree. This means, in the language of the industrial psychology profession, that the apprentice entrance examination as administered in those eight testing sessions had an "adverse impact" on non-whites, and that the impact was not likely to have happened by mere chance (i. e.: was likely to have occurred by chance on the basis of a random sampling no more than one time in twenty).

Dr. Katzell also separately analyzed each of the eight test batteries in the same manner. He found that on at least one level in each of seven batteries there was a statistically significant difference in percentage. There was no statistically significant difference on the eighth battery. The more conclusive result achieved in the aggregate analysis stems from the larger size of the applicant sample involved.

Dr. Katzell further testified that the presence of statistically significant adverse impact on the JAC battery is not surprising, as Blacks and other economically disadvantaged subgroups perform competitively less well than whites on verbally-based tests. This conclusion is generally shared by the psychology profession. See Cooper and Sobol, Seniority and Testing Under Fair Employment *478 Laws: A General Approach to Objective Criteria of Hiring and Promotion, 82 Harv.L.Rev. 1598 at 1639 (1969); indeed, JAC's expert witness, Dr. Judah Gottesman, senior consulting industrial psychologist at the Stevens Institute of Technology, testified in substantial agreement. See also Boston Chapter NAACP Inc. v. Beecher, 504 F.2d 1017, 1021 (5th Cir. 1974); Bridgeport Guardians, Inc. v. Members of Bridgeport Civil Service Comm'n, 482 F.2d 1333, 1337 (2d Cir. 1973).

Title VII proscribes standardized testing devices which, however neutral on their face, operate to exclude non-whites capable of performing effectively in the desired positions. McDonnell Douglas Corp. v. Green, 411 U.S. 792, 806, 93 S. Ct. 1817, 36 L. Ed. 2d 668 (1973). As the Second Circuit noted in United States v. Wood, Wire & Metal Lathers International Union, 471 F.2d 408, 414, n.11 (2d Cir.), cert. denied, 412 U.S. 939, 93 S. Ct. 2773, 37 L. Ed. 2d 398 (1973), "statistics may establish a prima facie case of discrimination" in violation of Title VII.[12] The court finds that plaintiffs' statistical analysis described above establishes such a case, and thereby shifts to defendants the burden of justifying their use of the discriminatory testing device. See Chance v. Board of Examiners, 458 F.2d 1167, 1175 (2d Cir. 1972). Since the touchstone in justifying a discriminatory practice is business necessity,[13] defendants' burden is to prove by professionally acceptable methods that the Local 28 apprentice entrance exam is significantly related to job performance. Albemarle Paper Co. v. Moody, 422 U.S. 405, 95 S. Ct. 2362, 45 L. Ed. 2d 280, 43 U.S.L.W. 4880 (1975); Griggs v. Duke Power Co., 401 U.S. at 431, 91 S. Ct. 849.

To sustain that burden JAC called Dr. Gottesman who testified as to three validation studies which he had performed on the JAC battery. A validation study is one made to determine if a test can with significant accuracy predict an individual's performance on the job. According to Dr. Gottesman, validity studies are of three kinds: A "construct validity" study establishes whether or not the exam determines the degree to which applicants possess characteristics important to job performance. A "content validity" study establishes whether or not the exam contents closely duplicate the actual duties to be performed on the job. Lastly, a "predictive validity" study establishes whether or not exam scores correlate with external variables considered to provide a direct measure of job performance. See American Psychological *479 Association, Standards for Educational and Psychological Tests at 25-31 (1974). Of the three methods, the predictive validity study is considered best. Bridgeport Guardians, 482 F.2d at 1337; United States v. Local 638, Enterprise Ass'n, 360 F. Supp. 979, 992 (S.D.N.Y. 1973), modified 501 F.2d 622 (2d Cir. 1974).

The first study performed by Dr. Gottesman fits within none of the three categories described above. At best it may be described as an "indirect validity study." It consists of no more than a comparison of the JAC battery to prevalidated aptitude tests of the same name and supposed subject matter in the United States Employment Service's General Aptitude Test Battery. Such a comparison shows nothing with respect to the JAC battery as a whole. Furthermore, it is predicated on the precarious assumption that aptitude tests with the same name are equivalent.[14] Dr. Gottesman's conclusion that the two sets of tests are "essentially identical" and therefore equally valid accordingly carries little, if any, probative force.

Dr. Gottesman's second study examined predictive validity. Although it also shows nothing with respect to the JAC battery as a whole, it indicates the separate predicative validity of each of the five tests. This study was made possible by the fact that JAC admitted to the apprentice program all of the applicants who took the April 1969 entrance examination (the April 1969 group) regardless of rank. As part of the validation study, when those in the April 1969 group were completing their four year apprenticeships they were examined as to job performance on a practical test ("hands-on") and a written test ("trade-information"). All parties agree that of the two tests, the hands-on sample, created by the apprentice program's coordinator of training, Robert Schluter, is the more direct measure of job performance. Plaintiffs, however, question the adequacy of that test as a scientific criterion of sheet metal success. See note 16 infra.

In comparing the apprentices' performance on the hands-on sample to their performance on each component of the JAC battery, Dr. Gottesman was able to determine the correlation between job performance and each aptitude test. He concluded that:

unfortunately, by reason of acceleration, natural attrition and other factors, the graduating apprentice class in June, 1973, did not contain a sufficient number of either Blacks (N=8) or Spanish-surnamed Americans (N=5) to represent statistically meaningful subgroups. Hence, no case can be made either for or against the existence of any differential validity. That is, the data does not support or negate any contention that the test predictors operate differently for minorities than for whites.

Ex. W. at 1-2 (emphasis added). His findings, however, show little evidence of validity for either whites or minorities, and largely support the expert opinion of Dr. Katzell that the apprentice entrance exam adversely affects non-white applicants. Dr. Gottesman's findings, in brief, are that for the April 1969 group as a whole, only one of the five tests in the JAC battery, the mechanical comprehension test, was highly and significantly correlated to the "hands-on" sample. That test, moreover, was the only one in which non-whites scored as well as or better than whites. The math computations and *480 concepts test correlated to a lesser degree, but the other three tests had no significant correlation at all. More importantly, for the non-white applicant group viewed alone, none of the five tests were significantly correlated to the hands-on sample.

Because this study produced only meagre evidence of validity, Dr. Gottesman recommended that JAC alter the apprentice entrance exam. He suggested the elimination of four out of five of the battery tests, leaving only the test on mechanical comprehension, and the addition of a basic arithmetic and a "read and follow directions" test. No action has been taken by JAC on his recommendations.

Dr. Gottesman's third validity study attempted to determine the predictive validity of the JAC battery as a whole. He compared the total weighted raw scores of applicants tested in April 1969 with their May, 1973 combined trade-information and hands-on scores in order to determine whether a significant predictive correlation existed between the JAC battery and the criteria used to determine performance level. The correlation coefficient he found was one of .25, which Dr. Katzell later demonstrated to be marginal, i. e.: "a weak depiction of relationship between the test on the one hand [and] what you are trying to predict by means of the test on the other." (Tr. 585).

The court has been unable to follow Dr. Gottesman's calculations in arriving at the .25 figure because the data on which it is based have not been made part of the record.[15] From the data available to the court, however, it appears that this study, like the previous predictive one, is seriously incomplete in that it fails to take into consideration forty-two apprentices in the April 1969 group. Those apprentices were not tested as to job performance because they graduated early from the apprentice program. The significance of this omission to the two validity studies premised on relative exam performance becomes apparent when one considers that approximately half of those who graduated early ranked in the bottom half of entrance exam scores. Thus they are evidence that persons who score poorly on the test battery can perform successfully as apprentices and journeymen. All of these apprentices readily could have been made available by the defendants for testing, as they had all been inducted, upon graduation, into membership in Local 28.[16]

In summary, then, the testimony produced by defendants as to the job-relatedness of the apprentice entrance exam is spotty and largely equivocal. Their validation studies have failed to demonstrate that the JAC battery as a whole is significantly job-related, or indeed, that any of its component parts other than the section on mechanical comprehension is capable of identifying and testing for characteristics necessary to adequate sheet metal performance. Defendants, therefore, have failed to sustain their burden of proving "that *481 the disproportionate impact was simply the result of a proper test demonstrating less ability of blacks and Hispanics to perform the job satisfactorily." Vulcan Society v. Civil Service Comm'n, 490 F.2d 387, 392 (2d Cir. 1973). Further use of the JAC battery will therefore be enjoined.

B. The Requirement of a High School or Equivalency Diploma

Until institution of the Corrected Fifth Draft, applicants were not excluded from the apprentice program for failure to attain a minimum educational level. Indeed, it is unclear to the court even after three weeks of trial testimony and the filing of post-trial memoranda, why the new requirement of a high school diploma was added other than to upgrade in general the median educational achievement level of Local 28. Neither JAC nor Local 28 produced evidence of a relationship between success as a sheet metal apprentice and completion of high school.

Plaintiffs' witness, Mrs. Roxee Joly, a high school superintendent and a former mathematics teacher in the New York City school system, reviewed on the witness stand most of the math examinations used in the apprentice program. She described the areas of skill tested in those exams as: decimals, fractions, trigonometry, basic arithmetic, basic arithmetic as to linear measurements, solid geometry and elementary algebra. She noted that students of average academic achievement in the New York City school system are taught those skills in grades four through nine, and defendants in no way rebutted her testimony. They merely emphasized on cross-examination what is a matter of common knowledge that some students have difficulty learning those skills and must continue to study them in later grades, and that some students never learn them at all. Apprentice program coordinator Robert Schluter furthermore testified that at least in so far as trigonometry is concerned, apprentices are not expected to enter the program with training in it; they are taught all the trigonometry needed to perform as sheet metal workers during their course of study. Thus, although the evidence indicates that knowledge of certain mathematical concepts is essential to adequate performance as an apprentice on written and practical examinations, defendants failed to demonstrate any nexus between that knowledge and a high school diploma.

It is a fact of which the court takes judicial notice that non-white persons obtain high school diplomas at a lower rate than do whites. Publications of the Bureau of the Census, for example, show that the median schooling possessed by all males, age 25 or older, in the New York City Standard Metropolitan Statistical Area is 12.1 years, while that possessed by Black males of the same age group and residence is 10.9 years, and that of Puerto Rican males of the same age group and residence, 8.3 years. 1970 Census of Population General Social and Economic Characteristics, New York, Tables 83, 91, 97. According to the EEOC Guidelines, a specific educational requirement such as a high school diploma is a "test" which must be validated like any other if it adversely affects persons protected by Title VII of the 1964 Civil Rights Act. 29 C.F.R. §§ 1607.2, 1607.3. While the Guidelines are not binding on the courts, they are entitled to great deference. Griggs v. Duke Power Co., 401 U.S. at 434, 91 S. Ct. 849.

In view of the fact that high school diplomas have never been required as a condition of attaining journeyman status in the union and that they have only recently been required for entry to the apprentice program, defendants' failure to produce any evidence tending to validate that requirement becomes all the more significant.

History is filled with examples of men and women who rendered highly effective performance without the conventional badges of accomplishment in *482 terms of certificates, diplomas, or degrees. Diplomas and tests are useful servants, but Congress has mandated the common sense proposition that they are not to become masters of reality.

Griggs v. Duke Power Co., 401 U.S. at 433, 91 S. Ct. at 854. This is not to say that no minimum level of education could or should be required as job-related in the sheet metal industry.[17] Defendants, however, have simply not produced convincing evidence that for Local 28's apprentice program the line should be drawn at twelve years. Cf. United States v. Sheet Metal Workers, Local No. 10, 6 E.P.D. ¶ 8715 (D.N.J.1973). Further requirement of a high school or equivalency diploma for entrance to the Local 28 apprentice program will therefore be enjoined.

C. The Application FormInquiry as to Arrest Record

Two types of application forms have been used by JAC for the apprentice program since 1965. Both contain a section which reads as follows:

POLICE RECORD: List below, giving all detailed information, any police record that you have. List each arrest. It is not necessary to list minor traffic violations. This information must be accurate and complete. Any information which is withheld will be cause for immediate dismissal from the Apprentice Program. If none, say "none."
  Name and Address  |        |           |
  of Police Station |        |           |
  or Court          |  Date  |  Offense  |  Outcome
                    |        |           |

It is evident that this section is intended to encompass both convictions and arrests. Plaintiffs object, however, only to the inquiry about arrests. Although they introduced evidence of five instances since 1965 in which JAC rejected applicants because of information contained in the Police Record section of their applications, the evidence is more remarkable for its paucity than for its probative weight; it is apparently unknown whether those five applicants were rejected because of conviction records or because of arrest records.

Plaintiffs also introduced into evidence two tables from Crime in the United States, Uniform Crime Report 1973 issued by Clarence M. Kelly as Director of the FBI, which indicate that non-whites are arrested both nation-wide and city-wide proportionately more often than whites. From this they argue that the presence of the Police Record section on the application form adversely affects non-whites, and therefore imposes a burden on the defendants to validate the arrest inquiry as job-related. The burden, however, cannot be shifted quite so easily; plaintiffs have not established a prima facie case that JAC has a policy of rejecting applicants who have suffered arrest without conviction, or even that JAC has rejected such applicants in the past. Cf. Gregory v. Litton Systems, Inc., 316 F. Supp. 401, 402 (C.D.Cal. 1970); aff'd in relevant part, 472 F.2d 631 (9th Cir. 1972). On the evidence presented, therefore, this court cannot determine that the apprentice program application form discriminates in practice against non-whites. Accord, Green v. Missouri Pacific R. R. Co., 381 F. Supp. 992, 996 (E.D.Mo.1974).


Non-whites presently are, as they have been in the past, conspicuously absent from Local 28. Although qualified non-white sheet metal workers exist in large percentages in other construction *483 locals within the New York metropolitan area, in particular Local Union 400 (Blowpipe Division) of the Sheetmetal Workers International Association, their rate of pay is substantially lower than that received by the Local 28 membership.[18] Defendants would explain this grouping of non-whites in the lower-paying union by arguing that the work performed by Local 28 is more complex and therefore requires a more skilled and adroit membership.

Plaintiffs, however, presented convincing evidence at trial, from members and contractors of both locals, that the skills and tools required to perform Local 400 work are largely the same as those required of Local 28 jobs. Although the trade jurisdiction of the two unions are separate and discreet,[19] they overlap to a significant degree. It is not unusual, for example, for members of Local 28 to perform blowpipe work.[20] Likewise, members of Local 400 are authorized by their collective bargaining agreement to perform up to 75 feet of the square duct work usually considered within Local 28's jurisdiction.

The Local 400 apprentice program, furthermore, is modeled after that of Local 28. Testimony of Thomas Carlough, a member of Local 28 who instituted and now coordinates Local 400's course of apprentice training, indicates that students in both programs are taught the same sheet metal skills. Thus, although a Local 28 member might be more adept at square duct work, or faster because of his daily familiarity with it, members of Local 400 are certainly equipped to, and can, perform the same work.[21] The fact that *484 Local 28 has initiated numerous complaints against Local 400 for infringement of its trade jurisdiction only corroborates this conclusion.

Why, then, is it that non-white sheet metal workers are not evenly distributed throughout the industry? Plaintiffs have argued, and the court finds that Local 28 has denied qualified non-whites direct access to membership in the union while granting such access to white persons by: (a) failing to administer yearly journeyman tests, and using as journeyman tests examinations not validated by EEOC Guidelines; (b) selectively organizing non-union sheet metal shops with few, if any, non-white employees, and/or admitting from those shops only white employees; and (c) accepting as transfer members whites from affiliated sister locals while refusing transfers of non-whites.

A. The Journeyman Tests

It is a matter of common knowledge that at least between 1967 and 1972 the urban areas of the United States experienced a construction boom. Yet since 1959, Local 28 has administered only two journeyman examinations. Both of these exams came about as a result of arbitration proceedings brought by the Contractors' Association to force the union to increase its manpower. In 1968 when ordered by Arbitrator Theodore Kheel to admit 100 new journeymen, Local 28 designed and administered a journeyman test which, admittedly, has never been validated in accordance with EEOC Guidelines. Of 330 individuals tested on the first, written portion of the exam, only thirty-four passed, and were allowed to proceed to the practical portion. Ten failed the practical portion, with the result that twenty-four journeymen, all white, were admitted to the union. The above statistics would seem to indicate that the test served more as an obstacle to, than a vehicle for, the admission of new journeymen. Indeed, one of the candidates for admission, then a fourth-year apprentice in Local 400, testified that "the test was pretty far out. In other words, you had to have more or less a college degree to really do anything on that test." (Tr. 1543).

Although Local 28 may have only intended to limit the number of new journeymen (white or non-white) admitted by way of the 1968 exam, the effect of that exam was to exclude non-whites. Robert Schluter, chairman of the local's examining board, testified that by visual observation 15% of those taking the exam were Black; he could not estimate the number of Hispanics. Even assuming that no Hispanics were tested, the exam clearly had an adverse impact on non-whites, and as such, without validation, was violative of Title VII. See Griggs v. Duke Power Co., 401 U.S. at 430, 91 S. Ct. 849.

In 1969, following another arbitration order, Local 28 administered a second journeyman test. According to Chairman Schluter the exam had been restructured since 1968 so as to eliminate questions involving mathematical concepts unrelated to sheet metal work, and replace them with questions of "shop math." As a result, 14 non-whites and 61 whites successfully passed the journeyman test and were admitted to the union. Because of Local 28's failure to keep records as to the numbers of whites and non-whites tested it is not possible to determine whether this exam also had an adverse impact on non-whites. In 1969 as in 1968 the exam had been advertised by sending a letter of notice to the New York State Employment Service, the Veterans Administration, the New York State and New York City Human Rights Offices, the Workers Defense League, members of Local 28, and Local 28 contractors. Application forms were not sent, however, as Local 28 required that these be obtained and filled out in person at union headquarters.

In 1970 Local 28 refused to administer another journeyman test. Instead, in response to pleas by the Contractors' *485 Association for more manpower, Local 28 recalled pensioners who on doctors' certificates were able to work, and issued hundreds of ID slips to members of affiliated locals and allied construction trades. Between July 1, 1969 and July 1, 1972, Local 28 issued the following numbers of ID slips:

    7/1/69        150-200
    1/1/70        200-250
    7/1/70        200-250
    1/1/71        250-300
    7/1/71        400-450
    1/1/72        400-450

Only one of those receiving ID slips has been identified as non-white. In addition, despite the fact that Local 28 saw fit to request ID men from sister locals all across the country, as well as from allied New York construction unions such as plumbers, carpenters and iron-workers, it never once sought them from Sheet Metal Local 400.[22] Furthermore, in 1969 when a group of apprentices and journeymen from Local 400 went to Local 28's offices to request ID cards, they were informed by Union President Mell Farrell that the union was not giving out ID slips.

By using the ID slip system of temporary manpower rather than continuing to administer journeyman tests, Local 28 restricted the size of its membership and thereby enabled its membership to earn substantial payment for overtime work. This had the illegal effect, if not the intention, of denying non-whites access to employment opportunities in the industry. Cf. United States v. Local 638, Enterprise Ass'n, 347 F. Supp. at 181; accord United States v. Local No. 357, et al., 356 F. Supp. 104, 116 (D.Neb.1973). Union President Farrell in 1971 at a Joint Adjustment Board grievance proceeding initiated by the Contractors Association justified the refusal to enlarge union membership by remarking that "overtime is expected in the Construction Industry" (Ex. 111 at 2). Self-serving expectations, however, do not constitute business necessity within the meaning of Title VII. Cf. United States v. Local 638, Enterprises Ass'n, 501 U.S. at 633.

B. Organization of Non-Union Shops

Prior to 1973 no non-white ever became a member of Local 28 through the organization of a non-union shop. The explanation for this proffered by the union is that they have never been aware of any non-union sheet metal shops owned by or employing non-whites. Such an assertion, aside from testing the credulity of the court, is clearly contradicted by the testimony of record. For example, Edward Carlough, former president of Local 28, testified by deposition that he had been aware of non-union shops employing non-white sheet metal workers as early as the 1940s. In addition, Rupert Jonas, a Black sheet metal worker in Local 295 of the Operating Engineers, testified that in 1971 Local 28 organized the Integrity Air Conditioning shop in which he worked with ten other non-whites.

Since 1973 non-whites have only gained membership in Local 28 through organization of shops because the parties to this litigation during settlement negotiations entered into an agreement whereby the union was to embark on an organizing campaign. Pursuant to that agreement Local 28 organized six shops in 1973 and 1974, and thereby admitted three non-whites to membership.

A large percentage of the shops with non-white workers which Local 28 had the opportunity but chose not to organize were blowpipe companies. In the late 1950's and early 1960's the Sheet Metal Workers' International Association *486 urged Local 28 to organize the blowpipe industry in the metropolitan region. The industry at that time consisted of approximately 265-365 workers, 60 to 75% of whom were non-whites. Local 28 refused to organize them despite insistent pressure from the International Association as well as from the Contractor's Association.[23] Its official reason was that the blowpipe contractors could not meet Local 28's pay scale, and the union could not under its collective bargaining agreement accept a wage differential for its members.[24] The unofficial reason, however, appears to be that Local 28 did not wish to admit as members the non-white blowpipe workers.

Seymour Zwerling, the principal of several contracting companies in signed agreement with Local 28, testified that it was a matter of "common knowledge in the industry" that the union did not want to organize the blowpipe workers because many of them were minorities. (Tr. 1145). Indeed, there appears to be no other reason for the union's refusal. In organizing an entire industry it would not have been able, as with individual shops, to admit only white workers and exclude non-white employees. In any case, whatever the reason, the effect of Local 28's refusal was the denial to non-whites of employment opportunities granted whites. "Congress directed the thrust of the [Civil Rights] Act to the consequences of employment practices, not simply the motivation." Griggs v. Duke Power Co., 401 U.S. at 432, 91 S. Ct. at 854. As a result of the union's refusal the International Association organized blowpipe workers on its own, forming them into a special building trades division of Local 400.

C. Transfers

Section 9(k) of Article 16 of the Sheet Metal Workers' International Association Constitution and Ritual provides:

Any member who has established a record of continuous good standing of five (5) years or more to and including date of issuance of transfer card shall be admitted by transfer card into any local union of this Association in accordance with the requirements of this Constitution, and without payment of any difference in initiation fee.

(emphasis added). Such a provision has existed in the Constitution since at least 1946. During the period from 1967 through 1972, Local 28 accepted fifty-seven transfers from sister locals. All of those accepted were white. Indeed, between May 1940 and February 1973, Local 28 accepted 153 transfers, all of whom were white. Only after commencement of this litigation did the union, in 1973, accept its first non-white transfers, two journeymen from Local 400.

The homogeneity of the transfer group, however, is not the only evidence of Local 28's purposeful discrimination against non-whites. Henry Woods, a Black member of Local 28 who gained admission to the union through the 1969 journeyman test, stated at trial that he and several other blowpipe workers from Local 400 inquired about transfer of Local 28 President Farrell. Farrell told them that transfer was impossible since they were not members of a building trades union like 28. (Tr. 1641.) Given his familiarity with the organization of the blowpipe industry, Farrell indubitably knew that his statement was false. Furthermore, only nine months *487 previously, four white blowpipe workers from Local 400 had been allowed to transfer into Local 28.

At trial present union officials testified that they had no records of, and could recall no non-whites ever requesting transfer into Local 28. However, traditionally all requests for transfer are formally made in person before the union's Executive Board. Non-whites, who were discouraged when they inquired informally as to transfer, were simply never given the opportunity to appear before the Board. Even had they been permitted to do so, however, Local 28 policy would have prevented their transfer. Union Recording Secretary Edward O'Reilly testified that for at least the past ten years, Local 28 has refused to accept transfers of all but former members, a policy in direct contravention of the International Association Constitution and Ritual.

This "members only" policy went into effect in the early 1960's when Local 28 was almost exclusively all white. It therefore effectively foreclosed transfer into Local 28 by non-whites. Although the International Association has on an occasional appeal overruled Local 28 and ordered it to accept a qualified transfer worker, this has never resulted in the admission of a non-white journeyman. The existence of an appeal procedure clearly cannot be viewed as justifying or in any way ameliorating the union's practice of denying to qualified non-whites the equal access to employment opportunities guaranteed them by the Civil Rights Act.


1. Local 28 is a union and labor organization within the meaning of § 701(d), Title VII of the 1964 Civil Rights Act and § B1-2.0, subd. 3, Title B of the New York City Administrative Code, and is engaged in an industry affecting commerce.

2. JAC is a joint labor-management apprenticeship committee within the meaning of § 701(d), Title VII of the 1964 Civil Rights Act and § B1-7.0(1) (a), Title B of the New York City Administrative Code, and is engaged in an industry affecting commerce.

3. Prior to the effective dates of Title VII and Title B, and continuing to the present, Local 28 has maintained clearly discernable discriminatory practices in recruitment, selection, training and admission to membership of non-white workers. As a result of this history of discrimination Local 28 has had a well-deserved reputation in non-white communities of discriminating in recruitment, selection training and admission. This reputation operated and still operates to discourage non-whites from seeking membership in the local union or its apprentice program, in violation of Title VII and Title B.

4. Prior to the effective dates of Title VII and Title B, JAC and Local 28 maintained standards and practices in selection of apprentices for admission to the Local 28 apprentice program which discriminated against non-whites. JAC and Local 28 have a clearly deserved reputation in non-white communities of discriminating in the administration of the Local 28 apprentice program. This reputation operated and still operates to discourage non-whites from seeking to enter the apprentice program in violation of Title VII and Title B.

5. Subsequent to the effective dates of Title VII and Title B, JAC and Local 28 adopted selection procedures and standards for admission to the Local 28 apprentice program, some of which are not demonstrably job-related. They operate individually and in combination to prevent non-whites from enjoying equal access to the program in violation of Title VII and Title B.

6. Local 28 and JAC, by virtue of the above-described discriminatory practices, have illegally denied non-whites access to lucrative employment opportunities in the sheet metal industry equal to that enjoyed by whites, and have thereby maintained Local 28 as a white "A" local to the Blowpipe Division's racially mixed "B" local.

*488 7. In order to remedy the effects of the above-described discrimination, Local 28 and JAC have been, and are under an obligation to take affirmative action to recruit, select, train, admit to the apprentice program and admit to membership in the local union substantial numbers of non-whites.


In determining what relief could most appropriately remedy the ongoing effects of defendants' discrimination, it is a relevant inquiry whether each defendant has "voluntarily `cleaned house' or taken any meaningful steps to eradicate the effects of its past discrimination." Rios v. Enterprise Ass'n, 501 F.2d 622, 631-632 (2d Cir. 1974). The record in both state and federal court against these defendants is replete with instances of their bad faith attempts to prevent or delay affirmative action. After Justice Markowitz ordered implementation of the Corrected Fifth Draft, with the intent and hope that it would create "a truly nondiscriminatory union."[25] Local 28 flouted the court's mandate by expending union funds to subsidize special training sessions designed to give union members' friends and relatives a competitive edge in taking the JAC battery. JAC obtained an exemption from state affirmative action regulations directed towards the administration of apprenticeship programs on the ground that its program was operating pursuant to court order; yet Justice Markowitz had specifically provided that all such subsequent regulations, to the extent not inconsistent with his order, were to be incorporated therein and applied to JAC's program. More recently, the defendants unilaterally suspended court-ordered time tables for admission of forty nonwhites to the apprentice program pending trial of this action, only completing the admission process under threat of contempt citations.

"Once a violation of Title VII is established the district court possesses broad power as a court of equity to remedy the vestiges of past discriminatory practices." Rios v. Enterprise Ass'n, 501 F.2d at 629. In light of Local 28's and JAC's failure to "clean house" this court concludes that the imposition of a remedial racial goal in conjunction with an admission preference in favor of non-whites is essential to place the defendants in a position of compliance with the 1964 Civil Rights Act. The use of such measures to compel compliance with the letter and spirit of civil rights legislation is well-recognized. See Patterson v. Newspaper and Mail Deliverers' Union, 514 F.2d 767, 2d Cir. 1975; Rios v. Enterprise Ass'n, supra and cases cited at 629; United States v. Wood Wire & Metal Lathers International, supra.

The purpose of setting a remedial goal is to place eligible non-whites in the position they would have enjoyed had there been no discrimination. To do so here the court must determine what percentage of union and apprentice program members would today be non-white had the defendants not engaged in discriminatory practices. The best available measure of that percentage is the percentage of non-whites in the relevant labor force existing today within New York City. Although there is some danger inherent in assuming the equivalence of these percentages, the court does so in an effort to "do the best it can" with the information available to it. Rios v. Enterprise Ass'n, 501 F.2d at 632.

The relevant labor force in this case consists of an aggregate of three groups: males, eighteen years of age and older[26] with zero to eight years of education, males eighteen years of age *489 and older with nine to twelve years of education, and males eighteen years of age and older with more than twelve years of education.[27] Since the only available labor force statistics come from the 1970 Census conducted by the Department of Commerce the court, to be as accurate as possible, should look to males who at the time of the census were thirteen years or older (i. e.: now eighteen years of age or older). Unfortunately, census figures speak of males sixteen years and older rather than males eighteen years and older. Thus to the extent that thirteen, fourteen and fifteen-year old males are excluded from our calculations, the total labor force figure arrived at is low.

The census figures, furthermore, while they reflect relevant data on total population, Black population, Spanish language population and Puerto Rican population, fail to isolate a Spanish surnamed group. The court therefore has used census statistics on Spanish language persons.[28] For these reasons, and others of similar nature, absolute precision in the calculation of a percentage goal is impossible. However, using figures provided by the Bureau of the Census to the best advantage, and adjusting for the fact that a percentage of Spanish language males are Black,[29] the court has determined that approximately 29% of the relevant labor force in New York City is non-white. See Appendix for an explanation of the court's method of determining percentage goal.

The court accordingly orders that by July 1, 1981, the combined union and apprentice program membership achieve a non-white percentage of 29%.[30] The recruitment, testing and admission procedure for arriving at that goal is to be agreed upon and developed by the parties, under the guidance of a court-appointed administrator, within the next two months. The court specifically requires, however, that the following be included as part of such procedure:

The union is to administer a nondiscriminatory hands-on journeyman's test, professionally developed and validated in accordance with EEOC Guidelines, at least once a year; the first such test is to be given in or before September 1975.
JAC is to administer a yearly apprentice entrance exam consisting solely of the mechanical comprehension aptitude test validated by Dr. *490 Gottesman and a "read and follow directions" test to be developed professionally and validated in accordance with EEOC Guidelines; the first such test is to be given in or before December 1975.
The union is to replace on of its present JAC trustees with a non-white.
Both Local 28 and JAC, in conjunction with the Administrator, are to develop recruitment practices specifically designed to dispel their reputation for discrimination in non-white communities and to guard against the recurrence of such reputation.[31]
Both Local 28 and JAC are to maintain separate lists of whites and non-whites who (a) apply to take the apprentice entrance exam and/or the journeyman's test; (b) take the apprentice entrance exam and/or journeyman's test; (c) pass the apprentice entrance exam and/or journeyman's test; (d) seek to transfer into Local 28 from a sister local; and (e) inquire about the possibility of transferring into Local 28.

The administrator named by the court is also requested to develop with the parties a plan aimed at protecting non-whites admitted into union and apprentice membership from bearing a disporportionate burden of the unemployment caused by current depressed conditions in the construction industry.

In light of the recent Supreme Court decision in Albemarle Paper Co. v. Moody, supra, this court is constrained to determine whether an award of back pay here pursuant to 42 U.S.C. § 2000e-5(g) is necessary and appropriate. In Albemarle the Court held that "given a finding of unlawful discrimination, backpay should be denied only for reasons which, if applied generally, would not frustrate the central statutory purposes of . . . [Title VII] ...." Those purposes are "to achieve equality of employment opportunities and remove barriers that have operated in the past to favor an identifiable group of . . . [whites] . . ." over others, Griggs v. Duke Power Co., 401 U.S. at 429-30, 91 S. Ct. at 853, and to make whole those persons who suffered injury on account of such unlawful discrimination. As the Court noted in Albemarle Paper Co., supra 422 U.S. at 417, 95 S. Ct. at 2371, backpay has an obvious connection with these purposes:

It is the reasonably certain prospect of a backpay award that "provide[s] the spur or catalyst which causes employers and unions to self-examine and to self-evaluate their employment practices . . .".

Plaintiffs in this case did not specifically request backpay in their complaints, and did not during trial attempt to define proposed classes of persons entitled thereto. However, they did include in their proposed post-trial findings the following paragraph:

As a result of the above-described discrimination, non-whites have suffered financial loss and are, therefore, entitled to receive backpay in amounts to be determined subsequent to the trial of this action.

The tardiness with which plaintiffs assert their demand for backpay does not preclude the court from awarding it where entitled, Rule 54(c) Fed.R.Civ.P., and in this case, does not prejudice the defendants, who have long been on notice of the discriminatory practices which are alleged as the basis for backpay relief. The difficulty with plaintiffs' proposal, however, is that no complete records exist of persons who would *491 be entitled under it to an award of backpay. Although the court hesitates to limit relief on this ground, thus in effect rewarding defendants for their failure to keep adequate records as required by the EEOC Guidelines, the alternative is clearly unacceptable. Any award of damages to those for whom records do not exist would at best be hypothetical. The court therefore concludes that backpay should be awarded by Local 28[32] only to non-white persons

a) for whom there exist records of application for direct entry into the union, either through the journeyman's exam or through transfer procedures;
b) who demonstrate before the administrator, in light of this court's conclusions of law on the merits, that they were discriminatorily excluded from union membership;[33] and
c) who show monetary damages suffered as a result thereof.

This class of persons is undoubtedly small. There is no risk, therefore, that it will inequitably drain the financial resources of the non-profit defendant association.

Damages suffered by persons denied entrance to the apprentice program and by blowpipe workers organized as part of Local 400 rather than as part of Local 28, are too highly speculative to merit backpay awards in this case. Likewise the award of damages to that class of non-whites who would have applied for direct admission to membership had Local 28's reputation for discrimination been less pervasive will also be denied as unascertainable.

Those non-whites who are entitled to backpay awards must file a claim with the administrator on or before January 15, 1976. They may recover proven damages from the date the discrimination occurred[34] to the date of filing of this decision on the merits, or to the date of union admission, whichever is earlier. Damages shall be computed on the basis of the average monthly wage earned in each calendar year by members of Local 28, and shall of course be adjusted to reflect other employment income or public assistance received by claimants. See 42 U.S.C. § 2000e-5(g); Memorandum Decision of Judge Bonsal, Rios v. Enterprise Ass'n, 400 F. Supp. 988, S.D.N.Y., June 27, 1975. Payment is to be made after determination by the administrator of all claims, and their discretionary review, if necessary, by this court.

The foregoing constitute the court's findings of fact and conclusions of law. The parties and the administrator are ordered to submit an agreed upon recruitment, testing and admission procedure *492 within 60 days of the filing of this decision. The court will maintain continuing jurisdiction over the parties to this action for purposes of ensuring implementation of appropriate relief.

So ordered.

      The Court's Method of Calculating Percentage Goal[a]
1. Total number of Spanish language males 18 and over:
   To find this total the court determined the ratio of the number of Spanish
   language males in the population to the number of Puerto Rican males in the
   population, and then multiplied the number of Puerto Rican males 18 and
   over by that ratio.
   Number of P.R.[b]          Number of Spanish language[c]     Total Number
   Males 18 and over   ×   Males in the Population       =      of Spanish
                                Number of P. R. Males in the[d]    Language Males
                                Population                            18 and Over
2. Adjusted number of Spanish language males 18 and over:
   The court then multiplied this figure by (1 minus .116)[e] or .884 in order to
   avoid counting Blacks as Spanish language males.
  | Number of P. R.            Number of Spanish lan- |            Adjusted
  | Males 18 and over  ×      guage Males in the Population|  × .884 =  Number of
  |                                                                Spanish Language
  |                            ---------------------  |            Males 18 and over
  |                            Number of P. R. Males  |
  |                            in the Population      |
3. Total number of non-white males 18 and over:
   Lastly, the court added the number of Black males 18 and over to arrive at
   the total for non-white males 18 and over. (T)
   Number of Black        |  Number of P. R.        Number of Spanish        |
   Males 18 and over   +  |  Males 18 and     ×  Language Males in × .884 | = T
                          |  over                   the Population           |
                          |                         ---------------------    |
                          |                         Number of P. R.          |
                          |                         Males in the             |
                          |                         Population               |
4. Total number of non-white males 18 and over in each educational category:
   For reasons explained in the opinion, supra at note 27, the court sought to
        (a) the total number of non-whites 18 and over who have completed
        zero to eight years of education;
        (b) the total number of non-whites 18 and over who have completed
        nine to twelve years of education; and
        (c) the total number of non-whites 18 and over who have completed
        more than twelve years of education.
   The court therefore conducted three series of calculations using the formula
   for T as adjusted to reflect education (T(a), T(b) and T(c)).
                     |---                                          ---|
  Number of Black[f]   | Number of P. R.[g]      Number of Spanish        |
  Males 18 and     + | Males 18 and        × Language Males in × .884 | = T( )
  over with ( )      | over with ( )         the Population           |
  education          | education             --------------
                     |                       Number of P. R.          |
                     |                       Males in the             |
                     |                       Population               |
                     |---                                          ---|
5. Non-white percent and percentage goal of the relevant labor force in each
   educational category:
   To learn what percent of the relevant labor force in each category is non-white,
   and therefore what the non-white percentage goal for that category
   should be (PG), the court simply compared T( ) to the total number of
   all males 18 and over with ( ) education (RLF).[h]
   T( )   = PG
   RLF ( )
6. Total non-white percent and percentage goal of the relevant labor force:
   The court multiplied PG(a) by 23.98%, PG(b) by 67.79% and PG(c) by
   8.22%[i] and then added the three results to achieve the total non-white percentage
   goal in this case of 29%.[j]

[1] For purposes of this case the parties have agreed that the term "non-whites" includes black and Spanish surnamed individuals. Pre-trial Order, p. 3.

[2] The reported decisions of those actions which have proceeded to trial can be found at 347 F. Supp. 169 (S.D.N.Y.1972) (Local 40 of the structural iron workers), and 360 F. Supp. 979 (S.D.N.Y.1973) (Local 638 of the steamfitters), modified 501 F.2d 622 (2d Cir. 1974); on remand, unreported decision of Judge Bonsal dated May 5, 1975.

[3] Intervention was sought during the pendency of an administrative proceeding for racial discrimination initiated by the City Commission on Human Rights against Local 28 for violation of Title B, Chapter I, of the New York City Administrative Code (City Human Rights Law). See 347 F. Supp. at 166.

[4] The Commission on Human Rights had found, after conducting administrative hearings, that the defendants discriminated against Negroes in the designation and approval of applicants for the Local 28 apprentice program. The Honorable Jacob Markowitz of the New York Supreme Court adopted all of the Commission's findings as to JAC and ordered implemented new standards for the admission of apprentices.

[5] Sheet metal workers fabricate and install ducts for ventilating, air-conditioning and heating systems.

"In heating and air-conditioning duct work, sheet metal workers plan the job to determine the size and type of metal needed before cutting it with hand snips, power-driven shears, and other tools. They shape the metal with machines, hammers, and anvils, then weld, bolt, rivet, solder, or cement the seams and joints .... To install ducts, components are fitted together, hangers and braces installed for support, and joints connected and soldered or welded. Some sheet metal workers specialize in shopwork or on-site installation, others do both." Occupational Outlook Handbook, 1974-75 Edition, published by the United States Department of Labor, at 279. This description appears accurate in light of testimony at trial as to the nature of work performed by members of Local 28.

[6] "The manufacture, fabrication, assembly, erection, installation, dismantling, reconditioning, adjustment, alteration, repairing and servicing of all sheet metal work (including ferrous or non-ferrous sheet metal) of No. 10 U.S. gauge or its equivalent or lighter gauge, or any and all substitute materials used in lieu thereof (any question of jurisdiction of substitute materials used in lieu thereof in accordance with the Joint Arbitration Plan between the Building Trades Employers' Associations and the Unions of the Building Trades of the City of New York), including all shop and field sketches used in fabrication and erection (including those taken from original architectural and engineering drawings or sketches and all other work included in the jurisdictional claims of Sheet Metal Workers International Association.

"Testing and balancing of all air-handling equipment and duct work contracted for after June 30, 1969. All internal linings for casing, plenum and ducts must be lined prior to erection. Supply casings, Supply air-shafts must be Sheet Metal.

"The manufacturing and erection of all sheet metal work in connection with buildings and structures as follows: hollow metal sash, frames, partitions, skylights, cornices, crestings, awnings, circular mouldings, spandrels (except stamping of same), sheet iron sheeting or roofing, package chutes, linen chutes, rubbish chutes, hoods, sheet metal fire proofing, ventilators, heating and ventilating pipes, air washers, conveyors, breeching and smoke pipes for hot water heaters, furnaces and boilers, laundry dryers and all connections to and from same, metal connections to machines in planning mills, saw mills and other factories (whether it be used for ventilating, heating or other purposes), sheet metal connections to and from fans, sepercators, sheet metal cyclones for shavings or other refuse in connection with various factories, sheet metal work in connection with or fastened to store fronts or windows, sheet metal work in connection with concrete construction and sheet metal columns and casings, covering all drain boards, lining of coil boxes, ice boxes and other sheet metal work in connection with bar furniture and soda fountains.

"Spot welding, electric are welding, oxyacetylene cutting and welding in connection with sheet metal work of No. 10 gauge or lighter covered by the collective bargaining agreement also; sheet metal work in connection with plain and corrugated fire doors of No. 10 gauge or lighter; also the erection of floor domes, the setting of registers and register faces in connection with sheet metal work, the cutting and bending of metal necessary for application and erection of metal ceilings and side walls (except stamping), the applying of metal to ceiling and side walls and the furring and sheathing of same. The assembling and erection of fans and blowers; also the erection of metal furniture, factory bins, shelving and lockers, corrugated iron on roofs and sidings, all metal shingles and metal slate, and tile, plain or covered with a foreign substance, the manufacture and erection of corrugated wire glass and accessories; the glazing of metal skylights. The installation of unit vents where there is sheet metal work in connection with the supply and discharge of air; the setting of radiator enclosures of sheet metal where it does not support the radiator.

"In the manufacturing of drawn metal work; the work of journeyman sheet metal workers shall be the cutting and forming of the metal before the same is applied to the wood, and all clipping and soldering that may be necessary in the finishing of the assembled parts and the covering of wood and composition door frames and sash with sheet metal. Also such other sheet metal work of No. 10 gauge or lighter, not herein specified that has been decided by the Executive Committee of the Building Trade Employers' Association to be, or is now, in the possession of the Sheet Metal Workers' Union shall be regarded as sheet metal worker's work.

"In the Kitchen Equipment Industry it shall be understood that the term "Sheet Metal Work" shall mean all work made of sheet metal No. 10 gauge or lighter including the making, mounting, erecting, cleaning and repairing of all steel and gas ranges, grid irons and oven racks, hoods, tables and stands, warming closets, plate warmers and plate shelves, bands, doors and slides for some, drip pans, urns, and percolators, kettles, revolving covers, meat dishes and covers, steam and carving tables and drainers for same, bain marie boxes and potato mashers and any other items or types of work that may be included in Article I, Section 5 of the Constitution and Ritual of the International Association.

"In the temporary operation of fans or blowers in a new building, or in addition to an existing building, for heating and/or ventilation, and/or air conditioning, the temporary operation and/or maintenance of such fans or blowers."

Description of Local 28's work jurisdiction, Stipulation of Facts, ¶ A(5).

[7] It also refused to work with the City and Federal government in negotiating an alternative, tailor-made affirmative action plan.

[8] The current Collective Bargaining Agreement provides in Rule XVII that if 300 Local 28 journeymen are unemployed, the industry shifts to a six hour day. When the six hour day is in effect, apprentices receive seven hours of pay for each class session. They are paid according to the following scale:

1st term 40% of the journeyman wage rate

2nd term 45% of the journeyman wage rate

3rd term 50% of the journeyman wage rate

4th term 55% of the journeyman wage rate

5th term 60% of the journeyman wage rate

6th term 65% of the journeyman wage rate

7th term 70% of the journeyman wage rate

8th term 80% of the journeyman wage rate

[9] Local 28's reputation for nepotism also prevented many non-whites who might have otherwise applied from even contacting JAC. One Black mechanic who had sought sheet metal work in New York City in the early 1960's testified that although aware of Local 28, he had not applied because he had "heard that it was a father and son thing, you had to be a relative to become a member ... That threw me out, I didn't have any relatives." (Tr. 1592).

[10] Neither Local 28 nor JAC maintained such records despite the Equal Employment Opportunity Commission Guidelines on Testing and Selecting Employees (EEOC Guidelines), which require:

Every employer, labor organization, and joint labor-management committee subject to Title VII which controls an apprenticeship program (regardless of any joint or individual obligation to file a report) shall, beginning August 1, 1967, maintain a list in chronological order containing the names and addresses of all persons who have applied to participate in the apprenticeship program, including the dates on which such applications were received. (See section 709(c), Title VII, Civil Rights Act of 1964). Such list shall contain a notation of the sex of the applicant and of the applicant's identification as "Negro," "Spanish Surnamed American," "Oriental," "American Indian," or "Other."

29 C.F.R. § 1602.20(b) (1975). However, despite the lack of records, plaintiffs managed to research at least in part the race and nationality of applicants for apprenticeship. Out of 3,490 applicants between 1969 and 1972 they were able to identify 489 as non-whites. Of 446 persons who became apprentices, 43 were non-white. Thus, to the extent that plaintiffs' attempts at identification were successful, it appears that 13.43% of whites who applied became apprentices while only 8.79% of non-whites did so, resulting in a success rate for whites of 1½ times that for non-whites. In Chance v. Board of Examiners, 458 F.2d 1167, 1171 (2d Cir. 1972) the court found that disparity enough to establish a prima facie case of discrimination.

[11] He confined his analysis to the top half and above because with the exception of applicants tested in April 1969 (see p. 479 infra) those ranking below have never been selected by JAC for admission to the apprentice program.

[12] Indeed, plaintiffs argue that the statistical disparity between the percent of union and apprentice members who are non-white (3.97%), and the percent of what they define as the "relevant labor force in New York City" that is non-white (36%), is by itself enough to establish a prima facie case of across-the-board discrimination. Defendants, of course, contest plaintiffs' definition of relevant labor force and thereby dispute plaintiffs' statistics. However, even accounting for errors and overinclusion on the part of plaintiff (see discussion infra at 488-490), the disparity between the number of non-whites available to Local 28 and JAC as a membership pool and those chosen may well be enough to establish a general prima facie case of discrimination. Cf. Vulcan Society v. Civil Service Comm'n, 360 F. Supp. 1265, 1269 (S.D.N.Y.1973); aff'd, 490 F.2d 387 (2d Cir. 1973). Rather than rely on this possibility, however, the court has made findings as to each allegation of discriminatory practice. See United States v. Ironworkers Local 86, 443 F.2d 544, 551 (9th Cir.), cert. denied, 404 U.S. 984, 92 S. Ct. 447, 30 L. Ed. 2d 367 (1971).

[13] In United States v. Bethlehem Steel Corp., 446 F.2d 652, 662 (2d Cir. 1971) the Second Circuit adopted the following definition of the business necessity doctrine:

When an employer or union has discriminated in the past and when its present policies renew or exaggerate discriminatory effects, those policies must yield, unless there is an overriding legitimate, non-racial business purpose.

The court then went on to state: "Necessity connotes an irresistible demand . . . If the legitimate ends ... can be served by a reasonably available alternative system with less discriminatory effects, then the present policies may not be continued."

[14] Standard F 4 of the American Psychological Association's Standards for Educational and Psychological Tests warns against making such an assumption without evidence to back it up, and rates the advice given as "essential." The EEOC Guidelines specifically rule out "assumptions of validity based on test names or descriptive labels." 29 C. F.R. § 1607.8 (1975). While the EEOC Guidelines are not binding on the courts, the Second Circuit has endorsed reliance on them "as a helpful summary of professional testing standards." Vulcan Society v. Civil Service Comm'n, 490 F.2d 387, 394 (2d Cir. 1973).

[15] As plaintiffs note in their Post-Trial Memorandum:

Appendix A of Exhibit W contains raw scores of . . . apprentices identified only by number. Plaintiffs were unable to do the same calculations as are contained in Dr. Gottesman's testimony because they had no way of correlating the raw scores of people identified only by numbers with the percentile scores of named individuals (with different identification numbers) which had previously been produced (P. Exh. 59, Appendices 208).

[16] These last two validity studies are also suspect in that they rely upon a hands-on test as a measure of job performance, yet that test itself was prepared, contrary to the EEOC Guidelines, without benefit of a professional job analysis. 29 C.F.R. § 1607.5(b) (3) (1975). See Vulcan Society v. Civil Service Comm'n, 490 F.2d 387, at 394 n. 8. As the district court in Chance v. Board of Examiners noted, the evaluation of a test's validity "depends upon the reliability and fairness of the field appraisal of performance on the job." 330 F. Supp. 203, 216 (S.D.N.Y.1971), aff'd, 458 F.2d 1167 (2d Cir. 1972).

[17] If the need for mathematical skills to adequately perform as an apprentice, however, is the only concern, it would seem that such skills could be better evaluated through a contemporary test of mathematical ability than by inference of that ability drawn from completion of a given educational level. Cf. Dobbins v. Local 212, Electrical Workers, 292 F. Supp. 413, 453 (S.D.Ohio 1968).

[18] Louis Commarato, president of Local 400, testified that in 1965-1966, members of his union earned approximately $2.25 an hour compared with Local 28's $5 an hour. As of July 1974, Local 400 men earn $7.10 per hour while members of Local 28 receive $12.05 per hour. Similarly, members of Local 295 of the Operating Engineers, who also perform sheet metal work, receive $4.60 per hour.

[19] The trade jurisdiction of Local 400 encompasses:

"Spray booth systems, including blowers, tanks, (duct work incidental to the system if 75 ft. or under), if over 75 ft., all duct work incidental to the system will be subcontracted to a sheet metal shop in agreement with the Sheet Metal Workers' International Association local having jurisdiction in the area in which the work is being done.

"Ovens and drying systems, including the manufacture of ovens, heaters, panels, etc., (and the duct work incidental to the system if 75 ft. or under), if over 75 ft., all duct work incidental to the system will be subcontracted to a sheet metal shop in agreement with the Sheet Metal Workers' International Association local having jurisdiction in the area in which the work is being done.

"Dust collecting systems, including the exhausts, blowers, fans, round pipe and cyclones.

"Conveyor systems; except no slide chutes or hoppers to be installed except at Building Trades rates by a sheet metal shop in agreement with a Building Trades local of the Sheet Metal Workers' International Association.

"Smoke houses.

"Plating and degreasing tanks, including the exhaust systems and round pipe work.

"Smoke stacks; breechings only when done as an accommodation while doing an installation covered in this Schedule in an existing plant.

"Retail bakery work where the duct work does not exceed 75 ft. on the exterior of the building if over 75 ft. all duct work incidental to the system will be sub-contracted out to a sheet metal shop in agreement with the Sheet Metal Workers' International Association local having jurisdiction in the area in which the work is being done.

"Furnish and install makeup air systems in industrial plants."

[20] Blowpipe work entails the creation of ducts to be used for removal of fumes, dust or indeed particles of any substance that can be conveyed by air. It usually involves fabrication and/or installation of round duct work. Local 28's sheet metal work, on the other hand, usually involves fabrication and installation of rectangular or square ducts. Blowpipe work tends to involve prefabricated or standard elbows, joints and forms whereas rectangular ducts for heating and air conditioning units must more frequently be custom made.

[21] The only skill found within Local 28 which is not shared by Local 400 is that of drafting. Not all members of Local 28 are drafters, however; drafting is considered the highest specialty in the union, and requires extra training over and above that acquired in the apprentice program.

[22] Local 28's asserted reason for not contacting Local 400 was its "knowledge" that members of the 400 Blowpipe Division were enjoying full employment. Yet Thomas Carlough, coordinator of the Local 400 apprentice program and himself a member of Local 28, testified that in 1970 the bankruptcy of a major blowpipe shop left many 400 sheet metal workers without jobs.

[23] The Contractors' Association wanted the blowpipe workers organized so as to have access to greater manpower, and so as to eliminate competition from the blowpipe contractors. (Tr. 1127).

[24] A dual wage scale, however, had for many years already existed within the union. Edward O'Reilly, Recording Secretary of Local 28, admitted that the Kalamein Workers (those union members specializing in applying sheet metal linings to doors) receive a lower rate of pay than other members of the local union. (Tr. 225).

[25] State Commission on Human Rights v. Farrell, 43 Misc.2d at 969, 252 N.Y.S.2d 649.

[26] The age requirements for admission to the Local 28 apprentice program were not challenged by plaintiffs.

[27] The three educational divisions were made in an attempt to replicate the educational mix which census figures show to exist among sheet metal workers nationwide. Those figures show that 23.98% of sheet metal workers nationwide have completed zero to eight years of education, 67.79% have completed nine to twelve years of education, and 8.22% have completed more than twelve years of education. See 1970 Census of Population, Subject Reports, Occupational Characteristics, PC(2)-7A, Table 5.

[28] Spanish language persons are defined by the Bureau of the Census as those individuals "of Spanish mother tongue and all other persons in families in which the head or wife reported Spanish as his or her mother tongue." Mother tongue is defined as the language spoken in the person's home when he or she was a child. 1970 Census of Population, General Social and Economic Characteristics, PC(1)-C34 (hereinafter "General Social and Economic Characteristics") Appendix B, at 7.

[29] Approximately 11.6% of males of Spanish origin are Black. Rios v. Enterprise Ass'n, 400 F. Supp. 983, S.D.N.Y.1975. Persons of Spanish origin are those who indicated to the census takers that they were of Mexican, Puerto Rican, Cuban, Central or South American, or "Other Spanish" descent. General Social and Economic Characteristics, Appendix B, at 7, 34. The parties have provided the court with no separate figures as to the percentage of Spanish language males who are Black. The Court has consequently assumed the percentage to be the same as that for males of Spanish origin.

[30] The court has chosen a six year period for implementation of the 29% goal after full consideration of the depressed economic condition of the construction industry, and in the firm belief that a gradual but steady influx of non-whites will produce the most stable membership.

[31] In Gresham v. Chambers, 501 F.2d 687, 691 (2d Cir. 1974) the Court of Appeals noted:

Where a pattern of past discrimination appears, recruitment procedures that might otherwise be classified as neutral will no longer be accepted as non-discriminatory. Additional methods must then be devised to compensate for the effects of past discriminatory practices and to guard against their perpetuation or recurrence.

[32] As JAC has no role in granting or denying direct admission to Local 28 it will not be held liable for backpay.

[33] This of course includes a showing that claimant is qualified for admission in accordance with Local 28's entrance requirements as modified by this decision.

[34] Discrimination for purposes of backpay computations will be deemed to have occurred on the date on which the next applicant for admission who does not qualify as a non-white, as defined for purposes of this case, is admitted to the union. See Rios v. Enterprise Ass'n, 400 F. Supp. 988, S.D.N.Y. 1975 memorandum decision of Judge Bonsal.

Title VII specifically provides that "back pay liability shall not accrue from a date more than two years prior to the filing of a charge with the Commission." 42 U.S.C. § 2000e-5(g). No such time limitation is imposed here, however, because unlike the Rios class action case, this action was not initiated by the filing of a charge with the EEOC. Rather, it was begun by the Attorney General in accordance with the procedures outlined in 42 U.S.C. § 2000e-6(a).

Ruling that accrual not extend more than two years prior to the filing of the Attorney General's complaint would appear a logical analogy under the circumstances. However, in formulating the accrual limitation quoted above, the United States Senate specifically rejected a provision that would have limited backpay liability to a date two years before institution of judicial proceedings. Albemarle Paper Co., supra 422 U.S. at 420, n. 13, 95 S. Ct. 2362. In light of this legislative history, and the small number of persons entitled to backpay in this case, the court choses not to limit accrual of Local 28's liability.

[a] The formula used in this appendix is substantially the same as that employed by Honorable Dudley B. Bonsal in Rios v. Enterprise Ass'n, supra, 400 F. Supp. 988. "Important national goals would be frustrated by a regime of discretion that `produce[d] different results for breaches of duty in situations that cannot be differentiated in policy.'" Albemarle Paper Co. v. Moody, 422 U.S. at 417, 95 S. Ct. at 2371.

[b] Taken from Table 131, General Social and Economic Characteristics at 663.

[c] Census data speaks only of Spanish language population. See Table 119, General Social and Economic Characteristics at 607. The court assumes, however, that the percentage of males among that population is equivalent to the percentage of males among the Puerto Rican population, i. e.: 47.5%. See Table 129, General Social and Economic Characteristics at 659.

[d] Taken from Table 129, supra.

[e] See note 29 supra.

[f] Census figures reflect educational levels for males 25 and over. See Table 125, General Social and Economic Characteristics at 643. The court assumes that educational levels of males 18 and over would be equivalent.

[g] Taken from Table 130, General Social and Economic Characteristics at 661.

[h] RLF stands for relevant labor force. This data is readily available from the census reports. See Table 120, General Social and Economic Characteristics at 613.

[i] See note 27 supra.

[j] The court finds it unnecessary to inflate the percentage goal because of the existence of a greater census undercount for Blacks than for whites. Any advantage the defendants thereby derive is counterbalanced by the advantage plaintiffs gain through the use of the more inclusive Spanish language data as a substitute for unavailable Spanish surname data.