Summary
denying motion to exclude expert testimony because the fact that some social science research is inconsistent goes to the weight of the evidence, not its admissibility
Summary of this case from T.B. v. Indep. Sch. Dist. 112Opinion
Civil No. 01-2086 (DWF/AJB).
June 20, 2006
Adam A. Gillette, Esq., and James H. Kaster, Esq., Nichols Kaster Anderson, PLLP; Samuel D. Heins, Esq., and Barbara J. Felt, Esq., Heins Mills Olson; and Joseph M. Sellers, Esq., Llezlie Green, Esq., and Christine E. Webber, Esq., Cohen Milstein Hausfeld Toll, PLLC — DC, counsel for Plaintiffs.
Holly S.A. Eng, Esq., Mark J. Ginder, Esq., Melissa Raphan, Esq., Michael Iwan, Esq., Paul B. Klaas, Esq., Ryan E. Mick, Esq., Shari L. Jerde, Esq., Emily L. Fitzgerald, Esq., and Heather C. Toft, Esq., Dorsey Whitney — Mpls, counsel for Defendant.
Peter N. Thompson, Esq., Hamline University School of Law, Special Master.
MEMORANDUM OPINION AND ORDER
Introduction
This matter came before the Court pursuant to Plaintiffs' Amended Motion for Class Certification and Defendant's Amended Motion to Strike Expert Reports and Opinions of Ronald L. Adler, William T. Bielby, and David W. Peterson. For the reasons set forth below, the Motion to Strike Expert Reports and Opinions is granted in part and denied in part and the Motion for Class Certification is denied.
Background
Plaintiffs, who are all salaried African-American current or former employees of Defendant Cargill, Inc. ("Cargill"), allege race discrimination in violation of 42 U.S.C. § 1981. Plaintiffs move the Court to certify a class consisting of:
Five Plaintiffs also assert disparate impact race discrimination claims under the Minnesota Human Rights Act, Minn. Stat. § A363.28, subd. 10, et. seq., but are not moving to certify a class on these claims.
[A]ll African-Americans who held salaried positions through salary grade 15 at Cargill and who:
A. were eligible to make claims alleging racial discrimination with respect to compensation and termination at any time between August, 1997[,] and the first date of trial;
B. were eligible to allege racial discrimination with respect to the denial of promotions that would have resulted in (i) a promotion from a non-exempt to an exempt position[,] (ii) a promotion from a non-management to a management position, or (iii) a promotion in which the employee's position in the chain of authority changes, at any time between August 22, 1995[,] and the first date of trial, and;
C. were eligible to allege racial discrimination with respect to the denial of promotions in any other respect at any time between August 22, 1997[,] and the first date of trial.
(Pls.' Am. Proposed Order regarding Mot. for Class Certification.)
Plaintiffs allege that Cargill has engaged in a "pattern and practice" of discrimination by fostering a pervasive company culture that has tainted its human-resource systems and produced a common discriminatory impact that affects promotions, compensation, and terminations. Plaintiffs' allegations relate to the processes that Cargill uses to evaluate employees and develop "high potential" employees. Such processes include the "Performance Management Process" ("PMP"), the "Key Employee Identification System" ("KEIS") and its successor, the "Leadership and Talent Management Program" ("LTMP"), the Cargill Leadership Development Program ("CDLP"), and the Strategic Business Development ("SBD") Plan. Additionally, Plaintiffs claim that Cargill has failed to effectively address widespread racial discrimination that the company knows is produced by its culture and the use of these systems.
To establish liability based on a pattern or practice of disparate treatment, Plaintiffs must prove that the defendant "`regularly and purposefully treated members of the protected group less favorably and that unlawful discrimination was the employer's `regular procedure or policy.'" E.E.O.C. v. McDonnell Douglas Corp., 191 F.3d 948, 951 (8th Cir. 1999) (quoting Int'l Bhd. of Teamsters v. United States, 431 U.S. 324, 335 360 (1977)). In other words, Plaintiffs must demonstrate that intentional discrimination was Cargill's "standard operating procedure the regular rather than the unusual practice." Teamsters, 431 U.S. at 336. Disparate treatment occurs where "[t]he employer simply treats some people less favorably than others because of their race, color, religion, sex or national origin." Id. at 335 n. 15. "Proof of discriminatory motive is critical, although it can in some situations be inferred from the mere fact of differences in treatment." Id. To prove a disparate treatment claim, plaintiffs typically offer statistical evidence of disparities between protected and unprotected employees who are otherwise similarly situated. Craik v. Minn. State Univ. Bd., 731 F.2d 465, 470 (8th Cir. 1984). But statistics alone are not dispositive. Catlett v. Mo. Highway Transp. Comm'n, 828 F.2d 1260, 1265 (8th Cir. 1987). And employers may show that the plaintiffs' "proof is either inaccurate or insignificant." Craik, 731 F.2d at 470 (quoting Teamsters, 431 U.S. at 360).
The named Plaintiffs request that the Court designate Roderick Arnold, Nii-Akwei Acquaye, Sean Allen, Hollis Branham, Toya Brown, Dawn Collins, Virginia Douglas, Cheneta Hughey, Jacqueline Jenkins, Keith Lewis, Vivian Little, Valerie Mason-Robinson, Anthony Mcdowell, Michael Mitchell, Phyllis Reece, Tonya Ross, Charles Scott, Clintonia Simmons, Tausha Tate, Emily Tyler, Jacqueline Williams, Cheryl Willis, Steve Wint, and Sean Wright as Class Representatives. Plaintiffs seek injunctive relief, punitive damages, and monetary relief.
Della Dickson was originally listed as a named plaintiff and class representative, but died on July 24, 2005. On March 2, 2006, the Court granted Plaintiffs' Motion for Substitution and ordered that Janice Lintz, Personal Representative of Della Dickson's estate, be substituted as a named Plaintiff in the place of Della Dickson. Dickson's estate now represents her legal interests in this case and Plaintiffs no longer propose Dickson as a class representative.
Cargill is a large, privately-held Delaware corporation headquartered in Wayzata, Minnesota. Cargill is an international marketer, processor, and distributor of agricultural, food, financial, and industrial products and services. Cargill employs approximately 101,000 employees in 60 countries, with more than 40,000 employees in more than 600 locations throughout the United States. Cargill operates through Business Units ("BUs"). Cargill has approximately 90 different BUs.
Cargill maintains a human resource ("HR") department at corporate headquarters that is overseen by an executive level Vice President. Cargill delegates authority to BU leaders to develop strategy and make operational decisions. Nancy Siska, Cargill's Vice President of Human Resources, testified that "We have a corporate process that is deployed within our business units. The decisions are made in the business units." (Aff. of Holly S.A. Eng in Opp. to Pls.' Mot. For Class Certification ("Eng Aff."), ¶ 4, Ex. B, Dep. of Nancy Siska ("Siska Dep.") at 277.) However, Cargill mandates that BUs adhere to the corporate headquarters' policies and standards for managing people. Cargill's HR department establishes and enforces company-wide policies that apply to salaried employees. Throughout the purported liability period, Cargill gave salaried employees an Employee Handbook that articulated those company policies and procedures.
Performance Management Process
In 2000, Cargill began requiring BU leaders and their management teams to conduct annual employee performance evaluations. The Performance Management Process ("PMP") is the principal tool that Cargill uses to evaluate the performance of its employees. The PMP includes three parts: a behavioral competencies section; a key results area ("KRA") section; and a development section. BU leaders customize the PMP to meet the needs of their particular industries and businesses. The behavioral competencies section includes six behaviors: "Discuss, Decide and Support," "Demonstrate Respect, Candor and Commitment," "Pursue and Reinforce Collaboration," "Ensure and Accept Accountability," "Challenge, Innovate, Change," and "Develop and Leverage Deep Customer Insight." (Decl. of Lawrence P. Schaefer in Supp. of Pls.' Mot. for Class Certification (Am.) ("Schaefer Decl."), ¶ 6, App. 4 at Px # 420.) The KRA section measures performance against job-related goals. Thus, KRAs are dependent on an employee's particular industry and the business strategies and goals of the particular business and location in which he or she works. If appropriate, technical competencies may be included in the PMP as well. The PMP template includes a rating for overall performance.
PMP implementation generally includes an employee self-assessment, followed by the employee's most immediate manager's assessment, negotiation by the employee and his or her manager over the rating, and a review by the manager's manager. Cargill's corporate HR department trains managers how to conduct PMPs, and specifically teaches managers how to make objective decisions about behavior. But Cargill does not make the training mandatory.
Employee Development
At the outset of the purported liability period until 2000, Cargill implemented a "high potential" employee-development process called the "Key Employee Identification System" ("KEIS"). KEIS has been described as a "people-planning process" to identify individuals who might be in a position to assume one of the top 100 positions within Cargill. When an opening occurred for one of these top positions, the KEIS program provided a list that division leaders could use as a tool for identifying replacement candidates. The manager or group of managers most closely associated with an employee determined ratings. The KEIS process involved using the employee's PMP rating and then assigning the employee a "potential" rating of either "A," "B," or "C." The KEIS program instructions define the factors that comprise a KEIS rating. Division heads' ratings were subject to review by their supervisors. Cargill's policy was to not provide employees with their KEIS forms or disclose KEIS ratings.
In 2001, Cargill replaced KEIS with the "Leadership and Talent Management Process" ("LTMP"). LTMP involved rating employees in one of four "high potential" categories: "Positional Leaders," "Next Generation Leaders," "High Impact Performers," and "Business Unit Others." A corporate selection department acts as a "coordinator" or "facilitator" in filling openings throughout the company with persons rated as high potential. Managers are instructed to consult with the selection department at corporate headquarters when filling open positions at the company. Employees not favorably rated are generally not competitive candidates for promotions or development assignments.
In addition to being used in selecting employees for promotions, KEIS and LTMP ratings were the primary factor used in selecting employees for participation in company-wide "leadership development" programs. Until 2001, Cargill's Organizational Development Group offered a training program called the Cargill Leadership Development Program ("CLDP"). CLDP was a year-long program in which about 50 "high potential" employees in salary grades 10-15 attended intensive leadership training. Another leadership program, "Strategic Business Development" ("SBD") involved an intensive two-year training program to groom participants for management positions.
Promotions
Decisions regarding promotions occur in BUs. The process by which promotions are made varies by BU and by business need. Plaintiffs contend that Cargill's promotion process is discretionary and shrouded in secrecy. Plaintiffs describe the promotion process as "secretive," "a mystery," "good old boys system," "smoke and mirrors," or a "tap on the shoulder system."
Compensation
Compensation plans at Cargill are designed and implemented at the BU level. BU managers have authority over individual salary and job-level decisions, and in some cases incentive awards. Compensation is linked to performance and external market rates. Salaried employees are compensated from different sources, ranging from straight salary to sales commissions, competitive bonuses, profit sharing bonuses, and industry-specific addons. Most BUs have a process in which managers will compile and review the results of their individual decisions.
Cargill's corporate HR office articulates a broad compensation philosophy that it implements by setting salary ranges, suggesting salary increase percentages for each pay grade, and requiring approval for increases outside specified parameters. However, BU leaders have the authority to implement compensation plans that meet their unique, competitive business needs as long as they meet certain standards of legal compliance and fairness, and they support Cargill's compensation philosophy.
Termination
Cargill delegates the authority to terminate employees "for cause" to division managers and local managers, with oversight and assistance by local HR. Cargill generates corporate-level policy guidelines to assist managers with the termination process. Cargill's "for cause" policy guidelines include the use of documentation, the provision of advanced warnings and opportunities for correction, the involvement of multiple decision makers, and, in many cases, multiple-level review. Managers are required to obtain prior authorization to terminate an employee with four or more years of service. The implementation of Reduction in Force ("RIF") guidelines is a division-level responsibility, with guidance and approval required through Cargill's Law Department. Cargill's RIF guidelines include documentation, consultation with local HR, preparation of job descriptions based on objective performance requirements, termination decisions based on employee qualifications as reflected on a selection grid, and approval and oversight.
Discussion
I. Motion to Strike
Cargill has moved to strike the testimony of Plaintiffs' experts, Ronald Adler, William Bielby, and David Peterson. The Court will address Cargill's motion to strike before reaching the issue of class certification because it finds the evidence at issue relevant to its determination of whether class certification should be granted.
The admissibility of expert testimony, whether in the form of an affidavit or a report, is governed in part by Rules 702 and 704 of the Federal Rules of Evidence. Rule 702 provides a standard of "helpfulness" that requires that expert testimony "assist the trier of fact to understand the evidence or to determine a fact in issue." Rule 704 provides that "testimony . . . otherwise admissible is not objectionable because it embraces an ultimate issue to be decided by the trier of fact."
The scope of the Rules has been illustrated by the Supreme Court's decisions in Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993) and Kumho Tire Co., Ltd. v. Carmichael, 526 U.S. 137 (1999). Daubert established factors to be considered when evaluating the principles and methodology of a scientific expert for validity, relevance, and ultimately, admissibility, and it held that general acceptance within the field was not a necessary precondition for admissibility. Kumho Tire extended the holding of Daubert to apply to non-scientific experts qualified by their experience, education, skill, or expertise in other fields.
The application of the Daubert test, however, is somewhat limited at the stage of class certification. Daubert is helpful to the extent that it can assist the Court in preventing the entrance of a methodology so apparently flawed. See In re Visa Check/Mastermoney Antitrust Litig., 192 F.R.D. 68, 76-77 (E.D.N.Y. 2000); In re Potash Antitrust Litig., 159 F.R.D. 682, 695-98 (D. Minn. 1995) (finding that plaintiffs are required only to make a "threshold showing" of whether proof will be "sufficiently generalized"). It would be inappropriate, however, for a court to look beyond the methodology and critique the prospective results of its application to a complete set of data. Id. A party and its experts should not be expected to have fully evaluated all data at the preliminary stage of class certification.
a. Ronald Adler
Ronald Adler is the president and CEO of an HR management consulting firm specializing in HR audits, employment practices risk management, strategic HR, employee surveys, and general HR management issues. Adler, who has more than 30 years of HR consulting experience, co-developed the Employment-Labor Law Audit™, ELLA® ("ELLA"). ELLA is an HR auditing and employment practices risk assessment tool, designed to help employers assess their exposure to employment practices liabilities and to help them identify the strengths and weaknesses of their strategic and transactional HR management programs and activities.
Here, Adler reviewed several current and former Cargill employees' deposition transcripts, the documents used as exhibits in these depositions, David Peterson's statistical report, Cargill's website, articles about Cargill, public and proprietary websites, and publications and articles regarding HR management employment policies and practices. In reviewing this material, Adler contends that he applied HR auditing and employment-related risk assessment techniques that he developed as an HR consultant. Adler concluded that "the design or implementation of [Cargill's] company-wide systems fail to eradicate and produce an ongoing vulnerability to systemic bias against its African-American employees in the area of performance management, promotion, talent management, compensation, and termination." (Schaefer Decl., ¶ 5, App. 3, Expert Report of Ronald L. Adler ("Adler Report") at 6.)
Cargill asserts that Adler's testimony should be excluded for several reasons. First, Cargill asserts that Adler lacks pertinent education, training, experience, or knowledge to qualify as an expert. Second, Cargill asserts that Adler failed to base his opinions on an adequate foundation. Third, Cargill asserts that Adler's methodology is not based on scientific or professional standards. Finally, Cargill asserts that Adler failed to reliably apply principles and methods to the facts of the case.
The Court finds that Adler's testimony and report are inadmissible at this preliminary stage of class certification. The Court finds that Adler's methodology does not meet scientific or professional standards of reliability. The Court agrees with Cargill that Adler fails to attribute his analysis to a recognized scientific basis. Although Adler claims to have used "employment related risk assessment techniques," Adler has not explained this process. On this record, the Court cannot determine what scientific method, if any, Adler used to reach his opinions. Instead, Adler's understanding of Cargill's corporate culture appears to be based solely on his own unfounded opinion as to what "culture" means and how it affects employment decision-making.
Plaintiffs' respond that Adler's analysis is not simply based on his own unfounded opinion because Adler's methodology has received feedback and endorsements from HR organizations and lawyers. The Court rejects Plaintiffs' assertion, however, because support from organizations and attorneys does not establish that Adler's methodology is reliable. The Court finds that endorsements are not an adequate substitute for scientific testing to determine reliability. Adler also relied on unreliable evidence. For example, in making his conclusions, Adler relied on commercial websites and popular publications that even Adler admitted were not scientific sources. (Decl. of Michael Iwan in Supp. of Mot. to Strike Expert Reports and Opinions of Ronald L. Adler, William T. Bielby, and David W. Peterson ("Iwan Decl.") at ¶ 2, Ex. A, Dep. of Ronald L. Adler ("Adler Dep.") at 240.)
Additionally, Adler's methodology has never been tested to establish reliability. Adler identifies his ELLA tool as the generally accepted standard for employment-related risk assessment techniques, but admits that ELLA has never been subjected to a scientific study to determine its validity. (Adler Dep. at 78-81, 28.) Moreover, Adler concedes that he did not complete an ELLA audit in evaluating Cargill. ( Id. at 32-33, 263-64.) Adler also admitted that he did not identify and eliminate any other possible causes of perceived disparities other than Cargill's policies and procedures. ( Id. at 114.)
Accordingly, the Court finds that Adler's testimony and report are inadmissible because Adler's methodology is not based on scientific principles. Because the Court has concluded that Adler's testimony and report are inadmissible at this stage of the litigation, the Court does not reach Cargill's additional contentions that Adler is not qualified as an expert, that Adler's opinions are not based on an adequate foundation, or that Adler's opinions are irrelevant and unhelpful.
b. William Bielby
William Bielby, Ph.D., a sociology professor, teaches courses and conducts research on organizational behavior, research methods for the social sciences, labor markets, quantitative methods, and social inequality. Bielby has specifically focused most of his research during the last twenty years on issues of workplace discrimination. Bielby reviewed the deposition testimony of Cargill's executives and managers regarding the company's personnel policies and practices, the corresponding exhibits, David Peterson's statistical report, and business-administration scholar Wayne G. Broehl, Jr.'s two-volume history of Cargill. Bielby also relied on social-science research on workplace bias. Bielby applied social framework analysis ("SFA"), in which he looked at "distinctive features of the organization's policies and practices and evaluate[d] them against what social-science research shows to be factors that create and sustain bias and those that minimize bias." (Schaefer Decl., ¶ 5, App. 3, Expert Report of William T. Bielby ("Bielby Report") at 4.) Bielby concluded that Cargill's strong organizational culture creates or sustains barriers to career advancement for African-Americans and that Cargill's Equal Employment Office ("EEO") policies are deficient in identifying and minimizing the effects of those barriers.
Cargill asserts that the Court should exclude Bielby's testimony on three grounds. First, Cargill contends that Bielby is unqualified to offer an opinion concerning the impact of culture or EEO accountability efforts on African-Americans. Second, Cargill contends that Dr. Bielby's methodology is unreliable. Third, Cargill contends that Bielby's opinions are irrelevant and therefore unhelpful.
The Court finds that Bielby is qualified as an expert. After reviewing Bielby's curriculum vitae and deposition testimony, the Court finds that Bielby has considerable experience with and expertise in issues of race in the workplace. Bielby has written reports, authored articles, taught classes, and given presentations on issues involving race. The Court rejects Cargill's assertion that Bielby's research and expert experience concerning race issues did not directly address the issues on which he opines in this case. Although Cargill contends that Bielby's race-related research primarily focused on the entertainment industry, Bielby testified that these studies also dealt with the issue of corporate culture as it relates to race issues. (Iwan Decl., ¶ 2, Ex. B, Dep. of William T. Bielby ("Bielby Dep.") at 24.) Bielby also testified that among the twelve cases in which he had testified in the four years preceding his deposition, four involved racial issues, and in two cases he addressed issues of corporate culture in his expert report. ( Id. at 16-18.) Further, since giving his deposition, Bielby submitted an expert report in connection with class certification in another race-discrimination case, Satchell v. FedEx Corp., 2005 WL 2397522, *3 (N.D. Cal. Sept. 28, 2005). Further, the Court rejects Cargill's assertions that Bielby conflates gender and race and that Bielby cannot claim expertise in the subfield of race simply by applying the conventions learned in the subfield of gender. The Court finds that Bielby has sufficient experience and expertise with the issues of race in the workplace, independent of his expertise in gender.
The Court also finds that Bielby's methodology is reliable. Here, Cargill acknowledges that courts have found SFA reliable, but asserts that Bielby violates the tenets underlying that methodology. Specifically, Cargill contends that Bielby does not premise SFA on a neutral statement of sociological knowledge and inappropriately advances unfounded theories of race effects as social fact. Cargill further contends that because sociological research does not unequivocally point to the conclusion that perceived unequal outcomes for African-Americans can be causally attributed to corporate culture or latent bias, Bielby was required to conduct empirical research on African-Americans' experiences at Cargill. Cargill also contends that Bielby's opinions cannot be quantified or assigned a potential rate of error. Finally, Cargill contends that Bielby cites no standards for measuring corporate culture.
Plaintiffs, on the other hand, contend that Cargill relies solely on its own expert's report to support its contention that Bielby's application of SFA is not generally accepted in the scientific community. Plaintiffs' expert concludes that Bielby's work departs from SFA because some published studies differ from those studies on which Bielby relied. The Court agrees with Plaintiffs that the fact that some social-science research is inconsistent with Bielby's testimony goes to the weight, not the admissibility, of Bielby's testimony. The Court also agrees with Plaintiffs that empirical research is not required under SFA. Furthermore, the Court agrees that Bielby had no way to access Cargill employees to conduct empirical research. Finally, the Court agrees with Plaintiffs that there is no expectation that an expert opinion be quantifiable. See Smith v. BMW North America, Inc., 308 F.3d 913, 920 (8th Cir. 2002). Thus, the Court finds that Bielby's methodology is reliable.
Finally, the Court rejects Cargill's assertion that Bielby's opinions are irrelevant. The Court finds that Cargill's assertion is based on the mistaken belief that Bielby must offer an opinion on the ultimate issue of whether Cargill engaged in intentional discrimination in order to be relevant. Instead, the Court agrees with Plaintiffs that Bielby's opinion that Cargill has a strong, common culture is relevant because it may make it more likely that common questions exist at the class certification stage. Further, a pattern or practice of intentional discrimination can be established by showing statistical disparities and common employment practices that could cause such statistical disparities. See Teamsters, 431 U.S. at 360 n. 46; Craik, 731 F.2d at 479. Thus, the Court finds that Dr. Bielby's opinions are relevant and therefore admissible. c. David Peterson
Noting Cargill's objection, the Court grants Plaintiffs' Motion for Leave to File a Supplemental Declaration of David W. Peterson, Ph.D., because the admission of the declaration makes no difference as to the outcome of the motion.
David Peterson, Ph.D., a statistician, conducted statistical analyses to determine whether statistical evidence supports the claim that Cargill discriminates against salaried African-American employees below the senior management level in performance rating, advancement, compensation, and termination. Peterson concluded that Cargill generally employs few African-Americans relative to their representation in the civilian labor force. Peterson also concluded that there is a pervasive pattern of statistically significant disparities that disadvantage African-Americans throughout Cargill. Further, Peterson concluded that these disparities persist from year to year, apply both to exempt and non-exempt employees, and are found across divisions and job groups, even after accounting for differences in education level and service.
Specifically, Peterson conducted multiple regression analyses. "A multiple regression analysis attempts to reveal relationships between explanatory variables and a dependent variable." Morgan v. United Parcel Serv. of America, Inc., 380 F.3d 459, 466 (8th Cir. 2004) (citing Daniel L. Rubinfeld, Reference Guide on Multiple Regression, in Federal Judicial Center, Reference Manual on Scientific Evidence 181 (2d ed. 2000). Explanatory variables those factors that one would expect to influence the dependent variable and for which the regression controls. Id.
Cargill contends that Peterson's analysis suffers from two fundamental flaws that render it inadmissible under the Daubert standard. First, Cargill contends that Peterson based his analytical framework on his generalized assumptions about how employers typically make employment decisions, rather than how Cargill actually makes such decisions. Thus, according to Cargill, Peterson presumed the existence of discrimination, thereby creating a circular and self-fulfilling hypothesis.
Second, Cargill contends that because Peterson does not know how Cargill makes employment decisions, he cannot know whether he has controlled for all reasonable, non-discriminatory explanatory factors. In particular, Cargill contends that Peterson failed to determine whether the distribution of performance ratings and salary-grade variables could explain, in whole or in part, the statistical disparities that Peterson found in pay, promotion, and termination decision-making and which Peterson attributes to racial discrimination. Cargill's expert, Robert Topel, Ph.D., an economist, analyzed the effect that controlling for performance ratings would have on the claimed disparities and concluded that this explains the outcomes in pay, promotion, and termination. Cargill also contends that Peterson ignored a host of "pre-market" factors correlated with race, that, according to Topel, impact employment outcomes. Cargill asserts that the facts here closely resemble those in Morgan v. United Parcel Serv. of America, Inc., 380 F.3d 459, 466-471 (8th Cir. 2004). In Morgan, the court rejected the conclusions reached by the plaintiffs' expert in support of an attempt to certify a nation-wide class of employees in a § 1981 action and held that a statistical analysis of employment outcomes cannot fail to at least consider obviously relevant explanatory factors simply because of the unproven claim that the explanatory variable is tainted by discrimination. Id. at 469-71.
The Court finds that Peterson's report and opinions are admissible at this preliminary class certification stage. The Court rejects Cargill's unsupported assertion that Peterson's statistical analysis must follow Cargill's actual decision-making process in order to be admissible. Such a contention does not render Peterson's accepted statistical methodology unreliable at this stage. The Court also rejects Cargill's assertion that Peterson's testimony should be excluded on the basis that Peterson failed to control for all reasonable, non-discriminatory explanatory factors.
Here, Peterson controlled for an employee's education level, years of service at Cargill, job categorization as exempt or non-exempt under the Fair Labor Standards Act, the job group to which the employee's position was assigned, and the division in which the employee worked. (Schaefer Decl., ¶ 5, App. 3, Expert Report of David Peterson ("Peterson Report") at 5.) Further, Plaintiffs allege that Peterson's omission of performance evaluation and salary grade variables was deliberate and necessary because these variables are tainted by the same discrimination that is alleged to affect the other challenged practices. Plaintiffs contend that Cargill's reliance on Morgan is misplaced because Peterson analyzed the data and found statistical evidence of discrimination in promotions and in performance ratings. (Peterson Report at 12-28.) The Court agrees that the facts here are distinguishable from those in Morgan, where the court found that the plaintiffs had "no evidence, statistical or otherwise, that past pay disparities were racially discriminatory." 380 F.3d at 471. The Court finds that Cargill's assertion that Peterson should have controlled for certain variables impacts the Court's determination of what weight to assign Peterson's report when considering whether to grant class certification, but does not impact the issue of admissibility. See Bazemore v. Friday, 478 U.S. 385, 400 (1986) (holding that the question of what explanatory variables should be included in a particular regression normally "affect[s] the analysis' probativeness, not its admissibility").
II. Motion for Class Certification
A. Standard for Certification
In order to proceed with a class action, a plaintiff has the burden of showing that the four requirements of Fed.R.Civ.P. 23(a) are met:
(1) the class is so numerous that joinder of all members is impracticable,
(2) there are questions of law or fact common to the class, (3) the claims or defenses of the representative parties are typical of the claims or defenses of the class, and (4) the representative parties will fairly and adequately protect the interests of the class.Coleman v. Watt, 40 F.3d 255, 258-59 (8th Cir. 1994); Kassover v. Computer Depot, Inc., 691 F. Supp. 1205, 1213 (D. Minn. 1987), aff'd, 902 F.2d 1571 (8th Cir. 1990). The plaintiff must then show that the action falls within one of the categories listed in Rule 23(b). Kassover, 691 F. Supp. at 1213.
In determining the propriety of a class action, the question is not whether the plaintiff or plaintiffs have stated a cause of action or will prevail on the merits, but rather whether the requirements of Rule 23 are met. Eisen v. Carlisle Jacquelin, 417 U.S. 156, 178 (1974). A trial court has broad discretion in determining whether a class action may be maintained. Bishop v. Comm. on Prof'l Ethics, 686 F.2d 1278, 1287 (8th Cir. 1982).
1. Numerosity
The first requirement of Rule 23(a) is that the class is so numerous that joinder of all members is impracticable. There is, however, no absolute number that satisfies the numerosity requirement. Paxton v. Union Nat'l Bank, 688 F.2d 552, 559 (8th Cir. 1982); In re Workers' Compensation, 130 F.R.D. 99, 104 (D. Minn. 1990). The Eighth Circuit has not established any rigid rules regarding the necessary size of a class and the question of what constitutes impracticability depends upon the facts of each case. Boyd v. Ozark Air Lines, Inc., 568 F.2d 50, 54 (8th Cir. 1977); In re Workers', 130 F.R.D. at 104. A number of factors are relevant to this inquiry, the most obvious of which is, of course, the number of persons in the proposed class. Paxton, 688 F.2d at 559. In addition to the size of the class, the court may also consider the nature of the action, the size of the individual claims, the inconvenience of trying individual suits, and any other factor relevant to the practicability of joining all the putative class members. Paxton, 688 F.2d at 559-60.
Here, the parties do not dispute whether the numerosity requirement is satisfied. There are approximately 1,600 potential class members. Given the large number of potential class members, the Court finds that joinder would be impracticable. Thus, the Court finds that Plaintiffs have met the "numerosity" requirement.
2. Commonality
The rule of commonality requires that there be common questions of law or fact among the class members. Paxton, 688 F.2d at 561. Class certification is "[p]eculiarly appropriate" when the "issues involved are common to the class as a whole" and when they "turn on questions of law applicable in the same manner to each member of the class." General Telephone Co. v. Falcon, 457 U.S. 147, 154-56 (1982) (quoting Califano v. Yamasaki, 442 U.S. 682, 701 (1979). The rule, however, does not require that every question of law or fact be common to every class member, and may be satisfied, for example, where the question of law linking the class members is substantially related to the resolution of the litigation even though the individuals are not identically situated. Paxton, 688 F.2d at 561.
Courts have held that allegations of similar discriminatory employment practices, such as the use of entirely subjective personnel processes that operate to discriminate, satisfy the commonality and typicality requirements of 23(a). See, e.g., Shores v. Publix Super Markets Inc., 1996 WL 407850, at *6 (M.D. Fla. Mar. 12, 1996) (citing cases). The Shores Court observed that the employer's "policy of delegating hiring and promotion decisions to managers, who make those decisions on the basis of subjective criteria, is a common course of conduct. Plaintiffs' allegations that this course of conduct results in a discriminatory practice is adequate to meet the commonality requirement of Rule 23." Shores, 1996 WL 407850 at *6.
In addition, courts have certified classes including employees of more than one location or division of a company where it has been found: (1) that a single policy prevails in several or all of defendant's subdivisions; (2) that the subdivisions are not autonomous; or (3) where there are specific allegations pertaining to more than one location. Avagliano v. Sumitomo Shoji America, Inc., 103 F.R.D. 562, 578-79 (S.D.N.Y. 1984) (citing cases). The commonality requirement is satisfied as long as the members of the class have allegedly been affected by a general policy of the defendant, and the general policy is the focus of the litigation. Mayo v. Sears, Roebuck Co., 148 F.R.D. 576, 580 (S.D. Ohio 1993).
a. Plaintiffs' "Culture" Claim
Plaintiffs first assert that the impact of Cargill's culture on African-American employees satisfies commonality. Plaintiffs assert that "Cargill's systemic race discrimination is rooted in a pervasive and powerful homogeneous company culture." (Mem. in Supp. of Pls.' Mo. for Class Certification (Am.) at 4.) Plaintiffs contend that Cargill used its culture to influence the behavior of managers and employees throughout the organization. Plaintiffs contend that "[p]roof of a corporate culture that emanates from CEO pronouncements and fosters discrimination establishes commonality." ( Id. at 7.)
Plaintiffs contend that former Cargill Chief Executive Officer ("CEO") and Board Chairman, Whitney MacMillan, encouraged homogeneity. In support of this contention, Plaintiffs point to a statement that MacMillan allegedly made in an annual recruiting meeting around 1993. Plaintiffs allege that MacMillan, a Caucasian, told a large group of Cargill recruiters, senior human resources professionals, and executives, including African-American employees Beverly Franklin and Duane Reed, that they should "recruit people that looked and talked like him." (Schaefer Decl., ¶ 9, App. 7, Dep. of Duane Reed at 80-85; Dep. of Beverly Franklin at 16.) Although MacMillan was not Cargill's CEO during the class period, Plaintiffs contend that MacMillan molded Cargill's culture well into the liability period.
MacMillan was Cargill's CEO and Board Chairman from 1977 to August 1995.
Cargill also contends that a company-wide "culture scan" that Cargill conducted in 2000 confirmed that homogeneity was a part of Cargill's culture. Plaintiffs note that the scan determined that among the "keys" to "Cargill's cultural core" were the descriptors "traditional," "inward looking," and "secretive." (Schaefer Decl., ¶ 6, App. 4, Px # 755 at C551966-67, C551969-70.) Plaintiffs also contend that the scan articulated "secondary cultural attributes" including "assimilation," "acculturation," and "unearned privilege." ( Id. at C551981-82.) Plaintiffs contend that Cargill's managers perpetuate these attributes by rewarding and advancing employees based on "sameness," "fit," "likeness," or similar qualities. Finally, Plaintiffs contend that sociological research establishes that institutional barriers to equal treatment frequently arise when a group comprises a small percentage of the workforce. (Bielby Report at 10-16.) Plaintiffs contend that Bielby concluded that these barriers "are especially pronounced" at Cargill given the company's culture of sustaining uniformity. ( Id. at 4.)
Cargill, on the other hand, contends that it is decentralized and requires many "cultures" in order to survive and thrive in the many facilities, industries, and countries in which it competes. Cargill contends that its many cultures vary from BU to BU. Cargill also contends that its decentralization was intentional. In support of this contention, Cargill notes that MacMillan testified that the only elements common to all of Cargill were "[t]eamwork, fun, [and] decentralized management system[s]." (Eng Aff., ¶ 4, Ex. B, Dep. of Whitney MacMillan at 53-54.)
Cargill contends that Plaintiffs incorrectly attempt to paint MacMillan as the impetus of a "common culture" of discrimination. Cargill asserts that even if MacMillan had stated that Cargill employees should recruit people who "looked and talked" like him, the statement was made long ago, and the two Cargill CEOs who cover the entire class period have supported and insisted upon building a diverse workforce at Cargill. For example, Cargill points out that Warren Staley, Cargill's CEO since 1998, regularly issues corporate-wide statements that "[v]aluing differences isn't just the nice thing to do, it's the smart thing to do," and that "[i]f we seek sameness in our people, we limit our ability to become a truly high performing organization." (Eng Aff., ¶ 5, Exh. C, Tabs 11 at C610567, 4, and 7 at C360340.) Finally, Cargill asserts that Plaintiffs' expert admits that Cargill does not have a centralized discriminatory culture.
MacMillan denies making this statement.
The Court rejects Plaintiffs' allegation that the impact of Cargill's culture on African-American employees satisfies commonality. Plaintiffs primarily rely on MacMillan's alleged statement that Cargill employees should recruit people who look and talk like him. But the Court finds that this statement is not sufficient to establish a common discriminatory culture against African-American employees, particularly given that MacMillan made the statement long before the liability period began and when the evidence shows that the two CEOs during the liability period valued diversity. The evidence also shows that although Cargill implements corporate policies throughout its company, it has varied cultures across BUs. The Court finds that Cargill's statement that "Cargill's systemic race discrimination is rooted in a pervasive and powerful homogeneous company culture" overstates the evidence in this case. The Court finds that such a conclusory statement does not satisfy the "commonality" requirement.
b. Plaintiffs' "Systems" Claims
Second, Plaintiffs assert that the design, centralized control, and impact of the systems at issue satisfy the commonality requirement. In particular, Plaintiffs contend that Cargill's "performance appraisal system," "systems for identifying talent and promoting employees," the "compensation system," and the "system for terminating employees" magnify the common impact of Cargill's culture on the proposed class. Plaintiffs assert that Cargill's corporate HR department centrally designed and administered on a company-wide basis each system that the Plaintiffs challenge. Plaintiffs also assert that statistics further establish commonality by demonstrating the company-wide impact of Cargill's culture and systems on the proposed class.
In particular, Plaintiffs contend that, throughout the liability period, Cargill has centrally mandated and managed the PMP process. Plaintiffs allege that the six behavioral factors included in the PMP are subjective. Plaintiffs also contend that although Cargill corporate HR offers PMP training to managers, such training is not mandated. Finally, Plaintiffs contend that Cargill did nothing in response to a 1997 study that showed that African-American employees received lower PMP ratings than non-minority employees. Specifically, in 1997, Cargill executives reviewed a study of PMP overall performance ratings showing that 9.6% of minorities were rated below expectations compared to 0.6% of non-minority employees. It further showed that minorities were more likely (25%) to receive no review at all than their non-minority peers (18.9%). Plaintiffs contend that by never repeating the 1997 analysis, Cargill ignored the alleged adverse impact of the PMP system on salaried African-American employees. Plaintiffs contend that Cargill's deliberate disregard of this impact raises the common issue of whether it intended to discriminate against African-American employees through the use of the PMP.
Plaintiffs further assert that the known negative impact of the PMP on African-American employees has continued throughout the liability period. In support of this contention and as an example, Plaintiffs contend that Plaintiff Michael Mitchell's completed 2001 PMP was lowered two rating levels, from "exceeds some" to "meets some," by his Caucasian manager after the review process was over, and the manager inserted criticism into the revised review that was inconsistent with his original review comments. (Schaefer Decl., ¶ 9, App. 7, Dep. of Michael Mitchell at 189-92, 197-98, 205.) Mitchell claims that, as a result of his PMP results, he was placed on a Performance Improvement Plan ("PIP"), and forced to resign a few months later. ( Id. at 105-6, 255-56, 321-330.) Additionally, Plaintiffs contend that Peterson's statistical analysis confirms widespread discrimination regarding the performance appraisal system. In analyzing the PMP ratings, Peterson concluded that "the overall patterns in which the treatment of African Americans is systematically dissimilar from that of people in their cohorts," is disadvantageous to African-American employees by 9.42 standard deviations. (Peterson Report at 285.)
The Supreme Court has explained that "[a]s a general rule . . ., if the difference between the expected value and the observed number is greater than two or three standard deviations," the disparity is not due to chance. Castaneda v. Partida, 430 U.S. 482, 496-97 (1977).
Plaintiffs also contend that the systems that Cargill has used to identify and develop "high potential" employees and to promote employees are centrally controlled and have harmed African-American employees throughout the purported liability period. Plaintiffs assert that both KEIS and LTMP were centrally managed and used to develop and perpetuate Cargill's culture. Plaintiffs allege that both processes involve categorizing employees' potential based on largely subjective assessments. Plaintiffs also allege that the KEIS and LTMP processes have been entirely secretive because employees never knew of the ratings exercises or how they were evaluated under the systems.
Plaintiffs contend that Cargill was aware that African-American employees received fewer high potential-ratings than their Caucasian peers, but did nothing to monitor these systems on African-American employees. Plaintiffs assert that because high-potential ratings are critical to advancement, African-American employees were not promoted. Plaintiffs also contend that, in addition to suppressing promotions awarded to African-American employees, KEIS and LTMP ratings kept African-American employees from being selected to participate in company-wide "leadership development" programs.
Next, Plaintiffs allege that Cargill's promotion process is discretionary and shrouded in secrecy as evidenced by documents describing the promotion process as "secretive," "a mystery," "good old boys system," "smoke and mirrors," or a "tap on the shoulder system." Plaintiffs contend that they and class members throughout Cargill's operations were passed over for promotions in favor of less qualified non-African-American peers.
Plaintiffs also assert that the fact that some of Plaintiffs' promotion claims are governed by a six-year statute of limitations period while other claims are subject to a four-year limitations period does not diminish the susceptibility of this action to class treatment. See the Court's March 22, 2005 Order (holding that Minnesota's six-year statute of limitations applies to Plaintiffs' promotions claims that were cognizable under 42 U.S.C. § 1981 before enactment of the Civil Rights Act of 1991, 42 U.S.C. § 2000e-2(k)). But the Court need not reach this issue because it denies Plaintiffs' motion for class certification on other grounds.
Plaintiffs contend that this evidence, coupled with Peterson's statistical analysis, compels a finding of commonality. Peterson calculated, within each cohort, "the perfect parity number of blacks who would be promoted," "the actual number of blacks who were promoted," and then the "variance that's associated with that actual number." (Peterson Report at 151-52.) Peterson then aggregated the results of each cohort analysis. ( Id.) Peterson found that the resulting disparity was -4.82 standard deviations, or less than one chance in a million. ( Id. at 22.)
Plaintiffs also contend that Cargill's failure to implement suggested remedial measures for this known problem provides further support for finding commonality. Plaintiffs allege that HR personnel recommended implementing a company-wide job posting system through salary grade 13 in order to advance diversity and remedy the widespread perception that promotions are conducted through a "smoke and mirrors" or "tap-on-the shoulder" method. Although Plaintiffs acknowledge that many BUs have implemented job-posting, Plaintiffs contend that the company should have required all BUs to post for open positions.
Further, Plaintiffs assert that throughout the liability period, Cargill has operated a centrally designed and controlled "pay for performance" compensation system. Plaintiffs acknowledge that Cargill allowed base pay ranges to vary by BU, but contend that Cargill's corporate headquarters otherwise maintained authority by mandating use of a single system for classifying jobs within pay grades and requiring each BU to report certain information. Plaintiffs admit that Cargill delegates authority to local managers to set individual pay increases and make discretionary pay decisions, and that Cargill holds the local managers accountable for such decisions. However, Plaintiffs contend that accountability is illusory because Cargill never required the BUs to conduct compensation analyses to examine how African-American employees fare at Cargill. Plaintiffs also contend that Cargill has not engaged in any corporate-wide analysis to determine whether the BUs have met their obligation to award pay increases and discretionary pay in a non-discriminatory manner.
Plaintiffs also assert that Cargill's compensation system reflects the company's alleged secretive culture. Plaintiffs contend that African-American employees' performance evaluations directly affect compensation, and because pay increases and discretionary pay are tied to these employees' performance evaluations, the alleged discriminatory impact of the PMP system directly infects Cargill's compensation system. Plaintiffs also maintain that the alleged discriminatory impact of KEIS and LTMP ratings also affect compensation decisions.
Plaintiffs further contend that Cargill has implemented its company-wide compensation system in a biased manner, resulting in lower pay increases and less discretionary pay than for similarly situated non-African-American peers. Plaintiffs and class members describe being paid less and receiving lower raises or bonuses than less qualified non-African-American peers doing similar or less responsible work. Plaintiffs also assert that Peterson's two — part statistical analysis supports their contention that Cargill's compensation system has operated to disadvantage the class throughout the liability period. Peterson first engaged in a pay progression study, which "offers a method for examining both pay and advancement practices, since pay increases often correspond to promotions and other advancement opportunities." (Peterson Report at 29.) Plaintiffs contend that the results show that African-American employees received smaller pay increases than others in their cohorts. Plaintiffs contend that the overall disparity equates to — 5.11 standard deviations, with a likelihood of less than one in six million that the imbalance occurred by chance. ( Id. at 29.) Peterson also found "a strong pattern in which African Americans have received smaller bonuses, awards or incentives than have their cohort peers, measuring 10.38 standard deviations." ( Id. at 38.)
Finally, Plaintiffs assert that Cargill's policies and system for terminating salaried employees are centralized. Plaintiffs point out that Cargill requires corporate approval of certain terminations. Plaintiffs contend that their descriptions of events surrounding their terminations establish the common nature of their complaints and the need to certify this claim on a class-wide basis. For example, Plaintiffs contend that they and other class members were told they were terminated because they did not "fit" at Cargill. (Schaefer Decl., ¶ 9, App. 7, Dep. of Keith Lewis at 203-05; Dep. of Michael Collier at 48-49.) Additionally, Sean Wright claims that he was terminated in an RIF when four Caucasian peers were retained. ( Id.; Dep. of Sean Wright at 454.) Wright contends that the managers involved in making this decision would not let Wright see copies of the selection grids allegedly used to make the RIF determination. ( Id. at 469-70.)
Plaintiffs also assert that Peterson's statistical evidence supports certifying the termination claims on a class-wide basis. Peterson concluded that ninety-six more terminations occurred than would be expected if African-American employees were terminated in the same proportions as non-African-American employees. (Peterson Report at 44-50; Schaefer Decl., ¶ 5, App. 3, Peterson Dep. at 151.) Plaintiffs contend that the excessive African-American involuntary terminations equates to — 7.96 standard deviations. (Peterson Report at 44.)
Cargill, on the other hand, asserts that Plaintiffs have not presented evidence of some common, discriminatory policy that affects members of the putative class in a consistent way. Cargill contends that the record demonstrates that Cargill does not have centrally designed and managed HR systems. Additionally, Cargill alleges that Plaintiffs have not presented evidence that Cargill's decision-making on personnel matters is "entirely subjective." Cargill also contends that Topel's statistical evidence indicates that the numerical performance of Cargill's African-American employees as compared to their white co-workers, varies tremendously across BUs: sometimes better, sometimes worse, sometimes exactly the same.
The Court agrees and finds that Plaintiffs have not satisfied the "commonality" requirement. After carefully considering the Plaintiffs' allegations and the evidence submitted by the parties, the court finds no identifiable pattern or practice of discrimination that affects the purported class in common ways. The evidence shows that there has never been a uniform, centralized system of performance evaluation in place across Cargill BUs and locations. Instead, BUs adopting the PMP have customized the PMP to meet the needs of their particular industries and businesses. Additionally, the KRAs are unique to the individual whose performance is being measured. Thus, KRAs are dependent on an employee's particular industry and the business strategies and goals of the particular business and location in which he or she works.
The Court also finds that the PMP is not entirely subjective. Conducting performance evaluations is a process that necessarily involves some discretion. See Abram v. United Parcel Serv. of America, Inc., 200 F.R.D. 424, 430 (E.D. Wisc. 2001) (finding that "subjective criteria necessarily and legitimately enter into personnel decisions involving supervisory positions.") Here, the behavioral factors in the PMP are objective as possible. Cargill also offered training on conducting PMP evaluations that addressed ways to make objective determinations. Moreover, many Plaintiffs testified that the PMP is a good tool and that its factors are job-related and reasonable.
Further, the Court finds that KEIS was not entirely subjective. Instructions defined the factors comprising a KEIS rating, and division heads' ratings were subject to review. Further, Plaintiffs have offered neither evidence of intentional discrimination through the KEIS process nor any meaningful statistical support for such a proposition. The Court also rejects Plaintiffs' claim of a common experience of discrimination under LTMP. Here, none of the named Plaintiffs has complained about LTMP, and Plaintiffs' expert also has no criticism of it and has performed no statistical analysis relating to it. Further, Plaintiffs have never challenged the CLDP program manager's assertion that granting affirmative preferences for African-Americans, not intentional discrimination, was standard procedure for the CLDP. Here, the program manager, Amy Cullen, testified that she selected every minority nominee for participation to maximize the diversity of the CLDP participant pool.
The Court also finds that Plaintiffs' claim that Cargill has engaged in a centralized pattern and practice of discrimination in the advancement of African-American employees ignores Cargill's economic and business realities. Here, the actual promotion of employees occurs at the BU level. Open positions are filled in a multitude of ways, and the process by which promotions are made varies by BU and by business need. Peterson's statistical analysis also weighs against a finding of commonality. Although Peterson claims that "Cargill promoted fewer African Americans than expected in 25 of the 48 job groups in which there are comparators and at least one promotion," Peterson's analysis demonstrates that in the other 23 job groups Cargill promoted more African-Americans than expected. (Peterson Report at 24, 26-27.)
Additionally, Plaintiffs' compensation claims do not establish commonality. The record demonstrates that Cargill does not have a centrally designed and managed compensation system. Rather, compensation plans at Cargill are designed and implemented at the BU level. Although Cargill articulates a broad compensation philosophy, the BUs and the BUs' various locations are free to develop and implement compensation plans that meet their unique, competitive business needs as long as they meet "minimum standards" of legal compliance, fairness, and support of Cargill's compensation philosophy. Further, Cargill limits managerial discretion by requiring that pay be linked to performance and external market rates.
The statistical evidence weighs against finding commonality with regard to Plaintiffs' compensation claims. Using Peterson's allegedly flawed definitions and methodology, rates of salary growth among African-American and Caucasian employees with similar performance reviews are statistically indistinguishable. Further, a proper analysis of change in total pay controlling for year, service, education, location, and type of work yields no statistically significant difference in total pay growth for African-American and Caucasian employees, even without controlling for performance, for both exempt and non-exempt employees. (Eng. Aff., ¶ 7, Ex. E, Expert Report of Robert Topel ("Topel Report") at 54-57, Exs. 43-47.)
Finally, Cargill's "termination system" does not support finding that Plaintiffs have satisfied the commonality requirement. Although Cargill issues corporate-level policy guidelines, the implementation, review, and accountability occurs at the BU level. The authority to terminate employees for cause has been delegated to division managers and, in turn, to managers at the local level, with oversight and assistance by local HR. Cargill's "for cause" policy guidelines include the use of documentation, the provision of advanced warnings and opportunities for correction, the involvement of multiple decision makers, and, in many cases, multiple-level review. Additionally, Cargill's RIF guidelines include documentation, consultation with local HR, preparation of job descriptions based on objective performance requirements, termination decisions based on employee qualifications as reflected on a selection grid, and approval and oversight. Thus, the termination process is not entirely subjective.
The statistical evidence also weighs against finding commonality with regard to terminations. Racial disparity is statistically eliminated in performance-related terminations when actual performance is taken into account. (Topel Report at 64.) Furthermore, even accepting and using Peterson's report, 82% of Cargill's total employee population worked in divisions without a statistically significant deviation from the expected number of African-American terminations and 100 of 105 divisions (95.2%) did not experience any deviation. (Peterson Report at Table 6.1, 6.5.) Additionally, in the six divisions represented by the seven named Plaintiffs with involuntary termination claims, none had a statistically significant deviation from expected African-American terminations, and in three, the disparity favored African-Americans. ( Id.)
In addition to finding that commonality is lacking for the above reasons, the Court also finds that Cargill's diverse and complex business structure weighs against finding that Plaintiffs have satisfied commonality. See Wright v. Circuit City Stores, Inc., 201 F.R.D. 526, 540-42 (N.D. Ala. 2001) (finding no commonality in a class action lawsuit against Circuit City, in part, because in Wright the purported class was spread across as many as fifteen states involving 125 to 160 different stores, warehouses, and service center facilities, each of which had varying degrees of autonomy over hiring, evaluation, and promotion decisions). The named Plaintiffs also come from only a small percentage of the total number of BUs, and even fewer of the actual locations that comprise Cargill's domestic businesses. Here, the named Plaintiffs only worked at 13 of 87 BUs at the time this lawsuit was filed.
The wide array of positions that the putative class members have held also weighs against finding that commonality is satisfied. Here, the Named Plaintiffs worked as slaughterhouse supervisors and commodities traders. There are also hundreds, if not thousands, of managers and supervisors who evaluated putative class members and made promotion and other relevant decisions during the relevant time period. Thus, the court finds that the purported class is comprised of a large group of diverse and differently situated employees whose highly individualized claims of discrimination do not lend themselves to class-wide proof. This diversity weighs heavily against certification.
Finally, Peterson's statistics do not establish a general practice of intentional discrimination. Peterson's statistics weigh against certifying a class that includes multiple BUs, job families, and geographies, because Peterson's data shows the wide statistical variation from BU to BU, from job family to job family, and from geography to geography. When examined individually, the BUs, job families, and geographies have no commonality whatsoever. Instead, they range from statistically significantly "unfavorable" to statistically significantly "favorable." Further, Peterson's analysis weighs against finding commonality because Peterson admits that the disparities he sees in performance ratings, promotions, compensation, and terminations may be explained by actual performance differences. Although Plaintiffs claim that Peterson was justified in omitting actual performance differences because this variable was allegedly tainted, Peterson's report does not support such a conclusion. Thus, the Court accords less weight to Peterson's report than Topel's because Peterson omitted performance data.
Accordingly, the Court finds that the Plaintiffs have not demonstrated that a company-wide policy or practice fostering discrimination affected the putative class in a consistent way. Instead, Plaintiffs have only established that they have individual claims of discrimination. Where the proposed class is spread across various BUs and was subject to various decision-making authority affecting promotions, compensation, and terminations, class treatment is not appropriate. Accordingly, the Court finds that Plaintiffs have not satisfied the "commonality" requirement on this basis.
c. Plaintiffs' "Failure to Remedy" Claim
Finally, Plaintiffs assert that Cargill's alleged failure to remedy the effect of its culture satisfies commonality. Specifically, Plaintiffs contend that Cargill failed to effectively exercise its centrally-managed diversity and EEO functions throughout the purported liability period.
First, Plaintiffs assert that before and throughout the liability period, Cargill established a pattern of acknowledging systemic discrimination, proposing specific remedial measures, and failing to adopt these measures. In support of this contention, Plaintiffs assert that in 1995, North American Operations President and current CEO Warren Staley sent a memo identifying steps Cargill was required to take in order to comply with a federal government agency's requirements in response to a company-wide "glass ceiling" audit. According to Plaintiffs, however, there is no evidence that these directives were followed. Plaintiffs also point to an EEO statement issued by former CEO MacMillan, in which MacMillan promised that Cargill would evaluate managers and supervisors based on their contributions to Cargill's achievement of its EEO objectives and that personnel decisions would be systematically audited to determine whether EEO practices are carried out in practice. Plaintiffs further allege that although MacMillan's successor, Ernest Micek, testified that this policy continued throughout his tenure as CEO, neither promise has been kept.
Plaintiffs also assert that HR personnel recommended that Cargill require job postings for certain types of jobs, but that Cargill has not mandated universal job posting. Finally, Plaintiffs point to a 1997 legal analysis conducted of Cargill's past and present salaried workforce that assessed Cargill's potential risks in the area of EEO/diversity. Plaintiffs contend that in response to this analysis, a diversity committee and two Cargill executives made certain suggestions to the company's most senior executives regarding measuring progress in advancing and retaining minority employees and evaluating managers on achieving goals in these areas. Plaintiffs point out that these recommendations were not adopted.
Second, Plaintiffs assert that a pattern of failing to address discrimination continued after Cargill launched the "Valuing Differences Initiative" ("VDI"). In May 1998, Cargill launched VDI in response to the results of the 1997 analysis. Plaintiffs point to a memo in which Cargill personnel warned that the hiring and promotion processes needed improvement in order to increase utilization of women and racial minorities. Plaintiffs contend, however, that Cargill did nothing to improve its promotion process relating to women or minorities in response to these warnings. Plaintiffs further contend that managers made numerous recommendations to measure diversity progress under VDI, but that Cargill did not adopt any of these recommended measures. Plaintiffs also contend that Bielby's report shows that Cargill's alleged hostility, lack of commitment, and repeated failure to follow through sent the message that diversity was not really valued or a priority at Cargill. In conclusion, Plaintiffs assert that this evidence of Cargill's awareness of the alleged pattern of race discrimination and its centralized failure to follow repeated recommendations designed to address and begin to remedy this pattern bears directly on commonality.
Cargill, on the other hand, contends that Plaintiffs' allegations concerning EEO accountability are not a proper substitute for proof of intentional discrimination. Cargill contends that Plaintiffs offer no evidence that Cargill took "inadequate" EEO accountability measures, let alone that it did so out of an intent to discriminate against its African-American employees. In support of its position, Cargill contends that its EEO policies and its workplace diversity programs are state of the art. Cargill also contends that Bielby ignores the differences between intent and impact. In particular, Cargill contends that Bielby does not suggest that Cargill's alleged failure to "monitor" or "measure" the results of its efforts concerning minority representation reflects or constitutes intentional race discrimination.
The Court finds that Plaintiffs' contention that Cargill's alleged failure to effectively exercise centralized control of its EEO function does not establish commonality. Although Plaintiffs point to various recommendations that were not adopted, Plaintiffs offer no evidence that Cargill took inadequate EEO accountability measures, let alone that it did so out of an intent to discriminate against African-American employees. Instead, in response to concerns, Cargill created VDI and sponsored the Ebony Council, a corporate initiative to gain feedback from Cargill's African-American employees. Further, Plaintiffs admit that, in response to VDI, Cargill asked BU Leaders in their annual "contract" with the CEO beginning in 2002-03, "what meaningful steps are you taking in the area of valuing differences?" (Schaefer Decl., ¶ 6, Px # 947.) Cargill also implemented various EEO policies and requirements. For example, Cargill requires BUs to create affirmative action plans in compliance with federal law, requires use of selection grids and recommends law department review for group RIFs, recommends that performance-improvement efforts be made prior to terminations for cause, and audits and benchmarks its HR policies and practices against the best practices of other international companies. Plaintiffs do not disagree that Cargill has such policies, but rather contends that Cargill did not do enough to measure the results of such policies on African-American employees. The Court agrees with Cargill that although Bielby contends that more EEO monitoring is a good idea, failing to collect such data is not discriminatory and does not constitute evidence of intentional discrimination. Accordingly, the Court finds that Plaintiffs have not satisfied the "commonality" requirement on this basis.
3. Typicality
The typicality rule requires that the claims or defenses of the representative parties be typical of the claims or defenses of the class. Paxton, 688 F.2d at 561. This requirement is generally satisfied if the claims or defenses of the representatives and the members of the prospective class are based on the same legal or remedial theory. Paxton, 688 F.2d at 561-62.
Plaintiffs contend that their legal theory — "that Cargill violated and continues to violate § 1981 through the impact of its culture on the centralized HR systems at issue and its failure to address the resulting systemic discrimination" — is the same as that of the class they seek to represent. (Mem. in Supp. of Pls.' Mot. for Class Certification (Am.) at 37.) Cargill, on the other hand, contends that Plaintiffs' claims are not typical because they do not arise from the same event, practice, or course of conduct. Instead, Cargill asserts that Plaintiffs have aggregated numerous individual complaints that require individualized proof and give rise to individualized defenses. Cargill further contends that the named Plaintiffs' testimony adds no support to their allegation of uniform discriminatory "systems" because each one of their claims concerns specific incidents such as specific denials of promotions, specific comments allegedly made by specific supervisors, and specific ratings provided by specific reviewers.
Having reviewed the affidavits and declarations of the named Plaintiffs and the prospective class members, the Court finds that the named Plaintiffs do not have the same or similar grievances as the proposed class members. Although Plaintiffs share the common position of being African-American salaried employees, their allegations are distinct and individualized. Their claims are not susceptible to generalized proof or defenses. Plaintiffs were employed in different BUs, in different geographical areas, and in different jobs. Plaintiffs seek to represent individuals who worked in different BUs and geographical areas, and who reported to numerous autonomous decision-makers. Thus, Plaintiffs' claims do not lend themselves to class treatment. Accordingly, Plaintiffs have not satisfied the "typicality" requirement.
4. Adequacy
The fourth requirement of Rule 23(a) provides that plaintiffs show that the representatives will fairly and adequately protect the interests of the class. In re Potash, 159 F.R.D. at 692. In order to satisfy this requirement, Plaintiffs must show that: (1) the representatives and their attorneys are able and willing to prosecute the action competently and vigorously; and (2) each representative's interests are sufficiently similar to those of the class that it is unlikely that their goals and viewpoints will diverge. In re Potash, 159 F.R.D. at 692. Otherwise stated, adequate representation turns upon the qualifications and experience of the plaintiffs' counsel to conduct the litigation and whether the plaintiffs have any interests antagonistic to the class. In re Potash, 159 F.R.D. at 692.
Regarding counsel's adequacy, there is no dispute that the Plaintiffs' counsel are experienced in the representation of plaintiff classes, particularly in the area of employment discrimination claims, and have the means to litigate this matter as a class action.
Regarding the representatives' adequacy, Plaintiffs assert that they are challenging the same unlawful conduct and seeking the same relief as the remainder of the class. Specifically, Plaintiffs contend that their right to relief, like that of the absent class members, depends on proving that Cargill engaged in a policy, pattern, or practice of discrimination against African-American salaried employees by paying them less than similarly-situated peers, failing to promote and develop them in the same manner as these peers, and terminating them far more frequently than these peers. Cargill, on the other hand, asserts that the named Plaintiffs are inadequate representatives for three primary reasons. First, Cargill asserts that, to the extent that the putative class alleges that the policies and practices of supervisory employees are discriminatory and hinder the promotion, advancement, and compensation of non-supervisory employees, such supervisory and non-supervisory employees do not belong in the same class. Second, Cargill asserts that a significant conflict of interest exists between current and former employees in a class seeking both monetary and injunctive relief. Third, Cargill asserts that this litigation was "law-firm inspired" because virtually all of the Named Plaintiffs were recruited by Plaintiffs' counsel or counsel's agent. (Cargill's Mem. in Opp. to Pls.' Mot. for Class Certification at 45.)
The Court finds that Plaintiffs cannot establish that each representative's interests are sufficiently similar to those of the class so that it is unlikely that their goals and viewpoints will not diverge. Here, named Plaintiffs supervised African-American employees, including putative class members, and/or made employment-related decisions of the type complained of in this lawsuit with respect to other African-Americans. Similarly, the named Plaintiffs were supervised by putative class members, including other named Plaintiffs, and/or they were decision makers with respect to decisions about which other putative class members complain. The named Plaintiffs even accuse other African-Americans, and fellow members of the putative class of engaging in discriminatory behavior.
Accordingly, the Court finds that Plaintiffs have not satisfied the "adequacy" requirement with regard to the class representatives. The Court need not reach Cargill's contentions that a conflict of interest exists between current and former employees in a class seeking both monetary and injunctive relief and that the Plaintiffs' counsel inspired this lawsuit.
Having failed to meet the prerequisites to the maintenance of a class action, the Court need not reach whether this action falls within the categories identified in Fed.R.Civ.P. 23(b). See Allen v. City of Chicago, 828 F. Supp. 543, 554 (N.D. Ill. 1993).
Conclusion
Accordingly, IT IS HEREBY ORDERED THAT:
1. Defendant's Amended Motion to Strike Expert Reports and Opinions of Ronald L. Adler, William T. Bielby, and David W. Peterson (Doc. No. 431) is GRANTED IN PART AND DENIED IN PART AS FOLLOWS:
a. Defendant's Motion to Strike the Expert Report and Opinions of Ronald L. Adler is GRANTED.
b. Defendant's Motion to Strike the Expert Reports and Opinions of William T. Bielby and David W. Peterson is DENIED.
2. Plaintiffs' Motion for Class Certification (Doc. No. 250) is DENIED.