Boyd v. Bechtel Corp., 485 F. Supp. 610 (N.D. Cal. 1979)

*611 *612 William E. Hoefs, Thelen, Marrin, Johnson & Bridges, San Francisco, Cal., for defendants.

Elizabeth G. Leavy, Carroll, Burdick & McDonough, Mark S. Rudy, San Francisco, Cal., for plaintiffs.

Cole & Scott, Rufus L. Cole, John Houston Scott, San Francisco, Cal., Oliver A. Jones, Regional Counsel, NAACP, San Francisco, Cal., Lois L. Johnson, Bushnell, Caplan & Fielding, Roderick P. Bushnell, San Francisco, Cal., for plaintiffs/objectors.

ORRICK, District Judge.

After more than three years of vigorous and thorough preparation, on the eve of trial, the parties to this employment discrimination suit against Bechtel Corporation, Bechtel International Corporation, Bechtel, Incorporated and Bechtel Power Corporation (hereinafter referred to as "Bechtel") brought on behalf of the black employees at Bechtel, presented to the Court for its approval a proposed settlement agreement. Following a lengthy and hotly contested Mandujano-type^[1] hearing, the Court, for reasons discussed herein, finds that the agreement is fair, reasonable and adequate.

Before embarking upon a detailed discussion of the reasons for approving the settlement, it is appropriate to offer some general observations concerning the difficulties now facing plaintiffs in trying or settling "second generation" Title VII suits of the type here at issue.

"First generation" Title VII suits involved overt intentional acts and policies which were not difficult to discern as discriminatory, such as the blanket exclusion of women from certain positions regardless of their ability to do the work, e. g., Rosenfeld v. Southern Pacific Co., 444 F.2d 1219 (9th Cir. 1971), or the use of tests with an obvious racial bias, e. g., Griggs v. Duke Power Co., 401 U.S. 424, 91 S. Ct. 849, 28 *613 L.Ed.2d 158 (1971). Today, however, after vigorous enforcement of our laws against employment discrimination by government agencies and "private attorneys general," employers are aware that their overall employment policies must, at the very least, comply with the letter of the law. Discrimination is likely to be found, if at all, in the enforcement, at lower levels, of company-wide policies which appear fair and reasonable on their face. Many employers have enacted affirmative action programs, and others, having contracts with the government, are periodically inspected by federal compliance agencies. Thus, if discrimination in employment exists in large organizations, it is likely to be subtle, detectable, if at all, only upon careful examination of sophisticated statistical analyses. As Judge Clark put it in Swint v. Pullman-Standard, 539 F.2d 77, 99 (5th Cir. 1976):

"If there ever was a time of facile Title VII litigation, it surely ended with the demise of intentional violations of equal employment opportunity. Today's parade of Title VII cases present more and more subtle manifestations of discrimination. Proof of invidious practices becomes more difficult as the ability to separate the real violation from the unfounded suspicion grows harder. This is especially so since many employers and unions * * * have made substantial good faith efforts toward eliminating racial distinctions for the work force."^[2]

It is in this context that the Court examines the settlement in the typical "second generation" case presently before it.

The original complaint in this case was filed on July 28, 1975, by plaintiffs Thomesine Lewis, Spotsel L. Boyd, and Richard J. Carter on behalf of themselves and a class of Blacks similarly situated, charging Bechtel, at the San Francisco Home Office, with engaging in a pattern of racial discrimination in employment in violation of the Civil Rights Act of 1866, 42 U.S.C. § 1981 (1974). Plaintiff Lewis subsequently withdrew from the case and has been dismissed. On June 16, 1976, the plaintiffs filed a First Amended Complaint against the defendants in which they added as a named plaintiff Sheila V. McDonald. This complaint was similar to the original one but added allegations under Title VII of the Civil Rights Act of 1964, 42 U.S.C. § 2000e, et seq., as amended (1974 & 1978 Supp.). On or about October 22, 1976, the plaintiffs filed a Second Amended Complaint against the defendants in which they added as a named plaintiff Alvino J. Hart. On January 3, 1978, the Court entered an Order that the action could be maintained as a class action under Rule 23(b) (2) of the Federal Rules of Civil Procedure.^[3] Defendants have denied all of the material allegations of all of the complaints.

Plaintiffs, seeking back pay, compensatory damages, and injunctive relief, contend that defendants have engaged in across-the-board discrimination^[4] against them as a class in connection with job assignment, promotion, salary increase, transfer, training *614 and development, discharge, layoff, and rehire. In addition, the four named plaintiffs raised particularized claims of racial discrimination: Spotsel Boyd, a black male employed at Bechtel as an engineer from May, 1974, until March, 1976, claims discrimination with respect to salary increase, promotion, work assignment, performance evaluation, transfer to an overseas assignment, and critical observations regarding absenteeism and tardiness; Sheila McDonald, a black female who has been employed by Bechtel in the Finance and Accounting Department since November, 1972, claims discrimination with respect to promotion and retaliation for filing a complaint with the Equal Employment Opportunity Commission ("EEOC"); Alvino Hart, a black male employed at Bechtel from April, 1973, until May, 1976, claims discrimination with respect to denial of training, promotion, tuition refunds, salary increases, transfer, layoff, and rehire.

The Bechtel group of companies consists of a number of separate corporations, chief among which are the four named defendants in this case: Bechtel Corporation, Bechtel International Corporation, Bechtel, Incorporated and Bechtel Power Corporation. The Board of Directors for each company is identical, and the stock of each of the companies is privately held by members of the Bechtel family and by members of the Board of Directors of the various companies. All of the Bechtel companies engage in engineering construction and related activities. They contract primarily with private parties to design and construct a variety of facilities, including refineries, petro chemical complexes, mining-metallurgical facilities, pipelines, and electrical power generating stations.

Bechtel is divided into a number of operating divisions, as follows: Hydro and Community Facilities Division, Mining and Metals Division, Pipeline and Production Service Division, Los Angeles Power Division, and Gaithersburg Power Division. In addition, within the Bechtel Corporation certain administrative operations are carried out for the whole group of companies in a variety of departments. These departments include: Finance and Accounting, Internal Auditing, Information and Planning Services, Procurement, Data Processing, Manpower Services, Special Management Services, and Labor Relations and Safety.

The defendants' workforce has a hierarchy of increasing levels of responsibility, skill and compensation, structured upon a system of salary grades which is currently composed of twenty levels. The lower eight grades, A through H, are called the "letter grades" and the upper twelve grades, grades 21 through 32, are referred to as the "number grades." Compensation, the level of responsibility, and the required skills progress in increasing alphabetical and numerical order. Numerous different jobs are lumped together in any one salary grade. There are presently over five hundred different job classifications which are held by the employees of the defendants. Broad categories have been devised to delineate groups of jobs having similar job content. Thus, Bechtel has divided its work into five job groups or "families": engineering, construction, administrative, technical support, and design drafting.

During the period in which this action was pending, the plaintiffs conducted thorough and wide-ranging discovery which was designed to investigate all of the dimensions of the alleged class discrimination. They propounded to the defendants and the defendants answered in excess of three hundred interrogatories. Plaintiffs served on defendants seven separate requests for production of documents which obligated the defendants to produce thousands of separate documents. They noticed and took the depositions of approximately forty of defendants' managers and supervisors. In addition, they obtained defendants' year-end personnel tapes and had them carefully reviewed and analyzed by computer and statistical experts. Through their discovery, the plaintiffs obtained voluminous information concerning the nature of the defendants' personnel procedures, policies, *615 and practices. For their part, defendants took the depositions of all of the named plaintiffs and all of their trial witnesses, with the exception of plaintiffs' statistical expert, and developed their own statistical analysis based upon their year-end personnel tapes.

Trial was scheduled to commerce on October 16, 1978. At the conclusion of the Second Pretrial Conference on October 12, 1978, counsel for the parties reached a tentative agreement or settlement. On October 17, 1978, counsel again met with the Court concerning the settlement and were advised that the Court would not be bound by any agreement they may have reached on the issue of attorneys' fees and costs, which would be set by the Court at or subsequent to the hearing on the fairness of the Consent Decree. On October 31, 1978, the parties lodged with the Court a mutually agreeable proposed Consent Decree, which provided that $120,000, to be distributed according to a formula based upon the duration of an employee's tenure at Bechtel, was to be set aside for distribution to all present and past employee class members. An additional $20,000 was allotted to the four named plaintiffs, Defendants also agreed to pay attorneys' fees and costs to opposing counsel in an amount to be determined by the Court, with an upper limit of $120,000 for fees, and $40,000 for costs.^[5] The Consent Decree contained no provisions for nonmonetary relief.

In considering the settlement, the Court followed procedures set forth in Section 1.46 of the Manual for Complex Litigation and prescribed by the United States Court of Appeals for the Ninth Circuit in Mandujano v. Basic Vegetable Products, Inc., 541 F.2d 832 (9th Cir. 1976). After addressing in written memoranda the eight inquiries regarding settlement listed in the Manual, counsel argued the same at a preliminary fairness hearing on November 2, 1978. At that time the Court determined that the proposed settlement was within the range of possible approval, and set a second fairness hearing for January 11, 1979. The Court also directed that defendants notify the class of the proposed settlement and Consent Decree on or before November 21, 1978, and approved the form of notice submitted by the parties.^[6] Notice was provided *616 to 1,127 class members: 717 notices were sent by first class mail to the last known address of the past employee class members; 410 notices were delivered to current employee class members by defendants' internal mail system; and the notice was also published in the San Francisco Chronicle, the San Francisco Examiner, and the Oakland Tribune.

By December 21, 1978, 153 "Notices of Objection" with the identical text and format were filed with the Court.^[7] The Court also received 7 handwritten or typed objections, 3 of which contained supplemental documentation sent by persons who had also filed the form Notice of Objection. In sum, a total of 160 objections, representing approximately 16 percent of the class, have been submitted to the Court.

On January 11, 1979, as scheduled, the Court held a fairness hearing, at which time 70 persons who appeared in Court to object to the Consent Decree were asked to submit their names and addresses for the record.^[8] After generally explaining the nature and purpose of the fairness hearing to the objectors and advising them of their right to retain counsel, the Court adjourned the fairness hearing to provide counsel for the objectors sufficient time to become acquainted with the case and to prepare their objections in an organized fashion. To facilitate *617 matters, the Court ordered that counsel for the objectors be associated with class counsel, and directed counsel for the objectors to file detailed narrative statements of the testimony which their witnesses intended to give at the next fairness hearing. The Court directed those objectors who were not represented by counsel to testify before a magistrate. The hearing was then continued until June 1, 1979.^[9] Prior to that time, counsel for the objectors filed several briefs in opposition to the Consent Decree, declarations of some 100 individuals objecting to the Decree, and proposed findings of fact and conclusions of law.^[10] At the final hearing, on June 1, 1979, the Court heard final argument and took the matter under submission.

In assessing a class action settlement, a district court must engage in a double-pronged inquiry. First, it must examine the proposed consent decree substantively and determine whether the relief it provides is fair, adequate and reasonable. In so doing, the court should consider several factors: the strength of the case; the reasonableness of the settlement in light of the attendant risks of litigation, and in light of the best possible recovery; the experience and views of plaintiffs' counsel; whether there is any evidence of collusion; the extent of discovery and the stage of the proceedings; the expected length, complexity and expense of further litigation; and objections to the settlement. Cotton v. Hinton, 559 F.2d 1326, 1330-31 (5th Cir. 1977); Flinn v. FMC Corp., 528 F.2d 1169, 1172-74 (4th Cir. 1975), cert. denied, 424 U.S. 967, 96 S. Ct. 1462, 47 L. Ed. 2d 734 (1976); Detroit v. Grinnell Corp., 495 F.2d 448, 463 (2d Cir. 1974); Alaniz v. California Processors, Inc., 73 F.R.D. 269, 278 (N.D. Cal.1976).

The court should independently weigh and evaluate each of these factors as applied to the case before it. A mere "rubber stamp" or "boilerplate" approval is inadequate. Detroit v. Grinnell Corp., supra, 495 F.2d at 462. On the other hand, the court is not to make a final determination of liability or damages, or to hold a trial or rehearsal of the trial. Nor is it to determine whether a better settlement might have been negotiated. Finally, the court should remain cognizant of the overriding public interest in settling large class actions, especially those involving employment discrimination. Alaniz v. California Processors, Inc., supra, 73 F.R.D. at 278; Flinn v. FMC Corp., supra, 528 F.2d at 1172-73; Cotton v. Hinton, supra, 559 F.2d at 1330; Detroit v. Grinnell Corp., supra, 495 F.2d at 462; Stull v. Baker, 410 F. Supp. 1326, 1333 (S.D.N.Y.1976); Saylor v. Lindsley, 456 F.2d 896, 904 (2d Cir. 1972); Pettway v. American Cast Iron Pipe Co., 576 F.2d 1157, 1214 (5th Cir. 1978), cert. denied, 439 U.S. 115, 99 S. Ct. 1020, 59 L. Ed. 2d 74 (1979).

In addition to these substantive concerns, the court must ensure that litigants' procedural rights, and especially those of any objectors to the settlement, as outlined in Mandujano are safeguarded throughout the settlement process. Thus, class members must be notified of a settlement in a manner that does not systematically leave any group without notice; notice must indicate that class members can object; each objection must be made part of the record; and each opponent raising a substantial objection has a right to be heard, to be represented by an attorney, and to have a reasoned response by the court on the record.

First, and most important, a settlement must be evaluated in light of the *618 strength of the case, and the court must balance the probable outcome of an action, if tried, against the provisions of the settlement. It is not necessary or expected that litigants will obtain through settlement all they might have realized through a victorious trial. Cognizant of the likely disappointment of class members who perceive they have been victims of discrimination in receiving less than the full recovery sought, the Court also recognizes that simply because a settlement may amount to only a fraction of the potential recovery does not in itself render it unfair or inadequate. Compromise is the very nature of settlement. Flinn v. FMC Corp., supra, 528 F.2d at 1173-74; Muldrow v. H. K. Porter Co., 20 Fed.R.Serv.2d 1069, 1075 (N.D.Ala.1975); Bryan v. Pittsburgh Plate Glass Co., 494 F.2d 799, 801 (3d Cir. 1974), cert. denied, 419 U.S. 900, 95 S. Ct. 184, 42 L. Ed. 2d 146 (1974); Grunin v. International House of Pancakes, 513 F.2d 114, 124 (8th Cir. 1975), cert. denied, 423 U.S. 864, 96 S. Ct. 124, 46 L. Ed. 2d 93 (1975); Detroit v. Grinnell Corp., supra, 495 F.2d at 455; Stull v. Baker, supra, 410 F. Supp. at 1332.

The case before the Court, an across-the-board attack on Bechtel's employment practices and policies, depends primarily upon statistics, and secondarily upon testimony by class members and documentary evidence.^[11] Having carefully examined the merits of this action, for reasons explained below, the Court finds that plaintiffs do not have a strong case, primarily because the results of their crucial statistical analyses are weak a shortcoming which cannot be overcome by their other evidence.

Statistical analysis is one of the most widely used and effective means of establishing the existence of a policy and pattern of discrimination. Dickerson v. United States Steel Corp., 439 F. Supp. 55, 65 (E.D. Pa.1977); Presseisen v. Swarthmore College, 442 F. Supp. 593, 599 (E.D.Pa.1977), aff'd 582 F.2d 1275 (3d Cir. 1978). Statistics are an especially important source of proof in cases involving employment discrimination, where plaintiffs attacking system-wide employment practices must establish that racial discrimination is the "standard operating procedure," something more than the mere occurrence of isolated or sporadic discriminatory acts. International Brotherhood of Teamsters v. United States, 431 U.S. 324, 339, 97 S. Ct. 1843, 1856, 52 L. Ed. 2d 396 (1977). See also Hazelwood School District v. United States, 433 U.S. 299, 97 S. Ct. 2736, 53 L. Ed. 2d 768 (1977).

On the other hand, "statistics are not irrefutable; they come in infinite variety and, like any other kind of evidence, they may be rebutted. In short, their usefulness depends on all of the surrounding facts and circumstances." International Brotherhood of Teamsters v. United States, supra, 431 U.S. at 340, 97 S. Ct. at 1856-1857. Thus, the significance of statistical findings may be diminished by such factors as small sample size or failure to consider relevant variables, or bolstered by the testimony of individual class members which brings "the cold numbers convincingly to life." Id. at 339-40, 97 S. Ct. at 1856. Presseisen v. Swarthmore College, supra, 442 F.Supp. at 599-600; Croker v. Boeing Co. (Vertol Division), 437 F. Supp. 1138, 1186 (E.D.Pa.1977); Parson v. Kaiser Aluminum & Chemical Corp., 575 F.2d 1374, 1387 (5th Cir. 1978).

In this "second generation" case, statistics are of vital importance, offering the most practical and comprehensive means of proving that discrimination is Bechtel's standard modus operandi. Although the Court cannot say that statistics are required in every case, the strength of this particular case is largely dependent upon them. In view of the emphasis the Court places upon statistics, below it will painstakingly analyze plaintiffs' statistical showing. First, the Court will explain the problems it perceives, *619 aided in some instances by the analysis of plaintiffs' expert. Next it will set forth the assessment of plaintiffs' attorneys. Interspersed throughout are comments by the objectors.

Initially, the Court will address several foundational problems which place additional burdens upon plaintiffs in proving the validity of their statistics. First, there is some question as to the appropriateness of certain samples. For example, the statistics on salary differences compare the salaries earned by Blacks and Whites at the same general salary grade within a job group.^[12] Defendants contend that plaintiffs should instead have compared salaries of persons holding the same individual jobs, who are engaged in literally "equal work." See, e. g., Smith v. Union Oil Co. of California, 17 FEP 960, 983-85 (N.D.Cal.1977); Agarwal v. McKee & Co., 16 EPD ¶ 8301 at 5580-81 (N.D.Cal.1977). The problem with defendants' view is that here, as in many cases, the resulting statistics would necessarily be predicated upon extremely small samples, which would make it virtually impossible for plaintiffs to develop significant data about company-wide practices. On the other hand, the Court could not permit plaintiffs to proceed with their present sample absent proof that in most cases various positions within a job family are sufficiently similar in skill, effort, and responsibility to warrant the use of job families as a proper ground for comparison of salaries, a proposition which defendants have already attempted to refute.^[13]See generally, Schlei & Grossman, Employment Discrimination Law 375-81 (1976), 104-05 (1979 Supp.).

A second, and related, problem, primarily affecting plaintiffs' statistics on promotion, is that many of the existing samples are too small to provide meaningful results. Although the value of statistics may be decreased by small sample size, even if the "universe" is relatively small, discrimination can still be shown if statistics show a great disparity in treatment and there is other strong evidence of discrimination. International Brotherhood of Teamsters v. United States, supra, 431 U.S. at 339 n.20, 97 S. Ct. at 1856 n.20; Ste. Marie v. Eastern R. R. Ass'n, 18 FEP 671, 683 (S.D.N.Y. 1978); Schlei & Grossman, Employment Discrimination Law at 328-29 (1979 Supp.). In this case, as explained infra, the statistics do not show a consistent pattern of discrimination or significantly great disparities in treatment. Nor does there appear to be other sufficient "strong evidence" of discrimination.

For example, this does not appear to be, as the objectors contend, a case involving totally standardless discretion in promotion, a factor which has in other cases remedied the potential handicap of small sample size.^[14] When such allegations are raised, the Court must examine the entire selection and promotion process and inquire whether it is totally discretionary or whether there are some guidelines or written instructions regarding qualifications. Ste. Marie v. Eastern R. R. Ass'n, supra, 18 FEP at 682-83. Here, in rebuttal to allegations of subjective performance evaluation, defendants have offered as trial exhibits policy manuals, instruction books, and training programs which specifically define the authority *620 of supervisors and the standards by which they operate. This rather impressive array of materials includes: a "Manual of Personnel Policies" (Exhibit 1521); a "Basic Supervisory Training" (Exhibit 1530); a "Supervisor's Workbook" and other materials prepared for Performance Review Training Programs (Exhibit 1543); a publication entitled "Equal Employment Opportunity and Affirmative Action Responsibilities of Bechtel Supervisors" (Exhibit 1545). Thus, this case is distinguishable from other, stronger, cases, in which no guidelines whatsoever exist with respect to promotion.^[15]

A third problem, affecting nearly all of the statistics, is that they fail to account for relevant individual variables. Both defendants and objectors, for different reasons, contend that absent consideration of, for example, the precise qualifications of an employee, statistics will be unreliable and skewed. Defendants point out that many personnel decisions affecting Whites as well as Blacks are based upon traditional nondiscriminatory reasons, so that statistical randomness cannot be assumed. The objectors contend that because the statistics fail to account for differences in authority or responsibility given to Blacks and Whites holding similar positions, they fail to adequately display the disfavorable treatment of Blacks. The omission of relevant factors does raise serious questions as to the weight to be afforded the statistics, and places a great burden on plaintiffs to illustrate, by testimony, the general absence of nondiscriminatory factors. The Court need not, however, reject the statistics in toto because of this problem, for it recognizes that it is virtually impossible to quantify all of the varied and individualized circumstances surrounding personnel actions. And in response to the objectors, some factors, such as "difference in authority," are simply too particularized and subtle to be quantifiable; it is thus doubtful that discrimination on such bases can be effectively handled in class litigation.

Finally, and most important, in addition to these foundational problems, which to some extent can be cured, plaintiffs' statistics do not show that Blacks received significantly disparate treatment from other employees. On this issue, the Court affords great weight to the affidavit of plaintiffs' statistical expert, in which he assesses the strength of the data he collected from Bechtel. He found that there were statistically significant differences in treatment between Blacks and Whites in some areas but not in others. Specifically, based upon a regression analysis for the years 1972-1976, taking into account sex, age, and education when hired, he determined that on the average, Blacks were assigned a salary grade when hired approximately .46 grade lower than were Whites. Based upon a regression analysis of salaries, taking into account sex, age, salary grade, time in salary grade, time since original hire, and job family, he found that for letter grade employees during 1972-1974 and for number grade employees in 1972, Blacks received significantly lower salaries than did Whites. For other years, the differences were not significant. His analysis of promotions in each year by job family and by letter or number grade showed disparities of statistical significance in only a few cases, although his analysis of employees promoted from a letter grade to a number grade demonstrated statistically significant differences for all years. His analyses of layoffs and firings by job family *621 and letter or number grade show statistically significant differences in few instances. His analysis of rehires of employees did not reveal statistically significant disparities. He concluded that the statistical results were mixed: plaintiffs' strongest issue was salary differences in early years; in regard to many other issues, though, he found plaintiffs' chances for recovery to be most uncertain. In fact, he recommended on several occasions that they attempt to settle the case. See generally Exhibits 128a-140.

The Court also notices that the statistics show an overall improvement in Bechtel's employment practices with respect to Blacks in recent years. Although this factor would not affect Bechtel's liability for discrimination in prior years, it, in addition to other evidence of Bechtel's efforts to counteract discrimination,^[16] renders uncertain the need for injunctive relief. See, e. g., Donnell v. General Motors Corp., 576 F.2d 1292, 1301 (8th Cir. 1978).

In addition to statistics, plaintiffs have prepared much individual testimony^[17] which, despite their commendable effort, would probably not be adequate to counteract the weakness of their statistics. Although such testimony is effective in bolstering statistics, it is a rare across-the-board case where, absent overt intentional discriminatory policies, it can effectively supplant the need for them.

The objectors, conceding that the statistics do not reflect a "strong" case, place a very high value on individual testimony, arguing that this case will stand or fall upon it. Thus, they contend that the one hundred affidavits they have submitted to the Court, alleging individual instances of discrimination, illustrate that black employees possess a very strong case against Bechtel. Even ignoring the foundational problems with the affidavits,^[18] the Court has already stated that testimony by class members, without more, cannot realistically reverse the uniform trend of data. As the class counsel somewhat facetiously suggest, literally thousands of witnesses would be required to duplicate the comprehensive nature of statistics. Moreover, in the face of the very thorough rebuttal already prepared by defendants,^[19] the Court cannot readily *622 assume that the testimony of class members would have a uniformly therapeutic effect on their case.

Although the class plaintiffs will also suffer from the statistics, individual testimony will be of greater importance to their case. It is difficult to assess the strength of their claims, but the Court suspects that they will have difficulty proving much damage, as three of the four plaintiffs spent only a few years at Bechtel, subsequently moving on to higher paying jobs. All plaintiffs are being compensated under the Consent Decree for all or a substantial part of their demands. See infra.

The conclusion that plaintiffs' statistics have numerous foundational problems, that they do not, in many areas, support their claims of disparate treatment, and that the other evidence prepared by plaintiffs is not likely to cure these problems, is corroborated by the views of the class attorneys, which the Court finds persuasive in light of their considerable experience litigating Title VII actions such as this one,^[20]Muldrow v. H. K. Porter Co., supra, 20 Fed.R.Serv. at 1074; Pettway v. American Cast Iron Pipe Co., supra, 576 F.2d at 1215; Blank v. Talley Industries, Inc., 64 F.R.D. 125, 132 (S.D. N.Y.1974), and the high caliber of their work on this case. Stull v. Baker, supra, 410 F. Supp. at 1332-33.

The recommendations of plaintiffs' counsel should be given a presumption of reasonableness. Id. at 1332. Attorneys, having an intimate familiarity with a lawsuit after spending years in litigation, are in the best position to evaluate the action, and the Court should not without good cause substitute its judgment for theirs. On the other hand, recognizing the potential conflict of interest between attorneys and the class they represent, the Court should not blindly follow counsel's recommendations, but give them appropriate weight in light of all factors surrounding the settlement. Pettway v. American Cast Iron Pipe Co., supra, 576 F.2d at 1215-16, 1176-77; Saylor v. Lindsley, supra, 456 F.2d at 900-01.

None of the circumstances surrounding this settlement detract from the presumption of reasonableness to which the class attorneys' views on this case are entitled. The settlement was reached after full discovery and full compliance with this Court's comprehensive pretrial order.^[21] Thus, the facts of this case have been sufficiently developed in an adversary context to permit a reasoned judgment of its merits. Moreover, there is no evidence whatsoever of collusion between the parties; in fact much of the negotiation occurred in the presence of the Court. Nor do we find any conflict of interest between plaintiffs' attorneys and the class which would operate to the detriment of the class. See infra; Pettway v. American Cast Iron Pipe Co., supra, 576 F.2d at 1215-16, 1176-77; Saylor v. Lindsley, supra, 456 F.2d at 900-01.

*623 It is the view of the class attorneys that the settlement affords the class as much as, if not more than, they would have received had the case gone to trial. They calculated the maximum liability of defendants to be about $750,000. To reach this figure they compared salary, layoff, firing, and promotion rates of Blacks and Whites, under the assumption that they would have been treated equally absent discrimination. They then calculated the total difference in pay received by similarly situated Blacks and Whites. They omitted one area of alleged discrimination from their calculations, salary grade when hired, because they thought it was virtually certain they would not prevail on that claim. Because the trial was to be bifurcated, with individual hearings on damages following the trial on class-wide liability, they recognized that the amount actually paid out by defendants would depend upon the number of individuals who successfully filed damage claims. In view of the difficulty of proving damages, the cost of hiring an attorney, and the fact that typically a very small percentage of class members initiate such claims,^[22] they thought it unlikely that defendants would have been required to pay out the full $750,000. Based upon their calculation that no more than eighteen percent of the class would prevail at the damage stage, they estimated that approximately $135,000 was a more realistic assessment of defendants' maximum liability. Given the risk that plaintiffs might lose on one or all issues at the liability stage, and the extreme unlikelihood that all claimants would prevail at the damage stage, they felt that the $120,000 set aside for the class, which represents eighty-nine percent of the maximum potential liability, was a reasonable figure. Another factor affecting their judgment was that the litigation would have been costly and time consuming: the trial on liability was estimated to last about six weeks; appeal was virtually certain; it is impossible to estimate how long it would have taken to complete the individual damage hearings.

Moreover, the class attorneys felt that the absence of nonmonetary relief should not invalidate the Consent Decree. Defendants were apparently unyielding in their refusal to accede to any such relief and, although class counsel would have preferred that the Decree contain injunctive relief, they did not think its absence warranted a refusal to settle the case. They reasoned that even if they had prevailed on the merits, an award of injunctive relief was unlikely, recognizing from research, experience, and admonitions from this Court, that courts in recent years have been very reluctant to order nonmonetary relief in cases such as this one.

The class attorneys' views were corroborated by John Low, an independent counsel who has had a great deal of experience litigating Title VII class actions.^[23] After reviewing the pleadings, motions, discovery, briefs, exhibits, proposed findings, and related documents on file in this case, Mr. Low stated that "the only reasonable inference that informed Title VII counsel could draw from the facts discovered was that should the case be tried, plaintiffs might lose on all counts," and under such circumstances might be liable for defendants' costs as well as their own. He also confirmed class counsel's view that few class members could be expected to file Stage Two (damage) claims, especially because the potential amount of back pay would be modest, at best, and that "a classwide settlement fund, although small in terms of individual awards, may far exceed the aggregate sum *624 obtained through the Stage Two `mini-trial' process." Affidavit of John Low.

Finding the settlement fair, adequate, and reasonable thus far, the Court must at last take into account the numerous objections. Opposition of a significant number of class members is a factor to be considered when approving a settlement, although it is not controlling. A settlement is not unfair simply because a large number or a certain percentage of class members oppose it, as long as it is otherwise fair, adequate, and reasonable. Mandujano v. Basic Vegetable Products, Inc., supra, 541 F.2d at 837; Cotton v. Hinton, supra, 559 F.2d at 1331; Flinn v. FMC Corp., supra, 528 F.2d at 1173-74; Muldrow v. H. K. Porter Co., supra, 20 Fed.R.Serv. at 1074-75; Bryan v. Pittsburgh Plate Glass Co., supra, 494 F.2d at 803; Pettway v. American Cast Iron Pipe Co., supra, 576 F.2d at 1214-15, 1217. Nor should a settlement be rejected merely because class plaintiffs oppose it. Flinn v. FMC Corp., supra, 528 F.2d at 1174; Saylor v. Lindsley, supra, 456 F.2d at 899-900. Thus, there are several cases in which the presence of a large number of objectors has not prevented a court from approving a settlement. See, e. g., Bryan v. Pittsburgh Plate Glass Co., supra (settlement approved despite opposition of more than twenty percent of the class); Muldrow v. H. K. Porter Co., supra (settlement approved over objection of twenty percent of the class); Cotton v. Hinton, supra (fact that objectors purported to represent fifty percent of class did not prevent settlement). To hold otherwise would put too much power in the hands of a few persons having no right to a preferred position in settlement, to thwart a result that might be in the best interests of the class. Saylor v. Lindsley, supra.

Cases in which large opposition has played a role in overturning settlements generally involve other significant factors as well. In the Mandujano case, in which five of nine named plaintiffs objected, there were serious due process violations: the court did not adequately notify class members of the settlement or enable them to voice their objections on the record. In Pettway v. American Cast Iron Pipe Co., notice to class members was also deficient. Moreover, in that case the vast majority of the class, seventy percent, and all three named plaintiffs objected. The court reasoned that in cases where all class members are affected equally by a settlement, where there is, for example, a settlement fund,^[24] the fact that a majority of the class objects is important. "[A]t some point objections from the class may become so numerous that in a very real sense it can be said that `the class' has not agreed to the proposal, that counsel's perceptions of the best interests of the class are faulty, and that approval of the settlement by the district court constitutes an abuse of discretion. No simple percentages are determinative, of course, and each case must turn on its own facts." Pettway v. American Cast Iron Pipe Co., supra, 576 F.2d at 1216.

In light of these principles, however, the Court does not find the quantum of objections, although impressive, sufficient in itself to overturn the settlement. Approximately sixteen percent of the class, including three of four named plaintiffs, have filed some opposition to the settlement.^[25] But, as in Pettway, all class members are affected equally by the settlement, and the Court finds persuasive the fact that eighty-four percent of the class has filed no opposition.

The opposition by class plaintiffs is at first glance more troublesome. But upon close examination, it appears that two of the three objecting plaintiffs, Richard Carter and Sheila McDonald, will recover under the Decree one hundred percent of their back-pay demands. The third objector, Spotsel Boyd, will receive fifty percent of *625 his maximum recovery.^[26] Moreover, two of the objecting plaintiffs are no longer employed at Bechtel, and thus have no real stake in whether the settlement includes nonmonetary relief. The weak standing of these plaintiffs to raise the objections they assert diminishes the initial impact of their numbers.

The Court therefore finds that plaintiffs have been represented by competent, experienced counsel, who, after a great deal of discovery and complete trial preparation, reasonably concluded that, due to the weakness of their case, the class would be better served by a fair settlement than by a costly, lengthy trial, which they could not expect to win. At arm's length, they negotiated a settlement with defendants, which, in light of the above factors, provided plaintiffs and the class with fair, reasonable and adequate relief. The size and the intensity of the opposition is not enough in itself to render the settlement inadequate. The Court additionally finds that more likely than not plaintiffs would have been unable to establish a prima facie case of class-wide discrimination with regard to most of the employment practices at issue, and that in light of the risks of litigation facing the individual plaintiffs, the $20,000 allocated to them represents a fair, adequate and reasonable compromise of their claims.

Under the second arm of its inquiry, the Court must respond to all substantial objections raised by opponents to the settlement. Mandujano v. Basic Vegetable Products, Inc., supra. In attacking the settlement, the objectors have attempted to establish the following: that the case is a strong one; that the need for injunctive relief is great; that the size of the back-pay award is extremely small, while the attorneys' fees are substantial; that there was inadequate notice of or class participation in the case; and that the case was inadequately prepared and negotiated by class attorneys. The Court has reviewed those objections in light of the objectors' obligation to provide the Court with concrete facts showing that vital information was ignored by the Court. Detroit v. Grinnell Corp., supra, 495 F.2d at 464. Under that standard, the Court can only conclude that the evidence proffered is insufficient to require rejection of the settlement.

The opponents' primary contention, one already discussed above, is that the proponents of the settlement have greatly under-estimated the true strength of the case which, when compared to the limited relief offered by the Consent Decree, renders the settlement unreasonable and unfair. They focus upon what they perceive to be ongoing discriminatory employment practices at Bechtel, namely subjective performance evaluations, unequal pay, overt abusive behavior (such as racial slurs, retaliation, acts of personal bias), high concentration of Blacks in nonauthoritative low-level positions and corresponding lack of participation by Blacks at managerial and supervisorial levels, high turnover rate for Blacks, practice of bypassing Blacks for promotion, or denial of earned or promised promotions, under or over utilization as a method of pressuring Blacks to leave the company, denial of access to or adequate credit for successful participation in training programs, and inadequacy of existing affirmative action programs.

The Court's uniform response is that absent proof these practices have a class-wide effect, the objectors' additional "evidence," in the form of affidavits and declarations, although disturbing, is not adequate to alter the Court's assessment of the case. Affidavits showing nine percent of Bechtel's current employees, who comprise only a part of the entire class, feel they have been victims of discrimination, cannot effectively persuade the Court that discrimination has a class-wide, rather than a merely scattered, individual, effect in the *626 face of contradictory competent statistical evidence. Significantly, the objectors have failed to impeach any of the proponents' statistics.

Moreover, many of the practices about which the objectors complain, such as under or over utilization, retaliation, and overt abusive behavior, either do not form a part of this lawsuit or are so individualized in nature as to be incapable of review on a class-wide basis. See supra and Croker v. Boeing Co. (Vertrol Division), supra, 437 F. Supp. at 1191-92. In any event, specific instances of such practices would have been used by class attorneys to bolster their statistics, and retaliation is already the focus of another lawsuit. Graves v. Bechtel Corp., C-79-0783 WHO (N.D.Cal., filed Apr. 6, 1979) (See note 9, supra).

With the rejection of these fundamental points, objectors' arguments regarding adequacy of relief must also fail. Despite their strenuous objections with respect to the need for injunctive relief, namely, protection against retaliation, appointment of a monitor, establishment of goals, timetables and other protection for affirmative action and upward mobility, for reasons stated above, it is highly unlikely that such relief, discretionary with the Court, would have been granted even had the case gone to trial.

Many objectors are strongly disappointed at the amount of back pay set aside for the class, which they inaccurately describe as $4 per month per employee. In view of the failure of many class members to submit claim forms, the estimated compensation is now approximately $10 per month and could ultimately be much more than that amount. As explained supra, class attorneys made a reasonable choice in opting for a formula approach to relief, which yields relatively small individual awards to a larger number of class members, and a greater overall recovery, than would individual damage trials. The aggregate sum allotted to the class, as also explained above, is fair and adequate in light of all the circumstances of this case.^[27]

The remaining issues raised by the objectors concern the class attorneystheir handling of the case, their relationship with the class, and their fees.

With respect to case preparation, the objectors contend discovery was inadequate in three respects: first, certain administrative personnel at Bechtel whom the objectors consider potential key witnesses were not deposed; second, there was not sufficient testimony proffered from class members; and, third, the information contained in Bechtel's year-end personnel tapes, from which class attorneys developed their statistics, was of questionable reliability and should have been verified. The response to all of these criticisms is that although the objectors have explained how they would have done things differently, they have not shown that the judgment of class attorneys was critically defective in any respect. Mere difference in judgment regarding case strategy, where reasonable minds can differ, is not a sufficient reason to overturn a settlement.

Plaintiffs engaged in a massive amount of discovery over a period of three years. This is not a case where there is any doubt that truly adversary discovery has occurred. Compare, Saylor v. Lindsley, supra, 456 F.2d at 904. They deposed all Bechtel personnel whom in their judgment would provide relevant testimony, and interviewed numerous class members. Mindful of the limitations on cumulative testimony imposed by this Court, they carefully selected for use at trial only the testimony they felt was strongest. The testimony objectors propose to develop is no different than that already prepared by class counsel.

The year-end personnel tapes were apparently incapable of verification because they *627 do not identify specific persons. It was reasonable to assume that Bechtel would not prepare fraudulent materials, and the objectors have not proffered any concrete facts supporting a contrary supposition. Moreover, the alternative to relying upon the tapes was to develop statistics from scratch from thousands of personnel filesclearly not a feasible approach.

Related to criticism of class counsel's discovery is the contention that throughout the litigation the class was not adequately notified of the case or encouraged to participate in its development or settlement. Objectors contend that had there been greater class participation, the case would have been much stronger. The first official notice of this action was sent to the class on November 21, 1978, informing them of the settlement. Because the class was originally certified pursuant to Rule 23(b) (2), and because it is clear that "[when] an action is certified under Rule 23(b) (2), * * * absent class members are not required to receive notice * *," EEOC v. General Electric Telephone Co. of Northwest, Inc., 599 F.2d 322, 334 (9th Cir., 1979), Wright & Miller, 7A Federal Practice and Procedure § 1786 (1st ed. 1972), notice was neither required nor sent at the time of certification.^[28] In light of the contemporaneous finding that the class attorneys could provide fair and adequate representation, the Court is confident that it was within its discretion in refusing to require notice. Moreover, many class members were "unofficially" aware of the litigation, and on several occasions met with class counsel.

The objectors also maintain that they, in particular the class plaintiffs, were not adequately informed of the settlement prior to its consummation. It scarcely needs reminding that counsel has a duty to keep the client informed of settlement prospects and negotiations. In Saylor v. Lindsley, supra, a shareholder's derivative action, class counsel apparently entered into a settlement without any notice to or authorization from the class plaintiff. Judge Friendly stated that:

"[W]e are not willing to * * * accept the view that the attorney for the plaintiff is the dominus litus and the plaintiff only a key to the courthouse door dispensable once entry has been effected. The attorney remains bound to keep his client fully informed of settlement negotiations, to advise the client before signing a stipulation of settlement on his behalf, and, if the client has objected, to inform the court of this when presenting the settlement, so that it may devise procedures whereby the plaintiff, with a new attorney, may himself conduct further inquiry if so advised." Saylor v. Lindsley, supra, 456 F.2d at 900 (emphasis added).

The question of how to allocate decision-making between the attorney and the class in class litigation is a more difficult one. This problem was discussed at length in Pettway v. American Cast Iron Pipe Co., supra, 576 F.2d at 1216:

"[T]he traditional notion of the `client' deciding important litigation questions is often problematic in the class action context because of the difficulty in identifying the client. * * * The class itself often speaks in several voices. * * * it may be impossible for the class attorney to do more than act in what he believes to be the best interests of the class as a whole. If the attorneys decision in the face of * * * disagreement affects each class member more or less equally, * * * the attorneys views must be accorded great weight * * *."

Elsewhere the court considers the role of class plaintiffs:

"[I]t remains unclear whether this model [the client deciding major litigation questions] can be carried over to the class action context, as no clear concept of the allocation of decision-making responsibility *628 between the attorney and the class members has yet emerged. Certainly it is inappropriate to import the traditional understanding of the attorney-client relationship into the class context by simply substituting the named plaintiffs as the client. * * * Were the class attorney to treat the named plaintiff as the exclusive client, the interests of other class members might go unnoticed and unrepresented." Id. at 1176 (citations omitted).

In the case at bar, attorneys met with class plaintiffs to discuss the possibility of settlement on five occasions, and active class members were kept well informed of negotiations.^[29] No stipulation of settlement was signed until all named plaintiffs and all active class members were informed of the terms of the settlement; and the Court was duly informed of the class' dissatisfaction. It is easy to appreciate the quandary facing the class attorneys, that is, whether to approve a settlement vociferously opposed by certain class members, or to prevent a disposition of the case, due to the opposition of some, which in their view would benefit the entire class. The Court finds that under the well-reasoned standards delineated by Judge Friendly and by the court in Pettway, class counsel adequately fulfilled their duty to provide advance notice of the proposed settlement, acted reasonably on behalf of the majority of the class in accepting the settlement, and adequately protected the procedural rights of the objectors.

Finally, several objectors were concerned with the amount of attorneys' fees which they erroneously thought were to be awarded to class counsel pursuant to the Consent Decree. This Court has made it abundantly clear that it alone has the authority to determine whether and how much attorneys' fees will be awarded in this case, and that it will not be bound by any agreements between the parties regarding fees. At any rate, the Decree merely sets forth the maximum amount defendants have agreed to pay. After approval of the settlement, class counsel may, in a separate hearing, fully document their entitlement to the requested fees. See, e. g., Norman v. McKee, 290 F. Supp. 29 (N.D.Cal.1968), aff'd 431 F.2d 769 (9th Cir. 1970); Foster v. Boise Cascade, Inc., 420 F. Supp. 674 (S.D.Tex. 1976), aff'd 577 F.2d 335 (5th Cir. 1978).

In conclusion, this case has presented the Court with an exceedingly difficult decision. Both the forceful, heartfelt objections raised by the able and experienced counsel for the opponents to the settlement, and the fact that the Consent Decree provides very modest relief to the class, have given the Court great pause. As a fiduciary to the class, the Court has scrutinized this settlement with utmost care to determine its fairness, principally in light of the strength of the case. The Court has found that there is substantial doubt whether plaintiffs could have prevailed in this "second generation" Title VII action. Thus, it is the Court's opinion, after examining the entire record, and in particular all of the materials prepared for trial and for the fairness hearings, that the settlement is fair, adequate and reasonable and is in the best interests of the class. Accordingly, the Court hereby approves the Consent Decree. As provided in the Consent Decree, the Court shall retain jurisdiction over this matter until the defendants have fully complied with the terms of the Decree.

[1] Mandujano v. Basic Vegetable Products, Inc., 541 F.2d 832 (9th Cir. 1976).

[2] See also, Croker v. Boeing Co. (Vertol Division), 437 F. Supp. 1138, 1192 (E.D.Pa.1977), and B. Nelson, New Directions in Settlement of Title VII Class Actions, paper presented at EEOC-MALDEF Employment Discrimination Law Conference, June 3, 1978.

[3] The class was defined as all past, present and future black employees at defendants' San Francisco offices who have been discriminated against on the basis of their race in violation of Title VII of the Civil Rights Act of 1964. On August 9, 1978, the Court amended the class certification order to allow the class to proceed under 42 U.S.C. § 1981 as well as under Title VII. There are three subclasses: a "clerical" subclass consisting primarily of letter grade employees in the administrative job family; an "administrative" subclass consisting primarily of employees in number grades in the administrative family; and a "technical" subclass consisting primarily of employees in letter and number grades in the engineering, construction, technical support, and design drafting job families. Bechtel's salary and job structure is explained infra.

[4] Bechtel's hiring practices are excepted from the otherwise comprehensive scope of this action.

[5] Other provisions of the Decree include a description of the case, a denial of liability by defendants, instructions regarding notice to the class, an explanation of how the settlement fund is to be distributed, and a provision donating nondistributed funds to the University of California.

[6] The notice sets forth the basic terms of the settlement and advises class members of their procedural rights, and states in relevant part:

"TIMETABLE FOR COURT ACTION

The Court has scheduled a hearing to consider approval of the Consent Decree, after class members have had an opportunity to review the Notice and make their views known to the Court. If you believe that the Consent Decree is fair, reasonable and adequate, there is nothing you need to do.

If you have any objections to the Consent Decree as proposed, you or your attorney must given written notice of these objections to [the Court and attorneys for plaintiffs and defendants]

* * * * * *

All objections to the Consent Decree must be signed and must contain your address and social security number. Objections must be postmarked or filed with the Court no later than December 21, 1978. Any class member who does not make an objection in the manner provided here will be deemed to have waived that objection and will be forever foreclosed from making any objection to the proposed settlement by appeal or otherwise.

A hearing will be conducted by this Court on January 11, 1979, at 4:30 p.m. in the courtroom of the Honorable William H. Orrick, Jr., 450 Golden Gate Avenue, San Francisco, California. At this hearing, the Court will decide whether or not to give final approval to the Consent Decree. You are entitled to have your attorney present at this hearing at your expense or you may personally appear at the hearing. If you or your attorney wish to appear at the hearing, you must file with the Court and with the individuals listed in the preceding paragraph by December 21, 1978, a written notice of intention to appear. Failure to file and mail a notice of intention to appear on time will mean that you or your attorney will not be heard at the hearing.

Plaintiffs and the class are represented by Elizabeth C. Leavy, Esq. and Mark S. Rudy, One Ecker Building, Suite 400, Ecker and Stevenson Streets, San Francisco, California 94105, (415) 494-0500. If you prefer to be represented by your own counsel, you may enter an appearance by your counsel on or before December 21, 1978; otherwise, for purposes other than objecting to the Consent Decree, you will be represented in this lawsuit, without charge to you, by Ms. Leavy and Mr. Rudy.

EFFECT OF THE SETTLEMENT AND THIS NOTICE

If the Consent Decree is approved by the Court, a Judgment will be entered in the lawsuit which will (1) entitle all members of the class to participate in the applicable benefits provided in the Consent Decree and (2) completely discharge the defendants and all of their directors, officers, employees, and agents and their predecessors and successors from any further liability to plaintiffs or the class members from any claims of race discrimination occurring on or before the date of the Consent Decree. This Judgment will bind all class members.

You may not exclude yourself from this class action. The Judgment which is entered pursuant to the Consent Decree after it is finally approved will prevent you from obtaining any relief in any judicial or administrative proceeding which you have brought or may bring against any of the defendants relating to race discrimination occurring on or before the date of the Consent Decree.

The proposed settlement contained in the Consent Decree is subject to approval by the Court and the issuance of this Notice shall not create any right in any class member or affect any issue in the lawsuit until and unless the Court approves the proposed settlement and then only in accordance with its final decree.

This Notice has been approved by counsel for all parties and by the Court."

[7] One hundred sixty such Notices were actually submitted, however, six were duplicates and one was filed by a person who was not a class member. The form reads as follows:

"As you are aware from the letter sent to you on December 4, 1978, three named plaintiffs did not agree to the terms of the Proposed Settlement of Class Action and Consent Decree. Additionally, the proposed settlement appears to provide a disproportionate sum for attorneys fees compared with those sums recommended for plaintiffs and members of the Class. Any statement alleging agreement to said documents by Plaintiffs Spotsel L. Boyd, Jr., Richard J. Carter and Sheila McDonald or by significant numbers of the Class is in error. On December 10, 1978 Richard J. Carter and Sheila McDonald, accompanied by several members of the named Class, met with you to discuss this matter. Despite stated objections by named plaintiffs and members of the Class therein present, you did not agree to pursue this matter further. Therefore, we request that you, as our representative, withdraw the Proposed Settlement and Consent Decree until such time as we, the clients, and you, our representatives, can agree on an appropriate course of action.

The Proposed Settlement and Consent Decree was initiated without specific consent of Spotsel L. Boyd, Jr., Richard J. Carter, Sheila McDonald or identified members of the Class. The signatories to this letter consider that the submittal of the Proposed Settlement and Consent Decree without consent of the named plaintiffs, and your lack of agreement to reconsider the matter, are possible grounds for submittal of necessary documents to the California State Bar Association charging you with not adequately representing those plaintiffs herein named and other members of the Class."

[8] Not all of these persons had filed written objections prior to the hearing, as required by the notice to the class.

[9] The hearing was initially set for April 23, 1979. On April 6, 1979, however, certain objectors filed a separate class action, Graves v. Bechtel Corp., No. C-79-0783 WHO (N.D.Cal.) seeking preliminary injunctive relief to prevent retaliation for participation in Boyd v. Bechtel. To better coordinate the two actions, the fairness hearing was postponed until June 1, 1979.

[10] Objectors also attempted, for the most part unsuccessfully, to obtain further discovery from Bechtel. The Court felt that further discovery was unnecessary in light of the massive amount of information already gathered by class counsel. See, e. g., Cotton v. Hinton, 559 F.2d 1326, 1329 (5th Cir. 1977); Detroit v. Grinnell Corp., 495 F.2d 448, 463 (2d Cir. 1974).

[11] Plaintiffs have submitted the following statistical exhibits: racial distribution of employees by grade and job family; regression analysis of salary when hired; comparison of salaries earned by Blacks and Whites by grade and job family; regression analysis of year-end salary; comparison of promotions of Blacks and Whites by grade and job family; actual versus expected number of Blacks promoted by year and job group; comparison of black and white hires; comparison of black and white layoffs and fires; comparison of rehires; etc.

[12] Bechtel's salary grades and job structure are described supra.

[13] For example, defendants point out that lawyers and public relations agents, employed at the same salary level within one job family, have dramatically disparate duties. Plaintiffs retort that this example illustrates the exception rather than the rule. The Court has no doubt that this issue would be hotly contested at trial.

[14] The objectors contend that Bechtel's use of subjective performance evaluations affects their ability to obtain promotions, as well as merit increases, training, and transfers, and causes them to receive a disproportionate number of layoffs and terminations. Specifically, they complain of a lack of adequately delineated standards or written qualifications; of a "buddy system" which operates to favor promotions of white employees over black employees by predominantly white supervisors; of job evaluations deliberately written to justify upcoming negative personnel decisions; and of the uneven application of the existing rules regarding performance evaluation.

[15] At the same time, because the objectors have not shown that subjective performance evaluations disproportionately affect Blacks (in, e. g., salary or promotion), proof that Bechtel engages in such practices could not itself provide a basis for any class-wide findings of discrimination. "The fact that certain elements of an employer's personnel practices and procedures involve subjective personnel practices, or determinations of employees' relative merit and qualifications, does not establish a violation of Title VII absent proof that subjective elements of the process are made in a manner which has a significant adverse impact on minorities." Agarwal v. McKee & Co., 16 EPD ¶8301 at 5586 (N.D.Cal.1977). Whatever method of proof is used, plaintiffs must show a disparity in their condition as compared to that of nonminority individuals. Dickerson v. United States Steel Corp., 439 F. Supp. 55, 64 (E.D. Pa.1977).

[16] Defendants have prepared testimonial and documentary evidence showing that they already engage in many employment practices which injunctive relief would presumably address, such as job posting, career counseling, uniform performance appraisal system, uniform salary administration system, training programs for supervisors. Exhibits 1521, 1536, 1540, 1541, 1547. They have also submitted into evidence copies of their affirmative action programs, which contain goals and timetables and outline policies and procedures applicable to all the Bechtel entities. Exhibits 1533A-1533U. Moreover, defendants, as government contractors, are subject to a panoply of regulations, audits, and reviews by the Office of Federal Contract Compliance Program. The Court rejects, in light of defendants' unrefuted statistics, the objectors' contention that Bechtel's affirmative action programs, especially those with respect to training, are merely cosmetic. See, e. g., Exhibits 2021-2022.

[17] In compliance with this Court's pretrial order, plaintiffs have listed detailed witness statements for all named plaintiffs and some fourteen Bechtel employees prepared to testify on behalf of the class. In excess of forty specific acts of discrimination would have been attested to at trial.

[18] Many of the allegations contained in the affidavits are based upon hearsay, opinion testimony, and highly subjective feelings and suspicions about discrimination. Although the objectors were not required to adhere to rules of evidence, such testimony cannot readily be considered by the Court as "factual" support for the presence of class-wide discrimination.

[19] Defendants have prepared "counter-statistics" and detailed witness statements for some eighty-two witnesses. They contend their statistics show that except for 1975, a greater percentage of Blacks were promoted than Whites (Exhibits 2016-2020); that their parity analysis shows that for each year and overall the actual number of black promotions did not differ significantly from the expected number (Exhibits 2084-2088); that their parity studies by salary grade prove that in later years Blacks received more training than would be expected from their percentage representation in various salary grades (Exhibits 2021-2022); that Blacks received favorable treatment with respect to transfer (Exhibits 2013, 2035-2039); that Blacks received on the average greater percentage of salary increase than comparable Whites (Exhibits 2062-2070); that there were no statistically significant differences in salary between Blacks and Whites in the same jobs from 1973-1977 (Exhibits 2047-2051); that their parity analysis of layoffs experienced by employees in the same job with the same length of service shows that there were no statistically significant differences between Blacks and Whites (Exhibits 2015, 2042-2046, 2077-2083); that their "turnover" analysis shows that on a job-by-job basis Black turnover percentages were comparable with non-black percentages (Exhibits 2014, 2037-2041). Where conflicting statistical conclusions are drawn from reasonably reliable statistics, plaintiffs usually fail to sustain their burden of proving discriminatory impact. Schlei & Grossman, Employment Discrimination Law at 329 (1979 Supp.).

[20] Elizabeth Leavy has practiced law since 1971. For three years she was a trial attorney with the San Francisco Regional Litigation Center of the EEOC, participating in six major actions initiated by the EEOC. For the past three years she has worked in a private law firm, where she has specialized in public sector labor law and employment discrimination law, representing both plaintiffs and defendants; while there she has been chief counsel for plaintiffs in five major Title VII cases. Mark Rudy has practiced law since 1970, during which time he has worked on eleven major lawsuits, chiefly in the areas of poverty law and employment discrimination.

[21] Each party filed trial briefs, proposed findings of fact and conclusions of law, narrative statements of witness' testimony, and exhibit lists, including statements of relevancy and sponsoring witnesses.

[22] See affidavit of Mark Rudy, filed May 18, 1979, and affidavit of John Low, filed May 18, 1979. In this particular case, only 50 of 717, or 7 percent of, the former employee class members sent in claims for a share of the proposed settlement. Even assuming a greater percentage of current employee class members participating in the suit, we cannot anticipate that overall a substantially greater number of class members would prevail at the damages stage.

[23] Prior to 1978, when he joined a private law firm, John Low was Senior Trial Attorney with the EEOC in San Francisco, where for five years he specialized in "pattern or practice" and class action litigation. He was lead government attorney in over fifteen major Title VII cases.

[24] In contradistinction are cases where the interests of class members conflict, involving, for example, disputes as to allocation of funds.

[25] In addition to the original objections, nine percent of the entire class, or nearly twenty-five percent of the current employee, filed affidavits.

[26] Alvino Hart, who has filed neither an objection nor a declaration, will also receive fifty percent of his maximum recovery. None of the plaintiffs will recover compensation for emotional distress. It is unlikely such relief, which is rarely awarded in Title VII or Section 1981 cases, would have been granted even if the case had gone to trial.

[27] Other substantive concerns of the objectors, including job training, different pay scales, and promotions, are discussed elsewhere. The remaining objections are insubstantial. These include, inter alia, the following complaints: that reserve funds will be donated to the University of California; that Bechtel does not admit discrimination; that the Consent Decree does not specifically identify discriminatory practices.

[28] In response to certain dissidents wishing to exclude themselves from the action, there is no right to "opt out" of a 23(b) (2) class action. EEOC v. General Telephone Co. of Northwest, Inc., 599 F.2d 322, 334 (9th Cir., 1979); Wright & Miller, 7A Federal Practice and Procedure § 1786 (1st ed. 1972).

[29] The first such meeting was held with approximately fifteen class members well before any settlement discussion had been held with Bechtel's attorneys. This and other meetings are described in detail in the Affidavit of Elizabeth Leavy filed January 4, 1979.