Full opinion text
MEMORANDUM AND ORDER GEORGE C. PRATT, Circuit Judge (sitting by designation): I. INTRODUCTION. Thirty-four current and former female employees of the State University of New York at Stony Brook (“Stony Brook”), as individuals and as representatives of a class of teaching and non-teaching professionals employed at Stony Brook on or after February 11, 1974, and applicants for those positions, brought this sex discrimination suit under Title VII of the Civil Rights Act of 1964, as amended, 42 U.S.C. § 2000e et seq. (“Title VII”), seeking monetary, declaratory, injunctive, and any other appropriate relief to remedy Stony Brook’s allegedly discriminatory employment practices. Defendants are the president of Stony Brook, the chancellor of the University of. the State of New York, and the members of the Board of Trustees of the state university. By memorandum and order dated June 22, 1977, the court granted plaintiffs’ motion for class certification to the extent of certifying under FRCP 23(b)(2) a class “solely with respect to plaintiffs’ claim that at Stony Brook there exists a system-wide pattern and practice of discrimination based on sex.” The class certified consisted of all women who in the past may have been or in the future may be discriminated against on the basis of sex by defendants’ practices with respect to recruiting, hiring, termination, job assignment, promotion, compensation, and other terms, privileges, and benefits of employment and who either (1) have been employed by Stony Brook as a teaching or non-teaching professional at any time on or after February 11, 1974, or (2) may be so employed by Stony Brook in the future, or (3) once unsuccessfully applied for employment as a teaching or non-teaching professional with Stony Brook at any time on or after February 11, 1974, or (4) who may apply for such employment in the future. Plaintiff’s claim of pervasive, system-wide discrimination on the basis of sex satisfied the requirement of FRCP 23(a)(2) for a question of law or fact common to the class. After the class was certified, the parties agreed to and the court approved of a bifurcated trial in which the class issue of a pattern or practice of system-wide discrimination would be tried first, it being intended that after resolution of that central issue, the parties would then work out with the court an appropriate method for resolving whatever issues remained. On the trial that has taken place, plaintiffs sought to prove a pattern and practice of discrimination against teaching and non-teaching professional (NTP) women with respect to virtually all terms and conditions of employment, including recruitment, hiring, initial placement, promotion, tenure, and salary. Plaintiffs also claim that women are excluded, or virtually excluded, from the highest ranks in the university, and are found in disproportionately high percentages in certain traditionally female departments and in the lowest NTP and faculty ranks. After extensive discovery that continued for over five years, the court tried, with the consent of the parties, the single class-oriented issue of whether defendants engaged in a pattern or practice of discriminating against female professionals at Stony Brook. The court has carefully reviewed the testimony of the witnesses who testified at trial, the exhibits admitted into evidence, and the arguments made by counsel, both at trial and in their pre-trial and post-trial briefs. This memorandum constitutes the court’s decision in accordance with Fed.R. Civ.P. 52. The court concludes that plaintiffs have failed to prove that Stony Brook has a university-wide pattern or practice of unlawful sex discrimination; consequently the class action portion of this case will be dismissed. A. Jurisdiction and Venue. The court has jurisdiction of this action under § 706 of Title VII of the Civil Rights Act of 1964, as amended, 42 U.S.C. § 2000e-5(f)(1970). Each named plaintiff filed charges with the Equal Employment Opportunity Commission complaining of various acts of discrimination arising out of her employment at Stony Brook, and each received her “right to sue” letter pursuant to § 706(f)(1) of Title VII. Within 90 days thereafter, plaintiffs commenced this action. The Board of Trustees of the University of the State of New York, the chancellor of the state university system and the president of Stony Brook are employers within the meaning of § 701(b). Venue is proper within this district under 28 U.S.C. § 1391(b), because many of de: fendants’ allegedly unlawful employment practices occurred at Stony Brook within this district. B. Theory of Action. Since the only issue to be tried was the class claim of a pattern or practice of sex discrimination, little evidence bearing on individual discrimination- claims of named class members has been presented. By agreement, those claims were to be addressed separately after determination of plaintiffs’ pattern or practice claim. Plaintiffs tried this case primarily on the strength of statistical evidence prepared by their expert witnesses, Drs. Killingsworth and Aboud, that was designed to show a university-wide work force imbalance and a treatment of women that was less favorable than of men with respect to specific terms and conditions of employment. Based on their statistics, plaintiffs ask the court to find that Stony Brook has certain facially neutral practices and policies that have a disparate impact on women, and from that, to infer that Stony Brook has a pattern or practice of discrimination against women. Plaintiffs also presented testimony by various Stony Brook administrators and former and present employees, in order to strengthen the inferences sought to be drawn from the statistical evidence. The court accepts plaintiffs statistics insofar as they show that Stony Brook’s work force is “sex-stratified”, in that women are distributed primarily in lower-level, and hold relatively few upper-level, teaching and administrative positions. The central issue in this case, therefore, was whether plaintiffs have shown that this sex-stratification is a result of illegal discrimination on account of sex, or that the stratification is otherwise the result of facially neutral criteria that fall disproportionately on women and that lack any business justification. Defendants challenged the validity of plaintiffs statistics, they questioned their underlying assumptions, they disputed the inferences plaintiffs urged the court to draw from the statistical picture, and they introduced independent statistical evidence of their own, prepared by Dr’s. Welch, Dillard, and Meyer. Defendants claim that since the beginning of the liability period, February 11,1974, Stony Brook has treated women and men equally. According to defendants, the workforce stratifications and imbalances that are shown by the raw numbers with respect to hiring, promotions, and salaries do not establish a pattern or practice of system-wide sex discrimination. C. Guiding Legal Principles. In order to prevail on a pattern and practice claim under Title VII plaintiffs must prove that sex discrimination was Stony Brook’s “standard operating procedure — the regular rather than the unusual practice.” International Brotherhood of Teamsters v. United States, 431 U.S. 324, 336, 97 S.Ct. 1843, 1855, 52 L.Ed.2d 396 (1977) (hereinafter “Teamsters ”). The legislative history of Title VII makes clear that plaintiffs must prove more than “isolated sporadic incident[s],” of discrimination. Id. Plaintiffs must show a repeated, routine, or generalized pattern of sexually discriminatory policies or practices. See id., quoting 110 Cong.Rec. 14270 (1964) (remarks of Senator Humphrey). In most Title VII cases plaintiffs present either a disparate treatment or disparate impact theory of discrimination. Either theory can be applied to a given set of facts. Teamsters, 431 U.S. at 335 n. 15, 97 S.Ct. at 1854 n. 15. Under either theory plaintiffs have the initial burden of establishing a prima facie case of discrimination. Id. at 336, 97 S.Ct. at 1855; Albemarle Paper Co. v. Moody, 422 U.S. 405, 425, 95 S.Ct. 2362, 2375, 45 L.Ed.2d 280 (1975); McDonnell Douglas Corp. v. Green, 411 U.S. 792, 802, 93 S.Ct. 1817, 1824, 36 L.Ed.2d 668 (1973) (disparate treatment); Guardians Association v. Civil Service Comm’n 633, F.2d 232, 239 (2d Cir.1980), aff'd — U.S. —, 103 S.Ct. 3221, 77 L.Ed.2d 866 (1980) (disparate impact). In order to prevail on a disparate treatment claim, plaintiffs must prove defendants acted with a discriminatory motive, although motive is inferable merely from differences in treatment. Teamsters, 431 U.S. at 335 n. 15, 97 S.Ct. at 1854 n. 15. In a disparate impact case, plaintiff must demonstrate that employment practices or policies that are facially neutral in their treatment of different groups in fact fall more harshly on one group than another and cannot be justified by business necessity. Teamsters at 335 n. 15, 97 S.Ct. at 1854 n. 15; Dothard v. Rawlinson, 433 U.S. 321, 329, 97 S.Ct. 2720, 2726, 53 L.Ed.2d 786 (1977). It is settled that “[wjhere gross statistical disparities can be shown, they alone may in a proper case constitute prima facie proof of a pattern or practice of discrimination.” Hazelwood School District v. United States, 433 U.S. 299, 307-08, 97 S.Ct. 2736, 2741, 53 L.Ed.2d 768 (1977); Grant v. Bethlehem Steel Corp., 635 F.2d 1007, 1015 (2d Cir.1980), cert. denied 452 U.S. 940, 101 S.Ct. 3083, 69 L.Ed.2d 954 (1981). The usefulness of statistics however, “depends on all the surrounding facts and circumstances.” Teamsters, 431 U.S. at 340, 97 S.Ct. at 1857. In this circuit, it is clear that plaintiffs can establish a prima facie case in a discriminatory impact case by statistical evidence alone. Guardians Ass’n v. Civil Service Comm’n, 630 F.2d 79, 88 (2d Cir.1980), cert. denied, 452 U.S. 940, 101 S.Ct. 3083, 69 L.Ed.2d 954 (1981); Grant v. Bethlehem Steel Corp., 635 F.2d 1007, 1017-18 (2d Cir.1980), cert. denied, 452 U.S. 940, 101 S.Ct. 3083, 69 L.Ed.2d 954 (1981). The Supreme Court has not defined with precision what showing of statistical disparity is required to make out a prima facie case, and has cautioned that the level is not inflexible — “statistics are not irrefutable” and may be rebutted. Teamsters, 431 U.S. at 339-40, 97 S.Ct. at 1856-57. Courts that have applied Title VII to universities have required plaintiffs show a gross statistical disparity with respect to each allegedly discriminatory claim in order to establish a prima facie case. See Wilkins v. University of Houston, 654 F.2d 388, 398 (5th Cir.1981). Once a plaintiff makes a prima facie showing of discrimination, the burden then shifts to the defendant to articulate a non-discriminatory explanation or justification for the inferences raised by plaintiff’s evidence. While the precise burden that shifts to the defendant necessarily varies according to the facts and theory of the case, generally the defendant must proffer a non-discriminatory explanation to rebut the inferences raised by plaintiffs’ evidence. The defendant may do so, for instance, by showing that plaintiffs’ proof is inaccurate or insignificant, Teamsters, 431 U.S. at 360, 97 S.Ct. at 1867; or that a facially neutral criterion that has a discriminatory impact is job-related, see Griggs v. Duke Power, 401 U.S. 424, 432, 91 S.Ct. 849, 854, 28 L.Ed.2d 158 (1971); Dothard v. Rawlinson, 433 U.S. 321, 329, 97 S.Ct. 2720, 2726, 53 L.Ed.2d 786 (1977); or that the claimed discriminatory pattern is a product of conduct that occurred before the act applied to defendants, Teamsters, 431 U.S. at 360, 97 S.Ct. at 1867, or is immunized by the applicable statute of limitations, United Airlines v. Evans, 431 U.S. 553, 97 S.Ct. 1885, 52 L.Ed.2d 571 (1977). Should defendants succeed in rebutting plaintiff’s prima facie case, the burden shifts back to the plaintiff to show that defendants’ presumptively valid reasons for rejection were in fact “a coverup for a * * * discriminatory decision,” McDonnell Douglas Corp. v. Green, 411 U.S. 792, 805, 93 S.Ct. 1817, 1826, 36 L.Ed.2d 668 (1973). Despite this elaborate technique for structuring the trial, once all the proof is in, the ultimate burden of persuasion remains with the plaintiff. Texas Department of Community Affairs v. Burdine, 447 U.S. 920, 100 S.Ct. 3009, 65 L.Ed.2d 1112 (1981). However, the Supreme Court recently emphasized that' the prima facie case method established in McDonnell Douglas was “never intended to be rigid, mechanized, or ritualistic.” United States Postal Service v. Aikens, 460 U.S. 711, —, 103 S.Ct. 1478, 1482, 75 L.Ed.2d 403 (1983) quoting Furnco Construction Co. v. Waters, 438 U.S. 567, 577, 98 S.Ct. 2943, 2949, 57 L.Ed.2d 957 (1978), and once defendant has presented evidence in a Title VII case, the duty of a district court is to decide whether the plaintiff has met its ultimate burden of pursuading the court that defendant has discriminated in violation of Title VII United States Postal Service v. Aikens, 460 U.S. at —, 103 S.Ct. at 1481. Applied to this trial, plaintiffs’ burden was to establish that after February 11, 1974 and with respect to its teaching and non-teaching professionals, Stony Brook has engaged in a pattern or practice of treating women less favorably than men, solely because they are women. To decide this question it is necessary to review and make findings concerning the structure of the university, the hiring procedures and policies, and the facts and statistical analyses pertaining to hiring, placement, promotion, tenure, and salaries. II. FINDINGS OF FACT. A. University Structure. 1. Academic Divisions. The State University at Stony Brook is a state-operated university founded in 1957 and designated in 1960 as one of four university centers of the State University of New York, pursuant to Article VIII of the Education Law of the State of New York. N.Y. Education Law § 352(3) (McKinney 1969 & Supp.1982). Stony Brook is a research university that offers degrees at the undergraduate, graduate, and professional levels. As chief administrative officer of Stony Brook, the president is directly responsible to the chancellor and trustees of the State University. With the advice of academic and professional employees, he is primarily responsible for the appointment of members of the teaching staffs, the recommendation and approval of salary increases, and the promotions in rank. During all relevant times during the liability period of this litigation John Toll was the president of Stony Brook. He has since been succeeded by John H. Marburger, III. Dr. Sheldon Ackley was special assistant to the president and assisted in many matters of personnel. Stony Brook is divided into a “Core Campus” and a “Health Sciences Center.” The core campus includes the College of Engineering and Applied Sciences, the W. Averell Harriman College for Urban and Policy Sciences, the College of Arts and Sciences, the Marine Sciences Research Center, and the Institute for Theoretical Physics. The Health Sciences Center (HSC) includes the Schools of Medicine, Dental Medicine, Social Welfare, Allied Health Professions, Basic Health Sciences (incorporated into the School of Medicine in 1980), and Nursing. Each of the academic units is headed by a dean, provost, or director, who is responsible to the academic vice president and who has primary responsibility for formulating academic policy within that unit. While the president retains primary responsibility for the policies and operation of Stony Brook, the academic vice-president has supervisory authority over all academic personnel and programs on campus. During the entire liability period, Dr. Sidney Gelber was the academic vice-president. 2. Work Force Structure. The employees of Stony Brook can roughly be divided into academic employees, non-teaching professionals (NTP’s), administrative personnel, and civil service personnel. Civil service and administrative personnel are not subjects of this litigation, which is brought on behalf of only academic employees (faculty) and NTP's. In some instances library employees have been dealt with by the parties separately, since they include both academic employees and NTP’s, and they present some special problems. All faculty, library, and NTP employees have been represented by collective bargaining agents since 1971. Academic employees are divided into five categories: professors, associate professors, assistant professors, instructors, and lecturers. A person is promoted to a full professorship after many years of distinguished university service marked by outstanding professional accomplishment. It is the highest academic rank within the university. An associate professor ranks immediately below full professor, and differs only in "the length and extent of service in academy or experience outside the university.” An assistant professor is the beginning level of full academic appointment. Stony Brook hires instructors infrequently and in unusual circumstances. “Lecturer” is a catch-all category often given to newly hired teachers who have not yet completed their PhD’s. After receiving their doctoral degrees, lecturers are eligible to become assistant professors. Other lecturers are hired to replace faculty on leave, or to assist a faculty member with research. From 1974 to 1977, Stony Brook hired 42 lecturers. People having the rank of lecturer, or an academic rank preceded by “clinical,” “visiting,” or “research,” hold a non-tenure track, “qualified” academic rank, and normally only for short periods of time. Except for the qualified ranks, the academic categories are linked for purposes of salary and promotion. For example, an instructor is normally promoted to the rank of assistant professor, an assistant professor to associate professor, and an associate professor to full professor. The library work force is made up of faculty employees and NTP’s. At the library faculty level are the director of the library and those holding titles of librarian, associate librarian, senior assistant librarian, and assistant librarian, all of whom are eligible for tenure. As with the regular faculty, these positions are also linked. Prior to 1977 librarians were classified into “PR” ranks along with NTP’s. They now have independent library ranks, but are still assigned to NTP ranks for salary purposes, although they have a separate salary schedule. Many of the NTP positions in the library require specialized language training. Some upper level NTP employees are “joint appointments” to the faculty, but are also assigned a PR rank. The remainder of the professional service consists of NTP’s. Each NTP rank consists of a wide variety of discrete jobs, each requiring different qualifications and experience. NTP positions are not linked, and while a few NTP employees are eligible for permanent appointment, most are not. Stony Brook has 146 NTP positions roughly divided into eight “PR ranks.” Each PR rank primarily establishes salary limits and is neither relevant for promotion nor, taken alone, indicative of responsibility except in the most general terms. Job applicants are interviewed for specific NTP jobs. Salary and rank are established before recruitment or interviewing begins. The highest PR rank is 8, which the president alone holds. Within a given PR rank, salaries cover a wide range, and the salary ranges of different PR ranks overlap. B. Affirmative Action Policies. Stony Brook has no official policies that explicitly operate to the disadvantage of women. On the contrary, Stony Brook’s administration is firmly committed to complying with the letter and spirit of Title VII. Stony Brook has acted in good faith, even before Title VII required it to do so, to ensure that it’s policies and practices have operated without discrimination based on sex. Equal treatment of women and minorities was a serious concern at Stony Brook throughout the 1970’s. Many of the policies and actions taken by Stony Brook administrators indicate an awareness of and a sensitivity to the treatment of women at Stony Brook, although it may be true that at least in part the impetus for these actions was due to the dissatisfaction voiced by various women’s groups on campus. In 1969, before Title VII applied to state universities, Stony Brook became subject to Executive Order 11246, which required it periodically to report regarding the status of women and minorities. In response, the president appointed Vera Rony to be the Affirmative Action Coordinator in the newly-created Office of Equal Opportunity, which reviewed all university affirmative action policies, prepared affirmative action reports, and processed discrimination complaints. Lee Yasamura, who was and continued to be the personnel director, replaced Rony as Affirmative Action Coordinator when Rony left that post in 1975. The president also created an Advisory Committee on Equal Opportunity (“president’s EOC”) to advise him on affirmative action policy within Stony Brook. He established equal opportunity committees (area EOC’s) in every vice presidential department within the core campus to define and monitor their respective affirmative action policies. He also required each university department to report to area vice-presidents on affirmative action matters. The chair-people of the area EOC’s were all members of the president’s EOC. Women have served , on these area EOC’s in substantial numbers. In the fall of 1971, anticipating a large amount of hiring at the HSC, Dr. Edward Pellegrino, Dean of the HSC, established policies and procedures for the recruitment of women and minorities and established an area EOC on that campus to monitor those procedures. In 1972, Academic Vice President Gelber established similar policies for the core campus. Official university pronouncements also indicated that affirmative action was a continuing goal of the administration. For instance, on April 16, 1971, President Toll stated: In nearly all university disciplines, the representation of minority groups and of women has been disproportionately low. Hence every operating unit of the university (academic, administrative, and staff) must make real efforts to recruit candidates from minority groups and women as replacements or occupants of new lines. The president termed compliance with this directive “both a legal and a moral requirement” and urged that Stony Brook “be in the vanguard of practical response to one of the major social concerns of our day.” Consistent with these policies, Stony Brook established specific directives, discussed below, regarding affirmative action in hiring and promotion. In December, 1973, the chairman of the Faculty Senate Executive Committee requested the president to appoint a “Salary Equity Task Force” (SETF) — comprised of four women and two men — to review the salaries of women and minorities at Stony Brook. The SETF recommended that the university make further studies regarding the differences of salaries between men and women. While Stony Brook published no formal study in response to this recommendation, it did conduct individual reviews of minority and women faculty whose salaries were more than one standard deviation below the expected salary, and awarded some employees retroactive salary increases. In 1975, Presidential Assistant Ackley prepared an affirmative action plan (1975 AAP), which found that women were “underutilized” at Stony Brook — that is hired in fewer numbers than predicted by their representation in the appropriate labor pools — by 8% to 9%. The 1975 AAP also presented a comprehensive university-wide program to comply with federal equal opportunity regulations. The president’s EOC, which also issued a report making suggestions regarding affirmative action, objected to the 1975 AAP. Similarly, after the 1975 Plan was submitted to the United States Commission on Civil Rights, the commission responded that it was dissatisfied with the plan because it failed to establish a time table or goal as to how the university would comply with various federal regulations. Stony Brook never formally replied to these objections. A second affirmative action plan, however, was prepared in 1980; it is discussed at greater length infra. Although women’s groups at Stony Brook expressed dissatisfaction with the administration’s policies regarding women, the evidence, taken as a whole, does not support any claim that the university acted pretextually or in bad faith in making its official pronouncements or attempting to establish and implement policies designed to ensure the non-discriminatory treatment of women. C. Hiring Procedures at Stony Brook. 1. In General. Stony Brook’s hiring procedures are facially neutral. The president appoints those faculty, who “... are, in his judgment, best qualified.” Although the president has nominal authority to hire faculty, NTP’s, and library staff, as a practical matter the hiring process at Stony Brook is decentralized, and many people, at various levels of the university, participate in these decisions. Faculty hiring decisions are primarily made by department chairmen acting on the recommendation of a hiring search committee. NTP decisions are made in the same manner, but by an area vice president acting on the recommendation of a search committee. Normally, the president ratifies these decisions through the appropriate vice president, and exercises only minimal and formal supervision. In the 1960’s, Stony Brook as a new university sought to establish quickly a national reputation by hiring faculty and administrators who themselves had national reputations.' For instance, Stony Brook hired C.N. Yang, a Nobel laureate in physics, in order to attract other outstanding faculty and students in the “hard” sciences. President Toll was hired from the department of physics and astronomy at the University of Maryland. H. Bentley Glass was a professor of biology at John Hopkins University before becoming Stony Brook’s academic vice president. These appointments and other similar ones were made outside of the normal hiring procedures. Their purpose was to enable Stony Brook to recruit in competition with other national universities. Stony Brook has acted in good faith to establish affirmative action policies which ensure that its hiring practices comply with the letter and spirit of the civil rights laws. While plaintiffs presented evidence of some instances where these procedures were not followed, the court finds these departures to be isolated instances and primarily made for non-discriminatory reasons, such as to attract well-known faculty. The clear weight of the evidence indicates that on the whole Stony Brook has complied with its declared internal policies for hiring both faculty and NTPs. . We turn next to an examination of these policies and to plaintiffs’ claims that they are “subjective,” “flexible,” and operate to the disadvantage of women. 2. Specific Hiring Directives. Sex is not a bonafide occupational qualification for any position at Stony Brook, and there is nothing in Stony Brook’s formal hiring policies that discriminates against women on account of sex. Stony Brook has actively recruited women and minorities during the liability period. On October 8, 1974, President Toll approved affirmative action policy statement P-9, which directed that: The university has a special obligation to take affirmative action to seek out candidates for employment from [groups that may be disadvantaged because they have not traditionally been represented in particular professional fields]. The university’s personnel appointment and recruiting procedures will provide for the implementation of this policy. P-9 established area EOC’s to monitor compliance with affirmative action directives within specific departments or areas within the university, and gave the president’s EOC general supervisory authority over the actions of the area EOC’s. P-9 required that all faculty appointments and promotions and all NTP appointments and promotions above the PR-4 level be submitted to an area EOC for advice on the extent and manner of recruitment, before actual recruitment began. This involved the area EOC’s in the hiring and promotion process at the earliest possible date. Although approval by an area EOC was not necessary for faculty appointments of less than one year or for NTP appointments at a level of PR-3 or below, P-9 nevertheless directed that the appropriate area EOC be notified of the short-term appointments in a “timely manner.” The procedures for faculty hiring, which are representative of the normal hiring procedures at Stony Brook, were announced July 21, 1975 in PR-320. The first step in recruitment is for the department or school seeking to fill a vacancy to prepare a “recruitment plan,” whose proposed procedures are reviewed by the appropriate area EOC prior to actual recruitment. These procedures include posting and advertising a vacancy in newspapers and professional journals. The evidence shows that no appointment could be made unless these procedures were followed, and plaintiffs presented no evidence to suggest these procedures were normally disregarded. PR-320 required that the person responsible for making the hiring recommendation receive authorizations to begin recruitment and to form search committees to assist in recruitment. It also required that each search committee prepare a “SUSB68-1 EEO Affirmative Action Worksheet” for each candidate interviewed and considered. Stony Brook generally followed these procedures. SUSB-68 was used in all recruitment procedures during the liability period. Form SUSB-68 required each search committee to record affirmative action data for each proposed appointment, including information about the race and sex of all applicants, advertisements placed, and responses received. It also required that each applicant meet certain minimal qualifications. In addition, SUSB-68 required documentation of each applicant’s interview and each applicant’s declination information. All of this information was forwarded to the area EOC along with the name of the applicant or applicants selected for the position. After the area EOC reviewed the SUSB-68 and approved the candidate, the candidate’s name was forwarded to the vice president’s office who reviewed the candidate’s file and notified the applicant. Stony Brook had similar procedures for NTP positions. P-36 governed NTP hiring, and required extensive advertising and posting for all NTP vacancies. It imposed broader advertising requirements than P-320, particularly in respect to posting vacancies at other SUNY campuses. The evidence before the court did not indicate that these procedures were disregarded for NTP or library hiring. Plaintiffs maintain that despite these elaborate policies to promote affirmative action and ensure non-discriminatory hiring, the procedures were flexible and frequently circumvented- Plaintiffs claim that department chairmen, most of whom were male, make all hiring decisions. Plaintiffs also claim that search committees “rubber-stamp” employment decisions made — presumably by department chairmen — before the selection process begins. The testimony of Dr. Entine and of plaintiff Miller shows that some search committees “pre-selected” candidates and then tailored recruitment to comply with procedural requirements. Plaintiff Miller admitted she had participated in three “phony” searches, each of which was to hire a full professor. She testified that two of the three searches were initially honest, but had been unsuccessful. She testified that “when it came to assistant professors, lecturers, instructors, then, of course, searches were probably much more honest.” Defendants respond that in the few instances that committees preselected candidates, they did so to attract top candidates who had received competing offers from other universities. Defendants claim that to compete effectively, Stony Brook needed to make a firm offer for a position, and as a result, search procedures were abbreviated. The court is satisfied that these were isolated, occasional events that usually involved searches for a full professor needed to fill a particular teaching need in a department. University personnel disapproved of such practices, and these isolated abuses, given the university’s reaction, are insufficient to prove that Stony Brook had a university-wide pattern and practice of ignoring formal hiring directives, or that search committees normally operated to the disadvantage of women. As to searches for positions other than full professor, plaintiffs presented no credible evidence to indicate that members of the search committees were discriminatorily selected or that the search procedures themselves were discriminatory. The evidence showed that the area EOC’s normally performed their supervisory responsibilities with diligence, and when necessary, refused to approve searches. Defendants showed, and plaintiff's witnesses conceded, that area EOC review was particularly vigorous for entry level positions such as assistant professor, but was admittedly less rigorous in searches for full professors, apparently in deference to the specialized teaching needs of particular departments. Furthermore, women — including several of the class representatives— regularly served on these search committees. The court accepts that the failure to comply with formal search guidelines in these isolated instances indicates that search procedures were flexible, but in view of the context in which the guidelines were circumvented and the infrequency of those departures, the evidence is insufficient to support an overall inference that all hiring procedures generally were “flexible” or routinely flouted. Nor do the departures support an inference that established policies and procedures were not followed in most searches. Nor do they support on inference that they were motivated by or resulted in unlawful sex discrimination. As a practical matter, the ordinary search procedures were not well-suited to the practicalities of hiring senior faculty and top level administrators, and the court accepts this as a non-discriminatory explanation for such searches. Moreover, without evidence that these procedures resulted in hiring a disproportionate number of men in preference to equally qualified women, or that search committees were in some other way biased against women — evidence that is totally absent in this case — these few departures from the search committee procedures simply are not probative of plaintiffs’ claim of a pattern and practice of sex discrimination. D. Statistical Proof of Discrimination. As already indicated, plaintiffs rest their claim primarily on the statistical evidence presented at trial. Through that evidence they have attempted to establish a pattern or practice of sex discrimination in the hiring, placement, promotion, tenure, and salary treatment of professional women at Stony Brook. They also argue that the overall work force picture at Stony Brook implies the existence of sex discrimination in the university’s employment decisions. Plaintiffs’ general presentation is superficially persuasive. Between 1972 and 1977 the professional work force was predominantly male, with faculty ranging from 13 to 19% female and NTP’s from 26 to 40% female. From 1972 to 1977 the majority of newly hired professionals were male, with 28% female faculty hired and 47% female NTP’s hired. The raw data also shows more males among the higher ranking faculty and NTP positions whether viewed as work force at a particular time, or as newly-hired personnel during the various periods. Plaintiffs also show that some ■ of Stony Brook’s departments have no females at all, that other departments such as library and nursing are predominantly female, and that generally females are concentrated at the lower end of the economic and promotion scales. All of the foregoing certainly implies that in abstract terms, women have not been treated equally with men as to those aspects of our society and economy that have resulted in being employed at Stony Brook. They do not, however, necessarily imply a pattern or practice of unlawful sex discrimination by the university. The problem of unlawful discrimination is far more subtle and elusive than raw numbers relating to the work force, rankings, and salaries at any particular time. As applied to this case, the issue of discrimination involves such questions as: What period of time do we consider? What is the impact of a separate trial of the pattern or practice issue? What actually happened during the liability period, the period for which defendants can be held legally responsible? How many women were actually available to be hired for the positions at Stony Brook? How many women were realistically available to be promoted and to be granted tenure? Does the conduct of defendants during the liability period suggest discrimination is the rule rather than the exception? In general terms, when one looks behind plaintiffs’ raw numbers, which are fairly impressive on first glance, one finds that the factors that have caused the present imbalances between men and women at Stony Brook are most likely a result of historical economic and social conditions that long predate the liability period, and that in many cases are unrelated to Stony Brook itself. If a work force imbalance results from intentional discrimination by Stony Brook, that occurred prior to 1972, when discrimination on grounds of sex was not unlawful for universities, it affords no basis for finding either a post-1972 pattern and practice of unlawful sex discrimination or a post-1974 liability to plaintiffs, United Airlines v. Evans, supra. In short, the raw numbers, taken alone, mean relatively little. With claims of hiring, placement, promotion and tenure, they must be compared with the applicable availability pools, i.e., the' percentage of women in the work force who are qualified and available to fill the positions in question. Since a university such as Stony Brook demands highly educated, specially qualified people to fill particular positions, its treatment of minorities cannot be evaluated in the same manner as that of, say a trucking company or factory that can hire many people with less differentiated qualifications. Instead, to fairly interpret plaintiffs’ claims it is necessary to evaluate the numerical data carefully in light of the available people and their relative qualifications, as well as to consider the external facts of history, the employment market and social factors over which defendants have no control. With these cautions in mind, we will consider plaintiffs’ claims and the evidence relating to hiring, placement, promotion, tenure, and salary. 1. Hiring. Plaintiffs claim that defendants have discriminated against women in hiring faculty; there is no claim of discrimination in hiring NTP’s. In general terms, in order to evaluate the claim of discrimination in faculty hiring we must compare the male/female ratios of hirings with the male/female ratios in the relevant labor force, the so-called availability pools. Hazelwood School District v. United States, 433 U.S. 299, 97 S.Ct. 2736, 53 L.Ed.2d 768 (1977). Defendants’ pre-act hiring conduct could have been relevant to the pattern or practice claim, but plaintiffs presented no evidence of pre-act availability pools against which to compare the hiring decisions. Stony Brook hired faculty for approximately ten years before 1972 when the act first applied to public employers and universities. During the same ten-year period of time there were significant changes in the numbers and percentages of women entering various teaching fields. Consequently, it would be improper to read back into the sixties availability pool figures based on conditions in the early and mid seventies. An additional conceptual problem is presented by two circumstances: (1) prior to 1972 sex discrimination at Stony Brook was not illegal, and (2) while sex discrimination at Stony Brook would have been illegal between March 24, 1972, the effective date of the statute, and February 11, 1974 the cutoff date of the statute of limitations, that discrimination could not be remedied in this action. The problem is, in light of these two circumstances, what weight if any, should the court give in this “pattern and practice” trial to evidence of discriminatory actions that occurred prior to February 11, 1974? Some help is offered by the majority opinion of the Supreme Court in Hazelwood School District v. United States, 433 U.S. 299, 97 S.Ct. 2736, 53 L.Ed.2d 768 (1977), a racial discrimination, pattern and practice suit wherein the court noted that work force statistics comparing the teaching staff with the racial composition of the qualified public school teacher population and the relevant labor market could make, out a prima facie statistical case. Indeed, the Court of Appeals had found that such work force statistics conclusively proved the government’s pattern and practice lawsuit. Criticizing that view, the Supreme Court stated: The Court of Appeals totally disregarded the possibility that this prima facie statistical proof in the record might at the trial court level be rebutted by statistics dealing with Hazelwood’s hiring after it became subject to Title VII. Racial discrimination by public employers was not made illegal under Title VII until March 24, 1972. A public employer who from that date forward made all its employment decisions in a wholly non-discriminatory way would not violate Title VII even if it had formerly maintained an all-white work force by purposely excluding Negroes. 433 U.S. at 309, 97 S.Ct. at 2742. In footnote 15 to this passage, the court noted: In light of the Supreme Court’s view and in the absence of any evidence of the relevant labor market before 1972, this court will concentrate its analysis of the statistics presented on those acts beginning in 1972, conscious of the fact that while defendants cannot be held liable to particular plaintiffs for acts that occurred outside the limitations period, i.e., before March 1974, conduct occurring between 1972 and 1974 is relevant on the issue of whether or not there existed a university-wide pattern or practice of unlawful sex discrimination. We do not have here a situation such as contemplated by the last few lines of Hazelwood’s footnote 15, supra, because beginning even before the effective date of Title VII Stony Brook adopted significant and extensive changes in its processes for arriving at employment decisions. The facts underlying plaintiffs’ claim of discrimination in faculty hiring are simple. Plaintiffs’ exhibits 147 and 183 summarize the raw data of hiring for each school year from 1972-73 to 1976-77 and break it down by departments and by academic rank at the time of hiring. That basic data sketches the hiring picture. As to the relevant labor market, plaintiff offered no independent evidence, but instead relied upon calculations of the availability pools that were included in the 1975 and 1980 affirmative action plans (AAPs) prepared by Stony Brook pursuant to governmental regulations. Plaintiffs used the AAP’s to show that Stony Brook had “underutilized” women in various departments. According to federal regulations released by the Office of Federal Contract Compliance Programs, (OFCCP), women are underutilized when the estimated percentage of the women in the labor market exceeds the percentage of women actually hired. If Stony Brook hired applicants without regard to sex, the percentage of women hired should be approximately equal to the percentage of women in the relevant availability pool. Women are therefore over or underutilized to the extent that Stony Brook hires more or less than the “predicted” number of women. The predicted number of women is obtained by multiplying the percentage of women in the relevant availability pool by the total number of people hired in that category. In the 1975 AAP, defendants reported that of the 828 members of the faculty they underutilized women by 14, or 1.7% of the total aggregated departmental units. Data in the 1980 AAP was more detailed. It showed that as of 1978 Stony Brook underutilized women by 25, spread over 16 departments and schools, and overutilized women by 28, spread over 18 departments and schools, leaving a university-wide net overutilization of 3. ,. Plaintiffs argue that these statistics are more probative of defendants’ discriminatory hiring practices than first appears, because both AAP’s used national availability pools that, which plaintiffs claim, understate the “actual” availability of women, and (2) those studies show that Stony Brook hired women in “traditionally female departments and grossly under-hired or do not hire at all in most other disciplines.” For instance, from 1974 to 1977, 16 of 66 (24%) of all female tenure track faculty hired by Stony Brook were hired into the school of nursing, and 47% of the women hired were hired into the school of nursing or the libraries — traditionally female departments. The record does not show an availability pool for these positions. Plaintiffs argue that when Stony Brook’s hiring decisions are analyzed without the “all-female schools of nursing and social welfare” the result is that only 78.9% of the female faculty expected to be hired were actually hired (i.e., expected 26.6 — actual 21), the percentage of women hired by the University decreases from the figure used by Dr. Welch, 25% (41/i64), to only 14.6% (21/l44).” Plaintiffs also argue that between the 1975 and 1980 AAP’s, the percentage of women on the Stony Brook faculty decreased, although the raw number of female faculty members increased slightly by seven. During this time, the number of men on the faculty increased by 106. While these figures are superficially striking, the court is unpersuaded by plaintiffs’ continued attempts to prove discrimination by relying on raw data that ignore availability pools. The court does not find either party’s statistical evidence to be conclusive proof of discrimination (or its absence) and weighs most heavily the data that “measures” discrimination by comparing the actual number of women hired with the expected number in an appropriate availability pool. Neither AAP evaluated the discrimination and hiring claims according to the actual number of applicants of Stony Brook (which theoretically was available from the SUSB-68 forms), but instead used availability pools which were an estimate of the percentage of qualified nationally-available females in the various disciplines. Plaintiffs attempt to draw favorable inferences from the slight underutilization shown in defendants’ AAPs — overall in the 1975 AAP and in selected departments in the 1980 AAP. The underutilizations in those reports are equivocal and unpersuasive, however, because (1) the slight adverse inference available is insufficient to overcome the weakness inherent in the imprecise availability data, (2) in 1980, at least, they are counterbalanced by equivalent overutilization, and (3) neither AAP can fairly be characterized as showing a “gross statistical disparity.” Neither does the court accept plaintiffs’ suggestion that the AAP’s underestimate the availability pools of women by using national, rather than regional, availability pools. The parties agree that regional pools would increase the apparent availability of women. The court finds the more appropriate labor pool for both the 1975 and 1980 AAP’s is national. Stony Brook is a highly regarded research university which attracted nearly 40% of its faculty (in 1976-77) from foreign countries or from states other than those from the “region” —New York, Pennsylvania, New Jersey, New England, California — suggested by the plaintiffs. Stony Brook attracts faculty applicants from throughout the country. Its faculty has earned a national reputation, particularly in the hard sciences, and it recruits through national publications. It would be inappropriate to categorize Stony Brook's hiring as limited regionally, despite plaintiffs’ evidence that from 1974-77, Stony Brook hired over 70% of its new faculty from California or the northeastern United States, and nearly 50% from New York, Pennsylvania, and New Jersey. The “national” availability pools in the 1975 AAP were a better measure of Stony Brook’s applicants than the regional predictions urged by plaintiffs. The court also accepts the “national” availability pools used in the 1980 AAP. Dr. Malcolm Agostini, Special Assistant to the President for Equal Opportunity, prepared the 1980 AAP availability pools using an 8-factor analysis according to the guidelines established in Revised Order No. 4 issued by the OFCCP. The availability pools were complied by dividing the faculty into two categories: full and associate professors, and assistant professors and below. The availability for full and associate professors was determined by calculating the percentage of women on faculty at research universities found by the Carnegie Council publication, “A Classification of Institutions of Higher Education, Revised Edition, 1976,” to be “similar” to Stony Brook. That percentage established the “expected” number of women on Stony Brook’s faculty. The availability pools for “assistant professors and below” was determined by calculating the percentage of Ph.D.’s awarded to women between 1973 and 1977 in each of eight academic disciplines. The Stony Brook faculty was subdivided into these eight categories and the expected number of faculty was based on the number of Ph.D.’s awarded nationally in each category. The 1980 AAP and its methodology was endorsed by the president’s AEOC and is accepted by the court. While the proper availability pool necessarily depends on the specific hiring practices of Stony Brook, other courts considering Title VII claims brought against universities have also found the appropriate faculty labor pool to be national. See, e.g., Wilkins v. University of Houston, 654 F.2d 388 (5th Cir.1981); Presseisen v. Swarthmore College, 442 F.Supp. 593 (E.D.P.1977) aff'd mem., 582 F.2d 1275 (3rd Cir.1978). Plaintiffs’ seek to excuse their failure to establish an appropriate availability pool by criticizing defendants for failing to analyze and summarize the information available on the SUSB-68 forms filled out by all applicants. Defendants respond, correctly, that the underlying burden was on plaintiffs and that it was their failing, not defendants’, which deprived the court of that information. Plaintiffs also argue that defendants mask the actual discrimination in the 1975 and 1980 AAP’s by overaggregating departments. In the 1980 AAP, women were underutilized in 16 departments and overutilized in 18 departments; university-wide, women were overutilized by 3. Plaintiffs claim the excessive aggregation enables Stony Brook to hire many more than the expected number of women in traditionally female departments and hire no women in traditionally male departments, yet still show no university-wide underutilization of women. The court finds to the contrary, however. The 1980 AAP showed that women were underutilized in the biological sciences, the HSC, and the libraries, but overutilized in the College of Arts and Sciences. There are 7 individual departments in which women are underutilized by more than 1 person, and an equal number of departments in which they are overutilized by more than 1. But taken as a whole the over and underutilizations cross cut departments and university divisions and do not follow a pattern of what is “traditionally” male or female. For instance, the 1980 AAP shows that women are underutilized in the departments of biology, political science, and physiology/biophysics, but overutilized in chemistry, history, and pathology. Indeed, plaintiffs refer to nursing and the library as traditionally female occupations, but there was neither over- nor underutilization in nursing, and in the library women were underutilized by 2. Thus, neither of those two departments could have skewed the picture in defendants’ favor. Finally, many of the departments in which women are underutilized combine few faculty members with a low availability percentage. There were 20 schools and departments that had no female faculty in 1978. In 9 of these, however, when the total faculty is multiplied by the availability pool percentage, the result is an expected female faculty component of less than .5, hardly a factor on which to rest claims of system-wide discrimination. This conclusion is buttressed by the report of defendants’ experts, Welch and Dillard (Welch Study), which, using a slightly different method of establishing a labor pool, showed that Stony Brook hired slightly fewer than the expected number of women. The Welch study concluded that the underutilization of women was “not statistically significant” and “could easily have occurred at random from a sex-blind hiring process.” While the Welch study made a variety of questionable exclusions and alone is not conclusive, it does tend to indicate that Stony Brook hired approximately the expected number of women at the higher faculty ranks. While the court does not attach great significance to this study and does not decide whether, taken alone, it would be sufficient to wholly rebut plaintiffs’ prima facie case had they made one, it does tend to corroborate the court’s conclusion. Plaintiffs argue that under Connecticut v. Teal, 457 U.S. 440, 102 S.Ct. 2525, 73 L.Ed.2d 130 (1982), evidence of defendants’ non-discriminatory university-wide hiring is no defense to plaintiff’s showing of underutilization of women in specific departments and divisions. But this argument ignores the implications of Hazelwood, supra, as to pre-act discrimination when applied to the early employment history at Stony Brook. Moreover, Teal does not apply here. In a Title VII action brought by an individual, disparate impact analysis is a two-step process. First, once the individual shows that a facially neutral criterion has an adverse impact on plaintiff’s class, the individual plaintiff is entitled to the inference that the employer discriminated against the individual. Then, if the employer is unable to come forward with a business justification for the criterion, or is unable to come forward with a non-discriminatory explanation as to the treatment of the individual, the individual prevails. Unlike the individual plaintiffs in Teal, plaintiffs here are a class of women seeking to prove by disparate impact analysis that Stony Brook has a pattern and practice of discrimination against women. If successful, that finding would then enable individual plaintiffs to rely on an inference of discrimination later when they seek to prove their individual claims. In Teal, four blacks proved they were not considered for promotion because they failed a written examination that had a proven adverse impact on blacks. The district court had dismissed their Title VII claims because the employer proved that although the examination had an adverse impact, blacks as a group were actually favored by the employer’s ultimate promotion decisions. The Supreme Court held that evidence that an employer’s “bottom line” practices had no adverse impact on blacks as a group was no defense to the proven claims of discrimination brought by individuals. The error of the district court in Teal was to foreclose proven and unrebutted individual claims of discrimination by looking to an employer’s treatment of a group. Here, the issue is whether Stony Brook’s neutral criteria have an adverse impact upon a group, and upholding Stony Brook’s defense against plaintiffs’ class action claims would not foreclose valid individual claims of discrimination, as the “bottom line” defense did in Teal. The court concludes that plaintiffs have failed to establish that Stony Brook has a pattern and practice of sex discrimination in hiring its faculty. 2. Placement at Hire. Plaintiffs claim that Stony Brook has discriminated against its female faculty NTP’s and librarians in terms of their placement at the time they were originally hired. These claims will be discussed separately. (a) Faculty Placement at Hire. On this issue plaintiffs introduced statistics to demonstrate that women with the same qualifications as men — measured in terms of age, level of education, field of degree, and total years of teaching experience — were placed in lower academic ranks. The percentage of hired male faculty initially placed as associate or full professors (26%) is consistently higher than the comparable percentage for women (13%). Conversely the percentage of hired women placed into the non-tenure track rank of lecturer (32%) was consistently higher than the comparable percentage for men (12%). Plaintiffs introduced a multiple logit regression analysis to show that from 1972 to 1976, women as compared with men with “comparable qualifications” were initially assigned to lower faculty ranks. That method, however, is incapable of recognizing linked variables such as the normal progression of faculty ranks. Instead, it considers each faculty rank as if it were unrelated to the others, and therefore did not consider prior faculty rank or experience among the variables to be analyzed. As a result of this defect plaintiffs’ statistics, while having superficial appeal, fail to consider important and legitimate factors relevant to initial placement. For instance, their statistics would group a tenured full professor awarded a Ph.D. in 1968 with a lecturer who also received a Ph.D. in ’68. Obviously the tenured full professor’s teaching experience would be a relevant factor in determining rank at hire. Moreover, prior academic rank is itself a measure of academic achievement, as defendants correctly note, and most faculty candidates would be reluctant to accept a position from Stony Brook below their rank at another university. In analyzing the same basic data defendants used an ordered logit regression analysis with control variables similar to plaintiffs’ but which additionally controlled for “rank immediately prior to hire”. An ordered logit analysis can consider a continuous dependent variable, such as prior rank, while a multiple logit analysis, used by plaintiffs, cannot. The variables employed in defendants’ statistics were: type of educational degree, lapsed time since receipt of highest degree, and academic rank achieved elsewhere. Their ordered logit analysis showed no statistically significant disadvantage to women faculty in initial placement. Defendants’ expert testified that had their studies “omitted information on prior rank, the result would have been a finding of slight statistical significance to suggest that women were placed into lower ranks than men.” The court recognizes that in general terms it is somewhat unrealistic to even attempt an evaluation of the initial placement of faculty on a university-wide basis, because the initial placement process is not carried out on such a basis. Stony Brook has no specific criteria for initial faculty assignment. Rather, Stony Brook recruits and interviews faculty for specific positions, and each separate department evaluates an applicant according to the vacancy it seeks to fill. Academic personnel are not first hired and then assigned to rank; instead, they are interviewed and hired to fill specific teaching positions. As a result, insofar as they attempt to isolate general factors governing initial placement both plaintiffs’ and defendants’ statistics are unrealistic because they seek to describe a condition that does not exist. More appropriately, their statistics should.be viewed as describing the backgrounds of facu