Sperling v. Hoffmann-La Roche, Inc.

924 F. Supp. 1346, 1996 U.S. Dist. LEXIS 6792, 76 Fair Empl. Prac. Cas. (BNA) 267, 1996 WL 249851

Opinion

OPINION

HAROLD A. ACKERMAN, District Judge.

This is a putative class action brought by former employees of Hoffmann-La Roche, Inc. (“Roche”), alleging violations of the Age Discrimination in Employment Act (“ADEA”), 29 U.S.C. § 621 et seq.

On February 4, 1984, Roche discharged or demoted approximately 1,100 employees pursuant to a reduction in force (“RIF”), known as Operation Turnabout. Based on Roche’s conduct during the RIF, Richard Sperling, one of the employees and a named plaintiff in this case, filed an age discrimination charge with the Equal Employment Opportunity Commission (“EEOC”) on behalf of himself and all employees similarly situated. Thereafter, on May 7, 1985, Sperling, along with the other named plaintiffs, filed this action on behalf of themselves and all others similarly situated, alleging, among other things, that the defendant discriminated against them in violation of the ADEA Subsequently, 476 of the over 1,100 employees affected in the RIF opted-in as members of the putative class.

Now before the court are motions by both parties. Plaintiffs move, pursuant to Federal Rule of Civil Procedure 42(b), for an order bifurcating the trial in this ease into two stages: (1) a Stage I trial regarding the issue of whether Roche engaged in a pattern or practice of age discrimination against the class; and (2) a series of Stage II trials that would address the liability and damages issues relating to the individual plaintiffs. Defendant Roche moves for summary judgment dismissing plaintiffs’ pattem-or-practice claim.

These matters were referred to the Special Master appointed in this case. 1 The Special Master submitted an opinion (1) granting plaintiffs’ motion for bifurcation; and (2) denying defendant’s motion for summary judgment dismissing plaintiffs’ pattem-or-practice claim. 2

On May 26,1995, oral argument was heard on Roche’s appeal from this decision. In Plaintiffs’ Brief in Opposition to Roche’s Appeal from the Special Master’s Denial of Roche’s Motion for Summary Judgment Regarding Plaintiffs’ “Pattern or Practice” Claims [hereinafter “Plaintiffs’ Pattern-or-Practice Opposition Br.”], plaintiffs “requested] a continuance, pursuant to [Federal Rule of Civil Procedure] 56(f), to complete discovery and present the full panoply of its pattern or practice evidence in opposition to this motion.” See Plaintiffs’ Pattem-orPractice Opposition Br. at 25. After being informed by the Special Master that such discovery would be completed in early December, I granted plaintiffs’ request in a Letter Opinion & Order dated November 2, 1995. The parties completed the filing of their supplemental submissions in early January 1996.

Because the resolution of these motions involve questions of law only, I must review the Special Master’s opinions de novo. See Prudential Ins. Co. of America v. U.S. Gypsum Co., 991 F.2d 1080, 1086 n. 11 (3d Cir. 1993) (citing Stauble v. Warrob, Inc., 977 F.2d 690, 697 (1st Cir.1992)). In addition, due to the granting of plaintiffs’ request pursuant to Federal Rule of Civil Procedure 56(f), there is an issue presently before the court that the Special Master did not have the opportunity to address, i.e., whether or not, given the “full panoply” of plaintiffs’ pattern-or-practice evidence, a reasonable jury could find that Roche engaged in a pattern or practice of discrimination. 3 When ruling on this issue, the standard of review is obviously the same standard that is applicable to the motion for summary judgment on which the Special Master did rule.

After summarizing the procedural history of these motions and the Special Master’s opinions, I will address the issues presented by Roche’s motion for summary judgment dismissing plaintiffs’ pattem-or-practice claim and the subsequent Rule 56(f) supplementation of the record. As will be discussed, the resolution of this motion moots plaintiffs’ motion for bifurcation.

I. Procedural History and the Special Master’s Opinions

Plaintiffs moved before the Special Master for an order, pursuant to Federal Rule of Civil Procedure 42(b), bifurcating the trial in this case into two stages: (1) a Stage I trial regarding the issue of whether Roche engaged in a pattern or practice of age discrimination against the class; and (2) a series of Stage II trials that would address the liability and damages issues relating to the individual plaintiffs. In response to this motion, Roche moved for summary judgment on plaintiffs’ pattern-or-praetice claim. Because the granting of Roche’s motion would moot plaintiffs’ bifurcation motion, the parties and the Special Master considered the two motions together. In an opinion, dated August 30, 1994, the Special Master held that plaintiffs’ motion for bifurcation should be granted and defendant’s motion for summary judgment dismissing plaintiffs’ pattem-or-practice claim should be denied.

The factual background on which the Special Master based his rulings in this opinion is as follows:

Roche is a large New Jersey drug manufacturer with a European parent. In the mid 1980s, Roche’s parent directed it to reduce costs, partly by reducing the size of its workforce. Roche carried this order out in two ways. First, it eliminated certain job types altogether. A number of plaintiffs apparently were terminated in consequence. Second, it reduced the number of employees in each department. A Roche task force prepared a set of guidelines to instruct managers in individual departments when these managers were deciding whom to release. The “line” managers then made termination decisions, apparently without further direction from senior Roche executives.

Roche claims that the RIF was a “one-shot” occurrence, and plaintiffs offer no evidence that Roche is contemplating another series of large scale dismissals. There is also little evidence that Roche systematically violated the ADEA before the date of the RIF or immediately thereafter. Thus plaintiffs’ case turns on whether Roche violated the ADEA in connection with the RIF.

Roche’s RIF guidelines explicitly directed its managers not to discriminate on the basis of age, race or sex when making termination decisions. Plaintiffs claim that the guidelines nevertheless left the individual managers with considerable discretion. The managers allegedly exercised this discretion, sometimes consciously and sometimes not, in contravention of the guideline’s explicit anti-discrimination directives and in contravention of the law; that is, Roche’s managers terminated plaintiffs on account of their age. Plaintiffs also claim that Roche violated the ADEA in connection with planning the RIF as well as executing it.

Opinion of Special Master, August 30, 1994, at 2-3 [hereinafter “Pattern-or-Practice Opinion”].

Roche made two arguments in support of its motion for summary judgment dismissing plaintiffs’ pattern-or-practice claim.

First, Roche argued that pattem-or-praetice claims are appropriate as a matter of law only where the class plaintiffs seek to enjoin the defendant from engaging in existing or threatened discriminatory behavior. Because Operation Turnabout has been completed and because the guidelines seemingly have no further application, there is, according to Roche, no illegal behavior to enjoin. Therefore, Roche argued that plaintiffs can challenge the RIF under the ADEA, but not in the form of a pattern-or-practice case.

The Special Master did not find this argument persuasive. If an employer is found to have engaged in a pattern or practice of discrimination, then class action plaintiffs are entitled to classwide, prospective injunctive relief. The Special Master correctly noted, however, that other benefits accrue to the plaintiffs upon a finding that an employer engaged in a pattern or practice of discrimination. Specifically, “[w]hen a pattern or practice of discrimination is found, the persuasion burden on the issue of whether the employer discriminated [against individual class members] shifts to the defendant.” Pattern-or-Practice Opinion at 6. Thus, the Special Master held that, in addition to obtaining injunctive relief, “a principal point of permitting class action plaintiffs to attack a pattern or practice of unlawful behavior is to shift the persuasion burden.” Id. at 9 (emphasis in original). In addition, “[t]he Supreme Court’s grounds for shifting the persuasion burden apply as strongly when class action plaintiffs primarily seek money as when they also seek an injunction against continuing violation of the law.” Id. at 10. Therefore, the Special Master rejected Roche’s argument that pattem-or-practice cases are not proper where there is no claim for injunctive relief against existing or threatened unlawful behavior.

Second, Roche argued that, as a matter of law, a “one-shot” event cannot constitute a pattern or practice of discrimination. Therefore, according to Roche, because Operation Turnabout was a one-shot event (i.e., there is no evidence that Roche has engaged in other large scale dismissals or that Roche has systematically violated the ADEA before or after the RIF), plaintiffs’ pattem-or-praetice claim must be dismissed and plaintiffs’ motion for bifurcation must be denied. The Special Master held that a one-shot event could not constitute a “practice” because “a ‘practice’ is the consistent application of a policy through time.” Pattern-or-Practice Opinion at 11. However, the Special Master also held that the element of duration that is implicit in a “practice” is “neither logically nor linguistically necessary to the finding of a ‘pattern’ of discrimination.” Id. This is because, in this case, approximately 1,100 employees were dismissed. Therefore,

if a sound statistical test establishes that Roche dismissed older workers at a rate that is statistically significantly higher than would have occurred by chance, and if there also is persuasive anecdotal evidence that the Roche managers who planned and conducted the RIF were motivated by age bias, then a jury could permissibly find that a “pattern” of discrimination on the basis of age likely existed.

Id. at 14. For these reasons, the Special Master denied Roche’s motion for summary judgment.

With respect to plaintiffs’ motion for bifurcation, Roche argued that, even assuming that plaintiffs have asserted a viable pattern- or-practice claim, bifurcation would be violative of Roche’s rights under the Due Process Clause of the Fifth Amendment and also under the Seventh Amendment.

Roche argued that it has a due process right “to prove that it relied on reasonable factors other than age when deciding to discharge ‘each of the plaintiffs’; and that it is entitled to ‘a single presentation of the evidence for each individual claim for relief.’ ” Pattem-or-Practice Opinion at 16. In other words, Roche claimed that if it were prohibited from introducing evidence regarding its reasons for terminating each individual plaintiff at the Stage I trial (where the only issue would be whether or not Roche engaged in a pattern or practice of discrimination and not whether or not each individual plaintiff was discriminated against), then Roche’s Fifth Amendment right to due process would be violated. The Special Master held that Roche’s Fifth Amendment rights would not be violated because, although Roche would be prohibited from introducing evidence regarding every plaintiff, Roche could “introduce direct statistical evidence, anecdotal evidence, illustrative evidence of individual dismissals and any other evidence that bears on the issue of whether a pattern of discrimination existed____” Pattern-orPraetice Opinion at 20. In addition, Roche could introduce evidence relating to individual claims at the Stage II trials where the issue of which individuals are entitled to relief would be litigated. Id. The Special Master also noted that bifurcated trials are the standard means by which pattern-or-practice claims are tried. Therefore, the Special Master rejected Roche’s Fifth Amendment due process challenge.

As mentioned above, Roche also argued that bifurcation would violate its rights under the Seventh Amendment. The Seventh Amendment provides, in part, that “no fact tried by jury, shall be otherwise re-examined in any Court of the United States, than according to the rules of common law.” The Special Master stated that “[t]he Seventh Amendment right derives from two goals, preventing jury confusion and avoiding inconsistent verdicts.” Pattern-or-Practiee Opinion at 17. Roche argued that bifurcation could result in inconsistent verdicts. More specifically, Roche argued that bifurcation could result in a finding at the Stage I trial that Roche engaged in a pattern or practice of discrimination, but at the Stage II trials Roche could still conceivably demonstrate that most, if not all, of the individual plaintiffs had not been the victims of discrimination. The Special Master held, however, that while this situation certainly presents a risk of inconsistent verdicts, that risk is slight in pattern-or-practice case. This is because, given the proofs required for a finding of a pattern or practice of discrimination, “if plaintiffs do prove at a Stage I trial that a pattern of illegal discrimination existed, the chances are low that Roche could show, in a large number of later stage trials, that it did not discriminate against individual plaintiffs.” See Pattem-or-Practice Opinion at 18-19. Therefore, the Special Master rejected Roche’s Seventh Amendment challenge.

For these reasons, the Special Master granted plaintiffs’ motion for bifurcation and denied defendant’s motion for summary judgment dismissing plaintiffs’ pattern-or-practice claim.

Although he noted that at this point in time he could not specifically state how a Stage I trial should be conducted, the Special Master did discuss in the opinion “some broad guidelines” regarding the conduct of the State I trial. See Pattern-or-Practice Opinion at 21-25. Plaintiffs will introduce (1) expert testimony regarding whether or not older employees were discharged at a statistically significant rate and (2) anecdotal evidence, through the testimony of individual plaintiffs and perhaps Roche managers, that individual Roche managers were animated by illegal bias. Roche would be permitted to introduce its own statistical evidence. In addition, Roche could introduce testimonial and documentary evidence to the effect that its managers acted lawfully when deciding what jobs to eliminate and whom to dismiss and that it was not Roche’s policy to discriminate. The Special Master also stated that

[t]he number of Roche witnesses Roche will need cannot be specified now____ Roche’s precise case will be a function of its view of the strength of plaintiffs’ case, a view that will become clearer as discovery concludes and a pre-trial order comes to be drawn____ [However,] Roche will not be permitted to offer evidence that each individual plaintiff was lawfully discharged or suffered little harm or should have mitigated damages because these are not issues in a Stage I trial.

Pattern-or-Practice Opinion at 23.

The Special Master revisited the question of what types of evidence Roche would be permitted to introduce at the Stage I trial in an opinion dated March 8, 1995 [hereinafter “Bifurcation Opinion”]. This opinion was prompted by Roche’s revelation of the strategy it intended to pursue at the Stage I trial. Roche informed the plaintiffs and the Special Master that it would attempt to disprove plaintiffs’ pattern-or-practice claim by showing that a majority of the plaintiffs, of whom there are 476, were dismissed lawfully. To establish this, Roche intended to call a large number of witnesses. Because Roche’s strategy seemed novel, the Special Master asked the parties to brief the issue of whether or not Roche could pursue this strategy.

Subsequently, the Special Master held that Roche is prohibited from pursuing this strategy, because allowing Roche to pursue it would vitiate the standard procedure of bifurcating pattern-or-practice cases into a Stage I trial at which the issue of whether the employer engaged in a pattern or practice of discrimination against the class is litigated and Stage II trials at which the liability and damage issues relating to the individual plaintiffs is litigated. This is because Roche’s strategy would force plaintiffs to contest Roche’s evidence as to a majority of the plaintiffs at the Stage I trial. Thus, “plaintiffs would have to establish the illegality of each individual discharge without the benefit of the presumption of discrimination to which their proof [might] otherwise entitle them.” Bifurcation Opinion at 7 (emphasis in original). Roche’s strategy would in effect collapse the Stage II trial into the Stage I trial. Therefore, the Special Master held that Roche could not pursue this strategy. Although the precise number of witnesses that Roche would be permitted to call still could not be determined, the Special Master identified the following relevant categories of evidence that Roche should be permitted to introduce:

(a) statistical evidence either original or in rebuttal; (b) documentary evidence, such as materials relating to the planning and conduct of the RIF; (c) anecdotal evidence in rebuttal and otherwise; and (d) illustrative testimony that Roche managers complied with the law when making termination decisions.

Bifurcation Opinion at 11-12 (emphasis in original).

The Special Master’s opinions make clear that two things are at stake in the resolution of these motions: (1) whether the plaintiffs will have an opportunity to demonstrate a pattern or practice of discrimination, and thereby, if so demonstrated, benefit from the resulting shift to Roche of the burden of persuasion on the issue of whether individual plaintiffs were victims of discrimination; and (2) assuming that plaintiffs have the opportunity to prove a pattern or practice of discrimination, in what manner will this issue be litigated. Thus, the threshold issue is whether or not, as a matter of law, there is a viable pattern-or-practice claim in this case.

As discussed above, before the Special Master, Roche made two arguments in support of its assertion that there is no pattern- or-practice claim in this case — (1) pattern-or-practice claims are only appropriate where the class plaintiffs seek to enjoin the defendant from engaging in existing or threatened discriminatory behavior; and (2) a “one-shot” event cannot constitute a pattern or practice of discrimination. Due to the granting of plaintiffs’ request pursuant to Rule 56(f), there is a third issue before the court— whether or not, given the “full panoply” of plaintiffs’ pattern-or-practice evidence, a reasonable jury could find that Roche engaged in a pattern or practice of discrimination.

II. Summary Judgment Standard

Federal Rule of Civil Procedure 56 provides that summary judgment may be grant ed only if the pleadings, supporting papers, affidavits, and admissions on file, when viewed with all inferences in favor of the nonmoving party, demonstrate that there is no genuine issue of material fact and that the movant is entitled to judgment as a matter of law. Fed.R.Civ.P. 56(c). See also Todaro v. Bowman, 872 F.2d 43, 46 (3d Cir.1989); Chipollini v. Spencer Gifts, Inc., 814 F.2d 893, 896 (3d Cir.), cert. dism’d, 483 U.S. 1052, 108 S.Ct. 26, 97 L.Ed.2d 815 (1987). The substantive law will identify which facts are “material.” Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 247-48, 106 S.Ct. 2505, 2509-10, 91 L.Ed.2d 202 (1986). Therefore, “[o]nly disputes over facts that might affect the outcome of the suit under the governing law will properly preclude the entry of summary judgment.” Id. An issue is “genuine” if a reasonable jury could possibly hold in the nonmovant’s favor with regard to that issue. Id.

At the summary judgment stage, a court may not weigh the evidence or make credibility determinations — these tasks are left to the factfinder. Petruzzi’s IGA Supermarkets, Inc. v. Darling-Delaware Co., Inc., 998 F.2d 1224, 1230 (3d Cir.), cert. denied, — U.S. -, 114 S.Ct. 554, 126 L.Ed.2d 455 (1993). Therefore, to raise a genuine issue of material fact, “ ‘the [summary judgment] opponent need not match, item for item, each piece of evidence proffered by the movant,’ but simply must exceed the ‘mere scintilla’ standard.” Id. See also Anderson, 477 U.S. at 252, 106 S.Ct. at 2512 (“The mere existence of a scintilla of evidence in support of the [nonmovant’s] position will be insufficient; there must be evidence on which the jury could reasonably find for the [nonmovant].”).

If the court determines that there is no genuine issue of material fact and that the movant is entitled to judgment as a matter of law, then summary judgment may be granted.

III. The Parameters of a Pattem-or-Practice Claim

As stated above, Roche made two arguments before the Special Master in support of its motion for summary judgment on plaintiffs’ pattem-or-practice claim — (1) pattem- or-practice claims are appropriate only where the class plaintiffs seek to enjoin the defendant from engaging in existing or threatened discriminatory behavior; and (2) a “one-shot” event cannot constitute a pattern or practice of discrimination. Roche also argues, in its brief submitted in opposition to plaintiffs’ Rule 56(f) submission, that plaintiffs’ class-wide claim is actually a disparate impact claim and not a pattem-or-practice claim. See Memorandum of Hoffmann-La Roche Inc. in Response to Plaintiffs’ Rule 56(f) Submission on Their “Pattem-or-Practice” Claim at 12 (“[I]f plaintiffs as a factual matter have any class-wide claim at all on their theory of ‘unplanned’ and ‘unconscious’ stereotyping, it is a disparate impact claim and net a pattem-or-practice claim”) (emphasis in original). All of these arguments are in effect alternative arguments in support of the proposition that, even assuming that plaintiffs’ allegations regarding the manner in which Operation Turnabout was conducted are true, these allegations are insufficient to bring this ease within the parameters of a pattem-orpractice case. Prior to addressing these arguments, I will discuss the .law regarding what constitutes a pattern or practice of age discrimination.

A Pattern or Practice of Age Discrimination

Initially it should be noted that the Age Discrimination in Employment Act (“ADEA”), the statute pursuant to which the action now before the court was brought, does not explicitly provide for pattem-orpractice claims. Rather, the statutory basis for pattem-or-practice claims is found in Section 707(a) of Title VII of the Civil Rights Act of 1964 [“Title VII”], which provides, in part, that

[w]henever the Attorney General has reasonable cause to believe that any person or group of persons is engaged in a pattern or practice of resistance to the full enjoyment of any of the rights secured by [Title VII], and that the pattern or practice is of such a nature and is intended to deny the full exercise of the rights [described in Title VII], the Attorney General may bring a civil action ... requesting such relief, including an application for a permanent or temporary injunction, restraining order or other order against the person or persons responsible for such pattern or practice, as he deems necessary to insure the full enjoyment of the rights herein described.

42 U.S.C. § 2000e-6(a). 4 Although this section does not explicitly authorize private litigants to bring pattern-or-practice claims, courts, including the Supreme Court, have implicitly approved of pattern-or-practice claims being brought by private litigants in the form of a class action. See, e.g., Cooper v. Federal Reserve Bank of Richmond, 467 U.S. 867, 876 n. 9, 104 S.Ct. 2794, 2799 n. 9, 81 L.Ed.2d 718 (1984) (stating that “Although Teamsters [v. United States, 431 U.S. 324, 97 S.Ct. 1843, 52 L.Ed.2d 396 (1977) ] involved an action litigated on the merits by the Government as plaintiff ..., it is plain that the elements of a prima facie pattern-or-practice case are the same in a private class action”). Furthermore, although the ADEA does not explicitly provide for pattern-or-practice claims, courts have applied the law developed in the context of Title VII pattern- or-practice claims to claims of age discrimination. See, e.g., King v. General Electric Company, 960 F.2d 617 (7th Cir.1992); EEOC v. Western Electric Company, Incorporated, 713 F.2d 1011 (4th Cir.1983); EEOC v. Sandia Corporation, 639 F.2d 600 (10th Cir.1980). Therefore, opinions discussing the meaning of “pattern or practice” in the context of Title VII are equally applicable to actions brought under the ADEA.

In an action alleging that an employer engaged in a pattern or practice of age discrimination, the ultimate factual issues are whether there was a pattern or practice of disparate treatment and, if so, whether the differences in treatment were due to age. See International Brotherhood of Teamsters v. United States, 431 U.S. 324, 335, 97 S.Ct. 1843, 1854, 52 L.Ed.2d 396 (1977) [hereinafter “Teamsters ”]. Before discussing what constitutes a “pattern or practice,” it is helpful to first discuss what constitutes disparate treatment based on age.

Disparate Treatment: An employer engages in disparate treatment where it treats some people less favorably than others because of a protected trait. See id. at 335 n. 15, 97 S.Ct. at 1854 n. 15. To succeed on a disparate treatment claim, “[p]roof of discriminatory motive is critical, although it can in some situations be inferred from the mere fact of differences in treatment.” Id. (citations omitted). 5

The language of the ADEA makes clear that a disparate treatment theory is available under the statute. See Hazen Paper Company v. Biggins, 507 U.S. 604, 609-11, 113 S.Ct. 1701, 1706, 123 L.Ed.2d 338 (1993) [hereinafter “Hazen Paper”]. The ADEA provides, in part, that

[i]t shall be unlawful for an employer ... to discharge any individual ... because of his age.

29 U.S.C. § 623(a)(1) (emphasis added). Under the ADEA, “a disparate treatment claim cannot succeed unless the employee’s [age] actually played a role in [the employer’s decisionmaking] process and had a determinative influence on the outcome.” Hazen Paper, 507 U.S. at 610, 113 S.Ct. at 1706. See also Miller v. CIGNA Corporation, 47 F.3d 586, 595 (3d Cir.1995) (holding that plaintiff in an ADEA ease “has the burden of persuading the trier of fact by a preponderance of the evidence that there is a ‘but-for’ causal connection between the plaintiffs age and the employer’s adverse action”).

Disparate treatment, thus defined, captures the essence of what Congress sought to prohibit in the ADEA. It is the very essence of age discrimination for an older employee to be fired because the employer believes that productivity and competence decline with old age---- Congress’ promulgation of the ADEA was prompted by its concern that older workers were being deprived of employment on the basis of inaccurate and stigmatizing stereotypes.

Hazen Paper, 507 U.S. at 610, 113 S.Ct. at 1706 (citations omitted).

In Hazen Paper, the Supreme Court emphasized that when an employer’s decision is entirely motivated by factors other than age, the problem of inaccurate and stigmatizing stereotypes is not present. See id. This is the ease even where the motivating factor is correlated with age. For example, pension status is typically correlated with age. However, if an employer fires an employee solely because the employee’s pension is about to vest, the employer does not violate the ADEA, even if older employees are more likely to be close to having their pensions vest than younger employees. See id. at 609-13, 113 S.Ct. at 1706-07. This is because

[t]he prohibited stereotype (“Older employees are likely to be —”) would not have figured in this decision, and the attendant stigma would not be the result of an inaccurate and denigrating generalization about age, but would rather represent an accurate judgment about the employee — that he indeed is “close to vesting.”

Id at 612, 113 S.Ct. at 1707. In other words, “[bjecause age and years of service are analytically distinct, an employer can take account of one while ignoring the other, and thus it is incorrect to say that a decision based on years of service is necessarily ‘age-based.’” Id.

If, however, an employer targeted for termination employees with a particular pension status based on the assumption that these employees would likely be older, then the employer would have discriminated based on age. Id. This is because the employer was actually motivated by age and merely used pension status as a proxy for age. Id.

Thus, if, in conducting Operation Turnabout, Roche terminated a member of the plaintiff class because of that plaintiffs age (and the plaintiff was at least 40 years old, see 29 U.S.C. § 631(a)), then Roche would have engaged in disparate treatment based on age against that plaintiff. At issue on this motion, however, is whether a reasonable jury could find that Roche engaged in a “pattern or practice” of such disparate treatment.

Pattern or Practice: In order to establish that Roche engaged in a pattern or practice of discrimination, plaintiffs must establish

more than the mere occurrence of isolated or “accidental” or sporadic discriminatory acts. [They must] establish by a preponderance of the evidence that [age] discrimination was the company’s standard operating procedure — the regular rather than the unusual practice.

Teamsters, 431 U.S. at 336, 97 S.Ct. at 1854. Title YII’s legislative history provides, in part, that

[a] pattern or practice would be present only where the denial of rights consists of something more than the isolated, sporadic incident, but is repeated, routine, or of a generalized nature. There would be a pattern or practice if, for example, a number of companies or persons in the same industry or line of businesses discriminated throughout all or a significant part of its system, or if a company repeatedly and regularly engaged in acts prohibited by the statute. The point is that single, insignificant, isolated acts of discrimination by a single business would not justify a finding of a pattern or practice____

110 Cong.Rec. 14270 (1964) (remarks of Senator Humphrey) (quoted in Teamsters, 431 U.S. at 336 n. 16, 97 S.Ct. at 1855 n. 16). In short, a company engages in a pattern or practice of discrimination when it systematically engages in intentional discrimination against a protected group, or in other words, when it maintains a regular, corporate policy of purposeful discrimination in some aspect of its employment practices. See Teamsters, 431 U.S. at 361-62, 97 S.Ct. at 1867-68. See also Cooper, 467 U.S. at 877-78, 104 S.Ct. at 2800-01 (referring to pattern or practice as “a companywide policy, or even a consistent practice within a given department”); King, 960 F.2d at 623 (stating that “plaintiffs in a pattern-or-practice ... class discrimination claim have the burden of showing that the unlawful discrimination was the employer’s regular policy” (quoting EEOC v. Sears, Roebuck & Co., 839 F.2d 302, 354-55 (7th Cir. 1988)); Dillon v. Coles, 746 F.2d 998, 1004 (3d Cir.1984) (stating that “the class representative’s prime facie case must establish that discrimination was the employer’s standard practice”); Segar v. Smith, 738 F.2d 1249, 1274 (D.C.Cir.1984) (stating that “[a] pattern or practice case challenges a host of employment decisions over time; in effect, it challenges an employment system”), cert. denied, 4 71 U.S. 1115, 105 S.Ct. 2357, 86 L.Ed.2d 258 (1985).

That “pattern or practice” means a consistent policy of purposeful discrimination is further supported by the consequences of a finding that a defendant engaged in a pattern or practice of discrimination.

Once the court finds that the defendant engaged in a pattern or practice of discrimination, then this finding, without any further evidence, justifies an award of prospective relief. Teamsters, 431 U.S. at 361, 97 S.Ct. at 1867-68.

Such relief might take the form of an injunctive order against continuation of the discriminatory practice, an order that the employer keep records of its future employment decisions and file periodic reports with the court, or any other order “necessary to ensure the full enjoyment of the rights” protected by Title VII.

Id. The particular prospective relief ordered in a case should be designed to ensure that employers found to have engaged in a pattern or practice of discrimination “eliminate their discriminatory practices and the effects therefrom.” See id. at 361 n. 47, 97 S.Ct. at 1868 n. 47. Thus, a finding of a pattern or practice establishes liability as to the class and justifies injunctive relief to eliminate the policy that was harming the class. See Price Waterhouse v. Hopkins, 490 U.S. 228, 266, 109 S.Ct. 1775, 1799, 104 L.Ed.2d 268 (1989) (O’Connor, J., concurring); Dillon, 746 F.2d at 1004. The provision for this type of relief clearly indicates that a “pattern or practice” is something more than random or sporadic discriminatory acts. Rather, a “pattern or practice” is a consistent policy of purposeful discrimination, the elimination of which requires the use of the court’s broad equitable powers. See Teamsters, 431 U.S. at 361 n. 47, 97 S.Ct. at 1868 n. 47. In addition, such a policy of discrimination may be demonstrated by showing either a consistent history of discriminatory decisionmaking in an employer’s employment practices or by showing that the employer has an openly declared discriminatoxy policy. See, e.g., United States v. Bd. of Educ. of School Dist. of Phila., 911 F.2d 882, 892-93 (3d Cir.1990) (stating in the context of a Title .VII pattern-or-practice claim that “[w]here the allegedly discriminatory policy is openly declared ... then proof that the policy was actually being followed consistently is not necessaxy in order to obtain an injunction against subsequent implementation” and therefore in this situation a plaintiff need not demonstrate “a consistent history of violation of Title VII” in order to prove a pattern or practice) [hereinafter “Phila, Bd. of Ed.”]; 110 Cong.Rec. 14239 (1964) (“It would be clear that an establishment or employer that consistently or avowedly denies rights under [Title II and Title VII] is engaged in a ‘pattern or practice of resistance.’ ”) (remarks of Senator Humphrey) (cited in Teamsters, 431 U.S. at 336 n. 16, 97 S.Ct. at 1855 n. 16).

Also probative of the meaning of “pattern or practice” is the effect of a finding that the employer engaged in a pattern or practice of discrimination on the individual class members’ claims for relief. If individual relief for the victims of the discriminatory policy is sought in addition to classwide injunctive relief, “then the district court must usually conduct additional proceedings after the liability phase of the trial to determine the scope of individual relief.” Teamsters, 431 U.S. at 361, 97 S.Ct. at 1867. See also Price Waterhouse, 490 U.S. at 277, 109 S.Ct. at 1804-05 (O’Connor, J., concurring) (stating that the finding that discrimination was the employer’s “standard practice” establishes li ability to the class, “[b]ut as to the individual members of the class the liability phase of the litigation is not complete”); Dillon, 746 F.2d at 1004 (same). At these additional proceedings, the individual class members can build on the finding that the defendant engaged in a pattern or practice of discrimination. More specifically, “[t]he proof of pattern or practice supports an inference that any particular employment decision, during the period in which the discriminatory policy was in force, was made in support of that policy.” Teamsters, 431 U.S. at 362, 97 S.Ct. at 1868. The individual class member need only show that he or she was “a potential victim of the proved discrimination.” Id. For example, if the class demonstrated that the employer engaged in a pattern or practice of discrimination in hiring persons for a certain position, then the class member need only demonstrate that he or she unsuccessfully applied for that position, and therefore, he or she was a potential victim of the discriminatory policy. 6 See id. Significantly, the employer thereafter has the burden of demonstrating that the potential victim of the discriminatory policy was denied an employment opportunity for lawful reasons. Id. In other words, once the class proves the existence of a discriminatory policy and once an individual class member demonstrates that he or she was a potential victim of that policy, then there is a rebuttable presumption that the employer discriminated against the class member. Id. at 359 n. 45, 97 S.Ct. at 1867 n. 45.

The burden of proof shifts to the employer in this situation because although the finding of a discriminatory policy does “not conclusively demonstrate that all of the employer’s decisions were part of the proved discriminatory pattern or practice, it [does] create a greater likelihood that any single decision was a component of the overall pattern.” Id. (emphasis added). Thus, the presumption shifting the burden to the employer is based on a “judicial evaluation” of evidentiary probabilities. See Teamsters, 431 U.S. at 359 n. 45, 97 S.Ct. at 1867 n. 45. Accord Price Waterhouse, 490 U.S. at 266-67, 109 S.Ct. at 1799-1800 (O’Connor, J., concurring). “Because the class has already demonstrated that, as a rule, illegitimate factors were considered in the employer’s decisions,” then it is more likely that an illegitimate criterion was a factor in the individual employment decision, and, therefore, the burden shifts to the employer to prove that the legitimacy of its decisions. See Price Waterhouse, 490 U.S. at 266-67, 109 S.Ct. at 1799-1800 (O’Connor, J., concurring) (emphasis added). 7

Thus, it is clear that a “pattern or practice of discrimination” is a policy of discrimination that is consistent enough, or pervasive enough, to justify the conclusion that it is more probable than not that the employer discriminated against any individual employee subject to that policy. See King, 960 F.2d at 624 (stating that “[i]n order to support the ‘judicial evaluation’ (that it is more probable than not that an employer that routinely discriminates, discriminated against a particular individual), there must be significant evidence of the alleged routine”) (emphasis added). It should be noted that “[a] pattern or practice of discrimination may be found even if a defendant does not discriminate uniformly.” United States v. Lansdowne Swim Club, 713 F.Supp. 785, 807 (E.D.Pa.1989) (citations omitted), aff'd, 894 F.2d 83 (3d Cir.1990). However, given the reasons underlying the shift in the burden of proof, there must be a certain level of consistency in the defendant’s discriminatory acts to support a finding of pattern or practice. See Price Waterhouse, 490 U.S. at 266-67, 109 S.Ct. at 1799-1800 (O’Connor, J., concurring) (stating that burden shifts “[b]ecause the class has already demonstrated that, as a rule, illegitimate factors were considered in the employer’s decisions”) (emphasis added).

B. Injunctive Relief, “One-Shot” Event, and Disparate Impact

As stated above, Roche makes alternative arguments in support of the proposition that, even assuming that plaintiffs’ allegations regarding the manner in which Operation Turnabout was conducted are true, these allegations are insufficient to bring this ease within the parameters of a pattern-or-practice ease.

Initially, it is important to set forth plaintiffs’ theory of the ease. In order to effectuate Operation Turnabout, Roche created a document entitled “Summary of Staff Reduction Guidelines” that contained a section called “Selection Guidelines for Staff Reduction Project” [hereinafter the “Guidelines”]. See Miller Aff. Ex. 7. Plaintiffs base both their pattern-or-practice claim and their disparate impact claim on the Guidelines. Specifically, with respect to the disparate impact claim, plaintiffs state that certain policies delineated in the Guidelines “allowed uncontrolled subjective decisionmaking, and because such subjective decisionmaking permits and/or encourages decisionmakers to allow their individual biases to enter into the decisionmaking process, [these policies] could cause a disparate impact on older workers.” See Flamm Aff. Ex. 2 (Plaintiffs’ Response to Interrogatory No. 1). With respect to plaintiffs’ pattern-or-practice claim, plaintiffs state that “[t]he same policy that resulted in a disparate impact on older Roche employees also led to intentional discrimination against older employees” and “Roche executives knew that age bias existed and that uncontrolled decisionmaking would, most likely, result in disproportionate terminations of older employees.” See id. (Plaintiffs’ Response to Interrogatory No. 11(a)).

In other words, plaintiffs’ theory is that Roche gave the line managers who performed the rankings, and thus made the actual termination decisions, complete discretion, see Plaintiffs’ Supplementation Br. at 7 (“Each step along the way furthered Roche’s goal of permitting untrammeled discretion to rule in the implementation of Operation Turnabout.”), and that Roche intended and/or knew that this discretion would result in both conscious and unconscious discrimination against older workers because there was “pre-existing age bias” among Roche’s managers, see, e.g., id. at 4-5 (stating that Roche’s “ad hoc employee evaluation system ... inevitably exploited the conscious and unconscious exercise of age bias among those managers making selection decisions” and that “Operation Turnabout’s selection policies in fact provided the fertile ground needed for the exercise of pre-existing age bias among Roche’s decision-making managers”).

The allegation that Roche intended and/or knew that giving the line managers discretion would result in intentional discrimination is the allegation that distinguishes plaintiffs’ pattern-or-practice claim from their disparate impact claim. However, there is insufficient evidence in the record to support a finding that Roche in fact intended and/or knew that uncontrolled subjective decision-making would result in disproportionate ter urinations of older workers. As the Special Master noted, “[t]here is ... little evidence that Roche systematically violated the ADEA before the date of the RIF or immediately thereafter.” Pattern-or-Practice Opinion at 2. 8 Because there is insufficient evidence to support a finding that Roche systematically violated the ADEA prior to or after Operation Turnabout, there is no circumstantial basis for finding that Roche knew that, if it gave the managers a great deal of discretion in deciding who would be terminated, the managers inevitably would exercise this discretion in a discriminatory fashion.

Also, there is no evidence, or even an allegation, that Roche used the Guidelines either before or after Operation Turnabout. See Plaintiffs’ Supplementation Br. at 4 (“Roche’s ad hoc employee evaluation system, devised and first used for this RIF, became established as the ‘standard operating procedure’ for Operation Turnabout, and that such standard operating procedure both promoted and depended on the anticipated application by Roche managers of age stereotyped thinking to the selection process.” (emphasis added)).

Thus, plaintiffs’ pattern-or-practice claim is essentially that Roche adopted a procedure, solely for use in Operation Turnabout, of giving its managers a great deal of discretion in determining who would be terminated, and this system happened to result in both conscious and unconscious discrimination against Roche’s older workers.

As stated above, Roche argues, in effect, that this claim does not fall within the parameters of a pattern-or-practice claim. For a number of reasons, I agree.

First, not all class actions alleging employment discrimination are pattern-or-practice eases. Teamsters, 431 U.S. at 360, 97 S.Ct. at 1867. Rather, a class action is a pattern-or-practice case only when the plaintiffs claim that “unlawful discrimination has been a regular procedure or policy followed by an employer or a group of employers.” Id. (emphasis added). In other words, in order to prove a pattern or practice of discrimination, plaintiffs must prove that unlawful discrimination is “the company’s standard operating procedure,” Teamsters, 431 U.S. at 336, 97 S.Ct. at 1855, and such standard operating procedure can be proven either by demonstrating a consistent history of discriminatory decisionmaking or by demonstrating that discrimination was the declared policy of the employer. See, e.g., Phila. Bd. of Ed., 911 F.2d at 892-93; 110 Cong.Rec. 14239 (1964).

The defect with plaintiffs’ pattern-or-practice claim in this case is that plaintiffs have not asserted that unlawful discrimination was Roche’s standard operating procedure. Rather, plaintiffs assert that “Roche’s ad hoc employee evaluation system” was the standard operating procedure and that this procedure happened to result in both conscious and unconscious age discrimination. See, e.g., Plaintiffs’ Supplementation Br. at 4 (“Roche’s ad hoc employee evaluation system, devised and first used for this RIF, became established as the ‘standard operating procedure’ for Operation Turnabout, and that such standard operating procedure both promoted and depended on the anticipated application by Roche managers of age stereotyped thinking to the selection process.”); id. at 4-5 (stating that Roche’s “ad hoc employee evaluation system ... inevitably exploited the conscious and unconscious exercise of age bias among those managers making selection decisions”). This is a disparate impact claim and not a pattern-or-practice claim.

Disparate impact claims “involve employment practices that are facially neutral in their treatment of different groups but that in fact fall more harshly on one group than another and cannot be justified by business necessity.” Teamsters 431 U.S. at 336 n. 15, 97 S.Ct. at 1854 n. 15 (citations omitted). Proof of discriminatory motive is not required under a disparate impact theory. Id. The Supreme Court has held that disparate impact claims can be brought, at least pursuant to Title VII, to challenge subjective or discretionary employment practices such as those being challenged by the plaintiffs in this case. See Watson v. Fort Worth Bank and Trust, 487 U.S. 977, 108 S.Ct. 2777, 101 L.Ed.2d 827 (1988). 9 Furthermore, unlike disparate treatment analysis (of which pattern-or-practice analysis is a species, see Teamsters, 431 U.S. at 335 n. 15, 97 S.Ct. at 1854 n. 15, disparate impact analysis addresses the effects of unconscious discrimination in addition to conscious or intentional discrimination. See Watson, 487 U.S. at 990, 108 S.Ct. at 2786-87 (stating that one of the reasons that disparate impact analysis should be applicable to subjective employment criteria as well as objective criteria is that “even if one assumed that any such discrimination can be adequately policed through disparate treatment analysis, the problem of subconscious stereotypes and prejudices would remain”). Thus, plaintiffs’ claim that the use of subjective decisionmaking procedures was Roche’s standard operating procedure during Operation Turnabout, and that such standard operating procedure happened to result in both conscious and unconscious discrimination is a disparate impact claim and not a pattern-or-practice claim.

This conclusion is bolstered by the fact that a number of the cases relied upon by plaintiffs in support of their assertion that “[t]he instant case is premised on substantial precedent derived from numerous Title VII cases involving undisciplined and ill-defined selection criteria coupled with uncontrolled decision-making by individual managers,” see Plaintiffs’ Supplementation Br. at 37, are disparate impact cases that do not discuss pattem-or-praetiee issues. See Crawford v. Western Electric Company, Inc., 745 F.2d 1373, 1384-86 (11th Cir.1984) (affirming district court’s finding that employer’s review system that was based in part on wholly subjective evaluations by white foremen of black employees’ “skill” had a disparate impact on black employees); Robinson v. Union Carbide Corp., 538 F.2d 652 (5th Cir. 1976) (holding, among other things, that promotion system for salaried employees that relied on supervisor’s subjective opinion concerning job candidates’ “adaptability,” “bearing, demeanor, manner,” “verbal expression,” “appearance,” “maturity,” “drive,” and “social behavior” violated Title VII because “[s]uch high-level subjectivity subjects the ultimate promotion decision to the intolerable occurrence of conscious or unconscious prejudice”; although the court did not specifically mention the term “disparate impact”, its based its holding on the seminal Supreme Court case on disparate impact, Griggs v. Duke Power Co., 401 U.S. 424, 431, 91 S.Ct. 849, 853-54, 28 L.Ed.2d 158 (1971), wherein the Supreme Court stated that Title VII “proscribes not only overt discrimination but also practices that are fair in form, but discriminatory in operation”), aff'd in part, rev’d in part on other grounds and remanded on reh’g, 544 F.2d 1258 (5th Cir.), cert. denied, 434 U.S. 822, 98 S.Ct. 65, 54 L.Ed.2d 78 (1977); Wade v. Mississippi Cooperative Extension Service, 528 F.2d 508, 518 (5th Cir. 1976) (holding, based on Griggs, that evaluation form used for evaluating professional workers violated Title VII where “the questions on the evaluation form were in part subjective and vulnerable to either conscious or unconscious discrimination by the evaluating supervisors,” “the evaluation scores themselves were not consistently used as a basis for ... promotion,” and “the defendants wholly failed to make a showing that the test was substantially related to the particular job of the individual being evaluated”); Rowe v. General Motors Corporation, 457 F.2d 348, 358-59 (5th Cir.1972) (holding, relying on Griggs, that procedure for promotion from hourly job to salaried job violated Title VII where (1) the foreman’s recommendation was the single most important factor in the promotion process, but the foreman was not given written instructions pertaining to the qualifications necessary for promotion; (2) the standards which were determined to be controlling were “vague and subjective;” (3) hourly employees were not notified of promotion opportunities or of the qualifications necessary to get the job; and (4) there were no safeguards in the procedure to avert discriminatory practices).

Other cases relied upon by the plaintiffs merely stand for the propositions that plaintiffs may challenge employment actions under both disparate impact and disparate treatment theories in the same case and that statistical evidence of a disparate impact is also relevant evidence regarding whether an employer engaged in intentional discrimination. See Mozee v. American Commercial Marine Service Company, 940 F.2d 1036, 1042 (7th Cir.1991); Green v. USX Corporation, 896 F.2d 801, 807 (3d Cir.), cert. denied, 498 U.S. 814, 111 S.Ct. 53, 112 L.Ed.2d 29 (1990). However, this does not mean that all disparate impact claims are also ipso facto pattern-or-practice claims. Rather, a pattern-or-practiee claim exists only when the plaintiffs claim that “unlawful discrimination has been a regular procedure or policy followed by an employer or a group of employers.” Teamsters, 431 U.S. at 360, 97 S.Ct. at 1867 (emphasis added). For example, in Mozee, in addition to their claim that the defendant’s subjective promotion and discipline system had a disparate impact on black employees, plaintiffs also specifically claimed that the defendant pursued “a general policy of [intentional] discrimination,” i.e., a pattern or practice of discrimination. Mozee, 940 F.2d at 1042. In Green, in addition to their claim that the defendant’s multi-component hiring system that utilized subjective criteria had a disparate impact on black applicants, see Green, 896 F.2d at 803, the plaintiffs also asserted a disparate treatment claim that was “based on the theory that [defendant’s] use of relevant labor market data, instead of applicant flow data, to set its minority hiring goals constituted the intentional setting of a discriminatory quota to limit black hires.” Green v. United States Steel Corp., 570 F.Supp. 254, 277 (E.D.Pa.1983). Unlike the plaintiffs in Mozee and Green, the plaintiffs in this case claim only that the use of subjective decisionmaking procedures was Roche’s standard operating procedure during Operation Turnabout. No where is it alleged the Roche’s standard operating procedure was intentional discrimination. Thus, plaintiffs’ are only asserting a disparate impact claim and not a pattem-orpractice claim.

Furthermore, the consequences of finding that plaintiffs’ claim fits within the pattern- or-practice framework would be that anytime a company gives managers discretion to make employment decisions that company potentially engages in a pattern or practice of discrimination. This is because anytime managers are given discretion they have the opportunity to exercise that discretion in a discriminatory manner. Thus, some may discriminate, while others may not. However, concluding that this situation, without more, is a pattern or practice would bring within the definition of a pattern' or practice employment practices that were not intended to be there. This is because a decision by a company to give managers the discretion to make employment decisions, and the subsequent exercise of that discretion by some managers in a discriminatory manner, is not tantamount to a decision by a company to pursue a systematic, companywide policy of intentional discrimination, i.e., a pattern or practice of discrimination. Rather, the exercise by some managers of their discretion to discriminate would constitute “the mere occurrence of isolated or ‘accidental’ of sporadic discriminatory acts,” which is not a pattern or practice of discrimination. Teamsters, 431 U.S. at 336, 97 S.Ct. at 1855.

I note that if a company gave managers the discretion to make termination decisions only as a ruse to conceal a systematic, discriminatory policy, then that company could be found to have engaged in a pattern or practice of discrimination. However, there is no such allegation in this case. Plaintiffs simply claim that “Roche’s ad hoc employee evaluation system” was the standard operating procedure and that this procedure happened' to result in both conscious and unconscious age discrimination. Therefore, accepting this claim as a pattern-or-practice claim would broaden the meaning of pattern or practice beyond that explained by the Supreme Court in Teamsters and the subsequent caselaw.

Another reason that plaintiffs’ claim does not fall within the framework of a pattem-orpractice case is that the employment practice which plaintiffs assert was Roche’s standard operating procedure was used only once, i.e., the Guidelines were used only during Operation Turnabout. Thus, because the Guidelines were used only during Operation Turnabout, and there is no evidence in the record to support a finding that Roche may use them in the future, there does not appear to be any discriminatory policy that the court should exercise its broad equitable powers to eliminate. This is because the “discriminatory” policy, i.e., the Guidelines, is no longer in effect and was in effect only once. Thus, because Operation Turnabout was a one-shot event, there is no basis to award classwide prospective injunctive relief — the main reason for bringing a pattern-or-praetice claim. Therefore, Roche’s arguments, that plaintiffs’ claim must be dismissed because pattern-orpraetice claims are only appropriate where the class plaintiffs seek to enjoin the defendant from engaging in existing or threatened discriminatory behavior and because a “one-shot” event cannot constitute a pattern or practice of discrimination, are also persuasive. This is another reason why this case does not fall into a classical pattern-or-practice framework.

In summary, plaintiffs allege that the ad hoc employee evaluation system was Roche’s standard operating procedure, that it was devised and first used for Operation Turnabout, and that it afforded managers the opportunity to act on their own individual biases. This is not a pattem-or-practice claim because plaintiffs are not claiming that age discrimination was Roche’s standard operating procedure and that such procedure happened to result in discrimination. Furthermore, because Operation Turnabout was a one-shot event, there is no basis to award class-wide prospective injunctive relief — the main reason for bringing a pattern-or-practice claim.

For these reasons, I find that plaintiffs’ claim is not within the parameters of a pattem-or-practice claim. Rather, plaintiffs are attempting to re-characterize a disparate impact claim as a pattern-or-practice claim. Therefore, I find that, as a matter of law, plaintiffs have not asserted a viable pattem- or-practice claim and Roche’s motion for summary judgment on plaintiff’s pattem-orpractice claim is granted.

TV. Pattem-or-Practice Evidence

Although the conclusion that plaintiffs are asserting only a disparate impact claim and not a pattem-or-practice claim is a sufficient basis on which to grant Roche’s motion for summary judgment, I also conclude in the alternative that, even assuming that plaintiffs’ claim could be construed as a pattem-or-practice claim, no reasonable jury could find, based on the evidence submitted by plaintiffs, that Roche engaged in such a pattern or practice of discrimination in conducting Operation Turnabout. Prior to discussing why this is so, I will set forth the evidence now in the record in the light most favorable to plaintiffs.

A Plaintiffs’Evidence

The evidence plaintiffs submitted in support of their pattern-or-practice claim can be divided into three areas: (1) evidence relating to the planning and execution of the RIF, i.e., Operation Turnabout; (2) expert reports that analyze Operation Turnabout; and (3) anecdotal evidence of alleged age-bias among Roche executives prior to Operation Turnabout.

1. Operation Turnabout

i) Development of the Plan

On August 1,1984, Roche’s then-president, Irwin Lemer, announced that Roche would undertake a comprehensive study of Roche’s operations. See Miller Aff. Ex. 10. This study was the initiation of what came to be known as Operation Turnabout.

The goal of the study was to determine how Roche could improve its profitability by, among other things, saving $50 million in annual expenses. See id. To assist in the study, Roche retained the management consulting firm of Cresap, McCormick & Paget. In addition, a steering committee and task forces, consisting of Roche personnel and the outside consultants, were formed to study Roche’s various divisions, and thereafter, make recommendations for a revised organizational structure. See Lemer at 40-41. 10 In developing the revised organizational structure, the main objective of the task forces was to find ways to reduce costs toward the goal of achieving $50 million in cost savings. President Lemer was aware that this “would undoubtedly lead to staff reductions.” See Lemer at 41.

The study took approximately five months to complete — from August 1, 1984 until December 31, 1984. At the completion of the study, the various task forces submitted binders with written recommendations to President Lemer and also made oral presentations in late December 1984 and early January 1985. The task forces recommended various ways to save money, including a significant reduction in force (“RIF”).

On January 2, 1985, Lemer appointed an “Implementation Task Force” (“ITF”), headed by Stanley Rosen, to develop a plan to carry out the recommendation of the various task forces who performed the studies. The ITF appointed “Implementation Coordinators” to serve as liaisons between the ITF and the line managers. See Rosen at 64. One Implementation Coordinator was appointed per division. Id.

Furthermore, from January 4 until January 15, 1985, the ITF, working with the Human Resources Department, specifically Michael Higgins, who was the director of Roche’s equal opportunity department, developed a plan to effectuate the implementation. On January 17, Lemer approved the final recommendations of the task forces and communicated such recommendations to the executive committee. Thereafter, the actual implementation of the plan began.

ii) The Guidelines

The process by which managers were to decide which employees would be fired is set forth in a section of a document prepared by Higgins and Mary Rose Swift, one of Roche’s in-house attorneys. The document is entitled “Summary of Staff Reduction Guidelines” and the relevant section of the document is entitled “Selection Guidelines for Staff Reduction Project.” See Summary of Staff Reduction Guidelines [hereinafter “Guidelines”] (attached as Exhibit 7 to Miller Affidavit). The Guidelines, which were completed in early January 1985, were intended to serve three functions: (1) to describe the staff reduction process to the managers making the termination decisions; (2) to provide a recommended, consistent methodology for all of the managers to use when selecting who would be terminated; and (3) to ensure that the managers making the decisions were cognizant of the various laws prohibiting employment discrimination. See Swift at 19.

In a prefatory section, the Guidelines specifically provide that the criteria on which the termination decisions are made “must be job-related, and in accordance with legal requirements.” See Guidelines at E002277. Then, in a section entitled “The Staff Reduction Process”, the Guidelines describe a two-step process for implementing the RIF.

Pursuant to step one, employees whose jobs were eliminated, were to be terminated on that basis. Id.

The second step of the staff reduction process “requires the involvement of supervisors and managers who must make individual staffing decisions within their particular work units.” Id. Thus, the manager or supervisor who'was directly in charge of the work group would make the initial decision regarding which employees would be terminated. These decisions, however, “require the approval and signature of the reviewing supervisor.” Id. at E002278. In addition, “[a]ll decisions will be reviewed by the appropriate Executive Committee member.” Id.

Thereafter, the Guidelines discuss the methodology and criteria to be used in determining who would be terminated. Justification for all terminations was required in writing, on a standardized form. On the standardized form

[a] brief narrative statement is required in response to the question: “Why was this employee selected for termination?” In some cases, the answer to that question may be simply, “This employee’s specific job (or function) was eliminated due to newly defined business requirements.” In other cases, employees will be terminated on the basis of factors directly related to their ability to perform in a job under review, as compared to other employees under consideration for the same job.

Guidelines at E002278. The suggested factors included skills and abilities, quality of work, knowledge, education, training, productivity, work habits, overall performance, versatility, flexibility, work experience, attendance, punctuality, “unique, critical capability”, “certification or licensure requirements,” and “other job factors as applicable.” Guidelines at E002278. The Guidelines then provided sample narrative statements.

The result of the evaluation of employees with respect to the above factors was to be a ranking of the employees, with respect to one another, within a given job. Termination decisions were to be based on the employees’ respective rankings. “In eases where overall assessments are the same, company length of service will be used to differentiate between the employees retained or terminated.” Id. at E002279.

The next section of the Guidelines contain a list of “Cautions.” The listed “cautions” included “[a]ge should not be considered as a factor in any decision” and “[pjension and retirements eligibility should not be considered. Such decisions could tend to discriminate against longer-service and likely older employees.” Guidelines at E002279-E002280.

Next, the Guidelines discuss the various forms that must be completed for each termination. For each terminated employee, the standardized form discussed above, which is entitled “Justification for Staff Reduction Selection Form” [hereinafter “Justification Form”], and an “Employee Profile” had to be completed. The Guidelines state that the “critical information” that would be entered on the Justification Form includes the reasons for termination, the signature of the supervisor completing the form, and the signature of the next higher level supervisor reviewing and concurring with its contents. See Guidelines at E002280. Sample Justification Forms, which included examples of possible justifications, were included in the Guidelines. The Employee Profile contained information regarding information such as number of accrued, unused vacation days and the number of weeks severance. See id. at E002281.

After these forms were completed by the relevant manager or supervisor and the next higher level supervisor, they were to be forwarded to the appropriate executive committee member. “Both forms will be reviewed for completeness and approved by the Executive Committee member.” Id. Thereafter, the forms would be forwarded to the implementation coordinator for the relevant division.

iii) The Selection Process

The Guidelines were first distributed during a series of meetings, each of which lasted for approximately 30 to 45 minutes, on January 21 and 22. At the meetings, Higgins and Swift distributed and discussed the Guidelines. 11 See Higgins ’94 at 62; Swift at 42. They also discussed various laws against employment discrimination. See Higgins ’94 at 62; Swift at 42.

A meeting was held for each division. Higgins ’94 at 57. However, attendance at these meetings was limited to the division heads, key managers, and implementation coordinators. See Higgins ’94 at 59-60. Thus, not every manager who would be involved in making the termination decisions attended the meetings. In fact, only 30% of the 215 managers or supervisors and higher level supervisors who signed the Justification Forms attended these meetings. See Miller Aff. at ¶ 7, Ex. 6. The managers who attended the meetings were supposed to disseminate the Guidelines to the lower level managers who would be doing the rankings of employees, and thus, making the initial termination decisions. See Swift at 23-24. However, there was no formal mechanism for distributing the Guidelines. See Higgins ’94 at 63; Swift at 24.

It appears, perhaps in part as a result of the lack of a formal procedure to distribute the Guidelines to all the managers making the ranking and termination decisions, that a number of such managers did not specifically use the Guidelines when the rankings and termination decisions were being made. This conclusion is supported by the following evidence in the record.

First, of the thirty-one managers or supervisors deposed in this ease who did rankings of their subordinates for the purposes of Operation Turnabout, twenty-one do not recall seeing the Guidelines. See Anton at 11, 27, 56; Berdux at 11-12, 62-63; Bills at 6, 22-23; Bledsoe at 6, 76; Daly at 9, 41-42; Data at 7, 52-53; Diller at 9, 46; Evangelista at 10, 55-56; Frenkl at 11-12, 24; Fryer at 13, 73-74; Koch at 7, 133-35; Kuntzman at 11,112-13; Lorentz at 18, 38-39; Oriel at 10, 36-37, 66-68; Sail at 10,123-24; Sheridan at 7, 59-61; Skalaban at 12-13, 33; Takahashi at 9-10, 77-78; Tsehopp at 8, 23; Weigele at 6, 80, 95; Zambrano at 9, 42-44.

Second, although the process of selecting individuals for termination was supposed to begin after the January 21 and 22 meetings where the Guidelines were first distributed, see Higgins ’94 at 55-56, a number of managers made their termination decisions prior to the Guidelines being distributed. See Anton at 51-52; Evangelista 52-53; Frohlieh at 77; Koch at 133-35; Sail at 47-57; Sunbury at 107-09. Some of these managers, however, reviewed their decisions after receiving the Guidelines or discussed the Guidelines with their superiors prior to performing their evaluations. See, e.g., Frohlieh at 77.

Third, some supervisors, rather than distributing copies of the Guidelines to the persons who would actually do the rankings, orally instructed their managers to rank their subordinates. This resulted, in some instances, in the rankings being done based on various, subjective criteria. See, e.g., Anton at 27, 56 (when told to do rankings was not given any criteria, therefore did rankings based on “performance”); Berdux at 40-41 (told by supervisor to compile “a list where you put the most competent at the top of the list and the least competent at the bottom”); Evangelista at 56 (told to rank employees based on performance, so judged performance based on “gut feeling”); Lorentz at 35 (told by supervisor to “rank the individuals based on peer ranking from best to worst, irrespéctive of years of service with the company or prior record with the company as to other areas in which they had been operating”); Meinhardt at 33 (stating that the “criteria were impressions, ... [t]here were no precise criteria”); Minafri at 22 (asked his managers “to force rank their personnel based upon skills, skills necessary to do the job”); Peterson at 72-73 (“subordinates were ranked from most desirable to least desirable of value to the department” where “most desirable” meant “that which can contribute or offer the most to the department in his performance” and “least desirable, that individual which would be of minimal value to the department in the long run”); Sail at 66 (told to do rankings by “gut feel”); Vanderwerff at 32-33 (stated that he ranked employees by how he “felt they were keeping up with the knowledge of the area that they were responsible for”, that “it was basically a judgment call”, and that he could not “pin down any specifics” with respect to exact criteria used).

In addition to a number of managers not being given specific criteria for ranking their subordinates, it appears that most managers doing the rankings did not review their subordinates’ prior written, annual performance review or annual merit ratings in conjunction with doing the rankings. (Plaintiffs assert that the merit ratings were objective, rehable measures of an employee’s past performance, and should have been used in performing the rankings. See Plaintiffs’ Supplementation Br. at 25-27) Roche concedes that the employees’ prior written annual performance reviews or annual merit ratings were relied on in making the termination decision for only 51 of the 334 plaintiffs who were terminated based on rankings. Miller Aff. ¶ 6, Ex. 5.

Lastly, although the Guidelines specifically provide that an employee whose job or function is eliminated should be terminated on that basis, in a number of instances, such employees were not terminated on that basis, but rather were considered for other jobs. See, e.g., Frohlieh at 54-55; Haid at 34; Koch at 56-57; Matriseiano at 89-92; Minafri at 45-47; Sheridan at 38-39; Sunbury at 85-86; Weigele at 82-83.

Thus, it is clear that some managers did not rely on the Guidelines, or at least did not completely follow the Guidelines, when performing the rankings and making the decisions regarding who would be terminated.

Furthermore, as stated above, the Guidelines “caution” that pension and/or retirement eligibility should not be a factor in any decision. In some situations, however, pension and/or retirement eligibility did factor into the decisions.

One department in which pension and retirement eligibility factored in the termination decisions was in the Chemical Production Department (“CPD”). In that department, the managers who ran it were “charged with reducing the number of people in chemical production,” Meinhardt at 23, by cutting approximately 126 of the 644 employees in the department at the Nutley facility, see Miller Aff. Ex. 11. According to A1 Meinhardt, who was a manager in the CPD, he and the other managers attended meetings where each manager would give a “synopsis” or a “snapshot” of the performance of the employees in their group. Meinhardt at 32. In giving this synopsis of their employees, the managers used “no precise criteria,” but rather just gave their “impressions” of the employees. Id. at 33. Meinhardt stated that, when ranking employees, he considered

family situations, and how old a guy is and where his kids are and his financial circumstances to the best you know it, and it all falls into the judgment, and that’s what it is. It’s a subjective judgment when you come down to it.

Meinhardt at 76. Meinhardt also stated that when performing the rankings he asked himself

“[wjould I want to keep somebody who is young and energetic and has got potential as opposed to somebody who has been around for whatever, 20, 25, 30 years,” and even though their performance currently meets expectations in accordance with the performance standards that we had____ It’s going to be a leaner tougher organization, expectations are going up.

Meinhardt at 45 (attached as Ex. 5 to Flamm Aff. which is attached as Exhibit 3 to Appendix of Hoffmann-La Roche Inc. on Appeal from Opinion of Special Master Dated November 28, 1994 Denying Partial Summary Judgment and Related Relief).

In addition, when the CPD managers were ranking their employees, they were working with documents on which the ages of the employees were listed. Meinhardt at 231. Thus, the managers “certainly were aware how old people were at the time [they] did this procedure.” Id.

Other managers in the CPD testified that proximity to retirement and/or eligibility for retirement benefits were specifically discussed. Fedor Frenkl, an associate manager who reported to Meinhardt, stated that

we were not as much concerned about people who were at the advanced age because they were covered by pension and suffering for these people would be less rather than losing their job than the younger people.

Frenkl at 28. Furthermore, Frenkl stated that an employee named Bob Czesniewski, a foreman, was terminated because “[h]e was at age 62 or 63 and [they] thought it wouldn’t be problem for him to leave” because “[i]t was just an earlier retirement.” Id. at 29. Thomas Anton, another associate manager in the CPD, stated that he, along with two other managers (Petrie and Kochling), considered the age of Joe Zelauskas when deciding to terminate him. Specifically, Anton stated that the three managers

felt that Joe Zelauskas was acceptable and better in the performance of his job but we felt that he would be the least hurt by being let go because he was somewhat near retirement anyway.

Anton at 28-29. Anton also stated that the ages of Mike Matta and A1 Papio were considered when they were terminated. Specifically, Kochling told Anton and Petrie that Matta and Papio “are near retirement age and, therefore, if we put them kind of low on the list it might save a younger guy a job if push came to shove.” Id. at 85.

In the Chemical Research Department, Edmund Tschopp, who was a supervisory foreman in the department at the time of Operation Turnabout, testified that although no one told him to use nearness to retirement as a factor in determining who should be terminated, he nonetheless did. Tschopp at 16-21. He believes that nearness to retirement is the major reason for the terminations of two of his subordinates, Theodore Gaida and Anton Wanio, because that was the reason he recommended them for termination to his supervisor. See Tschopp at 74-81.

Proximity to retirement was also considered as a factor in at least a part of the Research Department. Dr. Rodney Fryer, who was the director of medical chemistry at the time of Operation Turnabout, testified that where the employees he was ranking could not be differentiated based on performance, he took into account proximity to retirement. Fryer at 60, 62. He took this into account because he “was trying to find a way that would make the least impact on somebody’s career, somebody’s future” and not because he believed that if someone were close to retirement, then they were less competent. Id. at 62-63. Dr. Fryer further stated after he told his supervisor, Dr. Weigele, that it was difficult to perform the rankings without considering age, Dr. Weigele told him that he was well aware of who in the department was close to retirement and also gave Fryer a list of the employees with their birth dates on it. Id. at 67-69. In summarizing how he performed his rankings, Dr. Fryer testified that

when you have to play God and strike people from the register and everything else being equal, you try to do it with as little pain as possible; and age was a very useful factor in that for me.

Id. at 69.

iv) Documentation, Review, and the Carrying Out of the Decisions

After it was decided which employees would be terminated, the managers were required to provide written reasons for their termination decisions on the Justification Form. The Justification Form provided five blank lines on which managers were directed to “[pjrovide a brief, concise statement justifying th[e] decision.” See Swift at 20-21; Miller Aff. Ex. 7. In addition, neither the Justification Form nor the Guidelines required the managers filling out the form to identify the employees against whom the terminated employee was compared. See Guidelines (attached as Exhibit 7 to Miller Affidavit). Some managers, who, although they were subsequently fired in Operation Turnabout, had participated in the ranking process, see Anton at 27-30; Berdux at JO-41; Frenkl at 23-24, 28-29; Meinhardt at 32-33; Sail at 65-66; Skalaban at 29-30, 33, were told by their superiors what reasons for termination should be written on the Justification Forms, see Anton at 90; Berdux at 63-64; Capomaggi at 33-34; Frenkl at 49-50; Meinhardt 59-65; Sail at 120-23; Skalaban at 47-49.

The Justification Forms have a section for a “Higher Level Supervisor/Manager reviewing and concurring with th[e] decision” to sign. However, it appears that in some situations the higher level supervisor who signed the form did not substantively review the decisions that had been made. See, e.g., Muller at 52-53; Sheridan at 69-79. In addition, after the higher level supervisor signed, the form was sent to the appropriate executive committee member for signature. However, Stanley Rosen, who was head of the ITF, stated that executive committee members were given no instructions regarding how they should review the Justification Forms prior to signing. Rosen at 151-52. Executive committee member Ronald Kuntzman stated that he “sign[ed] with the assumption that [the line managers and higher level managers] were doing their job properly.” Kuntzman at 133. In addition, no executive committee member made any changes to or disagreed with any of the Justification Forms presented to them. Rosen at 152. Therefore, it appears that neither the higher level supervisor nor the executive committee member who signed the Justification Form reviewed the reasons why persons were being terminated.

After the executive committee member signed the Justification Forms, the forms were sent to the implementation coordinators. The implementation coordinators’ function was to (1) compile lists of who was being terminated, who was being retained, and who was being transferred, and (2) ensure that the Justification Forms, and the other documentation, were complete and bore the requisite signatures. See Rosen at 153. There is a conflict in the record as to whether the implementation coordinators were to review the reasons given for each termination. Roe-co Ricciardi, who was an implementation coordinator, stated that he did not believe that it was his job to review the reasons given on the forms for termination, and therefore, he did not closely review the reasons given. Ricciardi at 52. Rosen, however, testified that the implementation coordinators were responsible for ensuring that the terminations reflected on the Justification Forms were consistent with the Guidelines. Rosen at 153. Rosen further stated, however, that no implementation coordinator made any changes in the employees selected for termination. Id. Thus, it appears that the review function performed by the implementation coordinators, if any, was limited.

After the implementation coordinators completed their review, the forms were sent to Swift and Higgins for review. Swift merely reviewed the Justification Forms to ensure that the forms were completed and that the “justification read, ..., reasonably correctly.” Swift at 60. “It was not an intensive review.” Id. Higgins cheeked the justification forms to determine whether the general language on the form was consistent with the Guidelines. Higgins ’94 at 99.

After the completion of the Justification Forms, Roche conducted statistical, demographic analysis based on the recommended terminations. This analysis was completed by January 28,1985. See MiUer Aff. Ex. 9 at B010722; Higgins ’89 at 58-60. Roche has claimed the attorney-client privilege with respect to the results of this analysis. Higgins ’94 at 110-12. In any event, prior to the completion of the study, Roche had determined that regardless of the outcome, no managers would be asked to revise their selections for termination based in the statistical analysis. Rosen at 186-87, Ex. 6. Such revisions would be requested only if the supporting documentation indicated some “shortcoming.” Rosen Ex. 6. After the completion of the study, no decisions were changed. See Rosen at 186-87. Lastly, Roche did not have any contingency plans to delay the RIF if the study revealed any problems. Higgins ’94 at 112.

As previously stated, on February 4, 1985 Roche terminated over 1,100 employees at its Nutley, N.J. and Belvidere, N.J. facilities. Of the over 700 Roche employees over the age of 40 who were terminated in Operation Turnabout, 476 have joined this lawsuit as plaintiffs.

2. Expert Reports

In support of their claim that, when terminating these employees, Roche engaged in a pattern or practice of discrimination, plaintiffs submitted, in addition to the evidence just described, two expert reports. First, plaintiffs submitted the report of Dr. Farrell Bloch in which he discusses various statistical analyses that he performed generally addressing the question of whether a disproportionate number of employees over 40 years of age were terminated. Second, plaintiffs submitted the report of Dr. Joel Lefkowitz, an industrial and organizational psychologist, who was asked to opine on the cause of the disparities allegedly demonstrated in Dr. Bloch’s report, i.e., whether the age disparities were caused by age discrimination. 12

i) Report of Dr. Farrell Bloch

The statistical analyses presented in Dr. Bloch’s report address the following question: “If underlying termination rates were the same for older and younger workers, then what is the likelihood of observing the actual pattern of termination by age?” See The Relationship between Termination Rates and Age in Operation Turnabout at Hoffmann-La Roche, Inc. at 1 (attached as Exhibit 1 to Miller Affidavit) [hereafter “Bloch Report”].

In answering this question, Dr. Bloch performed a number of different statistical analyses, most of which are reflected in four tables located at the end of the report. Table 1 compares “the termination rates for young and old workers at Hoffmann-La Roche as of the close of business on February 3, 1985 (the day before the reduction in force).” Bloch Report at 2. As reflected in the table, 18.4% of the 4,301 employees at least 40 years of age were terminated, while only 9.3% of the 3,480 employees under age 40 were terminated. The difference between the 18.4% termination rate and the 9.3% termination rate is the equivalent of 11.45 standard deviations. Standard deviation is “a measure of predicted fluctuations from the ejected value of a sample.” Hazelwood School District v. United States, 433 U.S. 299, 309 n. 14, 97 S.Ct. 2736, 2742 n. 14, 53 L.Ed.2d 768 (1977). In other words, standard deviation may be used to measure the likelihood that the number of persons, at least 40 years old, actually terminated in Operation Turnabout is the result of something other than age discrimination. The Supreme Court has stated that “[a]s a general rule for such large samples, if the difference between the expected value and the observed value is greater than two or three standard deviations,” then the hypothesis that discrimination was not present is suspect. Id. (quoting Castaneda v. Partida, 430 U.S. 482, 496-97 n. 17, 97 S.Ct. 1272, 1281-82 n. 17, 51 L.Ed.2d 498 (1977)). Table 1 also compares termination rates based on other age dividing lines — employees at least 45, 50, 55, 65 years old. With these dividing lines, the standard deviation of the disparity in termination rates for younger and older workers range from 7.01 (where 65 years of age is the dividing line) to 14.60 (where 55 years of age is the dividing line).

In Table 2, Dr. Bloch compares termination rates for employees under 40 years of age and employees at least 40 years of age by division. In Divisions 1, 12 and 14, the disparities in termination rates are the equivalent of 1.0, 0.0, and 1.6 standard deviations. In addition, the disparity in termination rates for Division 8 is equivalent to 2.2 standard deviations. Therefore, for these four divisions, the disparity in termination rates is not particularly, statistically significant. The differences in termination rates in divisions 3, 5, 6, and 7 are more significant. In these divisions, the disparities in termination rates range from 4.3 standard deviations (division 7) to 8.1 standard deviations (division 5).

Table 3 compares the “increased termination probabilities for employees at least forty years old (with other factors held constant).” Bloch Report at 9. By holding constant factors that may influence termination decisions, Dr. Bloch sought to isolate the role age played in termination decisions. Dr. Bloch identified four factors to be held constant — “comporatio”, “percent of range”, length of service, and division. The compo-ratio and percent of range relate to an employee’s position in their salary range. 13 Dr. Bloch first held the factors constant individually, and then in various combinations. This resulted in disparities in termination rates ranging from 4.47 standard deviations (both comporatio and length of service held constant) to 11.38 standard deviations (division held constant). See Bloch Report at Table 3.

Dr. Bloch also compared termination rates for employees at least 40 years old while holding merit ratings constant. Arthur Bledsoe, Director of Compensation and Benefits at Roche at the time of Operation Turnabout, stated that “the Merit system compared employees to all others with similar duties and responsibilities.” Bledsoe at 49. Employees would then be rated in one of four categories: exceeds expectations, meets all expectations, meets most expectations, and does not meet expectations. Rosen at 113; Miller Aff. Ex. 13. Compensation was then based on these ratings. Bledsoe at 49. Dr. Bloch used the merit ratings as a proxy for performance. Although Dr. Bloch did a number of different tests using the merit ratings, his conclusions with respect to all of these tests can be summarized as follows: “Although the results show that termination rates are higher for those with poorer performance, termination rates are still higher for older workers, holding performance constant.” Bloch Report at 5.

Finally, as discussed above, Roche claims that the employees terminated in Operation Turnabout were terminated on the basis of how they were ranked against the other employees in their work group. Roche provided plaintiffs with pools of people who they claim were compared to each other in Operation Turnabout. The pools comprise 1697 of the 7781 Roche employees at Roche employee prior to Operation Turnabout and 385 of the over 1,100 terminated employees. Although the results are not reflected in a table, Dr. Bloch performed statistical analyses on these pools. See Bloch Report at 5-6. After taking into account the pools, the disparity in likelihood of being terminated between those older and younger than 40 years is the equivalent of 5.20 standard deviations. When last merit ratings are held constant, the resulting disparity is equivalent to 3.90 standard deviations for pools 1-115, 2.22 standard deviation for pools 117-143, and 4.62 standard deviations for the combined group. 14 When average merit ratings over the period 1982-84 are held constant, the respective standard deviations are 4.79, 2.44, and 5.48 standard deviations. See Bloch Report at 5-6.

Based on all of the analyses discussed above, Dr. Bloch concluded that “after analyzing Operation Turnabout in many different ways, holding constant various factors, and using different age dividing lines, I have consistently found that older employees are significantly more likely than younger employees to have been terminated.” Bloch Report at 6.

ii) Report of Dr. Joel Lefkowitz

As stated above, plaintiffs also submitted the report of Dr. Joel Lefkowitz, an industrial and organizational psychologist, who was asked to opine on the cause of the disparities allegedly demonstrated in Dr. Bloch’s report, i.e., whether the age disparities were caused by age discrimination. The “starting point” of Dr. Lefkowitz’s report was the “finding” by Dr. Bloch that Operation Turnabout had a “disparate impact” on older employees. See Critique of the Implementation of the Hoffmann-La Roche, Inc. Cost Reduction Plan “Operation Turnabout” at 2 (attached as Exhibit 4 to Miller Affidavit) [hereinafter “Lefkowitz Report”]. Dr. Lefkowitz “construed [his] task to be an evaluation of whatever evidence may be brought to bear concerning the likely explanation(s) or justification or that disparate impact.” Id.

Dr. Lefkowitz considered four possible explanations for the disparate impact that Operation Turnabout allegedly had on older workers. First, he considered whether the process by which Roche employees were se-„ lected for termination was ‘Valid.” Lefkowitz Report at 3. To determine whether this process was valid, Dr. Lefkowitz compared it to a number of “the professionally- and legally-accepted strategies for validating personnel procedures.” Id. He concluded that “[t]he employee evaluation ranking system employed by Hoffmann-La Roche as a basis for making the termination decisions fails to meet professional and legal standards of job-relatedness for employee performance appraisals.” Id. at 47. According to Dr. Lefkowitz, the system used by Roche did not meet these standards of “job-relatedness” because it was “ad hoc in nature.” See Lefkowitz Report at 7-10. It was “ad hoc in nature” because, among other things, “it was not developed in a systematic documented fashion so that its content would reflect job-related dimensions of work performance,” id. at 8, “none of the factors [listed in the Guidelines], including the more abstract and vague ones (versatility, flexibility, work habits) are defined in terms of the work behaviors that presumably reflect those attributes — for each applicable job,” id., “no instructions are provided concerning the relative ‘weighting’ of those factors nor how they are to be combined into a composite overall evaluation by which each individual would be ranked,” id. at 9, “no training or formal procedures were provided many managers and supervisors who would make the ratings and rankings,” id., “no empirical investigation was made of the reliability or consistency of the rankings (i.e. if repeated, would the same rankings result?),” id. at 10, and “[n]o appeals process was provided.” Id. Dr. Lefkowitz opined that “[a] professionally developed and legally acceptable performance appraisal system would not have had all of those omissions.” Id. Therefore, he concluded that “the disparate impact of the HLR terminations on older employees can not appropriately be ‘justified’ on the basis of the job-relatedness of the appraisal process.” Id. at 23.

Second, Dr. Lefkowitz considered whether age and job performance at Roche were inversely correlated, thereby justifying a positive correlation between age and termination in Operation Turnabout. Dr. Lefkowitz first noted that the Bloch Report found, using the Merit Ratings as a proxy for performance, that although termination rates are higher for those with poorer merit ratings, termination rates are still higher for older workers, holding performance constant. See Lefkowitz Report at 24. Then he discussed “a brief review of the behavioral science literature concerning the relationship between age and job performance.” Id. at 24. After reviewing this literature, he concluded that “age is not generally related to performance across a wide variety of jobs,” see id. at 32, and therefore, there is no reason to believe that age and job performance at Roche were inversely correlated, thereby, justifying a positive correlation between age and termination in Operation Turnabout.

Third, Dr. Lefkowitz considered whether “the disparate impact reflects, to some degree, unplanned (perhaps unconscious) discrimination in the form of age-stereotyping or age-bias.” Id. at 4. In considering this issue, Dr. Lefkowitz reviewed several studies that resulted in findings that differential treatment of older and wronger workers was the result of “unconscious age stereotyping rather than conscious discrimination.’’ See Lefkowitz at 40-41. Based on these studies, he concluded that

(i) age-stereotyping is widespread — perhaps especially so in the business world, (ii) it can be manifested in personnel actions such as terminations, (iii) it is more likely to occur when employees are compared directly with one another as was done by HLR, (iv) it is more likely to occur when not a lot of employee performance data or job description information is provided, as was true for many of the HLR rankings; and (v) researchers recommend that strong assertive steps be taken by management in order to prevent its occurrence — -which HLR appears not to have done.

Lefkowitz Report at 43. Therefore, according to Dr. Lefkowitz, “in the absence of support for either of the first two possibilities, the likelihood that the differential impact of the HLR RIF is due to the expression of age bias among those making the termination decisions must be seen as great.” Id. at 49.

Lastly, Dr. Lefkowitz noted that the disparate impact that Operation Turnabout had on older employees may be explained by intentional discrimination. However, with respect to the existence of intentional discrimination, Dr. Lefkowitz concluded that he was “not in a position to pass definitive judgment on this possibility” without knowing more about “the structure, functioning, history and culture” of Roche. Lefkowitz at 44 (emphasis added). Dr. Lefkowitz did, however, review deposition testimony of seven Roche employees and concluded that “[t]his testimony is consistent with the possibility of conscious, intended age discrimination.” Id. at 49.

Dr. Lefkowitz’s conclusions can be summarized as (1) Roche’s Guidelines did not meet professional standards for making performance appraisals, (2) according to behavioral science research, performance does not decline with age, (3) age-stereotyping, i.e., unconscious age discrimination, is prevalent in our society and is more likely to be acted on in employment settings when, as in the case of Operation Turnabout, the decisionmakers are not guided by job-related standards of performance, and (4) there is evidence in this case that, while not conclusive, is nonetheless consistent with intentional age discrimination.

3. Anecdotal Evidence of Intentional Age Discrimination 15

In support of their claim that Roche engaged in a pattern or practice of discrimination, plaintiffs also rely on evidence in the record regarding, in their words, “age discriminatory attitudes” by Roche executive committee members and other high level managers. See Plaintiffs’ Supplementation Brief at 30. This evidence is comprised of deposition testimony that several Roche managers, at various points in time prior to Operation Turnabout, made age-related comments and testimony that Roche compiled various reports on which employees’ ages were listed.

First, plaintiffs point to evidence in the record which, according to plaintiffs, demonstrates that “President Lerner himself expressed concerns regarding the company’s average age.” Plaintiffs’ Supplementation Brief at 30. Peter DePaolo, who was manager of analytical services at the time of Operation Turnabout, testified that he was told by Dr. Ann Goetz that, at a managers’ meeting, in about 1983 or 1984, attended by Dr. Goetz, Lerner stated that the average age of Roche was 44 years old and that was too old. See DePaolo at 49-52.

George Reilly, the Director of Finance and Administration for Roche at the time of Operation Turnabout, testified that in approximately early 1981, after either an executive committee meeting or a planning meeting, Anthony Maris, an executive committee member, went to Reilly’s office to discuss a topic that was brought up at the meeting. In the course of the discussion, “Maris indicated that he had just heard a comment from Mr. Lerner stating that the organization is loaded with gray-haired burnouts.” Reilly at 15-16. Reilly also testified in his deposition that “[he] always felt that Lerner had a history of terminating senior employees.” Id. at 134. Although at his deposition Reilly was unable to specify any senior employees who Lerner terminated, in response to a subsequent in terrogatory Reilly provided a list of seventeen Roche employees who were terminated by Lerner or one of his direct subordinates. See Miller Aff.Ex. 14. However, Reilly also stated in response to an interrogatory that he had “no personal, direct knowledge of whether Mr. Lemer’s decisions to terminate any particular individual was influenced by age.” See Miller Aff.Ex. 14.

Plaintiff Gabriel Saucy, who at the time of Operation Turnabout was the Director of Chemical Process Development in the Research and Development Division, testified that, after a tour in early 1983 of the two laboratories which Dr. Saucy directed, Lerner conveyed to Saucy his observation that a lot of the people in the two laboratories were “middle-aged.” See Saucy at 93-94. Thereafter, Dr. Saucy sent a memorandum to Lerner showing the age of the personnel in the labs. Id. at 98. The memo also states that “[t]he attached [the list of average ages] confirms that we have a rather ‘middle-aged’ department!” See Saucy Ex. 7.

Second, plaintiffs point to evidence in the record that, according to plaintiffs, demonstrates that other senior Roche executives were biased against older employees. For example, according to the deposition of Dr. Saucy, sometime between December of 1982 and March of 1983, at a luncheon meeting with upper level managers in the Research and Development division, Dr. Ronald Kuntzman proposed replacing senior level Ph.D.’s who were past 40 years old with younger Ph.D.’s. See Saucy at 69. Kuntzman gave the following reasons: First, he claimed that scientists over 40 were no longer were creative. Second, people over 40 in research were no longer knowledgeable in terms of state-of-the-art knowledge about new science education. Third, younger Ph.D.’s by contrast would be at the cutting edge of science education in terms of knowledge. Lastly, younger persons would be less expensive to employ. See Saucy at 69-70. In his deposition, Dr. Kuntzman stated that he sought to hire new Ph.D.’s, in part, because

[i]n any scientific organization you have to bring in people continuously that have just completed their training. They have new technology, they have the most modern thinking of the research in their field of anyone and so you constantly want to bring in new Ph.D.s into an organization. It is just important for the research effort to be able to do that.

Kuntzman at 103-04. He also testified that while new Ph.D.’s (i.e., Ph.D.’s who have recently completed their training) are generally younger than Ph.D.’s who completed their training some years ago, they are not necessarily younger. See id. at 104-05. Kuntzman was appointed by Lerner to the positions of Vice President of Research and Development and executive committee member in October of 1984. Id. at 16.

Dr. Herbert Weissbaeh, the head Roche’s Institute of Molecular Biology, was asked the following question at his deposition: “Was there a desire by you or by Dr. Kuntzman to focus your hiring on scientists who were young and talented?” Weissbaeh at 46. Weissbaeh answered:

I think it’s true in any organization that you want to always have a constant stream of young, dynamic individuals coming in, building up their career, leaving so that a place can stay vibrant. That’s true whether it be a university department, a pharmaceutical company, the RIMB [Roche Institute of Molecular Biology],

Id. Weissbaeh testified that the replacement of senior scientists in the RIMB by younger scientists was a “natural evolution,” id. at 51, and described the process as follows:

[A] good place, like the Roche Institute, has continually breeded [accomplished, senior scientists] who will become international figures and go off to chairmanships in medical schools, heads of institutes and it’s that turnover that when they leave allows the Institute to bring in younger people and hopefully the same cycle will take place.

Id. at 47. Weissbaeh also stated that he did not think that scientists grew “stale” with age, and that he hired one scientist at age 70 and one at age 65. Id. at 112. However, he also testified that he “think[s] in most professions there tends to be some physiological changes that perhaps leads to people not being quite as productive as when they were younger.” Id.

While at a party in approximately 1980 or 1981, John Kelly, Vice President of the Chemical Division, was overheard stating that he promoted Robert Sunbury, then age 37, to a high level managerial position in the Chemical Production Department because a younger person was needed in this position. Takahashi at 81; Miller Aff. at ¶ 20.

Other age-related comments relied upon by plaintiffs include the following: Betty Lochery, an associate scientist at Roche who was terminated in Operation Turnabout, testified that Dr. Sheridan, on two occasions, one in February 1983 and one in May 1984, told Ms. Lochery that she was too old for her job and that Roche should get somebody younger to perform it. See Lochery at 46-53. Fredric Berdux, who was Manager of Plant Engineering when he was terminated in Operation Turnabout, see Berdux at 5-6, testified regarding age-related comments made by Robert Sunbury at some unspecified time after Sunbury became General Manager of Roche’s Belvidere facility in 1980. See id. at 104-06. Sunbury did not make these comments in Berdux’s presence. Rather, Sunbury made the comments to Harry Kociencki, the Director of Engineering, who then told Berdux. See id. Sunbury told Kociencki, who told Berdux, that he noticed that Kocieneki’s staff supervisors were over the age of 40 or 50 and that older workers were just waiting around to retire and were reluctant to change. Id. Thereafter, Kociencki, on several occasions, stated to Berdux words to the effect to “keep in mind” “how Mr. Sunbury feels [about] people over 40.” Id. at 106.

Plaintiffs also point to the deposition testimony of Berton Haid, who at the time of Operation Turnabout was an associate personnel manager assigned to the Vitamin Fine Chemical Division in the CPD. See Haid at 9. Haid testified that persons in the human resources department had commented “[f]or years and decades” that the Roche workforce had a high average age. Id. at 46-47. The only specificity he could provide with respect to who made these comments was that the comments were made by “[b]enefits people, compensation people.” Id. at 47.

In addition, there is evidence in the record that indicates that certain reports were prepared, to which President Lerner had access, that contained information, both directly and indirectly, relating to the average age of the Roche workforce. Arthur Bledsoe, the assistant vice president and director of compensation and benefits, testified that he was “asked periodically to identify those employees who were eligible to retire for management planning purposes.” Bledsoe at 29. This information was passed on to Lerner. Id. at 30. In addition, as part of his job, Bledsoe compiled statistics regarding the average age of the work force. Id. at 29. This information was requested by Len Silverman, who was vice president of human resources and a member of the executive committee. Id. at 29-30. Bledsoe did not recall why or when this information was requested. Id.

Lastly, Gerhart Frohlich testified that, in connection with his duties on the benefit task force at Roche, he had “heard that the age profile in Roche was higher than in other companies.” Frohlich at 83. This information was provided to the benefit task force in reports by Hewit Associates, which is a benefit compensation consulting firm hired by Roche. Id. at 83-84. Hewit began preparing these reports for Roche beginning in the late 1970s, see id., and President Lerner received copies of the reports, id. at 85.

B. Analysis

The question now is: Based on this evidence, could a reasonable jury find that Roche engaged in a pattern or practice of discrimination in conducting Operation Turnabout?

As stated above, if the discriminatory policy were admitted or openly declared, then that is obviously sufficient to prove the existence of the policy. See, e.g., Phila. Bd. of Ed., 911 F.2d at 892-93; 110 Cong.Rec. 14239 (1964) (remarks of Senator Humphrey) (cited in Teamsters, 431 U.S. at 336 n. 16, 97 S.Ct. at 1855 n. 16). See also United States v. Gregory, 871 F.2d 1239, 1243 (4th Cir.1989) (“[I]f admissions [as to the existence of a policy of sex discrimination] are credited, the Title VII violation has been proven”), cert. denied, 493 U.S. 1020, 110 S.Ct. 720, 107 L.Ed.2d 740 (1990). This situation, however, is rare and is not present in the case now before the court. Most plaintiffs alleging a pattern or practice of discrimination must demonstrate a consistent history of discrimination and attempt to do so through a combination of statistical evidence and evidence of specific instances of discrimination in order to bring “the cold numbers convincingly to life.” Teamsters, 431 U.S. at 339, 97 S.Ct. at 1856.

For example, in Teamsters, the Government brought two actions, that were eventually consolidated, against T.I.M.E.-D.C., Inc. (“company”), a trucking company, and the International Brotherhood of Teamsters (“union”). “The Government’s theory of discrimination was simply that the company, in violation of [Title VII], regularly and purposefully treated Negroes and Spanish-surnamed Americans less favorably than white persons.” Teamsters, 431 U.S. at 335, 97 S.Ct. at 1854. The less favorable treatment' “allegedly involved the refusal to recruit, hire, transfer, or promote minority group members on an equal basis with white people, particularly with respect to line-driving positions.” M 16 The Government sought general injunctive relief and specific “make whole” relief for all individual discriminatees, which would allow them- an opportunity to transfer to line-driver jobs with full company seniority. 17 Id. at 330, 97 S.Ct. at 1852.

To demonstrate that the company engaged in a pattern or practice of discrimination, the Government relied on statistical evidence and “the testimony of individuals who recounted over 40 specific instances of discrimination.” Id. at 337-38, 97 S.Ct. at 1855-56. The Court summarized the statistical evidence as follows:

As of March 31, 1971, shortly after the Government filed its complaint alleging systemwide discrimination, the company had 6,472 employees. Of these, 314 (5%) were Negroes and 257 (4%) were Spanishsurnamed Americans. Of the 1,828 line drivers, however, there were only 8 (0.4%) Negroes and 5 (0.3%) Spanish-surnamed persons, and all of the Negroes had been hired after the litigation had commenced. With one exception ... the company and its predecessors did not employ a Negro on a regular basis as a line driver until 1969____ [E]ven in 1971 there were terminals in areas of substantial Negro population where all of the company’s line drivers were white.

Id. at 337, 97 S.Ct. at 1855. The Court further noted that 83% of African-Americans and 78% of Spanish-surnamed Americans who worked for the company held lower paying city-operations and servicemen jobs, while only 39% of the nonminority employees held jobs in these categories. Id. at 337-38, 97 S.Ct. at 1855-56.

The Government bolstered this evidence with testimony relating to over 40 specific instances of discrimination. Based on this testimony, the court found that

[n]umerous qualified black and Spanishsurnamed American applicants who sought line driving jobs at the company over the years, either had their requests ignored, were given false or misleading information about requirements, opportunities, and application procedures, or were not considered and hired on the same basis that whites were considered and hired.

Id. at 338, 97 S.Ct. at 1856. The company was unable to rebut this evidence, and thus, the company was found to have engaged in “systematic and purposeful employment discrimination,” see id. at 342-43, 97 S.Ct. at 1858-59, i.e., a pattern or practice of discrimination. This evidence justified a’ shifting of the burden of proof to the employer with respect to the claims for relief of the individual class members. Id. at 362, 97 S.Ct. at 1868.

In United States v. Lansdowne Swim Club, 894 F.2d 83 (3d Cir.1990), the Government asserted that Lansdowne Swim Club (“LSC”) engaged in a pattern or practice of discrimination by refusing membership to African-Americans because of their race, in violation of Title II of the Civil Rights Act of 1964, 42 U.S.C. § 2000a-5(a) [hereinafter “Title II”]. Although this case was brought pursuant to Title II, which prohibits discrimination in places of public accommodation, the Court of Appeals (and the district court) applied law relating to Title VII because the relevant sections of both Title VII and Title II “require proof of ‘pattern or practice’, ..., and the legislative history of the Civil Rights Act indicates that the phrases have the same meaning.” Id. at 88 n. 7.

The Court of Appeals affirmed the district court’s finding of a pattern or practice of discrimination based on the following evidence. First, “[s]ince its founding in 1957, LSC has granted 1400 full family memberships. Every white applicant has been admitted, although two as limited members. In that time, however, LSC has had only one non-white member.” Id. at 84-85. In addition, the Court of Appeals found the following uncontroverted experiences of three Lansdowne families to be significant.

In 1976, the Allisons wrote to LSC requesting an application but LSC did not respond. Dr. Allison is black; his three children are part black. In 1977, the Allisons twice again wrote for an application but LSC did not respond. The following year, the Allisons repeated the procedure with similar results. In 1983, the Allisons filed a timely application and otherwise qualified for membership but were rejected. The following year, the Ryans filed a timely application and otherwise qualified for membership. Nonetheless, they were rejected. Two of the Ryans’ adopted children are black. The Ryans then complained to the media and picketed LSC, joined by the Allisons. In 1986, the Iverys, who are black, filed a timely application and otherwise qualified for membership. Nonetheless, they were rejected (as were the Ryans and Allisons who had applied again).

Id. at 85. In addition, LSC deterred several applicants known to be black by ignoring their membership inquiries. Id. at 89. The Court of Appeals affirmed the district court’s finding that such inaction was intended to deter black applicants and was probative of a discriminatory practice. Id.

Based on this evidence, the district court found, and the Court of Appeals affirmed, that LSC engaged in a pattern or practice of discrimination by refusing membership to African-Americans because of their race.

In King v. General Electric Company, 960 F.2d 617 (7th Cir.1992), the Court of Appeals for the Seventh Circuit addressed the issue of what constitutes a pattern or practice of discrimination in the context of an ADEA case arising from a reduction in force. This case arose from a reduction in force at General Electric’s (“GE”) Bloomington, Illinois manufacturing plant between 1982 and 1985. 18 During that time period, GE laid off, demoted or transferred two hundred management and professional workers pursuant to the RIF. Id. at 618. Plaintiffs alleged that GE “pursued a policy of discrimination in its reduction in force.” Id. at 625. GE allegedly “gave employees false performance ratings, trumped up disciplinary charges against older employees, and terminated ‘many of its older employees while retaining younger employees whose performance and background were inferior to the older who were laid off and terminated.’ ” Id.

At the “pattern-or-praetice” trial, the plaintiffs’ expert testified that age was a statistically significant variable affecting the probability of an employee experiencing an adverse action during 1983 and 1984, even after controlling for the effect of perfor manee and assignment. Id. at 619. Plaintiffs also presented anecdotal evidence of discrimination. Six plaintiffs testified that they were demoted or laid-off while GE continued to employ younger employees in their respective departments. Witnesses also testified that some GE managers made disparaging comments about older workers. See id.

Based on this evidence the jury found that GE engaged in a pattern or practice of age discrimination in conducting the RIF between 1982 and 1985. However, the Court of Appeals held, among other things, that regardless of the court’s view of the statistical evidence provided by the plaintiffs, “they have failed to provide the amount of anecdotal evidence claimants in other eases have supplied to support a finding of pattern or practice.” Id. at 627. This ruling was based in part on the court’s finding that some the anecdotal evidence of discrimination related to events that were outside the time period that the alleged policy was in effect. See id. at 626. The cases on which the Court of Appeals relied to support the conclusion that the individual evidence of discrimination was' insufficient included Teamsters, 431 U.S. at 337-38, nn. 17-18, 97 S.Ct. at 1855-56, nn. 17-18, in which forty examples of individual discrimination were provided, and Chisholm v. United States Postal Service, 665 F.2d 482, 495 (4th Cir.1981), in which twenty employees testified about individual discrimination. The court stated that “[w]ithout significant individual testimony to support statistical evidence, courts have refused to find a pattern or practice of discrimination.” King, 960 F.2d at 624 (citing, e.g., Cooper, 467 U.S. at 878-80, 104 S.Ct. at 2800-02 (citing with approval the opinion of the court below, EEOC v. Federal Reserve Bank of Rich mondi 698 F.2d 633, 643-44 (4th Cir.1983) contrasting 20 or 40 class members testifying about individual discrimination (sufficient) to three or seven incidents of discrimination over a period of years (insufficient)); Goff v. Continental Oil Co., 678 F.2d 593, 597 (5th Cir.1982) (stating that “even if all three witnesses’ accounts of racial discrimination were true, this evidence would not have been enough to prove a pattern or practice of company-wide discrimination”); Ste. Marie v. Eastern Railroad Ass’n, 650 F.2d 395, 406 (2d Cir.1981) (finding that seven individual discriminatory acts coupled with problematic statistical evidence were insufficient to support a finding of pattern or practice of discrimination)). Therefore, the Court of Appeals for the Seventh Circuit held that there was not substantial evidence to support the jury’s finding that GE engaged in a pattern or practice of discrimination. 19 Id. at 627.

These cases, along with other pattem-or-practice cases which were reviewed, demonstrate that, in order to prove a pattern or practice of discrimination, plaintiffs generally must demonstrate widespread differences in treatment (usually through statistical evidence) and must also produce sufficient evidence of intentional discrimination (usually, but not always, through testimony regarding intentional discrimination against individual members of the class) so that the court or jury can infer that the statistical disparities are the result of the pursuit by the employer of a policy of intentional discrimination. See also EEOC v. Chicago Miniature Lamp Works, 947 F.2d 292, 297 (7th Cir.1991) (stating that a prima facie case for a pattern or practice of disparate treatment can be established by “statistical evidence demonstrating substantial disparities in the application of employment actions ..., buttressed by evidence of general policies or specific instances of discrimination”); Mozee v. American Commercial Marine Service Company, 940 F.2d 1036 (7th Cir. 1991) (affirming district court’s finding a pattern or practice based on statistical evidence of a disparate impact that was buttressed by the following “nonstatistical evidence of intentional discrimination”: (1) the defendant’s noncompliance with its various affirmative action plans; (2) discriminatory markings on the files of all African-Americans; (3) discriminatory signs posted in the work place; and (4) evidence supporting the conclusion that five individual plaintiffs were discriminated against), cert. denied, 506 U.S. 872, 113 S.Ct. 207, 121 L.Ed.2d 148 (1992); EEOC v. Western Electric Company, Incorporated, 713 F.2d 1011, 1019 (4th Cir.1983) (reversing district court’s finding of a pattern or practice where plaintiff was “unable to bolster its marginal statistical case with individual cases of discrimination because it failed to establish discrimination as to any of the 16 witnesses”).

In this case, plaintiffs’ theory is that Roche gave the line managers who performed the rankings, and thus made the actual termination decisions, complete discretion, see Plaintiffs’ Supplementation Br. at 7 (“Each step along the way furthered Roche’s goal of permitting untrammeled discretion to rule in the implementation of Operation Turnabout.”), and that Roche intended and/or knew that this discretion would result in both conscious and unconscious discrimination against older workers because there was “pre-existing age bias” among Roche’s managers, see, e.g., id. at 4r-5 (stating that Roche’s “ad hoc employee evaluation system ... inevitably exploited the conscious and unconscious exercise of age bias among those managers making selection decisions” and that “Operation Turnabout’s selection policies in fact provided the fertile ground needed for the exercise of pre-existing age bias among Roche’s decision-making managers”).

The portions of Operation Turnabout which the plaintiffs allege led to the managers who made the decisions having “untrammeled discretion” include (1) the number and content of the factors listed in the Guidelines to be used in performing the rankings, see id. at 8 (“Their overly generalized nature and sheer number render them almost useless in providing any real guidance for managers charged with actual decision-making.”); (2) the Guidelines’ failure to instruct the managers how to use the factors listed in the Guidelines, see id. at 11 (“no explanation was given as to how the several factors were to be combined and prioritized into a single composite and overall evaluation by which each individual employee would be ranked”); (3) the absence from the factors listed in the Guidelines of seniority and the prior written performance evaluations and merit ratings, which plaintiffs assert are objective measures of performance, see id. at 9; (4) the failure of Roche to ensure that the Guidelines were distributed to all managers who would perform the rankings, see id. at 12-14; (5) that a number of managers performed their rankings without using the Guidelines, and thus, used criteria, such as “gut feel,” that were even more subjective than that contained in the Guidelines, see id. at 15-16; (6) that the managers making the decisions only had a few days in which to perform this task, see id. at 17 (“The extremely limited time allotted to complete all these tasks fairly ensured that the underlying decisions themselves would be made in a haphazard way without even a pretense of methodological rigor.”); and (7) the lack of higher level managerial control or review of the termination decisions, see id. at 19 (“Another design defect in Roche’s implementation procedures was the lack of higher managerial control or review.”).

Based, for the most part, on these allegations, plaintiffs argue that

[d]uring the “one process” that constituted Operation Turnabout, Roche’s “standard operating procedure” for selecting individuals for termination was based on using vague, undefined and subjective criteria, coupled with no procedures for reviewing the exercise of managerial discretion in the selections____

Id. at 37. Given the evidence delineated supra in section IV.A., a reasonable jury could find that Roche’s standard operating procedure in conducting Operation Turnabout was the use of “vague, undefined and subjective criteria” coupled with a lack of higher level review of the managers’ selections.

Plaintiffs also allege that “[tjhese conditions, among other things, created fertile ground for exploiting pre-existing age bias and encouraging the discriminatory treatment of Roche’s older employees.” Id. at 37. It would be reasonable for a jury to infer that, ’if a manager has a great deal of discretion in determining which employees are to be fired, then that manager had the opportunity to exercise his discretion in a discriminatory fashion. See Ste. Marie, 650 F.2d at 405 (“While evidence of subjective and discretion ary promotion and transfer procedures may indeed strengthen the inference of a pattern or practice of purposeful discrimination that can be drawn from a valid statistical showing of disparities in the work force ... this is because the statistical disparities in such a case are evidence that the potential for disparate treatment created by loose procedures has been realized on a significant scale.”). However, this alone does not constitute a pattern or practice of intentional discrimination. See, e.g., id. at 405 (finding that where “relevant statistics are lacking and the probative evidence of discrimination is confined ... to seven individual incidents, subjective decisionmaking methods are not sufficient to establish a pattern or practice of discrimination; all that could be inferred is that their potential has been actualized in the same seven cases.”).

Plaintiffs must demonstrate that the managers making the termination decisions routinely took age into account when deciding who would be fired in Operation Turnabout, not that the managers had the discretion to do so. In other words, plaintiffs must demonstrate that the managers making the decisions systematically used their discretion to purposefully treat employees over 40 years of age less favorably, and, in addition, must demonstrate that such disparate treatment was regular enough, routine enough, or pervasive enough to justify the conclusion that it is more probable than not that Roche discriminated against any employee fired in Operation Turnabout who was over 40 years old. Therefore, without additional evidence, the decentralized, uncontrolled decisionmaking process used in Operation Turnabout is insufficient to support a finding that Roche engaged in a pattern or practice of discrimination. See, e.g., EEOC v. MCI Intern., Inc., 829 F.Supp. 1438, 1446 (D.N.J.1993) (“There is no single, company-wide action but, rather, layoffs by various supervisors of persons, on an employee by employee basis, of widely disparate ages, salaries, and positions over a two year period of time. The factual and/or employment setting of each person’s layoff is different, many varying questions of law are raised, and the defenses posited are individual to each or small groups of those persons. And while this action alleges that the putative class was adversely affected by a pattern or practice of discriminatory treatment, no such pattern or practice is the least bit evident. Indeed, the only common thread seen here is that the persons involved were forty years old or older and were laid off by MCII. That clearly is not enough.”).

The first possible common thread running through all the decisions is obviously the Guidelines. In fact, plaintiffs allege that the vague, subjective criteria delineated in the Guidelines were Roche’s standard operating procedure for conducting the RIF. The Guidelines, however, specifically state that “[a]ge should not be considered as a factor in any decision.” Guidelines at E002279. Therefore, as discussed above, this is not a situation where the discriminatory policy is explicitly stated. See, e.g., Phila. Bd. of Ed., 911 F.2d at 892-93; U.S. v. Gregory, 871 F.2d 1239, 1243 (4th Cir.1989). Furthermore, plaintiffs allege only that the Guidelines provided the individual managers making the decisions with the opportunity to exercise their own “pre-existing age bias,” see Plaintiffs’ Supplementation Br. at 10 (“Roche deliberately sought to maximize the opportunities for the exercise of subjective bias by its managers____”), not that a policy of discrimination was explicitly set forth in the Guidelines.

Thus, the crux of plaintiffs’ pattem-orpractice claim arguably is that age bias was so pervasive at Roche that the Roche executives knew that, if a decentralized decision-making process was used for conducting Operation Turnabout, then the individual line managers would exercise their discretion by terminating older workers because of their age.

In support of this claim, plaintiffs rely on two expert reports and anecdotal evidence of discrimination.

Dr. Bloch’s Report: The Supreme Court has stated that “a fluctuation of more than two or three standard deviations would undercut the hypothesis that decisions were being made randomly with respect to [a protected trait].” Hazelwood School District v. United States, 433 U.S. 299, 311 n. 17, 97 S.Ct. 2736, 2743 n. 17, 53 L.Ed.2d 768 (1977) (citation omitted). In performing various analyses in this ease, plaintiffs’ statistical expert, Dr. Bloch, found standard deviations as high as 14. See Bloch Report at 7 (comparison of number of terminated employees younger than 55 to those older than 55 resulted in a 14.60 standard deviations). However, the “usefulness” of statistical evidence “depends on all of the surrounding facts and circumstances.” Teamsters, 431 U.S. at 340, 97 S.Ct. at 1857. “Statistical comparisons, if they are to have any value, must be between comparable groups and free from variables which would undermine the reasonableness of discrimination inferences to be drawn.” Mazus v. Dept. of Transp., Com. of Pa., 629 F.2d 870, 875 (3d Cir.1980) (citation omitted), cert. denied, 449 U.S. 1126, 101 S.Ct. 945, 67 L.Ed.2d 113 (1981). See also King, 960 F.2d at 627 (stating that for statistical analyses to be probative, they must “properly control for factors that might lead to differences in raw percentages”). Therefore, given plaintiffs’ allegations and the evidence in the record that employees in each work group were ranked only against one another for the purposes of determining who would be fired (i.e., there was not a companywide ranking of employees), Dr. Bloch’s analyses, such as the one above that yielded a standard deviation of 14, in which he merely compared the termination rates of older and younger workers across the entire company, are of little probative value. Cf. Hazelwood, 433 U.S. at 308 n. 13, 97 S.Ct. at 2742 n. 13 (“When special qualifications are required to fill particular jobs, comparisons to the general population (rather than to the smaller group of individuals who possess the necessary qualifications) may have little probative value.”).

Dr. Bloch, however, did perform analyses comparing employees that Roche claims were ranked against one another. 20 After taking this into account, the disparity in likelihood of being terminated between those older and younger than 40 years is the equivalent of 5.20 standard deviations. When last merit ratings (which, because they are a measure of performance, are a nondiscriminatory factor that may explain the statistical disparity) were also held constant, the resulting disparity is equivalent to 3.90 standard deviations for pools 1-115, 2.22 standard deviation for pools 117-143, and 4.62 standard deviations for the combined group. 21 When average merit ratings over the period 1982-84 are held constant, the respective standard deviations are 4.79, 2.44, and 5.48 standard deviations. See Bloch Report at 5-6. Thus, when merit ratings are taken into account, whether the last merit rating or the average over the preceding years, and the standard deviation is calculated for the combined groups, the analysis yields standard deviations that would “undercut the hypothesis that decisions were made randomly with respect to [age].” See Hazelwood, 433 U.S. at 311 n. 17, 97 S.Ct. at 2743 n. 17 (citation omitted). However, when the groups are divided, the analysis yields standard deviations that undercut the age neutral hypothesis for groups 1-115 (standard deviations of 3.90 and 4.79), but yields standard deviations that are not as suggestive of age discrimination for groups 117-143 (standard deviations of 2.22 and 2.44). However, because these analyses all yield standard deviations of more than two or three, albeit in a number of cases only slightly more than two or three, a reasonable jury could find that this “reinforee[s] rather than rebut[s]” plaintiffs’ other pattem-or-practice evidence. Hazelwood 433 U.S. at 311 n. 17, 97 S.Ct. at 2743 n. 17 (stating that because “a fluctuation of more than two or three standard deviations would undercut the hypothesis that decisions were being made randomly with respect to [a protected trait],” if the government were able to demonstrate standard deviation of greater than five or six, then this would “reinforce rather than rebut the Government’s other evidence”).

Length of service is another factor that Dr. Bloch took account of in performing his analysis. This factor is particularly relevant given the testimony regarding the use of proximity to retirement and eligibility for retirement in making termination decisions. As will be discussed in more detail infra, terminating someone wholly based on proximity to retirement or eligibility for retirement benefits is not violative of the ADEA. Thus, statistical analyses that controlled for these nondiseriminatory factors would be more probative than those that did not. Dr. Bloch did not, however, perform any analyses in which he controlled for these factors. However, as will also be discussed in more detail infra, length of service is explicitly part of the definition of “ample retirement benefits,” as defined by plaintiffs, and it is reasonable to infer that it also correlates to some extent to “proximity to retirement.” Thus, length of service arguably can be used as a rough proxy for “ample retirement benefits” and “proximity to retirement.”

When length of service is held constant, both on its own and in connection with various other factors such as comporatio and percent of range (which are two measures of an employee’s compensation), this results in fluctuations from the expected number of between 4.47 standard deviations (comporatio and length of service held constant) and 5.51 standard deviations (percent of range and length of service held constant). See Bloch Report at 9 (Table 3). When division is also held constant the fluctuations range, with one exception, from 4.77 standard deviations (comporatio, length of service, and division held constant) to 5.52 standard deviations (length of service and division held constant). Id. When length of service, combined with percent of range and division, are held constant, the fluctuation jumps to 9.85 standard deviations. Id. Thus, with this one exception, when other factors that might lead to differences in raw percentages are properly controlled, the standard deviations are in the 4 to 5 range. Because these are greater than 3 standard deviations, a reasonable jury could also find that these statistics “reinforce rather than rebut” the plaintiffs’ other evidence of discrimination. Hazelwood, 433 U.S. at 311 n. 17, 97 S.Ct. at 2743 n. 17. 22

Turning to plaintiffs’ nonstatistical evidence, at least one court has stated that “[wjithout significant individual testimony to support statistical evidence, courts have refused to find a pattern or practice of discrimination.” King, 960 F.2d at 624 (citing Cooper, 467 U.S. at 878-80,104 S.Ct. at 2800-02; Teamsters, 431 U.S. at 337-38, nn. 17-18, 97 S.Ct. at 1855-56, nn. 17-18; Goff, 678 F.2d at 597; Chisholm, 665 F.2d at 495; Ste. Marie, 650 F.2d at 406). Anecdotal evidence, such as this, is useful to demonstrate that the statistical disparities are the result of intentional discrimination, and not the result of nondiseriminatory factors. See Teamsters, 431 U.S. at 339, 97 S.Ct. at 1856 (“The individuals who testified about their personal experiences with the company brought the cold numbers convincingly to life.”). See also Mozee, 940 F.2d at 1051 (affirming district court’s finding a pattern or practice based on statistical evidence of a disparate impact that was buttressed by the “nonstatistieal evidence of intentional discrimination”).

In this case, plaintiffs’ nonstatistieal evidence can be divided in three categories: (1) Dr. Lefkowitz’s Report; (2) anecdotal evidence of “discriminatory attitudes” by Roche executives and high level managers; and (8) evidence of intentional discrimination against individual plaintiffs in the conducting of Operation Turnabout.

Dr. Lefkowitz’s Report: As stated above, in his report, Dr. Lefkowitz concluded that (1) Roche’s Guidelines did not meet professional standards for making performance appraisals, (2) according to behavioral science research, performance does not decline with age, (3) age-stereotyping, i.e., unconscious age discrimination, is prevalent in our society and is more likely to be acted on in employment settings when, as in the case of Operation Turnabout, the decisionmakers are not guided by job-related standards of performance, and (4) there is evidence in this case that, while not conclusive, is nonetheless consistent with intentional age discrimination. These conclusions are only mildly probative of whether or not Roche engaged in a pattern or practice of discrimination. This is because Dr. Lefkowitz’ main conclusion is merely that the Guidelines did not properly direct the managers making the termination decisions to consider only job-related factors, and in such a situation the age-stereotyping that is prevalent in our society will likely be acted on. This is similar to the conclusion of the Court of Appeals for the Second Circuit in Ste. Marie, wherein the court stated that subjective and discretionary employment practices may strengthen the inference of a pattern or practice that can be drawn from statistical disparities, because such discretion gives the managers the opportunity to act in a discriminatory manner. See Ste. Marie, 650 F.2d at 405.

However, while this is somewhat probative of whether or not the Roche managers making the termination decisions discriminated, its value is limited because Dr. Lefkowitz explicitly states that he is “not in a position to pass definitive judgment” on the possibility that the statistical disparity in terminations between younger and older workers in Operation Turnabout was the result of intentional discrimination. See Lefkowitz Report at 44. Because plaintiffs are alleging a pattern or practice of discrimination, they must demonstrate that the statistical disparity is the result of a pattern or practice of intentional discrimination. See Teamsters, 431 U.S. at 335 n. 15, 97 S.Ct. at 1854 n. 15 (stating that “proof of discriminatory motive is critical”). Therefore, Dr. Lefkowitz’ Report is of limited probative value on the issue of whether Roche engaged in a pattern or practice of discrimination.

Anecdotal Evidence of ‘Discriminatory Attitudes” by Roche Executives: Plaintiffs also rely on evidence in the record regarding, in their words, “age discriminatory attitudes” by Roche executive committee members and other high level managers. See Plaintiffs’ Supplementation Brief at 30. This evidence is comprised of deposition testimony that several Roche executives and high level managers, at various points in time prior to Operation Turnabout, made age-related comments and testimony that Roche compiled various reports on which employees’ ages were listed and on which the average age of the Roche workforce was listed. Plaintiffs argue that this evidence demonstrates that there was “an age discriminatory corporate culture at Roche over a period of years prior to Operation Turnabout.” Plaintiffs’ Br. in Opposition to Roche’s Motion to Strike at 2.

First, plaintiffs rely on evidence of certain statements allegedly made by President Lerner. Specifically, Lerner allegedly stated at a managers’ meeting in approximately 1983 or 1984 that the average age of Roche employees was 44 and that was too old. Also, after a tour of a laboratory in a the Research and Development Division in 1983, Lerner allegedly told a director in the division that a lot of the people in the labs were “middle-aged.” See Saucy at 93-94. In addition, in approximately 1981, Lerner allegedly stated at either an executive committee or planning meeting that Roche was loaded with “gray-haired burnouts.” See Reilly at 15-16.

Generally, “stray remarks by non-decision-makers or by decisionmakers unrelated to the decision process are rarely given great weight, particularly if they were made temporally remote from the date of the decision.” Brewer v. Quaker State Oil Refining Corporation, 72 F.3d 326, 333 (3d Cir.1995) (citing Ezold v. Wolf, Block, Schorr & Solis-Cohen, 983 F.2d 509, 545 (3d Cir.1992), cert. denied, 510 U.S. 826, 114 S.Ct. 88, 126 L.Ed.2d 56 (1993)). There is no evidence in the case that Lerner made any termination decisions in Operation Turnabout. In addition, all of these statements were allegedly made prior to Operation Turnabout, ranging from one year to four years prior to the RIF. Also, there is no evidence of a causal link between these statements and any termination decision made during Operation Turnabout. Thus, these statements are properly classified as temporally remote, stray remarks made be a non-decisionmaker. See Brewer, 72 F.3d at 333. Statements such as these are generally not given great weight. 23 However, although the statements made by Lerner “should not be given significant or commanding weight,” they may provide some relevant evidence of discrimination. Id. (emphasis added). The Court of Appeals for the Third Circuit has held that

a supervisor’s statement about the employer’s employment practices or managerial policy is relevant to show the corporate culture in which a company makes its employment decision, and may be used to build a circumstantial case of discrimination.

Id. at 333-34 (holding that the following written statement in a company newsletter by defendant’s chief executive officer was circumstantial evidence of age discrimination because it was “evidence of the corporate culture in which the employment decision to discharge [plaintiff] was made”: the CEO stated in the newsletter “two of our star young men in their mid-40’s. That age group is our future.”) (citations omitted). See also Roebuck v. Drexel University, 852 F.2d 715, 733 (3d Cir.1988) (stating in ease where university president exercised a significant influence on the decisionmakers and had made the final tenure decision, that while two statements he made reflecting racial bias “standing alone, occurring as they did five years before the final denial of tenure, could not suffice to uphold a finding [of discrimination], they do add support, in combination with the other evidence, to the ultimate conclusion”). Thus, while Lemer’s statements should not be given significant weight, they are probative of Roche’s corporate culture. 24

Plaintiffs also assert that the following age-related comments by other high level Roche managers are probative of Roche’s corporate culture: (1) in late 1982 or early 1983, at a luncheon meeting with upper level managers in the Research and Development Division, Dr. Ronald Kuntzman allegedly suggested that senior level Ph.D.’s who were past 40 years old would be replaced with younger Ph.D.’s, because, among other reasons, scientists over 40 were no longer creative, see Saucy at 69-70; (2) Dr. Herbert Weissbaeh, the head of Roche’s Institute of Molecular Biology (“RIMB”) stated in his deposition that at RIMB the replacement of senior scientists with younger scientists was a “natural evolution” because accomplished senior scientists at RIMB move on to “chairmanships of medical schools, heads of institutes” and then younger scientists can replace them and keep RIMB “vibrant,” see Weissbaeh at 46-47; (3) at a party in 1981, John Kelly, then Vice President of the Chemical Division, was overheard at a party stating that he promoted Robert Sunbury, then age 37, to a high level managerial position because a younger person was needed in that position, see Takahashi at 81; (4) on two occasions, one in February 1983 and one in May 1984, Betty Lochery, an associate scientist at Roche who was terminated in Operation Turnabout, testified that Dr. Jane Sheridan, Vice President of Quality Control, told Ms. Lochery that she was too old for her job and that Roche should get somebody younger to perform it, see Lochery at 46-53; and (5) at some unspecified time after Robert Sunbury became General Manager of Roche’s Belvidere facility in 1980, he allegedly told Harry Koeiencki, the Director of Engineering, he noticed that KociencM’s staff supervisors were over the age of 40 or 50 and that older workers were just waiting around to retire and were reluctant to change, see Berdux at 104-06.

Like Lerner’s statements these are all temporally remote to Operation Turnabout and are either stray remarks made by nondecisionmakers or stray remarks made by decisionmakers unrelated to the decisionmaking process in Operation Turnabout. Thus, these statements are not significant evidence of age discrimination, however, they are relevant to show Roche’s corporate culture, and may be used to build a circumstantial ease of discrimination. See Brewer, 72 F.3d at 333-34.

Lastly, there is evidence in the record that certain reports were periodically prepared, to which President Lemer had access, that contained information identifying those employees who were eligible to retire. See Bledsoe at 29-30. Also, at some unspecified time, Bledsoe compiled statistics regarding the average age of the Roche workforce for executive committee member Len Silverman. Id. In addition, since the late 1970’s, Lerner was receiving reports from Hewit Associates, a benefits compensation consulting firm hired by Roche, which reflected that “the age profile in Roche was higher than in other companies.” Frohlich at 83.

Plaintiffs argue that this information is relevant to prove that “there was a general awareness at Roche that the age of its workforce was fairly high, as compared to other companies in the industry.” Plaintiffs’ Supplementation Br. at 30. This evidence is relevant to that issue, and thus, is probative of whether or not Roche engaged in a pattern or practice of discrimination. Cf. Armbruster v. Unisys Corporation, 32 F.3d 768, 781 (3d Cir.1994) (holding that documents prepared by the defendant that contained age notations, birth dates and/or dates of hire may have been related to the adverse employment decision, and therefore, the documents may be introduced as evidence of pretext); Sheerin v. NYS Div. of Substance Abuse Services, 844 F.Supp. 909, 915-17 (N.D.N.Y.1994) (holding that, although a chart containing the ages of employees was presumably compiled for a legitimate reason, plaintiff is not prevented from proving, based on a statistical disparity in a RIF between the termination rates for employees older than 50 years and younger than 50, that the chart was used in a discriminatory manner). But see Wilson v. Firestone Tire & Rubber Co., 932 F.2d 510, 514 (6th Cir.1991) (“absent direct evidence that [personnel documents containing all employees’ ages and years of service] was used in making adverse employment decisions, cannot create even a circumstantial inference of discrimination’’); Bowman v. Firestone Tire and Rubber Co., 724 F.Supp. 493, 506 (N.D.Ohio 1989) (“without further evidence of discriminatory use, [use of Age Data] forms [is] not supportive of plaintiffs’ age discrimination claim”).

Thus, plaintiffs’ anecdotal evidence of “discriminatory attitudes” by Roche executives and higher level managers is relevant to demonstrating the corporate culture in which the decisions regarding who would be fired in Operation Turnabout were made. This evidence, however, should not be given significant or commanding weight. This is particularly true in this case, given that the evidence supports plaintiffs’ allegation that the decisionmakers during Operation Turnabout were the lower level managers and their decisions were not subject to a great deal of review by the higher level managers, who made the age-related statements discussed above. See Plaintiffs’ Supplementation Br. at 19 (“Another design defect in Roche’s implementation procedures was the lack of higher managerial control or review.”). Thus, for the most part, plaintiffs do not allege that high level managers who made the age-related comments described above were directly involved in the decisionmaking process. Therefore, this evidence only goes to Roche’s corporate culture and may be used to build a circumstantial case of discrimination. See Brewer, 72 F.3d at 333.

Individual Instances of Discrimination: Lastly, plaintiffs rely on certain deposition testimony that they assert demonstrate that the “implementation procedures encouraged and permitted the application of age stereotypes and age biases.” Plaintiffs’ Supplementation Br. at 34.

First, there is testimony that in the Chemical Production Department the managers doing the rankings had documents that listed, among other things, employee’s ages and that therefore, “[the managers] certainly were aware how old people were at the time [they] did this procedure.” Meinhardt at 231-32. In addition to the employees’ age, this document contained, among other things, employees’ sex, race code, years of service, title, and years in that title. See Haid at 38-39, Ex. 4 (attached to Haid deposition). Also, Dr. Rodney Fryer, who was in the Research Division, testified that his superior gave him a list of employees that included the employees’ birth dates and race. Fryer at 66-68.

This evidence is probative of intentional discrimination based on age. See Armbruster, 32 F.3d at 781 (holding that documents prepared by the defendant that contained age notations, birth dates and/or dates of hire may have been related to the adverse employment decision, and therefore, the documents may be introduced as evidence of pretext) Sheerin, 844 F.Supp. at 915-17 (holding that, although a chart containing the ages of employees was presumably compiled for a legitimate reason, plaintiff is not prevented from proving, based on a statistical disparity in a RIF between the termination rates for employees older than 50 years and younger than 50, that the chart was used in a discriminatory manner).

Second, plaintiffs point to testimony by a number of managers that they took into account an employee’s proximity to retirement and eligibility for retirement benefits when determining who would be fired.

A1 Meinhardt, a manager in the Chemical Production Department, stated that he considered the following factors when performing the rankings:

family situations, and how old a guy is and where his kids are and his financial circumstances to the best you know it, and it all falls into the judgment, and that’s what it is. It’s a subjective judgment when you come down to it.

Meinhardt at 76. Fedor Frankl, who was also a manager in the Chemical Production Department, testified that

we were not as much concerned about people who were at the advanced age because they were covered by pension and suffering for these people would be less rather than losing their job than the younger people.

Frenkl at 28. With respect to one specific decision, Frenkl stated that an employee named Bob Czesniewski, a foreman, was terminated because “[h]e was at age 62 or 63 and [they] thought it wouldn’t be problem for him to leave” because “[i]t was just an earlier retirement.” Id. at 29.

Thomas Anton, who is another manager in the Chemical Production Department, testified that two employees, Mike Matta and A1 Papio, were fired because they were near retirement and would be less hurt by being fired than a younger employee. See Anton at 28-30; 53-56; 77-86. He also testified, with respect to a third employee fired in Operation Turnabout, that the managers

felt that Joe Zelauskas was acceptable and better in the performance of his job but we felt that he would be the least hurt by being let go because he was somewhat near retirement anyway.

Anton at 28-29.

In the Chemical Research Department, Ed Tschopp testified that he believed that nearness to retirement was the major reason for the terminations of two of his subordinates, Wanio and Gaida. See Tschopp at 74-81.

Lastly, Dr. Fryer, who was in the Research Division, testified that where the em ployees he was ranking could not be differentiated based .on performance, he took into account proximity to retirement. Fryer at 60, 62. He took this into account because he “was trying to find a way that would make the least impact on somebody’s career, somebody’s future” and not because he believed that if someone were close to retirement, then they were less competent. Id. at 62-63. In summarizing how he performed his rankings, Dr. Fryer testified that

when you have to play God and strike people from the register and everything else being equal, you try to do it with as little pain as possible; and age was a very useful factor in that for me.

Id. at 69. Fryer testified that he used the documents that contained the employees’ age to determine who was close to retiring. See id. at 66-67.

The plaintiffs and Roche agree that this testimony demonstrates that a number of Roche managers took into account proximity to retirement and/or eligibility for retirement benefits when determining who would be fired. However, the parties disagree as to whether this testimony constitutes evidence of disparate treatment in violation of the ADEA. See Roche Br. in Support of its Motion to Strike at 32-34; Plaintiffs’ Br. in Opposition to Roche’s Motion to Strike at 50-52.

Plaintiffs argue that consideration of these factors violates the ADEA because at Roche an employee’s proximity to retirement and eligibility for retirement benefits are a function of the employee’s age. Therefore, according to plaintiffs, if an employee were terminated based on one of these factors, then they were terminated because of their age. 25 See Plaintiffs’ Brief in Opposition to Hoffinann-La Roche Inc.’s Appeal from the Opinion of the Special Master Dated November 28,1994 Denying Partial Summary Judgment and Related Relief at 24 [hereinafter “Plaintiffs’ Hazen Paper Opposition Br.”]; Plaintiffs’ Br. in Opposition to Roche’s Motion to Strike at 51.

Citing the Supreme Court’s Hazen Paper decision, Roche argues that the ADEA is not violated when the employer’s decision is wholly based on factors other than age, even if the motivating factor is correlated with age. See Roche Br. in Support of its Motion to Strike at 34. Therefore, according to Roche, even if proximity to retirement and eligibility for retirement benefits are a function of age, firing employees based on these factors is not violative of the ADEA Id.

In Hazen Paper, the Supreme Court held that “an employer does not violate the ADEA just by interfering with an older employee’s pension benefits that would have vested by virtue of the employee’s years of service.” Hazen Paper, 507 U.S. at 613, 113 S.Ct. at 1707-08. Part of the reasoning underlying this holding was that

[i]t is the very essence of age discrimination for an older employee to be fired because the employer believes that productivity and competence decline with old age____ Congress’ promulgation of the ADEA was prompted by its concern that older workers were being deprived of employment on the basis of inaccurate and stigmatizing stereotypes.

Id. at 610, 113 S.Ct. at 1706 (citations omitted). In Hazen Paper, the employer’s motive in firing the plaintiff was to prevent plaintiffs pension benefits from vesting. Thus, the employer’s decision was wholly motivated by factors that were “analytically distinct” from age, and therefore, “the problem of inaccurate and stigmatizing stereotypes disappear[ed].” Id. at 611, 113 S.Ct. at 1706. Because the prohibited stereotype did not figure into the decision, the employer’s decision did not violate the ADEA. See id. at 611-15, 113 S.Ct. at 1707-08. This is the case even though the motivating factor— pension status — was correlated with age. Pension status is normally based on years of service. “Because age and years of service are analytically distinct, and an employer can take account of one while ignoring the other, and thus it is incorrect to say a decision based on years of service is necessarily ‘age-based.’ ” Id. at 611, 113 S.Ct. at 1707.

Plaintiffs argue that in this case proximity to retirement and eligibility for retirement benefits are not “analytically distinct” from age at Roche because both, by definition, are functions of age. “Ample retirement benefits” only applies to employees over 50 years old, and “proximity to retirement” only applies to employees within eight years of reaching age 65. See note 19 supra•; Plaintiffs’ Hazen Paper Opposition Br. at 24. Thus, according to plaintiffs, any decision made based on these factors violated the ADEA because “such a decision would per se have been made ‘because of age’ and it would not have occurred ‘but for’ the individual’s age.” Plaintiffs’ Hazen Paper Opposition Br. at 24.

Plaintiffs rely on Sheerin v. NYS Div. of Substance Abuse Services, 844 F.Supp. 909 (N.D.N.Y.1994) in support of their argument. In Sheerin, the court stated that

[b]ecause retirement eligibility is based on age (and years of service), an employer would violate the ADEA if it terminated an employee because he or she was eligible to retire. See [Hazen Paper, 507 U.S. at 611-13,] 113 S.Ct. at 1707 (leaving a similar question unanswered).[ 26 ]

Id. at 916 n. 11. Based on this case, plaintiffs argue that at Roche proximity to retirement and eligibility for retirement benefits are not analytically distinct irom age.

While the court in Sheerin correctly noted that the Supreme Court left a similar question unanswered in Hazen Paper, in that case the Supreme Court also stressed that the “very essence” of age discrimination was for an older employee to be fired because the employer believes the “inaccurate and stigmatizing stereotypes” “that productivity and competence decline with old age.” Hazen Paper, 507 U.S. at 610, 113 S.Ct. at 1706 (citations omitted). See also DiBiase v. SmithKline Beecham Corporation, 48 F.3d 719, 734 (3d Cir.) (statement of Judge Green-berg) (stating that “the primary purpose of [the ADEA] is to prohibit employers from acting on the assumption that ‘productivity and competence decline with old age,’ ” (quoting Hazen Paper, 507 U.S. at 610, 113 S.Ct. at 1706)), cert. denied, — U.S. -, 116 S.Ct. 306, 133 L.Ed.2d 210 (1995); EEOC v. Francis W. Parker School, 41 F.3d 1073, 1077 (7th Cir.1994) (interpreting Hazen Paper as holding that the employer did not violate the ADEA in that case “[b]ecause the decision to fire Biggins was not based on misperceptions about the competence of older workers”), cert. denied, — U.S. -, 115 S.Ct. 2577, 132 L.Ed.2d 828 (1995); Gokcen v. Babbitt, 999 F.2d 542, 1993 WL 283591 (9th Cir. July 27, 1993) (table) (holding that ADEA was not violated where “[r]etirement eligibility was considered only to minimize any adverse effect that closure [of the section where plaintiff worked] might have on the [employer] and its employees” and where “[n]othing in the record suggests the decision [to eliminate plaintiff’s position] was partially based on a belief that [plaintiffs] age prevented him from performing his duties competently”). In Sheerin, the court relied on the high, although not perfect, correlation between retirement eligibility and age in that case in finding that the defendant’s reliance on retirement eligibility violated the ADEA See Sheerin, 844 F.Supp. at 916 & n. 11. However, no where does the court discuss the primary purpose of the ADEA — the prohibiting of employers from acting on the assumption that productivity and competence decline with old age. For this reason, I find that Sheerin is not persuasive.

The Supreme Court in Hazen Paper clearly held that the ADEA is violated when employment decisions are the result of an “inaccurate and denigrating generalization about age,” Hazen Paper, 507 U.S. at 612, 113 S.Ct. at 1707, and that “[w]hen the employer’s decision is wholly motivated by factors other than age, the problem of inaccurate and stigmatizing stereotypes disappear” — even where the motivating factor is correlated with age, id. at 611, 113 S.Ct. at 1706. In this case, proximity to retirement and eligibility for retirement benefits are correlated with age. However, none of the deposition testimony relied on by plaintiffs suggest that the decisions made by those managers were based on denigrating stereotypes about older workers. 27 Rather, the testimony suggests that the managers were using proximity to retirement and eligibility for retirement benefits as a means to minimize the adverse effect that Operation Turnabout would have on Roche employees. See Gokcen v. Babbitt, 999 F.2d 542, 1993 WL 283591 (9th Cir. July 27, 1993) (table). In addition, there is no evidence that the managers making these decisions were targeting for termination employees who were close to retiring or who had ample retirement benefits based on the assumption that these employees would likely be older. See Hazen Paper, 507 U.S. at 611-13, 113 S.Ct. at 1707. Therefore, Mienhardt’s, Frankl’s, Anton’s, Tsehopp’s, and Fryer’s statements to the effect that some employees were terminated based on proximity to retirement and eligibility for retirement benefits are not evidence of disparate treatment based on age in violation of the ADEA

However, a reasonable jury could arguably infer age discrimination from two statements made by Meinhardt. First, he stated that, in performing the rankings, he considered “family situations, and how old a guy is and where his kids are and his financial circumstances to the best you know it, and it all falls into the judgment, and that’s what it is.” See Meinhardt at 76 (emphasis added). While this statement does not appear to reflect any denigrating stereotype regarding older workers (and seems to support the conclusion that managers were using age, as it related to eligibility for retirement benefits or proximity to retirement, as a means to minimize the adverse effect of Operation Turnabout), perhaps, looking at the statement in the light most favorable to plaintiffs, this statement evinces some animus against older workers. He also stated that when performing the rankings he asked himself

“[w]ould I want to keep somebody who is young and energetic and has got potential as opposed to somebody who has been around for whatever, 20, 25, 30 years,” and even though their performance currently meets expectations in accordance with the performance standards that we had____ It’s going to be a leaner tougher organization, expectations are going up.

Meinhardt at 45. This statement is more probative of age discrimination than the first. The reference to Roche becoming a “tougher organization” and that “expectations are going up” could easily be construed as reflecting the belief that older workers will not be able to perform in the new, tougher Roche simply because they are older. Thus, this statement evinces some animus against older workers. However, its probative value is limited because it is the statement of one manager and is not related to any specific employment decision.

Lastly, as stated above, the use of documents that listed, among other things, the employees’ ages is somewhat probative of intentional discrimination based on age. See Sheerin, 844 F.Supp. at 915-17. Thus, the anecdotal evidence of intentional discrimination during Operation Turnabout is not totally lacking, but is sparse.

The evidence just described constitutes the full panoply of plaintiffs’ pattern-or-practice evidence. In sum, plaintiffs have presented evidence of statistical disparities combined with circumstantial evidence of discrimination in the form of Dr. Lefkowitz’s Report, evidence relating to Roche’s corporate culture, and the existence and possible use of reports which contained information regarding, among other things, the ages of Roche employees. While all of this evidence is relevant to whether Roche discriminated in conducting Operation Turnabout, it is insufficient to establish that Roche regularly and purposefully engaged in disparate treatment based on age.

Particularly significant to this conclusion is the lack of any evidence regarding specific instances of discrimination against individual class members. In other cases, courts have found the number of examples of specific instances of discriminatory conduct against individual class members to be highly relevant to the determination of whether or not a defendant engaged in a pattern or practice of discrimination. See King, 960 F.2d at 624 (“Without significant individual testimony to support statistical evidence, courts have refused to find a pattern or practice of discrimination.”) (citing Cooper, 467 U.S. at 878-80, 104 S.Ct. at 2800-02; Teamsters, 431 U.S. at 337-38, nn. 17-18, 97 S.Ct. at 1855-56, nn. 17-18; Goff, 678 F.2d at 597; Chisholm, 665 F.2d at 495; Ste. Marie, 650 F.2d at 406).

For example, in Teamsters, the Supreme Court held that the Government plaintiff sustained its burden of proving a pattern or practice of discrimination where it bolstered its statistical evidence with the testimony of individuals who recounted over 40 specific instances of discrimination. See Teamsters, 431 U.S. at 337-39, 97 S.Ct. at 1855-56. Likewise, in Chisholm, the Court of Appeals for the Fourth Circuit affirmed the district court’s conclusion that the defendant committed racial discrimination in its promotion practices where statistical evidence was buttressed with the testimony of 20 former and present black employees of the defendant, who testified regarding their unsuccessful efforts to obtain promotions. Chisholm, 665 F.2d at 494-97. On the other hand, in Cooper, the Supreme Court summarized, with approval, the opinion of the Court of Appeals in that case as follows:

In essence, the Court of Appeals held that the statistical evidence, buttressed by expert testimony and anecdotal evidence by three individual employees ... was not sufficient to support the finding of a pattern of bankwide discrimination____ It is true that the Court of Appeals was unpersuaded by the anecdotal evidence; it is equally clear, however, that it did not regard two or three instances of discrimination as sufficient to establish a general policy. It quite properly recognized that a court must be wary of a claim that the true color of a forest is better revealed by reptiles hidden in the weeds than by the foliage of countless free-standing trees. Conversely, a piece of fruit may well be bruised without being rotten to the core.

Cooper, 467 U.S. at 878-80, 104 S.Ct. at 2801 (citations omitted) (internal quotation marks omitted). Likewise, in King, the Court of Appeals for the Seventh Circuit stated that, regardless of the statistical evidence indicating that the RIF at issue had a disproportionate effect on older workers, the testimony of five plaintiffs 28 that they were demoted or laid-off while the defendant continued to employ younger workers in their departments and testimony that some GE managers made disparaging comments about older workers was insufficient to support a finding that the defendant engaged in a pattern or practice of discrimination in conducting the RIF. King, 960 F.2d at 619, 627. See also Goff, 678 F.2d at 597 (stating that “even if all three witnesses’ accounts of racial discrimination were true, this evidence would not have been enough to prove a pattern or practice of company-wide discrimination”); Ste. Marie, 650 F.2d 395, 406 (2d Cir.1981) (finding that subjective decisionmaking procedures and seven individual discriminatory acts coupled with problematic statistical evidence were insufficient to support a finding of pattern or practice of discrimination).

In the case now before the court, the evidence that plaintiffs point to as evidence of specific instances of discrimination against class members demonstrates only that a number of persons over 40 years old were terminated because they were close to retiring or were eligible for retirement benefits. Firing employees for these reasons, however, is not violative of the ADEA. Thus, this case seems more like Cooper, King, Goff, and Ste. Marie than it does like Teamsters and Chisholm. This is not to suggest that there is a precise number of instances of discriminatory conduct against class members that must be proved in order for plaintiffs’ to sustain their burden in proving a pattern or practice of discrimination. See EEOC v. Chicago Miniature Lamp Works, 947 F.2d 292, 297 (7th Cir.1991) (stating that a prima facie case for a pattern or practice disparate treatment can be established by “statistical evidence demonstrating substantial disparities in the application of employment actions ..., buttressed by evidence of general policies or specific instances of discrimination”) (emphasis added); Coates v. Johnson & Johnson, 756 F.2d 524, 533 (7th Cir.1985) (“Neither statistical nor anecdotal evidence is automatically entitled to reverence to the exclusion of the other.”); Ste. Marie, 650 F.2d at 406 (“If there were evidence that a policy of discrimination had been adopted, perhaps two or even one confirmatory act would be enough [to establish a pattern or practice of discrimination].”). However, even when compared to cases in which a pattern or practice has been found based, in part, on only a few examples of discrimination against individual plaintiffs, the evidence in this case is still insufficient to support such a finding.

For example, in Lansdowne Swim Club, the Government was able to demonstrate only that three black families had been rejected for membership in the club. See Lansdowne Swim Club, 894 F.2d at 85. However, the district court stated that

[although the number of discriminatory rejections in this case is small, the rejections are not merely isolated and sporadic because the small number is attributable to LSC’s discriminatory reputation in the Lansdowne community and its deterrence of potential black applicants, which resulted in a small number of black applicants.

Lansdowne Swim Club, 713 F.Supp. at 818. Furthermore, the LSC did not merely reject each family once, rather, these families were repeatedly rejected over a ten year period. See Lansdowne Swim Club, 894 F.2d at 88-89 (holding that district court’s finding that “the repeated rejections of three qualified black applicants [were] highly probative of a pattern or practice of discrimination”). This evidence was also supported by the following additional evidence of discrimination. Over the LSC’s thirty year history, every non-black applicant obtained membership at some time, while only one part-black family was admitted. Id. at 89. At the time of the LSC’s founding, it did not reach its goal of having 300 members, therefore, the LSC solicited applications from nonresidents of Lansdowne. The LSC did not, however, solicit applications from black residents of Lansdowne. Id.; Lansdowne Swim Club, 713 F.Supp. at 818-19. Lastly, the district court specifically found that three families, in addition to the three families whose application were rejected, were deterred from applying for membership due to their race. Lansdowne Swim Club, 894 F.2d at 89; Lansdowne Swim Club, 713 F.Supp. at 819-20. Thus, in Lansdowne Swim Club, the plaintiff was able to buttress the small number of specific instances of discriminatory conduct with evidence that explained the reason for the small number of instances (i.e., blacks were deterred from applying) and with evidence that the LSC had a long history of discrimination. In the case now before the court, there is no similar reason as to why plaintiffs have been unable to demonstrate specific instances of discriminatory conduct in the conducting of Operation Turnabout. In addition, unlike the plaintiff in Lansdowne, the plaintiffs in this case have not demonstrated that Roche has a history of discriminating against older workers.

Another case in which the court found a pattern or practice of discrimination based, in part, on only a few examples of individual discrimination is Mozee v. American Commercial Marine Serv. Co., 940 F.2d 1036 (7th Cir.1991), cert. denied, 506 U.S. 872, 113 S.Ct. 207, 121 L.Ed.2d 148 (1992). In Mozee, the plaintiffs produced evidence supporting the conclusion that only five individual plaintiffs were discriminated against. See id. at 1051. However, this evidence was supported by (1) the statistical disparities that plaintiffs produced to support their disparate impact claim, which was based on the use of subjective criteria; (2) the defendant’s noncompliance with its various affirmative action plans; (3) the existence of discriminatory markings on the files of all African-Americans; and (4) discriminatory signs posted in the work place. Id. Unlike the plaintiffs in Mozee, the plaintiffs in this case have not produced evidence that any individual plaintiff was discriminated against, and have not produced evidence of past discriminatory conduct that is analogous to the Mozee defendant’s failure to comply with its affirmative action policies.

Plaintiffs assert that EEOC v. Sandia Corp., 639 F.2d 600 (10th Cir.1980) is an analogous case in which a court found a pattern or practice of discrimination in the conducting of a RIF. Specifically, plaintiffs assert that

[t]he Sandia Court relied heavily on the fact that (1) the procedures followed for selecting employees for involuntary termination relied on a highly subjective and invalidated performance evaluation system, id. at 612-14; (2) management was concerned about that increasing average age of the corporation’s employees, id. at 608; (3) the company’s management was convinced that “new blood” or young Ph.D.’s were the key to the company’s future, id., and (4) qualification for benefits under the retirement plan was a factor in choosing older employees for layoffs, id.. The Sandia Court further observed that management’s concern about the increasing average age of its workforce, and its emphasis on recruiting and advancing young Ph.D.’s, could “easily be reflected in subjective performance ratings.” Id. at 614. Thus, Sandia’s subjective evaluation system, which was the basis for selecting individuals for termination, id. at 604, 614 was found to have a “built-in bias against the protected age group.” Id. at 614. Based on these facts, as well as statistical proof, the Court found that Sandia had engaged in a pattern or practice of age discrimination in terminating employees in March, 1973. Id. at 621.

Plaintiffs’ Supplementation Br. at 40. Plaintiffs argue that the case now before the court is analogous because there is evidence that (1) Roche’s termination ■ procedures were , highly subjective and unvalidated; (2) Roche management was concerned about the increasing average age of Roche employees; (3) Roche management believed in the importance of bringing in younger Ph.D.’s and moving out older scientists; and (4) eligibility for retirement benefits was a key factor in termination decisions. Id.

Despite plaintiffs’ arguments, Sandia is distinguishable from the present case in a number of significant respects. First, although, like the defendant in Sandia, there is evidence in the record indicating that Roche was concerned with the average age of its workforce, compared to Sandia, there is an insufficient amount of evidence to support a finding that Roche set out to systematically remedy this situation. In this case, there is evidence from which a jury could find that Dr. Kuntzman suggested that older, senior level Ph.D.’s be replaced with younger Ph. D.’s. See Saucy at 69-70. There is, however, an insufficient amount of evidence to support a finding that this suggestion was implemented.

On the other hand, in Sandia, in addition to the evidence that in the late 1960’s and early 1970’s Sandia became concerned about the average age of its work force, there was evidence that Sandia took specific steps to remedy this situation. For example, recruiters were given age guidelines. Applicants who received their Bachelor’s Degree after the age of 23, their Master’s Degree after the age of 25-26, or their Ph.D. after the age of 30, were required to explain why they were delayed in receiving the degree. Sandia, 639 F.2d at 610. Furthermore, there were comments with respect to these applicants such as “age should not be a detriment or that Sve should overlook his age.’ ” Id.

In addition, there were a series of memoranda received into evidence that were designated “Bright Young Men Memos.” There were approximately 25 such memos covering the period of 1965-75. See id. at 611. In the text of these memos, there were references to the applicant as a bright or eager young man, and in some cases, “the writers observed that the individual is over 30 and then state why they are not concerned about his advanced age.” Id. ■

Furthermore, the court found that there was sufficient evidence regarding Sandia’s promotion policy “to cause ‘a reasonable suspicion’ that age was a factor in promoting members of the technical staff to the first supervisory level.” Id. at 612. Moreover, the court found that the system for evaluating the performances of individual employees, which was largely for the purpose of administering salary increases, was highly subjective. See id. at 612-13. Other salary policies included

the rejection of salary increases of less than five percent as anti-motivational and possibly insulting, and the requirement that salary increases be granted no more frequently than annually, the system known as “stretch-out.”

Id. at 613. The court found that the performance rating system had a built-in bias against the protected age-group because

“[m]anagement’s concern about the increasing age of its staff, reduced hiring, new technical developments, and emphasis on recruiting and advancing young Ph.D.’s might not violate ADEA in themselves, but these policies could easily be reflected in subjective performance ratings.”

Id. at 614 (quoting district court).

With respect to the documentary evidence, the district court stated:

The Secretary introduced numerous exhibits showing that Sandia’s management and lower line supervision tended toward stereotyping older non-supervisory employees as unproductive and becoming technically obsolete. Management also believed that the future of the Lab depended upon “new blood.” ... The end result was an arbitrary generalization that older employees did not have the ability to keep pace with new developments. This “attitude” evidence corroborates the statistical evidence and supports an inference that age was a factor in selection for layoff.

Id. (quoting district court). Based on this evidence, the Court of Appeals of the Tenth Circuit affirmed the district court’s finding that Sandia discriminated in conducting the RIF. Id. at 615.

Unlike Sandia, there is no evidence of specific steps taken by Roche to hire younger scientists prior to Operation Turnabout. Furthermore, there is no evidence that at the time of the RIF any Roche policy had a built in age bias like the performance evaluation system at Sandia. On the contrary, in this case plaintiffs assert that the merit ratings and performance evaluations that Roche used to determine salary increases were “reliable measures of employee performance.” Plaintiffs’ Supplementation Br. at 9. Lastly, unlike Sandia where age discriminatory attitudes were reflected in “numerous exhibits” and the concerted effort to hire younger employees, and where it was demonstrated that such attitudes were held by both management and lower line supervision, in this case, the evidence of age discriminatory attitudes is limited to a couple of stray remarks by Roche’s upper level management. In addition, other than this evidence of corporate culture, there is no evidence of such attitudes being held by the lower level management who the plaintiffs assert decided who would be fired in Operation Turnabout. For these reasons, Sandia is distinguishable from the case now before the court.

■ The other cases on which plaintiffs rely are also distinguishable. As discussed supra in section IV.B., plaintiffs rely on a number of cases for the proposition that evidence of “undisciplined and ill-defined selection criteria coupled with uncontrolled decision-making by individual managers”

supports a pattem-or-praetice claim. See Plaintiffs’ Supplementation Br. at 37. Many of these cases are inapposite to the issue now before the court because they are disparate impact cases and do not discuss pattern-or-practice issues. See Crawford, v. Western Electric Company, Inc., 745 F.2d 1373 (11th Cir.1984); Robinson v. Union Carbide Corp., 538 F.2d 652 (5th Cir.1976), aff'd in part, rev’d in part and remanded on reh’g, 544 F.2d 1258 (5th Cir.), cert. denied, 434 U.S. 822, 98 S.Ct. 65, 54 L.Ed.2d 78 (1977); Wade v. Mississippi Cooperative Extension Service, 528 F.2d 508 (5th Cir. 1976); Rowe v. General Motors Corporation, 457 F.2d 348 (5th Cir.1972). Others are inapposite because they only address individual disparate treatment claims, and no where discuss pattem-or-practice claims. See Franci v. Avco Corporation, 538 F.Supp. 250 (D.Conn.1982) (ADEA class action that resulted from RIF wherein the court analyzed the claim pursuant to the McDonnellr-Douglas framework; there was no mention of “pattern or practice” in the opinion); Marimont v. Califano, 464 F.Supp. 1220 (D.D.C.1979) (individual disparate treatment claim brought pursuant to Title VII).

Lastly, also as discussed supra in section III.B., plaintiffs also rely on Mozee v. American Commercial Marine Service Company, 940 F.2d 1036, 1042 (7th Cir.1991) and Green v. USX Corporation, 896 F.2d 801, 807 (3d Cir.), cert. denied, 498 U.S. 814, 111 S.Ct. 53, 112 L.Ed.2d 29 (1990). These cases merely stand for the propositions that plaintiffs may challenge employment actions under both disparate impact and disparate treatment theories in the same case and that statistical evidence of a disparate impact is also relevant evidence regarding whether an employer engaged in intentional discrimination.

In this case, however, even after considering plaintiffs’ statistical evidence of disparate impact, I find that no reasonable jury could conclude that Roche engaged in a pattern or practice of age discrimination in conducting Operation Turnabout. Plaintiffs allege that lower level managers decided who would be terminated in Operation Turnabout. In addition, plaintiffs allege that these managers had virtually complete discretion in making these decisions, and that their decisions were not subject to review by Roche’s higher level managers. There is clearly evidence in the record to support these allegations. However, this evidence implies that each manager, or perhaps, each department or division within Roche, went about determining who would be fired independently of the other managers, departments or divisions.

In order to connect these decisions into a pattern or practice of discrimination, plaintiffs assert that this discretion allowed Roche’s managers to act on their pre-existing age bias. The problem is there is an insufficient amount of evidence in the record to support a finding that there was widespread, pervasive pre-existing age bias among the Roche managers who made the decisions. The remarks by Roche executives, the awareness of the high average age of the Roche work force, and the use by some managers during the RIF of documents listing employees’ ages are all probative of intentional discrimination. However, even when reinforced with the statistical evidence, this evidence is far short of the “substantial proof of the [discriminatory] practice” that other courts have required to prove a pattern or practice of discrimination. See King, 960 F.2d at 624 (citing cases). In short, plaintiffs have failed to submit evidence comparable to that introduced in other cases in which courts have found a pattern or practice of discrimination. Particularly significant, is that all of the evidence that plaintiffs point to regarding specific instances of discrimination in Operation Turnabout consists of testimony that some employees were fired based on their proximity to retirement and eligibility for retirement benefits. Decisions motivated by these factors do not violate the ADEA. Therefore, after viewing the evidence in the light most favorable to the plaintiffs, I find that a reasonable jury could not find that Roche engaged in a pattern or practice of discrimination.

For these reasons, Roche’s motion for summary judgment on plaintiffs’ pattem-orpractice claim is granted. Because I am granting that motion, I must vacate the Special Master’s decision granting plaintiffs’ mo tion to bifurcate the trial in this matter. Given my ruling, the Special Master should now address the question of in what form will the trial of this action take place. I note, however, that it may be slightly premature to address this issue now. This is because plaintiffs have also asserted a classwide disparate impact claim. Roche has indicated that it is going to move to dismiss that claim on the grounds that disparate impact claims are not cognizable under the ADEA. The resolution of this motion will obviously affect how the case will be tried. Therefore, the parties should bring the disparate impact motion, and any other motion that may affect the form of the trial in this action, to the attention of the Special Master with alacrity.

V. Conclusion

For the reasons detailed above, Roche’s motion for summary judgment on plaintiffs’ pattem-or-practice claim is GRANTED, plaintiffs’ motion for bifurcation is DENIED, and Roche’s motions to strike are DENIED.

1 . After making a determination that this matter involved complex issues of fact and law, I appointed, pursuant to Federal Rule of Civil Procedure 53, Alan Schwartz, Sterling Professor of Law at the Yale Law School, as special master. The order appointing Professor Schwartz provided, in part, that

2. Professor Schwartz’s powers and duties as special master in this case shall be as follows:

B. To consider and resolve expeditiously any and all present and future disputes between the parties relating to discovery and other nondispositive motions made prior to the time of trial by jury of the legal issues in this action.

C. To fully consider and prepare reports to be submitted to the Court, including an exposition of all relevant facts and conclusions of law, concerning any and all present and future dispositive motions made prior to the time of trial by jury of the legal issues in this action.

See Order, filed on November 21, 1991, at 2.

2 . Roche has also moved (a) for partial summaiy judgment dismissing the individual disparate treatment claims of sixty plaintiffs, and (b) for an in limine ruling excluding certain evidence, based on the Supreme Court’s opinion in Hazen Paper Company v. Biggins, 507 U.S. 604, 113 S.Ct. 1701, 123 L.Ed.2d 338 (1993). These motions were also referred to the Special Master. In an opinion dated November 28, 1994, the Special Master held that due to the existence of a pattem-or-practice claim in this case, it would be inappropriate to dismiss the 60 class members’ individual disparate treatment claims until the pattem-or-practice issue was resolved. See Opinion of Special Master, dated November 28, 1994, at 7. Therefore, I will address Roche’s motions based on Hazen Paper in a subsequent opinion, after resolving Roche’s motion for summazy judgment on plaintiffs’ pattem-or-practice claim.

3 . Despite the fact that the plaintiffs clearly requested the opportunity to present the court with the “full panoply” of their pattem-or-practice evidence, plaintiffs assert that it would be improper for the court, on the record now before it, to determine if there is a genuine issue of material fact regarding whether Roche engaged in a pattern or practice of discrimination. See Plaintiffs’ Supplementation of the Record Relating to Hoffman-La Roche’s Appeal of the Special master’s Order Denying Roche’s Cross-Motion for Summaiy Judgment Regarding Plaintiffs' Pattern or Practice and/or Classwide Prospective Injunctive Relief Claims at 1 [hereinafter “Plaintiffs’ Supplementation Br.”] (stating that “[n]either Roche's original summaiy judgment cross-motion, nor its appeal of the denial thereof, sought to test the factual sufficiency of Plaintiffs’ entitlement to proceed on a pattern or practice basis.”); Flamm Letter, dated April 16, 1996, at 3 n. 2 (stating that "[o]n this record, it would be improper for the Court to even consider a summaiy judgment motion of the type requested by Roche,” i.e., Roche argues that plaintiffs' pattem-or-practice proofs do not raise a genuine issue of material fact).

Plaintiffs’ counsel does not, however, explain why it would be improper to consider this issue on this record. Plaintiffs state in the brief they submitted in connection with the Rule 56(f) supplementation that "the evidence [submitted] is of an. all-inclusive nature and will demonstrate that Plaintiffs have amply made a factual showing entitling them to proceed with their claims on a 'pattern or practice’ basis.” Plaintiffs’ Supplementation Br. at 2. Therefore, plaintiffs seem to concede that the factual sufficiency of their claim is at issue on this summary judgment motion.

In a recent letter to the court, plaintiffs' counsel states that "[pllaintiffs’ statistical expert ... has conducted additional analyses since receiving Roche’s expert's data,” but plaintiffs have not submitted these additional analyses to the court "because the Court’s decision on the pending appeals and motions does not turn on a detailed evaluation of the statistical evidence.” Flamm Letter, dated April 16, 1996, at 5 n. 5. Plaintiffs’ failure to submit these additional analyses to the court does not render a decision on this issue on the record now before the court improper. This is because in the November 2, 1995 Letter Opinion & Order granting plaintiffs’ Rule 56(f) request, I ordered Roche to have its expert report completed and disclosed to plaintiffs no later than December 1, 1995 so that plaintiffs would have an opportunity to address the information contained in that report in its Rule 56(f) supplementation papers. The deadlines for the parties' submissions were subsequently extended by a week. Thus, if plaintiffs needed extra time in order to complete these additional analyses after receiving Roche’s expert report, they should have brought this to the attention of the court in December. Therefore, the failure of the plaintiffs to submit these additional analyses does not render this record an improper basis on which to rule on this issue.

I also note that the Special Master has informed the court that the plaintiffs have not yet deposed Roche's expert. However, in concluding, among other things, that plaintiffs have made an insufficient factual showing on the issue of whether Roche engaged in a pattern or practice of discrimination, I did not rely on Roche's expert report. Therefore, that plaintiffs have not yet deposed Roche’s expert does not render this record an improper basis on which to rule in this issue.

In conclusion, the record now before the court provides a sufficient basis on which to determine if there is a genuine issue of material fact regarding whether Roche engaged in a pattern or practice of discrimination.

4 . The statute was amended in 1972 to authorize the Equal Opportunity Employment Commission ("EEOC”), rather than the Attorney General, to bring pattem-or-practice suits against private-sector employers. The Attorney General retains authority in the public sector. See 42 U.S.C. § 2000e-6(c).

5 . In addition to their pattern-or-practice claim, plaintiffs also assert that Operation Turnabout had a disparate impact on Roche's older employees. Disparate impact claims "involve employment practices that are facially neutral in their treatment of different groups but that in fact fall more harshly on one group than another and cannot be justified by business necessity.” Teamsters, 431 U.S. at 335 n. 15, 97 S.Ct. at 1854 n. 15 (citations omitted). In contrast to disparate treatment claims, proof of discriminatory motive is not required to succeed on a disparate impact claim. Id. (citations omitted). This claim is not before the court in this motion.

6 . In certain situations the individual class member may be able to demonstrate that he or she was a potential victim of discrimination even though he or she did not apply for the position. See Teamsters, 431 U.S. at 362-71, 97 S.Ct. at 1868-73. "When a person's desire for a job is not translated into a formal application solely because of his unwillingness to engage in a futile gesture he is as much a victim of discrimination as is he who goes through the motions of submitting an application.” Id. at 366, 97 S.Ct. at 1870. This situation is not present in the case now before the court because this is a discharge case rather than a failure to hire case, and therefore, because all the plaintiffs in this case were obviously fired in Operation Turnabout, they are all potential victims of the alleged pattern or practice of basing termination decisions on age.

7 . In Teamsters, the Court stated that shifting the burden of proof to an employer who is found to have engaged in a pattern or practice of discrimination is also justified because (1) "the finding of a pattern or practice changed the position of the employer to that of a proved wrongdoer” and (2) “the employer was in the best position to show why any individual employee was denied an employment opportunity.” Teamsters, 431 U.S. at 359 n. 45, 97 S.Ct. at 1867 n. 45. Although the presumption was also justified on these grounds, the evaluation of evidentiary probabilities seems to be the main reason for shifting the burden of persuasion to an employer found to have engaged in a pattern or practice. See Price Waterhouse, 490 U.S. at 266-67, 109 S.Ct. at 1799-1800 (O'Connor, J., concurring) (stating that a member of a pattem-or-practice class action "receives the benefit of a burden shift to the defendant based on the likelihood that an illegitimate criterion was a factor in the individual employment decision” and not mentioning any other justification for shifting the burden); King, 960 F.2d at 623 (stating that burden shifts because once a pattern or practice has been demonstrated it is "reasonable to infer that the employer discriminated against particular individuals”). This makes sense because the employer will always be in a better position to show why an individual was denied an employment opportunity, regardless of whether or not an employer is found to have engaged in a pattern or practice of discrimination.

8 . As will be discussed more in depth infra in section IV.B., plaintiffs rely on evidence that several Roche executives made discriminatory remarks in the years prior to Operation Turnabout in support of their proposition that there was pre-existing age bias among Roche managers. However, also as will be discussed more in depth infra in section IV.B., although these remarks are relevant to show the corporate culture in which a company makes its employment decisions, and may be used to build a circumstantial case of discrimination, they should not be given significant or commanding weight. See Brewer v. Quaker State Oil Refining Corporation, 72 F.3d 326, 333-34 (3d Cir.1995). Therefore, there is clearly insufficient evidence in the record to support a finding that Roche systematically violated the ADEA prior to Operation Turnabout.

Furthermore, “since Operation Turnabout only 11 claims of age discrimination have been filed against Roche, and in none of these cases has there been a finding that Roche engaged in age discrimination.” Chelel Aff. at ¶4 (attached at exhibit 8 to Zackin Supplemental Affidavit). Thus, there is also insufficient evidence to support a finding that Roche systematically violated the ADEA after Operation Turnabout.

9 . It should be noted that several courts have held that disparate impact claims are not cognizable under the ADEA. See Ellis v. United Airlines, Inc., 73 F.3d 999 (10th Cir.1996); EEOC v. Francis W. Parker School, 41 F.3d 1073 (7th Cir. 1994), cert. denied, - U.S. -, 115 S.Ct. 2577, 132 L.Ed.2d 828 (1995); Martincic v. Urban Redevelopment Authority of Pittsburgh, 844 F.Supp. 1073 (W.D.Pa.1994). See also DiBiase v. Smith-Kline Beecham Corporation, 48 F.3d. 719, 732-34 (3d Cir.) (statement of Judge Greenberg) (stating that “it is doubtful that traditional disparate impact theory is a viable theory under the ADEA”), cert. denied, — U.S. -, 116 S.Ct. 306, 133 L.Ed.2d 210 (1995).

10 . Citations herein that use the format of a proper name followed by page numbers refer to the excerpts from the deposition transcripts that are attached to the Miller Affidavit.

11 . It should be noted that Roche has claimed the attorney-client privilege with respect to any specific statements made by Swift at these meetings. See Swift at 42-43.

12 . Pursuant to Federal Rule of Civil Procedure 56(e), Roche moves to strike Dr. Lefkowitz's Report, which was submitted to the court attached as Exhibit 4 to an affidavit by Katherine Miller, Esq., because (1) Ms. Miller lacks personal knowledge of the matters set forth in Dr. Lefkowitz’s report; (2) any facts or opinions in that report would not be admissible to the extent they were offered into evidence by Ms. Miller; and (3) Ms. Miller is not competent to testify to the matters stated in Dr. Lefkowitz’s report.

Roche has also recently moved to strike Dr. Lefkowitz's report on the grounds that it is irrelevant to plaintiffs' pattem-or-practice claim. By letter dated, April 16, 1996, plaintiffs' counsel objected to this motion by Roche on the grounds of, among other things, that orderly case management would be enhanced by having Roche withdraw its latest motion to strike.

In any event, I am not going to consider either of Roche’s motions to strike the report of Dr. Lefkowitz. This is because even after considering the report, I am granting Roche’s motion for summary judgment. Therefore, Roche’s motions to strike are denied.

13 . The comporatio equals salary divided by the simple average of the maximum and minimum salary for the employee’s grade level. The percent of range equals the difference between salary and grade minimum divided by the difference between grade maximum and grade minimum.

Thus, if an employee was earning $58,000 within a grade with a salary minimum-of $50,-000 and a salary maximum of $60,000, the . comporatio would be $58,000 divided by $55,-000 (the simple average of $50,000 and $60,-000), or 1.05, and the percent of range would equal $8,000 ($58,000 minus $50,000) divided by $10,000 ($60,000 minus $50,000) or .8.

Bloch Report at 3.

14 . To avoid the problem of employees appearing in more than one pool, Dr. Bloch reduced the number of pools with non-overlapping employees to 143. Bloch Report at 5-6. The number of employees in the pools is further reduced when merit ratings are taken into account because Roche was unable to furnish merit ratings for all employees. Id. As indicated above, when including merit ratings in the analysis, Dr. Bloch divided the pools into two groups — pools 1-115 and pools 117-143. Id. No explanation appears in the record regarding what happened to pool 116.

15 . Roche has moved to strike a great deal of this evidence on the grounds that much of it is inadmissible hearsay, and also on other grounds that go primarily to the probative value of the evidence. Because even after considering this evidence I am granting Roche's motion for summary judgment on plaintiffs' pattem-or-practice claim, I am denying Roche’s motion to strike. However, I will consider the arguments Roche made in support of its motion to strike to the extent that they bear on the probative value of this evidence.

16 . Line drivers engage in long-distance hauling between company terminals. Teamsters, 431 U.S. at 329-30 n. 3, 97 S.Ct. at 1852 n. 3.

17 . Following the trial in Teamsters, but before the district court made a decision, the Government and the company entered into a consent decree, which provided for classwide prospective relief. See Teamsters, 431 U.S. at 330 n. 4, 97 S.Ct. at 1852 n. 4. The consent decree, however, did not constitute an adjudication on the merits. Id. Therefore, the district court still had to issue a decision regarding whether or not the company engaged in a pattern or practice of discrimination. Id.

18 . The case is actually the result of a consolidation of two separate cases arising from the RIF. The cases were consolidated to litigate the issue of whether General Electric engaged in a pattern or practice of age discrimination. See King, 960 F.2d at 619. The Court of Appeals held that the district court abused its discretion by consolidating these actions. Id. at 626. However, the court went on to address the question of whether the plaintiffs presented sufficient evidence to support the juiy's finding that a companywide practice of discrimination existed at GE between 1982 and 1983. See id. at 626-28.

19 . The case was remanded for new trials due to the problematic consolidation of the two cases. See note 14, supra.

20 . It should be noted that although Roche provided plaintiffs with pools of people who they claim were compared to each other in Operation Turnabout, the information provided by Roche only comprises 1697 of the 7781 Roche employees at Roche prior to Operation Turnabout and 385 of the over 1,100 terminated employees. See Bloch Report at 5-6.

21 . To avoid the problem of employees appearing in more than one pool, Dr. Bloch reduced the number of pools with non-overlapping employees to 143. Bloch Report at 5-6. The number of employees in the pools is further reduced when merit ratings are taken into account because Roche was unable to furnish merit ratings for all employees. Id.

22 . In addition, it should be noted that "[w]here gross statistical disparities can be shown, they alone may in a proper case constitute prima facie proof of a pattern or practice of discrimination.” Hazelwood, 433 U.S. at 307-08, 97 S.Ct. at 2741-42. In Hazelwood, however, the Supreme Court also stated that if the Government plaintiff in that case were able to demonstrate standard deviations of between five and six, then that evidence would “reinforce rather than rebut” the Government's other proof. Id. at 311 n. 17, 97 S.Ct. at 2743 n. 17. The Court did not suggest that standard deviations in this range by themselves would constitute prima facie proof of a pattern or practice of discrimination. In the case now before the court, plaintiffs’ most relevant statistics yield standard deviations of between 2 and 6. Thus, these results are not the "gross disparities” that would by themselves constitute a prima facie pattem-or-practice case. See also Lopez v. Laborers Intern. Union Local No. 18, 987 F.2d 1210, 1214 n. 13 (5th Cir.1993) (stating, without deciding, that “it is doubtful” that Supreme Court's statement, that standard deviations of greater than 2 or 3 undercut the hypothesis that decisions were made randomly, supports an "ironclad rule” "that anytime there is a difference of three standard deviations between the expected and observed number a prima facie case has been made out”; and also observing that in a prior Fifth Circuit opinion court found intentional discrimination based solely on statistics that yielded a result of 10.24 standard deviations, which constituted “a truly ‘gross' disparity”); Cota v. Tucson Police Department, 783 F.Supp. 458, 467 (D.Ariz.1992) (stating that cases where prima facie case of pattern or practice is demonstrated solely by statistics "are rare”). Rather, plaintiffs’ statistical evidence may be used to reinforce their other evidence.

23 . Arguably, Lerner was a decisionmaker in that he made the decision to approve the Guidelines. In that case, the result would still be the same because the statements would then be temporally remote, stray remarks made by a decisionmaker unrelated to the decision process. See Brewer, 72 F.3d at 333.

24 . Other evidence with respect to Lemer’s alleged age bias includes George Reilly's deposition testimony that "[he] always felt that Lemer had a history of terminating senior employees.” Reilly at 134. This statement is inadmissible because Reilly admitted in a subsequent interrogatory that he had "no personal, direct knowledge of whether Mr. Lemer’s decisions to terminate any particular individual was influenced by age.” Miller Aff. Ex. 14. See Fed.R.Evid. 602 (“A witness may not testify to a matter unless evidence is introduced to support a finding that the witness has personal knowledge of the matter.”).

25 . In connection with the parties’ submissions to the court regarding Roche's motion for partial summary judgment dismissing the individual disparate treatment claims of sixty plaintiffs based on Hazen Paper, the plaintiffs defined nine factors that they allege were impermissibly considered in Operation Turnabout. Two of the factors are “ample retirement benefits” and "proximity to voluntary retirement." See Brief of Hoffman-La Roche Inc. in Support of Appeal from Opinion of Special Master Dated November 28, 1994 Denying Partial Summary Judgment and Related Relief at 17-19, 21-22 [hereinafter "Roche Hazen Paper Br.”]. “Ample retirement benefits” is defined by the following three part test:

(1) Plaintiff was at least 50 years old or was 49 years old and was offered a one-year leave of absence;

(2) Plaintiff had 10 or more years of credited service with Roche; [and,]

(3) Plaintiff was entitled to receive aggregate retirement and/or insurance benefits amounting to at least 35% of plaintiff's base annual compensation.

Id. at 17 (emphasis added). "Proximity to Voluntary Retirement” is defined by the following two part test:

(1) Plaintiff's actual age at termination was 50 or above; [and]

(2) Plaintiff's number of years of remaining employment was less than 8 years — determined by subtracting plaintiffs age from his or her regular retirement age of 65 or stated retirement age.

Id. at 21 (emphasis added). The highlighted portions of the above definition indicate that these two factors are, at least in part, a function of the employee's age.

26 . The Supreme Court stated in that "we do not consider the special case where an employee is about to vest in pension benefits as a result of his age, rather than years of service ... and the employer fires the employee in order to prevent vesting.” Hazen Paper, 507 U.S. at 613, 113 S.Ct. at 1707.

27 . In fact, in the case of Thomas Anton, he specifically testified that he "felt that Joe Zelauskas was acceptable and better in the performance of his job but we felt that he would be the least hurt by being let go because he was somewhat near retirement anyway.” Anton at 28-29 (emphasis added).

28 . Six plaintiffs actually testified at the pattern- or-practice trial in King. Id. at 619. However, the Court of Appeals stated that the testimony of one of the plaintiffs was not probative of a pattern or practice because he testified regarding discrimination that took place outside the time the alleged discriminatory policy was in effect. See id. at 628.