# American Federation of State, County & Municipal Employees v. Nassau

> District Court, E.D. New York · August 24, 1992 · 799 F. Supp. 1370

URL: https://www.frixlaw.com/law-library/cases/1379222

## Case

- **Full name:** AMERICAN FEDERATION OF STATE, COUNTY AND MUNICIPAL EMPLOYEES, AFL-CIO, (AFSCME); The Civil Service Employees Association, Inc., Region I/Local 1000, (CSEA); Rita Wallace, Rachel Braver, Dorothy Garage, Linda Kelly, and Lois Whitely on Behalf of Themselves and All Others Similarly Situated, Plaintiffs, v. COUNTY OF NASSAU; Its County Executive; The Comptroller; The Members of the County Board of Supervisors; And the Members of the Civil Service Commission, Defendants
- **Court:** District Court, E.D. New York
- **Decided:** August 24, 1992
- **Citations:** 799 F. Supp. 1370; 1 Wage & Hour Cas.2d (BNA) 201; 60 Empl. Prac. Dec. (CCH) 42,004; 60 Fair Empl. Prac. Cas. (BNA) 465; 1992 U.S. Dist. LEXIS 12983; 1992 WL 206398
- **Precedential status:** Published
- **Opinion:** Opinion by Glasser
- **Judges:** Glasser
- **Cited by:** 7 later opinions in the Frix Law Library

## Citator (automated)

- No negative treatment found by the automated citator. That is not the same as a confirmation that the case is good law; read the citing cases.
- Full citator and citing cases: https://www.frixlaw.com/law-library/cases/1379222

## How later opinions describe it (automated extraction)

- noting that Dr. Howarth confirmed "that there is indeed a gap between the salaries of men and women in Nassau County that is not explained by job specifications alone"
- applying traditional Title VII intentional discrimination standards to Title VII intentional wage discrimination claim based on sex

## Opinion text

MEMORANDUM AND ORDER
GLASSER, District Judge:
This class action was brought pursuant to Title VII of the Civil Rights Act of 1964, 42 U.S.C. § 2000e et seq., and to the Equal Pay Act of 1963, 29 U.S.C. § 206 (d). This court has jurisdiction pursuant to 42 U.S.C. § 2000e — 5(f)(3), 29 U.S.C. § 216 (b), and 28 U.S.C. §§ 1331 and 1343(3), (4). The named plaintiffs are: the American Federation of State, County and Municipal Employees, AFL-CIO, (AFSCME), an international labor union; the Civil Service Employees Association Inc., Region I/Local 1000, (CSEA), a regional affiliate of AFSCME which represents approximately 13,800 employees of the County of Nassau; and several female employees of the County of Nassau. These plaintiffs represent a class comprised of all women employed at any time since July 28, 1982 in a Nassau County civil service job title with respect to which at least 70% of the incumbents have been female. The defendants are the New York State County of Nassau, the Nassau County Executive, the Nassau County Comptroller, the members of the Nassau County Board of Supervisors, and' the members of the Nassau County Civil Service Commission.
As filed, the first amended complaint alleged violations of Title VII and of the Equal Pay Act by the County as to its compensation system; more specifically, the unions and the individual plaintiffs alleged that the County “discriminates in compensation on the basis of sex by paying historically female job classifications less than historically male classifications which require an equivalent or lesser composite of skill, effort, responsibility and working conditions.” Amended Complaint 113(D). These allegations were advanced on behalf of a class “of all employees, male and female, employed within the applicable limitations period by the County of Nassau in historically female sex-segregated job classifications in the classified service."
Id.
¶ 3 (B). The complaint defined “historically female sex-segregated job classifications” as those “classifications in an occupational group or job family for which the entry level classification is or ever has been 70% or more female.”
Id.
Further, the individual plaintiffs alleged specific violations of Title VII and of the Equal Pay Act.
Id.
1Í1Í10-19. The amended complaint sought injunctive relief, back pay, liquidated damages, costs, and attorney’s fees.
Id.
H 24.
*1373
On motion of the defendants, this court on May 17, 1985 dismissed the Equal Pay Act claims of former plaintiffs Stephen Goldberg and Fred Jordan as well as the Equal Pay Act claim of plaintiff Lois Whitely; the court also dismissed the Title VII claims of all plaintiffs insofar as they sought relief under a disparate impact theory.
AFSCME v. County of Nassau,
609 F.Supp. 695 (E.D.N.Y.1985). On July 13, 1987, this court dismissed the Title VII claims of all the male plaintiffs. The court also certified at that time a class comprised of all female employees of the County of Nassau who, since July 28, 1982, have worked in Nassau County civil service jobs that have ever had at least 70% female employees. The court designated the named individual plaintiffs and the two union plaintiffs to represent the class. The certification did not apply to the Equal Pay Act claims.
AFSCME v. County of Nassau,
664 F.Supp. 64 (E.D.N.Y.1987). Subsequently, named plaintiff Odessa Colvin dismissed her Title VII claims with prejudice, and named plaintiffs Laurie Gillibertie and Erna Fluhr were removed by stipulation.
The parties stipulated to bifurcate the trial for this action; the liability segment began on November 27,1989 and concluded with final arguments on May 11,1990. For the reasons indicated below in the findings of fact and conclusions of law made pursuant to Federal Rule of Civil Procedure 52(a), this court holds that the plaintiffs have failed to prove violations of Title VII or of the Equal Pay Act by a preponderance of the evidence as to all claims except the Title VII claims of the class members employed as police detention aides. Accordingly, judgment shall be entered for the defendants on all other claims.
FINDINGS OF FACT
I. The Nassau County Compensation System
In order to consider the plaintiffs’ claims of sex discrimination by the County of Nassau, it is necessary first to provide a coherent view of the system under which Nassau County classifies jobs, determines salaries, and incorporates changes into its compensation system. Accordingly, these findings of fact must begin with an overview of the present structure of the Nassau County compensation system; further, it is necessary to review the process by which the present system was created as well as the manner in which it has been modified since its inception.
'' One significant terminological observation must be made at the outset: The parties and the witnesses at trial employed the terms “female-dominated” to refer to job titles for which 70% or more of the incumbents are women, “male-dominated” to refer to job titles for which 70% or more of the incumbents are men, and “mixed” to refer to job titles that are neither female-dominated nor male-dominated. Although counsel at times introduced confusion as to whether or not they were observing rigorous distinctions in their phraseology (see, e.g., Plaintiffs’ Post-Trial Brief at 1: counsel appear to equate the term “female-dominated” with “predominantly female”) the court will, for the sake of clarity, use the terms “female-dominated”, “male-dominated”, and “mixed” only as defined above. As used by the court, the terms “female-dominated” and “predominantly female” are not equated; as used by counsel, the possibility of equivalence between such phrases as “predominantly female” and “female-dominated” must be discerned from context.
A.
Overview: The Present Structure
Nassau County, New York, employs over 19,000 people in approximately 1,500 job titles. As with all local governments in the State of New York, employment by the County is regulated through the New York State Civil Service Law. The Civil Service System — administered by the Nassau County Civil Service Commission — divides county employees into “classified” and “unclassified” service. Tr. 1231. “Unclassified” positions are those of elected officials, department heads, and board members. “Classified” positions — the vast majority of County jobs — are subdivided into an “exempt” class, a “competitive” class, a
*1374
“non-competitive” class, and a “labor” class. Tr. 1231. The “exempt” class is generally comprised of deputy commissioners and of certain professionals. The “labor” class is that for which no training or experience requirements are established. The “non-competitive” class is that for which training and experience requirements are established but for which examinations are impracticable. Tr. 1231-32. Positions in non-competitive jobs are filled by discretionary hiring at individual County departments — subject to confirmation by the Civil Service Commission that the hired applicant satisfies the training and experience requirements for the particular position. Tr. 1257-58. Finally, the “competitive” class (in which the greatest number of Nassau County jobs are found) is that for which competitive examinations are required by the New York State Civil Service Law. Tr. 1230-33. The Civil Service Commission posts and distributes notices of forthcoming examinations. Tr. 1254; see, e.g., DX UUU-4. Pursuant to state law, open positions in a competitive job must be filled from an “eligibility list” of applicants compiled and ranked on the basis of examination results. Tr. 1233. Candidates on the list must be selected in a modified rank order; that is, any open position must be offered to one of the top three eligible applicants. Tr. 1249. This is known commonly as the “one-in-three” rule. Tr. 1249.
The salaries of Nassau County employees are determined by job titles and by seniority. All job titles other than those of police personnel and of college faculties— and other than those which are either unclassified or exempt — are assigned to one of 27 salary “grades.” PX 672. The minimum and maximum salaries for a particular grade are identical for all job titles at that grade. Further, an employee within a job title progresses through regular salary “steps” in a given grade; progression from one step to another is grounded on seniority. Tr. 1311-12. However, the number of steps through which a particular employee moves in the grade for her job title is not the same for all employees; rather, it is dependent on the first date of the employee’s tenure with the County. Employees who were hired on or before June 30, 1967 move through seven steps for their grades (“Plan A”); employees who were hired after June 30, 1967 but on or before December 31, 1976 also move through seven steps for their grades (“Plan B”); and employees who were hired on or after January 1, 1977 have 15 steps for their grades (“Plan C”). DX ZZZ-2; Tr. 1311-12.
Compensation at each grade and step has increased through the years pursuant to “across-the-board” pay increases negotiated in collective bargaining agreements between the County and the plaintiff CSEA (which represents most of the employees whose jobs fall within this salary grade system, Tr. 1301). Tr. 1310-11. Also, individual job titles are at times “upgraded" either through the collective bargaining process or through actions taken by the Nassau County Board of Supervisors on the recommendation of the Civil Service Commission. The Nassau County Civil Service Commission has primary responsibility for evaluating proposed job titles and for reevaluating job titles or salary grades that are sought to be modified. The Civil Service Commission may recommend such additions or changes, but final authority to create a new title or to change the title or the grade of an existing job rests with the Nassau County Board of Supervisors. However, without a recommendation from the Civil Service Commission, a proposed modification is, for all practical purposes, formally rejected.
B.
The 1967 Cresap Process
The present compensation system of Nassau County is a direct descendant of a comprehensive job and salary evaluation process conducted on behalf of the County between 1964 and 1967 by the management consulting firm of Cresap, McCormick and Paget (“Cresap”). The objectives of the Cresap process were to provide a rational classification of County job titles and to establish equitable and competitive salaries. See PX 621 at 1-1 (final report of Cresap to Nassau County, dated June 1966 and entitled: “Nassau County — A Proposed Salary Plan”); see PX 1104 at 2
*1375
(preliminary report of Cresap to Nassau County, dated September 1964 and entitled: “Nassau County: Salary Plan for County Employment”). Cresap determined the parameters and the methodology of the study, trained a team of four Nassau County employees to conduct the study, and exercised regular supervision over the project. PX 621, passim.
The job evaluation process began with the distribution of questionnaires to several thousand County employees. PX 621 at II — 1; Tr. 295. The County team of four and the Cresap professionals then set out to create a taxonomy of County job titles and of job characteristics. PX 621 at II — 1 to II — 2; Tr. 298. To supplement the information provided through the questionnaires, the team members and the consultants conducted approximately 1,500 “desk audits”; these audits entailed interviews with employees, observation of their duties, and consultation with their supervisors. PX 621 at II — 2; Tr. 362-66; but see Tr. 305. After all such information was gathered, the County team and the Cresap personnel together sorted jobs into occupational categories. Then, different jobs within occupational categories were distinguished and demarcated by hierarchical job titles. PX 621 at II — 2; Tr. 307. Thus, for example, the Cresap process ultimately proposed the related job titles of Probation Officer Trainee, Probation Officer I, Probation Officer II, Probation Officer Supervisor I, Probation Officer Supervisor II, Probation Officer Supervisor III, Deputy Director of Probation, and Director of Probation. PX 621. The County team set out approximately 650 such job titles or “job classes.” Tr. 307.
Next, the County team was trained by Cresap to write job descriptions or job specifications for each of the titles so created. PX 621 at II — 3; Tr. 308 and 310; see PX 620 (instruction guide prepared by Cresap for Nassau County team, dated July 1965 and entitled: “Nassau County Class Specification Writers’ Manual”). These job specifications were intended to set forth the essential characteristics of each title. PX 621; PX 620. The format developed by Cresap for the job descriptions required the following information: (1) a general statement of the duties of the job title; (2) a statement of the complexity of the duties of the job; (3) a statement of the typical duties of the job; and (4) a statement of the qualifications — specifically, the knowledge, skills, abilities, and experience — necessary for the job title. PX 620 at 6-8. The job specifications were not intended to be exhaustive descriptions of each aspect of the job; rather, the specifications were to set out the salient features of the particular title and to distinguish that job from other positions. PX 620; Tr. 309 and 1245. The job descriptions so drafted by the County team were reviewed and re-written both by the team members and by the Cresap professional staff. Tr. 308, 316, and 1452.
After the job specifications were completed, all Nassau County employees were notified of the proposed job titles and job descriptions. PX 621 at II — 4; Tr. 366-67. At that time, neither salaries nor salary grades had been assigned to these job titles; nonetheless, the employees were permitted to appeal the classifications of their jobs to an independent classification appeals board. PX 621 at II — 4 to II — 5; Tr. 367-68. The decisions of this appellate board were binding on the County. Tr. 367-68. Over one thousand employees ultimately sought such review of the job re-classifications. PX 621 at II — 4.
The Cresap process also entailed the assignment of job titles to salary grades. Toward this end, the County team evaluated the final job specifications on the bases of four factors: (1) the knowledge and the skills required by the job; (2) the complexity and the variety of the duties of the job; (3) the responsibility for independent action; and (4) the responsibility for supervision. PX 619 at 4 (instruction guide prepared by Cresap for Nassau County team, undated and entitled: “Nassau County: Job Evaluation Handbook”); PX 621 at III— 2 to III — 3. Each member of the County team rated designated jobs on each of these four factors. Again, the results of this process were reviewed and revised by the entire team and by the consultants from Cresap. Tr. 331; Tr. 333.
*1376
The numerical values assigned to these four factors for each job title were then weighted and totaled. PX 621 at III — 3 to III— 4. Although the final report submitted by Cresap to the County does not identify any factor other than these four on which grades were assigned to job titles, it is clear that the County team and the Cresap personnel did consider other criteria in grading particular jobs. Tr. 319-25, 1462. For example, the ultimate point scores for each title were revised in part on the basis of information gathered by the individual team members during the desk audit phase of the classification process. Tr. 322-23. Ultimately, ranges of total scores were organized to correspond to 27 salary grades set out by Cresap. PX 621 at III — 4 to III— 5. Thus, for example, if the total evaluation points of a particular title fell within the range of 336 to 370 points, that job was given a salary grade of nine; if the total points of a particular title fell within the range of 371 to 405 points, that job was given a salary grade of ten. PX 621 at III— 4.
Cresap also conducted a labor market survey as an aid to the salary determination aspect of the job evaluation process. Cresap collected salary data on 63 separate “benchmark” job classes from such public employers as the City of New York, Suffolk County, Westchester County, the State of New York, and the United States Government as well as from employers in private industry. PX 621 at IV-1 to IV-4; see also DX LLL-1. This labor market data was used primarily to develop “salary trend lines” by which to compare the salaries of Nassau County jobs with the market rates for similar positions. PX 621 at IV- 5 to IV-6. Although the salary survey was conducted at the same time as the grading process, Tr. 1612, the final report submitted to the County by Cresap does not indicate that the labor market salary data were ever used by Cresap as a basis on which to adjust the proposed grade of any particular job title. See Tr. 1609-10. Rather, the data formed the basis for general comparisons of the salaries paid by Nassau County for job evaluation point ranges. Thus, for example, a salary trend line drafted by Cresap indicates that the salary for Nassau County jobs with an evaluation point total of 250 compares favorably to the salaries offered by Suffolk County and by Westchester County for similar jobs; but the report does not specifically compare the salary offered by Nassau County for the particular job title of “Laborer I” with the salaries paid by those other counties for the job of “laborer.” PX 621.
Nonetheless, it is clear from papers drafted by Cresap as early as 1964 that a primary objective of the reclassification and evaluation process undertaken by Cresap was to ensure that the salaries offered by the County would be competitive with those of nearby communities. PX 1104 at 2; see also Tr. 1448. It is also clear that Cresap regarded the results of its labor market survey as confirmation that the final proposed salary plan would guarantee the competitiveness of the County. Thus, Cresap concluded that “[a]t the upper range of positions, the minimum Nassau County salaries are as high as, or higher than, the salaries of all of the public jurisdictions to which the County is compared.” PX 621 at IV-6. Similarly, “[a]t the middle range of positions, Nassau County salaries are higher than all but the New York City salary scale.” PX 621 at IV-6. And, finally, “[a]t the lower range of positions, the Nassau County minimum salary is slightly higher than the other salaries shown.” PX 621 at IV-6. These facts indicated to Cresap that “Nassau County is in a strong competitive position, compared with other public jurisdictions and with private industry.” PX 621 at IV-6.
Cresap made its final report of proposed job title classifications and of proposed salary grades to the County in June of 1966. PX 621. The proposed grades for approximately 38 job titles were not included in the body of the report but were placed in an addendum dated June 15, 1966. These job titles had not been initially proposed by Cresap, but they were created through the classification appellate process. Accordingly, Cresap assigned salary grades to these new titles and reported them to the
*1377
County. PX 621 at addendum. As with the initial determination of new job classifications and job specifications, County employees were provided with an opportunity to seek review of their proposed salary grades before an independent salary review board. Tr. 368-69. Again, the determinations of this appellate board were binding on the County. Tr. 368-69. Over 1,000 employees sought such review. The final job titles and salary grades were enacted by Nassau County ordinance in May of 1967. PX 591; Tr. 369.
C.
Modifications since the Cresap Process .
The Cresap system has continued to be the primary framework for the classification and for the evaluation of new and existing jobs in Nassau County. At the time of this trial, the County had not replaced the basic Cresap structure, and the County has retained most of the original job classifications and salary grades adopted in 1967. Tr. 397-98. Three-quarters of present Nassau County employees work in job titles that were established by the Cresap process, and the pay grades of three-quarters of those employees have not changed since 1967. Tr. 629-30. The Executive Director of the Civil Service Commission from 1966 through the time of this trial, Adele Leonard, indicated that the Cresap process was the “seed” for the entirety of the present Nassau County job classification and compensation system. PX 1096 at 56. Indeed, there have been only three principal mechanisms by which the results of the Cresap process have been altered: (1) the creation of new job titles through the Civil Service Commission; (2) upgrades of existing job titles through the Civil Service Commission; and (3) salary increases and upgrades through the collective bargaining process.
As to the first of these, when a Nassau County department requests the creation of a new classified position, the Civil Service Commission first reviews the request to ensure that the proposed title is not duplicative of an existing job. Tr. 1241-43. If tentative approval of the proposal is forthcoming, the Civil Service Commission establishes the new title and drafts, a job specification; such specifications are intended to correspond in form and in function to those specifications originally drafted during the Cresap process. Information for the specifications is gathered from the requesting department, from other departments, and, at times, through a job audit that is similar to the desk audits of the Cresap process. Tr. 1243-45. The Civil Service Commission must formally adopt the new title specification. Tr. 1243-45. The Civil Service Commission recommends the new title to the County Executive; the County Executive then refers the title to the Nassau County Board of Supervisors. Final approval for the creation of a new title is granted by the Board of Supervisors through enactment as ordinance. Tr. 1259.
However, Civil Service Commission personnel specialists who began working for the County after the Cresap process have not received the training in job evaluation and grading that was given to the Nassau County team at the time of Cresap; accordingly, the specifications of new titles are not assessed on the four-factor system used by Cresap and by the Nassau County team in the 1960s. Tr. 388-89. Rather, salary grades for new titles are fixed by comparing the job specifications to those of existing County jobs. Tr. 389-90, 393, and 1282-84; see also PX 1096 at 10-13. In order to facilitate this assessment and comparison process, the Civil Service Commission personnel specialists have developed informal “rules of thumb”; thus, for example, a new position that requires a baccalaureate degree is typically — although not invariably — assigned a salary grade of ten. Tr. 391-92. These “rules of thumb” were developed by inferences from salary grades that were originally assigned through the Cresap process. Tr. 393. In evaluating new positions, the Civil Service Commission personnel specialists consider training and experience to be among the most significant factors. Tr. 1282; but see PX 1096 at 11. However, other criteria considered by the personnel specialists include recruitment difficulties, the working conditions for the new position, and the market rate
*1378
for similar jobs. PX 1096 at 13-15, 57; Tr. 1246-1247. Again, the Board of Supervisors must enact by ordinance a proposed salary grade for a new job title. Tr. 402 and 1259.
As to formal upgrades through the Civil Service Commission, either an employee or the department in which that employee works may seek an upgrade of the employee’s job title. DX UUU-1; Tr. 1259. Often, upgrades are sought because the duties of a particular title have changed over the years or because an employee is performing “out of title work” — that is, work beyond the scope contemplated by the specification for that title. Tr. 1261. The upgrade request — whether made by employee or by department — is reviewed by the Civil Service Commission; additional information concerning the job title may be sought by Civil Service Commission personnel specialists from the department heads. Tr. 1263 and 1285. Not infrequently, Nassau County departments act as the primary advocates of their employees’ desire for upgrades. Tr. 1264 and 1271.
The Civil Service Commission considers two principal factors in assessing upgrade requests: (1) significant changes in the duties of a job title; and (2) significant problems in recruitment or retention of employees in a job title. Tr. 403 and 1264-65. However, the Civil Service Commission also considers such factors as the working conditions of a title as criteria for upgrades. PX 1096 at 9. With respect to the retention-and-recruitment factor, the personnel specialists at the Civil Service Commission regularly undertake market surveys in order to determine whether salaries offered by the County are competitive with those of nearby communities. PX 1096 at 13-15; DX ZZZ-4 at 27. In some instances, the Civil Service Commission will seek to address recruitment or retention problems through means less extensive than an upgrade. For instance, the Civil Service Commission may authorize a department to hire new employees at a higher salary step than that at which new employees typically begin (“above-step hiring”), Tr. 1268 and PX 1096 at 60; or the Civil Service Commission may authorize mass promotions of employees in one title to a higher related title. Tr. 1271.
Finally, the Cresap system set in place in 1967 has also been modified through the vehicle of collective bargaining. Since 1969, the County and CSEA have entered into nine separate collective bargaining agreements. Tr. 1319. These agreements have at times provided for “across-the-board” wage increases; at times they have provided for higher wages for particular groups of employees. Tr. 1327-28; DX ZZZ-2 at 11. The County and CSEA have also entered into memoranda of understanding that secured upgrades for particular job titles. Tr. 1318 and 1320-23. Throughout these negotiations, both the County and CSEA have relied extensively on labor market surveys. Tr. 1314-17, 1377, and 1389. Both AFSCME and the County conduct comprehensive labor market surveys in order to present persuasive arguments at the bargaining table that particular salary increases or upgrades should or should not be granted. Tr. 1377-79.
II. Evidence of Discrimination by Nassau County
With these findings as background, the court may now turn to its findings of fact on the ultimate issue in this case: whether or not the County of Nassau has intentionally discriminated against women in female-dominated jobs with respect to compensation. Although the plaintiffs pursued several theories at trial, they ultimately determined that their “primary claim in this case is that the defendants engaged in intentional discrimination by systematically giving to predominantly female job titles lower salary grades than they would have received if they were not female-dominated.” Plaintiffs’ Post-Trial Brief at 1. That is, the plaintiffs argue that, during the Cresap job evaluation process, the “County selectively departed from [its wage-setting] methodology in a deliberate effort to favor male-dominated job titles while disfavoring the most significant female-dominated titles.”
Id.
at 3. Further, the plaintiffs argue that, since the Cresap process, this intentional discrimination in
*1379
setting salary grades has been both perpetuated by the failure of the County to redress these selective discriminatory departures as well as repeated by the County in its classification and grading of new titles. In addition to the evidence regarding discrimination in the setting of salary grades, the plaintiffs also sought to show that there are sex-related disparities between the current salaries of male-dominated job titles and of female-dominated job titles— and that those disparities are the product of intentional discrimination. Finally, other evidence presented on the question of discriminatory intent concerned such matters as sex segregation of County job titles and specific instances of discriminatory behavior. The defendants presented evidence on all these matters as well as evidence concerning Nassau County affirmative action plans and the collective bargaining process between the County and its employees. Thus, on the critical question of discriminatory intent, the court makes the following findings of fact.
A.
Discrimination during the Cresap Process
The plaintiffs presented no direct evidence of intentional discrimination against women in female-dominated Nassau County jobs during the Cresap process. Rather, every witness who had participated in that job classification and evaluation project credibly denied that any decision or determination was made during the Cresap process on the basis of the sex of Nassau County employees. See, e.g., Tr. 348, 377, 435, 1455, and 1464. Every witness associated with the Cresap process denied that they even had data as to the number of women in any particular job title. See, e.g., Tr. 377, 1455, and 1610. Confronted with such testimony as to the absence of any intentional discrimination by the Nassau County personnel team or by the Cresap personnel, the plaintiffs sought to prove their allegations indirectly.
1.
The “Graduate Student” Study
The first element of evidence presented by the plaintiffs as to discrimination by the County in the conduct of the Cresap process was a study conducted by one of the plaintiffs’ expert witnesses, Dr. Donald J. Treiman. Dr. Treiman, a sociologist from the University of California at Los Angeles, undertook “to determine whether Nassau County had applied its own job evaluation system in a consistent way.” PX 1074 at 14 (report of Dr. Treiman to plaintiffs, dated September 1989 and entitled: “Job Classification and Salary Setting in Nassau County, N.Y. 1966-1986”); see also Tr. 583. His objective was to simulate the second half of the Cresap process with unbiased evaluators of job specifications.
Dr. Treiman hired five graduate students to rate Nassau County job specifications on the bases of the four factors used during the Cresap process, Tr. 589; through this exercise, Dr. Treiman sought to replicate the Nassau County job evaluation process. Because his aim was to achieve “unbiased” ratings, he gave the graduate students no information as to the Nassau County personnel system, no information as to the sex composition of the jobs that they were to score, and no information as to the existence and nature of this action. PX 1074 at 14; see also Tr. 590. Dr. Treiman reported that the results of this rating exercise by the graduate students were “unambiguous.” PX 1074 at 16. More specifically, Dr. Treiman found that:
When ratings of 297 Nassau County jobs evaluated by three raters working independently are summed and averaged, and the resulting point scores plus whether the job is female-dominated are used to predict the salary grade assigned by Nassau County, it emerges that employees in female-dominated jobs are, on average 2.4 salary grades lower than employees in other jobs
with the same evaluated worth according to the criteria used in Nassau County.
PX 1074 at 16 (emphasis in original); see also Tr. 584-85 and 591. In other words, Dr. Treiman found that the Cresap process had created a two-and-one-half grade disparity between female-dominated jobs and male-dominated jobs that were otherwise equivalent under the evaluation criteria used in the Cresap process. Dr. Treiman testified that he checked these results for
*1380
consistency among his students, Tr. 591, and he testified that he regarded this study as a “neutral measurement” against which to assess the results of the Cresap job evaluation process. Ultimately, Dr. Treiman concluded that this graduate student exercise demonstrates that “C[resap] and Nassau County applied the C[resap] criteria in a way that systematically discriminated against jobs performed mainly by women” and that “the job evaluation scores [used by Cresap and by the County during the Cresap process] were systematically manipulated to achieve a pay hierarchy that corresponded closely to the preC[resap] hierarchy.” PX 1074 at 16. Dr. Treiman concluded that the Nassau County team and the Cresap personnel had not applied the job evaluative criteria in a “good faith” manner. PX 1074 at 4.
However, the court is unable to credit the testimony and the report of Dr. Treiman as to the conclusions of this study. First, the court found Dr. Treiman himself to be extremely evasive and not entirely credible throughout much of his testimony. See, e.g., Tr. 643-46 and 787-89. Second, Dr. Treiman in a subsequent report recanted part of his conclusions from this “unambiguous” evidence, see PX 1075 at 1. Further, the court found his methodology in the conduct of this graduate student exercise to be so flawed as to render it unreliable and of no probative value in this case.
Dr. Treiman hired five graduate students about whom he knew virtually nothing to perform this exercise. Tr. 758. After he discharged two of the students from the exercise (because he determined that they were not capable of satisfactory execution of the project), Dr. Treiman was left with only three raters for the study — two of whom were in the first semester of their first year of graduate studies. Tr. 758. Dr. Treiman gave the graduate students minimal guidance in the exercise. Tr. 771-72.
More significantly, although Dr. Treiman instructed the students to rate each of 325 job specifications on the four factors used in the Cresap process, he did not give the students any information as to the number of employees supervised by the incumbents of these job titles. Tr. 768-69. Dr. Treiman told these students to do the “best that they could” without that information — information that Dr. Treiman himself conceded was essential to scoring the “supervisory responsibility” factor. Tr. 769. Thus, Dr. Treiman’s graduate students— his standard assessors against which the results of the Cresap process could be measured — were directed to improvise at least one-fourth of the data they needed for the project.
Next, Dr. Treiman instructed his students to spend no more than 50 to 60 hours on the evaluations. Tr. 761. The court cannot conceive that 50 to 60 hours devoted to this project could adequately replicate the efforts of the participants in the Cresap process who spent months evaluating these jobs. See, e.g., Tr. 333. Indeed, it became clear at trial that the time allotted by Dr. Treiman was terribly inadequate for any but the most hurried completion of this project. Tr. 762-67. Finally, in contrast to the procedure used in the Cresap process through which differences in point ratings among members of the County team were discussed and resolved by consensus, Dr. Treiman simply “averaged out” the different point ratings assigned by his graduate students. Tr. 781. Taken together, these procedural deficiencies significantly undermine the value of the graduate student exercise as a simulation of the original Cresap process.
The results of this exercise confirm that — far from tracking the Cresap job evaluation process — the exercise was a study in poor guidance, incomplete information, and inadequate time. The graduate students’ ratings were remarkably inconsistent: For example, the graduate students assigned to the job title of Housekeeping Supervisor grades that ranged from 7 to 20. DX KKK-2. Also, two of his three students gave grades that differed on average by three grades — a disparity greater than the two-and-one-half grade difference that Dr. Treiman identified between male-dominated and female-dominated jobs in the Cresap job process. See, e.g., Tr. 1561. Again, rather than
*1381
investigate these disparities among his students, he simply “averaged out” their results. Indeed, ultimately, Dr. Treiman’s graduate students rated the male-dominated job title of Custodial Worker I as equal to the female-dominated job title of Clerk-Stenographer I — a result for which Dr. Treiman himself argued there was no “rational basis.” Tr. 787-89. For all these reasons — because of Dr. Treiman’s problematic methodology and because of his less than persuasive testimony — the court must disregard entirely the results of this graduate student exercise. It is wholly without probative value, and it does not provide any basis for inferring discriminatory intent in the execution of the Cresap classification and evaluation process.
2..
The “Training and Experience” Study
The second item of evidence introduced by the plaintiffs on the question of intentional discrimination in the Cresap process was another study conducted by Dr. Treiman. In this study, Dr. Treiman sought to eliminate the subjectivity inherent in the grading exercise performed by the graduate students. Tr. 596; but see Tr. 688-89. To this end, he analyzed the job specifications prepared during the Cresap process for two “objectively measurable” factors: first, the amount of training and experience required by a particular position; second, whether or not the position entailed supervisory responsibility. Dr. Treiman testified that the first of these two factors constitutes a suitable proxy for the first three factors used by Cresap to evaluate jobs (that is, knowledge and skills, complexity of duties, and independence of action). Tr. 597-98. Thus, in conjunction with a measure for supervisory responsibility, Dr. Treiman believed he had two quantifiable variables that would permit objective evaluation of the Cresap job-evaluation process.
Dr. Treiman then analyzed Nassau County job specifications for these factors. He used data from 1967 as to the number and the sex of Nassau County employees, Tr. 598, and he weighted job titles by the number of employees in each title. PX 1074 at 10. As to training and experience alone, Dr. Treiman found that, “[o]n average ... employees in female-dominated jobs are 1.7 salary grades lower than employees in male-dominated jobs requiring the same training/experienee.” PX 1074 at 11; Tr. 598-99. When Dr. Treiman introduced the variable for supervisory responsibility into his study, he found that, on average, as to any given number of years for training and experience, nonfemale-dominated jobs without supervisory responsibility were better compensated than female-dominated jobs with supervisory responsibility. Tr. 606-07; PX 1074 at 14. From these comparisons, Dr. Treiman concluded that Nassau County could not have applied the Cresap job-evaluation procedures in a good faith manner. Tr. 607.
As with his graduate student exercise, however, this court again finds that Dr. Treiman’s conclusions are not reliable. First, this “objective” analysis of Nassau County job specifications rests on the critical subjective judgment of Dr. Treiman that the first variable he used — the training and experience variable — was an adequate proxy for the first three Cresap factors. Dr. Treiman testified that he was confident as to the adequacy of the training and experience variable because it is “well known” to be the most important aspect in determining compensation for job titles. Tr. 596. This rationale completely ignores the fact that the two variables used by Dr. Treiman plainly account for only
half
the job evaluation factors used by Cresap; it also indicates that his study is only half complete. Along these lines, Dr. Treiman further testified that the first three Cresap factors are “very highly correlated” and are “alternative measurements of the same thing.” Tr. 597-98; see PX 1074 at 33 n. 9. He ultimately conceded, however, that this determination was entirely subjective. Tr. 695, 696.
Quite apart from the question as to the bases on which Dr. Treiman’s confident assertions may rest as a general matter, it is clear that the particular “verification” of this assumption as to the'adequacy of the two variables used is untrustworthy. Dr. Treiman’s report indicates that the data on which he confirmed the high correlations
*1382
among the first three Cresap factors and the adequacy of his proxy variable were the data produced by his graduate students in their ratings of Nassau County job specifications. , PX 1074 at 33 n. 9 and at Appendix I (entitled
“An
Unbiased Application of the CMP Job Evaluation Criteria to Nassau County Jobs”); Tr. 695. That is, in order to confirm his critical assumptions about these variables, Dr. Treiman calculated correlations among the point scores assigned by his graduate students. The problems with this procedure are as patent as the flaws in the graduate student exercise are serious: The unreliable results generated by the graduate student evaluations are simply imported into the second study — a study intended in part to confirm the first study. Tr. 612-13. The use of the graduate-student point scores as the underpinning for the training and experience study demonstrates the untrustworthiness of that second project.
Additionally, the fact that Dr. Treiman collapsed the first three Cresap factors into one variable resulted in his ignoring entire sections of the Nassau County job specifications. Tr. 749. And Dr. Treiman also weighted job titles by the number of employees — even though the participants in the Cresap process had not done so. Tr. 742-43. The reasons stated by Dr. Treiman for this methodological move of weighting titles (“ultimately we are interested in what happens to people and not to jobs”, PX 1074 at 34 n. 11) is plainly inconsistent as well with the allegations of the plaintiffs’ complaint. See Amended Complaint ¶ 3(D) (alleging discrimination on the basis of sex domination of jobs). Dr. Treiman conceded that this aspect of his study could be expected to yield results different from those of the Cresap process. Tr. 743. The court is deeply troubled that Dr. Treiman so readily and so regularly departed from the stated methodology of the Cresap process in order to determine whether or not the Cresap personnel and the Nassau County team had themselves departed from that methodology. The court simply cannot give any weight to this terribly flawed study.
The court is also unable to find that Dr. Treiman’s conclusions as to the sex composition of job titles in 1967 are entirely reliable. Dr. Treiman himself had no data as to the sex composition of these job titles; rather, paralegals at the office of the plaintiffs’ counsel reviewed the names of many of the incumbents of these job titles during the year 1967, and those paralegals drew inferences as to the sex of each employee from those names. Tr. 685; PX 1074 at 32 n. 6. Indeed, the court notes in this regard that
every
data base used by Dr. Treiman in each of the studies he prepared for the plaintiffs was created by the plaintiffs’ attorneys. Tr. 680-85. As to this data base, although the drawing of inferences as to sex from an individual’s name is not very significant in itself, these paralegals appear to have proceeded on the assumption that only 6,000 people were employed by Nassau County in 1967, see PX 1075 at Table 1. But, at that time, there were between 8,300 and 10,000 people employed by Nassau County. See PX 621 at II — 1 and V-2; compare with Tr. 714 (Dr. Treiman testifying that the number of Nassau County employees in 1967 “was somewhere between five and ten thousand”). There is no indication in the record as to how the subset of 6,000 employees was derived by the plaintiffs — and there is no indication that those 6,000 employees are representative of the 1967 County work force as a whole. Indeed, the plaintiffs have provided precious little information as to how the sex domination of Nassau County jobs in 1967 was determined, Tr. 716-17, and that determination cannot now be adequately reviewed either by the defendants or by this court. Accordingly, to the extent that the conclusions as to the male or the female dominance of Nassau County job titles in 1967 is predicated on inferences made by agents of the plaintiffs from less than all the relevant data — and to the extent this court has been given no basis on which to find that the method for selection of those 6,000 names was a trustworthy one — the court finds that the reliability of Dr. Treiman’s assumptions as to sex dominance of job titles in 1967 is somewhat limited. Ultimately, however, the reserva
*1383
tions that this court may have as to these data are simply collateral concerns to the deep methodological flaws indicated above. It is on those grounds — coupled with the evasive and unpersuasive testimony of Dr. Treiman — that the court finds the “training and experience” study by Dr. Treiman to be of no weight.
The court must note as well, however, that the problematic aspects of Dr. Treiman’s evidence are not found only in his written reports. His testimony before this court was consistently evasive — particularly when he was asked about latent (or patent) flaws in his studies. Thus, it often required extensive questioning either by counsel or by the court to secure from Dr. Treiman a candid response about whether a particular methodological error on his part affected the results of his study. For example, after Dr. Treiman acknowledged that he had erroneously coded a job requirement of “elementary education” as six years rather than as eight years because he was drawing on his experience as “a boy in California” rather than on the actual school system in Nassau County, Tr. 700-09, Dr. Treiman was initially unwilling to admit that his mistake had skewed other aspects of his training and experience study:
THE COURT: Just let me see if I understand this for a minute. If your assumption is wrong with respect to elementary school being six when it should be eight, then the ten would be wrong, too, wouldn’t it, or would it? How did you get to ten? You built in two more years of junior high and added on top of that two more years.
THE WITNESS: I am sorry, ten on page—
THE COURT: I am looking at footnote 33. You made no provision for what your experience was in California of junior high. Is that it?
THE WITNESS: Sir, that was what I was working from was the job specifications and the job specifications in Nassau County refer to elementary education. They refer to some high school. They refer to high school graduation. I made an error. It's now, obviously an error in coding elementary education as six instead of eight; but it would have no impact on the rest of the codes.
THE COURT: Okay. Well, if a person had eight years of elementary school education and two years of experience, where would you code that?
THE WITNESS: It’s coded by the job. The job says elementary education and if it says elementary education plus two years of experience, I would have coded it as an eight on page ten.
THE COURT: Suppose that a person had an elementary school education and four years of experience, where would you code that?
THE WITNESS: That would be coded as ten.
THE COURT: It should be 12, right, assuming that an elementary school education should be an eight. So, it might [ajffect some other numbers. Is that right?
THE WITNESS: Yes, that’s correct.
Tr. 709-10. A colloquy such as this reveals either that Dr. Treiman is unable to appreciate and to assess the effects of the methodological defects in his study or that he is unwilling to testify candidly about those effects. Neither possibility secures the confidence of this court in his evidence.
For their part, the defendants attempted to demonstrate through one of their expert witnesses, Dr. David P. Jones, that Dr. Treiman’s study would not have indicated a sex-related disparity in 1967 salary grades had he accounted for additional factors in the Cresap process. DX GGG-1 (report of Drs. Jones and Haworth to defendants, dated November of 1989 and entitled: “Summary of Statistical Employment and Compensation Patterns by Gender in the Nassau County Government: 1980 and 1986”). Specifically, Dr. Jones undertook to demonstrate that the salary grades assigned by Cresap were affected by the local labor market and by the number of hours worked each week for particular job titles; Dr. Jones concluded that, once these two factors are taken into consideration, the sex-
*1384
related disparity in job grades is statistically insignificant.
Dr. Jones repeated Dr. Treiman’s “training and experience” study with several modifications. First, in order to obtain statistically significant examples, Dr. Treiman restricted his own study to those Nassau County job titles that had 15 or more incumbents. Tr. 1583-84. That is, Dr. Jones wanted to eliminate the possibility that the sex of only two or three employees would radically alter the sex-dominance of a particular title. This restriction left Dr. Jones with 101 Nassau County job titles for his study. Tr. 1583-84. Next, Dr. Jones rated these 101 titles not only on the two factors used by Dr. Treiman but also on three factors derived from the Dictionary of Occupational Titles. Tr. 1581-82. Notably, Dr. Treiman was the staff director of the committee that developed the Dictionary of Occupational Titles. Tr. 581. These factors were used to provide Dr. Jones with a job description analysis that he expected to parallel that of Dr. Treiman. Tr. 1632-33. Dr. Jones then collected current labor market data for as many of these 101 titles as he could. Dr. Jones and his staff attempted to match Nassau County job specifications with jobs in the New York counties of Rockland, Suffolk, and Westchester as well as with jobs in New York City. Tr. 1593-96; DX GGG-4 and DX GGG-5. These procedures left Dr. Jones with 80 usable job titles. Tr. 1622-23. However, Dr. Jones used the same data as to the sex composition of Nassau County jobs in the year 1967 that Dr. Treiman had used. Tr. 1537.
Dr. Jones then conducted regression analyses without a variable for labor market data. Tr. 1625. He found that, on the bases of the variables from Dr. Treiman’s “training and experience” study, there was a significant correlation between salary and sex in Nassau County jobs. Tr. 1626; DX GGG-1 at 45. Similarly, Dr. Jones found similar results when he considered the job characteristic variables derived from the Dictionary of Occupational Titles. Tr. 1632-33; DX GGG-1 at 47. Thus, whether he used the variables derived by Dr. Treiman or the variables from the Dictionary of Occupational Titles, Dr. Jones was able to confirm the conclusion of Dr. Treiman that the sex composition of Nassau County jobs does have a high correlation to the salary grade midpoints of those positions. Tr. 1633-34. However, Dr. Jones found that when he introduced into these regression analyses an independent variable to represent the market data he had gathered, the sex composition of jobs was of virtually no statistical significance as to the original salary grade disparity. Tr. 1635-36, 1641-42; DX GGG-1 at 54; DX GGG-1 at 57. Dr. Jones achieved similar results upon the introduction of a variable to represent the number of hours worked by employees of different Nassau County job titles: Again, Dr. Jones found that a variable for the length of the work week reduced the effect of sex on salary (although not as much as did the variable for labor market data). Tr. 1647-49; DX GGG-1. Thus, Dr. Jones concluded and testified that Dr. Treiman’s “training and experience” study had omitted important variables and that, once those variables are considered, Dr. Treiman’s study would not suggest any relationship between sex composition and the setting of salary grades during the Cresap process. Tr. 1655; see also DX GGG-1 at 60.
Although the court found Dr. Jones’ study to be procedurally sound (with certain possible exceptions such as the use of the 1967 sex-composition data), the weight that the court is able to attribute to the conclusions of Dr. Jones is significantly limited by the paucity of corroborating evidence that the County of Nassau ever considered either the labor market or the length of the work week when it determined the initial Cresap grades. First, as to length of the work week, the County adduced absolutely no evidence that the participants in the Cresap process ever considered this factor — or even had data as to it — when they established salary grades. See, e.g., Tr. 326. Next, as to the labor market data, although the Cresap team did collect market salary information on 63 “benchmark” jobs, the final Cresap report does not indicate that the market data were
*1385
used to set or to modify the salary grades of particular job titles. Rather, the report indicates that the market data were used to draft broad salary “trend lines” for comparisons as to the ranges of salaries offered by Nassau County and by other communities. PX 621 at IV-4 to IV-6; Tr. 1605-06. Further, no participant in the Cresap process testified that market data were used to adjust salary grades. See, e.g., Tr. 327-29 and 1609-12. On the other hand, concern with the competitiveness of Nassau County salaries recurs in almost every segment of every report from Cresap to the County. It would not be an implausible inference that the consultants whose task it was to design a competitive salary plan might have amended salary grades for certain job titles in order to ensure such competitiveness. However, there was no evidence introduced to support this conclusion beyond the findings of the regression analyses of the defendants’ expert. Accordingly, the court is unable to conclude that the Cresap team actually considered market data in setting individual salary grades.
Because there is no satisfactory support for a finding that the variables introduced by Dr. Jones actually affected the assignment of salary grades to particular job titles during the Cresap process, Dr. Jones’ conclusions are only marginally relevant to the question of discriminatory intent. Nonetheless, the court will note that it specifically rejects two challenges posed by the plaintiffs to Dr. Jones’ methodology. First, the plaintiffs attacked the procedures followed by Dr. Jones and by his staff to collect labor market data, see, e.g., Tr. 1718-39; however, the court found these procedures to be both reasonable and reliable. Accordingly, the court credits the results of Dr. Jones’ labor market survey to the extent those results are relevant to issues in this action. Second, the plaintiffs attacked Dr. Jones’ restricted sample of 101 Nassau County job titles as unrepresentative of Nassau County job titles generally. Tr. 1748. However, the court found that the decisions made by Dr. Jones as to the bases on which job titles would be éliminated from his study were sound. The court also notes that Dr. Jones was able to obtain with that sample — and with other data used by Dr. Treiman — results that were similar to the results obtained by Dr. Treiman in his “training and experience” exercise. Tr. 1633-34. The plaintiffs thus stand in an awkward posture when they attack Dr. Jones’ procedures.
Again, without even considering the refinements suggested by Dr. Jones’ report, this court does not find that the “training and experience” analysis of Dr. Treiman provides evidence of discriminatory intent in the Cresap process. Rather, the questionable — and poorly verified — assumptions on which that study is based as well as the flawed methods with which it was executed erode completely what little confidence this court might have had in Dr. Treiman’s conclusions after observing his testimony. This court credits neither Dr. Treiman’s testimony nor the results of his “training and experience” analysis.
3.
Plaintiffs’ Exhibit 616/616-A
The next item of evidence introduced by the plaintiffs to demonstrate discrimination in the execution of the Cresap process was plaintiffs’ Exhibit 616-A. This exhibit is a color photocopy of a thirty-three page document retrieved from the files of the Nassau County Civil Service Commission. See PX 1082 (stipulation of counsel that Exhibit 616-A was found in a Nassau County file folder marked: “CMP — Point Evaluation to Determine Grades”); but see PX 1096 at 48-49 (exhibit retrieved from “junk” files). Exhibit 616-A contains job titles typewritten on the left side of each page; these job titles are grouped under general headings, and they are alphabetically arranged within each grouping. Next to these job titles are five columns that are headed (at least on page one of the exhibit): “K-S”, “C-V”, “IA”, “S”, and “T”. Numerical values that correspond to job titles are handwritten in the columns under these headings. However, many of these numerical values have been crossed out by hand in different colors of ink; some columns show several numbers crossed out, rewritten, and recrossed out. Most of the crossed out values fall within the “T” column. Further,
*1386
there are other job titles handwritten onto the document in red ink; many of these job titles do not have numerical values associated with them. Some of the typewritten job titles are crossed out, some are changed, and some have handwritten “check” marks next to them. Stray handwritten marks found on the document include question marks, dates, apparent deletions of entire sections, the letter “x”, words such as “ok”, “wrong”, and “out”, and illegible scribbling. Two of the pages have sections of yellow legal paper pasted or taped onto them; handwritten on those pieces of legal paper are job titles for which there are handwritten columns but no numerical values.
Vito Competiello — a former employee of the Nassau County Civil Service Commission who participated in the Cresap job evaluation process, Tr. 288 — testified that Exhibit 616 (a monochromatic photocopy of Exhibit 616-A, Tr. 838) is a draft for a master compilation of Nassau County job titles and of the job evaluation point scores for those titles. Tr. 334, 376. Accordingly, the notations at the tops of the columns on Exhibit 616-A denote the four Cresap evaluation factors: knowledge and skills (“KS”), complexity and variety of duties (“CV”), responsibility for independent action (“IA”), and supervisory responsibility (“S”). The “T” column is intended to represent the total of these point values for each job title. Competiello testified that he recognized some of the handwriting on Exhibit 616 as his own, but he could not identify many of the notations. He did not unequivocally indicate that he could identify when the information on Exhibit 616 was compiled or what use, if any, was made of it.
The plaintiffs have assumed that Exhibit 616-A contains the final results of the Nassau County job evaluation process — that is, that it lists “the point scores assigned to Nassau County jobs in 1966 by the team that applied the four-factor job evaluation methodology” of the Cresap process. Tr. 613; see also PX 1075 at 1 (supplemental report of Dr. Treiman to plaintiffs, dated November 1989 and entitled: “Job Evaluation Points and Ordinanced Salary Grades: Discriminatory Changes in Salary Grade Assignments, Nassau County 1966-67”). On this assumption, Dr. Treiman, on behalf of the plaintiffs, calculated the salary grades “that would have resulted from the[ ] point scores” on Exhibit 616, PX 1075 at 1; he then compared these grades to the grades that were proposed by Cresap in its final report to the County and to the grades that were eventually enacted by the County in 1967. The purpose of this exercise was “to ascertain whether changes in salary grades subsequent to the evaluation reported in Exhibit 616 were related to the gender composition of jobs.” PX 1075 at 1.
Again, Dr. Treiman found evidence of discrimination. He recanted the finding of his earlier report to the effect that “[b]ased on Exhibit 616, it now appears that the job evaluation procedure itself may have been carried out in good faith ... but that subsequent to the evaluation procedure the salary grades of a large number of Nassau County jobs were altered in a discriminatory manner.” PX 1975 at 2. That is:
Specifically, in the final report to Nassau County submitted to Nassau County by Cresap [Exhibit 621] the salary grade of the average employee in male-dominated jobs was increased by nearly two grades (1.9) while the salary grade of the average employee in female-dominated jobs was not increased at all____ Subsequent to the Cresap report, there was a minor reduction in this gender differential, so that the ordinanced salary grade for the average employee in a male-dominated job was 1.6 grades higher than the salary grade derived from the point scores in Exhibit 616, while the ordinanced salary grade for the average employee in a female-dominated job was one-tenth of a salary grade higher than the salary grade derived from Exhibit 616.
PX 1074 at 2.; Tr. 619-20. Thus, for example, after he converted the point scores on Exhibit 616 into salary grades, Dr. Treiman found that the mean salary grade for jobs that were 81% to 90% male in 1967 would have been 8.1 as scored on Exhibit 616. But as recommended in the final Cresap report, the mean salary grade for those
*1387
jobs was 9.5. On the whole, he found that the difference in mean salary grades for male-dominated jobs between the Cresap final report and Exhibit 616 was 1.9; the difference in mean salary grades between the Cresap final report and Exhibit 616 for female-dominated jobs was 0.0. PX 1075 at Table 1.
Dr. Treiman concluded that the existing differential in salary grades between male-dominated jobs and female-dominated jobs “is mainly attributable to changes in the salary grade of male dominated jobs between the [Exhibit] 616 point scores and the Cresap final report.” Tr. 620. Finally, Dr. Treiman determined that these changes in mean salary grades strongly supported an inference of sex discrimination:
Although I do not know who is responsible for these alterations, their intent and effect are completely clear — to increase the salary grades of the largest male-dominated jobs relative to what these jobs should have been paid on the basis of an application of Nassau County’s
own criteria for determining compensation.
PX 1075 at 2-3 (emphasis in original).
However, the court is unable to agree with the plaintiffs as to the nature of Exhibit 616-A. Most importantly, the court cannot find that Exhibit 616-A constitutes a master compilation of the final scores from the Cresap job evaluation process. Rather, Exhibit 616-A is clearly a working draft that was not a list of the ultimate scores from the job evaluation process but rather part of the job evaluation process itself. The lists in Exhibit 616-A are incomplete. For example, there are typewritten job titles with point values on Exhibit 616-A that do not appear at all in the final Cresap report. Compare, e.g., PX 616-A at 18 (“Hospital Administration Trainee”) with PX 621. Conversely, there are approximately one hundred job titles handwritten onto Exhibit 616-A that have no point values assigned at all. The plaintiffs argue that these were jobs added to the Nassau County system after the classification process, and Competiello initially confirmed this view. Tr. 361. However, a comparison between these titles and the addendum to the final Cresap report that lists the titles created through the appellate process demonstrates that the plaintiffs’ argument is groundless — and that Competiello’s recollection was inaccurate. Tr. 376.
Moreover, perhaps one-quarter to one-third of all the numerical values in the “T” column are crossed out; a small number of those have new values written in either above or next to the old values. Almost invariably, where new numbers are recorded in the “T” column, at least one of the values in the other four columns has been changed. This fact strongly suggests that Exhibit 616-A records many of the moments of review and revision in the job evaluation process. Similarly, numerous notations are scattered throughout Exhibit 616-A; entire groups of job titles are crossed out or are inserted; question marks, “stars”, and “cheek” marks all indicate that Exhibit 616-A represents anything but the end of the job evaluation process.
The plaintiffs have attempted to make much of the fact that Competiello identified as his handwriting a notation on page 15 of Exhibit 616 that reads “6/8/66.” Tr. 336. The plaintiffs argue that this notation— dated in the same month that Cresap issued its final report to Nassau County— demonstrates that Exhibit 616-A was a master list that recorded all the point scores for the Nassau County titles from which Nassau County selectively departed in the assignment of salary grades. But that conclusion is not supported by the exhibit when considered as a whole: As indicated, there are over one hundred job titles listed on Exhibit 616-A for which no point scores at all have been recorded. Thus, if the fact that Competiello wrote on the document a date contemporaneous to the issuance of the final Cresap report demonstrates anything of significance, it shows that Exhibit 616-A could not have been a catalogue of the “final results” of the job evaluation process. Similarly, other dates recorded on Exhibit 616-A are inconsistent with the plaintiffs’ understanding of that document. For example, another no
*1388
tation on page 15 of the exhibit records a date in September of 1965. Tr. 336. But Competiello testified that the job evaluation process did not even begin until early 1966. Tr. 308, 317. Such inconsistencies between Exhibit 616-A and the testimony of Competiello about the Cresap process further obscure the meaning and the significance of that document.
Moreover, the court notes that Competiello could not recall clearly the nature and the purpose of Exhibit 616 — a document that he had not reviewed in the 24 years between the Cresap process and this litigation. Competiello was unable to identify the document with confidence as anything more than a draft, and he did not satisfy the court that the numerical values recorded on Exhibit 616 were the “final results” of the job evaluation process. Tr. 333-34, 343; compare Tr. 356 with Tr. 376. Curiously in this regard, the plaintiffs never asked Competiello to identify Exhibit 616— A — the color copy of the original document which clearly revealed different markings in different colors of ink. Rather, they asked him only to identify Exhibit 616, a monochromatic photocopy of that document. The reluctance of the plaintiffs to provide Competiello with an exhibit which more closely approximates the original document about which he was to testify is troubling to this court, and it underscores the degree to which Competiello was easily led by the questions asked of him as to the nature and the ■ content of Exhibit 616. Compare Tr. 356 with Tr. 376. Quite frankly, the court found that Competiello improvised many of his “recollections” about Exhibit 616. Compare with Tr. at 305 (Competiello inaccurately recalling number of desk audits conducted during Cresap process). For this reason, the court does not credit Competiello's testimony as to the nature of Exhibit 616.
More credibly, Competiello testified that the job evaluation process — through which points were assigned to job titles — was a protracted undertaking. Tr. 333. He also testified that many of the early “results” were revised on the basis of information gathered during the desk audit phase of the classification process; for example, a member of the Cresap team who thought that a particular job had been inadequately scored might suggest aspects of the job that would merit a change in that score. Tr. 333. The markings of Exhibit 616-A confirm this testimony. In any event, Competiello testified that, after initial scores were assigned to jobs by individual members of the team, all four members reviewed and discussed the preliminary scores in order to make informed and consistent modifications. Tr. 333. Exhibit 616-A plainly captures one — or perhaps many — of the moments in that process; but the court has no satisfactory basis on which to conclude that 616-A is the
last
moment — or even a significant moment of that process. Indeed, the fact that Exhibit 616-A was found in the files of Nassau County strongly suggests that it was not ever transmitted to Cresap for incorporation into the final report. Interestingly, the plaintiffs did not present the exhibit to those witnesses who had worked for the Cresap firm during the job evaluation process. The court simply cannot conclude that Exhibit 616-A is in any manner a reliable indication of what the final job evaluation scores were. In other words, the probative value of this exhibit is limited to confirming that the Nassau County team did in fact score job titles, did in fact total those point scores, and did in fact make numerous revisions to the points throughout the evaluation process. But those facts were never in dispute in this action.
Even if this court could have agreed with the plaintiffs that Exhibit 616-A lists the final results of the job evaluation process, the court would nonetheless be compelled once again to disregard the conclusions drawn from that document by Dr. Treiman. As indicated above, Dr. Treiman undertook an analysis of the salary grades that would have been generated by the point scores on Exhibit 616-A (again, on the assumption that the exhibit lists the final results of the job evaluation process), and he compared those salary grades to the salary grades ultimately proposed by Cresap in its final report to the County. He found that the average salary grade for male-dominated
*1389
jobs in the final Cresap report was 1.9 grades higher than the average salary-grade for male-dominated jobs that would have been the result of the point score totals on Exhibit 616-A. Dr. Treiman also found that the average salary grade for female-dominated jobs in the final Cresap report was the same as the average salary grade for female-dominated jobs that would have been compelled by the point score totals on Exhibit 616-A.
But Dr. Treiman’s study is unreliable. First (and of less importance), the basis upon which Dr. Treiman determined whether a particular job title was male-dominated or female-dominated in 1967 — the time of the Cresap process and, hence, the time relevant to studying “selective departures” in that process — is weak. As indicated above with respect to his “training and experience” study, Dr. Treiman’s data as to the sex composition of Nassau County jobs in 1967 are somewhat suspect, and the plaintiffs have provided no basis on which to evaluate the trustworthiness of the process by which those data were derived. Again, however, the problems with these data are most telling insofar as they simply reveal the casual procedures that characterize Dr. Treiman’s work as a whole. Similarly, the data base used by Dr. Treiman as to the point scores on Exhibit 616 was prepared for him entirely by the plaintiffs’ attorneys. However, he testified that he himself is unable to verify the reliability of those data:
Q: Well, did you ever make any inquiry to find out how this information had been assembled for you?
A: I am sure that we discussed it. I do not recall the discussion.
Q: Wouldn’t you agree with me, Doctor Treiman, that the accuracy of the data that you’re analyzing is something that’s very important to an expert like yourself?
A: Yes.
Q: And, in fact, if the data isn’t right, the analysis surely isn’t right. Isn’t that correct?
A: Yes.
Q: It’s the old story of bad in and bad out, right?
A: Yes.
Tr. 716-17. That Dr. Treiman would adopt as his own data which he is unable now to validate illustrates the cavalier — or, worse, partisan — character of his endeavors in this action. The court is unable to credit completely the reliability of quantitative evaluations made by the plaintiffs’ attorneys’ paralegals and accepted at their face value by an expert — particularly to the extent that the expert acknowledges that his conclusions are only as reliable as his data.
Further, the assumptions made by Dr. Treiman — or, more accurately, by the plaintiffs’ attorneys who compiled the data used by Dr. Treiman in all his studies — as to the point totals on Exhibit 616 are dubious at best and inexcusably deficient at worst. Dr. Treiman grounded his analysis on the point totals as they are found on Exhibit 616; but he admitted that, wherever Exhibit 616 contains a job point total that is crossed out but that does not have a new value written in, Dr. Treiman used the crossed out value as the point total. Thus, for example, Dr. Treiman determined that the salary grade for the title of Cashier II would have been a grade of six on the basis of Exhibit 616; but a review of either Exhibit 616 or of Exhibit 616-A reveals that the only point total written in for Cashier II has been crossed out. Dr. Treiman nonetheless assumed that this crossed-out value represented the “final result” of the job evaluation process for the title of Cashier II. Compare PX 616-A at 3 with PX 621 at III — 4 and with PX 1075 at Table 2. He made similar assumptions as to at least
one hundred
more job titles for which numbers in the “T” column had been crossed out.
As though this aspect of his methodology were not sufficiently questionable, Dr. Treiman did not even apply it consistently. In those instances for which a value had been crossed out in the “T” column and for which a new value had been recorded by hand, Dr. Treiman used the second value— not the first — as the basis for his analysis. PX 1075 at 2 n. 1; Tr. 799. Thus, rather
*1390
than proceed uniformly on the assumption that the initial values were the “final results” of the Cresap job evaluation process, Dr. Treiman determined that a later recorded value was the “final result” of the job evaluation process. This inconsistency obviously takes no account of the very compelling inference that, as with the job titles for which new point totals had been recorded, new values were contemplated and intended for the job titles as to which the initial point totals had been crossed out. Thus, for Dr. Treiman, a point total was “final” if it had been simply crossed out; if a point total had been crossed out and revised, the newer total was the “final result.” The court simply cannot agree that any aspect of Exhibit 616-A warrants Dr. Treiman’s casual methodological distinctions. Indeed, Dr. Treiman did not even undertake to determine whether these revisions on the face of Exhibit 616 would support an inference of discriminatory modifications — even though he conceded that, given the purpose of his study, it would have been a “reasonable thing to do.” Tr. 847-48.
To the extent that Dr. Treiman’s analysis relies on hundreds of numerical values of uncertain significance or finality — and to the extent that Dr. Treiman treated these values inconsistently — the court cannot accept his findings as reliable analyses of Exhibit 616-A. Pages and pages of numerical values on Exhibit 616-A have been stricken, and yet Dr. Treiman — with no apparent justification — treated those values as “final results.” Moreover, as noted above, many of the job titles on Exhibit 616-A have no point totals indicated at all; Dr. Treiman, of course, had to omit those titles from his study. For all these reasons — and, again, because his testimony on these points was simply unpersuasive — the court finds Dr. Treiman’s analysis of this document to be of no probative value: It provides no basis for a finding of discriminatory intent in the execution of the Cresap process.
One of the experts who testified on behalf of the defendants, Dr. Jones, indicated that his own analysis of Exhibit 616-A was inconsistent with any inference of discriminatory intent in the Cresap evaluation process. Dr. Jones first examined the changes that appear on the face of Exhibit 616-A to determine whether or not they bore any statistical relationship to the sex composition of Nassau County jobs; he found that they did not. Tr. 1541-44.
Next, Dr. Jones undertook to determine whether or not the results found by Dr. Treiman — that the differences in mean salary grades between Exhibit 616-A and the final Cresap report suggested discrimination against women — would still hold after the introduction of variables other than the sex composition of jobs. More specifically, Dr. Jones evaluated Dr. Treiman’s results with the introduction of one variable to represent the market salary data collected by Cresap and of a second variable to represent the pre-Cresap salaries of Nassau County jobs. Tr. 1547-48. Dr. Jones found that, after a regression analysis with these variables, the market-data variable and the existing-salary variable helped to explain the salary grade differential between Exhibit 616-A and the final Cresap report. Tr. 1550. He also found that the sex composition of these job titles was no longer statistically significant after the introduction of these new variables. Tr. 1550. In other words, Dr. Jones reported that, on a statistical analysis, the differences in mean salary grade that Dr. Treiman identified are better explained by the market information collected by Cresap and by the pre-Cresap salaries of Nassau County jobs than by the sex composition of those titles.
A similar regression analysis conducted by a second expert for the defendants, Dr. Joan G. Haworth, an economist, achieved results consistent with the findings of Dr. Jones. Tr. 2020-25; see also DX KKK-8. Dr. Haworth found that the factors of the 1967 salary market and of the pre-Cresap pay rates exhibit a stronger statistical relationship to the “changes” between Exhibit 616-A and the final Cresap report than does the factor of sex composition. Tr. 2023; DX KKK-8. She testified that the “gender coefficient" identified by Dr. Treiman in his analysis of Exhibit 616 and of
*1391
the Cresap final report becomes statistically insignificant when market and pre-Cresap salaries are introduced. Tr. 2023. Thus, from these reports, the defendants would have the court infer that these salary grade differentials are not the product of sex discrimination but of an attempt to adjust salary grades so as to track market salaries and so as not to depart from preCresap salaries.
For the most part, the court finds Dr. Jones’ and Dr. Haworth’s modifications to Dr. Treiman’s study to be reliable (although the court has some reservations about the methods by which Dr. Jones narrowed the relevant data base, see Tr. 1673-80); that is, the court generally accepts the results of the regression analyses of Drs. Jones and Haworth. Nonetheless, the court finds their conclusions in this regard to have little weight. First, as indicated above, the court has determined that Exhibit 616-A cannot be read as a reliable source of data concerning the job evaluation survey. To the extent that Dr. Treiman’s data as to point score values are not trustworthy, the analyses of Drs. Jones and Haworth (which parallel Dr. Treiman’s study) are equally dubious representations of what happened during the Cresap process. Even though Drs. Jones and Ha-worth made changes to Dr. Treiman’s data base for their studies, Tr. 838, 842, and 1538-39, their data bases were nonetheless derived from the same unreliable source as was Dr. Treiman’s — that is, from Exhibit 616-A.
Further, the court finds that there is at best only an incomplete basis in the record to support a conclusion that market data and pre-Cresap salaries were actually considered in the setting of salary grades during the Cresap process. Although the final Cresap report indicates that jobs were evaluated and grades were assigned on the bases of the four factors indicated above, the report does not state that no other factors were considered. Nonetheless, the evidence that these two particular factors — labor market data and pre-Cresap grades — were also used in the determination of specific salary grades is not compelling.
First, as indicated above with respect to the training and experience study of Dr. Treiman, the court finds no satisfactory basis on which to conclude that the Cresap process entailed the use of labor market data to set specific salary grades. Next, as to consideration of pre-Cresap salaries, the evidence is even less supportive of the defendants’ argument. Not only was there absolutely no testimony that the County team or the Cresap personnel modified salary grades in order to maintain pre-Cresap salaries, the final Cresap report strongly implies that such was not the case. The cover letter for the Cresap report indicates that, in those cases for which the proposed salary grade would generate a salary lower than the present salary of a job title, “[n]one of these individuals should have his salary reduced so long as he holds the position, but such employees should not continue to receive salary increases in the future unless they are moved to higher level positions.” PX 621 (cover letter). This proposed process for handling disparities between existing salaries and proposed salaries — known as “redlining” of job salaries, Tr. 852 — is simply inconsistent with the argument of the defendants that salary grades were altered so as not to effect significant changes in the Nassau County salary status quo.
For all these reasons, the court is unable to attribute more than minimal weight to the modifications to Dr. Treiman’s report that were conducted by Drs. Jones and Haworth. Indeed, the court finds it more than a little remarkable that so much expert time and so much expert money were devoted to an exegesis of Exhibit 616-A — a document of obscure content and of dubious significance. As indicated above, Exhibit 616-A is not at all probative of the “final results” of the job evaluation process; as such, the studies and analyses of Exhibit 616-A are not probative of discriminatory intent in that process.
4.
Conclusion as to Cresap Process
Finally, the court finds that — beyond the studies of Dr. Treiman and beyond Exhibit 616-A — the plaintiffs introduced no other
*1392
evidence that was probative of discriminatory intent in the Cresap process. Indeed, as indicated above, testimony adduced by the defendants indicated that no one associated with that process had any information as to the sex composition of Nassau County jobs in 1967. See, e.g., Tr. 377, 1455, and 1610. This fact — along with the difficulty encountered by the plaintiffs in collecting such data — confirms that the participants in the Cresap process had no particular knowledge as to the sex composition of individual Nassau County job titles at the time of the salary grading process. Again, every witness associated with that process denied that any decision or determination was ever made on the basis of the sex of job incumbents or on the basis of the sex domination of any particular title. See, e.g., Tr. 348, 377, 435,1455, and 1464. The court found those denials credible and consistent with the evidence adduced by the plaintiffs as to the Cresap process. Accordingly, the court finds that there was no discriminatory intent in the Cresap job classification and evaluation process.
B.
Discrimination after the Cresap Process
1.
Perpetuation of the Cresap System
The plaintiffs argue that the discriminatory aspects of the Cresap system have been perpetuated by the County. Specifically, the plaintiffs point out that the Cresap system remained in place through the time of this trial and that new job titles in Nassau County are assimilated to old job titles. See, e.g., Tr. 389-96 and 1282-84; PX 1074 at 17-19 (report of Dr. Treiman); see also PX 1096 at 10-13. However, the fact that Nassau County continues to use the Cresap structure as the basis for its classification and compensation system is only probative of discriminatory intent to the extent that the initial execution of the Cresap process involved discriminatory intent or to the extent that it is perpetuated
because
that system has adverse effects on women in female-dominated jobs. The court has already determined, however, that there is no evidence of discriminatory intent in the Cresap process. Further, the plaintiffs introduced absolutely no evidence that the County continues to use the Cresap system because it treats these plaintiffs less favorably than it does other employees. Accordingly, the mere fact of the perpetuation of that system — either through preserving existing titles and salary grades or through assimilating new titles to old ones — does not indicate that the County of Nassau has in the past or does now discriminate against its female employees in compensation.
2.
Requests for Upgrades since Cresap
The plaintiffs sought also to show that the manner in which the County has considered requests for upgradings of particular titles since the Cresap process is probative of discriminatory intent by the County. That is, the plaintiffs attempted to show that the County has generally denied requested upgrades for female-dominated job titles but that the County has expeditiously granted upgrades for male-dominated titles. See, e.g., PX 1074 at 19-28 (report of Dr. Treiman). However, the court finds that the evidence on this issue is not as simple as the plaintiffs have presented it; rather, the evidence reveals that the County has undertaken a variety of legitimate responses for non-discriminatory reasons to the problems that drive requests for upgrades. That is, the court finds that the evidence on this point does not support an inference of discriminatory intent.
As a threshold observation, this court was disturbed by the report and the testimony of Dr. Treiman on this question of the County response to upgrade requests. See PX 1074 at 19-28; Tr. 634-36. With the exception of a brief “quantitative analysis” in his report, Dr. Treiman’s conclusions are based on a review of County memoranda and of deposition testimony; his report details little more than his inferences and “impressions” from those sources. See, e.g., PX 1074 at 23. A study of the evidence evaluated by Dr. Treiman readily reveals that it presents factual issues properly resolved by this court — not by an expert witness. Indeed, it is plain that Dr. Treiman’s expert qualifications as a sociologist and as a statistician were en
*1393
tirely irrelevant to this segment of his report.
As indicated above, employees and departments often request upgrades for particular job titles. These requests are reviewed by the Nassau County Civil Service Commission; although the Civil Service Commission either recommends or disapproves of the requested upgrade, final authority to change the salary grade of a job title rests with the Board of Supervisors. Further, the two principal criteria on which the Civil Service Commission considers requests for upgrades are changes in the complexity of job duties over time and problems with recruitment or with retention of personnel.
Almost all the job titles in the caseworker career line — the incumbents of which handle adoptions, foster care, and protective services — are female-dominated. Tr. 450-51; PX 672. Despite the fact that the Department of Social Services has perceived increasing complexity of job duties in these titles, PX 1086 at 23-28 — and despite the fact that there have been clear recruitment and retention problems for the lower caseworker job titles, Tr. 451 and 947 — the County has denied requests for upgrades of these titles even when recommended by the Civil Service Commission. Tr. 456-57 and 924-25. However, two important facts undermine any inference of discriminatory intent: First, retention problems in the caseworker title series have arisen largely from the fact that the male and female incumbents of those jobs move almost invariably and always freely into the more highly paid job titles of the probation officer job series. PX 126; Tr. 452, 481, 925-26, 1277, and 1412-13. That the County has never sought to prevent the female caseworkers from effectively “upgrading” themselves into more well compensated positions with the County belies the notion that the County has intentionally underpaid them because they are women. Second, in lieu of specific upgrades, the County has encouraged mass promotions of caseworkers into higher titles within that job series — thereby increasing their salaries. PX 1086 at 31.
Similarly, the Department of Social Services has frequently sought upgrades for the job titles in the female-dominated social welfare examiner job line. Tr. 449-450 and 496. The Department has demonstrated to the Civil Service Commission that it has had severe retention problems in the Social Welfare Examiner I title and that the salary for that title was below market rates. Tr. 445-50. All requests for upgrades have been denied by County officials — even when recommended by the Civil Service Commission. Tr. 450 and 924. However, to alleviate the retention problems, the Civil Service Commission permitted the Department to institute rapid mass promotions of employees in the title of Social Welfare Examiner I to the title of Social Welfare Examiner II. Tr. 474-76, 479 and 1271. The result of this policy is that employees in the entry level position of Social Welfare Examiner I are now promoted after only six months to Social Welfare Examiner II — and that they receive a three-grade increase upon promotion. Tr. 477. Thus, in effect, the lowest level employees of the social welfare examiner job series have received a three-grade upgrade. Tr. 475-77. This has reduced the retention problems in that job series. Tr. 477. To the extent that this job series is female-dominated, the policy of early and regular promotions — with its attendant increase in salary — is plainly inconsistent with discriminatory intent.
The dietitian job series is female-dominated. PX 672. The plaintiffs argue that two aspects of the salary grades for these jobs reveal discriminatory intent. First, although one of the “rules of thumb” of the personnel specialists at the Civil Service Commission is that job titles that require a baccalaureate degree are given a salary grade of ten, the title of Dietitian I (which does require such a degree) was given a salary grade of nine through the Cresap process, and that grade has remained unchanged. PX 672. Second, repeated requests for upgrades because of recruitment and retention problems have been denied. Tr. 457-58, 934; PX 262; but see DX ZZZ-3 at 10. However, as to the first of these, the court notes that “training and
*1394
experience” was only one of four factors on which a job was evaluated and on which a salary grade was assigned during the Cresap process. Thus, that Dietitian I was given a salary grade of nine and other job titles that required a college degree were given a grade of ten is not in itself particularly probative of discriminatory intent. Cf. Tr. 392. Next, although the County has denied upgrade requests for the dietitian series, it has undertaken to hire new employees for these jobs at higher salary steps. Tr. 482. In other words, the County now pays those employees — who are, by hypothesis, mostly women — more than it did before. Tr. 482-83. Discriminatory intent cannot be inferred from the fact that the County has refused to grant an increase in salary grade while effectively granting an increase in the salary itself.
The plaintiffs make similar arguments as to the job titles in the career series of physical therapists and of occupational therapists. The court notes first that not all the job titles in these career lines are female-dominated. For example, Physical Therapist I has six male incumbents, five female incumbents, and one incumbent whose sex is not reported in the record. See PX 672. In any event, the County has refused to grant upgrades to these titles despite severe recruitment and retention problems. Tr. 518-19. Again, however, the County has begun to hire employees for these titles at higher salary steps— thereby providing higher compensation for these jobs. Tr. 518-19.
The registered nurse job titles are female-dominated. PX 672. Since at least the early 1980s, recruitment and retention problems in these job titles have been severe and persistent. Tr. 525-26. Nonetheless, the County has increased compensation for these employees more than once. In 1982, the County began to higher nurses at above entry-level salary steps. Tr. 917-18; in 1987, the County granted (through the collective bargaining process) a one-grade “vertical” upgrade — pursuant to which nurses were moved to the next higher salary grade at the same step they had previously occupied. Tr. 1362-64 and 1372. In that the County has increased the pay of these job titles, the court is unable to find that the response of the County to recruitment and retention problems with registered nurses reveals any discriminatory intent toward women.
In contrast to all these job titles for which upgrades have been denied but for which special pay increases have nonetheless been granted, the female-dominated job series of police communications operator appears to have suffered from chronic and severe recruitment and retention problems for the better part of a decade. PX 912-M. Nevertheless, requests for upgrades have been denied. Tr. 1122-26. The County has adduced no evidence that these job titles received increases in compensation to address these recruitment and retention problems. This failure by the County to upgrade these titles or otherwise to provide additional pay to the job incumbents is inconsistent with the County policy of providing salary increases under such circumstances; however, the court is unable to conclude on the basis of one such example that the County systematically and regularly discriminates against female-dominated jobs in considering requests that job titles be upgraded. Rather, the singularity of this example indicates that it is an aberration.
With respect to the job series of nurses’ aides, patient care assistants, and clerical positions, all female-dominated, the court is unable to find evidence of discriminatory intent by the County. As to nurses’ aides, recruitment problems have not been significant. Tr. 459-60. As to patient care assistants, there was no evidence that any recruitment or retention problems have ever been encountered. Rather, the plaintiffs adduced evidence that the duties of patient care assistants have become more complex. Tr. 572-73. The court found the testimony on this point to be vague and unpersuasive. Also, as to the clerical job titles, the court did not find that the recruitment problems about which the plaintiffs adduced testimony were so clear as to suggest that the County was disregarding its own criteria in denying upgrades to these positions. Compare Tr. 528-31 with PX 1096 at 36. The
*1395
treatment of these three job series does not furnish a basis for an inference of discriminatory intent.
The plaintiffs contrast all these instances of denied upgrade requests in female-dominated job titles to instances of granted requests in male-dominated and in mixed job titles. Thus, for example, titles in the job series of veterans counselor have been upgraded — as have job titles for pharmacists, public health engineers, environmental chemists, ambulance medical technicians, computer programmers, and automotive maintenance coordinators. See PX 587, PX 608, Tr. 922, Tr. 920, PX 608, DX C-40, Tr. 1336-40, and PX 389. However, these upgradings are probative only of the fact that the County does change salary grades — and, hence, salaries — when recruitment and retention problems or when changes in job duties so merit. That these male-dominated and mixed titles have been upgraded would be probative of discriminatory intent by the County if female-dominated jobs that had similar justifying reasons were denied upgrades or the equivalent of upgrades — that is, increases in salary. But the fact that such female-dominated jobs as caseworkers, social welfare examiners, dietitians, and therapists have experienced salary
increases
— albeit not through the form of upgrades — in response to similar problems does not indicate any discrimination by the County. Rather, it indicates precisely what one would expect: that when the County experiences difficulty in hiring and retaining qualified employees, the County adjusts the compensation of those employees in order to remain competitive in the local labor market. The fact that some of these pay increases are through the form of upgrades while others are in forms more narrowly tailored to address a particular problem (for instance, the institution of above-step hiring to meet recruiting needs) is clearly of no consequence: In either event, salaries are increased. Indeed, the County employees who have been responsible for the evaluation of upgrade requests flatly denied that the sex of job title incumbents has ever entered into any such compensation decision. See, e.g., Tr. 435, 946, 1279-80, 1858-59; see also PX 1096 at 70-71. Again, this testimony is consistent with the other evidence presented.
C.
Statistical Analyses of Current Salaries
The plaintiffs introduced extensive expert testimony to demonstrate that there is at present a significant disparity in salaries between male-dominated jobs and female-dominated jobs in Nassau County. Their expert also testified that this compensation disparity correlates to a statistically significant degree with the sex composition of the jobs. On the bases of multiple regression analyses, the plaintiffs’ expert attempted to demonstrate that the compensation disparity was understandable only with reference to the sex of Nassau County job incumbents. The defendants, however, introduced expert testimony in order to demonstrate that the present compensation disparity in Nassau County jobs was better explained by such factors as the labor market and the number of hours worked each week by employees. The court makes the following findings of fact with respect to the testimony of these experts.
1.
Dr. Michelson’s Report
The plaintiffs’ expert witness on the matter of current salaries in Nassau County was Dr. Stephan Michelson, an economist and consultant who has appeared as an expert witness in many Title VII actions. Notwithstanding that in opening remarks plaintiffs’ counsel promised that Dr. Michelson would provide compelling evidence of discrimination, Tr. 13-14, and notwithstanding that the plaintiffs paid Dr. Michelson over $850,000.00 in fees even before this trial began, the plaintiffs have since tempered their estimation of the comparable worth of Dr. Michelson’s services. See, e.g., Plaintiffs’ Post-Trial Brief at 30 (statistical analyses by Dr. Michelson “not critical to the outcome in this case”); see also Tr. 2154 (argument of plaintiffs’ counsel to court that “[t]his case ultimately shouldn’t turn on statistics”). The plaintiffs' reevaluation of Dr. Michelson is prudent: Dr. Michelson proved to be as evasive a wit
*1396
ness as Dr. Treiman. See, e.g., Tr. 113-14, 132-138, and 144-48.
Dr. Michelson initially submitted a report to the plaintiffs on February 20, 1989; on November 1, 1989, he revised that report by means of appendices “F” and “G” to the original report. PX 1057 (report of Dr. Michelson to plaintiffs, dated November 1989 and entitled: “A Study of Salary Rate Differences Between Male and Female Employees , of Nassau County, N.Y.”). Dr. Michelson himself appears to believe that the revisions of his new report greatly enhance the quality of the initial study: “The result of these refinements is to make this presentation ... the most exacting model of a salary system, using the most error-free data set, ever presented in litigation.” PX 1057 at G-3. In any event, the revisions to Dr. Michelson’s first report certainly had the effect of reducing the salary differentials that he attributed to sex.' Tr. 132. To the extent that these reports are inconsistent,- the court will consider his revised report as superseding his initial report.
Dr. Michelson undertook to evaluate the salaries of “historically” male-dominated jobs and “historically” female-dominated jobs (although he conceded to the court that he had no adequate bases on which to characterize job titles as “historically” male- or female-dominated). Tr. 13. Dr. Michelson proceeded, as have all persons involved in this case, on the arbitrary demarcation established by the plaintiffs as to sex domination: That is, Dr. Michelson accepted as given that a job title comprised of 70% women is “female-dominated” but that a job title with 69% women is “mixed.” Tr. 13-14 and 114.
Dr. Michelson considered the mean salaries for Nassau County employees in the years 1983 and 1986. Tr. 9. He found that the raw difference in the average salaries of male-dominated jobs and of female-dominated jobs was $7,873 in 1983 and $9,166 in 1986. Tr. 15. Dr. Michelson then undertook to subject these “raw differences” to multiple regression analysis. Principally, Dr. Michelson sought to determine how much of these basic disparities were attributable to differences in the descriptions for the Nassau County job titles — the bases on which the Cresap team had ultimately evaluated jobs and assigned salary grades. In his final analysis — with Nassau County police officers considered separately (because they have a separate pay system, Tr. 29-30) but with job descriptions accounted for — Dr. Michelson found that the difference between the average salary of male-dominated jobs and the average salary of female-dominated jobs was $2,473 in 1983 and $3,298 in 1986. Tr. 16; PX 1057 at G-20. Cast in slightly different terms, Dr. Michelson found — again, with police considered separately and with job descriptions accounted for — that every percentage point increase of females in a Nassau County career line corresponded to a decrease in salary of $31 in 1983 and of $42 in 1986. PX 1057 at G-21. Dr. Michelson testified that these sex-associated differences in salaries are not random. Tr. 104-05. Rather, Dr. Michelson found that the salary disparity in Nassau County is clearly related to the sex of Nassau County employees. Tr. 104-05, PX 1057 at 2.
The methodology of Dr. Michelson’s study was to analyze Nassau County job descriptions and to try to create an independent variable for every aspect of these specifications that have a bearing on salaries. He noted that, although the Cresap team formally evaluated each job on the basis of four factors, these factors could be subsumed, in his opinion, under two groups: (1) factors that concern characteristics of the employee sought for the job (such as education and experience); and (2) characteristics of the job itself (such as complexity of duties and supervisory responsibility). Tr. 31. Dr. Michelson also recognized, aside from the job descriptions, that the seniority of employees has an effect on salary, and he attempted to account for that fact in his regression analyses. PX 1057 at 133. But his principal task was to evaluate the effect that the characteristics of the Nassau County job descriptions have on the critical dependent variable against which all aspects of this lawsuit are measured — salary. It was therefore the objective of Dr. Michelson to account
*1397
for as many of these characteristics as was both possible and appropriate. PX 1057 at G — 21 to G — 22.
In the formulation of his data base, Dr. Michelson began with the decision to analyze “career lines” rather than separate job titles. Tr. 34-35. Thus, rather than consider separately the jobs of Probation Officer Trainee, Probation Officer I, Probation Officer II, Probation Officer Supervisor I, and the rest of the probation officer jobs separately, he evaluated the entire probation officer career line as one unit. Dr. Michelson also testified that, in Nassau County job titles, most women are in female-dominated career lines and most men are in male-dominated career lines. Tr. 53.
As noted, the most significant aspect of Dr. Michelson’s study was his derivation of independent variables from the Nassau County job descriptions. With the aid of the “writer’s manual” produced by Cresap to train the Nassau County team to write job descriptions, PX 620, Dr. Michelson translated the language of these title specifications into several dozen independent variables. Tr. 60. For example, many job specifications indicate that a particular job title requires a certain degree of knowledge about certain subject matter. If the specification indicated that the title required only some or no knowledge about certain subject matter, Dr. Michelson recorded the variable “KNOWOl” for that job. If the job specification indicated that “knowledge” simplieiter was required, he coded it as “KNOW2”. The variable “KNOW3” and “KNOW4” represented job specification language of “considerable knowledge” and of “thorough knowledge” respectively. PX 1057 at G-5; Tr. 60; see also PX 1057 at 100-01 and at G-4 to G-5 (comprehensive lists of independent variables). Dr. Michelson conceded that he did not attempt to code every “minute variation” in the language of the job specification and also that such coding of language required subjective judgments, Tr. 73-75; nonetheless, he testified that, in his opinion, he had accounted for all variables that had explanatory power with respect to the salary differentials he had earlier identified. Tr. 77.
Again, after coding the language of the job specifications — both the characteristics of the job itself and the characteristics of the person required for the job — into independent variables and after accounting as well for seniority, Dr. Michelson concluded that:
[Sjome of the salary differential is explained by characteristics held differently by males and females, and gained (arguably and assumedly) independent of the actions of Nassau County. Some of the salary differential is explained by differences in the specifications of the jobs males and females are in. However, ... neither individual nor job characteristics explain all of the sex-related salary differences. Salary appears to be related to gender itself.
PX 1057 at 1-2.
Although the scope of Dr. Michelson’s undertaking in this project was ambitious, certain methodological shortcomings prevent this court from assigning more than minimal weight to his conclusions. As a threshold matter, the court must discount somewhat Dr. Michelson’s own assertions that he accounted for all statistically significant variables in his regression analysis. He himself testified that he had to revise his initial report because he concluded that important variables had not been considered. Certainly he would not have submitted his initial report as “final” in February of 1989 did he not believe then that it was comprehensive. He also characterized the process of identifying and of coding significant aspects of the job specifications as “unending.” Tr. 20.
Perhaps one of the most troubling aspects of Dr. Michelson’s study is that, even though he himself did not personally code any of the job specifications, and even though he acknowledges that the coding process required subjective judgments, he did not record any of these judgments for subsequent review or evaluation. Tr. 216. In fact, despite his insistence to the contrary, Dr. Michelson made clear on cross-examination that he cannot now determine whether or not his staff followed his in
*1398
structions in the coding process because he does not know what instructions he gave. Thus, for example, on the question of how to code the phrase “prepares progress reports”, the following exchange took place:
Q. Do you know whether in every case where your staff saw the language “prepares progress reports,” they coded it as communication?
A. No, I only know they either did or didn’t.
Q. In fact, you are not sure as you sit here today whether it was coded at all, isn’t that correct?
A. If it was coded at all, either as zero or one, that is, it was consistently handled upon instruction from me but I can’t recall that instruction or most of them.
Q. You can’t as you sit here tell us today whether any particular instruction was followed or not, isn’t that correct?
A. Oh, no, we could look at volume 4 of the supplement [to the November 1, 1989 report] and find out exactly how this was coded and then the question is whenever we see these records, prepares program [sic] reports, will it be coded the same. It will be coded the same but I can’t tell you which way.
Q. And you can’t tell what instruction was given?
A. Yes, I can tell. If we look at the supplement, we will see whether it was or was not coded and that is the instruction that was given.
Tr. 216-17. In other words, because he failed to record instructions to his staff, Dr. Michelson has no way to determine what a particular coding instruction was other than to review the results of the work of his staff; but he also has no way to verify that his staff followed that instruction other than to review the same information. In order to investigate, then, whether Dr. Michelson’s staff consistently followed the instructions that he gave them on certain “judgment calls”, it becomes necessary
to assume
that Dr. Michelson’s staff consistently followed the instructions that he gave them on those matters. Apparently, Dr. Michelson does not even entertain the possibility that his staff may have consistently misunderstood or misapplied his directions on a given question. In any event, without records of those directions, it is not possible to assess the quality of such codings by Dr. Michelson’s staff — even though Dr. Michelson acknowledged that the results of the coding process depend in part on who does the coding. Tr. 237. Indeed, Dr. Michelson himself concedes that, without those records of his coding rules, there is no way to know that any other expert could replicate — and, therefore, verify — his study. Tr. 235-36.
Furthermore, the court found that the manner in which Dr. Michelson — or, more precisely, his staff, Tr. 189-90 — coded certain specifications to be of dubious legitimacy. For example, the job specifications for Nassau County police officers indicate that an incumbent must “develop skill in the use of firearms.” Tr. 175. Accordingly, Dr. Michelson did not code this job for the ability to use a gun but for the ability to learn to use a gun. Tr. 178. However, again following the precise language of the job specifications, Dr. Michelson did code
supervisory
police positions for the attribute of ability to use a gun. Tr. 191. In other words, on Dr. Michelson’s model, a police officer who patrols the streets of Nassau County is deemed to have the attribute of the capacity to learn to use a gun, but his captain at headquarters is recorded for the attribute of active use of a gun. Insofar as Dr. Michelson acknowledged that he and his staff had at times made subjective judgments in the coding process, the court cannot help but wonder whether a modicum of common sense in instances such as this might have produced a more “exacting model” of the Nassau County salary system.
Next, Dr. Michelson testified that he had used only two sources of data: job specifications and personal information about employees. Thus, other than sex, seniority, and salary, Dr. Michelson claimed to have considered nothing that was not contained in a Nassau County job specification. Tr. 160-61. But Dr. Michelson was not consistent in the application of this methodological rule: At times, he coded his own inferences about the characteristics of a job
*1399
rather than the language of the specification; at other times, he coded phrases of effectively identical significance with different values. Thus, a job title might receive the variable for stress — even if the word “stress” appears nowhere on the job specification. Tr. 185. And, when a career line is considered, such a variable as “stress” may appear in some job codings but not in others. Tr. 246-47. Alternatively, Dr. Michelson coded differently the phrases “receives training” and “ability to learn” — because, as he testified, he inferred a significant distinction between the two. Tr. 224-226. And, at times, Dr. Michelson simply did not coded certain phrases in job specifications that clearly have significance as to the setting of salaries. Tr. 209-10 (phrase “ability to carry out oral and written instructions” not coded — and, therefore, not evaluated — as a salary determinant). Such selective departures from his stated methodology undermine the confidence that any reasonable person can have in his conclusions — particularly if one accepts Dr. Michelson’s own characterization of his study as an exercise in the coding of key words of uncertain reliability. Tr. 251.
Furthermore, to the extent that he did not stray beyond the four corners of the job specifications, Tr. 267-70, he failed to account for obvious and important influences on the salaries of particular job titles. For example, Vito Competiello, a participant in the Cresap process, testified that the members of the Nassau County job evaluation team had considered such nonCres'ap factors as the working conditions of a particular job, Tr. 322-23; and Adele Leonard, the Executive Director of the Civil Service Commission throughout the entire period relevant to this action, indicated that working conditions affect decisions as to salary upgrades, PX 1096 at 9. But Dr. Michelson’s study took no account of this aspect of a particular job title or career line. Thus, Dr. Michelson coded the job specification for the title of Morgue Attendant I with the variables that captured these factors: “graduation from high school”, “use of electric saw”, and “direct supervision.” Tr. 213-14. Dr. Michelson disregarded entirely the fact that a morgue attendant works with — and on — cadavers. Tr. 214. The court does not doubt that such circumstances of the job were carefully considered in the setting of salary grades during the Cresap process and continue to influence salary grade determinations. Tr. 325, 365-66; see also PX 1096 at 9 (testimony of Adele Leonard as to working conditions of morgue attendant: “It’s never been too desirable, bring a body in and see the maggots running out of it.”). Similarly, when Dr. Michelson coded the job description for the job of assistant detective investigator, he did not assess the fact that an incumbent of this job would be responsible for confidential information. Tr. 215. Dr. Michelson’s model took no account of these important considerations — such as whether a career line required work indoors or outside, whether it required work in a prison or underground, or whether it required working with the living or the dead.
Dr. Michelson’s multiple regression analysis also did not account for the fact that salaries in Nassau County are affected by the collective bargaining process — even though Dr. Michelson himself acknowledged that such negotiations are an “influence” on Nassau County salaries. Tr. 147; see also Tr. 145. In fact, Dr. Michelson revealed on cross-examination that he knew nothing about the collective bargaining process in Nassau County other than that it affects the salaries of the jobs which his model was intended to study:
Q: Do you strive to have your model reflect the real world as close as you possibly can?
A: Yes.
Q: Dr. Michelson, do you understand that over the last 22 years, salaries in Nassau County were arrived at through a collective bargaining process?
A: At the dollar level, yes, it’s my understanding.
Q: Are you aware that there have been a series of collective bargaining agreements that have governed the relation
*1400
ship between the CSEA and the County of Nassau throughout the period?
A: Yes, there are collective bargaining agreements.
Q: Have you ever examined any of those collective bargaining agreements?
A: No.
Q: Did you ever ask for any information about the collective bargaining process in Nassau County?
A: No.
Q: Did you ever receive any information as to how the collective bargaining process impacts on salaries in Nassau County?
A: No.
Tr. 144-45. And the evasiveness with which Dr. Michelson ultimately indicated that he had failed entirely to include the collective bargaining process among the dozens of independent variables in his multiple regression analysis is telling not only of the obvious limitations on his study but also of the quality of his testimony generally:
Q: So as you sit here today you have absolutely no idea of the impact of collective bargaining negotiations on the county’s wage setting, is that correct?
A: No, that is completely incorrect.
Q: Isn’t the collective bargaining process a variable that could affect the real world model of salaries in Nassau County?
A: I can’t answer that question as phrased.
Q: You didn’t examine any variable for collective bargaining in you model, did you?
A: Well, effectively, I did.
THE COURT: Dr. Michelson, with all due respect, sir, I’m having difficulty understanding your answers. Did you or did you not consider the collective bargaining process as one of the variables?
THE WITNESS: There are two things.
THE COURT: Let me ask the question again[.] [I]f you can answer it yes or no, please do so. Did you consider the collective bargaining process as a variable in the study which you’ve conducted?
THE WITNESS: No.
THE COURT: Do you regard it as a variable in determining the real world status of wage determinations?
THE WITNESS: I regard it as an influence, of course.
THE COURT: Is influence the same as variable?
THE WITNESS: No, I don’t know how to measure it as a variable.
THE COURT: Doesn’t a variable have an influence on the result of your study? THE WITNESS: I hope so but I have to be able to measure.
THE COURT: Do[] variable and influence mean essentially the same thing[.] That is, if you introduce a variable, it influenced the result, is that right? THE WITNESS: Yes.
Tr. 146-48. In other words, although Dr. Michelson was aware of the critical role that the collective bargaining process has had in the Nassau County compensation system, he did not undertake even a brief investigation of it, and he deliberately omitted the results of that process from his study because he did not “know how to measure it as a variable.” With such a significant — and, indeed, intentional— omission from his study, this court cannot conclude that Dr. Michelson’s multiple regression analysis is a reliable study of salary determinants in Nassau County job titles.
Along the lines of omitted salary determinants, the defendants sought to make much of the fact that Dr. Michelson had included no variable to represent the number of hours that Nassau County employees work each week. Tr. 150-51. In fact, Dr. Michelson was completely unaware until just a few days before the start of trial that employees in different Nassau County departments work 33 and three-quarters hours each week or 35 hours each week or 40 hours each week. Tr. 153. However, the evidence produced later in the trial unequivocally established that the number of hours worked each week in no way affects the salaries of County employees: A clerk in one department who works 35
*1401
hours each week is paid the same salary as a clerk in another department who works 40 hours each week. Tr. 1442; see also Tr. 326 and 936-37. For this reason, the omission of a variable to represent the length of County work weeks — as well as Dr. Michelson’s ignorance of the fact of different work weeks — is of absolutely no significance.
What is of significance — indeed, of great significance — is the complete failure of Dr. Michelson’s study to account for the effect of market forces o

[Text truncated at 120,000 characters. The full text is on the page linked above.]

---

Source: Frix Law Library, https://www.frixlaw.com/law-library/cases/1379222. Public record. Not legal advice.