No. 03-12230 IN THE UNITED STATES COURT OF APPEALS FOR THE ELEVENTH CIRCUIT Cornelius COOPER, Michael EDWARDS, Charcella GREEN, Patricia HARRIS, Sarah Jean HARRIS, Irene McCULLERS, and Carolyn WILSON, Plaintiffs-Appellants, v. SOUTHERN COMPANY, GEORGIA POWER COMPANY, SOUTHERN COMPANY SERVICES, INC., and SOUTHERN COMPANY ENERGY SOLUTIONS, INC., Defendants-Appellees. ____________________________________ On Appeal from the United States District Court for the Northern District of Georgia (Atlanta) Hon. Orinda D. Evans ____________________________________ BRIEF OF THE EQUAL EMPLOYMENT OPPORTUNITY COMMISSION AS AMICUS CURIAE IN SUPPORT OF PLAINTIFFS-APPELLANTS AND REVERSAL ____________________________________ NICHOLAS M. INZEO Acting Deputy General Counsel PHILIP B. SKLOVER Associate General Counsel CAROLYN L. WHEELER Assistant General Counsel DORI K. BERNSTEIN Attorney EQUAL EMPLOYMENT OPPORTUNITY COMMISSION Office of General Counsel 1801 L Street, N.W., Room 7046 Washington, DC 20507 (202)663-4734 Cooper, et al. v. Southern Company, et al., No. 03-12230 CERTIFICATE OF INTERESTED PERSONS Pursuant to 11th Cir. R. 26.1-1, I hereby certify that the following persons or entities have an interest in the outcome of this case: 1. Chris Arbery, Esq. 2. Emmet Bondurant, Esq. 3. Dori Bernstein, Esq. 4. Johnnie Cochran, Jr., Esq. 5. Cornelius Cooper 6. Von DuBose, Esq. 7. Michael Edwards 8. Equal Employment Opportunity Commission (EEOC) 9. The Honorable Orinda Evans 10. Georgia Power Company, Inc. 11. Richard Gerakitis, Esq. 12. J. Keith Givens, Esq. 13. Charcella Green 14. Ashley Hager, Esq. 15. Patricia Harris 16. Sarah Jean Harris Cooper, et al. v. Southern Company, et al., No. 03-12230 17. Nicholas Inzeo, Esq. 18. Irene McCullers 19. Ray Persons, Esq. 20. Stephen Riddell, Esq. 21. Steven Rosenwasser, Esq. 22. Hezekiah Sistrunk, Jr., Esq. 23. Philip Sklover, Esq. 24. Southern Company, Inc. 25. Southern Company Energy Solutions, Inc. 26. Southern Company Services, Inc. 27. Michael Terry, Esq. 28. Joshua Thorpe, Esq. 29. Carolyn Wheeler, Esq. 30. Carolyn Wilson Pursuant to Fed.R.App.P. 26.1, the EEOC, as a government agency, need not file a corporate disclosure statement. _________________________ Dori K. Bernstein TABLE OF CONTENTS CERTIFICATE OF INTERESTED PERSONS i TABLE OF CONTENTS iii TABLE OF CITATIONS iv STATEMENT OF INTEREST 1 STATEMENT OF THE ISSUES 1 STATEMENT OF FACTS 2 SUMMARY OF ARGUMENT 15 ARGUMENT 17 PLAINTIFFS SATISFY ALL REQUIREMENTS OF RULE 23(b)(2) TO LITIGATE CLASS CLAIMS OF SYSTEMIC RACE DISCRIMINATION IN PAY AND PROMOTIONS AT SOUTHERN COMPANY, AND HAVE PRODUCED SUFFICIENT EVIDENCE TO TRY THEIR PATTERN-OR-PRACTICE AND DISPARATE IMPACT CLAIMS TO A JURY 17 CONCLUSION 32 CERTIFICATE OF COMPLIANCE 34 CERTIFICATE OF SERVICE TABLE OF CITATIONS CASES Anderson v. Liberty Lobby, Inc., 477 U.S. 242 (1986)31 Bazemore v. Friday, 478 U.S. 385 (1986)24, 25, 26, 30, 31 Cooper v. Federal Reserve Bank of Richmond, 467 U.S. 867 (1984)17 Cooper v. Southern Co., 205 F.R.D. 596 (N.D. Ga. 2001)passim Cooper v. Southern Co., 2003 WL 1889268 (N.D. Ga. 2003)14 Cooper v. Southern Co., 2003 WL 1889273 (N.D. Ga. 2003)14 Cooper v. Southern Co., 2003 WL 1889274 (N.D. Ga. 2003)14 Cooper v. Southern Co., 2003 WL 1889275 (N.D. Ga. 2003)14, 15, 25 Cooper v. Southern Co., 2003 WL 1889282 (N.D. Ga. 2003)14 Cooper v. Southern Co., 2003 WL 1889283 (N.D. Ga. 2003)14 Cooper v. Southern Co., 2003 WL 1889284 (N.D. Ga. 2003)14 Cox v. American Cast Iron Pipe Co., 784 F.2d 1546 (11th Cir. 1986)29 Eisen v. Carlisle & Jacquelin, 417 U.S. 156 (1974)22 General Telephone v. Falcon, 457 U.S. 147 (1982)18, 19, 22 Hazelwood School Dist. v. United States, 433 U.S. 299 (1977)10, 18, 32 Holmes v. Continental Can Co., 206 F.2d 1144 (11th Cir. 1983)17, 21 International Brotherhood of Teamsters v. United States, 431 U.S. 324 (1977)29, 30 Kilgo v. Bowman Transp. Co., 789 F.2d 859 (11th Cir. 1986)29 Nelson v. United Steel Corp., 709 F.2d 675 (11th Cir. 1983)19, 23 Pettway v. American Cast Iron Pipe Co., 494 F.2d 211 (5th Cir. 1974)21 Reeves v. Sanderson Plumbing Prods., 530 U.S. 133 (2000)32 Robinson v. Metro-North Commuter R.R. Co., 267 F.3d 147 (2d Cir. 2001), cert. denied, 535 U.S. 951 (2003)22 VisaCheck/Mastermoney Litigation, 280 F.3d 124 (2d Cir. 2001)26 Watson v. Fort Worth Bank and Trust, 487 U.S. 978 (1988)18, 32 STATUTES Title VII of the Civil Rights Act of 1964, 42 U.S.C. §§ 2000e et seq.1 42 U.S.C. § 19817 RULES AND REGULATIONS Fed.R.App.P. 26.1ii Fed.R.App.P. 29(a)1 Fed.R.App.P. 29(d)34 Fed.R.Civ.P. 231, 19, 21, 22, 23 Fed.R.Civ.P. 23(a)13, 16, 18, 23, 27 Fed.R.Civ.P. 23(b)21 Fed.R.Civ.P. 23(b)(2)iii, 1, 13, 16, 17, 21, 22, 32 Fed.R.Civ.P. 32(a)(7)(B)34 11th Cir. R. 26.1-1i STATEMENT OF INTEREST The EEOC administers and enforces Title VII of the Civil Rights Act of 1964, 42 USC §§ 2000e et seq. This appeal involves the proper application of Fed.R.Civ.P. 23 to Title VII systemic discrimination claims, and the quantum of evidence required to proceed to trial on pattern-or-practice and disparate impact claims challenging a system of subjective, discretionary decisionmaking. Because the ability of private litigants to challenge systemic discrimination on a class-wide basis is vital to the effective enforcement of Title VII, and because the legal standards and evidentiary burdens for proving systemic discrimination are the same in government enforcement actions as in private class litigation, the EEOC offers its views to the Court. See Fed.R.App.P. 29(a). STATEMENT OF THE ISSUES Whether the district court abused its discretion in denying class certification under Fed.R.Civ.P. 23(b)(2) to litigate claims of systemic racial discrimination caused by defendants’ common pay and promotion policies. Whether the district court erred in granting summary judgment against plaintiffs’ claims that defendants’ common policy of affording managers unmonitored discretion to use subjective criteria in making compensation and promotion decisions fostered a pattern or practice of disparate treatment, or caused an adverse disparate impact, against African-American employees. STATEMENT OF FACTS Defendants-appellees are Southern Company and subsidiaries Georgia Power Company (GPC), Southern Company Energy Solutions (SCES), and Southern Company Services (SCS).1 See Cooper v. Southern Co., 205 FRD 596 (N.D. Ga. 2001).2 In 1995, SoCo consolidated the human resources (HR) functions - including staffing, compensation, and EEO compliance - of each of its subsidiaries into a centralized operation at SCS “to implement common compensation and promotion policies and practices between the companies.” DE141, Ex.B, ¶ 3; 205 FRD at 600 (finding GPC, SCES, and SCS “utilize common compensation and promotions policies promulgated and administered by ... SCS”). Under SoCo’s personnel policies, non-union jobs3 are normally filled internally through posting on SoCo’s “computer intranet system known as ‘JobNet,’ which is accessible to virtually all [SoCo] employees.” Id. at 600. Mark Wolf, SoCo’s Staffing Director, oversees the SCS department that provides staffing “infrastructure” for all SoCo subsidiaries. DE117, Ex.9 at 5-7. While SCS has developed detailed instructions and procedures for job posting, applicant screening, interviews, and candidate selection, see DE119, Ex.36-42, Wolf testified that SoCo’s policy is to “encourage” managers to follow these “guidelines” for filling vacancies, but “there is no requirement” that they do so. DE117, Ex.9 at 67. Wolf was unaware of any effort by SoCo or its subsidiaries to monitor or control the internal hiring and promotion process to ensure managers exercise their discretion in a non-discriminatory manner and use job-related criteria to select employees for promotion. Id. at 69-70. SoCo’s internal hiring policies afford each hiring manager discretion to decide: whether to consult HR in filling a vacancy;4 whether to post an opening on JobNet;5 what job qualifications to list in the posting, DE117, Ex.9 at 59; what criteria to apply in screening applicants for interviews, id. at 52-53; whether to use a selection committee, id. at 84; whether to follow “structured interview” guidelines, id. at 67; what to ask interviewees, id. at 75, 82; whether to ask the same questions of each candidate for the same position, id. at 77; and which candidate to hire. “Even in cases where a selection committee is used, the manager is not required to accept the committee’s recommendation.” 205 FRD at 60-0. SoCo further entrusts individual managers with discretion to promote employees without posting a vacancy by: awarding “promotions within a progressive job family”; identifying an employee with “demonstrated leadership potential” and providing him “with opportunities to obtain experience in other areas” through a “developmental move” to a new position; and “re-evaluat[ing] jobs within their departments if they determine that the actual skills are above or below those anticipated by the job criteria.” See 205 FRD at 600-01; DE119, Ex.36 at E212-13. SoCo employees in non-CBA jobs are compensated pursuant to “a common job, salary or pay grade system ... formulated by ... SCS for non-exempt employees (grades NE1 through NE9) and exempt employees (grades E1 to E15).” 205 FRD at 600.6 SoCo’s pay grade system features broad and overlapping salary ranges. In 1998, for example, “grade E5 ranged from $42,396 to $67,837. Grade E3 was approximately $24,020 to $54,432.” Id. Under SoCo’s pay grade system, each job may be slotted at more than one grade, authorizing compensation within multiple salary ranges. Id. at 612 n.29. A job posted in Grades 3-5 in 1998 could thus carry an annual base salary of anywhere between $24,020 and $67,837. Steven Wilkinson, SoCo’s Compensation Manager, heads the SCS department that develops and administers compensation policies for SoCo subsidiaries, including GPC and SCES. See DE117, Ex.8 at 18-21. According to Wilkinson, SoCo delegates to individual managers discretion to assign job grades and set base salaries, id. at 42-45, and “there are no set standards, there’s no formula as such to follow,” and no controls in place “to ensure a manager does not discriminate” by assigning lower job grades to black employees. Id. at 195-96. Nor are there standards or controls to ensure managers assign the same pay grade to jobs that entail the same duties. Id. at 201. Each manager also has discretion to decide the percentage increase over base salary an employee will receive upon promotion to a different job. Id. at 98, 103. In addition to base salary, employees in nearly all non-CBA jobs are eligible for annual merit pay increases and lump sum incentive bonuses under SoCo’s common Performance Pay Plan (PPP). See 205 FRD at 601. While SoCo provides “general considerations” and “helpful hints” for awarding compensation under these common plans, see DE119, Ex.73 at F664, F669 & Ex.74, “managers and supervisors within individual departments have discretion to determine how to allocate individual merit increases” and incentive bonuses among subordinate employees. 205 FRD at 601. Wilkinson testified that SoCo affords each manager “broad ... discretion to establish a methodology” to distribute bonuses and merit increases by ranking the comparative contributions of employees based on performance. DE117, Ex.8 at 147-48, 161, 202-03. Under SoCo’s policy, Wilkinson agreed, a manager “could establish different criteria to evaluate” five different employees holding the same job under his supervision. Id. at 151. According to Wilkinson, “there’s no review” by HR “of the performance evaluations or the documentation” to determine whether managers are using job-related factors to evaluate employees. Id. at 153. Aside from requiring a manager to stay within his departmental budget, Wilkinson testified, “there are no absolute limits” on each manager’s discretion to determine the size of each employee’s annual pay increase. Id. at 201-02 Seven African-American employees sued SoCo in July 2000, asserting individual and class claims of race discrimination in violation of Title VII and 42 U.S.C. §1981. See 205 FRD at 597-98. The named plaintiffs variously allege that they were denied positions, promotional opportunities, or compensation based on race pursuant to SoCo’s common pay and promotion policies. See id. at 602-05; DE141, Ex.E. The complaint sought the full range of declaratory, injunctive, equitable, and monetary relief provided by law, and demanded a jury trial. DE62 at 59. Plaintiffs sought class certification to litigate claims of systemic race-based disparate treatment and adverse impact caused by SoCo’s policies for awarding compensation in and promotion to non-CBA jobs. The proposed class included approximately 2,400 African-Americans employed by SoCo at any time from July 1998 to the present. 205 FRD at 598-99. To support class certification, plaintiffs submitted documentary and testimonial evidence of SoCo’s policies and practices on internal hiring, promotions, and compensation, see DE117 & 119, and SoCo-generated documents indicating racial disparities in these areas during the relevant time period. A 1998 audit of SoCo’s internal hiring practices disclosed that “interview pools had a majority white male population; records indicate that when diverse applicant pools are provided, hiring managers are interviewing primarily white male candidates.” DE119, Ex.57 at L1314; see also Ex.63 at J5841 (internal SoCo spreadsheet analysis showing that of 1342 positions filled, 13.6% of selections were from a “diverse pool” of candidates). The audit further showed that 98% of “promotions resulting from job reevaluations/developmental assignments” went to white males, 2% to females, and 0% were awarded to minority employees. Ex.57 at L1314. “Diversity Report Cards” for GPC (dated September 1999) and SoCo (dated August 1999), reported the following racial disparities in compensation and workforce composition: a “glass ceiling effect” on African-American representation at all levels of SoCo management;7 a “glass wall effect” in “minority leadership by function,” Ex.86 at L373; a substantial racial disparity in the average pay of exempt employees at SoCo, id. at L364, L374; substantial racial disparities in the average salaries of employees in both exempt and non-exempt jobs at GPC and SCS, id. at L341-42, L352-53; and a substantial racial disparity in the distribution of annual PPP incentive bonuses. Id. ay L364, L371. Dr. Janice Madden, a labor economist, analyzed workforce data supplied by SoCo and prepared an expert report that confirms the statistical significance of the racial disparities reflected in SoCo’s internal reports. See DE119, Ex.1; DE141, Ex.B. Using a “multiple pools model” and multi-regression analyses to compare the promotion rates of employees in the same salary grade, subsidiary, and year, Dr. Madden found “statistically significant shortfalls in the numbers of promotions awarded to African-American employees, when compared to Caucasian employees ... for 1996 through 1999.” Id. at 5. Dr. Madden further analyzed the data, controlling for education level, experience (measured by both tenure at SoCo and years since completing education), and job function8 at hiring (or in 1995 if hired before 1995). See DE119, Ex.1 at 12-15; DE141, Ex.B at 2-6. The results of these regression analyses “show increases in both the coefficient and the statistical significance of estimated racial disparities in promotions when credentials (experience, education and job function) were added to the model,” leading Dr. Madden to “conclude that African-Americans are at least as qualified [or “more qualified”] for promotion as Caucasians in the same salary grade, subsidiary, and year.” Id. at 5. “The racial differential in promotions for the entire period,” Dr. Madden calculated, “is 7.07 standard deviations beyond no difference (or zero). The probability that this racial difference could have occurred by chance is less than 26 in ten trillion.”9 Id. Dr. Madden’s multi-regression analyses of SoCo’s data further revealed that “African-American employees have statistically significant lower compensation than Caucasians during the period 1995-1999.” Id. at 9; see DE119, Ex.1 at 15-20; DE141, Ex.B at 6-9. “African-Americans receive 2.56% to 3.52% less than Caucasian employees in the same salary grade and subsidiary with the same time since completing schooling, tenure at Southern, level of education and job function.” Id. at 9. Dr. Madden further calculated that once “the effects of the racially differentiated salary grades are removed, the salary disadvantage of African-Americans grows to a range of 13.17% to 14.33%.” Id. In each year, Dr. Madden found, “the statistical significance of the race differences in compensation, after considering the effects of any racial differences in subsidiary, education, work experience, job function, and salary grade ... strongly rejects the possibility that they could have been generated by chance.” Id. at 7. Dr. Madden determined that “the overall pattern of lower compensation for African-Americans ... is at levels well beyond those required to dismiss random variation,” and that the results of her analyses “are consistent with racial discrimination.” Id. at 8. Plaintiffs offered evidence that the pattern of racial disparities in promotions and compensation at SoCo is nothing new. In a compliance review of GPC’s operations covering the period November 1991 to July 1994, the U.S. Department of Labor’s Office of Federal Contract Compliance Programs (OFCCP) found numerous EEO violations in awarding promotions and compensation to minority employees of SoCo’s largest subsidiary, including: salary disparities between minorities and non-minorities in similar positions; deviations from job posting procedures to the disadvantage of minority employees; “numerous examples ... of non-minorities ... selected for promotion with less or equal qualification than the minority ... applicants”; failure to ensure minorities “receive career progression moves comparable to non-minorities”; and underrepresentation of minorities in key areas, including management. DE119, Ex.14 at 3-7.10 Plaintiffs also submitted the following evidence of racial animus at SoCo facilities: photographs and documentary proof, generated in 2000, of at least 13 nooses displayed in eight GPC plants, some over a period of years, DE119, Ex.3-5; a compendium of racially offensive “jokes” posted on bulletin boards in company facilities, Ex.6; a “KKK cartoon” placed in employee mailboxes, Ex.7; a photograph of a hat with a racial epithet worn in the workplace, Ex.8; and testimonial accounts by numerous black employees of racially offensive slurs, “jokes,” epithets, and harassment at work. Ex.11. Finally, 111 prospective class members signed affidavits recounting instances of alleged racial discrimination during their employment at SoCo. See DE120. The district court denied class certification on October 11, 2001. Despite finding that SoCo subsidiaries “utilize common compensation and promotions policies promulgated and administered by ... SCS,” 205 FRD at 600, the court concluded that plaintiffs’ systemic claims that these common policies have fostered a pattern or practice of racially disparate treatment or adverse impact do not raise questions of law or fact common to the class. Id. at 626. The court further found that none of the named plaintiffs, each of whom claims a denial of pay or promotional opportunities pursuant to these policies, has “claims typical of the class as a whole.” Id. at 610. There was no dispute that the proposed class met Rule 23(a)’s numerosity requirement, id. at 608, and the court found plaintiffs’ “counsel are capable and experienced lawyers who would more than adequately represent the class.” Id. at 627. The court also decided that “this case cannot be certified under Rule 23(b)(2),” because plaintiffs’ requests for compensatory and punitive damages would entail “highly individualized fact findings” that were not “merely incidental to an injunction or declaration in the class’s favor.” Id. at 627. The court never addressed plaintiffs’ alternative request to certify a class under Rule 23(b)(2) solely to determine SoCo’s liability for a pattern or practice of disparate treatment or adverse impact against African-Americans in promotions and compensation, and to award appropriate class-wide declaratory, injunctive, and equitable relief. On March 31, 2003, the court issued seven opinions granting summary judgment against the claims advanced by each named plaintiff. See Cooper v. Southern Company, 2003 WL 1889268 (N.D. Ga., Mar. 31, 2003) (Edwards); 2003 WL 1889273 (S.J. Harris); 2003 WL 1889274 (P. Harris); 2003 WL 1889275 (McCullers); 2003 WL 1889282 (Green); 2003 WL 1889283 (Wilson); 2003 WL 1957484 (Cooper).11 In resolving the individual claims, the court included virtually identical text in each opinion rejecting plaintiffs’ claims that SoCo’s policies and practices for awarding promotions to and compensation for non-CBA positions have fostered a pattern or practice of race-based disparate treatment by individual managers, see Edwards at *13-*14; S.J. Harris at *10-*11; P. Harris at *7; McCullers at *6-*7; Green at *10-*11; Wilson at *12; Cooper at *14, or resulted in an adverse disparate impact against black employees. See Edwards at *14-15; McCullers at *8-*9; Green at *11-*12; Wilson at *12-*13. The court included two paragraphs in each opinion that referenced only plaintiffs’ statistical evidence and SoCo’s “internal personnel documents,” and rejected plaintiffs’ pattern-or-practice claims of racially disparate treatment under SoCo’s promotion and compensation policies. See, e.g., McCullers at *6-*7. Without mentioning the additional evidence plaintiffs submitted to support their pattern-or-practice claims (i.e., proof of nooses and racial hostility in various SoCo facilities; OFCCP findings of racial disparities in promotions and compensation at GPC during the early 1990s; anecdotal evidence of 111 SoCo employees recounting specific incidents of racial discrimination and harassment), the court summarily concluded that each plaintiff “failed to come forward with evidence sufficient to prove that intentional discrimination was [SoCo’s] ‘standard operating procedure,’” and granted summary judgment against plaintiffs’ pattern-or-practice claim. Id. (internal citation omitted). The court further ruled that each plaintiff “failed to establish that manager discretion in compensation and promotion decisions caused a statistically significant disparity” adverse to African-Americans, and granted summary judgment against plaintiffs’ disparate impact claims. Id. at *8. SUMMARY OF ARGUMENT The district court abused its discretion by effectively deciding the merits of plaintiffs’ claims of systemic race discrimination in the guise of denying class certification. The court further erred in holding plaintiffs’ evidence insufficient for a jury to find that SoCo’s common policies of committing pay and promotion decisions to the unmonitored discretion of individual managers have fostered a pattern or practice of disparate treatment, or caused an adverse disparate impact, against African-Americans. Viewed under the appropriate legal standards established in Supreme Court precedent, plaintiffs plainly satisfy the prerequisites of Rule 23(a) and have presented evidence sufficient to warrant a trial on their systemic discrimination claims. Plaintiffs’ request to certify a class solely to determine SoCo’s liability for systemic discrimination and award appropriate declaratory, injunctive, and equitable relief - an alternative the district court did not even address - is precisely the type of action contemplated for class treatment under Rule 23(b)(2). This Court should therefore reverse the summary judgment for SoCo and the denial of class certification, and direct the court on remand to certify a class under Rule 23(b)(2) to litigate at trial the questions of whether SoCo’s pay and promotion policies foster a pattern-or-practice of disparate treatment, or an adverse disparate impact, against African-Americans, and, if appropriate, to award declaratory, injunctive, and equitable relief. ARGUMENT PLAINTIFFS SATISFY ALL REQUIREMENTS OF RULE 23(b)(2) TO LITIGATE CLASS CLAIMS OF SYSTEMIC RACE DISCRIMINATION IN PAY AND PROMOTIONS AT SOUTHERN COMPANY, AND HAVE PRODUCED SUFFICIENT EVIDENCE TO TRY THEIR PATTERN-OR-PRACTICE AND DISPARATE IMPACT CLAIMS TO A JURY. The district court abused its discretion in refusing to certify a class pursuant to Rule 23(b)(2), and erred in granting summary judgment against plaintiffs’ claims of systemic race discrimination. These rulings betray this Court’s longstanding “conviction that Title VII and the class action rule should be construed so as to further the strong public policy of eradicating all vestiges of racial discrimination in employment.” Holmes v. Continental Can Co., 206 F.2d 1144, 1152 (11th Cir. 1983). The Supreme Court in Cooper v. Federal Reserve Bank of Richmond, 467 U.S. 867, 876 n.9 (1984), explained that [t]he crucial difference between an individual’s claim of discrimination and a class action alleging a general pattern or practice of discrimination is manifest. The inquiry regarding an individual’s claim is the reason for a particular employment decision, while at the liability stage of a pattern-or-practice trial the focus often will not be on individual hiring decisions, but on a pattern of discriminatory decision-making. Id. at 867. A policy or practice of committing employment decisions to the subjective discretion of individual managers or supervisors may be challenged under Title VII as fostering a pattern or practice of disparate treatment, Hazelwood School Dist. v. United States, 433 U.S. 299, 301-02 (1977) (challenging pattern or practice of intentional racial discrimination in “relatively unstructured procedures for hiring teachers” that delegated to “each school principal ... virtually unlimited discretion ... to hire the ‘most competent’ person available”), or resulting in an unjustified disparate impact. See Watson v. Fort Worth Bank and Trust, 487 U.S. 978, 992 (1988) (“subjective or discretionary employment practices may be analyzed under the disparate impact approach”). “The commonality and typicality requirements of Rule 23(a) tend to merge” because [b]oth serve as guideposts for determining whether under the particular circumstances maintenance of a class action is economical and whether the named plaintiff’s claim and the class claims are so interrelated that the interests of the class members will be fairly and adequately protected in their absence. These requirements therefore also tend to merge with the adequacy-of-representation requirement, although the latter requirement also raises concerns about the competency of class counsel and conflicts of interest. General Telephone v. Falcon, 457 U.S. 147, 157 n.13 (1982). Here, these three requirements are plainly met. Plaintiffs’ pattern-or-practice and disparate impact claims raise questions of law and fact common to every African-American employee whose compensation and opportunities for advancement may be limited under SoCo’s common system of delegating to individual managers virtually complete discretion to make pay and promotion decisions for employees in non-CBA jobs. The resolution of plaintiffs’ claims of systemic discrimination rests on the same evidence and identical legal precedent. See Nelson v. United Steel Corp., 709 F.2d 675, 679 (11th Cir. 1983) (Falcon requires plaintiff to show “proof or legal arguments involved in her claims and class claims would be overlapping so that ... class suit would advance Rule 23's goal of judicial economy”). Ironically, the district court in effect acknowledged the commonality of the legal and factual questions raised by plaintiffs’ pattern-or-practice and disparate impact claims by including virtually identical passages in each summary judgment opinion to resolve them. Furthermore, the individual claims of each named plaintiff that s/he was denied a promotion or compensation because of race pursuant to SoCo’s common policies are typical of the claims of absent class members who were discriminatorily denied promotions or compensation under the same policies. In finding that individual plaintiffs’ claims were typical only of portions of the class (i.e., employees holding positions at his/her grade level, or in the same non-exempt or exempt status), and that none of the named plaintiffs raised claims typical of management-level employees, 205 FRD at 610, the court disregarded undisputed evidence (and its own findings, id. at 600) that the same policies and practices under which each plaintiff claims s/he was denied promotion to a non-CBA job, or received a lower salary or bonus in such a job, apply to all promotion and compensation decisions for non-CBA jobs at every grade level, whether exempt or non-exempt. Finally, every class member shares with named plaintiffs the same interest in establishing whether SoCo’s common pay and promotion policies foster systemic racial discrimination and, if so, in securing declaratory and injunctive relief to eliminate the discriminatory practices and ensure equal opportunity in the areas of compensation and advancement. Because the named plaintiffs have every incentive to litigate vigorously the systemic discrimination claims, and have retained able and experienced counsel, they have demonstrated their adequacy as class representatives. Plaintiffs thus satisfy all four prerequisites of Rule 23(a). Plaintiffs’ request to certify a class solely to determine SoCo’s liability for systemic discrimination and award appropriate declaratory, injunctive, and equitable relief - an alternative the district court did not even address - is precisely the type of claim contemplated for class treatment under Rule 23(b)(2).12 “The Advisory Committee Note cited cases enjoining violations of civil rights as ‘illustrative’ of the types of cases properly cognizable under subsection (b)(2), ... as did the reporter of the rules revision committee in his personal commentary on the 1966 amendments” to Rule 23. Holmes, 706 F.2d at 1155. The sole basis for the district court’s denial of certification under Rule 23(b) was the predominance of individualized determinations required to resolve individual class members’ claims for monetary damages. Because plaintiffs’ alternative suggestion to certify only the claims for class-wide declaratory, injunctive, and equitable relief eliminates entirely this concern, this Court should reverse and remand with directions to certify plaintiffs’ claims of systemic discrimination to determine liability and appropriate declaratory, injunctive, and equitable relief. See Robinson v. Metro-North Commuter R.R. Co., 267 F.3d 147, 172 (2d Cir. 2001) (vacating dismissal of Title VII systemic racial discrimination suit challenging company-wide policy of delegating to department supervisors discretionary authority to make discipline and promotion decisions, and directing district court to certify plaintiffs’ disparate impact claim and, at a minimum, the liability stage of plaintiffs’ pattern-or-practice claim under Rule 23(b)(2)), cert. denied, 535 U.S. 951 (2002). The district court abused its discretion by effectively deciding the merits of plaintiffs’ pattern-or-practice and disparate impact claims - without a hearing, let alone a jury trial - in the guise of denying class certification. While “the class determination generally involves considerations that are enmeshed in the factual and legal issues comprising the plaintiff’s cause of action,” Falcon, 457 U.S. at 160 (internal citations omitted), the Supreme Court in Eisen v. Carlisle & Jacquelin, 417 U.S. 156, 176 (1974), found “nothing in either the language or history of Rule 23 that gives a court any authority to conduct a preliminary inquiry into the merits of a suit in order to determine whether it may be maintained as a class action.” Although a plaintiff seeking certification is “obligated to show, in at least a preliminary fashion, the required commonality between her claims and those of the putative class,” her “‘burden of proof to demonstrate the existence of this common question entails ... less than a prima facie showing of liability.’” Nelson, 709 F.2d at 680 (citations omitted). The court in this case plainly violated these principles. In addressing Rule 23(a)'s commonality requirement, the court discussed at length “the relative strength and quality” of evidence supporting plaintiffs’ systemic discrimination claims, 205 FRD at 610, and its opinion is replete with judgments on the weight and probative value of plaintiffs’ evidence that are within the sole province of the jury. The opinion denying class certification thus demonstrates a clear abuse of discretion in applying Rule 23, and exposes the court’s error in granting summary judgment against plaintiffs’ systemic claims. With respect to plaintiffs’ statistical evidence, for instance, the court concluded that while the “methodologies” of plaintiffs’ expert “may be valid, the analysis has some limitations which undermine its usefulness in measuring whether [SoCo’s] employment practices are racially neutral.” Id. at 613. At the same time, the court rejected some arguments and analysis advanced by SoCo’s expert, Dr. Joan Haworth, to counter plaintiffs’ statistical proof. For instance, the court rejected SoCo’s argument that Dr. Madden improperly lumped together progressive and competitive promotions “because both types of promotions involve evaluative choices by management.” Id. at 614. Further, while Dr. Haworth’s analysis of salary and promotion data included many more variables and concluded that “small gaps” in the promotion rates and compensation of black and white employees “were not statistically meaningful,” id., the court found “some truth” in Dr. Madden’s response “that the lack of statistical significance was due to the fact that Haworth had ‘carved up’ the employees into pools so small that it would be unlikely that the results would be statistically significant.” Id. In addition, the court found Dr. Haworth’s omission of “the so-called ‘developmental moves’” from her analysis of promotions “is a factor limiting the usefulness of [her] report to the extent it is intended to prove that [SoCo’s] treatment of African-Americans is the same.” Id. Having reviewed both experts’ reports, the court was “left with the impression that there could be some salary gap and promotions gap between African-American and white employees,” but found it “impossible to determine what the gaps are, whether they are statistically significant, or whether factors other than race are involved.” Id. at 614-15. The Supreme Court’s decision in Bazemore v. Friday, 478 U.S. 385, 400 (1986), establishes that the district court’s “view of the evidentiary value of [plaintiffs’] regression analysis was plainly incorrect.” The court’s primary criticism of plaintiffs’ statistical evidence was a perceived “failure to adequately measure the promotion and compensation experiences of similarly situated employees.” 205 FRD at 613. Although Dr. Madden’s multi-regression analyses controlled for year, subsidiary, job grade, level of education, years since completing school (as a proxy for time in the labor force), tenure at SoCo, and job function when hired (or in 1995), the court faulted her failure to control for “specific job categories and job progressions,” while conceding “it would be very difficult to factor those considerations into a statistical analysis.” Id. at 614. Having concluded that Dr. Madden did not adequately “account for differences in the type or level of the employees’ applied skills,” see McCullers, 2003 WL 1889275 at *6-*8, the court refused to consider plaintiffs’ statistics “evidence of a pattern and practice of discrimination.” 205 FRD at 615. As in this case, plaintiffs in Bazemore claimed systemic race discrimination in compensation and “relied heavily on multiple regression analyses designed to demonstrate that blacks were paid less than similarly situated whites.” 478 U.S. at 398. The Supreme Court held a lower court “erred in stating that [plaintiffs’] regression analyses were ‘unacceptable as evidence of discrimination,’ because they did not include ‘all measurable variables thought to have an effect on salary level.’ ... While the omission of variables from a regression analysis may render the analysis less probative than it otherwise might be,” the Court explained, “it can hardly be said, absent some other infirmity, that an analysis which accounts for the major factors ‘must be considered unacceptable as evidence of discrimination.’” Id. at 400. “Normally, failure to include variables will affect the analysis’ probativeness, not its admissibility,” unless the regression is “so incomplete as to be inadmissible,” an argument SoCo never made. Id. & note 10. The Court further emphasized, Importantly, it is clear that a regression analysis that includes less than ‘all measurable variables’ may serve to prove a plaintiff’s case. A plaintiff in a Title VII suit need not prove discrimination with scientific certainty; rather, his or her burden is to prove discrimination by a preponderance of the evidence. ... Whether, in fact, such a regression analysis does carry the plaintiffs’ ultimate burden will depend in a given case on the factual context of each case in light of all the evidence presented by both the plaintiff and the defendant. However, as long as the [factfinder] may fairly conclude, in light of all the evidence, that it is more likely than not that impermissible discrimination exists, the plaintiff is entitled to prevail. Id. at 400-01. The credibility of the parties’ experts, and the weight and probative value of their statistical analyses, is for the jury to decide, and plaintiffs’ evidence exceeds the legal standards for submission to a jury. See VisaCheck/Mastermoney Litigation, 280 F.3d 124, 135 (2d Cir. 2001) (“a district court may not weigh conflicting expert evidence or engage in ‘statistical dueling’ of experts” in resolving Rule 23 motion). The district court similarly usurped the jury’s function in evaluating plaintiffs’ additional evidence of systemic discrimination at SoCo. In finding plaintiffs’ evidence of racial animus in SoCo facilities insufficient to establish the commonality required by Rule 23(a), the court improperly discounted the probative value of this evidence to permit an inference of discriminatory motive and to demonstrate managerial indifference to the rights of black employees. The court reviewed documentary, photographic, and testimonial evidence that “at numerous times at various locations knotted rope was present which resembled a noose,” and acknowledged that “[t]hese ‘nooses’ did offend and intimidate some employees, who attributed a racial significance to the nooses.” 205 FRD at 615. Citing the “sworn statements of [SoCo] employees [who] ... disavowed any knowledge of racial meaning or racial animus,” id., and “voluminous” memoranda prepared by SoCo during investigations conducted shortly after this lawsuit was filed, id. at 615-18, however, the court found it “impossible to say whether any of the nooses were made or displayed out of a desire to offend or intimidate African-American employees.” Id. at 618. The court also reviewed sworn accounts of “various racial slurs, epithets, jokes, and harassment” in nearly all of the 111 affidavits submitted by prospective class members, and found “the conduct referenced therein are [sic] unacceptable and ... demeaning to African-Americans.” Id. at 619. While such evidence of racial hostility is relevant “to determine the existence of a pattern and practice of discrimination,” id. at note 39, the court improperly discounted its probative value “given the size of [SoCo’s] workforce (over 12,000 employees) and the large number of work locations,” and therefore concluded that “the noose evidence and evidence of racial slurs, jokes and epithets ... are inadequate to establish the existence of a pattern and practice of discrimination.” Id. (footnote omitted). The court used a similarly inappropriate rationale to reject the anecdotal evidence of discrimination submitted by 111 prospective class members “who recite (in addition to the noose testimony and the racial jokes and slur testimony ...) their [unsuccessful] efforts to obtain promotions or greater compensation, and their belief that [SoCo has] discriminated against them” based on race. Id. at 620. The court remarked that “resolving the validity” of these allegations “would be a fact-intensive process,” id. at 624, and observed that “111 affiants out of a group of 2400 individuals is too low a proportion to prove a pattern and practice of discrimination.” Id. at 624. The district court’s treatment of the non-statistical evidence reflects a fundamental misunderstanding of the purpose such evidence serves in proving systemic discrimination. In deciding an employer’s liability for systemic discrimination, “the focus is on the broad pattern or practice at issue, not the merits of individual claims.” Cox v. American Cast Iron Pipe Co., 784 F.2d 1546, 1556 (11th Cir. 1986). The function of offering the testimony of individual employees “about their personal experiences with the company” is to bring the “cold numbers” provided by a statistical analysis “convincingly to life,” not to establish an employer’s liability for individual claims of discrimination. International Brotherhood of Teamsters v. United States, 431 U.S. 324, 339 (1977); Kilgo v. Bowman Tansp. Co., 789 F.2d 859, 874 (11th Cir. 1986) Plaintiffs’ anecdotal evidence far exceeds the quantity submitted in Teamsters, where the government introduced testimony “recount[ing] over 40 specific instances of discrimination.” Id. at 338. The credibility of the testimony offered by these employees, and the weight to be accorded the other evidence of racial hostility at SoCo, are for the jury to resolve at trial, not for the district court to determine at summary judgment, much less on a motion for class certification. Similarly, the court’s failure to grasp the probative significance of uncontested evidence showing a “significantly smaller representation of African-Americans in lower, middle and upper management than exists at non-management levels” at SoCo, 205 FRD at 619, reflects a profound misconception of the law governing statistical proof of systemic discrimination. “[T]his data,” the court ruled, “does not in itself show an intent to discriminate.” Id. The Supreme Court has held otherwise. “Statistics showing racial or ethnic imbalance” in an employer’s workforce “are probative in a case such as this one,” the Court explained, “because such imbalance is often a telltale sign of purposeful discrimination; absent explanation, it is ordinarily to be expected that nondiscriminatory [promotion] practices will in time result in [managerial or supervisory ranks] more or less representative of the racial and ethnic composition” of the workforce from which this staff is promoted. See Teamsters, 431 U.S. at 340 n.20 (absent discrimination, hiring practices should be expected to produce a workforce that mirrors the general population). The court further erred in discounting the probative value of SoCo’s documented lack of progress in rectifying racial disparities in pay and promotions, identified by OFCCP, to show “the indifference of [SoCo’s] senior management to race discrimination.” 205 FRD at 625. Evidence that “changes implemented” under the OFCCP conciliation agreement “did not ... [close] the salary and promotion gap between Caucasians and African-Americans,” and that SoCo’s “own self-study (‘The Diversity Report Card’) was critical of the extent of improvement,” the court decided, did not show “the indifference of senior management is at such a level” to support plaintiffs’ systemic claims. Id. Once again, the district court’s view conflicts with Supreme Court precedent. The Court in Bazemore considered proof of an employer’s history of “salary discrimination, and ... ineffectual attempts to eradicate it,” part of “an impressive array of evidence to support [plaintiffs’] contention that [defendant] engaged in a pattern or practice of discrimination with respect to salaries.” 478 U.S. at 401. Finally, the district court criticized plaintiffs’ reliance on Staffing Director Wolf’s testimony to show that SoCo’s policy permits individual managers, in their discretion, to ignore “numerous, detailed written policies and procedures pertaining to filling job vacancies, making promotions, and determining compensation levels.” 205 FRD at 619-20. In the court’s view, “Wolf’s testimony is more fairly characterized as asserting lack of knowledge,” and a “fairer interpretation of the evidence would be that [SoCo has] numerous written procedures and policies ..., that there is some variation in the manner in which various managers use these procedures, and that the policies and procedures were never intended (nor could they be) to dictate particular outcomes in individual cases.” Id. at 620. It was not the court’s role, however, to decide on the “fairest” interpretation of Wolf’s (or any other witness’s) testimony. Such determinations are a quintessential jury function. “Credibility determinations, the weighing of the evidence, and the drawing of legitimate inferences from the facts are jury functions, not those of a judge.” Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 255 (1986). While a jury might regard the evidence as the district court did, this is certainly not the only permissible view of the record, and certainly not the view most favorable to plaintiffs, as required to sustain summary judgment for SoCo. See Reeves v. Sanderson Plumbing Prods., 530 U.S. 133, 150-51 (2000). In holding plaintiffs’ evidence of systemic discrimination insufficient to go to trial, the district court ignored Supreme Court precedent and imposed a legally unattainable standard to challenge, pursuant to Hazelwood and Watson, systemic discrimination arising from a policy of affording individual managers unmonitored discretion to engage in subjective decisionmaking. CONCLUSION The EEOC urges this Court to reverse the denial of class certification and entry of summary judgment, and direct the court on remand to certify a class under Rule 23(b)(2) to litigate at trial the questions of whether SoCo’s pay and promotion policies foster a pattern-or-practice of disparate treatment, or an adverse disparate impact, against African-Americans, and, if appropriate, to award declaratory, injunctive, and equitable relief. Respectfully submitted, ____________________________ DORI K. BERNSTEIN, Attorney NICHOLAS M. INZEO, Acting Deputy General Counsel PHILIP B. SKLOVER, Associate General Counsel CAROLYN L. WHEELER, Assistant General Counsel EQUAL EMPLOYMENT OPPORTUNITY COMMISSION CERTIFICATE OF COMPLIANCE I certify that this brief complies with the type-volume limitation set forth in Fed.R.App.P. 32(a)(7)(B), and the length authorized for briefs as amicus curiae by Fed.R.App.P. 29(d). This brief contains 6,993 words. ____________________________ Dori K. Bernstein, Attorney CERTIFICATE OF SERVICE I hereby certify that two copies of this brief were mailed, first class, postage prepaid, on this 13th day of June, to the following: Michael B. Terry Joshua F. Thorpe Steven J. Rosenwasser Bondurant, Mixson & Elmore, LLP 3900 One Atlantic Center 1201 West Peachtree Street, N.W. Atlanta, Georgia 30309-3417 W. Ray Persons King & Spaulding 191 Peachtree Street, N.E. Atlanta, Georgia 30303-1763 Walter Christopher Arbery Hunton & Williams 600 Peachtree Street, N.E. Bank of America Plaza, Suite 4100 Atlanta, Georgia 30308-2216 Johnnie L. Cochran, Jr. J. Keith Givens Hezekiah Sistrunk, Jr. Cochran, Cherry, Givens,Smith & Sistrunk, P.C. Suite 800 The Candler Bldg 127 Peachtree Street Atlanta, Georgia 30303 Stephen William Riddell Troutman Sanders Bank of America Plaza 600 Peachtree Street, N.E. Suite 5200 Atlanta, Georgia 30308-2216 ______________________________ Dori K. Bernstein Attorney EQUAL EMPLOYMENT OPPORTUNITY COMMISSION Office of General Counsel 1801 L Street, N.W., Room 7046 Washington, DC 20507 (202)663-4734 1. Except where identification of a particular defendant company is relevant, defendants will be referenced collectively as "SoCo." 2. References to the district court's rulings will cite the published decisions. All other record references correspond to numbered docket entries (DE) on the district court docket. 3. Some of GPC's employees are covered by a collective bargaining agreement (CBA) with the International Brotherhood of Electrical Workers. The terms of the CBA govern promotion to positions covered by the CBA, and compensation of employees holding CBA-covered positions. See 205 FRD at 599. Plaintiffs do not claim CBA-covered promotions and compensation are racially discriminatory. See DE141 at 5-6. 4. DE119, Ex.57 at L1314 (1998 SoCo audit found "HR business reps have minimal involvement in selection process either through facilitation or consultation services"). 5. Internal SoCo audits reveal that in practice many jobs are filled without posting. See Ex.57 at L1314 (1998 audit revealed "high level of inconsistency in hiring process, including the use of JobNet and/or retaining interview records, and selection documentation"); Ex.62 at N1895 ("recent audit" of "approximately 1300 job transactions" found "50% were not posted and 50% had incomplete or no documentation"); Ex.63 at J5841 (spreadsheet showing 39.9% of 1342 jobs filled were posted). 6. The "non-exempt" and "exempt" designations indicate whether a position is subject to the wage and hour requirements of the Fair Labor Standards Act. See 205 FRD at 599. SoCo further divides non-exempt employees into "technical" and "non-technical" designations, and has established job grades and salary ranges for each. See DE119, Ex.72 at E675-79. 7. Ex.86 at L372, L364 (17.4% of workforce is African-American, compared to 7.4% of "officials & managers"). 8. Dr. Madden derived job function descriptions (such as accounting, customer services, engineering services, generation, human resources, marketing, security) from SoCo's database. DE141, Ex.B at App.A. 9. Dr. Madden measured the standard deviation for each year as follows: 1996 = 3.75; 1997 = 1.72; 1998 = 3.71; 1999 = 4.25. DE141, Ex.B at Table R4'. The Supreme Court has long approved standard deviation calculations as a "precise method of measuring the significance of ... statistical disparities" based on race. Hazelwood School Dist. v. United States, 433 U.S. 299, 309 n.14 (1977). The Court has recognized that "as a general rule for such large samples, if the difference between the expected value [i.e., the expected number of African-Americans promoted] and the observed number [i.e., the actual number of African-Americans promoted] is greater than two or three standard deviations, then the hypothesis that [employees] were [promoted] without regard to race would be suspect." Id. (internal citation omitted). 10. In 1994, GPC and OFCCP negotiated a conciliation agreement to resolve these violations. Following GPC's final report under the agreement, OFCCP informed SoCo in 1997 that "[t]he placement results for protected groups, particularly minorities and minority females, into [mid-level manager to executive level] positions over the two-year reporting period was poor." DE119, Ex.17. 11. Official citations to the court's summary judgment opinions are not yet available. For ease of reference, citations to each opinion will reference the name of the plaintiff whose claims are resolved and the WestLaw page number. 12. Under this proposal, the class claims of systemic discrimination (i.e., pattern-or-practice and disparate impact), and the individual claims of the seven named plaintiffs, would be resolved in this action. A judgment of liability on either systemic claim would require an award of class-wide declaratory and injunctive relief, and entitle absent class members to litigate separately their individual claims that they were denied promotions or compensation pursuant to SoCo's discriminatory policies. The district court acknowledged "controlling precedent" holding that "back pay is considered equitable relief and therefore can be awarded to class members in a case certified under Rule 23(b)(2)." 205 FRD at 627 (citing Pettway v. American Cast Iron Pipe Co., 494 F.2d 211, 257 (5th Cir. 1974)).