Opinion
No. 3:12-cv-01334-CRB
01-15-2016
GREGORY P. BARNES, ET AL., Plaintiffs, v. THE HERSHEY COMPANY, Defendant.
ORDER GRANTING DEFENDANT'S PARTIAL SUMMARY JUDGMENT MOTION ON PATTERN AND PRACTICE AND DISPARATE IMPACT CLAIMS
Plaintiffs argue that Hershey employed an ageist "slot blocker" policy designed to replace older employees with younger workers within the company. See Opp'n (dkt. 295). Plaintiffs further argue that this policy culminated in a 2009 Career Planning Meeting that assigned employees Talent Indicator Rankings. See id. The three remaining Plaintiffs here—who did not sign group terminations waivers—allege pattern and practice, disparate impact, and disparate treatment claims premised on these events. See id.; see also Group Waiver Order (dkt. 276). This Order addresses the pattern and practice and disparate impact claims; another Order addresses Hershey's separate Partial Summary Judgment Motion on Plaintiffs' Individual Claims (dkt. 289). See Order on Individual Claims (dkt. 338).
For the following reasons, the Court concludes that Plaintiffs have not produced sufficient statistical or anecdotal evidence to establish a pattern and practice or disparate impact claim. Plaintiffs have also failed to establish the "causation" element of their disparate impact claim. The Court thus GRANTS Hershey's Partial Summary Judgment Motion on the Pattern and Practice and Disparate Impact Claims. See Motion (dkt. 288).
I. FACTUAL AND PROCEDURAL BACKGROUND
These facts are viewed and presented in the light most favorable to Plaintiffs, the non-moving party. See Villiarimo v. Aloha Island Air, Inc., 281 F.3d 1054, 1061 (9th Cir. 2002).
A. Hershey Employment Structure
Hershey's Domestic Sales Department is made up of a Retail Division and a Category and Customer Division. See Declaration of Brian J. Malloy in Support of Plaintiffs' Motion for Partial Summary Judgment ("April Malloy Decl.") (dkt. 237) Ex. 9 (Magnuson Depo.) at 14-15. Plaintiffs were Customer Sales Executives ("CSEs") and Category Development Managers ("CDMs"), which are Category and Customer Division management positions in pay grades 206 to 213, although the Plaintiffs here were in pay grades 206-210. See April Malloy Decl. Exs. 46 and 47 (Data Disks). Hershey employees in these positions manage from one to six clients and are responsible for developing client relationships through direct interaction with their accounts and creating and communicating sales plans. April Malloy Decl. Ex. 8 (Burkett Depo.) at 33-35.
B. Hershey's Alleged Ageist "Slot Blocker" Policy
Viewing the record in the light most favorable to Plaintiffs for purposes of this motion, beginning as early as 2003, Hershey employed a "slot blocker" policy designed to "manage out," or terminate, longer-tenured workers who were unwilling to relocate for jobs in different cities at Hershey's request. See, e.g., Bolle Decl. (dkt. 299). A number of witnesses provide evidence stating that these "slot blockers" were older workers who were fired and replaced with younger workers. Id. Plaintiffs rely in large part on the testimony of Plaintiff David Bolle to establish that Hershey employed an ageist slot blocker policy. Id. Bolle worked in the Hershey Sales Department's upper management as a "Manager Category Development" from September 2004 until June 2009. Bolle Decl. ¶ 6.
Plaintiffs also reference excerpts from statements made by Hershey executives in earlier age discrimination lawsuits to establish that they employed a "slot blocker" policy. See Opp'n (dkt. 295) at 4-5. Hershey responds that the executives' testimony is inadmissable because it took place years ago and did not involve the "same subject matter" at issue here. See Hub v. Sun Valley Co., 682 F.2d 776, 777 (9th Cir. 1982). The Court previously denied Plaintiffs' request to depose these executives in this action because they lacked "personal knowledge" regarding the relevant subject matter and time periods, see Order (dkt. 190); Order (dkt. 202), so this Order does not consider that evidence. Id. ("The decision whether to admit a deposition from a prior lawsuit is vested in the district court's sound discretion.")
Bolle declares that he personally observed Hershey's management (a) target older workers based on their age, (b) refer to those workers as "slot blockers" in meetings and conference calls, and (c) "manage out" those employees by giving false performance reviews that were a pretext for age discrimination. See id. ¶¶ 9-10. Bolle states that he heard high-level Hershey Sales Department executives—including Ben Stoffel (Director of West Grocery Customers), Mike Weinstock (Vice President), Mike Weldon (Director of Category Management), Scott Cole (Vice President of Food Customers), and Tom Smuda (Vice President of North American Sales)—refer to older workers as "slot blockers." Id. ¶ 10.
According to Bolle, a high-level manager told him that "older people were not able to get the job done, not [] able to keep up," and that the manager wanted "younger up-and-coming people" on his team. Id. ¶ 14. Bolle also states that he heard Hershey executives joke about putting older employees "out to pasture." Id. ¶ 10. Finally, Bolle declares that Hershey's management directed him to "manage out" Greg Barnes, one of Bolle's co-Plaintiffs, because Barnes was "old and slow" and a "slot blocker." Id. ¶¶ 21, 23. Bolle declares that Hershey would demand that older workers uproot their families and relocate, and that Hershey refused to reassign these older workers to positions that did not require them to relocate. Bolle Decl. ¶¶ 7-9.
Bolle testified that the term "slot blocker" could refer to someone who was unwilling to relocate and take a job in a different city at Hershey's request. Bolle Deposition (dkt. 290-9) at 23:3-7. When asked whether it was "possible to be a slot blocker and under the age of 40," Bolle responded affirmatively. Id. at 8-12. Bolle also declares that upper management would examine how long CSEs and CDMs had been in their respective jobs—examining an employee's length of tenure in a position to determine which employees were "slot blockers." Bolle Decl. ¶ 12.
The record contains evidence relating to a "slot blocker" policy that extends beyond Bolle's statements. Barnes declares that he heard managers in his chain of command discuss retirement with two of his colleagues in their 50's and heard at least one manager use the term "slot blocker." Barnes Decl. ¶¶ 47-49. Barnes, however, also answered "yes" to the question "is it possible to be a slot blocker and be under the age of 40?" Barnes Deposition (dkt. 290-14) at 47:6-11.
A Hershey's employee named James Bombeck declares that his manager used the term "slot blocker" in his presence, terminated him, and replaced him with a younger worker. Bombeck Decl. (dkt. 302) ¶¶ 9-20. Another employee named Mary Wasson reported to management and human resources that the company was targeting older CSEs as "slot blockers," but she declares that her report was never investigated. Wasson Decl. (dkt. 301) ¶¶ 36-41. Jerry Chapman, another Hershey salesperson, declares that his managers repeatedly asked him about his retirement plan and then finally forced him to retire. Chapman Decl. (dkt. 295) ¶¶ 52-53. Finally, employee Lori DeLaRue declares that her managers repeatedly referenced her tenure with the company before firing her and replacing her with a younger worker. DeLaRue Decl. (dkt. 303) at ¶¶ 15-22.
C. Hershey Implements a Talent Indicator Ranking Program
In their opposition to the summary judgment motion at issue here, Plaintiffs refer back to their prior filings to describe the background of this employment discrimination action. See Plaintiff's Opposition (dkt. 295) at 3 ("A list of the cast of characters, including job titles, and essential Sales Department career path at Hershey were discussed at Docket 176 at 10:1-14:7 and Docket 236 at 4:22-6:2 and incorporated here by reference.). Therefore, in this subsection, the Court draws from its Order addressing those prior filings. See Order (dkt. 276) at 2-6.
Tom Smuda, Vice President of the Hershey Retail Division, made a presentation on June 23, 2009, to a group of Sales Department executives known as the Sales Leadership Team. See April Malloy Decl. Ex. 93 (Ibberson Depo.) at 8-9; April Malloy Decl. Ex. 44 (Hershey's Response to Plaintiffs' Amended Request for Admissions) Nos. 1-2. The presentation included a series of PowerPoint slides entitled "The Who?" that outlined a wide variety of workforce management challenges, objectives, and suggestions. See April Malloy Decl. Ex. 50 (The Who? Presentation). Among many other points, the presentation reminded managers that "certain roles are developmental or pass through roles and not intended for someone to remain in place for an extended period of time." Id. at 5. The presentation discussed ways in which managers can identify talented employees as well as help their teams gain the experiences and skills they need to move into higher-level positions. See generally id. Hershey was facing difficulty getting qualified employees to apply to higher-level management positions, in part because some employees lacked key career experiences and many were reluctant to relocate, which coupled with a limited number of entry-level 206/207 grade CSE positions due to prior reorganization and headcount reductions. Id. at 2. The presentation proposed that these challenges could be addressed through a range of options, including mentoring, talent development workshops, increasing the client-facing opportunities available to high-potential employees, and "[e]stablish[ing] a term limits program for 206-209/210 positions." Id. at 7-9. The record does not suggest that a term limits program was actually implemented.
Two days later, Robert Ibberson, an executive who reported to Smuda, circulated an email to representatives of various business segments within Hershey about a career planning project. See April Malloy Decl. Ex. 93 (Ibberson Depo.) at 10, 64-76 and Ex. 51 (June 25 Email). The email listed, among other "follow-up" items, that "Steve[n Wojcik] will investigate certain salary grades by organization, and average tenure by grade/organization" and "will communicate some career and performance management BDPs and processes." See April Malloy Decl. Ex. 51 (June 25 Email). The email also scheduled two conference calls for June and July, and Ibberson asked the team members to review Smuda's "The Who?" PowerPoint presentation ahead of an in-person meeting in July. April Malloy Decl. Exs. 51 and 53 (July 10 and 25 Emails).
The following month, in August 2009, Ibberson gave a presentation to the Sales Leadership Team. April Malloy Decl. Ex. 93 (Ibberson Decl.) at 111-12. Slides for the presentation, entitled "The Career Pathing Project," stated among other things that the goal was to "further develop and enhance people strategies for the sales department;" that the team had been "tasked to explore ways to better meet our people needs;" that "cultural, economic and generational trends have limited mobility of employees, resulting in a need for creative solutions;" and that current processes were failing to "develop necessary skills to enable employees to advance within the organization," such that the existing candidate pool lacked the skills needed for higher level positions. April Malloy Decl. Ex. 55 (Presentation Slides) at 3-5. The presentation suggested that Hershey "[u]pdate [the] Career Development Action Plan[, i]ncluding career path timeline, training history, awards and recognition," and that these be "[i]ntegrated into ongoing performance and feedback discussion"; it further suggested the creation of an interactive career website and mentoring program. Id. at 7.
The presentation also called for greater communication and accountability, including "[c]lear and measurable expectations commiserate [sic] with time in role," and that managers should "[c]larify[] time in role expectations" through a "sliding scale of job expectations reflecting time in role" with a detailed list of qualities that reflected each level of performance. Id. at 14-16. One slide stated "Restructure account assignments to allow for more entry level account management positions to enhance the pipeline of talent," and another listed the raw numbers of employees in each paygrade broken down by the amount of time in their current role. Id. at 24-25. The presentation then proposed several "Big Ideas," including a time limit at each paygrade from 206 to "213+," at the end of which the position would be reposted, along with field office hubs, internships, and development tools. Id. at 28-32. Updated presentations on the "Career Planning Project" in November and December 2009 reiterated a focus on "career timelines and greater emphasis on developmental activities," improved mentoring programs and website resources, and enhanced feedback channels for employees. See April Malloy Decl. Ex. 63 (Nov. 2009 Slides) at 4-5 and Ex. 64 (Dec. 2009 Slides) at 2-5.
This preparation led up to a Sales Department Career Planning Meeting on January 26-28, 2010, which approximately fifty Sales Department executives attended. See, e.g., April Malloy Decl. Ex. 69 (Email); April Malloy Decl. Ex. 94 (Zerphy Decl.) at 147. Ahead of the meeting, each supervisor was instructed to evaluate each of his or her direct subordinate employees across a number of dimensions, including time in role, annual performance ratings for the previous two years, a "Talent Indicator" that reflected that employee's "promotability" or ability to take on more advanced roles, what those possible future roles might be, and developmental activities. See April Malloy Decl. Ex. 70 (Jan. 19 Email) at 3 and Ex. 72 (Slides) at 15. The definitions associated with this "Talent Indicator" ranking explained that a ranking of 5 meant the employee had the ability to move up two levels in the company within 3-5 years; a ranking of 4 meant the employee had the ability to move up one level in that time frame; a ranking of 3 meant the employee had high potential but was unable to advance due to external factors, such as relocation limitations; a ranking of 2 meant the employee was an expert in his or her current role but did not have the potential to advance; and a ranking of 1 meant the employee was "developing or unsuccessful." See, e.g., April Malloy Decl. Ex. 74 (Presentation) at Appendix.
By the end of the first day of the meeting, each attendee was told to submit to Human Resources a "Team Scorecard," which were compiled and loaded onto a presentation template for the next day's meeting. See April Malloy Decl. Ex. 74 (Presentation) at Appendix. This "Team Scorecard" was a one-page roster for each supervisor that listed, for each employee that reported to that supervisor, the employee's 2009 performance rating, time in role, the aforementioned "Talent Indicator," and new hires and turnover for the team. See, e.g., id. at 5. Separate slides listed this same information by total numbers (rather than individual names) for divisions as a whole. See, e.g., id. at 7. These evaluations were not limited to CSEs and CDMs in pay grades 206-210, but rather extended to a variety of positions, pay grades, and regions. See, e.g., id. at 7, 10. The Team Scorecards were loaded onto an overhead projector, and each supervisor presented to the attendees about "their talent" and "their development strategies as far as high potentials." See April Malloy Decl. Ex. 93 (Ibberson Decl.) at 155. Although promotions of "high potential" employees were discussed extensively, there is no evidence of any discussion or decisions regarding employee terminations at the meeting. See, e.g., id.; April Malloy Decl. Ex. 74 (Presentation); April Malloy Decl. Ex. 44 (Response to RFA) No. 43.
This meeting was the only time that Hershey ranked its CSEs and CDMs using Talent Indicator Scores; the record does not indicate that management shared these scores with the CSEs and CDMs. See April Malloy Decl. Ex. 93 (Ibberson Depo.) at 94.
D. Terminations in the Years Following the Career Planning Meeting
1. Statistical Data on Terminations
Plaintiffs present evidence regarding the Talent Indicator Rankings assigned to 154 CDMs and CSEs for whom Hershey provided "complete data." See Opp'n at 15 (citing April Malloy Decl. Exs. 45-47, 75). Of those 154 employees, 114 employees fell within the pay grades at issue in this action: pay grades 206-210. Of those 114 employees, 23 workers were assigned a "4 to 5," or promotable, ranking. Id. Only one worker out of those 23 was over 40 years old, indicating that about 95% of the employees assigned "high, promotable rankings in the 206-210 grade level" were workers under age 40. Id.
The summary judgment record further establishes that the percentage of workers over the age of 40 employed by Hershey remained roughly similar through the years in question here—2009 through 2011:
| YEAR | % CSE = 40 | % CSE = 40 | % CDM = 40 | % CDM = 40 |
|---|---|---|---|---|
| 2009 | 56% | 44% | 26% | 74% |
| 2010 | 52% | 48% | 27% | 73% |
| 2011 | 56% | 44% | 21% | 76% |
Plaintiffs offer statistical evidence prepared by Dr. Tom Means to "calculate the statistical significance between the frequency of terminations for employees under 40 and those over 40" for CSEs and CDMs in a number of pay ranges, including the 206-210 pay range at issue here, during the years following the 2009 Career Planning Meeting. Means Decl. (dkt. 297) ¶ 5. The parties do not dispute that Hershey terminated 6 CDMs and 24 CSEs between 2009 and 2011. See Motion (dkt. 288) at 17 (citing dkt. 237-46; 237-49). Hershey employed approximately 8,871 total employees during the years in question. See Smuda Decl. (dkt. 11) at 3. The undisputed record thus shows that Means studied a sample of approximately ten terminations per year over three years. Id.
Means' report offers three conclusions regarding terminations that occurred in the years following the Career Planning Meeting. Means Decl. (dkt. 297). First, the results reported for 2010 "reveal a strong positive and statistically significant relationship between the probability of an involuntary termination and those individuals over 40 years of age," or in "other words, the data support the hypothesis that those employers over age 40 were involuntarily terminated at a higher probability than those under age 40." Id. ¶ 20. Second, there "is an even stronger statistical relationship when the data is restricted to Pay Grades 206-210." Id. ¶ 21. Third, "the results are not dependent on the functional form used." Id. ¶ 22. Fourth, "the results also show a strong negative and statistically significant relationship between the age of an employee and their talent indicator score," or in "other words the data suggest that older employees received lower talent indicator scores on average." Id. ¶ 23.
The undisputed summary judgment record, even viewed in the light most favorable to Plaintiffs, identifies a number of additional facts about Means' report. Means made a number of computational errors in preparing his report. See Means Decl. ¶¶ 37-39. He ran calculations by hand, a practice that he testified he would not have employed in preparing a peer-reviewed report. See Means Depo. (dkt. 290-66) at 67:3-13. Means also included some duplicate records in the data he analyzed. Id. Means used a one-tailed test for statistical significance, defining any disparity greater than 1.65 units of standard deviation as significant. See Means Decl. ¶¶ 31-35. Finally, he found no statistically significant hiring disparities in 2009 or 2012. See Means Depo. (dkt. 290-66) at 89.
Hershey has produced rebuttal statistics, see Siskin Decl. (dkt. 317-1), to which Means has declined to offer a "point-by-point" response, see Means Decl. ¶ 25, leaving a significant amount of statistical evidence undisputed on this record. Plaintiffs do argue that Siskin initially failed to file his report under oath, but he has subsequently cured that defect. See Siskin Aff. (dkt. 317); Jacobsen v. Katzer, No. C 06-1905, 2009 WL 4823021, at *4 (N.D. Cal. Dec. 10, 2009) ("Although Defendants object to the admission of this expert report because it is unsworn, the Court finds the later submission of [a] sworn declaration sufficient to overcome the objection.").
In conducting his analysis, Means did not control for educational background, tenure with the company, or an employee's particular time in pay grade or time in job. See Means Depo. at 30, 108-09. He admitted that he would have preferred to control for these variables, that including such information in his report might have been significant, and that he was under appreciable time constrains in performing his analysis. Id. Additionally, Means did not control for performance history in analyzing the terminations at Hershey.
Plaintiffs respond that Means analyzed the relationship between Talent Indicator Rankings and age—noting that the rankings were based in part on performance history—but Means did not reach a conclusion regarding the relationship between Talent Indicator Rankings and terminations, so Plaintiffs do not meaningfully contest this point. Means Decl. ¶¶ 20-23; Means Depo. at 102.
Hershey's rebuttal expert presents undisputed evidence that controlling for such factors would have (a) undermined the statistical significance of Means' results and (b) allowed Means to account for "duration dependence," or the explanation that
(i) better performing older workers tend not to be in the same grade or job as better performing younger workers, because the better performing older workers have progressed up the career ladder, and (ii) poorer performing younger workers tend not to be in the same grade (level of job) as poorer performing older workers, because the poorer performing younger workers have not yet progressed to the level of their older counter-parts while, at some point, the better performing younger workers catch up with the poorer performing older workers.Siskin Decl. (317-1) at 16.
2. Anecdotal Evidence On Terminations
The record provides anecdotal evidence regarding nine different terminations in the years following the Career Planning Meeting. Plaintiffs reference allegations from a tenth employee in passing, so viewing the record in the light most favorable to Plaintiffs, the Court will assume they present ten anecdotal examples for purposes of this motion. See Reply (dkt. 315) at 6 n.3; Opp'n at 13. Each of the nine terminated employees for whom Plaintiffs have provided substantive information performed successfully at some point during their long careers with Hershey. See, e.g., Barnes Decl. ¶¶ 2-13; Bolle Decl. ¶¶ 2-7, 24, 28, 30, 35, 40; Chapman Decl. (dkt. 300) ¶¶ 2-14; DeLaRue Decl. (dkt. 302) ¶¶ 2-15; Frazier Decl. (dkt. 304) ¶¶ 3-4, 8-13; Nelson Decl. (dkt. 305) ¶¶ 3-5; Wasson Decl. (dkt. 301) ¶¶ 2-14, 25-29; Weeks Decl. (dkt. 306) ¶¶ 4-10; Bombeck Decl (dkt. 302) ¶ 2. Between late 2009 and 2011, these nine employees faced restructuring or declining performance reviews and were either terminated or resigned. See, e.g., Barnes Decl. ¶¶ 13-42; Bolle Decl. ¶¶ 17-18, 20-31; Chapman Decl. ¶¶ 15-73; DeLaRue Decl. ¶¶ 15-21; Frazier Decl. ¶¶ 13-33; Nelson Decl. ¶¶ 6-24; Wasson Decl. ¶¶ 2-3, 14-45; Weeks Decl. ¶¶ 11-23; Bombeck Decl. ¶¶ 10-11. The record indicates that employees Chapman, Wasson, and Ringstad—the employee mentioned in passing by the parties—all made complaints to human resources regarding potential age discrimination, but human resources declined to take action that satisfied them. See Opp'n at 13-14 (discussing Ringstad); Wasson Decl. ¶¶ 36-41; Chapman Decl. ¶¶ 37-40, 45-49.
The parties dispute whether these individual terminations stemmed from age discrimination, as discussed in more detail in the separate Order on Plaintiffs' individual claims and in this Court's prior Orders. See Individual Claims Order (dkt. 338); see also Group Waiver Order (dkt. 276). For purposes of the pattern and practice and disparate impact claims at issue in this motion, it is sufficient to state that Hershey presents evidence that these employees were terminated or resigned due to performance or restructuring issues; Plaintiffs dispute whether Hershey's stated reasons for these employment actions were a pretext for age discrimination. See, e.g., Motion (dkt. 288) at 20; Opp'n (dkt. 295) at 11-12.
Although not relevant here, there are no disputes of material fact regarding one of these plaintiffs'—Chapman's—individual claims. See Individual Claims Order (dkt. 338).
E. Procedural Posture
Plaintiffs filed this action in 2012, alleging that Hershey employed an ageist "slot blocker" policy designed to replace older employees with younger workers. See Opp'n (dkt. 295). Plaintiffs now argue that this policy culminated in a 2009 Career Planning Meeting that assigned employees Talent Indicator Rankings. See id. The three remaining Plaintiffs—who did not sign group terminations waivers—allege pattern and practice, disparate impact, and disparate treatment claims. See id.; Order (dkt. 276). Hershey has filed this Partial Summary Judgment Motion on Plaintiffs' Pattern and Practice and Disparate Impact Claims (dkt. 288). Another Order analyzes Hershey's separate Partial Summary Judgment Motion on Plaintiffs' Individual Claims (dkt. 289). See Order (dkt. 338).
For the following reasons, the Court concludes that Plaintiffs have not produced sufficient statistical or anecdotal evidence to establish a pattern and practice or disparate impact claim. The Court further concludes that Plaintiffs have failed to establish the "causation" element of their disparate impact claim. The Court thus GRANTS Hershey's Partial Summary Judgment Motion on the Pattern and Practice and Disparate Impact Claims.
II. LEGAL STANDARD
The Court can grant a motion for summary judgment "if the movant shows that there is no genuine dispute as to any material fact and the movant is entitled to judgment as a matter of law." Fed. R. Civ. Proc. 56(a). A principal purpose of summary judgment "is to isolate and dispose of factually unsupported claims." Celotex Corp. v. Catrett, 477 U.S. 317, 323-24 (1986). A dispute is genuine "if the evidence is such that a reasonable jury could return a verdict" for either party. Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 248 (1986). A fact is material if it could affect the outcome of the suit under the governing law. Id. at 248-49 (quoting First Nat'l Bank of Ariz. v. Cities Serv. Co., 391 U.S. 253, 288 (1968)). Moreover, if the evidence presented is "merely colorable" and not "significantly probative," the court may decide the legal issue and grant summary judgment. Id. at 249-50 (citations omitted). To determine whether a genuine dispute as to any material fact exists, the court must view the evidence in the light most favorable to the non-moving party. Id. at 255.
In determining whether to grant or deny summary judgment, it is not a court's task "to scour the record in search of a genuine issue of triable fact." Keenan v. Allan, 91 F.3d 1275, 1279 (9th Cir. 1996) (internal citation omitted). Rather, a court is entitled to rely on the nonmoving party to "identify with reasonable particularity the evidence that precludes summary judgment." See id.
III. DISCUSSION
Defendant moves for summary judgment here on two types of claims raised by Plaintiffs: their pattern and practice claims and their disparate impact claims. See Partial Summary Judgment Motion on Plaintiffs' Pattern and Practice and Disparate Impact Claims (dkt. 288). For the following reasons, the Court concludes that Hershey is entitled to summary judgment in its favor on the claims at issue in this motion.
Another Order addresses Hershey's separate Partial Summary Judgment Motion on Plaintiffs' Individual Claims. See Order (dkt. 338).
It is important to note that pattern and practice claims typically arise in the context of class actions. See, e.g., Wal-Mart Stores, Inc. v. Dukes, 131 S. Ct. 2541, 2556 (2011). There is a circuit split regarding whether individual plaintiffs, like Plaintiffs here, may shift the burden of proof by demonstrating a pattern and practice of discrimination. See Christine Tsang, Uncovering Systemic Discrimination: Allowing Individual Challenges to A "Pattern or Practice," 32 Yale L. & Pol'y Rev. 319, 326 (2013). The majority position, adopted by five circuits, "allows pattern or practice suits only when brought by a certified class" given that "the Supreme Court has never before extended Teamsters to an individual, non-class suit." Id. Two circuits nevertheless disagree with the majority position and would permit Plaintiffs to proceed on a pattern and practice theory. Id. Given that the Ninth Circuit has not yet addressed this question, this Order addresses the merits of Plaintiffs' pattern and practice claim.
A plaintiff advancing a pattern and practice claim bears "the initial burden of making out a prima facie case of discrimination." Obrey v. Johnson, 400 F.3d 691, 694 (9th Cir. 2005). Plaintiffs, in proceeding on this theory, must "prove more than the mere occurrence of isolated or accidental or sporadic discriminatory acts," and instead establish that discrimination was Hershey's "standard operating procedure—the regular rather than the unusual practice." Id. (internal quotation marks omitted) (citing Int'l Bhd. of Teamsters v. United States, 431 U.S. 324, 335 (1977)). By "demonstrating the existence of a discriminatory pattern or practice," Plaintiffs would establish "a presumption that [they] had been discriminated against on account of their age. Id.
Plaintiffs may establish a pattern or practice of discrimination wholly or in part through the use of statistical evidence. See Obrey, 400 F.3d at 694. They may also rely on anecdotal evidence to support their pattern and practice claim. Id. at 698 ("Like statistical evidence, anecdotal evidence of past discrimination can be used to establish a general discriminatory pattern in an employer's hiring or promotion practices."). Plaintiffs here offer both statistical and anecdotal evidence to support their claim. See, e.g., Opp'n at 18-20. For the following reasons, the Court concludes that Plaintiffs' proof on their pattern and practice claims fails to establish a prima facie case of discrimination.
1. Plaintiffs' Statistical Proof Fails as a Matter of Law
"In a case in which the plaintiff has alleged that his employer has engaged in a pattern or practice of discrimination, statistical data is relevant because it can be used to establish a general discriminatory pattern in an employers hiring or promotion practices." Obrey, 400 F.3d at 694. Plaintiffs' statistical proof here—based on Means' reports and statements—fails to create a triable issue of material fact under Ninth Circuit case law for three reasons: (a) the sample size underlying Means' report is too small; (b) Means fails to account for obvious variables; and (c) Means' conclusions are undermined by the relevant facts and circumstances of the case.
a. Inadequate Sample Size
The Ninth Circuit has stated that "statistical evidence derived from an extremely small universe . . . has little predictive value and must be disregarded" because "slight changes in the data can drastically alter the result." Aragon v. Republic Silver State Disposal Inc., 292 F.3d 654, 663 (9th Cir. 2002), as amended (July 18, 2002). The undisputed record here shows that Means—whose report represents Plaintiffs' statistical proof of a pattern and practice of discrimination—studied the link between age and the terminations of about ten employees per year over three years. See Motion (dkt. 288) at 17 (citing dkt. 237-46; 237-49). The Ninth Circuit addressed a similar situation in Palmer v. United States, 794 F.2d 534, 539 (9th Cir. 1986), concluding that an ADEA plaintiff failed to establish a triable issue of fact on a disparate impact claim because his statistical analysis only studied a sample of 13 employees in one year and 15 employees in another. See id. The court stated that the plaintiff's "statistical proof is unpersuasive; his sample size is too small to have any predictive value and we must disregard it," further noting that "[s]tatistics used to establish a prima facie case are not irrefutable and their effectiveness depends on all of the surrounding facts and circumstances." Id.
Means' report specifically breaks down its analysis of disparity and statistical significance calculations by year. See, e.g., Means Decl. at 5-6.
Plaintiffs respond that because there were about 150 employees in the overall pool of CSEs and CDMs, Means' report analyzing the small number of terminations at issue here relied on an appropriate sample size. This argument fails for three reasons. First, Means testified that his report analyses the "probability of termination for someone over 40," explaining that "someone" referred to "those that were terminated"—a group of around ten persons per year for the years studied. Means Decl. at 84:16-24. Second, Plaintiffs do not support their argument with any citations to Ninth Circuit authority. See Group Termination Reply (dkt. 261) at 10-11. The Court thus notes Aragon's statement that small sample sizes have "little predictive value" because "slight changes in the data can drastically alter the result" and concludes that criticism applies to the small number of terminations studied here. See 292 F.3d at 663. Third, the record contains unrebutted expert testimony stating that Means' analysis was "limited due to the small sample size." See Siskin Decl. at 6; see also Means Decl. at 8-11 (addressing Siskin's report and omitting any response to the sample size criticism).
Plaintiffs do not substantively respond to Hershey's sample size argument in their opposition briefs here—the only explanation that the Court can find for Plaintiffs' decision to analyze a set of ten terminations per year in a company that employees approximately 8,871 employees, see Smuda Decl. (dkt. 11) at 3, is in a reply brief to an earlier summary judgment motion on group terminations. See Group Termination Reply (dkt. 261) at 10-11.
Furthermore, the First Circuit has squarely rejected the population size argument that Plaintiffs advance here, affirming summary judgment against a plaintiff who presented statistics "based on the dismissal of only five employees . . . over two years from an average annual employee population . . . of approximately 215 people" because "[s]uch a small statistical sample carries little or no probative force to show discrimination." See LeBlanc v. Great Am. Ins. Co., 6 F.3d 836, 849 (1st Cir. 1993) (collecting cases that reject similarly small sample sizes). Given the undisputed record, the Court concludes that Plaintiffs' statistical evidence fails due to its sample size and "must be disregarded." Palmer, 794 F.2d at 539; Aragon, 292 F.3d at 663; see also Shutt v. Sandoz Crop Prot. Corp., 944 F.2d 1431, 1433 (9th Cir. 1991); Sengupta v. Morrison-Knudsen Co., 804 F.2d 1072, 1075-76 (9th Cir. 1986); Morita v. S. Cal. Permanente Med. Grp., 541 F.2d 217, 220 (9th Cir. 1976).
b. Failure to Account for Obvious Variables
Statistical analysis purporting to show a disparity in employment discrimination cases is "not enough to take [a] case to trial" if it fails "to account for obvious variables," which can include "education, previous position at the company, and distribution of age groups by position"—"that would have affected the results of the analysis." Coleman v. Quaker Oats Co., 232 F.3d 1271, 1283 (9th Cir. 2000). The Court concludes that Means failed to consider at least two obvious variables that would have affected the results of his analysis. See id.
First, given that Means did not link Talent Indicator Rankings to terminations in his report, the Court concludes he failed to control for terminated employees' performance histories, see Means Decl. ¶¶ 20-23; Means Depo. at 102, even though "job performance" may have been the variable "most likely to have offered a legally appropriate explanation of why certain employees were selected for lay-off." See Pottenger v. Potlatch Corp., 329 F.3d 740, 748 (9th Cir. 2003). The Ninth Circuit has concluded that such "statistical analysis is insufficient to raise a triable issue of fact." Id.
Means admits that he could have separately analyzed the pay grades at issue here, but Plaintiffs' counsel asked him to lump pay grades 206-210 together, so he did not account for such a variable. See Means Depo. at 46, 83. Hershey's rebuttal expert presents undisputed evidence that controlling for pay grade would have allowed Means to account for the "basic statistical concept" of "duration dependence," which identifies that
(i) better performing older workers tend not to be in the same grade or job as better performing younger workers, because the better performing older workers have progressed up the career ladder, and (ii) poorer performing younger workers tend not to be in the same grade (level of job) as poorer performing older workers, because the poorer performing younger workers have not yet progressed to the level of their older counter-parts while, at some point, the better performing younger workers catch up with the poorer performing older workers.Siskin Decl. (317-1) at 16. The undisputed record shows that if Means had controlled "for an employee's job and or pay scale, the results confirm that there is no valid statistical evidence that age is a factor in the termination decision process." Id. at 20; Means Decl. at 8-11 (addressing Siskin's report and omitting any response the "duration dependence" criticism). Given the undisputed evidence that "duration dependence" is a "basic statistical concept," the Court concludes that Means failed "to account for obvious variables . . . that would have affected the results of the analysis." See Coleman, 232 F.3d at 1283. Thus, under Ninth Circuit case law—and independent of the sample size problem discussed above—Plaintiffs' statistics are "not enough to take [the] case to trial." Id.
The Ninth Circuit relied on what the expert here calls "duration dependence" to affirm the grant of summary judgment for an ADEA defendant in Rose v. Wells Fargo & Co. See 902 F.2d 1417, 1423 (9th Cir. 1990). The Court reasoned that a "statistical disparity in regards to termination rates" identified by the plaintiff there was "explained by the fact older persons tend to occupy [higher] management positions." See id.
c. Surrounding Facts and Circumstances
Finally, "[s]tatistics used to establish a prima facie case are not irrefutable and their effectiveness depends on 'all of the surrounding facts and circumstances.'" Palmer, 794 F.2d at 539 (citing Teamsters, 431 U.S. at 340). The Ninth Circuit—when presented with deficient statistics like those offered by Plaintiffs and Means here—has examined the percentage of employees over and under 40 who remained in a company's work force before and after the actions challenged in an ADEA lawsuit as "relevant surrounding facts and circumstances." See Rose, 902 F.2d at 1427 ("36.47% of the Crocker work force and 32.09% of the Wells Fargo work force were 40 years or older while 34.8% of the work force was 40 or older after the consolidation").
The summary judgment record here establishes that the percentage of workers over the age of 40 employed by Hershey remained roughly similar throughout the years in question—2009 through 2011:
| YEAR | % CSE = 40 | % CSE = 40 | % CDM = 40 | % CDM = 40 |
|---|---|---|---|---|
| 2009 | 56% | 44% | 26% | 74% |
| 2010 | 52% | 48% | 27% | 73% |
| 2011 | 56% | 44% | 21% | 76% |
Consequently, the Court concludes that Plaintiffs' statistical evidence fails to support a prima facie case, let alone establish pretext, on a pattern and practice theory—a theory which requires them to "prove more than the mere occurrence of isolated or accidental or sporadic discriminatory acts," and instead establish that discrimination was Hershey's "standard operating procedure—the regular rather than the unusual practice." Obrey, 400 F.3d at 694.
2. Plaintiffs' Anecdotal Proof Fails as a Matter of Law
"Like statistical evidence, anecdotal evidence of past discrimination can be used to establish a general discriminatory pattern in an employer's hiring or promotion practices." See Obrey, 400 F.3d at 698. Given the deficiency of Plaintiffs' statistical evidence here, the Court must turn to the anecdotal evidence they have presented. The Court concludes that Plaintiffs' have not produced "the amount of anecdotal evidence claimants in other cases have supplied to support a finding of a pattern or practice of discrimination." King v. Gen. Elec. Co., 960 F.2d 617, 627 (7th Cir. 1992); see Mundy v. Household Fin. Corp., 885 F.2d 542, 546 (9th Cir. 1989).
Plaintiffs have produced ten anecdotal examples of alleged discrimination here. See, e.g., Opp'n at 18-19; Motion at 6-7. Each of the nine terminated employees for whom Plaintiffs have provided substantive information had performed successfully at some point during their long careers with Hershey. Between 2009 and 2011, these nine employees faced restructuring or declining performance reviews and were either terminated or resigned. Hershey presents evidence that these ten employees were terminated or resigned due to performance or restructuring issues; Plaintiffs dispute whether Hershey's reasons for these employment actions were a pretext for age discrimination. See, e.g., Motion (dkt. 288) at 20; Opp'n (dkt. 295) at 11-12; see also Individual Claims Order (dkt. 338).
See, e.g., Barnes Decl. ¶¶ 2-13; Bolle Decl. ¶¶ 2-7, 24, 28, 30, 35, 40; Chapman Decl. ¶¶ 2-14; DeLaRue Decl. ¶¶ 2-15; Frazier Decl. ¶¶ 3-4, 8-13; Nelson Decl. ¶¶ 3-5; Wasson Decl. ¶¶ 2-14, 25-29; Weeks Decl. ¶¶ 4-10; Bombeck Decl. ¶ 2; Opp'n at 13-14 (discussing Ringstad)
See, e.g., Barnes Decl. ¶¶ 13-42; Bolle Decl. ¶¶ 17-18, 20-31; Chapman Decl. ¶¶ 15-73; DeLaRue Decl. ¶¶ 15-21; Frazier Decl. ¶¶ 13-33; Nelson Decl. ¶¶ 6-24; Wasson Decl. ¶¶ 2-3, 14-45; Weeks Decl. ¶¶ 11-23; Bombeck Decl. ¶¶ 10-11; Opp'n at 13-14 (discussing Ringstad)
Although not relevant here, there are no disputes of material fact regarding one of these plaintiffs'—Chapman's—individual claims. See Individual Claims Order (dkt. 338).
Ninth Circuit authority, supported by authority from two other circuits, establishes that these ten anecdotes are insufficient to support a pattern and practice claim that must "prove more than the mere occurrence of isolated or accidental or sporadic discriminatory acts," and instead establish that discrimination was Hershey's "standard operating procedure—the regular rather than the unusual practice." See Obrey, 400 F.3d at 694. The Ninth Circuit considered similar anecdotal evidence in affirming the grant of summary judgment for an ADEA defendant in Mundy. See 885 F.2d at 546. In that case, the court explained that in "a large corporation with branch offices throughout the country, the discharge of seven older employees over a two and a half year period alone does not establish a pattern or practice of discrimination." Id. The Seventh Circuit, in reversing a jury verdict entered in favor of ADEA plaintiffs, similarly concluded that six anecdotal examples of alleged discrimination in one year did not represent "the amount of anecdotal evidence claimants in other cases have supplied to support a finding of a pattern or practice of discrimination." See King, 960 F.2d at 627. The Second Circuit has similarly stated that "when, as here, relevant statistics are lacking," then "seven individual incidents . . . are not sufficient to establish a pattern or practice of discrimination." See Ste. Marie v. E. R. Ass'n, 650 F.2d 395, 405 (2d Cir. 1981).
Here, Plaintiffs present ten anecdotal examples of alleged discrimination that occurred over four separate years in a company that employed approximately 8,871 total employees during the years in question. See Smuda Decl. at 3; see supra notes 12-13. Given the deficient statistical evidence presented by Plaintiffs, their attempt to establish a pattern and practice claim with the anecdotal evidence here is squarely foreclosed by Mundy. See 885 F.2d at 546; see also Ste. Marie, 650 F.2d at 405; King, 960 F.2d at 627.
In the last clause of a footnote to their opposition to Hershey's motion for summary judgment, Plaintiffs' suggest that if their anecdotal evidence is lacking, they should be given leave under Rule 56(d) to conduct additional discovery regarding terminations. See Opp'n at 14. The discovery they reference has already been the subject of prior Orders in this case, see, e.g., Discovery Orders (dkt. 182) at 5; Discovery Order (dkt. 235) at 1-2 ("Here, the Court finds that plaintiffs had more than ample opportunity to obtain the information they now seek, and their requests are unreasonably cumulative."), and Plaintiffs' have failed to identify any reason to revisit those rulings. See, e.g., Blough v. Holland Realty, Inc., 574 F.3d 1084, 1091 (9th Cir. 2009).
B. Disparate Impact Claim
Plaintiffs also bring a disparate impact claim; the Court concludes that it fails for many of the same reasons that Plaintiffs' pattern and practice claim fails. "A disparate impact claim challenges employment practices that are facially neutral in their treatment of different groups but that in fact fall more harshly on one group than another and cannot be justified by business necessity." Pottenger, 329 F.3d at 749. To "establish a prima facie case of disparate impact, the plaintiff must: (1) identify the specific employment practices or selection criteria being challenged; (2) show disparate impact; and (3) prove causation." Rose, 902 F.2d at 1424. The plaintiff must "isolat[e] and identify [] the specific employment practices that are allegedly responsible for any observed statistical disparities." Stockwell v. City & Cty. of San Francisco, 749 F.3d 1107, 1114 (9th Cir. 2014). A plaintiff "must also demonstrate a causal connection between those specific employment practices and the asserted impact on those of a particular age." Id. "Statistical disparities must be sufficiently substantial that they raise such an inference of causation." Rose, 902 F.2d at 1424.
Plaintiffs' disparate impact claim here fails for two reasons. First, as stated earlier, the undisputed summary judgment record indicates that Plaintiffs' statistical analysis fails to establish evidence of "statistical disparities." See id. Second, Plaintiffs have not established a causal link between the only "specific employment practice[]" they have described—the purported use of Talent Indicator Scores to implement a "Slot Blocker" policy—and any of the terminations at issue here. See Rose, 902 F.2d at 1424.
Hershey also contends that Plaintiffs' disparate impact claim fails because Plaintiffs omitted a disparate impact theory from their EEOC charges. See B.K.B. v. Maui Police Dep't, 276 F.3d 1091, 1099 (9th Cir. 2002) ("Allegations of discrimination not included in the plaintiff's administrative charge may not be considered by a federal court unless the new claims are like or reasonably related to the allegations contained in the EEOC charge."). But the Ninth Circuit "construe[s] the language of EEOC charges with utmost liberality," see id., and given the language contained in the charges about a "blocker policy," see Opp'n at 21, the Court will reach the merits of Plaintiffs' disparate impact claim here.
Plaintiffs' expert did not reach a conclusion regarding the relationship between Talent Indicator Rankings and terminations. Means Decl. ¶¶ 20-23; Means Depo. at 102. Furthermore, although promotions of "high potential" employees were discussed extensively at the Career Planning Meeting, there is no evidence of any discussion or decisions regarding employee terminations at the Career Planning Meeting. See, e.g., id.; April Malloy Decl. Ex. 74 (Presentation); April Malloy Decl. Ex. 44 (Response to RFA) No. 43.
Plaintiffs' only response is that the Court, in a prior summary judgment order, conditionally stated "for purposes of that motion" that "directly or indirectly related to [the Career Planning Meeting], Plaintiffs were identified as being long-tenured in their roles with little ability to advance and were terminated." See Opp'n at 25; Order (dkt. 276) at 20. Even if that conditional statement from the Court's prior Order applied with force for purposes of this separate motion—and it does not—the statement does not establish any causal link between the Talent Indicator Scores and employee terminations. See id. The Court concludes that Plaintiffs' have failed to establish either "statistical disparities" or "causation," and thus their disparate impact claim fails as a matter of law. See Stockwell, 749 F.3d at 1114; Rose, 902 F.2d at 1424.
The Court's conditional statement indicates that two separate events occurred, chronologically, after the Career Planning Meeting: Plaintiffs were identified as being long-tenured in their roles, and later, for various reasons, they were terminated. See Opp'n at 25; Order (dkt. 276) at 20. --------
IV. CONCLUSION
For the foregoing reasons, the Court concludes that Plaintiffs have not produced sufficient statistical or anecdotal evidence to establish pattern and practice or disparate impact claims. The Court further concludes that Plaintiffs have failed to establish the "causation" element of their disparate impact claims. The Court thus GRANTS Hershey's Partial Summary Judgment Motion on the Pattern and Practice and Disparate Impact Claims.
IT SO ORDERED. Dated: January 15, 2016
/s/_________
CHARLES R. BREYER
UNITED STATES DISTRICT JUDGE