From Casetext: Smarter Legal Research

E.E.O.C. v. International Business Machines Corp.

United States District Court, D. Maryland
Apr 13, 1984
583 F. Supp. 875 (D. Md. 1984)

Opinion

Civ. A. No. R-80-1408.

March 29, 1984. As Amended April 13, 1984.

Earl Harper, Jr., John T. Irick, Isaac Joe, Jr., Frederick P. Charleston, Larry Newell, Barry I. Bank, E.E.O.C., Baltimore Dist. Office, Baltimore, Md., for plaintiff.

J. Hardin Marion, Peter J. Sommer, A. Thomas Pedroni, Jr., Tydings Rosenberg, Baltimore, Md., Jerome Ackerman, Michael S. Horne, Jeffrey G. Huvelle, Covington Burling, Washington, D.C., for defendant.


OPINION


Plaintiff Equal Employment Opportunity Commission ("EEOC") filed the complaint in this action on June 3, 1980, alleging racial discrimination in employment in violation of Title VII of the Civil Rights Act of 1964, as amended, 42 U.S.C. § 2000e, et seq. ("Title VII"). The complaint challenges that defendant International Business Machines Corporation ("IBM") discriminated in its (1) performance planning, counseling and appraisal program, (2) compensation program and (3) promotion program as these programs applied to black managerial and professional employees at defendant's Data Processing Division ("DPD") facilities in Maryland.

Plaintiff presented its case-in-chief from April 15 until May 15, 1983, and the defense spanned the period from May 16 until June 14, 1983. The EEOC offered rebuttal from June 14 until June 17, 1983, at which time the trial record was closed. After the parties submitted proposed findings of fact and conclusions of law, the Court heard final argument on August 12, 1983. Pursuant to Fed.R.Civ.P. 52(a), the Court makes the following findings of fact and conclusions of law.

I. Procedural Background

On August 16, 1976, George W. Hunter, Jr. submitted a charge of discrimination (# 033-627871) with the EEOC alleging that IBM had discriminated against him on the basis of race with respect to performance plans and evaluations, promotions and pay. The EEOC initially referred the charge to the Montgomery County Human Relations Commission for processing, but later rescinded the referral, and the charge was officially filed with the EEOC on October 14, 1976. Hunter amended the charge on October 26, 1976, and the EEOC first notified IBM of the charge on November 1, 1976.

On August 1, 1977, Hunter submitted a second amended EEOC charge alleging retaliation for the filing of the original charge. The EEOC issued a determination letter on November 3, 1977, finding reasonable cause to believe that IBM had discriminated against Hunter and other black professionals and managers in Maryland. The EEOC also asked IBM to engage in conciliation efforts, including the granting of back wages and/or promotions to members of the class, reinstatement of Hunter and promotion to a branch manager position with back pay, and modification of certain employment practices at DPD. After IBM refused to accept these proposals, the EEOC notified IBM on May 23, 1978, that conciliation efforts had failed.

Hunter had resigned from IBM on April 30, 1977. In the August charge, he claimed that his resignation constituted a constructive discharge and that IBM had retaliated against him for filing his 1976 charge by evaluating him before he had sufficient time to achieve the objectives outlined in a performance plan, and by giving him a low appraisal.

Among the proposed changes were the adoption of job posting in Maryland, the basing of promotions on education and/or seniority, and the elimination of "subjectivity" in IBM's performance appraisal process.

The EEOC complaint was not thereafter filed until June 3, 1980. The complaint alleges that IBM had discriminated against black professionals and managers employed in IBM's Maryland DPD in the categories of pay, promotions, and performance evaluations. The EEOC cites IBM's failure to develop non-discriminatory criteria to measure the effectiveness of their managers and professionals, to post notices of position vacancies, and to take affirmative action to eliminate alleged discriminatory employment policies and procedures. As to George Hunter specifically, the complaint alleges that IBM discriminated against him by refusing him promotion to a branch manager position, by paying him discriminatory wages, and by constructively discharging him.

IBM's answer denied all of the EEOC allegations and maintained that its standards for compensation, promotion, and evaluation constituted a "bona fide merit system" as outlined in 42 U.S.C. § 2000e-2(h). In addition, IBM raised laches and the statute of limitations as affirmative defenses, both of which were dismissed by the Court by letter dated March 2, 1983, without prejudice to defendant's right to raise these defenses at the close of plaintiff's case or at the close of all evidence.

By the same letter, the Court confirmed that the claims to be tried were those of black managerial and professional employees only and not those of marketing or sales personnel, which plaintiff had belatedly attempted to raise. Although the Court did not dismiss the claims pre-dating December 18, 1975 on limitations, because the parties did not engage in discovery on these claims, they were not deemed appropriate for presentation at trial. The Court did, however, give plaintiff the opportunity to show continuing violations and accord it the option to move at the end of the case to reopen the liability issue on the pre-1975 claims. The plaintiff did not request to reopen discovery on the earlier claims arising in 1965-75 until the eve of trial, when it was estimated that it would take six months to two years of further discovery to prepare these earlier claims for trial. Although the Court denied the plaintiff's request to reopen, it did allow plaintiff to offer at trial any background material that it deemed relevant.

IBM consistently took the position that December 18, 1975, was the cutoff date for substantive claims and, allowing for a reasonable background period, did not respond to discovery requests prior to January 1, 1974. The EEOC never sought to compel discovery pertaining to earlier events and did not file any timely opposition to IBM's stance on the time frame of discovery.

In making its findings of facts, the Court first turns to George Hunter's individual claims. The Court will then consider the anecdotal testimony offered by the class witnesses and finally the statistical evidence offered by both parties.

II. Findings of Fact

A. Individual Claims of George Hunter

George Hunter began working for IBM on November 14, 1966, as a systems engineer ("SE") trainee. His initial compensation was $9,000 per year. Upon completion of training, Hunter qualified as an associate systems engineer, a level 52 position. IBM allowed Hunter to switch to a marketing job in November of 1968. He was the "team leader" of the NASA-Goddard marketing team and oversaw sales to the orbiting astronomical observatory. In July, 1971, IBM promoted Hunter to a level 56 position: marketing programs administrator for the medical industry. In July, 1972, Hunter was again promoted, this time to marketing manager in National Federal Marketing ("NFM"), a level 59 position.

Each job in IBM is assigned a two-digit "salary level." At each salary level there is a set compensation range.

Under Hunter's management, his marketing team won a significant account, that of the National Oceanographic Atmospheric Administration (NOAA). This "win" provided the impetus for his selection as NFM's outstanding Marketing Manager for 1973. Up to this point, Hunter's success and attendant advancement within IBM was exceptional. His performance ratings were all satisfactory or better. (The lowest was a "3".)

IBM uses a numerical rating system (except for training periods) by which a manager rates an employee's effectiveness in carrying out responsibilities enumerated in individual performance plans. The manager measures the employee against the objectives of the plan and assigns ratings for each important area, in addition to an overall score. During the relevant period between 1975-1981, there existed four possible levels of satisfactory performance, with a "1" rating the highest and a "4" rating the lowest level of satisfactory performance. At various times there were either one or two levels of unsatisfactory performance. As of 1981, the four categories of satisfactory performance were:
Rating
1 The results achieved far exceeded the requirements of the job in all areas.
2 The results achieved consistently exceeded the requirements of the job in all key areas.
3 The results achieved consistently met the requirements of the job and exceeded the requirements in many areas.
4 The results achieved consistently met the requirements of the job.
As a general rule, IBM employees are to receive an evaluation after six months in a new job and annually thereafter, subject to certain exceptions.

On January 16, 1974, Hunter was appointed administrative assistant ("A.A.") to C.E. (Chuck) McKittrick, Jr., DPD Vice President for Public Sector Industry Marketing ("PSIM"). This position was generally regarded as a stepping stone into upper level management at IBM. It is a temporary assignment of approximately one year's duration, during which a rising employee is given the opportunity to gain a breadth of experience and exposure to the office of the vice-president and its duties. Practically speaking, the position is "unleveled" because of its unique training nature, although it is formally assigned a level 50 holding code. It was customary that no performance plan be prepared for the position, and McKittrick did not prepare one for Hunter.

The performance plan issue, however, was a primary focus of Hunter's testimony at trial. Hunter asserted that the lack of a performance plan adversely affected both his chances for promotion and his salary as determined by company guidelines. The EEOC asks the Court to find that McKittrick refused to put a performance plan in place for Hunter during the fifteen-month period he served as his A.A. despite "Hunter's repeated requests." According to McKittrick, Hunter never complained about the absence of a plan, nor did he request one. McKittrick had never prepared a formal performance plan for any of his previous A.A.'s, all of whom were white. Because Hunter as McKittrick's A.A. worked under his close daily supervision, there was little or no need for a performance plan. Hunter discussed his duties with McKittrick as well as with his predecessor as A.A., Max McCullough. McKittrick evaluated Hunter on an ongoing basis and the Court finds that the absence of a performance plan was neither irregular nor did it hinder Hunter in carrying out his duties.

IBM uses a formal merit pay system to compensate its employees. As previously noted, each position in IBM is assigned a two-digit salary level. At each salary level there is a set range of possible salaries, which is adjusted annually for industry competitiveness. The maximum figure in the range is 60% higher than the minimum figure which allows the company to reward the more effective employees. The salary range at each salary level is divided into quarters that correspond to the four satisfactory performance ratings (1-4). Thus, in 1980 for example, the monthly pay ranges for Levels 52-55 were:
Fourth Third Second FirstLevel Quarter Quarter Quarter Quarter 52 $1595-1832 $1833-2069 $2070-2307 $2308-2545 53 1765-2029 2030-2294 2295-2559 2560-2825 54 1965-2259 2260-2554 2555-2849 2850-3145 55 2185-2512 2513-2839 2840-3167 3168-3495
The difference within each pay quarter allows room for an increase when an employee, who remains at the same salary level, sustains his performance rating over time. When first promoted to a higher salary level, employees typically are compensated near the bottom range for that new level until sustained performance raises it. This may result in employees being "below minimum" at the beginning of the year when a new salary range is established to account for inflation. It is expected that by year end the employee's salary will be brought up to minimum.

Although technically unleveled, Hunter was paid a salary equivalent to a level 59. Nevertheless, it was Hunter's opinion that his position as A.A. was commensurate to an industry marketing manager's position, a level 60 or 61 position. This assumption has no basis in fact. It is apparent from the record that A.A.s were then and are now routinely paid within the salary range of the level they previously held and that this compensation rate was acceptable because of the valuable experience the job provided to the employee. Hunter vehemently asserted that he was paid "below minimum" as an A.A. However, McKittrick gave Hunter a 13.6% pay increase upon his elevation to the job of A.A. The raise was consistent with McKittrick's treatment of past A.A.s and brought Hunter within the appropriate range of compensation for a level 59 employee.

Hunter maintained that the guidelines call for a 20% increase because he received a "1" rating as a marketing manager. This evaluation, however, was not made until after Hunter had accepted the A.A. job and had received the raise.

With respect to promotion, George Hunter had a desire almost amounting to an obsession to attain the position of branch manager. Administrative assistant to a corporate vice-president was very often the springboard to the branch manager position. It was Hunter's dissatisfaction over his failure to attain a branch manager's job that triggered a deterioration in the employee-company relationship and ultimately led to this lawsuit.

IBM utilizes an elaborate system of identifying and developing candidates for promotion. One part of the system is the Employee Development Planning program. Employees are given the opportunity to discuss their career aspirations and development needs with their immediate managers who then prepare a written individual development plan for the employee. It is the manager's primary responsibility, when he believes a subordinate is ready to assume a position of greater responsibilities and/or would benefit from transfer, to locate a suitable opening and alert the hiring manager of the employee's candidacy. Another program is Executive Resources, which identifies possible replacements for each incumbent executive (level 63 and above). Executive Resources also calls for the identification of three groups of high potential employees: (1) those with high management potential ("HMP"), minorities with high potential ("MHP"), and women with high potential ("WHP"). Individual development plans were prepared for each person in Executive Resources, covering each employee's strengths and development needs, an assessment of the employee's potential, and projections as to possible future positions.

Initially, McKittrick was satisfied with Hunter's performance as an A.A. Within the framework of Hunter's individual development plan ("IDP"), McKittrick assessed Hunter as having a "B" potential and evaluated his performance in May, 1974, as a "2" He estimated that Hunter should be ready to compete for a branch manager position in six months. At one point later in 1974, however, McKittrick did question Hunter about his objectivity in investigating charges against a black manager. McKittrick testified that he did so only after complaints from other employees. McKittrick also noticed some significant deficiencies in Hunter's management capabilities, specifically a tendency to over-delegate and an avoidance of analytical tasks. In an effort to address these weaknesses, McKittrick arranged for Hunter to participate in IBM's corporate and division middle management school programs.

McKittrick, nevertheless, supported Hunter in his candidacy for a branch manager position. In December of 1974, he arranged for an interview between Hunter and Ed Lucente, Regional Manager in Boston, for a position as manager of the IBM "location" in Providence, Rhode Island, which had recently been downgraded to a location from a branch during reorganization. Backed by McKittrick's recommendation, Hunter was offered the job. He rejected it because he felt that the position was inferior to that of branch manager, the job for which he felt qualified and was convinced he deserved.

Within IBM's field hierarchy, the location manager reports directly to the branch manager. The branch manager reports to the regional manager. However, there is little other difference between branch and location managers. Both are level 60 positions in second-line management with the same career development opportunities.

McKittrick asked Hunter to reconsider his decision. McKittrick believed the position to be the best one available at the time and was very much aware of the shrinking number of branches due to IBM's reorganization. McKittrick viewed the location manager position as a positive step for Hunter and believed that the opportunity to work under Lucente would be extremely beneficial. McKittrick testified that he equated the Providence location to a medium-sized branch office and that Hunter was in effect exalting form over substance in his refusal to take the job. There were 40 employees at the Providence location, which is more than at some medium sized branch offices.

The number of DPD branch offices was substantially reduced by the 1974 reorganization while locations significantly increased. From 1973-76, branches decreased from 219 to 153 and locations increased from 15 to 28.

Hunter testified that he was also troubled by the future of the Providence location — the indications were that it was to be phased out altogether. Lucente testified, however, that he told Hunter that phase out was only one of three options. The other two were to maintain its location status or to return it to the status of a branch. The latter was his preference. He planned to review its status within the next year. Lucente told Hunter that although Hunter would not report directly to him, he would personally monitor his performance, salary, and career development. Hunter was to be included in all branch manager meetings. He told Hunter that he would be "Mr. IBM" in Rhode Island. Lucente also assured Hunter that in the event that the Providence office was closed, a significant position on regional staff would be available until a branch manager position opened up. However, during the course of the meeting it became clear to Lucente that Hunter was not interested in the position and that Hunter had so concluded prior to the meeting.

Hunter refused to reconsider his decision. He believed he was overqualified for a location manager position, and that he deserved to be a branch manager particularly because of his impression that all of his predecessors as A.A.s had made such a move. Ironically, after Hunter's rejection of the position, the Providence office again became a medium-sized branch in September, 1975, after Lucente's scheduled review. According to McKittrick, Hunter's performance declined after the Providence decision, and he lost his drive.

Although many A.A.s did become branch managers, Hunter was incorrect in his assumption that all of McKittrick's A.A.s had been made branch managers. Grabe had not. He also failed to consider that many of the A.A.'s had been made branch managers prior to the company's reorganization which reduced available branch manager opportunities.

The individual who took the location manager job after Hunter's rejection of it, J.A. Lopiano, became branch manager. He earned over $50,000 in 1975, which Hunter would have earned had he accepted the job, while Hunter who remained as A.A. earned $33,834.

In February, 1975, McKittrick was promoted and David Fine succeeded him. Hunter remained as Fine's A.A. Fine is currently an independent consultant and was an extremely candid, credible witness. It was Fine who had originally recommended Hunter for the position of A.A.

Fine met with Hunter to discuss Hunter's career developments. Hunter had a number of grievances: he had been an A.A. for too long, his career had not progressed in the manner he had expected, he was not being paid enough, he had not been appraised by McKittrick, and he wanted to be a branch manager. Fine immediately asked McKittrick to appraise Hunter formally, but McKittrick told Fine that he felt it would be detrimental to Hunter for him then to evaluate Hunter formally. McKittrick reported to Fine that Hunter had been an inconsistent performer and that his most recent performance rating and morale were low. McKittrick, however, did assess Hunter's performance as a low "2" to support a salary increase. McKittrick testified that Hunter's performance was really that of a "3" level at that time.

Fine testified that if Hunter was below minimum in salary it did not surprise him because Hunter had been moving up so quickly through the salary levels. Typically, when someone is moving up fast, he will start at the bottom of the range of salaries of a particular level.

Fine found Hunter to be erratic in his performance, and testified that Hunter even admitted that his attitude was poor. Fine found that Hunter appeared preoccupied with the frustrations arising out of his apparent inability to be promoted to branch manager. Although Fine actively investigated the possibility of a branch manager position for Hunter, he frankly told Hunter that the probabilities were low because of the tight business environment and the reduction of the number of branches arising from the reorganization. Fine had several counseling sessions in an attempt to motivate Hunter. Although he assured Hunter that he still had the potential to become a branch manager, he cautioned Hunter that he would have to perform consistently at a high level if he expected to obtain the job.

For many of the same reasons assigned by McKittrick, Fine also never formalized a performance plan for Hunter. He was confident that Hunter was well aware of his duties as A.A. There was no testimony elicited at trial that indicated that Hunter was treated differently from any other A.A. From July to August, 1975, Fine offered Hunter promotion to a level 61 position as industry marketing manager in the finance industry organization in Princeton, New Jersey. Hunter refused to even consider this offer.

The reorganization of IBM in 1975 resulted in the restructuring of DPD's industry marketing organization. Two new groups were established, one to be responsible for market planning which Fine was to manage, and the other to be responsible for marketing support to the field which H.F. (Bud) Enright would manage. IBM was placing particular emphasis on the reduction of headquarter staff and expenses and the increase in sales. This meant that headquarters staff would be reduced with many managers and professionals being required to return to the field.

Fine and Enright decided that Hunter was best suited to manage the new public sector industry marketing ("PSIM") procurement support organization within the marketing support group. Hunter had knowledge of the Federal procurement process and his background as administrative assistant would be helpful. Hunter first declined the position, but Fine eventually talked him into taking it. Fine told Hunter that he had little alternative because Fine's old position had been abolished as was the A.A. job. Hunter began as procurement support manager, a level 60 position, on September 1, 1975, and reported directly to R.G. Richmond, the new Public Sector Industry Marketing manager.

In his new position, Hunter's responsibility was to develop a procurement support function. It was Hunter's belief that Richmond assigned a co-worker, Jerry Addicott, to perform the same function. It is unclear to what extent there was overlap between the work committed to Addicott and Hunter's assignment. It appears, however, that management was simply taking advantage of Addicott's experience in state and local government and his familiarity with the issues. No discriminatory or malicious motive can be gleaned from any possible similarity of assignment. It is clear that Hunter disagreed with Richmond over the role that he and Addicott should play in developing the strategy and organization of the procurement support function. Hunter went off to work on his own plan which focused on a centralized organization and utilized twelve government marketing advisors ("GMAs") who would report directly to Hunter. Richmond's task force recommended instead a decentralized organization of GMAs. This conclusion was presented at Richmond's request by Addicott to Enright in White Plains, New York on November 11, 1975. Richmond directed Hunter to attend also, but he declined to do so. Instead, Hunter wired Enright with his own plans.

Enright was concerned when Hunter did not attend the meeting and it was apparent to him that Hunter's work attitude was deteriorating rapidly. Enright summoned Hunter to White Plains for a face-to-face meeting. Hunter testified that at this meeting, Enright accused him of being insubordinate and told him that he would be driven out of the business. Enright disputes this account, recalling that he and Hunter discussed what it was like to be black in an organization such as IBM. Enright, who now works for Prime Computer, was adamant that he had never told Hunter that he would drive him out of the business. He did acknowledge that Hunter complained that Enright and Richmond had misled him.

Enright recalled that McKittrick initially felt that he had made a "brilliant" choice when Hunter was elevated to the A.A. job. Enright observed, however, that Hunter began to misuse the power of the office and that he lacked objectivity. He became what Enright termed a "surrogate equal opportunity manager." Enright testified that the procurement support job assumed by Hunter was the most important position in the sector other than that held by Randy Richmond. Nevertheless, Hunter was unhappy because his sights had been set on a branch manager position.

Hunter's dissatisfaction culminated in an "open door" letter to Frank T. Cary, Chairman of IBM's Board of Directors, asking that the Chairman investigate a number of Hunter's concerns. Hunter sought redress for being placed in what he termed a "meaningless" position instead of promoting him to branch manager. He also questioned IBM's commitment to moving blacks into higher management positions. The complaint was initially referred to J.F. Akers, DPD president, for investigation, but Hunter subsequently requested that the investigation be centered in the Chairman's office. With Hunter's consent, George Youngdale, director of IBM's Resident Manager program, was named as the investigator on behalf of the Chairman.

The open door policy is one of several safeguards built into IBM's Performance Planning, Counseling and Evaluation (PPC E) program. It is one of the methods by which employees bring grievances to the attention of upper level management. According to Frank Cary, IBM averaged between 800 to 1,000 open doors a year. In theory, an employee may jump over the head of his next line manager(s) all the way to the Chairman of the Board in requesting investigation of a problem.

On November 25, 1975, Youngdale met with Hunter for more than four hours at Hunter's office in Bethesda. At that meeting, Youngdale reviewed Hunter's complaints in detail. Youngdale made the following findings based on interviews with ten persons concerning Hunter's complaint: Hunter had not been discriminated against because of his race, he had been fairly paid by McKittrick, he had not been unfairly denied a branch manager position, he had been offered two promotional opportunities which he had chosen to reject, he was still a viable candidate for a branch manager position, and his present job was meaningful and offered him an opportunity to succeed in IBM.

Youngdale met with Hunter on December 18, 1975, to review his investigation, conclusions, and recommendations. Hunter did not believe that the investigation was thorough and asked Youngdale to answer additional questions and make further findings. Youngdale did so and reported back to Hunter in January of 1976. At that time he informed Hunter that Division President Akers wished to meet with him. Hunter met with Akers for an hour in White Plains and discussed, according to Akers, the Providence job offer, Akers belief and commitment to IBM's equal opportunity program, the process by which IBM selects branch managers, and Hunter's concerns about the job he then held. Hunter testified that Akers used obscene language, impliedly directed at him, when discussing the failure of someone to take advantage of a location manager position as a stepping stone to a job as branch manager. Hunter also reported that Akers used obscene language in maintaining that none of his managers discriminated. Akers flatly denied using such language. The Court is doubtful that a person in Akers' position would exhibit such extreme bad judgment as to direct obscene language at a disgruntled employee during a counseling session. The Court accepts the Akers' testimony, and expressly finds that there was no such verbal conduct on Akers' part during the interview.

Akers directed his own investigator, Jack Quinn, to look into the viability of Hunter's job as well as his future. Quinn's written report concluded that it was reasonable to offer the job to Hunter, it was a viable position, Hunter could still obtain a branch manager position, and there were alternative positions available if Hunter wished transfer. Youngdale reviewed the conclusions of this study with Hunter on January 30, 1976. At that point Hunter indicated that his working relationship with Richmond and Enright made the job untenable even though he acknowledged that the position was viable. Youngdale indicated to Hunter that there were other jobs available, but Hunter wished to await the Chairman's report. On February 6, 1976, Cary sent a letter to Hunter, informing Hunter that he believed that he had been fairly treated.

Hunter remained in the procurement support job under the supervision of Randy Richmond. Richmond counseled Hunter about his lack of leadership and negative attitude. By this point in time, Hunter had allowed his dissatisfaction to affect his work and was not fully carrying out his duties. Richmond gave Hunter an informal "3" rating — "exceeds at times" — on May 26, 1976. Although Hunter conceded that his attitude was a problem, he did not agree with the evaluation. In July, 1976, Richmond resigned from IBM and was replaced as manager of PSIM by T.E. Hobbs.

Without informing Hobbs, Hunter submitted his charge of discrimination to the EEOC on August 16, 1976. Hobbs did not learn of this charge until IBM was officially notified in November. Hobbs and Hunter had several informal discussions about Hunter's dissatisfaction and attempted to start anew. Hobbs followed up on Hunter's interest in a branch manager position and discovered that Hunter was on the branch manager candidate list, but because of his performance was not being considered a "highly recommended" candidate. Hobbs relayed this information to Hunter.

T.R. Lautenbach, DPD president, assigned M.J. Mullin to meet with Hunter and discuss Hunter's charges of discrimination. Lautenbach in the past sought to communicate directly to employees who had filed discrimination charges. Mullin met with Hunter for three hours on February 2, 1977 at the office of the IBM resident manager in Arlington, Virginia. Hunter testified that Mullin gave him the impression that filing the EEOC charge had adversely affected his career and that he really did not have a future in the corporation. Mullin's notes and his testimony, however, indicated that he assured Hunter that the filing of the EEOC charge would not affect his career and that if he performed at a consistently high level, he could still become a branch manager. He did inform him, however, that IBM believed that it had not discriminated against him and that he would not be denied promotion simply because he had filed the charge. On April 5, 1977, Hunter resigned from IBM, effective April 30, 1977. Hunter testified to a general feeling of despair and helplessness that led to his resignation.

B. Class Witnesses

The plaintiff called nine class member witnesses in addition to George Hunter. These witnesses offered anecdotal testimony concerning their experiences with IBM which was intended to support the allegations in the complaint. Each of these nine persons (except as hereinafter mentioned) were either a manager or a professional in DPD Maryland sometime between December 18, 1975 and December 31, 1981. In addition, the EEOC called two black IBM employees, Kenneth W. Branch and Raymond J. Pitts, who had never worked for DPD in Maryland.

Plaintiff originally listed 39 individuals as "potential class members witnesses."

The Court initially notes that much of the testimony of these witnesses was both irrelevant to the issues in the case and was grossly repetitive. The Court will summarize, seriatum in the order that they appeared at trial, those aspects of the anecdotal testimony presented by class members which is of even marginal relevance.

1. Frank Sartor

Frank Sartor began working for DPD in January, 1969 in the systems engineering career path, but became a sales employee on January 1, 1970. On September 1, 1974, he was promoted to a Level 56 professional position as a marketing program administrator in Maryland. He remained a professional in Maryland until December 31, 1981, when IBM reorganized. At that time he was transferred to the National Accounts Division ("NAD").

Sartor indicated dissatisfaction with IBM beginning with his very first starting salary. During the course of his employment, Sartor worked under two black managers, in addition to a number of white managers. He testified that several of the appraisals that he received did not accurately reflect his performance. He believed that race played a role in certain appraisals and that he was not allowed to perform up to his potential because he received assignments of little importance.

All of the ratings received by Sartor were either "2"s or "3"s. Although he consistently felt that his ratings were low, Sartor did not point to any specific instances of bias. It is difficult to discern any discriminatory pattern from the ratings because one black manager rated Sartor a "2" and another rated him a "3" while similarly, some white managers rated Sartor a "2" while others rated him a "3".

Sartor did interview for at least two management jobs, one in Syracuse and one in Houston, but was not promoted. He listed the names of other white co-employees who were promoted to managerial positions. On cross-examination, it was established that Sartor limited his options because he wished to stay in the Baltimore-Washington area. He also apparently expressed reservations about taking other managerial jobs because he was under the impression that he might lose money. Sartor admitted that not all whites with whom he had worked had become marketing managers and that he knew blacks who had attained that position. Indeed, a number of marketing and SE manager positions in the Baltimore-Washington area were filled by blacks beginning in 1976. Sartor was unable to offer any concrete reasons for believing that he was not considered for these positions because of his race.

IBM called Richard Klett, one of Sartor's managers, to respond to Sartor's testimony. He explained his reasons for appraising Sartor a "2" instead of a "1". There were several business losses, in Klett's opinion, that Sartor could have dealt with better. Sartor made significantly fewer branch/customer visits than other members of Klett's team. Klett testified that Sartor maintained a relatively low visibility with branch managers and that he needed to improve his skills in dealing with people. Klett did, however, arrange for Sartor's marketing manager interview in Houston and endorsed him as a viable candidate, even though other candidates were more highly qualified in a technical sense. Overall, Klett did not consider Sartor to be a highly competitive candidate for a marketing manager's job.

Klett also testified that about half of the approximately dozen candidates that he sponsored for marketing manager did not attain that position. Almost all of those who did not were white.

Sartor testified in rebuttal and, although his testimony differed significantly from that given during direct examination, he failed to give any concrete examples as to how he was treated less fairly due to race. On direct, Sartor had testified that Klett failed to explain the reasons for his ratings, but on rebuttal, Sartor admitted to many conversations with Klett as well as a formal appraisal session. Sartor provided the Court no basis for finding that any of his appraisals were in any way based on race.

The Court makes a general finding that although Sartor was displeased with some of his managers and with some of his ratings, there was no evidence presented that in any way indicated race as a factor. His ratings, administered by both blacks and whites, were consistent throughout his career, and valid reasons for less than exceptional ratings were provided. Although Sartor did not obtain the marketing manager position to which he aspired, he was promoted consistent with his capabilities in the opinion of his managers. As of 1979, he was earning almost $36,000 a year as a senior marketing support representative.

2. David H. Swann

David Swann joined IBM in November, 1968, as an Assistant SE trainee assigned to a DPD office in Washington, D.C. In February, 1969, he transferred to a Maryland facility. He was DPD professional or manager from February 1969 through the IBM reorganization of January, 1982, when he was transferred to NAD. In March, 1982, he was promoted to his present position in NMD (National Marketing Division).

Swann indicated dissatisfaction with IBM dating back to 1970 when his manager appraised him as a "4" performer. Swann attributed this appraisal at least in part to his inability to work more than "normal" working hours because of a personal family problem. Swann implied racial bias on the part of his manager and indicated that he disliked his "bush" hairstyle.

George Anderson, a black, was Swann's next manager. Swann initially received a "3" rating from Anderson and subsequently a "2" rating. He received various awards and commendations and on January 16, 1976, was promoted to a level 57 specialist position as an advisory marketing support representative, a position which he had actively sought.

Swann received a "3" appraisal in September, 1977, and although initially dissatisfied, he testified that it was a fair appraisal once his manager added comments to it upon his request. His closeout appraisal from Anderson, a "2", was satisfactory.

On August 1, 1978, IBM promoted Swann to systems engineering manager ("SEM"). Swann had three white managers while an SEM: Don Swedenborg, Jeff Samuelson and David Nichols. Swann received "3"s from both Swedenborg and Samuelson, but did not consider them unfair. Nichols gave Swann a "2" appraisal in March, 1981, and a "3" appraisal in April, 1982 for the period ending March 1, 1982, when Swann was promoted. Swann took issue with this last appraisal. Swann believed that Nichols' attitude toward him shifted dramatically because of his involvement with the Black Worker's Alliance ("BWA").

The April, 1982 appraisal, with which Swann disagreed, contained the following subratings: "2" in marketing, "2" in equal opportunity, "3" in customer satisfaction, "3" in personnel, and a "4" on general management. On cross-examination, however, Swann acknowledged that most of the statements made in support of the ratings were accurate. He admitted to receiving an unfavorable customer satisfaction survey. Nichols testified that his assessment of Swann changed significantly after one of Swann's employees, Wayne Priesman, "open doored" to Rooker's administrative assistant, complaining about Swann's poor management of him and about unfair promotion delays. After investigating the situation, Nichols concluded that Priesman had been mismanaged and discussed the problem with Swann. Another of Swann's employees, Lu Nguyen, came to Nichols with similar complaints.

As to Swann's feeling that his membership in the Black Worker's Alliance was a liability to his IBM career, on cross-examination Swann admitted that Nichols did not criticize him for attending BWA meetings, never told him not to attend such meetings, and never expressed any concern about Swann's BWA membership. There was no corroboration of Swann's claim that Nichols was greatly influenced by Swann's membership in the BWA.

Swann interviewed for a number of jobs while he was under Nichols' management. Not all of them interested him and he received two job offers, one of which he accepted. During one interview in Dallas, a manager made a racially insensitive remark which Swann reported to Nichols. Nichols investigated it and apologized for the manager's statement. That position was one of several that Swann elected not to pursue.

Swann testified that although he expressly wished a promotion to a level 59 position, some of the job interviews arranged by Nichols were for level 58 positions. However, Nichols testified that he discussed fully with Swann that while they would seek a job that would "technically" be a promotion, they should also look at the level 58 jobs that are available. In rebuttal to Nichols, Swann did not dispute his testimony that "a good career step might not always be a level promotion and it is very often not, as a matter of fact." During their initial promotion discussions, Swann did not tell Nichols that he would not consider a level 58 job.

Swann never made a claim that Nichols tried to force him to take a level 58 job and, indeed, the only job Nichols tried to persuade him to accept was a level 59 job in White Plains, which Swann turned down. At no time did Swann allege that Nichols treated him differently than others under his management because of his race.

Swann did point to a level 60 job, that of branch marketing support manager, which opened up in April 1982, after he had been promoted. Francine Yaker, another SCM in the branch, moved into this position. Swann maintained that in August, 1981, he and Nichols had discussed the possibility that this position would open up, and Nichols had assured him that neither he nor anyone else in the branch would get the job because it was two levels higher and required general staff experience. Swann testified that Yaker lacked the staff experience and was not qualified.

Nichols testified that he never told Swann that he nor anyone else in the branch would be a candidate for the job. Nichols chose Yaker for the job because she had been an outstanding SE in two different branches, had held regional staff positions in two different regions, had been a headquarter staff manager twice and had been an outstanding SEM. Nichols did not consider Swann because he had just been promoted and would not otherwise have considered him because of a perceived lack of leadership skills. Indeed, he and Swann had already discussed that Swann's next job should not be in management. In summary, the Court finds that the "3" appraisal about which Swann complained had sufficient business justifications, and was not a product of race or Swann's BWA activities. In addition, Swann was given significant promotional interviews and opportunities consistent with his experience and performance.

3. Milton J. Edmonds

Milton Edmonds was one of the co-founders of the BWA and was its first chairman. He testified that blacks were at a disadvantage because (1) IBM did not recognize contributions to black-oriented fund drives, (2) blacks were not placed on visible accounts, (3) blacks dress differently, and (4) the location of plant facilities was not convenient to blacks. Edmonds felt that he was harrassed by IBM management because of his involvement in BWA. He testified that Chuck McKittrick advised him that he could lose his job if his BWA involvement continued, it being inappropriate for a manager to be active in a union.

Edmonds discussed his position as level 57 manager in DPD's products scheduling department in Maryland under the management of J.W. Shelton. Edmonds made varied assertions that he thought the salary increase he received should have been higher and that he had received three appraisals that were attributable to his race. At his deposition, however, Edmonds had stated that Shelton had not treated him unfairly because of his race. While he did not contend that his appraising manager was a racist, he pointed to the cumulative effect of the racist policies of the company as responsible for his lower rating.

Other than conclusory statements, plaintiff presented no evidence that would support a finding that Edmonds received any performance evaluation under Shelton that was unfairly affected by his race. Neither was there any evidence that Edmonds was denied any promotion or paid unfairly because of his race during that time.

4. Betty J. Watkins

Betty Watkins began working for IBM on December 3, 1973, as a general clerk in the federal systems division in Virginia. From May 1978 until March 1982 Watkins worked as a professional in Bethesda, Maryland, first for DPD and after the 1982 reorganization, for the informational services group.

Watkins expressed dissatisfaction with numerous appraisals that she received during her IBM career. She exhibited, however, a general lack of understanding for the way IBM measured their employees against performance plans. While employed by DPD in Maryland, Watkins worked for two managers, Chuck Rosenberry, a white, and Ray Mosely, a black. She was unhappy with her treatment by both of these managers, but admitted on cross-examination that she had no basis for believing that it was due to her race.

Watkins questioned the subjectivity of the rating process and eventually wrote an open door letter because "she had a very uneasy feeling about management." She eventually sought a transfer because of her unhappiness. Watkins was promoted and rated a "2" before being transferred to the location of her choice.

Watkins made some general allegations that white employees were getting more assistance than blacks. She testified that she did not receive sufficient help because the black staff programmer assigned to assist her was very busy. She also testified that she felt physically isolated from other members of her department because their offices were on the other side of the building.

Mosely, a black, testified that Watkins' office location, which was only four doors from his was determined by the space available when she joined his department.

The Court finds Watkins' testimony to be without credibility and lacking any basis in fact. Her testimony was generally conclusory and vague and every specific point she made was refuted by cross-examination and IBM's rebuttal. It was established on cross-examination that Mrs. Watkins and her husband, who also works for IBM, had been relocated at the expense of IBM several times at their request. There is absolutely no evidence that Watkins was treated unfairly in any respect because of her race while she worked in DPD Maryland.

5. Russell T. Crockett

Russell Crockett was a professional DPD employee in Maryland from his date of hire, June 9, 1969, until July 1, 1974, when he became a sales employee, and again from May 1, 1978 until August 1, 1979, when he returned to sales.

Crockett testified that his career goal was to become a marketing manager. At one point he was promoted to a professional position as a marketing support-procurement administrator, a stepping stone for marketing manager. He complained, however, that this promotion prevented him from earning bonuses from a significant sale he had recently made to the Air Force Data Services Center. He was advised by Brad Bryan, his second level manager, that if he wanted to be a marketing manager, he should capitalize on the Air Force sale by accepting the promotion.

Crockett hesitated to make the move because, while the initial sale to the Air Force represented four million dollars, the projected revenue from the total sale package was another six million dollars, which would result in further substantial bonuses. Crockett believed that he should remain in his sales position for another two and one-half years to obtain these bonuses. Crockett testified that Bryan pushed his move to national federal marketing staff because Bryan wished his removal from the Air Force account. He was subsequently interviewed by John Adourian for the position of marketing support administrator and encouraged by Fred Harllee, a black EOP representative, to accept the position. When Adourian agreed that Crockett should pursue a marketing manager career path and that the job would be at a level 57, Crockett accepted the job.

On December 15, 1978, Adourian gave Crockett a "six-month" "3" appraisal with which Crockett agreed. This appraisal was given shortly before Adourian left for a new IBM assignment and after his departure, Crockett's unit had no manager for about five months. Consequently, Crockett did not have a performance plan during this time, but he did not bring this to anyone's attention. It was not established whether others in the unit had performance plans in place at the time. Adourian was eventually replaced by Don Dean who became Crockett's manager.

One of Crockett's primary complaints was that Dean appointed Leah Solat, a white female who was also in Crockett's group, to be the defense branch marketing procurement manager. She had joined the procurement staff after Crockett. Crockett believed that from January, 1979, to the time of her appointment, considerable direction, attention, and exposure was given to Solat which was not accorded him. Although the appointment would be a lateral move, Crockett considered it desirable because it was a high visibility position. Crockett complained that Solat was sent to White Plains by herself to make business presentations, while he was not.

It was Crockett's belief that Solat had no marketing or procurement background and had worked primarily in the scientific/social branch, not the defense branch. Crockett, on the other hand, had been in marketing since 1974 and had prior experience with the defense industry. Dean testified that Solat was a much better performer and had much greater potential than Crockett. Solat had only a few months less experience as a procurement administrator than Crockett. David Nichols, Solat's branch manager when she was in SEM, testified that Solat was a "very good performer" and that she had the initiative, ambition, aggressive leadership skills, and people management skills that could carry her into several additional levels of management beyond the SE manager job.

Subsequent to the Solat appointment, Crockett decided to return to the field in a sales capacity. He concluded that Bryan had been very dishonest with him and that he had pressured him to accept the staff position to get him off the Air Force account.

Crockett, however, provided no support for a finding that Bryan was motivated by race in promoting Crockett to a staff position. On cross-examination, Crockett agreed that the best time to make a move to management was after a significant sale. He admitted that he did not feel that Bryan was forcing him to move to the staff position at the time of the promotion.

The Court finds that Crockett may have lacked direction for the five-month period after Adourian left as his manager. This situation was unfortunate and may have hampered Crockett's movement along his career path. There is no evidence, however, that others did not suffer from the same lack of direction or that this was the result of racial discrimination. IBM and any similarly structured company which has a policy of promotion from within would, at any given time, have vacancies which might cause some employees to suffer from a lack of proper management. Certainly there is no way in which a racial connotation can be put on such occasional management shortcomings.

Plaintiff fails to support its proposed finding that Crockett was promoted to a staff position as part of a scheme to discriminate against him because of his race. IBM offered a perfectly valid business reason for his promotion — that of trying to help him attain his goal of becoming a marketing manager. Although Crockett did not ultimately become a marketing manager, it was his choice to return to sales. Crockett's disappointment at not attaining Leah Solat's position is understandable, but IBM presented the testimony of three managers who believed that Leah Solat's selection was appropriate. Plaintiff failed to rebut the testimony that Solat was significantly more effective as a procurement administrator than Crockett. There is absolutely no showing of pretextual treatment.

6. Eugene S. Hughes

Eugene Hughes worked for DPD as a professional or manager from the time he was hired in 1966 until the January, 1982 reorganization.

Hughes harbored a number of complaints against IBM, beginning with the "4" rating he received on his early performance plan in 1973 and 1974. He claimed that many of his performance plans were "overloaded" with unrealistic and unattainable objectives. At trial, Hughes testified that it was his belief that IBM discriminated against him because of his race with respect to every performance evaluation he received during the 17 years of his IBM employment. He felt that he should have received a "1+" on a particular plan even though he had received the highest rating of 1. Hughes filed an EEOC charge against IBM in October, 1978, in which he alleged that his manager, Al Bissell, unlawfully discriminated against him with regard to job assignments and a May, 1978 performance appraisal.

No determination was made by the EEOC on this charge.

Hughes' testimony contradicted his deposition testimony of November, 1982. At his November deposition, Hughes testified that only five of the fourteen appraisals he received reflected racial discrimination on the part of IBM. Three of the appraisals given by Bissell during the relevant period since 1976 fell into the category of those not claimed to be discriminatory. Hughes did, however, change his testimony at a later deposition held in January, 1983. Hughes gave three reasons for changing his testimony: (1) he was surprised by the scope of depositions, (2) defense counsel harassed him; and (3) he did not review the document carefully.

As to Hughes' first explanation, he stated that he expected the deposition examination to be limited to the Bissell appraisal that was the subject of his EEOC charge. However, Hughes brought to the deposition copies of his appraisals dating to the 1960's, his own analysis of his appraisal history back to 1966, and other documents predating 1978. In addition, he had submitted to the EEOC in connection with this charge information relating to his entire IBM career.

As to Hughes' second explanation, he was unable to specify any instance of harassment on the part of the attorneys. The EEOC counsel who was present during the deposition acted in the capacity as advisor to Hughes and did not object to the conduct of the proceedings. Indeed, Hughes was allowed to frequently confer with EEOC counsel. Having observed the defense counsel in question for the length of the laborious forty-day trial, the Court finds it extremely unlikely that this explanation has any basis in fact. Lastly, Hughes was given every opportunity to review the documents during his deposition, even if, as he asserts, he was unable to review the documents in advance. The Court accepts none of Hughes' explanations and can assign little if any weight to his testimony.

In addition to the direct contradiction in his testimony, Hughes' credibility was eroded significantly by cross-examination. He admitted that his assignment to the Resource Management Systems ("RMS") Task Force in early 1977 was a high priority assignment, but in his 1978 EEOC charge, he asserted that he had been denied assignment to high priority work. Hughes made bald accusations of unfairness and constantly attempted to rationalize his performance. His increasing nervousness on the stand was very noticeable, and he began rambling uncontrollably as his claims became more extreme.

Hughes discussed one incident with Bissell that he thought exemplified Bissell's discriminatory attitude towards him. In late May or early June, 1977, Hughes began to develop a survey to be used by IBM health industry specialists around the country to gather the necessary information concerning the needs of hospitals in their areas. Although Hughes apparently consulted with Bissell during the development of this survey and prior to his going on vacation, Hughes presented the survey package to Bissell's secretary for typing and mailing without Bissell's knowledge. The package was mailed out to the industry specialists on June 23, 1977. Hughes returned from vacation the second week of July and discovered that although the package had been mailed, Bissell had notified the industry specialists by telegram that they should not go forward on the survey until they received further word from him.

Bissell testified that he never authorized the release of the survey to the field industry specialists. When Bissell returned from vacation, he learned that Hughes had sent it out in his absence. In telling health industry specialists not to proceed on this survey, he was responding to their complaints that it would require an enormous amount of work to complete. Hughes admitted at trial that it would take three to five days to complete the initial survey. Bissell told Hughes that the health industry specialists could not be asked to undertake elaborate surveys without first contacting their managers.

In addition to Hughes' lack of credibility, there also was no showing that Hughes was treated less favorably than whites either under Bissell or any other managers. On the RMS project about which Hughes had several complaints, he received a "3" rating from Bissell, while the team leader, a white, received a "4" because a higher level of performance was expected from him. Only three employees in health industry have received "4" ratings from Bissell and all three were white. In addition to Hughes, Bissell has managed three other blacks, Annette Clayborne, Irving Justice, and Jacqueline Cole. Bissell rated Clayborne an initial "3" and then mostly "2"s, Justice a "3" and then a "2", and Jacqueline Cole, a "1". Hughes complained about the lack of promotion, but no employee in Bissell's department has been promoted from level 58, the level at which Hughes joined the department, to a higher level.

In summary, the Court gives little credence to Hughes' testimony and does not find that he was the victim of any racial discrimination with respect to salary, promotions or appraisals. At the time of trial, Hughes was making $50,000 a year with IBM and there was no showing that he was treated any less favorably than similarly situated whites.

7. Kenneth W. Branch

Although Kenneth Branch was listed by plaintiff as one of its class member witnesses, IBM filed a motion in limine to exclude his testimony because he had not been employed by DPD and allegedly had no personal knowledge of facts relevant to the litigation. The Court denied the motion in limine with the admonition that Branch's testimony be restricted to relevant areas. Branch's testimony related to his tenure as equal employment opportunity program ("EEOP") manager at the Federal Systems Division facility in Gaithersburg in 1975 and 1976. There were some DPD employees who were housed at that facility during that period.

The rest of his testimony centered around the paucity of employees and the number of blacks in the Gaithersburg facility. His figures, however, included employees in job categories other than professionals and managers. To the best of his recollection, in 1974 and 1975 none of the six or seven black DPD employees at the facility were at level 58 or above. Nonetheless, it was established that a black, Oscar Bunche, held a level 61 DPD managerial position during that time.

Branch's brief testimony, based on thin recollection of intermingled facts and figures, will support no findings of fact relevant to resolution of the issues in this case. His primary assertion, that no blacks held managerial positions, proved unfounded and there were no other aspects of his testimony to which the Court can affix any particular significance.

8. Jacqueline W. Cole

Plaintiff's next class witness, Jacqueline Cole, was very articulate and straight forward. She is obviously extremely capable and has been an exceptional performer with IBM. Without question, she deserved to be on the fast track that IBM had kept her on during her career. However, given that she is a remarkable performer, IBM has treated her most favorably and indeed her testimony may be taken to support a finding that IBM's appraisal system is effective.

Cole began working for DPD in Maryland in July 1965. During her career she has requested and received transfers on at least two occasions consistent with her husband's movement as a college professor. From the period beginning May, 1972 until the trial of this case, Cole has been accorded forty-four months of leave. On three occasions she has left for maternity leave, and in addition, she was granted fourteen months social service leave to work in the Washington, D.C. government. Cole worked in the DPD facility in Maryland for a total of 31 1/2 months between December 1, 1975 and September 1, 1980, the date she transferred to the Federal Systems Division. She was on leave for the remaining 28 months during that period. Her salary as of the time of trial was $67,500 a year.

The primary focus of her testimony was the evaluation she received. Although she did not directly contend that she was discriminated against because of her race, she felt that certain of her appraisals were low and complained that she had been locked into a level 59 position since 1976. She pointed to a disparity in the per cent of whites rated at "1" or "2" as compared to the per cent of blacks given these same ratings. She also alluded to some minor salary problems, but generally was able to correct those with her manager.

In July, 1976, Cole was promoted to a level 59 marketing manager under Don Dean. Cole was unhappy that Dean did not give her performance plan until December. She did have a performance plan still in place from her prior manager, although Cole maintained that this plan was not effective under Dean. It is not, however, an established IBM practice for a new manager to rewrite an employee performance plan. Cole was unhappy with Dean's initial appraisal of her, a "3", and complained to her second level manager, Lee Noel. She worked within the IBM system and was successful in obtaining a "2" reappraisal. The "3" appraisal never became part of her IBM record and was not used to administer her salary.

At that time she also had a complaint with her initial marketing management bonus ($2,000.00). Dean considered her complaints and increased the bonus to between $2,700 and $2,800, admitted that a mistake had been made, and gave her the formula necessary to calculate a correct amount. Cole could not assign any racial bias as a motive behind this mistake.

Cole next moved to the position of a level 59 product support manager in June, 1978, reporting to Larry deLorimier. deLorimier's first appraisal of Cole was initially a "3", but after her objection, he rewrote the appraisal changing it to a "2". The Court finds that the initial appraisal was a draft, consistent with IBM practice, and after an appropriate employee/management meeting, the appraisal was changed to the satisfaction of the employee. The draft does not become part of an employee's record at IBM and does not inhibit the employee's career. Cole was given a "2" closeout appraisal by deLorimier, and although Cole testified at trial that she disagreed with this rating, no complaint was made at the time.

Cole did speculate that she had suffered repercussions from protesting both the Dean and deLorimier appraisals. This notion was not substantiated or detailed at trial. Neither was there any evidence that Dean or deLorimier appraised similarly situated whites more favorably.

Cole also complained that when she was granted social service leave, deLorimier asked her to sign a letter of understanding stating that she was not guaranteed "the same level or position" when she returned. Cole did not sign this letter until deLorimier had deleted the word "level." deLorimier testified that the letter was simply intended to inform her that in light of potential business changes, a position at a particular job level necessarily could not be guaranteed at the conclusion of leave. deLorimier added that he did not expect Cole to return from social service leave with a promotion because she was not increasing her product skills while she was away.

deLorimier did arrange for some interviews for Cole in preparation for her return to IBM. She had certain restrictions because she wished to remain in the Washington, D.C. metropolitan area and had been out of the main stream of IBM's product line during a period of significant business change. George B. (Rick) Richardson took over as Cole's career manager and advised her of the availability of level 59 jobs in DPD as a marketing manager, but Cole indicated she did not want such a job. After he was unable to find Cole the kind of job she wanted in DPD, Richardson arranged for interviews for her with FSD (Federal Systems Division), and she was offered two positions. She accepted a level 59 position in FSD working for Hal Horowitz in the advanced energy systems area on synthetic fuels. Richardson testified that he was aware of no jobs in DPD above a level 59 that were open in the Washington area at the time.

Although Cole may not have progressed to the level that she would have liked at IBM, it was apparent that if anything held her back it was the amount of time that she spent away from the business. It is difficult to conceive that an employee earning almost $70,000 a year who has been away from the business for almost four years out of the past ten and who continued to be offered challenging positions could have a credible complaint against a company. IBM offered into evidence the leave record of other employees on the Maryland "tape" who have been promoted to either level 60 or higher positions between 1978, when Cole returned from maternity leave, and the end of 1981 — only twelve had any significant amount of leave from IBM and of those twelve, twenty-four months was the most.

The Court finds that while Jackie Cole is a most capable employee who would excel in any endeavor she chose to pursue, there is no evidence that she was discriminated against on the basis of race. She has been liberally treated with respect to leave and has been able to work within the IBM system to correct what she perceived as unfair evaluations. No evidence was presented that whites were treated any more favorably by her manager than was she. Her managers appear to have made every effort to keep her in the mainstream of the business, despite the extraordinary amount of time she spent away from it. If anything, her experience demonstrates the great lengths to which IBM was willing to go to accommodate an exceptional employee.

9. Carl H. Brown

Carl Brown started working with IBM in 1969 in the Federal Systems Division. In 1976, he transferred into DPD as a level 54 program analyst and was assigned to DPD services in Maryland. He remained in Maryland DPD until December 10, 1981, when he was discharged from the company for what IBM termed "unsatisfactory job performance." Subsequently, effective December 14, 1982, Brown was rehired into ISG as a staff programmer analyst, a level 55 position, as a result of an open door investigation.

Brown's early career with IBM was quite successful, marked by steady advancement and predominently "2" ratings. Brown continued to perform well in DPD. In December, 1977, he was promoted into a level 55 position as industry applications specialist under Clete Roh in Public Sector Industry Marketing. Roh initially appraised him a "3" after six months, and then gave him a "2" in the 1979, 1980 and 1981 appraisals. The latter was a closeout appraisal given after Brown had accepted a job as a system engineer in the Washington North Branch office.

Brown was experienced and had developed expertise on IBM's 3790 computer, a predecessor to IBM's 8100 computer. Brown was initially brought into Roh's organization to help support the introduction of the 8100. Once the 8100 became a viable product, however, Roh's work force was reallocated from the 8100 to the health industry area. The 8100 support group was being deemphasized or eliminated. At this point, Roh planned to transfer Brown and three others in the 8100 support group. Roh asked Brown to interview for a position in the advanced communications center as an instructor or a supporter to instructors in the 8100 area. Brown, however, was adamant that he deserved a promotion after three years of steady "2" ratings and refused to interview for the position.

Roh suggested other job possibilities, and Brown interviewed with Allan Berg for a systems engineer position. Brown had never worked as a systems engineer before, but testified that he agreed to interview for the job because Roh indicated that he could try to persuade the branch management to give him a promotion. Berg informed Brown that a promotion was not justified in view of his lack of branch experience, but Brown accepted the job anyway.

It was undisputed that Brown was unqualified for this position. Even though it was a lateral position and he lacked the necessary experience, Brown took the job because he felt it was his best opportunity to obtain a promotion. Because Brown so desperately wanted to be promoted, Berg wrote his performance plan as he would for an advisory systems engineer, the next higher level. Brown testified that Berg explained that if he met all the plan's requirements, his promotion would be justified. Brown found this acceptable and as he stated at trial, "I was willing to go along with anything to get promoted." However, his subsequent evaluation was unsatisfactory "5" — a rating with which he was obviously unhappy. He believed that he was asked to do too much and did not receive adequate training or education for his job. Brown refused to sign the appraisal and did not discuss it with Berg because he saw no point in doing so. He believed that this appraisal was racially biased.

After receiving the "5" rating, Brown was placed on an improvement plan, which is standard procedure at IBM when an employee is performing at an unsatisfactory level. However, another unsatisfactory rating followed and Brown was given the option of resigning or being terminated. He chose to resign and then "open doored" to his regional manager, Dennis Hearon, a black, who upheld the decision to terminate Brown. Brown then "open doored" to the division president who appointed Robert Griffin, a black, to investigate Brown's complaint. Griffin also upheld the termination decision. In December, 1982, Brown finally "open doored" to the Chairman's office. The investigator, Catalano, a white, concluded that Brown should indeed be reinstated because his assignment to the branch as an SE was a mistake. Brown was rehired with backpay to the day he sent his open door letter to the chairman. Catalano did not find that Brown had been treated unfairly because of his race and never told Brown that he had been the victim of racial discrimination.

The Court finds that Carl Brown was so eager for a promotion, that he plunged headlong into a job for which he was unqualified. IBM unquestionably made poor management decisions in placing him in the SE job and then writing a plan which clearly exceeded his talent and ability. Although Brown was pushing hard for a promotion, it made little sense for the company to acquiesce if the employee is lacking experience and qualification. It is apparent that the company realized it had made a mistake and it sought to rectify the error by acting favorably on Brown's third open door. Brown is currently a staff programmer with IBM.

Although the company is certainly partly to blame for many of Brown's difficulties, there was absolutely no showing of race as a motivating factor. IBM presented extensive testimony on its efforts to improve Brown's performance as an SE, although obviously the effort was unsuccessful. The Court finds that IBM's errors in handling Brown's career was due primarily to its efforts to accommodate Brown. There is no factual basis for finding that racial animus played any role whatsoever. Brown's successful pursuit of his open door also showed that the open door process is more than a cathartic exercise or cosmetic remedy.

10. Daniel L. Bailey

Daniel Bailey began working for IBM on September 1, 1966. Since December 1, 1974, Bailey has been a manager for DPD and NAD in Maryland. He is an articulate individual who is obviously highly talented.

Bailey joined IBM in September, 1966, as an SE trainee. He received numerous awards as he moved up through the ranks and obtained eight 100% club awards for meeting his annual sales quotas. He had notable success in the marketing area.

In December, 1974, Bailey received a four level promotion from level 55 marketing representative to a level 59 marketing manager NFM. It was during his tenure in this position that he experienced difficulties with management. Bailey received six ratings as a marketing manager, three 3s and three 2s and only took exception to one "2" rating that he received in 1980 from Dave Nichols. He stated that he had never done a better job and that he had attained 200% of quota during this appraisal period.

On cross-examination, however, Bailey admitted that he did not know whether Nichols was motivated by racial bias, and that there was no evidence that Nichols applied different standards of evaluation to Bailey as compared to white managers. Nichols testified that of the 40 appraisals he gave as NFM manager, only two were "1" ratings. It was Nichols' judgment that Bailey performed at a "1" level in only one area, marketing volumes. Nichols testified at length on other areas of Bailey's performance that were either a "2" level or a "3" level. According to Nichols, Bailey placed great emphasis on securing orders and closing out sales, but did not pay sufficient attention to other areas, such as the details of the sales contracts or his staff work. Although Bailey did "a very fine job" in Nichols' estimation, his work did not reach the level of "far exceeds."

Bailey often complained about his failure to advance beyond the level 59 position he assumed in 1974. He admits, however, that in 1978 his branch manager, Rooker, arranged an interview for a level 60 regional marketing job in California for him, a job which he was offered. Rooker had counseled Bailey on his career development and advised him that he had been identified as an employee with high management potential and that DPD planned to move him rapidly through a series of promotions so that he would have the opportunity to obtain high executive level positions. Rooker arranged for the offer of the Los Angeles regional marketing manager position to Bailey. Rooker explained to Bailey that he might move from that position to branch manager. Bailey, however, declined the position, explaining that he did not wish to relocate because (1) his wife had a career in Washington, D.C. and (2) he did not believe he had yet firmly established his performance record as a marketing manager. Rooker also testified that Bailey was offered a regional marketing manager position in Dallas, which he rejected. Bailey did not recall this offer.

Thereafter, Bailey had difficulty obtaining a new job and most for which he interviewed were lateral positions. Bailey still is in a level 59 position and took his current job for a lack of better opportunities. His current salary is $5,000 a month.

On cross-examination, Bailey could assign no racial bias as the cause for his difficulties. He acknowledged that his decision not to relocate was a significant stumbling block for promotions. He also agreed that Rooker had told him he could achieve a high level in IBM including the "run for the roses," the IBM terminology for the race to be president. Rooker had made it clear that Bailey was on an excellent career path, and that Rooker was trying to move him ahead quickly.

There is no question that Bailey's career in IBM upper level management stalled. The Court finds that the primary reason for Bailey's lack of promotion was his refusal to relocate. In reviewing the progress of countless careers of IBM managers during the course of the forty-four day trial, one aspect of IBM employee dynamics was obvious: as an international firm, one must be willing to relocate in order to maximize one's job opportunities and career potential. Bailey's refusal to do so obviously lost him what had once been Rooker's fervent support.

There is no basis for a finding that Bailey was discriminated against by IBM. An extremely competent employee, Bailey himself limited his opportunities in a company that requires an employee to make sacrifices in attaining career goals. Plaintiff presented no evidence, once IBM had provided business explanations for Bailey's treatment, that any of IBM's decisions with regard to Bailey's evaluation, salary, or promotional opportunities were pretextual in any way.

11. Raymond J. Pitts

Raymond Pitts was the second witness whose testimony IBM sought to exclude as irrelevant because Pitts was not on the EEOC witness list. The Court allowed the testimony, but reserved the right to exclude it from consideration.

Pitts has been a very high achiever at IBM and has only recently, in 1983, encountered difficulty with a low appraisal. The appraisal has been wiped off his record and there is an ongoing open door investigation into the matter. Pitts testified that his career at IBM has been jeopardized because the only record of what he termed an exceptional year's performance is a letter stating that his performance appraisals had been invalidated.

Although Pitts did not agree fully with the resolution of the situation, efforts were still being made at the time of trial to promote Pitts to a desirable position. Resolution of the open door complaint was still pending. The substance of Pitts' complaint also occurred after the discovery cut-off in this action. In addition, no evidence was presented that his difficulties were in any way racially motivated. The Court, therefore, needs not and does not make any findings of fact with regard to Raymond Pitts as respects the merits of this case.

C. The Statistical Case

1. Dr. William E. Sedlacek

Plaintiff called as its statistical expert in its case-in-chief, William E. Sedlacek, Ph.D. Dr. Sedlacek is an Associate Professor at the University of Maryland, having taught courses and published articles in the area of statistics. Dr. Sedlacek has previously testified in four Title VII cases. He does not consider himself an expert in the fields of economics or personnel administration.

Dr. Sedlacek conducted his analysis based on computerized personnel records which contained data for the period 1973 through 1981 for every manager or professional who actively worked at an IBM DPD Maryland facility. The personnel record, referred to as the "Maryland tape" or the "Basic Maryland File," was extracted from IBM's Personnel Data System and contains up to ninety different fields of year end information for every one of the aforementioned DPD employees. The information, however, is not limited to the period that the employee was assigned to a DPD facility as a professional or manager, rather it includes data for the period 1973 through 1981 regardless of the location or job category to which the employee was assigned during that period. Thus, the tape includes data in any given year on employees who were not managers or professionals, not in Maryland, or not in DPD in that year.

The Maryland tape contains data for 2,227 persons who were listed as DPD managerial or professional employees as of any year-end date from 1973 through 1981.

In plaintiff's exhibit 88 and "88 corrected," Dr. Sedlacek compared personnel data on black and white employees who fell within four groups: (1) all employees; (2) employees assigned to position codes having the numbers "04," "14," and "24" as the first two digits (first, second and third line managerial positions including branch, marketing, and SE managers); (3) employees assigned to position codes with the numbers "43" and "44" as the first two digits (system engineering and staff support positions); and (4) employees assigned to position codes having the numbers "54" and "55" as the first two digits (sales, sales support and other staff professional positions).

Dr. Sedlacek performed two types of statistical analyses of the data on the Maryland tape. First, Sedlacek conducted a "t" test which compared mean values for black and white employees for the years 1975 through 1981. The object of his analysis was to determine whether any difference in the separate means calculated for whites and blacks was statistically significant at the .05 confidence level or was so small as to be attributed to chance.

Second, Dr. Sedlacek performed multiple regression analyses runs for different years and groups of employees. In one analysis, Dr. Sedlacek developed a regression equation or formula to predict the salaries of a grouping of white employees based on current appraisals and time in current positions. He then applied the same formula to a corresponding group of black employees utilizing their appraisals and time-in-position data and reached a predicted mean salary for blacks and a standard deviation. He then used a "t" test to determine if there was a statistically significant difference between the predicted and actual mean salaries of blacks. Dr. Sedlacek's other regression analysis attempted to predict salary level rather than salary, also using current appraisal and time in current position data.

a. "t" Test Analysis

On the basis of his "t" test, Dr. Sedlacek testified that he could conclude that the following tended to be lower for blacks than whites for the years 1975 through 1981: (1) current monthly salary, (2) latest salary change amount, (3) prior salary change amount, (4) second prior salary change amount, (5) percent of salary above minimum range, (6) current salary level, (7) EOP salary level, and (8) salary at entry to level. Not all of the mean differences upon which Sedlacek based his findings were statistically significant and some favored blacks. Dr. Sedlacek also found that as compared to whites, blacks generally had higher salaries and salary levels at time of hiring. Findings inconsistent with his ultimate conclusions were that blacks had higher salary levels than whites in 1976, blacks have larger percent salary change in 1979 and 1980, blacks had more frequent salary changes than whites and blacks had a shorter time in months since latest salary change.

Sedlacek also concluded that blacks consistently received performance appraisal ratings that were significantly lower statistically than those received by similarly situated whites. This conclusion did not vary whether appraisal ratings examined included employees' current appraisal ratings, their prior appraisal ratings, their appraisal ratings at time of last salary change on their second prior appraisal ratings.

b. Multiple Regression Analysis

As previously noted, in conducting his regression analyses, Dr. Sedlacek developed an equation that he believed would allow him to predict the current monthly salary of each white included in one of the four groups. Once developed and tested, this equation was applied to black employees with the same time in level to see if they were paid the same as their white counterparts. It was Sedlacek's thesis that if treatment of blacks and whites were equal, there should be no difference in the combination of variables (i.e., current performance appraisal ratings and time in current position) used to determine the current monthly salary of black employees in the same group.

Based on the multiple regression analyses, Dr. Sedlacek testified that he found that the actual monthly earnings of black employees were generally statistically less to a significant degree than monthly earnings predicted by the regression equations. In approximately three-quarters of his analyses, the monthly salary for blacks was less than that for whites by a statistically significant amount.

In addition, Dr. Sedlacek also conducted a multiple regression analyses using race as a dummy or binary variable, assigning a value of "0" to whites and "1" to blacks in the regression equation. He inserted the variables in his equation in such an order as he believed would permit him to look at the effect of each variable separately. His conclusion was that there was a significant relationship between race and both monthly salary and salary level.

There are a number of infirmities in and drawbacks to the Sedlacek analysis. First, the employee groupings devised by Dr. Sedlacek or by EEOC counsel included a substantial number of positions that the Court previously had ruled were not encompassed by the lawsuit. Each grouping includes jobs at the very top and bottom of the managerial hierarchy of IBM. For example, the 04, 41, and 24 "managers" group includes persons whose responsibilities involve the supervision of 10 clerical employees, branch managers who may supervise 100 or more employees, and the Director of National Marketing who has ultimate responsibility for over 700 IBM employees. The 43, 44 grouping combines newly appointed systems engineers with experienced advisory or senior systems engineers as well as a diverse group of employees outside of systems engineering. The 54, 55 grouping combines sales trainees with upper level sales personnel (all of whom are not encompassed in this suit) with a variety of jobs not involving sales. Within each grouping, given the wide range of jobs, one would expect a significant variability in salaries paid.

Dr. Sedlacek testified on direct that he was responsible for creating the groupings and discussed them with EEOC counsel. At his deposition, it was his recollection that EEOC counsel had suggested the groupings.

Second, as previously noted, Dr. Sedlacek did not separate employees on the Maryland tape who at any given year end were with another division other than DPD, in another state other than Maryland, or in some other type of job category besides managerial or professional. Data submitted by IBM indicated that typically more than half the people included in Dr. Sedlacek's analysis for each year were not managers or professionals at DPD facilities in Maryland. Dr. Sedlacek attempted to correct his groupings in presenting plaintiff's exhibit 88A (excluding persons located outside Maryland and those not managers and professionals), 88 corrected (excluding persons for whom data were missing in certain fields and persons with salary levels below 44 or above 69), and 88A corrected (also excluding persons for whom data missing and persons with current salary levels below 44 or above 69). However, Dr. Sedlacek still failed to exclude employees working with other IBM divisions and also excluded some class members and their white peers. There is an enormous difference between the IBM count of class members and white peers and that in each of Dr. Sedlacek's four exhibits. Dr. Sedlacek's groupings were numerically grossly inflated making a statistically significant finding more probable.

The chances of finding statistical significance increase as the numbers become larger.

Third, Dr. Sedlacek failed to take into account a critical factor, seniority, in calculating the mean differences between blacks and whites. On cross-examination, Sedlacek conceded that he would expect salaries to be lower in a group whose seniority was significantly lower. In each of his four studies, Dr. Sedlacek reported that blacks had less IBM seniority than whites in the same grouping and that blacks were younger than whites. Sedlacek's groupings fail to distinguish between relatively new hires and employees with considerable experience and consequently, his comparisons of appraisal ratings across these groupings fail to distinguish between persons new to their current position and those in the position significantly longer. Also, because of the wide range of jobs in each grouping, a "2" or "3" rating at a complex upper level job was compared unfavorably to a "1" at a comparatively simple, lower level job.

The differences in the means, both unfavorable and favorable to blacks, could easily be explained by the considerable difference in seniority and the great disparity in job level within the groupings. It would be irresponsible, in light of these factors for which there is no control, to attribute to racial discrimination the differences in the means.

Further support for alternative explanations for the mean differences lies in the following trends noted in Dr. Sedlacek's analyses. The relative difference between the seniority of blacks and whites has decreased between 1975 and 1981 and as it has, the differences between black and white salary levels and salary amounts have diminished. These declines were demonstrated to be statistically significant. The average amount of time spent in current salary level after promotion increased for both whites and blacks from 1975 to 1981, but more so for whites.

Fourth, Dr. Sedlacek's regression predictions were based on the assumption that salaries and salary levels are determined by one or two independent variables, current appraisal ratings and time in current position. This methodology is utterly unreliable given the wide disparity between job levels and seniority within each grouping. A senior upper level employee with a "3" rating would be predicted to have a lower salary than a relatively new low-level employee with a "2" appraisal as long as the two have been in their current jobs about the same length of time. Indeed, the senior employee would be predicted to be at a lower job level. As discussed infra, IBM's statistical expert factored in salary level, seniority, time in salary level, current appraisal and education all as being essential variables in determining salary and salary level.

Besides seniority and job level, Dr. Sedlacek failed to control for another key factor that may substantially influence salary and salary level: education. An employee with a Ph.D. was not distinguished from one with only a high school education. All three factors are of critical importance. In particular, because IBM has had a strong affirmative action program, many of its increasing percentage of black employees are recent hires relative to whites. Dr. Sedlacek failed to take this factor into account in any way.

In addition, the predictive power of his regression equations, which may be measured by the squared multiple correlation coefficient ("R2"), were extremely low. The vast majority of the R2 values calculated were below .02, indicating that current appraisal and time in position are poor predictors of salary and salary level.

When the independent variables in a regression equation fully explain the outcome or dependent variable, R2 will approach a value of unity or 1.0. When there is little predictive value, R2 approaches 0.

Dr. Sedlacek also presented a "supplementary analysis" which attempted to take into account seniority. It still failed to control for education and job level. Even controlling for just one additional factor, seniority, eliminated half of the instances of statistical significance that Dr. Sedlacek was able to generate with his primary analyses.

None of Dr. Sedlacek's analyses, either the simple mean comparisons or his regression analyses, support a finding of race discrimination. Too many critical factors, seniority, education, and job level, are not controlled and his employee groupings are close to meaningless. Any one of a number of factors may explain his findings, many of which support IBM's position that it has one of the most advanced affirmative action programs in the world.

2. Dr. James L. Medoff

IBM called Dr. James L. Medoff as its expert in the fields of labor economics and econometrics. Dr. Medoff is an Associate Professor at Harvard University, teaching both graduate and undergraduate courses, and conducting research and supervising graduate work in labor economics. The EEOC previously solicited Dr. Medoff's participation in a seminar in his capacity as an expert and leading researcher in labor market issues relevant to employment discrimination.

Dr. Medoff worked with two data bases: the Maryland tape and what will be referred to as the "peers tape." The peers tape contained data relating to each black employee who was a manager or professional at a Maryland DPD facility. During the relevant period for each of these blacks, the tape contained a white peer group who met the following criteria: (1) were hired by IBM in the United States within the period plus or minus six months of the black employee's hiring date; (2) were hired into the same salary level or job; and (3) had the same highest level of educational attainment.

Medoff performed a series of multiple regression analyses on the Maryland tape in which he controlled for race and four other predictors that he regarded as reasonable predictors of an employee's success: (1) salary level, (2) years of IBM service, (3) time in salary level, and (4) level of educational attainment. Dr. Medoff ran a separate regression for each grade level and year in which both whites and blacks were found.

He also performed a "matched pairs" analysis on the Maryland tape. Again, Dr. Medoff used salary level, length of IBM service, time in level and educational attainment, this time to find matches of similar black with white employees. Medoff created a total number of 178 matched pairs.

Based on his peers analysis, the regression analyses, and the matched pairs analysis, Dr. Medoff refuted the EEOC's contention that IBM had discriminated against the class. With respect to promotion, 72 of 79 regression equations found no statistically significant differences between the promotion probabilities of blacks and whites. Of the seven where statistical significance was found, four favored whites and three favored blacks. In assessing the frequency of promotion, Medoff found blacks to have a higher probability of promotion, but the difference was both statistically and in practical terms insignificant. The matched pairs analysis showed a slightly higher, but statistically insignificant probability of promotion for blacks in five of eight years, whites in two and none in the remaining year. Dr. Medoff's peers analysis on promotion found statistically significant differences in four of the nine years tested and in all instances the results favored blacks.

As to salary growth, Dr. Medoff's regression analysis found no statistically significant differences in 73 out of 81 equations, and the remaining eight were split, four favoring whites and four blacks. Over all years and all grades, blacks had a small, statistically insignificant salary growth advantage. Matched pairs revealed no statistically significant salary growth difference.

The peers analysis on salary growth found a statistically significant difference only in 1977, which favored whites by 3.7 percent. Dr. Medoff testified that this one result, in the context of his overall salary growth findings, was not indicative of discriminatory salary advancement for class members.

Dr. Medoff also looked at salary, and of 94 regressions run on blacks and whites in the same grade level, 79 were not statistically significant. Of the remaining 15, 13 regressions found whites with statistically significant higher salaries and two found blacks higher. For all year-ends and grade levels combined, controlling for the same factors — grade level, seniority, time in level and education — the mean salary for blacks was found to be 0.88 percent lower than that for whites, a statistically and in practical terms insignificant amount. The matched pairs analysis only resulted in a statistically significant difference in 1979, favoring blacks. The peers finding on salary yielded significant differences in two of nine years, both favoring blacks.

This percentage is the equivalent of roughly $23 per month when applied to an overall mean salary per month of approximately $2,600.

The Court finds that Dr. Medoff's approach to the statistical aspects of this case was both reasonable and comprehensive. He controlled for the relevant variables which Dr. Sedlacek did not, which could cause or explain statistical differences between outcome variables found for blacks and whites. His ultimate conclusion, based on findings that were rarely statistically significant and often favored blacks, was that the black class members had suffered no disparate treatment. The record supports this conclusion.

3. Dr. Stephan Michelson

As part of its rebuttal case, the EEOC called Dr. Stephan Michelson, an expert in statistics, in an attempt to refute the studies and testimony of Dr. Medoff. Dr. Michelson is the President and Director of Research of Econometric Research, Inc., a company which provides statistical and economic data for litigation. His appearance on behalf of the plaintiff in the current case was controversial because he had met with IBM counsel in 1981 to discuss possible employment as a defense expert. IBM moved to exclude his testimony because he had not been listed as a witness and because he had earlier had access to defense counsel's work product. The Court allowed his testimony after noting that Dr. Michelson's claim that he had no recollection of the 1981 meeting would bear on his credibility and that his testimony would be discounted if it appeared that he used any improperly obtained information.

Dr. Michelson generally stated that Dr. Medoff's report was neither a strong affirmative defense nor was it particularly helpful to plaintiff. He then criticized Dr. Medoff's methods, although he did agree with some of Medoff's approaches.

Dr. Michelson criticized Medoff's regression analyses because they did not have a high explanatory power and were not reflective of the number of persons for whom data were included in the equation. Dr. Michelson also criticized Dr. Medoff's promotion regressions, asserting that regression analysis is inappropriate when the outcome variable (promotion) is dichotomous, i.e., yes or no.

Michelson believed that the equations should be weighted according to the number of persons studied in each regression rather than simply taking the mean of the regression result across each year as Medoff did. Under Medoff's approach, regressions done with few subjects were given the same weight as those with many.

With respect to Michelson's criticism that Medoff's regressions had a low explanatory power, Dr. Medoff had testified that his focus was not on explaining the promotional process within IBM, but rather to ensure that race was not a factor when other relevant variables were controlled. Dr. Medoff acknowledged his relatively low R2 values and stated that a random promotion system, not affected by race, would result in low values. The Court is satisfied that Dr. Medoff's concentration on isolating the effect of race was proper and that it would not be expected that an equation could be devised that would be highly predictive of promotion.

Dr. Michelson measured the explanatory power with a device called an "F" statistic, similar in many respects to the R2 value previously discussed.

As to Dr. Medoff's weighting of his regressions, his statistical method may not give full recognition to the number of subjects in each equation. However, his method is nevertheless statistically acceptable because within a given year the results for one grade level were entirely independent of the results for another grade level. Dr. Michelson was unable to show that weighting by sample size, as he suggested, would have any meaningful effect on Dr. Medoff's conclusions.

Dr. Medoff anticipated Dr. Michelson's criticism of his regression analysis of promotion. Dr. Medoff testified that he used more "esoteric" means to test this dichotomous variable and that it did not result in any different conclusion. Dr. Medoff had also used a simple mean comparison between blacks and whites of probability of promotion within grade levels without significant result. Dr. Michelson could not refute these techniques.

Dr. Michelson thought that Dr. Medoff's matched pairs analyses should be completely ignored because of the small sample size. Dr. Medoff had acknowledged the lesser weight that should be accorded these analyses because of the small size, but stated that it was used as further confirmation of the results of his other methods.

Although Dr. Michelson offered some criticisms of the peers tape, he did little to shake Dr. Medoff's conclusions. He complained that the results of the peers analyses were overly influenced by the experience of recently hired blacks, thereby clouding the treatment of blacks with relatively high seniority. However, Dr. Medoff did analyze the peers tape to see how blacks with relatively high seniority were treated and found that, if anything, they were treated more favorably relative to their white peers than were blacks with little seniority.

The Court did not find the Medoff conclusions to be refuted in any meaningful fashion by the Michelson presentation. Dr. Michelson conceded that it was appropriate to control for seniority and time in level and to conduct studies within grade level. Michelson was unable to show that Medoff's study was conducted in a statistically unreasonable manner and that his analysis was to be preferred.

D. General Findings Concerning IBM

IBM presented extensive evidence pertaining to the strength of the company affirmative action program. It was apparent that the Maryland DPD division had a particularly strong program. Also, the number and percentage of blacks employed by DPD as managers and professionals in Maryland significantly increased between 1974 and 1981.

Managers Professionals Year Total Blacks Total Blacks

1974 213 17 (8.0%) 659 36 (5.5%) 1975 193 21 (10.9%) 683 44 (6.4%) 1976 178 19 (10.7%) 734 44 (6.0%) 1977 196 17 (8.7%) 799 60 (7.5%) 1978 200 21 (10.5%) 907 75 (8.3%) 1979 219 27 (12.3%) 899 78 (8.7%) 1980 226 28 (12.4%) 922 105 (11.4%) 1981 227 31 (13.7%) 975 117 (12.0%)

Defendant's Exhibit 398. The number of black managers and professionals in DPD Maryland nearly tripled (51 to 148) between 1974 and 1981 while the overall number of managers and professionals in Maryland only increased by 38% (872 to 1,205).

Also, IBM offered evidence of black representation in its employee mix relative to other private industry and its competitors. For instance, in 1981 the following was the black representation:

Black Representation Managers Professionals

Office and Computer Industry 2.7% 3.0% Private Industry Nationwide 4.2% 4.3% IBM — All Divisions 6.1% 5.6% IBM — DPD 9.8% 8.0% DPD — Maryland 13.7% 12.0%

Defendant's Exhibit 396.

The federal government has audited IBM more than 600 times to test its compliance with equal opportunity and affirmative action requirements. IBM has never been required to change any personnel policy and always has been found to comply. IBM maintains an affirmative action program in each of its facilities and establishes timetables for increasing minority and female representation in job groups in which there is underutilization.

IBM called Professor Fred K. Foulkes as an expert witness on the subject of personnel management, personnel administration and merit systems. Dr. Foulkes, a Professor of Management Policy and Acting Chairman of the Management Policy Department at the Boston University School of Management, has conducted numerous studies on employers who utilize merit systems. He has researched IBM personnel policies and developed a case study based upon that work.

Dr. Foulkes found IBM's performance evaluation system to be of "very high quality." Although it is impossible to have a system that does not include some subjectivity, he was unable to identify any mechanisms by which managers could discriminate against employees. Safeguards include checks and balances on individual managers (second line management supervision etc.), individual manager training, appeal procedures, and other methods by which employees could express displeasure. Government agencies possess some of the same performance evaluation systems.

Foulkes gave IBM's performance evaluation system an "A" because it is well designed and fits IBM's business objective. He noted that IBM possesses an excellent reputation in the academic community and that other members of the industrial community copy its merit system. He testified that he was unaware of any other performance evaluation system that would provide as reliable and useful information or less opportunity for bias in evaluations.

On cross-examination Foulkes testified that it would not be useful for IBM to engage in job posting for upper level positions. It would be harmful to morale, raising peoples' expectations unnecessarily and would interfere with IBM's ability to place employees in openings for developmental reasons. It also might interfere with IBM's affirmative action goals.

Dr. Foulkes was an extremely credible witness whose testimony was unrefuted. There is little doubt from his testimony and that of numerous other witnesses that IBM is one of the most advanced in its performance evaluation system and its regard for affirmative action.

III. Conclusions of Law

Plaintiff has asserted that IBM has engaged in unlawful employment practices and policies against its black employees at its Maryland Data Processing Division facilities in violation of Title VII, 42 U.S.C. § 2000e-3. Trial was limited to the claims of black professional and managerial employees dating to December 18, 1975, with background material dating to January 1, 1974. Plaintiff asks the Court to find that defendant engaged in a continuing pattern, practice and policy of discrimination against its black managerial and professional employees in the Maryland DPD back to at least July 2, 1965.

An initial issue is whether plaintiff's case rests on a disparate treatment or disparate impact theory. Plaintiff contends that this action is grounded upon both disparate impact and disparate treatment theories. Defendant maintains that plaintiff's claims are based strictly on the disparate treatment theory.

The United States Court of Appeals for the Fourth Circuit discussed the difference between the impact and treatment theories in Stastny v. Southern Bell Telephone Telegraph Co., 628 F.2d 267, 274 n. 10 (4th Cir. 1980):

The disparate impact "pattern or practice" is typically based upon an objective standard, applied evenly and automatically to affected employees: an intelligence or aptitude test, e.g., Griggs v. Duke Power Co., 401 U.S. 424, [ 91 S.Ct. 849, 28 L.Ed.2d 158] (1971); an educational requirement, id; a physical requirement, e.g., Weeks v. Southern Bell Telephone Telegraph Co., 408 F.2d 228 (5th Cir. 1969). . . . On the other hand, the disparate treatment pattern or practice must be one based upon a specific intent to discriminate against an entire group, to treat it as a group less favorably because of sex (or other impermissible reason).

It is quite apparent that plaintiff's case relies upon a disparate treatment theory of discrimination. Plaintiff made no suggestion on the record of a Griggs-type objectively imposed practice having discriminatory disparate impact. See Stastny. The allegations throughout the case were that blacks employed as DPD managers and professionals in Maryland were subjected to less favorable treatment with respect to promotions, pay and appraisals because of their race. This case is in no way akin to the typical discriminatory impact type fact situation where, for example, a height requirement that adversely impacts females is evenly applied to all job applicants. The claims before the Court relate to a pattern and practice of intentional discrimination against an entire group by treating it less favorably because of race. See EEOC v. Federal Reserve Bank of Richmond, 698 F.2d 633 (4th Cir. 1983). The Court notes, however, that the result would not be materially different whether the class action was considered a disparate impact or a disparate treatment case.

As first set forth by the Supreme Court in McDonnell Douglas Corp. v. Green, 411 U.S. 792, 93 S.Ct. 1817, 36 L.Ed.2d 668 (1973), there is a specific allocation of burdens in a Title VII action. The complainant must carry the initial burden under the statute of establishing by a preponderance of the evidence a prima facie case of racial discrimination. If plaintiff meets this burden, the burden then shifts to the employer to articulate some legitimate, nondiscriminatory reason for its action or inaction. Id. at 802, 93 S.Ct. at 1824. Should the defendant carry this burden, the plaintiff must then have an opportunity to prove by a preponderance of the evidence that the legitimate reasons offered by the defendant were not its true reasons, but were a pretext for discrimination. Id. at 804, 93 S.Ct. at 1825. See also Texas Department of Community Affairs v. Burdine, 450 U.S. 248, 253, 101 S.Ct. 1089, 1093, 67 L.Ed.2d 207 (1981).

The Court's analysis need not extend beyond the initial burden on the plaintiff to prove a prima facie case of discrimination. Because plaintiff alleged a companywide pattern or practice of discrimination, its burden was to prove "more than a mere occurrence of isolated or `accidental' or sporadic discriminatory acts. It had to establish by a preponderance of the evidence that racial discrimination was the company's standard operating procedure — the regular rather than the unusual practice." International Brotherhood of Teamsters v. United States, 431 U.S. 324, 336, 97 S.Ct. 1843, 1854, 52 L.Ed.2d 396 (1977). In a discriminatory treatment case, proof of discriminatory motive is critical, although it can in some instances be inferred from differences in treatment. Id. at 335 n. 15, 97 S.Ct. at 1854 n. 15; see also Arlington Heights v. Metropolitan Housing Dev. Corp., 429 U.S. 252, 265-66, 97 S.Ct. 555, 563, 50 L.Ed.2d 450 (1977). Plaintiff may rely on either direct or circumstantial evidence to sustain its burden of proof. United States Postal Service Board of Governors v. Aikens, 460 U.S. 711, 103 S.Ct. 1478, 1481 n. 3, 75 L.Ed.2d 403 (1983). Statistics alone or in combination with circumstantial evidence or specific incidents of discrimination may establish a prima facie case of disparate treatment. See, e.g. Teamsters, 431 U.S. at 339, 97 S.Ct. at 1856; EEOC v. Federal Reserve Bank, 698 F.2d at 645.

As fully detailed in its extensive findings of fact, the Court finds that plaintiff utterly failed to prove a prima facie case of discriminatory treatment against blacks employed as DPD managers and professionals in Maryland. There was absolutely no evidence of any systemic pattern or practice of discrimination within IBM. Quite to the contrary, even the testimony of the class witnesses was replete with evidence of IBM's active and effective affirmative action program.

Plaintiff first failed to support its claims that IBM discriminated against George Hunter individually by (1) failing or refusing to promote him to branch manager; (2) failing or refusing to pay him wages equal to those earned by similarly situated whites; and (3) constructively discharging him and retaliating against him for filing his EEOC charge of discrimination. The EEOC never adduced any testimony that Hunter was treated unfairly and no racial motive was attributed to his treatment nor could it be inferred.

Hunter unquestionably was on a very fast track in his quest to become a branch manager until he was derailed. Although Hunter was qualified to compete for a branch manager position, the EEOC made no showing of the degree of Hunter's qualifications for any particular branch and that he was rejected from such a branch, as is their burden. See Wright v. National Archives Records Service, 609 F.2d 702, 714 (4th Cir. 1979). Instead, it was presented at trial that to his career detriment, Hunter rejected two significant promotional opportunities, the location manager job in Providence, Rhode Island and the finance industry job in Princeton, New Jersey. Hunter freely admitted that his attitude became poor and that he did not fully perform his administrative assistant duties. Even had Hunter met his burden of proving a prima facie case of discrimination in promotion, IBM presented extensive evidence on the branch manager selection process and fully explained why Hunter did not receive one of the branch manager vacancies filled during the relevant period.

There were eleven medium sized branch manager positions filled. Hunter was never seriously considered for any of these positions because of his poor attitude and performance. Of the eleven, two were filled by blacks who were clearly more qualified than Hunter.

There also was no prima facie case of salary discrimination against George Hunter. IBM showed that his salary was consistently higher than that of his white peers. Hunter as well as EEOC counsel, demonstrated a lack of understanding of the IBM compensation system, and it was fully explained that when an employee moves up the ladder quickly, as did Hunter, it takes some time in a new level before the rate of compensation fully catches up. Indeed, Hunter never complained about the manner in which his salary was administered after December 18, 1975, during the relevant time period. Even if a prima facie case of salary discrimination had been made, IBM detailed every step of Hunter's career, and no inference of any racial animus is possible.

As to Hunter's constructive discharge and retaliation, Hunter's testimony reveals his inability to meet his burden on these claims. To establish constructive discharge under Title VII, "the employee must [have been] subjected to employment practices which are discriminatory and which make the working conditions intolerable, thus forcing the employee to quit. Further, the employer's actions must be intended by the employer as an effort to force the employee to quit." EEOC v. Federal Reserve Bank, supra 698 F.2d at 672, quoting Irving v. Dubuque Packing Co., 689 F.2d 170, 172 (10th Cir. 1982). It was obvious that IBM made every effort to accomodate Hunter's promotional wishes during a difficult business period for the company. There was no evidence that Hunter was forced out; rather, the company made good faith efforts to investigate his complaints and resolve internally his grievance. His ultimate resignation was voluntary and despite his manager's serious efforts to persuade him to reconsider. Hunter's related claim of retaliation is without basis and no prima facie case of such a claim was established.

Similarly, there was no prima facie case of discrimination established by the class member claims. Many of the witnesses could assign no improper racial motive to their treatment by the company. Any that did claim discriminatory treatment were shown on cross-examination either to lack credibility or to be complaining of what can only be characterized as fair treatment. Given the extended time accorded to the direct testimony of the class witnesses, the Court finds astounding the absolute lack of substance to their claims. Many seemed intent on self-destructing careers that, if anything, were enhanced rather than limited by their race.

Presented for the Court's consideration was the testimony of but nine members of a putative class of over 220 during the relevant time period. At best, the class testified about only isolated incidents of perceived discrimination which is insufficient to establish that discrimination was the "standard operating procedure" of the employer without a valid statistical showing. Ste. Marie v. Eastern Railroad Ass'n, 650 F.2d 395, 405 (2d Cir. 1981). Even if the nine witnesses were regarded as credible, their testimony did not establish systemic discrimination. As discussed in detail in the findings of fact, no witnesses that the Court found credible could assign any racial motive to the actions of the company.

Finally, plaintiff's statistical case, offered by Dr. Sedlacek, fails to establish even an inference of disparate treatment by IBM. The infirmities in his analysis are many. He failed to work with the relevant class grouping. His comparisons of disparate groupings of employees are largely irrelevant. His multiple regression analyses omit seniority and job level, two key variables that may be significant determinants of success in IBM. Ignoring key variables may render the results of a multiple regression analysis useless. EEOC v. Federal Reserve Bank, supra, 698 F.2d at 656-60. Using current appraisals and time in current position as his key variables in no way filters out the other variables that may have caused any differences he found. See, e.g., Wilkins v. University of Houston, 654 F.2d 388, 402 (5th Cir. 1981), vacated on other grounds, 459 U.S. 809, 103 S.Ct. 34, 74 L.Ed.2d 47 (1982).

The Court finds that Dr. Sedlacek's analyses cannot be relied upon as indicative of a pattern of discrimination within the IBM Corporation. His analysis was flawed from its inception and cannot be considered competent for consideration. Although the Court need not consider IBM's statistical defense or the EEOC's rebuttal, Pouncy v. Prudential Insurance Co., supra, 668 F.2d at 803 n. 10, the Court notes that Dr. Medoff conducted properly controlled studies which rejected the notion that IBM engaged in a pattern or practice of discrimination. The differences found between blacks and whites were small and frequently favored blacks. The Court also notes that Dr. Michaelson, the EEOC's rebuttal witness, did not materially erode Dr. Medoff's analysis.

Although the Court also need not consider IBM's extensive testimony and exhibit presentation on its affirmative action program and performance evaluation system, the Court briefly points out that IBM has developed an elaborate system of safeguards to protect against the possibility of race or sex bias. Obviously, there is subjectivity built into IBM's evaluation system, but that is unavoidable and is checked by elaborate guidelines and procedures. IBM's system is the standard for its industry and for private business nationwide, and it is hard to imagine a system with any tighter controls.

The Court reserved ruling on defendant's motion to dismiss under Rule 41(b), of the Federal Rules of Civil Procedure, which was filed at the close of plaintiff's case. Because of the importance and length of the case, the Court wished a full record to be made. Although ruling on defendant's motion to dismiss still would be appropriate, see United States Postal Service Board of Governors v. Aikens, ___ U.S. ___, 103 S.Ct. 1478, 1482 n. 4, 75 L.Ed.2d 403 (1983), the Court shall decide this case on the entire record. The plaintiff failed to meet its initial burden of establishing a prima facie case. There has been no showing that IBM violated Title VII, 42 U.S.C. § 2000e-3, under either a disparate impact theory or a disparate treatment theory. In addition, although the burden to counter a prima facie case, see McDonnell Douglas, supra, never shifted to the defendant, the Court notes that IBM has articulated legitimate nondiscriminatory reasons for the actions and inactions of which the plaintiff complains. There is no showing that any of the defendant's explanations were pretextual. A full review of all the evidence on the record further establishes plaintiff's total failure to prove every allegation in the complaint by a preponderance of the evidence. A separate order incorporating the decision herein will be entered contemporaneously with the filing of this Opinion.


Summaries of

E.E.O.C. v. International Business Machines Corp.

United States District Court, D. Maryland
Apr 13, 1984
583 F. Supp. 875 (D. Md. 1984)
Case details for

E.E.O.C. v. International Business Machines Corp.

Case Details

Full title:EQUAL EMPLOYMENT OPPORTUNITY COMMISSION, Plaintiff, v. INTERNATIONAL…

Court:United States District Court, D. Maryland

Date published: Apr 13, 1984

Citations

583 F. Supp. 875 (D. Md. 1984)

Citing Cases

United States v. Johnson

The Government's suggested inference that ACSO applies a different search standard to Hispanics, even in…

United Food and Com. Workers U. Loc. 120 v. Wal-Mart Stores

The court also recognized that even with the serious flaws infecting plaintiffs' regression analysis, it…