Chen-Oster et al v. Goldman, Sachs & Co. et alMEMORANDUM OF LAW in Opposition re: 246 MOTION to Certify Class . . DocumentS.D.N.Y.July 25, 2014UNITED STATES DISTRICT COURT SOUTHERN DISTRICT OF NEW YORK -------------------------------------------------------------------- x H. CRISTINA CHEN-OSTER; LISA PARISI; and SHANNA ORLICH, Plaintiffs, v. GOLDMAN, SACHS & CO. and THE GOLDMAN SACHS GROUP, INC., Defendants. : : : : : : : : : : : : 10 Civ. 6950 (AT) (JCF) -------------------------------------------------------------------- x DEFENDANTS’ MEMORANDUM OF LAW IN OPPOSITION TO PLAINTIFFS’ MOTION FOR CLASS CERTIFICATION Barbara B. Brown (admitted pro hac vice) Robert J. Giuffra, Jr. Neal D. Mollen (admitted pro hac vice) Theodore O. Rogers, Jr. Carson H. Sullivan (admitted pro hac vice) Robin D. Fessel PAUL HASTINGS LLP Suhana S. Han 875 15th Street, NW Jeffrey B. Wall Washington, DC 20005 SULLIVAN & CROMWELL LLP 125 Broad Street New York, New York 10004 (212) 558-4000 Attorneys for Defendants, Goldman, Sachs & Co. and The Goldman Sachs Group, Inc. July 3, 2014 Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 1 of 84 TABLE OF CONTENTS Page PRELIMINARY STATEMENT ..................................................................................................1 BACKGROUND ............................................................................................................................8 A. The Parties ...............................................................................................................8 1. The Named Plaintiffs ...................................................................................8 2. Goldman Sachs ............................................................................................9 B. The Non-Centralized and Non-Uniform Method of Evaluating, Compensating, and Promoting Professionals Within Goldman Sachs ..................12 1. Performance Evaluation Decisions ............................................................13 a. 360 Reviews .................................................................................. 13 b. Manager Summaries and Quartiling ............................................. 15 2. Compensation ............................................................................................16 3. Promotion ...................................................................................................20 C. Proceedings in This Case .......................................................................................21 1. Pleadings, Motion Practice, and Discovery ...............................................21 2. Plaintiffs’ Experts ......................................................................................23 a. Dr. Cascio ..................................................................................... 23 b. Dr. Farber ...................................................................................... 24 LEGAL STANDARD ..................................................................................................................31 ARGUMENT ................................................................................................................................32 I. PLAINTIFFS HAVE NOT SHOWN THAT THEY SATISFY EACH OF RULE 23(A)’S REQUIREMENTS. .........................................................................32 A. Plaintiffs Have Not Shown a Question Common to the Class ...............................32 1. Plaintiffs Have Not Shown a General Policy of Discrimination. ..............36 Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 2 of 84 -ii- a. The Record Establishes that Pay and Promotion Decisions Are Decentralized. ........................................................ 37 b. Plaintiffs’ Expert Evidence Is Either Irrelevant or Circular ......................................................................................... 39 c. Plaintiffs’ Anecdotal Evidence Is Extremely Misleading and Woefully Inadequate on its Own Terms ................................ 43 2. Plaintiffs Have Not Shown a Classwide Practice With a Classwide Disparate Impact. ......................................................................54 a. Plaintiffs’ Aggregate Statistical Evidence Cannot Show the Requisite Uniformity Across and Among Different Businesses. .................................................................................... 55 b. Plaintiffs’ Evidence Shows the Extent of Discretion and Decentralized Decisionmaking in the Evaluation, Compensation, and Promotion Processes. .................................... 57 B. Plaintiffs Are Not Adequate Class Representatives and Their Claims Are Not Typical. ....................................................................................................62 a. Plaintiff Orlich .............................................................................. 63 b. Plaintiff Chen-Oster ...................................................................... 64 II. PLAINTIFFS HAVE NOT ESTABLISHED THROUGH SUFFICIENT EVIDENTIARY PROOF THAT THEY SATISFY ANY OF RULE 23(b)’s REQUIREMENTS..................................................................65 A. Plaintiffs Have Not Met the Requirements of Rule 23(b)(3).. ...............................65 1. Plaintiffs Have Not Shown Predominance ................................................65 a. Liability Requires an Inherently Individualized Inquiry. ............. 66 b. Damages Require an Inherently Individualized Inquiry. .............. 67 2. Plaintiffs Also Have Not Shown Superiority.. ...........................................70 B. Plaintiffs Cannot Avoid the Requirements of Rule 23(b)(3).. ...............................72 1. Plaintiffs Cannot Proceed Under Rule 23(b)(2).. .......................................72 2. There is No Basis for Certifying a Class Under Rule 23(c)(4).. ................73 CONCLUSION ............................................................................................................................75 Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 3 of 84 -iii- Table of Authorities Page(s) Cases Abram v. UPS, 200 F.R.D. (E.D. Wisc. 2001) .................................................................................................42 Amchem Prods., Inc. v. Windsor, 521 U.S. 591 (1997) ...........................................................................................................65, 70 Am. Express Co. v. Italian Colors Rest., 133 S. Ct. 2304 (2013) .............................................................................................................31 Babbitt v. Albertson’s, Inc., No. 92 Civ. 1883, 1992 WL 605652 (N.D. Cal. 1992) ............................................................52 Bennett v. Nucor Corp., 656 F.3d 802 (8th Cir. 2011) ..................................................................................1, 34, 35, 59 Bickerstaff v. Vassar Coll., 196 F.3d 435 (2d Cir. 1999).....................................................................................................41 Blyden v. Mancusi, 186 F.3d 252 (2d Cir. 1999).....................................................................................................71 Bolden v. Walsh, 688 F.3d 893 (7th Cir. 2012) .............................................................................................42, 56 Byrd v. Lynch, No. 10 Civ. 247, 2011 WL 2680572 (D.N.J. July 8, 2011) .....................................................51 Chen-Oster v. Goldman Sachs & Co., 877 F. Supp. 2d 113 (S.D.N.Y. 2012) ......................................................................................72 Cokely v. N.Y. Convention Ctr. Operating Corp., No. 00 Civ. 4637, 2004 WL 1152531 (S.D.N.Y. May 21, 2004) ............................................52 Comcast Corp. v. Behrend, 133 S. Ct. 1426 (2013) .........................................................................................................3, 31 Cuevas v. Citizens Fin. Grp., Inc., 526 F. App’x 19 (2d Cir. 2013) ...............................................................................................33 Davis v. Cintas Corp., 717 F.3d 476 (6th Cir. 2013) ...............................................................................................1, 34 Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 4 of 84 -iv- Table of Authorities (continued) Page(s) DiLaura v. Power Auth. of State of N.Y., 982 F.2d 73 (2d Cir. 1992).......................................................................................................72 Dukes v. Wal-Mart Stores, Inc., No. CV 01–02252 CRB, 2013 WL 3993000 (N.D. Cal. Aug. 2, 2013) ..................................58 Easterling v. Connecticut Dep’t of Corr., 278 F.R.D. 41 (D. Conn. 2011)............................................................................................2, 69 Ellis v. Costco Wholesale Corp., 285 F.R.D. 492 (N.D. Ca. 2012) ..........................................................................................2, 39 Espenscheid v. DirectSat USA, LLC, 705 F.3d 770 (7th Cir. 2013) ...................................................................................................68 Faulk v. Sears Roebuck & Co., No. 11 Civ. 2159, 2013 WL 1703378 (N.D. Cal. Apr. 19, 2013) ...........................................51 Fernandez v. Wells Fargo Bank, N.A., Nos. 12 Civ. 7193, 12 Civ. 7194 (PKC), 2013 WL 4540521 (S.D.N.Y. Aug. 28, 2013) .........................................................................33 Godoy v. Maplehurst Bakeries, Inc., 747 F. Supp. 2d 298 (D.P.R. 2010) ..........................................................................................51 Green v. Borg-Warner Protective Services Corp., Nos. 95 Civ. 10419(RPP), 96 Civ. 0038(RPP), 95 Civ. 5558(RPP), 96 Civ. 8740(RPP), 1998 WL 17719 (S.D.N.Y. Jan. 16, 1998) ..............................................................................61 Griffin v. Bd. of Regents of Regency Univ., 795 F.2d 1281 (7th Cir. 1986) .................................................................................................30 Guadiana v. State Farm Fire & Cas. Ins. Co., No. 07 Civ. 326, 2010 WL 5071069 (D. Ariz. Dec. 7, 2010) ..................................................52 Gulino v. Bd. of Educ., 2013 WL 4647190 (S.D.N.Y. Aug. 29, 2013) ...........................................................................1 Halliburton Co. v. Erica P. John Fund, Inc., No. 13-317, 2014 WL 2807181 (June 23, 2014) ...........................................................3, 31, 35 In re Front Loading Washing Mach. Class Action Litig., No. 08 Civ. 51, 2013 WL 3466821 (D.N.J. July 10, 2013) .....................................................52 Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 5 of 84 -v- Table of Authorities (continued) Page(s) In re Literary Works in Elec. Databases Copyright Litig., 654 F.3d 242 (2d Cir. 2011).....................................................................................................63 In re Nassau County Strip Search Cases, 461 F.3d 219 (2d Cir. 2006).....................................................................................................73 Jones v. National Council of YMCA, No. 09 C 06437, 2014 WL 1305153 (N.D. Ill. Mar. 31, 2014) ...............................................58 La Day v. Catalyst Tech., Inc., 302 F.3d 474 (5th Cir. 2002) ...................................................................................................51 Lapin v. Goldman Sachs & Co., 254 F.R.D. 168 (S.D.N.Y. 2008) ............................................................................................63 Levitt v. J.P. Morgan Sec., Inc., 710 F.3d 454 (2d Cir. 2013).....................................................................................................31 Levitt v. PricewaterhouseCoopers, LLP, No. 04 Civ. 5179, 2007 WL 2106309, *1 (S.D.N.Y. July 19, 2007) ......................................51 Leyva v. Medline Industries Inc., 716 F.3d 510 (9th Cir. 2013) ...................................................................................................70 Lienhart v. Dryvit Sys., Inc., 255 F.3d 138 (4th Cir. 2001) ...................................................................................................67 Lott v. Westinghouse Savannah River Co., 200 F.R.D. 539, 560-61 (D.S.C. 2000) ...................................................................................42 Marisol A. v. Giuliani, 126 F.3d 372 (2d Cir. 1997).....................................................................................................62 Mason v. Stock, 955 F. Supp. 1293 (D. Kan. 1997) ...........................................................................................52 Mattenson v. Baxter Healthcare Corp., 438 F.3d 763 (7th Cir. 2006) ...................................................................................................48 McLaughlin v. Am. Tobacco Co., 522 F.3d 215 (2d Cir. 2008).....................................................................................................73 McReynolds v. Merrill Lynch, Pierce, Fenner & Smith, 672 F.3d 482 (7th Cir. 2012) ...............................................................................................2, 56 Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 6 of 84 -vi- Table of Authorities (continued) Page(s) McPartland v. Am. Broadcasting Cos., Inc., 1987 WL 14907, *1 (S.D.N.Y. July 21, 1987) .......................................................................73 Moore v. The Boeing Co., No. 4:02CV80 CDP, 2004 WL 3202777, *1 (E.D. Mo. Mar. 31, 2004) ................................42 Morgan v. UPS, 380 F.3d 459 (8th Cir. 2004) ...................................................................................................41 Oakley v. Verizon Commc’ns Inc., 09 CIV. 9175 (CM), 2012 WL 335657 (S.D.N.Y. Feb. 1, 2012) ............................................33 Ottaviani v. State Univ. of N.Y. at New Paltz, 875 F.2d 365 (2d Cir. 1989).....................................................................................................41 Panzirer v. Wolf, No. 79 Civ. 3445, 1980 WL 1395 (S.D.N.Y. Apr. 21, 1980) ..................................................63 Parisi v. Goldman, Sachs & Co., 710 F.3d 483 (2d Cir. 2013).......................................................................................................8 Penk v. Oregon State Bd. of Higher Educ., 816 F.2d 458, 465 (9th Cir. 1987) ....................................................................................42, 43 Personnel Adm’r of Mass v. Feeney, 442 U.S. 256 (1979) .................................................................................................................44 Puffer v. Allstate Ins. Co., 255 F.R.D. 450 (N.D. Ill. 2009), aff'd, 675 F.3d 709 (7th Cir. 2012) ................................43, 48 Randall v. Rolls-Royce Corp., No. 1:06-cv-860, 2010 WL 987484 (S.D. Ind. Mar. 12, 2010) .....................................1, 42, 49 Robinson v. Metro-North Commuter R.R. Co., 267 F.3d 147 (2d Cir. 2001).................................................................................................2, 70 Saleem v. Corporate Transp. Grp., Ltd., No. 12 Civ. 8450 (JMF), 2013 WL 6061340 (S.D.N.Y. Nov. 15, 2013) .........................................................................33 Schear v. Food Scope Am. Inc., 297 F.R.D. 114 (S.D.N.Y 2014) ..............................................................................................69 Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 7 of 84 -vii- Table of Authorities (continued) Page(s) Serrano v. Cintas Corp., Nos. 04-40132, 06-12311, 2009 WL 910702 (E.D. Mich. Mar. 31, 2009) ........................................................................36 Shayler v. Midtown Investigations, Ltd., No. 12 Civ. 4685(KBF), 2013 WL 772818 (S.D.N.Y. Feb. 27, 2013) ....................................33 Sheehan v. Purolator, Inc., 839 F.2d 99 (2d Cir. 1988).......................................................................................................48 Siddiqi v. New York City Health & Hospitals Corp., 572 F. Supp. 2d 353 (S.D.N.Y. 2008) ......................................................................................59 Smith v. City of Jacksonville, 544 U.S. 228 (2005) .................................................................................................................54 Smith v. Va. Commonwealth Univ., 84 F.3d 672 (4th Cir. 1996) ...................................................................................................43 Sobel v. Yeshiva Univ., 839 F.2d 18 (2d Cir. 1988).......................................................................................................41 Spagnola v. Chubb Corp., 264 F.R.D. 76 (S.D.N.Y. 2010) ..............................................................................................63 Stagi v. National R.R. Passenger Corp. 391 Fed.Appx. 133 (3d Cir. 2010) ...........................................................................................42 Stockwell v. City and County of San Francisco, 749 F.3d 1107 (9th Cir. 2014) ...........................................................................................35, 57 Tabor v. Hilti, Inc., 703 F.3d 1206 (2013) ...........................................................................................................1, 34 Teamsters v. United States, 431 U.S. 324 (1977) .....................................................................................................48, 66, 67 Trachtenberg v. Dep’t of Educ. of City of New York, 937 F. Supp. 2d 460 (S.D.N.Y. 2013) ......................................................................................59 Trezza v. The Hartford Inc., No. 98 Civ. 2205, 1999 WL 511673 (S.D.N.Y. July 20, 1999) ..............................................46 Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 8 of 84 -viii- Table of Authorities (continued) Page(s) United States v. City of New York, 717 F.3d 72 (2d Cir. 2013)...................................................................................................7, 44 United States v. City of New York, 276 F.R.D. 22 (E.D.N.Y. 2011) ...........................................................................................2, 69 Valentine v. Standard & Poor’s, 50 F. Supp. 2d 262 (S.D.N.Y. 1999), aff’d, 205 F.3d 1327 (2d Cir. 2000) .............................59 Wal-Mart Stores, Inc. v. Dukes, 131 S. Ct. 2541 (2011) ..................................................................................................... passim Wang v. Hearst Corp., 293 F.R.D. 489 (S.D.N.Y. 2013) .............................................................................................68 Wards Cove Packing Co. v. Atonio, 490 U.S. 642 (1989) .................................................................................................................54 Watson v. Fort Worth Bank & Trust, 487 U.S. 977 (1988) .................................................................................................................54 Statutes 42 U.S.C. § 2000e-5(g)(2)(A) ........................................................................................................67 42 U.S.C. § 2000e-8(e) ..................................................................................................................52 Rules Enabling Act, 28 U.S.C. § 2072(b) ......................................................................................68 Other Authorities Federal Rule of Civil Procedure 23 ....................................................................................... passim U.S. Const. Amend. VII .................................................................................................................71 Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 9 of 84 - PRELIMINARY STATEMENT Three years ago, in Wal-Mart Stores, Inc. v. Dukes, 131 S. Ct. 2541 (2011), the Supreme Court refused to certify a Title VII employment class whose members held “a multitude of different jobs, at different levels of Wal-Mart’s hierarchy, for variable lengths of time, in 3,400 stores, sprinkled across 50 states, with a kaleidoscope of supervisors (male and female).” Id. at 2557 (internal quotation marks omitted). Through their motion, Plaintiffs ask this Court to go further and to take the extraordinary—indeed, unprecedented—step of certifying a Title VII class not of retail store employees, but of more than 2,300 highly paid women professionals in one of the most complex financial institutions in the world. Both before and after Wal-Mart, federal courts have repeatedly denied class certification in employment discrimination cases where, as here, plaintiffs cannot show the existence of common questions suitable for classwide resolution. In fact, no fewer than four federal courts of appeals have rejected attempts to cobble together a class of different types of employees, performing different functions in different departments or facilities.1 Tellingly, the cases that Plaintiffs cite (see Pl. Br. 2-3) all involved a single job or a small number of similar jobs: Gulino v. Bd. of Educ., No. 96-cv-8414 (KMW) 2013 WL 4647190 (S.D.N.Y. Aug. 29, 2013), one job—school teacher—and an allegedly biased standardized test; 1 See, e.g., Davis v. Cintas Corp., 717 F.3d 476, 489 (6th Cir. 2013) (rejecting putative class consisting of “a number of women, who failed to obtain employment at many places, over a long time, under a largely subjective hiring system”); Tabor v. Hilti, Inc., 703 F.3d 1206, 1229 (10th Cir. 2013) (rejecting putative class of female employees with “highly individualized facts and circumstances”); Bennett v. Nucor Corp., 656 F.3d 802, 815 (8th Cir. 2011) (rejecting putative class of employees with “stark inter-departmental variations in job titles, functions performed, and equipment used”); Randall v. Rolls-Royce Corp., 637 F.3d 818, 821-23 (7th Cir. 2011) (rejecting putative class “sprawl[ed] over twenty different compensation grades, including supervisory and nonsupervisory positions and encompassing starting salaries ranging from $40,050 to $190,750”). Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 10 of 84 -2- United States v. City of New York, 276 F.R.D. 22 (E.D.N.Y. 2011), one job— firefighter—and two allegedly biased standardized tests; Easterling v. Conn. Dep’t Corr., 278 F.R.D. 41 (D. Conn. 2011), one job— corrections officer—and an allegedly biased standardized fitness test; Ellis v. Costco Wholesale Corp., 285 F.R.D. 492 (N.D. Ca. 2012), retail store- level employees—challenging promotions to one of two store-level manager positions; and McReynolds v. Merrill Lynch, Pierce, Fenner & Smith, 672 F.3d 482 (7th Cir. 2012), one job in one business unit—financial advisors in Merrill’s retail business—challenging two allegedly biased rules, on account assignment and team formation.2 Here, by contrast, the members of Plaintiffs’ proposed class—present and former Vice Presidents and Associates at Goldman, Sachs & Co. (“Goldman Sachs” or the “Firm”) over more than a decade—did not perform the same job or even the same handful of jobs. Instead, these professionals worked in three separate Divisions of the Firm comprising 140 separate businesses (the “Business Units”), 3 each run by managers with broad discretion over performance evaluation, compensation, and promotion. Across these Business Units, the proposed class members advised on mergers and acquisitions in the healthcare industry, traded petroleum futures, structured financial derivatives, ran investment funds based on mathematical algorithms, and so on. And, even within the same Unit, they performed a wide array of functions—for example, client relationship professionals sat side-by-side with strategists and desk analysts who may have never met a client. 2 Plaintiffs also rely on the pre-Wal-Mart decision in Robinson v. Metro-North Commuter R.R. Co., 267 F.3d 147 (2d Cir. 2001), but the portion that they cite merely describes standards of proof on the merits. Plaintiffs recognize that Wal-Mart has changed the standards governing certification of employment class actions in the Second Circuit. (See Pl. Br. 9-11.) 3 Although the Business Units numbered from at least 90 up to 110 in any particular year, there were 140 in all during the proposed class period. This number understates the total Business Units by excluding those that existed only prior to 2005. (Expert Report of Michael P. Ward, Ph.D. (“Ward Report”) 54-58 (Figures 8-10).) Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 11 of 84 -3- Because class actions are “an exception to the usual rule that litigation is conducted by and on behalf of the individual named parties only,” Comcast Corp. v. Behrend, 133 S. Ct. 1426, 1432 (2013) (internal quotation marks omitted), “plaintiffs wishing to proceed through a class action must actually prove—not simply plead—that their proposed class satisfies each requirement of Rule 23.” Halliburton Co. v. Erica P. John Fund, Inc., No. 13-317, 2014 WL 2807181, at *12 (June 23, 2014) (emphasis in original). In Wal-Mart, the Supreme Court made clear that to certify a class, plaintiffs must “demonstrate that the class members have suffered the same injury.” 131 S. Ct. at 2551 (emphasis added). In a Title VII action, where “the crux of the inquiry is the reason for a particular employment decision,” a plaintiff must prove that there is “some glue holding the alleged reasons for all [the challenged] decisions together.” Id. at 2552. Unless plaintiffs make that showing, “it will be impossible to say that examination of all the class members’ claims for relief will produce a common answer to the crucial question why was I disfavored.” Id. In their original Complaint, Plaintiffs alleged that Goldman Sachs permitted too much manager discretion in employment decisions. (See Class Action Complaint filed Sept. 16, 2010, Dkt. 5 ¶ 48.) After Wal-Mart, Plaintiffs shifted course completely and now allege that individual managers (men and women) uniformly applied the Firm’s procedural frameworks for assessing performance—“360 reviews” and “manager quartiling”—to discriminate against women professionals across 140 separate Business Units. In denying as premature Goldman Sachs’ motion to strike Plaintiffs’ class allegations, this Court underscored the need for “significant proof” of “a common mode of exercising discretion.” Report and Recommendation of M.J. Francis, dated Jan. 19, 2012 (“January 19 Order”) at 12-13 (quoting Wal-Mart, 131 S. Ct. at 2554). Fatally for their motion, Plaintiffs provide no such proof. Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 12 of 84 -4- To try to overcome Wal-Mart, Plaintiffs propose to certify a disparate treatment class by claiming that there is “‘[s]ignificant proof that [Goldman Sachs] operated under a general policy of discrimination. . . .’” Id. at 2553. Plaintiffs speculate that some “central” group of senior executives either itself discriminates against the Firm’s women professionals or has issued “some common direction” that individual managers do so. (Pl. Br. 3, 39.) But, after extensive discovery, Plaintiffs have not identified any supposed group of “central” wrongdoers, much less introduced any proof of an institutional Firm policy of gender discrimination spanning 140 separate businesses, many of them run by women managers. Instead, Plaintiffs and their statistical expert, Dr. Henry Farber, try to show that Goldman Sachs has a general policy of discrimination by impermissibly aggregating data about compensation and promotion for professionals who worked in 140 different Business Units (Pl. Br. 38, 40). But the Supreme Court has made clear that “information about disparities at the [Firm] level does not establish the existence of disparities at individual [Business Units], let alone raise the inference that a company-wide policy of discrimination is implemented by discretionary decisions at the [Business Unit] level.” Wal-Mart, 131 S. Ct. at 2555 (brackets and internal quotation marks omitted) ); see infra at 33-34 & n.22 (collecting cases). And, making matters worse, Dr. Farber ignores the highly individualized factors affecting compensation and promotion, including the professional’s position, his or her performance, the performance of the professional’s Business Unit, market demand for the professional’s services, and the discretionary judgments of individual managers. As a result, Dr. Farber’s statistics have so little explanatory power that they are essentially meaningless. But, even taking Dr. Farber’s analyses at face value, and not correcting for his multiple errors, the differences between male and female professionals are either Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 13 of 84 -5- statistically insignificant or statistically significant and favorable to women in the vast majority of Business Units, confirming that there is no classwide issue for resolution Specifically, the differences in compensation are insignificant or favorable for women Associates in 90% of the Business Units and for women Vice Presidents in 76% of those Units. (See Ward Report, Appendix A (Chart 1).) Similarly, the differences in 360 review results and manager quartiling are insignificant or favorable to women in more than 80% of the Firm’s Business Units that Dr. Farber studied. (See id. Charts 2 and 4.) And, as to promotion, even Dr. Farber’s flawed statistics show a “shortfall” of only 19 women over his cherry-picked five-year period, when 384 Vice-Presidents were promoted. In short, “[e]ven if they are taken at face value, [Dr. Farber’s] studies are insufficient to establish that [Plaintiffs’] theory can be proved on a classwide basis.” 131 S. Ct. at 2555. Having no proof—much less significant proof—of central control or a general policy of discrimination, Plaintiffs revert (Pl. Br. 40) to a handful of anecdotes of supposed hostility toward women. To be clear: Goldman Sachs does not tolerate discrimination of any kind. Tellingly, after almost four years of highly publicized litigation, the two named Plaintiffs and their lawyers have not found a single other woman to join in this suit. In fact, on this motion they have put forward only six other declarants, four of whom are not even members of the putative class. At a firm of Goldman Sachs’ size and complexity, this purported evidence of discriminatory treatment is less than paper-thin; it is virtually nonexistent. As in Wal-Mart, Plaintiffs’ anecdotes are “too weak to raise any inference that all the individual, discretionary personnel decisions are discriminatory.” 131 S. Ct. at 2556. The remainder of Plaintiffs’ evidence consists largely of mischaracterizations of Goldman Sachs’ extensive efforts to achieve workplace diversity. Plaintiffs also rely on a series of unsubstantiated and inadmissible Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 14 of 84 -6- allegations, the vast majority of which are irrelevant because This case is a textbook example of why class certification requires proof, not a broad brush and a bucket of mud. As an alternative to their unsubstantiated contention that Goldman Sachs operates under a general policy of discrimination, Plaintiffs seek to certify a disparate impact class, claiming that Goldman Sachs’ procedural framework for assessing the performance of Vice Presidents and Associates—360 reviews and manager quartiling—disparately impacts women in compensation and promotion. But every company of any size in America has a framework for evaluating its employees. What Plaintiffs must show to certify their proposed class is not merely the existence of some procedural framework, but that the Firm’s undeniably gender-neutral processes have been uniformly applied—across a class of more than 2,300 women professionals in 140 Business Units—to disparately impact those professionals. As the Supreme Court made clear in Wal-Mart, “demonstrating the invalidity of one manager’s use of discretion will do nothing to demonstrate the invalidity of another’s.” 131 S. Ct. at 2554. The record is uncontroverted that the 360 reviews are not comparable to a uniform business practice, such as a standardized test, that may have a direct adverse impact on all members across a class. Rather, the 360 reviews simply consist of co-workers’ subjective assessments of each professional’s performance, taking into account the requirements of his or her position and their own observations. For example, co-workers necessarily apply different criteria in evaluating a client-facing investment advisor and a mathematician who develops algorithmic trading strategies, even if they both work in the same Business Unit. And, in making quartile assignments, individual managers consider many factors, including the 360 reviews and Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 15 of 84 -7- their own observations. Again, Plaintiffs’ own statistics confirm that Plaintiffs’ proposed class cannot be certified: in more than 80% of the Business Units, the 360 review and manager quartile results were either statistically insignificant or favorable to women professionals. In short, Plaintiffs cannot satisfy their burden to show some “glue” that binds “all those decisions together” by pointing to 360 reviews and manager quartiling. Wal-Mart, 131 S. Ct. 2552. In addition to not proving commonality, Plaintiffs have not established adequacy and typicality as required by Federal Rule of Civil Procedure 23(a). Indeed, it would be difficult to imagine two more idiosyncratic claimants. Ms. Orlich, who worked for just over a year as an entry-level Associate in the Goldman Sachs’ Securities Division and was paid in lockstep with her male colleagues, was not at the Firm long enough to have been affected by the pay and promotion procedures challenged here. Ms. Chen-Oster, who resigned voluntarily from the Securities Division in 2005, advances an individualized retaliation claim. Finally, Plaintiffs cannot remotely show, as they must to satisfy Federal Rule of Civil Procedure 23(b), how class issues predominate over individual ones. Even if a class were certified here and Plaintiffs could prove at trial a pattern or practice of discrimination or an overall disparate impact, that would not materially advance this litigation. Goldman Sachs still would have the right to show—as to each and every class member—that its compensation or promotion decisions were not discriminatory. Given the variety of positions held by putative class members, and the multiplicity of factors influencing the Firm’s highly discretionary compensation and promotion decisions, the Court still would have to hold a series of lengthy individual trials on the particulars of what happened to each class member. This confirms that a class action is neither a necessary nor superior means of resolving the claims of over 2,300 highly paid professionals in 140 different Goldman Sachs’ Business Units. Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 16 of 84 -8- BACKGROUND A. The Parties 1. The Named Plaintiffs The two remaining named Plaintiffs were collectively employed at Goldman Sachs for just a third of the putative class period (2002 to present).4 Both worked in only one Division—the Securities Division, which itself had more than 35 separate Business Units since 2005. (Ward Report 54-58 (Figures 8-10).) Cristina Chen-Oster worked exclusively in the Equities area of the Securities Division from 1997 until she voluntarily resigned in 2005. (Chen-Oster Dep. at 440:2-19 and Ex. 19)5 For most of that time, she sold convertible bonds in the Convertible Sales Business Unit. (Id. at 440:2-19.) At the end of 2004, after the Firm disbanded that Unit, Ms. Chen-Oster worked briefly in U.S. Research Sales, another Business Unit within Equities, before voluntarily resigning. (Id. at 421:5-10.) Ms. Chen-Oster’s individual allegations bear no relationship to her class allegations. She claims that her employment was adversely affected by an alleged 1997 incident involving a co-worker (a peer with no direct supervisory authority over her). (Plaintiffs’ First Amended Complaint, Dkt. 104 ¶¶ 74-75.) Even though her compensation increased by 68% from 1998 to 2000, Ms. Chen-Oster claims that her 4 Lisa Parisi, the other Plaintiff named in the Amended Complaint, was a party to an arbitration agreement with the Firm. The Second Circuit enforced her promise to arbitrate her claims. See Parisi v. Goldman, Sachs & Co., 710 F.3d 483 (2d Cir. 2013). 5 Unless otherwise noted, all references to “Ex.” are to the exhibits accompanying the Declaration of Theodore O. Rogers, Jr. Excerpts of deposition transcripts cited herein are attached to the Rogers Declaration as Exhibits 1 through 17, in alphabetical order. Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 17 of 84 -9- manager retaliated against her for reporting this alleged incident, in making such routine supervisory decisions as assigning accounts. (See Ex. 18).) Shanna Orlich was a short-term, junior employee in a sub-unit of the Americas and Credit Trading Business Units within the Securities Division. She joined the Firm out of business school in the fall of 2007. As a new Associate, Ms. Orlich was paid the same as her peers, male and female, and was never a candidate for promotion. (Orlich Dep. at 48:1-11, 509:3-9.) Ms. Orlich thus was never subject to the compensation or promotion processes that she seeks to challenge on behalf of a class. Her employment ended in 2008, in the middle of the financial crisis, as part of an extensive reduction in force. (Id. at 10:4-7, 458:22-459:19.) 2. Goldman Sachs Plaintiffs seek to represent a class of women professionals, in a multitude of highly specialized jobs, from three of Goldman Sachs’ four revenue-generating Divisions: Securities, Investment Banking, and Investment Management. These Divisions operate a multiplicity of different businesses, with many Business Units and subunits within each Division. The Securities Division, comprised of the Equities and Fixed Income, Currencies, and Commodities (“FICC”) sub-divisions, transacts in stocks and other equity-related products, options and futures, interest rate products, credit products, mortgages, currencies, and commodities. (Miller Dep. at 51:10-23; 58:22-59:20; see also Declaration of John Levene (“Levene Decl.”) ¶ 2; Declaration of Mary Louise Kirk (“Kirk Decl.”) ¶¶ 2-3.) The Investment Banking Division (“IBD”) has two large Sub-Divisions, classic banking and financing. Classic bankers specialize in particular industries, such as industrial, consumer/retail, technology/media, and telecommunications, and they advise boards and senior executives of clients on a variety of strategic issues, such as whether to pursue a merger, a debt offering, or an equity offering. Each banker must develop a specialized knowledge of her Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 18 of 84 -10- coverage industries and develop and maintain long-term client relationships. (Declaration of Sue Benz (“Benz Decl.”) ¶¶ 6, 10.) Financing bankers, by contrast, become involved in transactions when financing is needed, and they specialize in developing bespoke financial products, such as leveraged finance issuances, structured finance issuances, and equity issuances. (Declaration of Craig Packer (“Packer Decl.”) ¶ 4.) The Investment Management Division (“IMD”), which is separated by informational barriers from the Securities and IBD Division, also has two broad Sub-Divisions: Private Wealth Management (“PWM”) and Goldman Sachs Asset Management (“GSAM”). (Felix Dep. at 7 at 47:5-48:22; 83:16-85:3; 205:17-206:12.) PWM provides wealth advisory and investment management services to high net-worth individuals and families, and selected foundations and endowments. (Declaration of Megan Taylor (“Taylor Decl.”) ¶ 3.) GSAM manages asset funds for the Firm’s institutional and PWM clients. (Felix Dep. at 83:25-84:16; see also Declaration of James McNamara (“McNamara Decl.”) ¶ 3.) Each of these Divisions includes many individual Business Units. (Attached as Appendix A is a schematic of the Business Units in the Investment Banking, Investment Management and Securities Divisions since 2005. The two named Plaintiffs worked in the Business Units (or their successors) circled in red.) The diversity of the positions within this complex set of businesses and the corresponding diversity of skills required of the professionals within the putative class are described fully in the accompanying declarations.6 Here are four examples of this diversity: 6 See Benz Decl. ¶¶ 3-13; Declaration of Donald Casturo (“Casturo Decl.”) ¶¶ 3-11; Declaration of Darren Cohen (“D. Cohen Decl.”) ¶¶ 3-11; Declaration of Celeste Guth (“Guth Decl.”) ¶¶ 4-12; Kirk Decl. ¶¶ 3-9, 14-17; Levene Decl. ¶¶ 3-17; Declaration of Todd Lopez (“Lopez Decl.”) ¶¶ 3-8; McNamara Decl. ¶¶ 3-17; Packer Decl. ¶¶ 3-12; Declaration of Stephen Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 19 of 84 -11- 1. The Prime Brokerage Business Unit within the Equities Sub-Division of the Firm’s Securities Division provides services to Goldman Sachs’s hedge fund clients to help them manage risk, monitor their portfolios, maintain liquidity, and build their businesses. (Levene Decl. ¶¶ 2-5.) Prime Brokerage consists of 10 separate departments, with professionals performing a different role in each department. (Id.) For example, professionals in Hedge Fund Consulting work with clients on such varied issues as technology and infrastructure, real estate, human capital management, and operations, while those in Capital Introduction evaluate client capital needs and introduce clients and investors. (Id. ¶¶ 6, 11.) 2. Professionals in the Strategies Business Unit within FICC (“FICC Strats”) collaborate with sales and trading desks across FICC to develop quantitative and technological solutions for various businesses and their clients. (Kirk Decl. ¶ 3.) FICC Strategies professionals themselves fall into four general categories—Sales Strats, Desk Strats, Core Strats and Divisional Strats. Sales Strats, for example, work with the Firm’s sales force and clients on matters such as risk analysis and transaction structure. (Id. ¶¶ 5, 15-17.) Sales Strats are divided into departments focusing on different areas, including mortgages, foreign exchange, Latin America, interest rates, commodities, and e-products. (Id. ¶ 4.) Strats in the Mortgages department often have Ph.Ds in mathematics. In the Foreign Exchange and Commodities departments, professionals frequently have advanced financial engineering degrees. E-Products Strats typically have experience in graphic design or computer programming. (Id. ¶ 6.) 3. Within the GSAM sub-division of IMD, professionals in the U.S. Institutional Business Unit sell and service the Firm’s own financial products—including equities, fixed income, private equity and hedge funds—to institutional clients. (Russell Decl. ¶¶ Pierce (“Pierce Decl.”) ¶¶ 3-12; Declaration of Peter Craig Russell (“Russell Decl.”) ¶¶ 3-9; and Taylor Decl. ¶¶ 3-8. Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 20 of 84 -12- 3-4.) Professionals in this Unit include sales professionals, who pitch products directly to clients; internal strategists, who issue investment recommendations to client-facing sales teams based on risk assessments and financial modeling, and who often have advanced degrees and expertise in specialized fields such as pension regulation and insurance strategy; and professionals responsible for client relationship management. (Id. ¶¶ 6, 7-9.) 4. Professionals in the Leveraged Finance Business Unit (“Leveraged Finance”) within the Financing Group of IBD, work with below-investment-grade corporate clients to raise debt financing, including capital bank loans, “junk” bonds, and capital restructurings. (Packer Decl. ¶ 7.) Professionals in Leveraged Finance fall into two categories: Originators, who structure and execute transactions; and the Syndicate Desk, whose members distribute financial products and work with the Firm’s sales force to sell the products to investors. (Id. ¶¶ 10-11.) Originators must excel at credit analysis and client management, while professionals on the Syndicate Desk, who commit the Firm’s capital, must have extensive market understanding and sound judgment of risk. (Id.) There are a myriad of additional examples across the 140 Business Units covered by Plaintiffs’ proposed class of more than 2,300 women professionals. Nothing about the diverse roles of these professionals—the skills required, the criteria for success, or the market value of the role—is uniform. B. Goldman Sachs’ Non-Centralized and Non-Uniform Decision-Making Process for Evaluating, Compensating, and Promoting Professionals The record shows that the evaluation, compensation and promotion decisions that Plaintiffs seek to challenge were not (and are not) centralized or uniform. Rather, these decisions turn on the substantial discretion of hundreds of individual managers. Plaintiffs point to the Firm’s procedural frameworks—for example, that each Division goes through “rounds” in Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 21 of 84 -13- setting compensation as the fiscal year draws to a close (Pl. Br. 9)—but they studiously ignore the different bases on which individual managers measure performance and make compensation decisions about individual professionals. Assessing the performance of a banker who specializes in healthcare industry mergers and acquisitions is simply not the same as assessing the performance of a mathematician who develops algorithms to drive computerized trading. As women professionals across the putative class have confirmed, the Firm recognizes and rewards the talents of women.7 1. Performance Evaluation Decisions To assist individual managers in evaluating the Firm’s professionals, Goldman Sachs, like many businesses, uses so-called “360 reviews” to collect co-worker assessments and asks individual managers to rate professionals in their Business Units in broad categories (“quartiling”). a. 360 Reviews Each year, every Goldman Sachs professional performs a self-evaluation and proposes 8-12 evaluators—subordinates, peers, more senior colleagues, and internal clients— 7 See Declaration of Jean Altier Bohm (“Bohm Decl.”) ¶ 16 (“In my experience, Goldman Sachs is a meritocracy, where hard work and strong performance are rewarded, including through compensation and career advancement”); Declaration of Stephanie Cohen (“Cohen Decl.”) ¶ 5 (“I have worked hard and my contributions to the Firm have been recognized.”); Declaration of Arianne Criqui (“Criqui Decl.”) ¶ 11 (“In my experience, Goldman Sachs has a team-oriented, supportive culture where I have been rewarded for my hard work.”); Declaration of Carolyn Sabat (“Sabat Decl.”) ¶ 10 (“In my experience, Goldman Sachs is a meritocracy where hard work is rewarded and where working mothers, like myself, can succeed.”); Declaration of Clare Scherrer (“Scherrer Decl.”) ¶ 11 (“I believe the process of assessing relative performance is an important one, particularly at Goldman Sachs where we hire the very best and brightest individuals.”); see also Declaration of Ritu Kalra (“Kalra Decl.”) ¶ 7; Declaration of Nora Creedon (“Creedon Decl.”) ¶ 13; Declaration of Lora Robertson (“Robertson Decl.”) ¶ 13; Declaration of Merrill Alice Chester (“Chester Decl.”) ¶ 9; Declaration of Samantha Davidson (“Davidson Decl.”) ¶ 6; Declaration of Ashley Eckhard (‘Eckhard Decl.”) ¶ 4; Declaration of Susie Scher (“Scher Decl.”) ¶ 13. Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 22 of 84 -14- with whom the professional has worked over the prior year. (Landman Dep. I at 64:12-16.) Based on their own experiences, these evaluators rate the professional on a numerical scale across a range of topics and provide narrative explanations for their ratings. Unlike a standardized test score or a pass-fail eligibility test, the results of these 360 reviews represent collections of disparate co-workers’ ratings and comments on broad criteria that may apply with varying degrees to each job—including “Technical Skills,” “Communication Skills,” “Teamwork,” and “Leadership.” (Ex. 21, Kung Dep. II at 307:2-8.) The Firm also permits managers to revise the proposed panel of reviewers to ensure well-rounded feedback. (See Landman Dep. I at 63:14-64:16; 134:8-20.) Goldman Sachs has devoted substantial resources to developing the 360 review process. (See Landman Dep. I at 57:22-58:2.) The Firm’s in-house professional Talent Management staff implements the 360 process, and two respected external consulting firms have formally validated it. (Id. at 8:1-9, 18:16-19:5.) Dr. Michael Campion of Purdue University, an expert retained by Goldman Sachs in this case, found that “Goldman Sachs’s [360 review] practices meet nearly every one of [the 56] best practices [identified in the professional literature]. The firm has a process designed with evident care that surpasses other such systems I have encountered.” (Expert Report of Dr. Michael Campion (“Campion Report”) at ¶ 15.) Dr. Wayne Cascio, Plaintiffs’ industrial and organizational psychologist, agrees that 360 review systems can be “very, very valuable” and that one very important purpose of such reviews is to obtain feedback about employees from a variety of different perspectives, so that one gets “a better rounded view of the employee’s performance.” (Cascio Dep. at 46:5-47:24.) Dr. Cascio also acknowledges that 360 reviews are “pretty good tool[s] to foster teamwork.” (Id. at 162:5-7.) Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 23 of 84 -15- b. Manager Summaries and Quartiling After receiving the 360 reviews, supervising managers in each of the Business Units add their own summaries and make their own evaluations of their reports. Like companies that evaluate performance with narrative categories such as “exceeds expectations” or “needs improvement,” Goldman Sachs asks managers to sort professionals into “quartiles,” with the first quartile designating the most successful professionals. In making these decisions, individual managers may consider many factors, including the 360 review feedback, diversity and citizenship-related activities, long-term commercial contribution, potential to assume greater responsibility, and any other information that they consider relevant, such as individual production metrics. (See Exs. 10B, 10C.) The Firm does not require managers to follow any formula in weighting the significance that they place on their own observations or the feedback that they receive, including from the 360 reviews. (Landman Dep. II at 37:22-38:11.)8 To ensure a useful distribution, each Business Unit manager (or, depending on the size of the Unit, each desk head within the Unit) is asked to place a percentage of the Unit’s professionals in the first quartile, a larger percentage in the next two groupings, and so on. (See, e.g., Kung Dep. II at 316:19-318:24.) The Business Unit recommendations may be adjusted to a small degree within the Division and by Human Capital Management (“HCM”) principally to 8 See also McNamara Decl. ¶ 18 (“When making manager quartile decisions, the factors we consider vary by channel [sub group within the Business Unit]”); Levene Decl. ¶ 18, n.2 (“I could never assign uniform weight to the 360 results due to the wide variation in the jobs across the Departments [in the Business Unit]”; Casturo Decl. ¶ 17, n.4 (“[W]e may place more weight on the 360 feedback for a junior sales representative whose role it is to support more senior representatives … than on the 360 feedback for a more senior representative who is more directly accountable to clients.”); Russell Decl. ¶ 15, n.2 (“[W]hen making manager quartile assignments … I generally place more emphasis on production numbers for those in the sales role than I do on 360 feedback from internal Goldman Sachs professionals. However, for my internal strategists, who do not have production and are not client facing, I generally give more weight to how those with whom they work internally have reviewed them.”). Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 24 of 84 -16- ensure quartile recommendations meet the distribution goals. (Heller Dep. II at 320:6-322:10.)9 This “managed distribution” prevents “grade inflation” from affecting individual managers’ discretionary compensation and promotion decisions. 10 As Goldman Sachs’ expert Dr. Campion explained, asking managers to distribute their evaluations of their reports “mitigates the tendency to homogenize evaluation scores”; in the absence of some such guideline, there is a strong tendency to rate nearly all professionals as “above average.” (Campion Report ¶¶ 50, 59.) Notably, Plaintiffs’ expert, Dr. Cascio, acknowledges that he has “absolutely” recommended such managed distribution to employers in order to avoid homogenized ratings, and that he has never recommended that an employer eliminate such a process. (Cascio Dep. at 45:15-46:4.) 2. Compensation Most professionals at Goldman Sachs receive a salary and a discretionary year- end bonus (although some professionals in the proposed class are paid sales commissions instead of bonuses). Year-end bonuses generally are the largest portion of compensation and frequently amount to multiples of salary. (See Mehling Dep. at 227:1-18.) 9 In IBD and IMD, those teams will ask managers for further explanations if they see significant discrepancies between 360 review scores and the recommended quartiles. (See Larson Dep. at 161:10-13; 188:9-188:21; Kung Dep. II at 322:4-23; 332:1-12.) In the Securities Division, 360 review scores and manager quartiles are provided by what is called a “Franchise Manager” for discussion during meetings held at the Business Unit level to discuss compensation decisions within the Business Unit. (See Heller Dep. at 65:21-67:15; 68:3-69:6. ) 10 In referring to manager quartiling as “forced ranking,” Plaintiffs misstate the record. Managers do not “force rank” their direct reports one by one from first to last. (See Landman Dep. II at 15:3-11; Heller Dep. at 84:15-85:2, 90:17-91:1, Larson Dep. at 191:23-192:10; Kung Dep. II at 311:16-312:8.) Instead, quartiling groups professionals into large buckets. And this process has “soft shoulders”—flexibility at the borders of each quartile to ensure that professionals are not “forced” into a category in which they do not belong. (Landman Dep. II at 52:2-17, 131:9-16; Heller Dep. II 323:17-25.) Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 25 of 84 -17- The Firm’s Chief Financial Officer (“CFO”) allocates bonus pools to each revenue-producing Division based largely on the profitability of the Firm and of the Division itself. Division leaders then make allocations to managers in each Business Unit, based principally on the Units’ relative performance and contribution to the Division during the year. (Kung Dep. I at 58:21-25.) In making such allocations, Division leaders may take into account other factors, such as the Unit’s potential for growth, risk management, use of Firm capital, and competitive pressures affecting the Unit. The money available for each manager to allocate to professionals in his or her Business Unit varies from year to year and from one Business Unit to another in each Division, because the financial performance of each Unit varies, sometimes dramatically, based on market conditions and other factors. (See Kung Dep. I at 58:21-25; Mehling Dep. 59:8-12, 147:16-19.) After budgets are allocated to Business Units, managers propose compensation for their reports. (Kung Dep. I at 58:1-9, 58:21-59:2; Larson Dep. at 61:10-19.) 11 Typically, this process takes place within sub-units (the department level or below) of each Business Unit. (Heller Dep. 65:21-66:9.) As a result, professionals in a Unit that had a very successful year would be expected to receive larger bonuses than professionals in a less successful Business Unit, even if other factors affecting pay were the same. (See Kung Dep. I at 71:23-72:19.) Thus, 11 See Packer Decl. ¶ 15 (“Every year, as the Business Unit Leader, I receive a separate budget from IBD leadership to allocate among the professionals in my Business Unit.”); Lopez Decl. ¶ 11 (“When making my compensation decisions, I have to stay within the budget I am given for my Business Unit. I do not compare those in GSET with individuals outside of the Business Unit when making compensation decisions.”); Casturo Decl. ¶¶ 12-13 (“each desk head (all of whom are Managing Directors) recommends to us during a meeting with the head of the Business Unit the year-end bonuses they want to pay each of the individuals on their desks.”); McNamara Decl. ¶ 5; see also Kirk Decl. ¶ 10; Russell Decl. ¶ 13; D. Cohen Decl. ¶ 12; Benz Decl. ¶ 17-18; Pierce Decl. ¶ 17; Guth Decl. ¶ 16; Taylor Decl. ¶ 9 (noting that the budgets for the PWM Sales Business Unit do not apply to a large percentage of sales employees who are paid on a straight commission basis). Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 26 of 84 -18- professionals within each Business Unit essentially compete with each other for the finite dollars available in that Unit’s bonus pool, but they do not compete for dollars with professionals in any other Business Unit. (See Heller Dep. I at 85:22-86:2, 111:2-5; Larson Dep. at 204:4-15.) Each of the Divisions within Goldman Sachs has a compensation committee that reviews individual managers’ compensation recommendations. (Banerjee Dep. at 224:6-12; Kung Dep. I at 40:17-42:6; Heller Dep. I at 57:9-58:1, 58:8-60:13, 113:14-114:2.) These committees may question the Business Unit leaders to understand the considerations that drove their decisions. The compensation committees typically do not involve themselves in individual compensation decisions, and to the extent that individuals are discussed at all, those discussions are generally focused on only the most highly compensated professionals. (Kung Dep. I at 86:17-87:4; Larson Dep. at 96:18-22; Heller Dep. I at 26:7-13.) The most common reason for a change in a manager’s recommended compensation is that either more or less money is allocated to the Business Unit by the Division as the fiscal year draws to a close. (Heller Dep. I at 23:11-25:20; Larson Dep. at 62:11-63:23; Kung Dep. I at 112:3-115:4.) If those allocations change, individual Business Unit managers, who are in the best position to assess the relative contributions of their professionals, again apportion the allocations. (Heller Dep. I at 30:8-31:3.)12 This iterative process, the so-called “rounds,” occurs within each Division, not across them, and changes made at successive rounds have been favorable—not adverse—to women. (See Ward Report at 71-72.) The Firm does not direct compensation in any formulaic way based on a professional’s 360 reviews or manager quartile. (Heller Dep. I at 158:4-159:10; 170:1-13; 12 During this “rounds” process, managers may also make recommendations on individual compensation decisions if there has been a change in performance toward the end of the year. (Heller Dep. I at 28:5-29:3.) Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 27 of 84 -19- Larson Dep. at 142:22-143:4; Kung Dep. I at 227:7-22). Instead, managers make compensation decisions based on their overall assessments of professionals’ contributions, particularly to the Business Unit’s financial performance. (Kung Dep. I at 80:13-25.) Across the Firm’s many Business Units, individual managers consider varying factors in setting compensation for professionals. For example, in the Commodities Business Unit within the FICC Sub-Division of the Securities Division, traders’ compensation is tied directly to their profit and loss statistics. (Casturo Decl. ¶ 15.) By contrast, in the Principal Strategic Investment Business Unit within the Equities Sub-Division of the Securities Division, Business Unit managers do not analyze profit and loss on an individual level in determining compensation. Instead, they assess whether a professional has shown creative thinking, delivered critical market structure solutions, and contributed to improvements in the Firm’s infrastructure. (D. Cohen Decl. ¶ 13.) In the PWM Sub-Division of IMD, certain professionals, called Private Wealth Advisors (“PWAs”), are paid entirely on commission. Non-PWAs, who do not receive commissions, are compensated through base salaries that can be augmented by “supplemental compensation” designated by the PWAs on a discretionary basis from the revenues they have generated. (Taylor Decl. ¶ 9; ¶ 11.) In IBD, compensation decisions often turn heavily on qualitative considerations relating to the skill sets required of professionals performing different roles, which vary across and within Business Units. For example, some managers focus on a professional’s analytical, modeling, and accounting skills. (See Guth Decl. ¶ 20.) For other managers, market judgment and risk assessment are particularly important. (Packer Decl. ¶ 19.) In the Public Sector & Infrastructure (“PSI”) Business Unit, as another example, the professional’s understanding of Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 28 of 84 -20- applicable tax rules and attention to legal and compliance issues bear heavily in compensation decisions. (See Benz Decl. ¶ 22.) 3. Promotion The process for promoting Goldman Sachs professionals from Vice President to Extended Managing Director (“EMD”) 13 involves hundreds of different nominators and reviewers each year. (Heller Dep. I at 204:4-14; Larson Dep. at 227:5-11, 229:6-21, 230:18- 231:2.) Through this decentralized process, any Managing Director in the Firm can nominate any Vice President for promotion to EMD. (See Heller Dep. I at 232:7-8, 233:6-9; Kung Dep. II at 432:16-20; Larson Dep. at 300:19-301:7.) With the assistance of professionals in HCM, leaders in each Division narrow the nominations list to a manageable number of candidates. (Heller Dep. I at 205:18-23, 234:1-3; Larson Dep. at 227:15-21; Kung Dep. II at 419:5-24.) Each Division then engages in a vetting process known as “cross-ruffing.” (See Larson Dep. at 240:15-21; Heller Dep. I at 213:25-214:8; Kung Dep. II at 398:25-399:5.) One member of a cross-ruffing team is designated to obtain feedback on each candidate, and that team member conducts between 10 and 15 interviews to determine the candidate’s readiness for promotion. (See Heller Dep. I at 218:8-19, 224:11-16; Kung Dep. II at 434:15-20, 441:17-20; Ex. 11B at GS0109329.) The cross-ruffing team then confers and ranks its list of the candidates. (See Heller Dep. I at 226:11-16; Kung Dep. II at 448:25-449:3.) The Divisional leaders receive the cross-ruffing team’s ranked list and focus on which business areas need more senior professionals, more than on individual assessments of candidates. (See Heller Dep. I at 230:13- 21, 231:3-7; Larson Dep. at 247:6-15.) Without making decisions about individual candidates, 13 Managing Director is the highest level title at the Firm. Within that category are Participating Managing Directors, who are more senior, and EMDs. (See Heller Dep. I 203:19- 204:3; Kung Dep. II at 397:15-398:10.) Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 29 of 84 -21- the Management Committee then approves, in conjunction with the Executive Office,14 how many EMDs to promote each year, taking into account the Firm’s business and geographical needs. (See Larson Dep. at 249:20-250:4; Heller Dep. I at 229:24-230:3; Kung Dep. II at 452:19-453:23, 454:17-455:2; Boyle Dep. at 82:16-18 (“Does the management committee make decisions about promotions? A. No.”), 89:1-14).) In short, as with compensation, the Firm’s promotion decisions are decentralized, with decisions made through a broad-based, bottom-up process bearing no resemblance to a top- down, centralized scheme. C. Proceedings in This Case Since the start of this litigation, Plaintiffs have shifted positions to try to find a theory supporting the certification of a massive Title VII class of women professionals performing varying functions in 140 Business Units within Goldman Sachs. But after two complaints and extensive discovery, Plaintiffs simply cannot do so. 1. Pleadings, Motion Practice and Discovery The centerpiece of Plaintiffs’ original complaint, filed in September 2010, was the charge that Goldman Sachs permitted excessive manager discretion in setting compensation and in promoting Vice Presidents to EMD. For example, Plaintiffs alleged that individual Goldman Sachs managers “have a wide amount of discretion in setting an employee’s compensation” and “[m]anagers are simply given a total compensation pool for their team of employees and allowed to distribute it as they see fit.” (Class Action Complaint filed Sep. 16, 2010, Dkt. 5 ¶ 48.) Similarly, Plaintiffs alleged “Goldman Sachs managers enjoy extremely broad discretion in choosing who to nominate for promotion. . . . Its managers select individuals based on personal 14 See Heller Dep. I at 242: 5-11, Larson Dep. at 280:18-25. Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 30 of 84 -22- preferences, personal relationships, and subjective and biased views of their aptitude and performances.” (Id. ¶ 53.) In June 2011, after the Supreme Court in Wal-Mart rejected the notion that a policy of manager discretion could support class certification, Plaintiffs changed direction to try to save their putative class action. Thus, in August 2011, Plaintiffs amended their complaint to mask their core allegation—excessive individual manager discretion—by emphasizing the 360 review process and manager quartiling. In its January 19, 2012 Order denying as premature Goldman Sachs’ motion to strike the Amended Complaint’s class allegations, the Court stated that Plaintiffs had identified three processes they alleged were uniform and discriminatory: (1) the “360-degree review process”; (2) “forced quartile-ranking of employees”; and (3) the “tap on the shoulder” promotion process. (January 19 Order at 13.)15 The Court concluded that it could not determine “from the pleadings alone the extent to which any of these practices contribute to a common mode of exercising discretion by Goldman Sachs’ managers.” (Id. at 13.) Since February 2011, Plaintiffs have conducted extensive discovery, including deposing more than 20 Goldman Sachs executives with knowledge of each of the challenged processes and obtaining more than 350,000 pages of documents and enormous banks of data about hundreds of thousands of individual employment decisions. At great expense, Goldman Sachs reviewed millions of emails. In their Amended Complaint, Plaintiffs sought to represent women professionals from all four Goldman Sachs’ revenue-producing Divisions. Now, however, they seek to represent professionals in only three Divisions, even though the Firm employs the same 15 In their motion for class certification, through silence, Plaintiffs have seemingly abandoned their claim that Goldman Sachs has an arbitrary “tap on the shoulder” promotion process. In fact, as shown supra at 20-21, the Firm’s process is broad-based and decentralized. Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 31 of 84 -23- challenged procedural frameworks in its Merchant Banking Division. (See, e.g., Donovan Dep. at 282:13-15; 421:4-8; 431:24-432:10.) Plaintiffs’ recognition that these processes have no actionable effect in the Merchant Banking Division confirms that they are not applied uniformly, do not have a uniform impact, or both. 2. Plaintiffs’ Experts To try to sustain their proposed class, Plaintiffs rely heavily on the testimony of Dr. Wayne Cascio, their industrial psychologist, and their statistician, Dr. Henry Farber. a. Dr. Cascio As Goldman Sachs has shown in its June 13, 2014 Daubert motion, Dr. Cascio’s opinions about Goldman Sachs’ performance evaluation and compensation processes are not relevant here, because he cannot causally connect those processes to gender disparities. Specifically, Dr. Cascio opined that Goldman Sachs’s processes “lack key safeguards to ensure procedural justice,” but he does not claim that those alleged “deficiencies” adversely affect women professionals: Q. Are you offering an expert opinion that the absence of any one of those items would have — would result in gender bias or gender differences? A. I am not. (See Cascio Dep. at 181:19-182:4.) Although Dr. Cascio claims that Goldman Sachs employees should, but do not, have access to their appraisal results (which is itself inaccurate), he expressly acknowledged that he draws no causal connection between the access to appraisal results and any gender differences in outcomes in evaluation or compensation: Q. Do you have an expert opinion to offer in this case, that failing to give employees access to review their appraisal results has a disparate impact on women? Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 32 of 84 -24- … A. I do not. (See Id. at 149:5-10.) And, although claiming that Goldman Sachs does not maintain “thorough and consistent documentation” in employment processes, Dr. Cascio admitted that he draws no causal connection between this purported flaw and gender differences: Q. Do you have an expert opinion to offer in this case that failing to provide thorough and consistent documentation across raters, that includes specific examples of performance based on personal knowledge, has a disparate impact on women? . . . A. I can’t say that that caused any disparate impact, that’s for sure. (See Id. at 149:11-19.) b. Dr. Farber The many flaws in Dr. Farber’s statistical analyses are detailed in Goldman Sachs’ Daubert motion to strike his report. Summarized below are some of the more glaring errors. 1. Dr. Farber’s failure to account for compensation differences among Business Units and professionals in those Units. To begin with, Dr. Farber groups all Goldman Sachs professionals together as if they are paid from a single compensation pool. (Farber Dep. at 134:11-135:7.) As shown, supra, however, each Business Unit has a separate compensation pool, and pools available to one Business Unit have no effect on compensation in another. Although recognizing that a professional’s Business Unit could explain compensation disparities, Dr. Farber “didn’t control for department, desk or business unit” in any of his studies. (Id. at Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 33 of 84 -25- 103:22-23.) Indeed, Dr. Farber testified: “I’m not sure I understand what you mean by business unit.” (Id. at 228:13-14.) Moreover, Dr. Farber did nothing to account for the impact on compensation of the many different functions performed by Goldman Sachs professionals across 140 Business Units—for example, the Firm’s rewarding of traders based on the profitability of trades they execute, or Ph.D. computer scientists based on their success in improving the way a particular Business Unit reports on and reacts to market developments. Like any sophisticated business, Goldman Sachs pays more for some positions than for others, in response to market and other factors, and evaluates its professionals in different positions based on their different skills. Dr. Farber purported to account for the undeniable differences in the myriad of professional positions within 140 Goldman Sachs’ Business Units by relying on Affirmative Action Plan (“AAP”) position group codes. But these codes have no bearing on the Firm’s compensation, performance evaluation, or promotion decisions. (Farber Dep. at 138:3-9 (“Q: And you’re not aware of any testimony in the record or any documents in the record that suggests that affirmative action plan job groups are used or connected in any way to compensation setting at Goldman Sachs, are you? A: That’s correct.”).) These codes are generic occupational classifications used for mandatory governmental reporting purposes. In fact, these codes are so broad that the three most populous job groups cover about 70% of the entire population that Dr. Farber studied. (See Declaration of Cathy Obradovich” (“Obradovich Del.)” ¶ 6.) Tellingly, Dr. Farber admitted that he never actually understood what these “AAP job groups” meant at Goldman Sachs. (Farber Dep. at 110:5-11 (“I never really had a satisfactory— I never really—never managed to find a satisfactory description of what the job groups actually were. They were numeric codes, but I don’t—I couldn’t tell you what any code was.”).) Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 34 of 84 -26- For many of the professionals whom Dr. Farber studied, “production”—for example, trades made, sales closed, or the value of deals worked on—is the single most significant determinant of compensation. (Heller Dep. I at 74:25-78:4; 137:9-138:4.) But Dr. Farber ignored production data, testifying that he would have been willing to use it only if such data was available “across all workers” in the putative class: “[t]he population as a whole [is what] I'm interested in, I can’t just have a productivity measure for some subset of them.” (Farber Dep. at 122:16-18.) But this view is akin to ignoring available productivity data for workers on a manufacturing floor, because no similar body of data exists for executives or support personnel. Dr. Farber also ignored compensation differences among professionals promoted internally and those hired laterally. When Goldman Sachs hires laterally, the Firm negotiates compensation differently than for incumbent “home-grown” professionals at that same level. (Declaration of Kirstine Wilson (“Wilson Decl.”) ¶¶ 6-7.) The differential treatment of lateral and home-grown professionals is common in the securities industry. (See Expert Report of Michael P. Curran, dated December 11, 2013 ¶ 28.) Because Dr. Farber ignores the factors critical to understanding compensation at Goldman Sachs—including the Business Unit allocation process, the work individual professionals perform, and the quality of their work—his models are so incomplete that they cannot explain compensation at all. This shortcoming is illustrated by applying his models to the compensation of male professionals alone. If his models were sound, they would explain why particular male Vice Presidents received different levels of compensation. But they do not; applying Dr. Farber’s models can explain only 42% of the variation among male Vice Presidents. (See Ex. 30 (Ward Backup Data).) Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 35 of 84 -27- 2. Dr. Farber’s own statistics confirm that women professionals are not disfavored in performance assessments or compensation in the overwhelming majority of Business Units. Dr. Michael Ward, the expert statistician Goldman Sachs retained, followed Dr. Farber’s methodology, but simply organized the results by Business Unit. Dr. Ward’s organization of Dr. Farber’s results by Business Unit shows that there is no pattern of discrimination against Goldman Sachs women professionals in the three challenged processes: compensation—differences in pay between men and women are statistically insignificant or statistically significant and favorable to women in 90% of the Business Units for Associates and in 76% of Business Units for Vice Presidents; 360 review scores—differences in scores are either statistically insignificant or statistically significant and favorable to women in 87% of the Business Units for Associates and in 83% of Business Units for Vice Presidents; and manager quartile—differences in quartile placement are statistically insignificant between men and women or statistically significant and favorable to women in 83% of the Business Units for Associates and in 84% of Business Units for Vice Presidents. (See Ward Report, Appendix A (Charts 1, 2 and 4).)16 These results answer the question that Plaintiffs have avoided—whether there is any statistical basis for concluding that there is pattern of discrimination across the putative class. 16 These percentages are derived from Dr. Ward’s charts cited above. Plaintiffs complain about the “small sample sizes” and the purported large margins of errors in Dr. Ward’s studies. But, just taking Dr. Farber’s own results by Business Unit and then simply counting up the Units in which the pay differences are favorable to women—without doing any statistical analysis or examining statistical significance—shows that 42% of the Business Units are favorable to women. (See Ward Report at 5-6.) Moreover, that some of the Business Units may be small does not excuse ignoring them altogether, as Dr. Farber does, in a statistical analysis purporting to establish commonality. Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 36 of 84 -28- The answer to that question—based on Dr. Farber’s own models—is a resounding “no.” Indeed, the same lack of commonality shown at the Business Unit level for compensation repeats itself in Dr. Farber’s analyses of the 360 process and manager quartile placement across the putative class. (See Ward Report at 7-8; Appendix A (Charts 2 and 4).)17 Dr. Ward confirmed the absence of a common pattern of discrimination in compensation across Business Units by preparing compensation studies that, unlike Dr. Farber’s analyses, accounted for relevant variables, such as Business Unit, and the professional’s function and performance. (See Ward Report, Figures 56-60.) 3. Dr. Farber’s flawed analysis of promotion rates of Goldman Sachs professionals. Dr. Farber makes three key errors in studying promotions. First, he does not account for performance, even though superior performance is the most likely reason a professional will be promoted. For example, if Dr. Farber had taken into account manager quartile results, his reported results would not have been statistically adverse for women. (See Ward Report at 29.) Second, Dr. Farber misunderstood the Firm’s promotion process. He calculated the number of promotions men received in each year from 2004 to 200818 and then estimated the number of additional female promotions needed to reach parity. In doing so, he assumed incorrectly that Goldman Sachs does not, like any large business, limit the overall number of 17 In these statistics, the unit of observation is a person-year, i.e., a year of employment for one employee. 18 Plaintiffs erroneously claim that Dr. Farber should not study 2009, because “there were no promotions that year.” (Pl. Br. 12 n.46). In fact, Goldman Sachs changed its fiscal year-end in 2009 from November 30 to December 30, and the 2009 promotions were made effective in January 2010. Plaintiffs offer no explanation at all for Dr, Farber’s failure to study 2010 and 2011. Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 37 of 84 -29- promotions in any given year to take into account financial and other considerations. (See Farber Dep. at 270:17-273:16.) Third, Dr. Farber’s method for studying promotions was inconsistent with his testimony about the “appropriate” method to use. He testified that his preferred method for analyzing a binary, yes/no process like promotions was a “probit model.” (Farber Report ¶ 60 (probit model “appropriate for cases where the outcome variable takes on two discrete values.”).) Notably, if Dr. Farber used the same probit model as he did when studying the other binary decisions, he would have reported no significant adverse results for women. Instead, he elected to use a model that generated a shortfall, but he offered no explanation for his switch in methodologies. In any event, even applying his flawed method, Dr. Farber’s statistics show a “shortfall” of only 19 women out of 384 Vice-Presidents that were promoted to EMD over his five-year period. By contrast, Dr. Ward examined promotions across the three Divisions for the entire putative class period (that is, without cherry picking certain years as Dr. Farber did) and he separately analyzed lateral and non-lateral hires.19 Overall, for all Divisions, Dr. Ward found no pattern of promotion adverse to women. (See Ward Report, Table 10.) 4. Basic statistical tests confirm that Dr. Farber’s analyses are inadequate and unreliable. Statisticians rely on a standard test—the “Chow” test—to determine whether the factors affecting outcomes in a personnel process are sufficiently similar for two (or more) seemingly disparate groups that those factors can be considered through a single statistical model. 19 The factors included in Dr. Ward’s regression model include fiscal year, years as Vice President, manager quartile in the current and prior year, production in the current and prior year, 360 score, and education. (Ward Report at 52-53.) Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 38 of 84 -30- Dr. Farber ignored this universally accepted test.20 Dr. Ward performed the Chow test on Dr. Farber’s model, and the test rejected the notion that professionals in all Divisions could be combined in the same aggregated model to test the challenged Goldman Sachs processes—360 review, manager quartiling, compensation and promotion—including combining lateral and non- lateral hires in the same model. (Ward Report at 11-12.) Moreover, standard statistical tests confirm that Dr. Farber’s analyses have little or no ability to explain the compensation of male or female Goldman Sachs professionals. The standard measure of the explanatory power of a statistical model is the “R-squared” of the model. A model incorporating important variables affecting outcomes will have a much higher “R- squared” than a study that excludes those variables. Dr. Farber’s preferred compensation studies for Associates show an R-squared of 0.33, meaning that only 33% of the variation in Associate compensation, among both men and women, is explained by the variables in his model. For Vice Presidents, his model has an R-squared of only 0.23.21 (Ward Report at 20.) By contrast, Dr. Ward’s compensation analyses, which do not show significant differences between male and female professionals, have R-squared results exceeding .70. These values confirm the very low explanatory power of Dr. Farber’s models. See Griffin v. Bd. of Regents of Regency Univ., 795 F.2d 1281, 1291 (7th Cir. 1986) (R-squared of .45 too small to support an inference of discrimination). 20 See Farber Dep. at 135:18-137:21. 21 The same is true of Dr. Farber’s 360 models; the R-squareds for Dr. Farber’s 360 models for Vice Presidents are 27% and 43% for his pre-and post-2010 models, respectively, and for Associates, the R-squareds for his models are 34% and 45%, respectively. (See Exs. 30 and 31 (Backup produced by Dr. Ward and Dr. Farber).) Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 39 of 84 -31- LEGAL STANDARD To prosecute a class action, Plaintiffs must meet all of the prerequisites of Federal Rule of Civil Procedure 23(a)—numerosity, commonality, typicality, and adequacy—and, in addition, at least one of the requirements of Rule 23(b). That is a high bar, because class actions are “an exception to the usual rule that litigation is conducted by and on behalf of the individual named parties only.” Comcast Corp. v. Behrend, 133 S. Ct. 1426, 1432 (2013) (internal quotation marks omitted). Rule 23 thus demands that its requirements be satisfied by “evidentiary proof,” and that this proof be tested “rigorous[ly].” Id. As a result, the “stringent requirements” that Rule 23 imposes “in practice exclude most claims.” Am. Express Co. v. Italian Colors Rest., 133 S. Ct. 2304, 2310 (2013). In the particular context of employment discrimination claims, the Supreme Court has recognized that “there is a wide gap between (a) an individual’s claim that [she] has been denied a promotion [or higher pay] on discriminatory grounds . . . and (b) the existence of a class of persons who have suffered the same injury as that individual.” Wal-Mart, 131 S. Ct. at 2553. Plaintiffs incorrectly suggest that they can bridge that “wide gap” merely by posing a question common to the class. (Pl. Br. 34) To the contrary, “plaintiffs . . . must actually prove—not simply plead—that their proposed class satisfies each requirement of Rule 23.” Halliburton, 2014 WL 2807181, at *12; Comcast Corp., 133 S. Ct. at 1432 (“[E]videntiary proof” must be sufficient to convince the Court, “after a rigorous analysis, that the prerequisites of Rule 23 have been satisfied.”); Levitt v. J.P. Morgan Sec., Inc., 710 F.3d 454, 464-65 (2d Cir. 2013) (internal quotation marks omitted) (“[T]he district court is required to make a definitive assessment of Rule 23 requirements, notwithstanding their overlap with merits issues, and must resolve material factual disputes relevant to each Rule 23 requirement.”) (internal quotation marks omitted). Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 40 of 84 -32- ARGUMENT Plaintiffs here seek to challenge the way in which more than 2,300 female Vice Presidents and Associates across 140 different Goldman Sachs’ Business Units were evaluated, compensated, or promoted over twelve years. After extensive discovery and numerous statistical and other expert reports, Plaintiffs have not shown that individual Goldman Sachs managers uniformly applied the Firm’s unremarkable procedural frameworks for assessing performance— 360 reviews and manager quartiling—to discriminate against women professionals. Indeed, Plaintiffs’ own evidence shows the opposite: there are no statistically significant differences in pay or performance among male and female professionals in three-quarters of Goldman Sachs’ Business Units. (See Ward Report at 2, 5, 6; Ward Report, Appendix A.) In the remaining one-quarter of Goldman Sachs’ businesses, Plaintiffs’ own statistical analyses show that women sometimes do better and men sometimes do better. Because Plaintiffs have not satisfied Rule 23(a)’s prerequisites of commonality, typicality, or adequacy, nor satisfied any of Rule 23(b)’s additional requirements, this Court should deny certification of what would be an unprecedented class. I. PLAINTIFFS CANNOT SATISFY EACH OF RULE 23(a)’S REQUIREMENTS. A. Plaintiffs Have Not Shown the Existence of a Question Common to the Class. Rule 23(a)(2) requires Plaintiffs to show that “there are questions of law or fact common to the class,” but the rule “is easy to misread, since any competently crafted class complaint literally raises common questions.” Wal-Mart, 131 S. Ct. at 2550-51 (internal quotation marks and brackets omitted). Certification is appropriate only if the claims of all class members “depend upon a common contention—for example, the assertion of discriminatory bias on the part of the same supervisor.” Id. In addition, “[t]hat common contention . . . must be of such a nature that it is capable of classwide resolution—which means that determination of its Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 41 of 84 -33- truth or falsity will resolve an issue that is central to the validity of each one of the claims in one stroke.” Id. at 2551. In Wal-Mart, the Supreme Court identified two ways in which Plaintiffs might prove the existence of a question amenable to classwide resolution in this action: (i) for a disparate treatment claim, by showing a “general policy of discrimination” against women professionals, and (ii) for a disparate impact claim, by demonstrating a “testing procedure or other companywide evaluation” that operated uniformly and adversely on that group of professionals. Id. at 2553. As explained below, Plaintiffs have shown neither. Here, as in Wal- Mart, “it will be impossible to say that examination of all the class members’ claims for relief will produce a common answer to the crucial question why was I disfavored.” Id. at 2552 (emphasis in original). Plaintiffs’ evidence—both statistical and anecdotal—does not withstand the rigorous analysis that Rule 23 requires. Far from having a general policy of discrimination, Goldman Sachs has a policy of antidiscrimination that the Firm has expended great effort to implement and to enforce. And extensive discovery confirms that the challenged employment practices, far from being uniform, are decentralized, discretionary, and highly variable. Courts in the Second Circuit have regularly refused to certify classes with these characteristics.22 22 See Cuevas v. Citizens Fin. Grp., Inc., 526 F. App’x 19, 20 (2d Cir. 2013) (summary order) (reversing certification of class asserting claims under New York Labor Law); Saleem v. Corporate Transp. Grp., Ltd., No. 12 Civ. 8450 (JMF), 2013 WL 6061340, *4-5 (S.D.N.Y. Nov. 15, 2013) (denying certification where propriety of classification of the class members turned on individualized inquiries into degree of control employer exercised over those employees); Fernandez v. Wells Fargo Bank, N.A., Nos. 12 Civ. 7193, 12 Civ. 7194 (PKC), 2013 WL 4540521, *5-6 (S.D.N.Y. Aug. 28, 2013) (denying certification for lack of commonality where plaintiffs were unable to identify a common policy); Shayler v. Midtown Investigations, Ltd., No. 12 Civ. 4685 (KBF), 2013 WL 772818, *7-9 (S.D.N.Y. Feb. 27, 2013) (denying certification of class under New York Labor Law where, among other things, individualized inquiries would predominate over common questions); Oakley v. Verizon Commc’ns Inc., 09 CIV. 9175 CM, Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 42 of 84 -34- Other federal courts also have denied certification in similar circumstances. In Tabor v. Hilti, Inc., 703 F.3d 1206 (2013), for instance, the Tenth Circuit declined to certify a class of female sales representatives at a tool manufacturing company, because plaintiffs, like Plaintiffs here, were challenging “a highly discretionary policy for granting promotions.” Id. at 1229. Even if that policy caused “an overall disparate impact on women,” the court reasoned, “Plaintiffs ha[d] not shown that the facts and circumstances involved in [the defendant’s] promotion choices [were] common across the class of female employees.” Id. Likewise in Davis v. Cintas Corp., 717 F.3d 476 (2013), the Sixth Circuit declined to certify a class of female applicants for sales representative positions, because plaintiffs, again as in this case, “could not show that a number of women, who failed to obtain employment at many places, over a long time, under a largely subjective hiring system, shared a common question of law or fact.” Id. at 489; see Jones v. Nat’l Council of YMCA, No. 09-C- 06437, 2014 WL 1305153, at *6 (N.D. Ill. Mar. 31, 2014) (“A company-wide policy that vests managers with the discretion to evaluate employees based on subjective criteria cannot be the commonality predicate for a class action,” even where the policy required a “forced” numerical grade). Here, Plaintiffs seek to certify a class that is orders of magnitude more complex than the classes rejected in Tabor, Davis, and Jones. Indeed, Plaintiffs’ proposed class in this case bears some resemblance to the class proposed in Bennett v. Nucor Corp., 656 F.3d 802 (8th Cir. 2011), where the Eighth Circuit refused to certify a class of all African-American employees at a single steel manufacturing plant. On an extensive record (like the one in this case), the court 2012 WL 335657, *9 (S.D.N.Y. Feb. 1, 2012) (denying certification of class bringing claims under federal Family and Medical Leave Act, because of individualized questions of causation and damages). Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 43 of 84 -35- of appeals cited “stark inter-departmental variations in job titles, functions performed, and equipment used.” Id. at 815. Just like Dr. Farber, plaintiffs’ expert conducted a “statistical analysis that did not differentiate between the hiring and promotion decisions made in each department.” Id. The Eighth Circuit deemed that “bottom-line analysis . . . insufficient to demonstrate that any disparate treatment or disparate impact present in one department was also common to all the others.” Id. at 815-16. The court similarly rejected plaintiffs’ anecdotal evidence of a hostile work environment, because their allegations about their own department “[did] little to advance a claim of commonality across the entire plant.” Id. at 816. Plaintiffs have simply recycled the same flawed theory from Bennett but expanded it massively to cover 140 businesses across the country. Under this settled law, Plaintiffs cannot avoid their burden to provide “significant proof” of their claims and cannot establish commonality merely by offering a classwide statistical analysis. Plaintiffs say that “whatever the failings of the class’s statistical analysis,” those failings “affect every class member’s claims uniformly.” (Pl. Br. 35 (quoting Stockwell v. City and Cnty. of San Francisco, 749 F.3d 1107 (9th Cir. 2014)).) But the Ninth Circuit recognized in Stockwell that “affirmative evidence” of commonality must be subjected to “rigorous analysis.” 749 F.3d at 1111 (quoting Wal-Mart, 131 S. Ct. at 2551). As explained above, three other federal courts of appeals have subjected similar claims to such analysis and found them insufficient for Rule 23 purposes. And, the Supreme Court reiterated just last week in Halliburton Co. v. Erica P. John Fund, Inc., ___ S. Ct. ___, No. 13-317, 2014 WL 2807181, at *12 (June 23, 2014), that “plaintiffs wishing to proceed through a class action must actually prove—not simply plead—that their proposed class satisfies each requirement of Rule 23.” Id. (second emphasis added). Thus, Plaintiffs’ failure to offer sound evidentiary proof, by statistical Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 44 of 84 -36- evidence or otherwise, means that they may not invoke the class action device; it does not merely postpone their day of reckoning to trial. 1. Plaintiffs Have Not Shown a General Policy of Discrimination. Plaintiffs “provide no convincing proof of a companywide discriminatory pay [or] promotion policy, [and thus] they have not established the existence of any common question” with respect to their disparate treatment claim. Wal-Mart, 131 S. Ct. at 2556-57. Plaintiffs do not dispute that Goldman Sachs has a widely disseminated policy that strictly forbids any discrimination against employees on the basis of sex or other protected characteristics. The record here is replete with evidence of the Firm’s past and current efforts to hire, retain, and promote women professionals. These facts alone severely undercut Plaintiffs’ assertion that Goldman Sachs has engaged in a pattern or practice of discrimination against women.23 Moreover, Plaintiffs have not substantiated their conclusory claim that Goldman Sachs’ anti-discrimination mandate exists side-by-side with an unspoken policy of discrimination against women. (See Pl. Br. 38-40.) Without identifying the precise group that supposedly makes these centralized decisions, Plaintiffs speculate that the unnamed wrongdoers have made discrimination against women “standard operating procedure” at Goldman Sachs. (Pl. Br. 40.) Plaintiffs urge that such a general policy of discrimination also may be inferred from their expert’s statistical analyses and anecdotal evidence of gender bias. (See Pl. Br. 39-40.) But Plaintiffs’ purported evidence does not withstand scrutiny. 23 See, e.g., United States v. City of New York, Nos. 11-5113, 12-0491, 2013 WL 1955782, *9 (2d Cir. 2013) (emphasizing that “non-statistical evidence, including an employer’s affirmative action efforts, are ‘both relevant to and probative of absence of intent to discriminate’”) (quoting 1 Lex Larson, Employment Discrimination § 9.03[2][c], at 9–20 (2d ed. 2011) (footnote omitted)); Serrano v. Cintas Corp., Nos. 04-40132, 06-12311, 2009 WL 910702, at *6-7 (E.D. Mich. Mar. 31, 2009) (rejecting argument that employer maintained a “white male business culture” when employer “presented evidence of sincere attempts to achieve greater diversity in its company”). Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 45 of 84 -37- a. The Record Establishes that Pay and Promotion Decisions Are Decentralized. Plaintiffs’ notion of top-down decision-making is a complete fiction. The record shows that hundreds of managers (men and women) who oversee the Firm’s 140 disparate Business Units, make the challenged compensation and promotion decisions. See supra at 12-21. Ignoring the record evidence, Plaintiffs instead speculate that various groups might be the source of centralized compensation decisions, but that is demonstrably incorrect. First, Plaintiffs point to the fact that “a central Management Committee, composed of only 34 of Goldman’s top partners, oversees the entire firm.” (Pl. Br. 3.) But every company of any significant size in America—including Wal-Mart—has upper management and thus could be claimed to be “run from the top down.” (Pl. Br. 3.) What matters for Rule 23 and commonality purposes is whether upper management made the decisions of which the putative class members complain. Here, the undisputed record shows that the Firm’s Management Committee reviews aggregate compensation numbers, to ensure that the Divisions have operated within their budgets. (Mehling Dep. at 86:21-87:4.) It plays no role in individual compensation decisions for Associates or Vice Presidents. (See Boyle Dep. at 82:13-15 (“Q: Does the management committee make decisions about compensation? A: No.”); Mehling Dep. at 86:6-25.) Plaintiffs’ contention that compensation at Goldman Sachs is “overseen by . . . the Chief Financial Officer” (Pl. Br. 9) is misleading. The Firm’s CFO plays no role at all in determining individual compensation. Indeed, as Plaintiffs’ evidence (Pl. Br. 9 n.31) itself demonstrates, the CFO communicates to Division heads only their per capita compensation targets, which each Division then allocates among its Business Units. The CFO is not involved in how the Divisions allocate their funds among individual Business Units, let alone how any Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 46 of 84 -38- particular manager compensates Vice Presidents and Associates. (See, e.g., Mehling Dep. at 58:6-60:10.) Moreover, divisional compensation committees do not determine Vice President and Associate pay, and to the extent the committees discuss individuals at all, their focus is only on the highest compensated individuals, without regard to gender or any other factor. (See Kung Dep. I at 84:25-85:8, 86:17-87:4; Larson Dep. at 96:18-22; Heller Dep. I at 26:7-13.)24 To be sure, the Firm implements the final pay decisions, but the decisions themselves are made by the managers of individual Business Units, not any senior officer or committee.25 Promotions tell the same story as compensation. Plaintiffs speculate about the role of “the firmwide executive office” and the “Management Committee” (Pl. Br. 11), but neither of those groups makes the promotion decisions at issue in this case. As in other large companies, central management ratifies decisions made by others. (See Boyle Dep. 82:16-18 (“Q: Does the management committee make decisions about promotions? A: No.”), 89:1-14; Kung Dep. II at 453:7-23.) As described supra at 20-21 , committees of cross-ruffers interview and rank Vice Presidents for promotion. The head of each Division reviews the Division’s cross- ruffing list, but the focus is on which geographical areas and businesses are in need of additional 24 Plaintiffs mischaracterize the role of “the firm compensation team.” (Pl. Br. 10.) As the record shows, within HCM, there is a compensation team that provides logistical support to the revenue Divisions. For instance, the compensation team runs reports that summarize how Business Units are allocating compensation. (See Mehling Dep. at 59:13-21.) But the compensation team does not have any authority over the Divisions’ allocation of budgets to Business Units or Business Units’ subsequent decisions about individual compensation. (See id. at 65:10-22; Larson Dep. at 63:15-23.) 25 To the extent that individual managers’ initial recommendations on compensation were modified in a subsequent round, such changes generally helped women Vice Presidents and Associates. Dr. Ward examined the changes between the initial recommendation and the final compensation decision in the Securities Division between 2006 and 2011 for male and female Associates and Vice Presidents. (See Ward Report at 71-73.) He found that women professionals benefitted from changes between “rounds” of compensation decisions for four out of the six years he studied. (See id.) Plaintiffs have not taken issue with Dr. Ward’s conclusions. Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 47 of 84 -39- leadership, rather than on individual candidate assessments. (See Heller Dep. I at 231:3-7; Larson Dep. at 246:20-247:23.) The Management Committee and Executive Office similarly review each Division’s list for geographic and functional distribution and may make minor adjustments to the lists. (Heller Dep. I at 229:24-230:3.) 26 In the face of this record, Plaintiffs’ reliance (Pl. Br. 11, 39) on Ellis v. Costco Wholesale Corp., 285 F.R.D. 492 (N.D. Cal. 2012), could not be more misplaced. In Ellis, a small group of senior managers—including the CEO—was personally involved in meeting the candidates and making the challenged promotion decisions: “Top management’s involvement in the promotion process is . . . consistent, and pervasive, class-wide.” Id. at 511-512. Here, even after extensive discovery, Plaintiffs have not pointed to anything remotely akin to that top-down approach. This Court made clear in ruling on Defendants’ motion to strike the class allegations that Plaintiffs would have to prove that Goldman Sachs “operated under a general policy of discrimination” to certify a disparate treatment class. (January 19 Order at 22.) In response, Plaintiffs have pointed to nothing more than the small degree of rational centralized oversight and guidance present in any large business and routine human resources administration. b. Plaintiffs’ Expert Evidence Is Either Irrelevant or Circular. Plaintiffs next contend that this Court should infer the existence of a general policy of discrimination because, according to their paid experts, there are “flaws” in Goldman Sachs’ performance evaluation processes and significant disparities in “performance, pay, and 26 Plaintiffs also refer to the “firmwide group that creates and distributes the firmwide cross- ruffing manual and trains all interviewers” (Pl. Br. 11), but HCM merely provides administrative support to the Divisions. For example, it suggests people to serve on the cross-ruffing teams, keeps track of prior participants on the teams, and compiles information on the candidates for promotion. (See Kung Dep. II at 399:6-24, 401:12-402:1; Larson Dep. at 235:13-20.) In short, the talent assessment group—as with HCM more generally—assists with administrative functions related to promotion, but has no authority over the promotion decisions themselves. Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 48 of 84 -40- promotion rates for female Associates and Vice Presidents.” (Pl. Br. 39.) That evidence, however—testimony by Dr. Wayne Cascio, Plaintiffs’ industrial psychologist, and statistics prepared by Dr. Henry Farber—shows no indication, let alone proof, of a general policy of gender discrimination. i. Dr. Cascio’s Irrelevant Testimony Dr. Cascio criticizes certain aspects of the Firm’s 360 review and manager quartiling processes, but he offers no evidence linking those supposed defects to any alleged gender biases. Indeed, Dr. Cascio concedes that he cannot draw any causal connection between the details of the Firm’s evaluation processes and gender disparities in pay or promotion.27 Needless to say, Dr. Cascio does not even attempt to take the further step of saying that any flaws in the 360 review and manager quartiling processes were intended to harm women professionals as part of a firmwide general policy of discrimination. In fact, Dr. Cascio’s criticisms all center on the fact that hundreds of Goldman Sachs’ managers were not uniform in the way that they evaluated professionals across the Firm during the class period: “There is no requirement that the composition of [360] reviewer [panels] be uniform for similarly situated employees.” (Cascio Report ¶ 35(b).) “[T]he weights of variables that inform Quartile and compensation decisions can vary.” (Cascio Report ¶ 27.) “[T]here is no requirement that raters be consistent in their application of factors.” (Cascio Report ¶ 27.) 27 Dr. Cascio candidly admits that he has no opinion on whether the handful of “defects of omission” he claims to have identified can be linked to any perceived gender disparities in outcomes. (Cascio Dep. at 181:19-182:4 (“Q: Are you offering an expert opinion that the absence of any one of those items would have — would result in gender bias or gender differences? A: I am not.”).) Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 49 of 84 -41- ‘“Potential’ as used by Goldman Sachs [as a criterion in the manager quartiling process] is not a reliable, valid performance measurement [because] [a]ccording to the deposition testimony of divisional corporate representatives in the divisions at issue in the class, the assessment of potential varied considerably.” (Cascio Report ¶ 88.) Looking at their own expert testimony, Plaintiffs cannot escape the Supreme Court’s holding in Wal-Mart: “allowing discretion by local supervisors over employment matters . . . is just the opposite of a uniform employment practice that would provide the commonality needed for a class action; it is a policy against having uniform employment practices.” 131 S. Ct. at 2554. ii. Dr. Farber’s Flawed Statistics The Second Circuit has made clear that the power of a statistical model turns “on how accurately the model mimics the actual factors influencing the dependent variable.” Sobel v. Yeshiva Univ., 839 F.2d 18, 22 (2d Cir. 1988); see also Bickerstaff v. Vassar Coll., 196 F.3d 435, 449 (2d Cir. 1999) (statistician’s “starting point [should have been] to understand [and account for the fact] that salary increases [are usually] based upon merit”). Only by “identifying [the] legitimate criteria that affect the decision making process [can a plaintiffs’ statistician] make predictions about what job or job benefits . . . similarly situated employees should ideally receive, and then . . . measure the difference between the predicted treatment and the actual treatment of those employees.” Ottaviani v. State Univ. of N.Y. at New Paltz, 875 F.2d 365 (2d Cir. 1989). 28 Here, as shown supra at 24-30, and in the Daubert motion, Dr. Farber’s studies assume what they seek to prove. Dr. Farber lumped together all of the professionals—from 28 Dr. Farber testified that he does not use “employer controlled” variables in his regressions because they “might” be “tainted,” but he admitted that he did nothing to determine whether any of the variables he omitted were, in fact, “tainted.” (See Farber Dep. at 177:21-178:10 (“I did not do any [taint] analyses”)); Ottaviani, 875 F.2d at 372-73 (plaintiffs were required to show that academic rank was a “tainted” variable before omitting it from analysis); see also Morgan v. UPS, 380 F.3d 459, 470 (8th Cir. 2004) (court rejected plaintiffs’ expert’s failure to consider prior pay in his regression model, because there was no evidence that past pay disparities were, in fact, discriminatory and tainted). Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 50 of 84 -42- traders to computer scientists—across the Firm’s 140 separate businesses. See, e.g., Randall v. Rolls-Royce Corp., 637 F.3d 818, 822-23 (7th Cir. 2011) (rejecting statistician’s analyses that failed to account for “differences in the jobs performed”).29 He did not even attempt to consider whether there were common patterns of results among those professionals across Divisions or Business Units. See Farber Dep. at 103:22-23 (“I didn’t control for department, desk or business unit.”). Having impermissibly aggregated the data across the putative class, Dr. Farber then speculated that unexplained classwide disparities in compensation and promotion were caused by gender discrimination. (See Farber Report ¶¶ 16-17.) As the Supreme Court has held, “information about disparities at the [Firm level] does not establish the existence of disparities at individual [Business Units], let alone raise the inference that a company-wide policy of discrimination is implemented by discretionary decisions at the [Business Unit] level.” Wal-Mart, 131 S. Ct. at 2555 (brackets and internal quotation marks omitted). 29 See also Davis v. Cintas Corp., 717 F.3d 476 (6th Cir. 2013) (no commonality based on plaintiffs’ aggregated statistics; “although some Cintas locations under-hired women and racial minorities, other locations over-hired women and racial minorities during the same period”); Bolden vs. Walsh, 688 F.3d 893, 896 (7th Cir. 2012) (rejecting aggregated statistical evidence; Plaintiffs’ expert . . . assumed that the appropriate unit of analysis is all of Walsh’s Chicago-area sites. He did not try to demonstrate that proposition.”); Abram v. UPS, 200 F.R.D. 424, 431 (E.D. Wisc. 2001) (“plaintiffs’ somewhat tendentious reliance on aggregate data illustrates the perils and misuses of statistical analysis. If Microsoft-founder Bill Gates and nine monks are together in a room, it is accurate to say that on average the people in the room are extremely well-to-do, but this kind of aggregate analysis obscures the fact that 90% of the people in the room have taken a vow of poverty”); Stagi v. National R.R. Passenger Corp. 391 Fed.Appx. 133 (3d Cir. 2010) (quoting Abram with approval); Penk v. Oregon State Board of Higher Education, Civil Action No. 80-436, 1985 WL 25631 (D. Or. Feb. 13, 1985) (“aggregating statistical results over all institutions in order to prove that a pattern or practice of discrimination exists at each institution “puts the cart before the horse”); Lott v. Westinghouse Savannah River Co., 200 F.R.D. 539, 560-61 (D.S.C. 2000) (“[a]ggregated statistics . . . are of scant utility to the court in conducting a certification analysis”); Moore v. The Boeing Co., No. 4:02CV80 CDP, 2004 WL 3202777, at *12 (E.D. Mo. Mar. 31, 2004) (“[a]lthough the aggregation of all the job groups within a site may suggest a pay disparity favoring men, within many job groups there is not a statistically significant disparity, and in some groups, women are actually favored”). Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 51 of 84 -43- Moreover, as shown supra at 27-28, taking Dr. Farber’s results as they are (without any correction for his methodological errors), but calculating their results by Business Unit, there plainly is no general and common pattern of discrimination against women professionals at Goldman Sachs. Goldman Sachs’ expert, Dr. Michael Ward, performed an analysis that accounted for the relevant variables, including the differences among Business Units. See supra at 29-30. Dr. Ward found that there is no statistical basis for concluding that there is any common pattern of results across the putative class of Goldman Sachs’ female Vice Presidents and Associates. (Ward Report at 38-41, 55-71, 75, (Figures 8-11).)30 This Court, however, does not have to resolve a battle of experts. It is enough that Plaintiffs’ statistical evidence is all directed at the firmwide level rather than at the Business Unit level, where compensation and performance decisions are actually made. As the Supreme Court put it in Wal-Mart, an alleged firmwide disparity “cannot by itself establish the uniform, [unit-by-unit] disparity upon which the plaintiffs’ theory of commonality depends.” 131 S. Ct. at 2555. c. Plaintiffs’ Anecdotal Evidence Is Misleading and Inadequate. In their Amended Complaint, Plaintiffs conclusorily allege that senior officials at Goldman Sachs purposefully designed the Firm’s compensation, evaluation, and promotions policies to discriminate against women professionals. (See Am. Compl. ¶58.) This assertion is 30 Dr. Farber’s failure to account for production—the most directly relevant measure of performance for roughly one-third of the class—renders his models meaningless. See, e.g., Puffer v. Allstate Ins. Co., 255 F.R.D. 450, 465 (N.D. Ill. 2009) (plaintiffs’ regression analysis was “only minimally probative” on commonality because it failed to account for measures of skill or productivity, including actual measures of the quantity or quality of an employee’s production); Smith v. Va. Commonwealth Univ., 84 F.3d 672, 676-77 (4th Cir. 1996) (“common sense require[s] that any multiple regression analysis used to determine pay disparity must include all the major factors on which pay is determined”); Penk v. Oregon State Bd. of Higher Educ., 816 F.2d 458, 465 (9th Cir. 1987) (multiple regression analysis not probative where “highly determinative quality and productivity factors” were missing). Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 52 of 84 -44- baseless, and Plaintiffs’ own expert admitted that nothing in the design of the Firm’s policies indicates such a malevolent intent. (See Cascio Dep. at 194:15-195:3.) As a result, Plaintiffs have shifted gears, claiming that the Firm’s senior executives knew about but ignored a widespread and pervasive culture of gender bias. (Pl. Br. 40) This claim too is factually baseless and, even if proved, would not show that executives wanted the problem to persist, let alone that they had intentionally implemented a general policy of discrimination.31 See, e.g., Personnel Adm’r of Mass v. Feeney, 442 U.S. 256, 279 (1979) (Disparate treatment “implies that the decisionmaker . . . selected or reaffirmed a particular course of action at least in part ‘because of,’ not merely ‘in spite of,’ its adverse effects upon an identifiable group.”); U.S. v. City of New York, 717 F.3d 72, 93-95 (2d Cir. 2013) (failure of City officials to “validate” exam, even after knowing of statistical disparities, is not evidence sufficient to “support a reasonable inference that they declined to act because they wanted to discriminate”).32 31 Plaintiffs try to goad Goldman Sachs into waiving the attorney-client privilege, by speculating that the Firm knew of differences but took inadequate steps, in a “secret” process, to study them. (Pl. Br. 30.) As the Court knows, the process conducted by the Employment Law Group at Goldman Sachs over the years provided privileged legal advice to its clients, and the content of those consultations and advice is, quite rightly, not in the record. See, e.g., Order dated June 18, 2013 (Dkt. No. 183). Plaintiffs’ insinuations are purposely misleading and ignore the Court’s careful analysis of that process during discovery in the case. 32 Yet another misrepresentation is Plaintiffs’ statement that “Goldman [has] itself . . . acknowledge[d] the 360’s overall unreliability [and admitted] that the 360 scores are . . . decoupled from performance in such a way that they cannot be relied on.” (Pl. Br. 5-6.) The document Plaintiffs cite says no such thing. It does not anywhere say that 360 reviews “cannot be relied on”; rather, it expressly instructs managers to rely on them. The author of the message—Goldman Sachs’s global head of HCM—urges managers to consider both numerical and narrative comments. Managers, she urged, should “interpret[] [the 360 scores for themselves] giv[ing] appropriate weight to the review commentaries, objective metrics, and [the manager’s] own first-hand observations.” (Pl. Br. 6 n.15 (citing GS143793 at 794).) Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 53 of 84 -45- i. Goldman Sachs Has Made Extensive Efforts To Achieve a Diverse Workplace. Perhaps the most troubling part of Plaintiffs’ brief is its distortion of documents concerning Goldman Sachs’ training and diversity initiatives. For example, Plaintiffs assert that “in 2005 the Firm cautioned new associates in their orientation that while clients will ask to go to strip clubs, they should merely not ‘expense’ that entertainment.” (Pl. Br. 25.) What that document actually tells new Associates is that they may not entertain at such places at all: “You may not entertain colleagues or clients for business purposes (1) at establishments which restrict use of their facilities on the basis of race, gender, religion or other similar inappropriate factors; or (2) with sexually oriented entertainment, such as entertainment at topless bars or lap-dancing bars.” (Ex. 25 at GS0177838 (emphasis added).) The guide then instructs Associates to decline any client suggestion that they attend such establishments. (Id.)33 Goldman Sachs has conducted diversity training programs for over 20 years, dating back to 1990. (Hendricks Decl. ¶ 4.) Every Goldman Sachs employee is required to complete training on diversity and harassment, and employees take an additional course addressing diversity when they become managers. Nonetheless, Plaintiffs misleadingly assert that although “employees are supposedly required to complete two hours of diversity training per 33 Plaintiffs’ emphasis on supposed “strip club activity” is flatly contradicted by numerous accounts from women who have worked at the Firm for years. (See Sabat Decl. ¶ 9 (“I have never felt pressured to go to a strip club.” 13-year employee); Robertson Decl. ¶ 12 (“I cannot recall hearing the possibility of going to a strip club being discussed” 15-year employee); Davidson Decl. ¶ 16 (same; 10-year employee); Creedon Decl. ¶ 12 (same; 7-year employee); Bohm Decl. Decl. ¶ 15 (6-year employee); Chester Decl. ¶ 8 (6-year employee); Eckhard Decl. ¶ 8 (6-year employee); Criqui Decl. ¶ 10 (4-year employee); see also Scherrer Decl. ¶ 10 (in her 18 years with Goldman Sachs, only once did someone—a client in the 1990s—mention a strip club; she reported it to her manager who helped her address the situation). And while Plaintiffs cite their declarant, Denise Shelley, Shelley Decl. ¶ 10, she admitted that her statement in this regard was based on a “general understanding”; that she could not remember the source of her understanding; and that she was not aware of even one instance in which anyone took clients to strip clubs. (See Shelley Dep. at 118:8-119:4.) Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 54 of 84 -46- year, internal audits suggest that . . . between one half and three quarters of managers attended no diversity training at all between 2002 and 2007.” (Pl. Br. 30.) But the particular two-hour diversity training requirement Plaintiffs refer to was not instituted until 2007. In every year since, between 96% and 99% of the Firm’s employees have met this requirement. (Declaration of Aime Hendricks (“Hendricks Decl.”) ¶ 7.) Moreover, to the extent that training or other diversity materials say anything about potential issues, they evidence Goldman Sachs’ commitment to solving those issues. These materials were created by diversity staff as part of manager training or were developed to generate discussion within the Firm’s committee devoted to diversity. 34 They constitute “precisely the type of evaluative and analytical exercise in which the public has a strong interest in encouraging corporations to engage.” Trezza v. The Hartford Inc., No. 98 Civ. 2205, 1999 WL 511673, at *2-3 (S.D.N.Y. July 20, 1999). Without that willingness for self-examination, the Firm—and indeed any company—would be less successful in developing its diversity programs. Far from being indifferent to women, the Firm has developed programs and networks to help its women professionals succeed and advance at every stage of their careers.35 Female professionals who have succeeded at the Firm view these programs as helpful to their and other women’s success. 36 In addition, the Firm offers extensive benefits for expectant and working 34 See Hendricks Decl. ¶¶ 19-20 (describing the missions of the Firm’s Office of Global Leadership and Diversity and the Americas Diversity Committee). 35 An example is the Women’s Career Strategies Initiative (“WCSI”), an annual six-month program designed to assist female Associates with a successful transition to Vice President. WCSI helps participants develop their leadership profile; and it provides tools and guidance on career management, work/life balance, and how to use feedback from the Firm’s performance review process. (See Hendricks Decl. ¶ 21 Ex. 5 (attaching GS0112929 (2011 WCSI kickoff presentation).) 36 See, e.g. Robertson Decl. ¶11 (“[T]he Securities Women’s Network provides women across the Division with invaluable opportunities to learn more about commercial topics relevant to our businesses and to network with clients, other women and senior leaders in the Division.”); Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 55 of 84 -47- mothers, putting the lie to Plaintiffs’ hearsay accusations of animus toward pregnant women. Indeed, one of the critical junctures for many women is when they contemplate return from maternity leave or time off for parenting, 37 and Goldman Sachs has pioneered innovative programs to reintegrate women returning to work.38 Publications have recognized Goldman Sachs’ efforts, naming it among the “100 Best Companies to Work For.”39 ii. Plaintiffs’ Few Conclusory Declarations Are Weak, Vague, and Legally Insufficient. In Wal-Mart, the plaintiffs submitted “some 120 affidavits reporting experiences of discrimination—about 1 for every 12,500 class members—relating to only 235 out of Wal- Mart’s 3,400 stores.” 131 S. Ct. at 2556. The Supreme Court deemed that anecdotal evidence “too weak to raise any inference that all the individual, discretionary personnel decisions are discriminatory.” Id. “Even if every single one of these accounts [were] true,” the Court reasoned, “that would not demonstrate that the entire company ‘operate[s] under a general policy of discrimination.’” Id. (quoting General Telephone Co. of Southwest v. Falcon, 457 U.S. 147, 159 n.15 (1982)). The court contrasted the thin factual record in Wal-Mart with the level of Chester Decl. ¶ 5 (“I was selected by Division leadership to participate in the [WCSI] program, and I believe it was very helpful to my career development.”); Criqui Decl. ¶ 9 (“[T]he Securities Division Women’s Network provides important mentoring opportunities. Because of my own experience as a mother of three, I have volunteered to serve as a mentor through the Network to women in the Division who are about to or who recently had a baby. I will meet with my mentees to share my experiences and provide advice on how to succeed as a working mother, and will also be available whenever they have questions.”). 37 Hendricks Decl. ¶¶ 22, 26, Ex. 8-10 (attaching Affirmative Action Program for Minorities and Women (2010) (GS0114543). The Firm provides four months’ paid maternity leave. It also provides Parenting Leave of two weeks for non-primary caregivers and four weeks for primary caregivers. The Firm provides additional leave for events such as family emergencies. (See id. ¶ 22.) 38 Hendricks Decl. ¶ 25, Ex. 7 (attaching GS0225871 (2011 Returnship program); GS0114535 (2010 AAP narrative)). 39 Hendricks Decl. ¶¶ 27-32 (citing descriptions of these awards in 2008, 2010, and 2011 AAP narratives). Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 56 of 84 -48- anecdotal evidence put forward in Teamsters v. United States, 431 U.S. 324 (1977), in which plaintiffs’ 40 accounts of discrimination “represented roughly one account for every eight members of the class.” Wal-Mart, 131 S. Ct. at 2556 (internal quotation omitted). Here, Plaintiffs have put forward just six declarants other than themselves out of more than 2,300 women professionals during the class period. Of these, four are not even members of the putative class. See, e.g., Sheehan v. Purolator, Inc., 839 F.2d 99, 102-03 (2d Cir. 1988) (complaints made by employees who are not members of the proposed class have “little probative value”). As in Wal-Mart, even if this handful of accounts were credible, it would not demonstrate that Goldman Sachs as an entire organization “operate[s] under a general policy of discrimination,” which is what Plaintiffs “must show to certify a companywide class.” Wal-Mart, 131 S. Ct. at 2556. Nor do these declarations provide any probative or persuasive evidence. They are largely boilerplate statements that mimic Plaintiffs’ allegation of a “boys’ club” culture or atmosphere. (See Pl. Br. 20 (“Goldman maintains a ‘boy’s club’ atmosphere.”) (capitalization and emphasis omitted).) Courts have routinely held these types of vague and unspecific accusations to be inadmissible and unhelpful.40 Not a single declarant points to any specific evidence of systematic bias in evaluation, compensation, or promotion. To the extent that these few declarations have any specific content at all, they only highlight the dissimilarities in the experiences of women in different positions, within different businesses, working under different managers, at different 40 See, e.g., Mattenson v. Baxter Healthcare Corp., 438 F.3d 763, 770 (7th Cir. 2006) (holding that “testimony about a work-place ‘culture’” was “too vague” to be admissible); Puffer v. Allstate Ins. Co., 255 F.R.D. 450, 467 (N.D. Ill. 2009) (“While some of these declarants allege generally that Allstate has a ‘male-dominated’ or ‘boys’ club’ culture, these allegations, unaccompanied by a time frame or specific discriminatory act, are too vague to be helpful.”), aff'd, 675 F.3d 709 (7th Cir. 2012). Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 57 of 84 -49- times. See Wal-Mart, 131 S. Ct. at 2551 (“Dissimilarities within the proposed class are what have the potential to impede the generation of common answers.”) (internal quotation marks and citation omitted). For example: Lisa Parisi was promoted to EMD two years after joining the Firm. (Parisi Decl. ¶ 3.) She directs her discrimination claims against her female manager. (Parisi Dep. at 40:23- 41:15, 47:10-15.) She then speculates that women professionals are unwilling to complain to the Employee Relations Group, but admits having no knowledge of any woman who had suffered retaliation. (Parisi Dep. at 274:20- 275:8, 277:9-12.) Leslie Baggett, a summer associate at Goldman Sachs for about 8 weeks during the summer of 2005, was not subject to the compensation or promotion processes challenged in this action; her compensation was lockstep with other summer associates. (Baggett Decl. ¶ 3.) Allison Gamba (Gamba Decl. ¶ 3) was nominated for Managing Director in 2008 and 2009. She claims that she should have been nominated again in 2010, when her Business Unit was undergoing a massive downsizing, and even though no one from her Unit—male or female—was nominated that year. (Gamba Dep. at 103:17-21, 105:3-106:11.) Denise Shelley was promoted to Vice President within 18 months of joining the Firm. (Shelley Decl. ¶ 3.) She testified that she had no complaints about several of her male managers and acknowledged receiving fair feedback from the 360 reviews. (Shelley Dep. at 26:7-19, 65:5-13, 67:16-68:13, 76:6-77:21, 97:10-20.) Putting aside the veracity of these allegations, the point is that they are not united by any common contention amenable to classwide proof or resolution. To the contrary, these anecdotes reflect highly individualized accounts of alleged discrimination by different people, confirming that here there can be no “common answer to the crucial question why was I disfavored.” Wal-Mart, 131 S. Ct. at 2552.41 41 See Davis v. Cintas Corp., 717 F.3d 476, 489 (6th Cir. 2013) (“Even assuming that these statements constituted particular instances of discrimination, commonality is not satisfied; on the contrary, these statements illustrate that the circumstances of discrimination are highly individualized and cannot be adequately treated in a generalized class action setting.”); Randall v. Rolls-Royce Corp., No. 1:06-cv-860, 2010 WL 987484, at *8 (S.D. Ind. Mar. 12, 2010) (“[R]ather than support a common nucleus of claims, these declarations actually provide anecdotal evidence of distinct differences, as opposed to similarities, among each of the Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 58 of 84 -50- Finally, Plaintiffs’ handful of anecdotes pales in comparison to the accounts of many women, provided in the declarations submitted with this brief, who detail the course of their careers and illustrate that Goldman Sachs is a meritocracy where professionals with the requisite ability, drive and commitment rise to the top. These accounts also squarely contradict Plaintiffs’ attempt to denigrate the Firm’s support of women with pregnancies.42 The Second Circuit has emphasized that “the quality of any anecdotal evidence or the absence of such evidence” must be evaluated. Sheehan, 839 F.2d at 102. Here, it is notable that so few women have stepped forward with any specific accounts of gender bias, and that so many women have stepped forward with highly detailed accounts of the Firm’s support for women. employee’s circumstances.”); Puffer, 255 F.R.D. at 467 (“[N]on-class member declarations do not support plaintiff’s argument for commonality” and “highlight the differences and uniqueness in the individual plaintiffs’ claims,” because “[t]hese individuals worked in different departments within Allstate Protection, had different supervisors, and allegedly experienced different forms of discrimination.”). 42 See Robertson Decl. ¶ 5 (“I have two children and have taken maternity leave twice while working at the Firm. In fact, I was promoted to PMD when I was pregnant with my first child. Additionally, when I was eight months pregnant with my second child, I was promoted to be Global Co-head of FX Sales.”); Criqui Decl. ¶¶ 4-6 (“I have three children and have taken one maternity leave while at Goldman Sachs”; describing flexible schedule, support from managers and continued expanding role after maternity leave); Sabat Decl. ¶ 5 (“[I]n 2014, after I had taken my two maternity leaves, my responsibilities increased to include leading the Investment Grade Credit desk in London [in addition to leading the desk in the United States].”); Davidson Decl. ¶ 5 (“I have received significant support from my managers and other colleagues at Goldman Sachs both before, during and after my maternity leave.”); Creedon Decl. ¶ 4 (“I learned of my promotion to Managing Director at the end of 2013, not long after my second maternity leave.”); Bohm Decl. ¶ 5 (“[My male manager] went out of his way to help me before I went on maternity leave and when I returned to work; that support made the transition much easier for me and facilitated my continued career success (including my later promotion to EMD).”); Guth Decl. ¶ 23 (“I was promoted to the position of PMD in 2002, less than 18 months after taking my first maternity leave.”); Benz Decl. ¶ 24 (“I feel that the Firm has supported all of my choices regarding work-family balance, and I have never met with resistance when I needed to take time off to care for my children or my parents.”). Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 59 of 84 -51- iii. Goldman Sachs’ Internal Complaint Record Shows a Culture of Compliance, Not of Bias. Plaintiffs try to bolster their spare declarations with selective quotations from portions of the Firm’s internal investigation files. All of this so-called evidence is inadmissible hearsay.43 (Pl. Br. 40.) Plaintiffs do not cite any court decision permitting unsworn hearsay allegations to be presented in a Title VII class action as competent evidence of the truth of those allegations. They do not brief the issue here, but in a letter brief to the Court seeking to unseal these materials, 44 Plaintiffs cited inapposite cases where, by contrast to Plaintiffs’ offer of hearsay, parties presented verified documents (like declarations) or submitted documents for purposes other than the truth of the matters asserted. 45 Plaintiffs’ reliance on newspaper 43 These materials also are excludable under Federal Rule of Evidence 403. Injecting into this proceeding a debate over the truth of various hearsay allegations and the adequacy of the Firm’s response, extending back 12 years, would represent just the sort of undue delay and wasted time Rule 403 was designed to avoid. 44 The parties submitted letter briefs concerning the application of the Protective Order to these materials, in the course of which they also briefed admissibility. The Court, in its June 10, 2014 Order, directed that “[t]he parties may address the issue of admissibility in their briefing” of the class certification motion. (June 10, 2014 Order at 2.) 45 See La Day v. Catalyst Tech., Inc., 302 F.3d 474, 481 n. 7 (5th Cir. 2002) (notes concerning plaintiff’s own complaint in individual discrimination case); Byrd v. Lynch, No. 10 Civ. 247, 2011 WL 2680572, at *6 (D.N.J. July 8, 2011) (declaration summarizing documents of plaintiff fired for sexual harassment were “offered not for their truth but to show that accusations of sexual harassment were made . . . and that Defendant decided to terminate Plaintiff as a result.”); Godoy v. Maplehurst Bakeries, Inc., 747 F. Supp. 2d 298, 305 (D.P.R. 2010) (memoranda at issue not “hearsay because the author is relying on her own personal recollection of events” and they “[do] not rely on another declarant’s statements.”); see also Levitt v. PricewaterhouseCoopers, LLP, No. 04 Civ. 5179, 2007 WL 2106309, *1 (S.D.N.Y. July 19, 2007) (declaration allowed, noting challenged entry “merely lists exhibits attached to the declaration”); Faulk v. Sears Roebuck & Co., No. 11 Civ. 2159, 2013 WL 1703378, at *6 n.5 (N.D. Cal. Apr. 19, 2013) (product liability case, document allowed on issue of predominance); In re Front Loading Washing Mach. Class Action Litig., No. 08 Civ. 51, 2013 WL 3466821, at *10 (D.N.J. July 10, 2013) (publications by other companies that “may or may not contain hearsay”); Guadiana v. State Farm Fire & Cas. Ins. Co., No. 07 Civ. 326, 2010 WL 5071069, at *5 (D. Ariz. Dec. 7, 2010) (insurance claim files included as part of a declaration); Cokely v. N.Y. Convention Ctr. Operating Corp., No. 00 Civ. 4637, 2004 WL 1152531, at *2 (S.D.N.Y. May 21, 2004) (filed grievances and charges, not internal complaints). Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 60 of 84 -52- accounts is similarly unavailing. See, e.g., Green v. Borg-Warner Protective Servs. Corp., Nos. 95 Civ. 10419(RPP), 96 Civ. 0038(RPP), 95 Civ. 5558(RPP), 96 Civ. 8740(RPP), 1998 WL 17719, at *3 (S.D.N.Y. Jan. 16, 1998) (denying motion for class certification) (“Plaintiffs’ counsel should be aware that news articles are of no evidentiary value to courts and that only on television shows or in newspapers are news articles cited as evidence.”); Mason v. Stock, 955 F. Supp. 1293, 1312 (D. Kan. 1997) (newspaper articles “have no evidentiary value”).46 Beyond their inadmissibility, Plaintiffs’ attempt to portray the Firm’s internal complaint process as evidence of a “culture of bias” turns reality on its head. That Goldman Sachs has invested in a professional Employee Relations staff and operated a robust internal complaint resolution process shows that the Firm is deeply committed—not indifferent—to equal opportunity and to an appropriate working environment. Plaintiffs repeatedly point out (Pl. Br. 2, 40) that, over the course of 10 years, out of a group of over 6,000 women in the Securities, Investment Management, and Investment Banking Divisions, there were 133 instances in which women in those Divisions spoke to Employee Relations Advisors about various situations. If anything, that fact reflects an environment in which women were able to raise concerns, not a culture of intimidation or retaliation.47 Indeed, the complaints averaged just 13 a year, which from a female workforce of approximately 6,000 professionals and support staff amounts to a 46 Plaintiffs also seek to expose copies of a few non-public EEOC charges they received in discovery. Plaintiffs do not dispute that they seek to disclose charges that were never disclosed by the individuals who made them. Plaintiffs cite Babbitt v. Albertson’s, Inc., No. 92 Civ. 1883, 1992 WL 605652, at *3-4 (N.D. Cal. 1992), for the proposition that 42 U.S.C. § 2000e-8(e) only prevents the EEOC itself from disclosing the contents of EEOC charges and does not apply to private litigation. That misses the point. In Babbitt, the defendant was objecting to producing the discovery documents to the plaintiff, not to disclosing them broadly to the public. Indeed, the Babbitt Court only ordered defendants to produce documents subject to a “well-crafted protective order.” Id. at *5. The materials at issue would also be excludable under Federal Rule of Evidence 403. 47 See, e.g., Scher Decl. ¶ 12, Robertson Decl. ¶ 12, and Eckerd Decl. ¶ 8. Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 61 of 84 Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 62 of 84 Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 63 of 84 -55- Because decisions about evaluation, compensation, and promotion are highly discretionary and decentralized throughout Goldman Sachs, Plaintiffs’ evidence—whether statistical or anecdotal—does not show that any practice had a uniform, adverse effect on women Associates and Vice Presidents in all 140 Business Units. Indeed, the evidence shows precisely the opposite: that the impact of the challenged practices varies widely across the proposed class. a. Plaintiffs’ Aggregate Statistical Evidence Cannot Show the Requisite Uniformity Across and Among Different Businesses. As Plaintiffs acknowledge, they must show that any practice they challenge operates in a “uniform” way that “systematically disadvantage[s] female Associates and Vice Presidents across the [c]lass.” (Pl. Br. 3, 12.) But Plaintiffs’ statistical expert did not even try to show that any particular process impacted women professionals within Business Units in the same way. He acknowledged that, by design, he was only interested in “the population as a whole”—i.e., all female Associates and Vice Presidents throughout the three relevant Divisions. (Farber Dep. at 122:14-18.) He was explicit that he “didn’t want to be involved in doing an analysis for [any] subgroup.” (Id. at 134:11-16; 165:12-15.) Plaintiffs have thus abandoned any effort to show that the processes at issue had the same effect in one Business Unit as in another. In Wal-Mart, the Supreme Court rejected exactly that type of lump-sum, aggregate analysis. It explained that even if plaintiffs’ “statistical proof” established “a pay or promotion pattern that differs from the nationwide figures or the regional figures in all of Wal- Mart’s 3,400 stores, that would still not demonstrate that commonality of issue exists.” 131 S. Ct. at 2555 (emphasis omitted). “A regional pay disparity, for example, may be attributable to only a small set of Wal-Mart stores, and cannot by itself establish the uniform, store-by-store disparity upon which plaintiffs’ theory of commonality depends.” Id. In order for Title VII liability to rest on a disparate-impact theory, the Court thus observed, “merely proving that [a] discretionary Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 64 of 84 -56- system has produced a racial or sexual disparity is not enough.” Id. (emphasis in original). Or, as the Court put it, “when a class of plaintiffs is sought to be certified,” “[m]erely showing that [an employer’s] policy of discretion has produced an overall sex-based disparity does not suffice.” Id. Here, all Plaintiffs have attempted is to show an overall, nationwide sex-based disparity across 140 separate Business Units. This case is a perfect example of why aggregate nationwide data does not suffice to establish commonality for purposes of Rule 23(a). Even if a particular manager within a given business employed 360 review data or his or her own quartile assessments in a manner that disparately impacted women, Plaintiffs have not remotely shown that holds true across the putative class. See Bolden v. Walsh, 688 F.3d 893, 896 (7th Cir. 2012) (“Plaintiffs’ expert . . . assumed that the appropriate unit of analysis is all of Walsh’s Chicago-area sites. He did not try to demonstrate that proposition. If Walsh had 25 superintendents, 5 of whom discriminated in awarding overtime, aggregate data would show that black workers did worse than white workers—but that result would not imply that all 25 superintendents behaved similarly, so it would not demonstrate commonality.”). Remarkably, Plaintiffs make no effort to explain how their nationwide statistical evidence across 140 Business Units is different from the nationwide evidence the Supreme Court rejected in Wal-Mart. The relevant section of their brief, which is less than two pages (Pl. Br. 37-38), cites only general propositions from Wal-Mart with no analysis. Instead, Plaintiffs rely on McReynolds v. Merrill Lynch, Pierce, Fenner & Smith, Inc., 672 F.3d 482 (7th Cir. 2012), but that case dealt with a single business, a single job classification (financial advisor in Merrill Lynch’s retail operation), and two specific and substantive nationwide policies (for forming teams of brokers and for distributing accounts). The plaintiffs in McReynolds did not have to Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 65 of 84 -57- show that those policies had a uniform impact across more than a hundred different businesses— because there was only one business; nor across hundreds of different jobs—because there was only one job; nor across different levels within the company—because there was only one level. Needless to say, the Seventh Circuit did not consider, let alone approve, the certification of classes that included professionals in all jobs at varying levels within all of Merrill Lynch’s divisions throughout the country. That type of certification would have presented the same aggregation problem as in Wal-Mart. In the end, Plaintiffs do not cite a single post-Wal-Mart case certifying a disparate impact class encompassing 140 separate businesses based on aggregate statistical evidence of the sort that they have put forward here.51 b. Plaintiffs’ Evidence Confirms the Extent of Decentralized Decision-Making in the Firm’s Evaluation, Compensation, and Promotion Processes. Far from showing uniformity, the record in this case actually shows that the reasons for evaluation, compensation, and promotion decisions vary dramatically across the putative class. The existence of a procedural framework does nothing to “produce a common answer to the crucial question why was I disfavored.” Wal-Mart, 131 S. Ct. at 2552 (emphasis in original). What Plaintiffs must show is a pattern of decision-making to the disadvantage of women across the 140 Business Units in the proposed class. Plaintiffs have not made that showing, nor could they in light of the discretionary and decentralized way in which professionals in different Goldman Sachs businesses are evaluated. See Cintas, 2009 WL 910702, at *4-5 (affirming district court’s denial of certification despite some “common . . . 51 Plaintiffs cite Stockwell v. City & County of San Francisco, No. 12-15070, 2014 U.S. App. Lexis 74056 (9th Cir. Apr. 24, 2014). In that case, police officers who had been eligible for promotion based on one examination claimed that the city’s decision instead to promote based on a second examination caused a disparate impact on older officers. Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 66 of 84 -58- hiring policies and procedures”; “[h]iring decisions are made for a diverse range of reasons and depend on widely differing circumstances at each facility, such as the available applicant pools and the hiring needs at a particular facility”).52 In fact, Wal-Mart had the same types of frameworks and guidelines as exist here: on remand from the Supreme Court, plaintiffs sought to certify a California-only class challenging the use of Wal-Mart’s Firmwide Compensation Guidelines as a basis for compensation decisions. These Guidelines listed numerous factors for managers to consider in making such decisions like those in the 360 reviews, including “leadership ability,” “interpersonal skills,” and “communication skills.” The district court rejected the argument that these criteria constituted a “specific employment practice,” reasoning that plaintiffs’ “repackaged discretion theory” amounted to nothing more than “challenging Wal-Mart’s practice of delegating discretion to local managers, which the Supreme Court specifically held was not a specific employment practice supplying a common question sufficient to certify a class.” Dukes v. Wal Mart Stores, Inc., No. CV 01–02252 CRB, 2013 WL 3993000 at *9-10 (N.D. Cal. Aug. 2, 2013) (emphasis in original). This Court should similarly reject Plaintiffs’ attempt to challenge Goldman Sachs’ framework for discretionary and decentralized decision making. 52 In Jones v. National Council of YMCA, No. 09 C 06437, 2014 WL 1305153 (N.D. Ill. Mar. 31, 2014), plaintiffs challenged the YMCA’s compensation, performance review, and promotion policies and sought to certify a class of African-American employees company-wide. Plaintiffs cited the YMCA’s “forced performance grading policy” as the specific company policy that they claimed satisfied commonality. Id. at *5. Relying on Wal-Mart, the court concluded that “a company-wide policy that vests managers with the discretion to evaluate employees based on subjective criteria cannot be the commonality predicate for a class action.” Id. at *6. Although the YMCA’s procedures were company-wide and required supervisors to assign numerical grades to employees, that procedural uniformity “did not make the process either less subjective or discretionary.” Id. at *7. Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 67 of 84 -59- i. Evaluation and Compensation Title VII applies to adverse employer actions. As numerous courts have made clear, performance evaluations like the 360 reviews and manager quartiles are not themselves adverse employment actions and, thus, cannot be the basis for a disparate-impact claim absent a demonstrable link to some term or condition of employment.53 It is not enough for Plaintiffs vaguely to assert that “gender-based pay differences” were “tied” to 360 reviews and manager quartiles. (Pl. Br. 9.) Here, Plaintiffs cannot show how alleged gender bias in those processes affected members of the putative class members commonly with respect to actual employment actions. The record shows that 360 reviews are not used in any formulaic way to determine what quartile an individual manager assigns to the professionals in his or her Business Unit. (See Exs. 22 at GS0109355, 23 at GS0109425.) Similarly, the Firm imposes no formula for determining compensation based on manager quartile results. (See Banerjee Dep. at 295:10- 16, Cohen ¶ 15 n.2; Lopez Decl. ¶ 10 n.2; Levene Decl. ¶ 18.) Simply put, individual managers vary widely as to how they weight 360 reviews and quartile results. Some managers view 360 reviews as primarily a developmental tool and do not give them significant weight for quartiling and compensation purposes; other managers see such results as useful, but more so for inwardly facing professionals than for those who focus on external client services. (See Lopez Decl. ¶ 10; McNamara Decl. ¶ 18 n.4; Russell Decl. ¶ 15 n.2.) Similarly, some managers do not refer 53 See, e.g., Siddiqi v. New York City Health & Hospitals Corp., 572 F. Supp. 2d 353, 367 (S.D.N.Y. 2008) (“[N]egative evaluations, standing alone without any accompanying adverse results, are not cognizable.”) (quoting Bennett v. Watson Wyatt & Co., 136 F. Supp. 2d 236, 247 (S.D.N.Y. 2001)); Valentine v. Standard & Poor’s, 50 F. Supp. 2d 262, 283 (S.D.N.Y. 1999), aff’d, 205 F.3d 1327 (2d Cir. 2000) (same); see also Trachtenberg v. Dep’t of Educ. of City of New York, 937 F. Supp. 2d 460, 469 (S.D.N.Y. 2013) (“[I]n the absence of negative consequences attaching to the alleged negative performance evaluations or other reprimands, courts in this district have repeatedly found no adverse employment action.”). Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 68 of 84 -60- directly to quartiling results in recommending compensation (they are not listed as a factor at all in “Junior Banker” compensation guidelines for IBD); other managers view quartile assignments as an integral part of their compensation recommendations. (See Larson Dep. at 68:19; 71:9- 73:15 and Ex. 11A; Guth Decl. ¶ 17) Here, far from proving “a common mode of exercising discretion that pervades the entire company,” Wal-Mart, 131 S. Ct. at 2554-55, Plaintiffs’ own evidence shows that managers do not exercise their discretion in a common and uniform way. In fact, Ms. Chen- Oster actually complains that managers have “unchecked discretion” in setting compensation. (Chen-Oster Dep. at 436:24-437:4). According to Ms. Chen-Oster: There’s so much information and he has complete discretion to pick out a comment that may be completely irrelevant and blow that out of proportion and weight that as more important than the actual production of the employee or to, you know, choose a metric that may be creativity where there’s no real way to gauge. (Id. at 476:17-25 (emphasis added).) In Ms. Chen-Oster’s view, the problem with the 360 review process is that managers have “unbridled discretion to choose metrics, to choose comments, to choose where those comments and metrics come from.” (Id. at 488:25-489:4 (emphasis added); see also Pl. Br. 10 (arguing that the firm uses “poorly defined factors in determining employee compensation”; “does not train managers with respect to a common understanding of how to evaluate these factors”; and “does not weight the factors”). This is the very antithesis of a common mode of exercising discretion. The Court cannot certify a disparate impact class here, because Plaintiffs have no evidence showing how some alleged gender bias in the 360 reviews and manager quartiles adversely affected putative class members commonly with respect to compensation. Tellingly, as shown supra at 24-26, Dr. Farber did not study how those procedural frameworks could have a classwide impact across 140 Business Units. In fact, Dr. Farber studiously avoided examining Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 69 of 84 -61- the highly individualized factors affecting compensation and promotion, including the professional’s position, his or her performance, the performance of the professional’s Business Unit, market demand for the professional’s services, and the discretionary judgments of individual managers. Similarly, Dr. Cascio testified repeatedly that he had not done any research to show that his complaints about the 360 review process and manager quartiling are linked to gender disparities. See supra at 23-24. By contrast, Dr. Ward did examine the factors affecting compensation. Dr. Ward found that while, not surprisingly, both manager quartile and 360 review results are correlated with pay—higher-rated employees typically make more money than otherwise similarly situated employees with lower ratings—neither 360 results nor manager quartiles have a uniform effect on compensation across the class: “[t]he impact of these variables is not the same across Divisions, levels and lateral status.” (Ward Report at 26.) In short, “[t]here is no statistical evidence to substantiate the claim that these processes have a common impact on pay across the putative class,” (id.) (emphasis added) much less that these factors uniformly affect women’s pay differently than men’s across the class. On this record, Plaintiffs simply have no evidence supporting their disparate- impact claim as to compensation. ii. Promotion Plaintiffs’ support for their disparate-impact claim as to promotions is similarly nonexistent. Their discussion of the Firm’s promotion procedures—which occupies roughly one page of their brief (Pl. Br. 10-12)—does not identify any particular aspect of those procedures that allegedly results in a disparate impact.54 As with compensation, Plaintiffs’ own allegations 54 Plaintiffs note, in passing, that individual employees have to be nominated to be considered and cannot “apply,” but there is no evidence in the record that women “commonly,” across the 140 Business Units at issue, have been unable to find a Managing Director willing to nominate them. (Pl. Br. 10-11.) Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 70 of 84 -62- about the promotion process, in fact, center on the lack of uniformity in the process: “While Goldman [Sachs] does not require Vice Presidents to attain any particular performance level to qualify for promotion, it does consider a variety of poorly defined and unreliable criteria, such as whether the Vice President is a ‘role model’ or ‘an effective coach.’” (Pl. Br. 11.) Again, that is the exact opposite of “a common mode of exercising discretion.” Wal-Mart, 131 S. Ct. at 2554. Plaintiffs’ statistician, Dr. Farber, offers no support for Plaintiffs’ disparate- impact challenge to the Firm’s promotion process. He carefully limited his analysis only to 2004 to 2008, and even then, he identified a supposed “shortfall” of only 19 women out of 384 Vice-Presidents who were promoted to EMD. (See supra at pages 28-29.) This alleged “overall sex-based disparity” is not only illusory, but also it tells this Court nothing about whether there was a common reason why more women were not promoted from one level to the next. Wal- Mart, 131 S. Ct. at 2556. Different managers may have used “sex-neutral, performance-based criteria for hiring and promotion that produce[d] no actionable disparity”; rewarded “various attributes that produce[d] disparate impact,” such as “scores on general aptitude tests or educational achievements”; or been “guilty of intentional discrimination that produce[d] a sex- based disparity.” Id. at 2554. Because Dr. Farber’s data sheds no light on that inquiry, Plaintiffs cannot “show that all the [putative class members’] Title VII claims will in fact depend on the answers to common questions.” Id. Simply put, no disparate impact class can be certified on this record. B. Plaintiffs Are Not Adequate Class Representatives and Their Claims Are Not Typical. In addition to failing to demonstrate any common question, Plaintiffs have not shown that they meet the typicality or adequacy requirements of Rule 23(a). Rule 23(a)(3) requires that “the claims or defenses of the representative parties [be] typical of the claims or Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 71 of 84 -63- defenses of the class,” meaning that “each class member’s claim arises from the same course of events.” Marisol A. v. Giuliani, 126 F.3d 372, 376 (2d Cir. 1997). Rule 23(a)(4) requires that “the named plaintiffs must possess the same interests and suffer the same injuries as the class members” to ensure that “the representative parties will fairly and adequately protect the interests of the class.” In re Literary Works in Elec. Databases Copyright Litig., 654 F.3d 242, 249 (2d Cir. 2011). “[F]actors that the Court should consider to determine whether the absent class is adequately represented” include “whether the proposed plaintiffs are credible,” “whether the proposed plaintiffs have adequate knowledge of the case and are actively involved,” and “whether the interest[s] of the proposed plaintiffs are in conflict with those of the remainder of the class.” Spagnola v. Chubb Corp., 264 F.R.D. 76, 95-96 (S.D.N.Y. 2010). Of particular relevance here, named plaintiffs are not adequate if their claims “involve[] problems that could become the focus of cross-examination and unique defenses at trial, to the detriment of the class.” Lapin v. Goldman Sachs & Co., 254 F.R.D. 168, 177 (S.D.N.Y. 2008). Such “unique defenses” could divert “the plaintiff[s’] attention from the case as a whole” and thus “unnecessarily prejudice the class’s probability of success.” Panzirer v. Wolf, No. 79 Civ. 3445, 1980 WL 1395, at *3, *5 (S.D.N.Y. Apr. 21, 1980). Both named Plaintiffs assert claims atypical of the putative class. a. Plaintiff Orlich Ms. Orlich was never subject to the compensation system she challenges. As a first-year Associate and recent business school graduate, Ms. Orlich was paid in lockstep with others of her tenure—both women and men—and, as she confirmed at her deposition, having been an Associate for about 16 months she was not a candidate for promotion to EMD. (See Orlich Dep. at 47: 14-18; 48:2-11; 85:11-14; 10:4-7.) Accordingly, Ms. Orlich cannot identify a Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 72 of 84 Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 73 of 84 Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 74 of 84 -66- Here, those additional proceedings necessarily would involve detailed examination of the particular circumstances surrounding the employment decisions made by different managers concerning different class members. Thus, there is no possibility that resolution of any common question would materially advance this litigation. Goldman Sachs would be entitled “to raise any individual affirmative defenses it may have, and to ‘demonstrate that the individual [class members were] denied an employment opportunity for lawful reasons.’” Id. (quoting Teamsters, 431 U.S. at 362). a. Liability Requires an Inherently Individualized Inquiry. Plaintiffs assert that the issue of “liability for [their] disparate impact and/or pattern-or-practice claims” predominates, “because it is determinative of every class member’s ability to move forward with the claims.” (Pl. Br. 45-46.) That is incorrect for at least two reasons. Most obviously, even assuming there were a question “determinative of every class member’s ability to move forward,” that would meet only Plaintiffs’ threshold burden under Rule 23(a)(2) to demonstrate a common question; it would do nothing to meet their burden under Rule 23(b)(3) to show that such common questions predominate. In any event, Plaintiffs are wrong that liability for their claims can be determined on a classwide basis. Success in proving a pattern or practice of discrimination would create only a presumption of discrimination, which Goldman Sachs would be entitled to rebut with respect to each individual class member. The putative class members worked in different jobs, at different levels of the Firm, for variable lengths of time, in different offices throughout the country, with hundreds of different supervisors (both male and female) making the challenged decisions. Goldman Sachs has the statutory right to introduce evidence regarding each class member’s individual job responsibilities, performance (a key component in making Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 75 of 84 -67- compensation and promotion decision), and anything else relevant to determining whether each class member was the victim of intentional discrimination. Wal-Mart, 131 S. Ct. at 2561. Moreover, to resolve Plaintiffs’ claims that 360 reviews and manager quartiling caused a disparate impact on compensation, the Court would have to hold individual trials to resolve whether and to what extent those processes affected each class member’s compensation during each year of the class period. See Teamsters v. United States, 431 U.S. 324, 360-62 (1977). In each trial, the parties would litigate such individualized issues as (i) each manager’s reliance on 360 reviews or quartile results in setting compensation for each class member; (ii) the particular class member’s performance; (iii) the performance of her Business Unit; and (iv) market demand for her services. Similarly, individualized inquiries would pervade any promotion claims. As a result, following any class trial, the matters remaining for individual trials inevitably would be virtually identical to those necessary if the class members had brought individual actions in the first place. b. Damages Require an Inherently Individualized Inquiry. Moreover, even if any liability were proven, there would be individualized evidence on damages with respect to each class member. Here, again, such determinations necessarily would involve the unique circumstances of each class member. Under Title VII, the employer “is entitled to individualized determinations of each employee’s eligibility for backpay.” Wal-Mart, 131 S. Ct. at 2560–61; see 42 U.S.C. § 2000e-5(g)(2)(A); United States v. New York, 276 F.R.D. 22, 47 (E.D.N.Y. 2011) (“[T]he determination of what particular forms of relief an individual claimant will receive” is an “obviously . . . individual question.”). Individualized determinations of damages, under either a theory of disparate treatment or disparate impact, inevitably would swamp any common questions that Plaintiffs have managed to identify. See Lienhart v. Dryvit Sys., Inc., 255 F.3d 138, 149 (4th Cir. 2001) (“The functional equivalent of a Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 76 of 84 -68- full-blown trial on damages causation for each putative class member would be required to determine to which individuals [the company] is liable as a result of the class suit’s result. We do not believe that such a class can meet the prerequisites of Rule 23(b)(3).”). Recognizing that individual damages determinations defeat certification here, Plaintiffs claim that they can use “a [statistical] model for estimating the aggregate backpay damages across the class.” (Pl. Br. 46) (emphasis added). That is clearly incorrect. That would be just the type of “Trial by Formula” barred by Rule 23 and the Rules Enabling Act, 28 U.S.C. § 2072(b). Wal-Mart, 131 S. Ct. at 2561. Plaintiffs cannot estimate some type of “average backpay award” and then apply that across the class “without further individualized proceedings.” Id. As the Supreme Court has put it, “a class cannot be certified on the premise that [Goldman Sachs] will not be entitled to litigate its statutory defenses to individual claims.” Id. If Plaintiffs succeeded in showing a pattern or practice of discrimination or a classwide disparate impact, and if liability were thereafter established with respect to any individual class member, that member would be entitled only to whatever individual damages she could prove. See, e.g., McLaughlin, 522 F.3d at 231 (“Roughly estimating the gross damages to the class as a whole and only subsequently allowing for the processing of individual claims would inevitably alter defendants’ substantive right to pay damages reflective of their actual liability.”). Plaintiffs say “the Supreme Court clarified in Dukes [that] ‘individualized monetary claims belong in Rule 23(b)(3).’” (Pl. Br. 47 (quoting Wal-Mart, 131 S. Ct. at 2558).) That quote is ripped out of context. The Supreme Court held only that individualized monetary claims may not be brought under Rule 23(b)(2). The Court’s point was that such claims belong, if at all, under Rule 23(b)(3); the Court nowhere suggested that individualized monetary claims always can be brought under Rule 23(b)(3), without regard to predominance. For that reason, Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 77 of 84 -69- several courts have ruled since Wal-Mart that Rule 23(b)(3) precludes certification of a large class seeking individual damages.56 Plaintiffs are thus wrong in contending that “[i]t is well-established that ‘individualized calculations of damages do not defeat the predominance requirement.’” (Pl. Br. 47 (quoting Schear v. Food Scope Am. Inc., 297 F.R.D. 114, 126 (S.D.N.Y 2014)).) Of course, individualized calculations do not necessarily defeat the predominance requirement. In certain cases individualized monetary claims may be pursued under Rule 23(b)(3), but this is not such a case. The cases Plaintiffs cite all involved circumstances in which damages could be readily calculated based on the existing record.57 No such easy or mechanical approach is even arguably possible here. The putative class members occupied different jobs, at different pay levels, at different times, and for varying periods. Finally, Plaintiffs’ reliance (Pl. Br. 47) on United States v. City of New York, 276 F.R.D. 22 (E.D.N.Y. 2011), is misplaced. That case involved an eligibility test for firefighters; candidates who passed the test were eligible for hire, but not all who passed would be hired. In the court’s view, aggregate damages could be awarded because it would be “impossible to determine exactly which non-hire victims would have received job offers and which delayed-hire 56 See, e.g., Espenscheid v. DirectSat USA, LLC, 705 F.3d 770, 773 (7th Cir. 2013) (affirming decertification of a class of more than 2000 satellite TV technicians who brought state law minimum wage and overtime clams because “determining damages would . . . require 2341 separate evidentiary hearings”); Wang v. Hearst Corp., 293 F.R.D. 489, 497 (S.D.N.Y. 2013) (Plaintiffs cannot show predominance when, “[t]o make liability determinations, the Court would have to evaluate . . . individualized evidence, in addition to individualized testimony” concerning the “content” of plaintiffs’ positions.”). 57 See Schear, 297 F.R.D. at 126 (claims were for tips allegedly withheld unlawfully, and damages “could be mechanically calculated with relative ease”) (internal quotation marks omitted); Gulino v. Bd. of Educ., No. 96-8414, 2013 WL 4647190, at *8 (S.D.N.Y. Aug. 29, 2013) (damages could be readily calculated “based on salary tables”); Leyva v. Medline Industries Inc., 716 F.3d 510, 514 (9th Cir. 2013) (defendant’s “computerized payroll and time- keeping database would enable the court to accurately calculate damages and related penalties for each claim”). Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 78 of 84 -70- victims would have been hired sooner in the absence of discrimination.” Id. at 44; see Easterling v. Connecticut Dep’t of Corr., 278 F.R.D. 41, 48 (D. Conn. 2011) (same). As the court was careful to explain, however, such damages are appropriate only in those cases where “the number of qualified class members exceeds the number of openings lost to the class through discrimination and identification of the individuals entitled to relief would drag the court into a quagmire of hypothetical judgments and result in mere guesswork.” City of New York, 276 F.R.D. at 44 n.11 (quoting Robinson v. Metro-North Commuter R.R. Co., 267 F.3d 147, 161 n.6 (2d Cir. 2001) (internal quotation marks omitted)). Plaintiffs offer no argument why individual calculation of damages would be impossible here, and it would not be. The damages of any class member who proves she was paid less or not promoted because of intentional discrimination, or disparately impacted by the 360 review or manager quartile processes, are fully capable of individual determination. 2. Plaintiffs Also Have Not Shown Superiority. Plaintiffs also have failed to satisfy the Rule 23(b)(3) requirement that a class action be “superior to other available methods for fairly and efficiently adjudicating the controversy.” Rule 23(b)(3)(A) directs that in making this determination a court should consider potential class members’ interest in having individual control of their own causes of action. Here, the putative class members are highly paid professionals who are capable of deciding for themselves whether they have a claim they wish to pursue individually. Indeed, over 95% of the putative class members were paid more than annually, and a significant number of them were paid far more. Rogers Decl. ¶ 8. More than three years after the initial filing of this highly publicized lawsuit, it is telling that not a single additional person has sought to join Plaintiffs in this case or brought individual claims. As the Supreme Court explained in Amchem Prods., Inc., Rule 23(b)(3) was designed to vindicate “the rights of groups of people who Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 79 of 84 -71- individually would be without effective strength to bring their opponents into court at all.” 521 U.S. at 617. That is a far cry from this case. Plaintiffs seek to explain the lack of interest among putative class members by asserting that women professionals at Goldman Sachs are “highly reluctant to sue the company for fear of retaliation.” (Pl. Br. 48.) But Plaintiffs’ only purported evidence of such reluctance are the conclusory statements made by three of their declarants. And, one of those women—Ms. Gamba—felt sufficiently free from the prospect of retaliation that she was willing to make her allegations in this lawsuit while still employed by Goldman Sachs.58 Plaintiffs assert without support that the “costs of Title VII litigation are prohibitive, even for a Goldman professional,” and that a class would be superior because “Plaintiffs have already conducted extensive discovery.” (Pl. Br. 49.) Both are untrue. Individual Title VII claims are brought in droves in federal courts—with concomitant discovery—including by plaintiffs whose financial resources pale in comparison to those of the highly compensated putative class members here. And, Title VII provides for an award of reasonable attorney’s fees for prevailing parties. Nor have Plaintiffs conducted “extensive discovery” relating to any individual claims other than their own. 59 58 Ms. Gamba’s supervisor was supportive of her decision to testify against Goldman Sachs, offering to help with “anything [Ms. Gamba] need[ed],” and Ms. Gamba does not believe that her participation will harm the working relationship with her supervisor. (Gamba Dep. at 21:7- 8.) 20:25-23:9.) 59 The Seventh Amendment also gives Goldman Sachs the right to a jury for these “additional proceedings,” and the prospect of multiple juries trying individual liability and damage issues following a first phase of this litigation raises significant Reexamination Clause implications that Plaintiffs have not addressed. See Blyden v. Mancusi, 186 F.3d 252, 268-72 (2d Cir. 1999) (in class case, after a first jury found a pattern of excessive force by corrections officers, the Seventh Amendment does not permit so-called “damages juries” to determine whether specific confrontations between prisoners and officers resulted in excessive force warranting damage award). Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 80 of 84 -72- Finally, although Rule 23(b)(3)(D) directs the Court to consider “the likely difficulties in managing a class action,” Plaintiffs ask the Court to ignore this requirement, claiming that it is “unnecessary to adopt a specific trial plan now.” (Pl. Br. 49.) As the Advisory Committee has recognized, however, “determin[ing] how the case will be tried [is a] critical need” in the Rule 23 analysis. There is a reason Rule 23(b)(3)(D) does not postpone that question for another day: the inherently individualized trials on liability and damages would render proceeding on a class basis here utterly unmanageable. There would be enormous difficulties in managing the claims of 2,300 dissimilarly situated professionals as a class action, because individual questions concerning a particular supervisor’s discretionary decisions as to a particular professional in a particular unit or group of a particular Business Unit could necessitate literally hundreds of complex mini-trials. B. Plaintiffs Cannot Avoid the Requirements of Rule 23(b)(3). 1. Plaintiffs Cannot Proceed Under Rule 23(b)(2). Recognizing the difficulty they face in seeking certification under Rule 23(b)(3), Plaintiffs attempt to relitigate whether they can seek classwide injunctive relief under Rule 23(b)(2). (Pl. Br. 43-44.) But that issue was resolved against Plaintiffs in 2012 when Judge Sand ruled that Plaintiffs lack standing to seek injunctive relief. As Judge Sand explained, the Supreme Court’s decision in Wal-Mart “forecloses certification under [Rule] 23(b)(2),” because the named plaintiffs, as former rather than current employees, have at best a “metaphysical interest in an injunction” against Goldman Sachs’ personnel practices. Chen-Oster v. Goldman Sachs & Co., 877 F. Supp. 2d 113, 121 (2012). That decision is the law of the case absent compelling circumstances such as “an intervening change of controlling law, the availability of new evidence, or the need to correct a clear error or prevent manifest injustice.” DiLaura v. Power Auth. of State of N.Y., 982 F.2d 73, 76 (2d Cir. 1992) (quoting Virgin Int’l Airways, Ltd. v. Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 81 of 84 -73- Nat’l Mediation Bd., 956 F.2d. 1245, 1255 (2d Cir. 1992)). Plaintiffs do not even claim that such circumstances exist here. Moreover, Plaintiffs’ proffered reason for revisiting this issue is disingenuous. Plaintiffs claim they should be allowed to seek injunctive relief because they sought reinstatement in their prayer for relief. (See Pl. Br. 44 & n.121.) Not only did Judge Sand reject this argument (July 17, 2012 Order, Dkt. 158, at 13), but Plaintiffs did not plead a claim entitling them to reinstatement. “[R]einstatement is appropriate … only upon a finding that the plaintiff was unlawfully or retaliatorally discharged.” McPartland v. Am. Broadcasting Cos., Inc., 1987 WL 14907 at *1 (S.D.N.Y. July 21, 1987). Nor does either Plaintiff truly seek reinstatement; rather, both assert only that they have a “keen interest” in making Goldman Sachs a better place for women to succeed. Ms. Chen-Oster says that she has this interest because she is a Managing Director at another financial services company, and “Goldman Sachs remains a leading company on Wall Street and is an opinion leader in many areas.” (Declaration of Cristina Chen-Oster ¶ 16.) Ms. Orlich, who has been practicing law in Arizona since 2010, parrots this language. (Declaration of Shanna Orlich ¶ 14.) 2. There Is No Basis for Certifying a Class Under Rule 23(c)(4). Plaintiffs suggest vaguely (Pl. Br. 42) that they could invoke Rule 23(c)(4), notwithstanding their failure to satisfy Rule 23(b)(3), citing the Second Circuit’s holding in In re Nassau County Strip Search Cases, 461 F.3d 219 (2006). That is incorrect. The Second Circuit has made clear that Rule 23(c)(4) may not be employed where certification “would not materially advance the litigation, because it would not dispose of larger issues such as . . . injury and damages.” McLaughlin v. Am. Tobacco Co., 522 F.3d 215, 234 (2008). In Nassau County, the Second Circuit permitted certification under Rule 23(c)(4) where individual questions would be “de minimis”—because it was undisputed that individual Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 82 of 84 -74- issues would need to be determined only “regarding a limited number of plaintiffs.” 461 F.3d. at 229-30. Resolving the common questions in that case (there was no dispute that such questions were presented) would materially advance the litigation. In McLaughlin, the Second Circuit rejected certification under Rule 23(c)(4) in a case involving alleged misrepresentations concerning tobacco products, explaining that “proof of misrepresentation—even widespread and uniform misrepresentation—only satisfies half of the equation; the other half, reliance on the representation, cannot be the subject of general proof.” 522 F.3d. at 223. In the face of the need for individualized proof as to reliance, injury, and damages, the court rejected plaintiffs’ attempt to certify under Rule 23(c)(4): “given the number of questions that would remain for individual adjudication, issue certification would not ‘reduce the range of issues in dispute and promote judicial economy.’” Id. at 234 (quoting Robinson, 267 F.3d at 168). The facts of this case equally bar certification. There is no issue that, once resolved, would materially reduce the individual issues that would remain for adjudication. Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 83 of 84 -75- CONCLUSION For the foregoing reasons, this Court should deny Plaintiffs’ motion for class certification. Respectfully, By /s/ Robert J. Giuffra, Jr. Barbara B. Brown (admitted pro hac vice) Robert J. Giuffra, Jr. Neal D. Mollen (admitted pro hac vice) Theodore O. Rogers, Jr. Carson H. Sullivan (admitted pro hac vice) Robin D. Fessel PAUL HASTINGS LLP Suhana S. Han 875 15th Street, NW Jeffrey B. Wall Washington, DC 20005 SULLIVAN & CROMWELL LLP 125 Broad Street New York, New York 10004 (212) 558-4000 Attorneys for Defendants, Goldman, Sachs & Co. and The Goldman Sachs Group, Inc. July 3, 2014 Case 1:10-cv-06950-AT-JCF Document 264 Filed 07/25/14 Page 84 of 84