Earl v. Nielsen Media Research, Inc.
The modern template for comparator-based pretext: a 59-year-old recruiter fired for a paperwork mistake defeated summary judgment by showing that significantly younger recruiters with similar violations kept their jobs — and that the company skipped, for her alone, the disciplinary step its own HR chief called required for “consistency.”
Earl v. Nielsen Media Research, Inc. (9th Cir. 2011) 658 F.3d 1108
Federal authority — with subsequent-development note. A Ninth Circuit panel decision adjudicating California FEHA claims in diversity: binding on federal courts in the circuit, persuasive (not binding) in California state courts, and unusually useful there because it applies the same FEHA framework state courts use. Note: Earl recites Godwin’s “specific and substantial” standard for circumstantial pretext evidence while expressly flagging that Cornwell v. Electra Central Credit Union (9th Cir. 2006) 439 F.3d 1018, 1029–1031, questioned that standard’s continued viability after Desert Palace, Inc. v. Costa (2003) 539 U.S. 90 — see the Significance section.
Discrimination
FEHA — age discrimination
Comparator evidence / procedural deviation
Summary judgment reversed
In brief. Nielsen fired Christine Earl, 59, for recording the wrong house address on a recruitment form, citing her history of policy violations. The Ninth Circuit reversed summary judgment on her FEHA age discrimination and wrongful termination claims: recruiters aged 36 to 42 with similar — and arguably more serious — violation histories were not fired, one of them precisely because the head of HR wrote that termination without a formal warning “would not be consistent with our procedure,” a step Nielsen skipped for Earl. (Earl v. Nielsen Media Research, Inc. (9th Cir. 2011) 658 F.3d 1108, 1113–1114, 1117–1118.) The opinion is the circuit’s most practical guide to building a comparator record, and it holds that age comparators need not be outside the over-40 protected class — only “significantly younger.” (Id. at pp. 1116–1117.)
By Jonathan J. Delshad
Founder & Editor-in-Chief
Facts
Christine Earl worked more than a dozen years as a Membership Representative — a “recruiter” — for Nielsen Media Research in Northern California, signing up households with specified demographics to host the monitoring devices behind Nielsen’s television ratings. Nielsen hired her in 1994, when she was 47. (Earl, supra, 658 F.3d at p. 1110.)
Her difficulties began in August 2005, when she violated a policy forbidding recruiters from leaving gifts at unoccupied households; after a verbal warning and a company-wide reminder e-mail, she violated the gift rule again in January 2006, and the next month violated a policy requiring recruiters to carry a company map — answering a supervisor’s question about how she signed a home without one with a single word: “Magic?” (Earl, supra, 658 F.3d at p. 1110.) Nielsen responded in February 2006 by placing her on a Developmental Improvement Plan (“DIP”) — “an informal, nondisciplinary tool that Nielsen uses to notify an employee that his or her performance fell below company standards” — which is distinct from a Performance Improvement Plan (“PIP”), “part of Nielsen’s disciplinary process.” Earl’s DIP warned that future failures “may result in the implementation of the disciplinary process”; a PIP warns of “further disciplinary action up to and including termination.” “At no point during her time at Nielsen did Earl receive a PIP.” (Id. at pp. 1110–1111.)
Her annual review for the year ending August 31, 2006 was, on the whole, favorable: her supervisor recorded “a sign average of 1.7,” a “basic rate” that “was an outstanding 72%,” noted the DIP and areas needing improvement, and concluded: “Overall Christine had a good year with her production and she is always consistent with signing homes.” (Earl, supra, 658 F.3d at p. 1111.) In September 2006 Earl was diagnosed with peripheral neuropathy, a progressive hereditary nerve condition, and told coworkers and her supervisor about it. (Ibid.)
The terminal event came in October 2006, on assignment in Texas: Earl obtained a demographically qualifying household’s consent at 319 Lake Forest Drive but mistakenly wrote 327 Lake Forest Drive on the signed form and entered the wrong address into Nielsen’s system, violating the address-verification policy. When a technician arrived at the wrong house the following month, the owner objected — and the technician “located 319 Lake Forest Drive two doors down the street and successfully installed the equipment there.” (Earl, supra, 658 F.3d at p. 1111.) Nielsen learned of the mistake in December 2006 and terminated Earl in January 2007. She was 59. “In the months before and after Earl’s termination, Nielsen hired five new recruiters for her region: four in their 20s, and one in his early 30s,” one of whom filled her position; the new recruiters were paid “a salary less than half Earl’s salary.” (Ibid.)
Procedural history
In October 2007 Earl sued Nielsen in California Superior Court, alleging age and disability discrimination under the FEHA and wrongful termination in violation of public policy; Nielsen removed on diversity grounds. (Earl, supra, 658 F.3d at p. 1111.) In September 2009 the district court granted summary judgment across the board. It found Earl had established a prima facie case of age discrimination — protected class, adverse action, a satisfactory evaluation months before termination, replacement by a substantially younger employee — and that Nielsen had articulated a legitimate reason (multiple policy violations), but concluded Earl’s pretext evidence failed because her comparators had not committed the identical violation and because two of them, being over or near 40, fell within the protected class. (Id. at pp. 1111–1112, 1114, 1116.) The Ninth Circuit reversed as to the age discrimination and wrongful termination claims, affirmed as to disability discrimination (unbriefed and therefore waived), and remanded. (Id. at pp. 1110, 1118.)
Issue
The “central dispute” was single and dispositive: did Earl present sufficient evidence that Nielsen’s stated reason for firing her — the address error atop a history of policy violations — was a pretext for age discrimination? (Earl, supra, 658 F.3d at p. 1112.) That question subdivided into three: whether recruiters with similar but not identical violations can serve as comparators; whether comparators must fall outside the over-40 protected class, or need only be significantly younger; and whether Nielsen’s failure to give Earl the pre-termination PIP it gave younger recruiters — as policy or as practice — independently evidenced pretext. (Id. at pp. 1113–1118.)
Holding
Reversed as to the FEHA age discrimination and wrongful termination claims. Earl “presented specific and substantial evidence that Nielsen did not terminate significantly younger recruiters with similar histories of multiple policy violations,” raising a triable issue of pretext. (Earl, supra, 658 F.3d at p. 1114.) Comparators must be “similar in material respects,” not identical, and here the court could “fail to see any material difference” between Earl’s violation and those of her comparators — one of whom made the same address-verification mistake two months later with a worse record and was not fired. (Id. at pp. 1114–1115.) In an age case, “comparison with younger employees within the protected class is not improper as a matter of law”; the proper inquiry is whether the comparators “are significantly younger than Earl.” (Id. at p. 1116.) Independently, Earl raised a genuine dispute that Nielsen deviated from its normal disciplinary procedure — terminating her without the PIP its own head of human resources described as required for consistency — or at minimum “applied a more forgiving disciplinary process for younger recruiters,” which “raises a triable issue of pretext, regardless of the company’s formal policy or practice.” (Id. at pp. 1117–1118.) Summary judgment on the disability claim was affirmed as waived; the wrongful termination claim survives with the FEHA claim. (Id. at p. 1118.)
Reasoning
The FEHA framework, in federal court. Because Earl’s claims arose under California law, the panel began with the FEHA: it “prohibits employers from discharging or dismissing any employee over 40 years old based on the employee’s age” (Gov. Code, §§ 12926(b), 12940, subd. (a)), and “[b]ecause state and federal employment discrimination laws are similar, California courts look to federal precedent when interpreting FEHA,” applying the McDonnell Douglas burden-shifting test to disparate treatment claims. (Earl, supra, 658 F.3d at p. 1112, citing Guz v. Bechtel National, Inc. (2000) 24 Cal.4th 317, 354.) The court’s summary judgment posture was equally explicit: “[S]ummary judgment should be used prudently in [age discrimination] cases involving motivation and intent,” and “[w]e require very little evidence to survive summary judgment in a discrimination case, because the ultimate question is one that can only be resolved through a searching inquiry — one that is most appropriately conducted by the factfinder, upon a full record.” (Ibid., quoting Coleman v. Quaker Oats Co. (9th Cir. 2000) 232 F.3d 1271, 1282, and Lam v. Univ. of Hawaii (9th Cir. 1994) 40 F.3d 1551, 1564.) With the prima facie case and the employer’s articulated reason both established, everything turned on step three — and the ultimate burden of persuasion remained with the plaintiff. (Id. at p. 1112, citing Reeves v. Sanderson Plumbing Prods., Inc. (2000) 530 U.S. 133, 143.)
The pretext standard — recited with a visible asterisk. Quoting Chuang v. University of Cal. Davis (9th Cir. 2000) 225 F.3d 1115, 1127, the court restated the two paths: directly, “by showing that unlawful discrimination more likely than not motivated the employer,” or indirectly, “by showing that the employer’s proffered explanation is unworthy of credence because it is internally inconsistent or otherwise not believable.” (Earl, supra, 658 F.3d at pp. 1112–1113.) Earl had no direct evidence — no supervisor “comments . . . betraying bias or animus against older workers” — so the circumstantial standard governed: such evidence must be “specific” and “substantial” under Godwin v. Hunt Wesson, Inc. (9th Cir. 1998) 150 F.3d 1217, 1222. But the panel immediately flagged the standard’s contested status — “But see Cornwell v. Electra Cent. Credit Union, 439 F.3d 1018, 1029–31 (9th Cir. 2006) (questioning the continued viability of Godwin after Desert Palace, Inc. v. Costa, 539 U.S. 90, 100 . . . )” — and softened it: “That standard is ‘tempered’ by our observation that a plaintiff’s burden to raise a triable issue of pretext is ‘hardly an onerous one.’” (Id. at p. 1113, quoting Noyes v. Kelley Servs. (9th Cir. 2007) 488 F.3d 1163, 1170.) Of Earl’s four pretext arguments, the court reached only two: comparators and procedural deviation. (Ibid.)
The comparator showing. “A plaintiff may raise a triable issue of pretext through comparative evidence that the employer treated younger but otherwise similarly situated employees more favorably than the plaintiff” — an idea as old as McDonnell Douglas itself, which called “[e]specially relevant” any “evidence that white employees involved in acts against [the employer] of comparable seriousness . . . were nevertheless retained or rehired.” (Earl, supra, 658 F.3d at p. 1113, quoting McDonnell Douglas, 411 U.S. at p. 804.) Earl’s discovery produced three Nielsen recruiters, identified by number. Employee 33071 (age 37): DIPs in 2005 and 2006 for asset maintenance and signing a home outside required demographics — then, in December 2006, the same address-verification failure as Earl, two months after Earl’s, for which she “either received a PIP or suffered no disciplinary consequence at all” and was not terminated. (Id. at p. 1113.) Employee 36082 (ages 39–40): three incorrect-demographics violations in nine months, forcing removal of a monitoring device for “corrupted” data — and after a “final warning” threatening termination, his third violation earned him a PIP, not discharge. (Id. at p. 1114.) Employee 46432 (age 42): wrong demographics and inaccurate data in his first three weeks, plus chronic lateness, inflated overtime, a “trashed” company car, and a fabricated appointment — his supervisor reporting “[t]here have been too many to list” and asking permission to fire him. The head of HR, Bob Burns, refused: “As much as it sounds reasonable to terminate him without a PIP, it would not be consistent with our procedure. My recommendation is to place him on a Final Warning PIP immediately for all these things, and the next time he fails to meet our expectations, terminate him.” (Ibid.) On that record, “[t]his evidence raises a triable issue of pretext”: Earl had shown “specific and substantial evidence that Nielsen did not terminate significantly younger recruiters with similar histories of multiple policy violations.” (Ibid.)
“Similar in material respects” — not one-to-one identity. The district court had demanded an exact match between Earl’s violation and the comparators’. The panel rejected that as a matter of law: comparators must “have similar jobs and display similar conduct” (Vasquez v. County of Los Angeles (9th Cir. 2003) 349 F.3d 634, 641), “need not be identical, but must be similar in material respects” (Hawn v. Executive Jet Management, Inc. (9th Cir. 2010) 615 F.3d 1151, 1157), with materiality a context-dependent question of fact that “cannot be mechanically resolved” — and, quoting the Seventh Circuit, “[i]t is not an unyielding, inflexible requirement that requires near one-to-one mapping between employees.” (Earl, supra, 658 F.3d at pp. 1114–1115.) Four points made the claimed distinctions immaterial here. Employee 33071 had committed the same violation with a worse record. Nielsen itself had characterized the comparators’ violations as similar — its HR manager, answering Earl’s discovery, canvassed for discipline imposed for “signing the wrong household,” and the court could “fail to see any material difference” between “signing the wrong home” and “enrolling the wrong household for equipment installation.” (Id. at p. 1115.) All the policies at issue served the same data-integrity purpose. And the violations were of “comparable seriousness” — Nielsen’s own internal memorandum called a demo-mismatch removal “a very serious matter,” and Earl’s error, unlike her comparators’, caused neither corrupted data nor a device removal: “viewed in the light most favorable to her, their violations were more serious.” (Id. at pp. 1115–1116.) The court also disposed of Nielsen’s same-supervisor argument — “It was error for the district court to impose a strict ‘same supervisor’ requirement.” (Id. at p. 1115, quoting Hawn.) — and closed with the reminder that similarity “is ordinarily a question of fact.” (Id. at p. 1116.)
Age is relative: comparators need only be “significantly younger.” The district court had reasoned that because employee 36082 turned 40 in the relevant period, his retention “refutes rather than supports” Earl’s claim, and Nielsen argued a 42-year-old “cannot be a proper comparator as a matter of law.” The panel disagreed in a passage with doctrinal reach beyond the Ninth Circuit: “Rigid insistence that a comparator be a member of the protected class overlooks a key difference between age and other forms of discrimination. Whereas sex and race discrimination rely on an individual’s membership in a particular class, age discrimination is relative.” (Earl, supra, 658 F.3d at p. 1116.) Drawing on O’Connor v. Consolidated Coin Caterers Corp. (1996) 517 U.S. 308, 312–313 — the ADEA “does not ban discrimination against employees because they are aged 40 or older; it bans discrimination against employees because of their age,” and a replacement’s being “substantially younger . . . is a far more reliable indicator of age discrimination” than class membership — the court extended the logic from the prima facie stage to pretext, noting that “neither Nielsen nor the district court offered any reason why the logic of O’Connor should not apply with equal force to pretext.” (Id. at pp. 1116–1117; see Begnal v. Canfield & Assocs., Inc. (2000) 78 Cal.App.4th 66, 73.) Earl, 59, could therefore measure herself against recruiters aged 36 to 42. (Id. at p. 1117.)
Deviation from procedure — or a double standard; either way, triable. Finally, “[a] plaintiff may also raise a triable issue of pretext through evidence that an employer’s deviation from established policy or practice worked to her disadvantage.” (Earl, supra, 658 F.3d at p. 1117, citing Diaz v. Eagle Produce Ltd. Partnership (9th Cir. 2008) 521 F.3d 1201, 1214.) Earl was terminated after a single informal DIP, never having received a PIP — while Burns had blocked the termination of a far worse-performing younger recruiter precisely because skipping the PIP “would not be consistent with our procedure,” and later explained that he required the PIP so the company “can demonstrate that we treated him no differently than we have anyone else,” since he “assumed that [employee 46432] is a protected class individual, and so consistency is of utmost importance.” (Ibid.) “Earl has raised a triable issue why consistency was not of similar importance when Nielsen terminated her — another ‘protected class’ employee — seven months later.” (Ibid.) Nielsen answered with its written at-will policy and testimony that Burns’s e-mail misstated policy; the court held the conflict itself created “a genuine dispute whether issuing a PIP prior to a termination constitutes an internal company practice or procedure” — and then supplied an independent alternative: “[e]ven if issuance of a pre-termination PIP was not Nielsen’s formal policy,” the evidence showed “a more forgiving disciplinary process for younger recruiters,” which “raises a triable issue of pretext, regardless of the company’s formal policy or practice.” (Id. at pp. 1117–1118.) The disability claim, unargued on appeal, was waived; and because the FEHA age claim survived, “so too does her claim for wrongful termination in violation of public policy.” (Id. at p. 1118, citing Stevenson v. Superior Court (1997) 16 Cal.4th 880, 897.)
Significance
Earl is the working manual for the most common pretext theory in age cases. Three of its holdings do recurring duty. First, the materiality standard for comparators — similar, not identical; a question of fact; no same-supervisor requirement; and, most usefully, measured in part by the employer’s own characterizations, since Nielsen’s discovery responses and internal e-mails grouping the violations together defeated its litigation position that they differed. (Earl, supra, 658 F.3d at pp. 1114–1116.) Second, the relativity holding: comparators (like replacements) need only be significantly younger, not under 40 — a rule that aligns Ninth Circuit FEHA practice with O’Connor and with California authority such as Begnal, and that forecloses a mechanical defense argument still made routinely. (Id. at pp. 1116–1117.) Third, the procedural-deviation holding, with its important fallback: even where the employer disputes that the skipped step was formal “policy,” selectively lenient process for younger employees is itself specific and substantial evidence of pretext. (Id. at pp. 1117–1118.)
For California litigators the case has a further layer of utility: it is a FEHA decision. The panel applied Guz’s framework, cited California comparator authority, and carried the wrongful-termination tort along with the statutory claim under Stevenson — so its reasoning transplants into state-court oppositions with minimal translation, and comparator-driven reversals in California’s appellate courts continue in the same vein (see, e.g., Wawrzenski v. United Airlines, Inc. (2024) 106 Cal.App.5th 663 [comparator and me-too evidence in a FEHA summary judgment reversal]). Finally, Earl is a load-bearing citation in the “specific and substantial” debate: it is the decision that both recites Godwin’s standard and openly cross-references Cornwell’s questioning of it after Desert Palace, resolving the tension in practice by declaring the standard “tempered” and the plaintiff’s burden “hardly an onerous one.” (Earl, supra, 658 F.3d at p. 1113.) Since Earl satisfied even the stricter formulation, the case did not resolve which standard governs — it demonstrated instead that a well-built comparator record makes the debate academic.
Key quotes
“That standard is ‘tempered’ by our observation that a plaintiff’s burden to raise a triable issue of pretext is ‘hardly an onerous one.’” (Earl, supra, 658 F.3d at p. 1113, quoting Noyes v. Kelley Servs. (9th Cir. 2007) 488 F.3d 1163, 1170.)
“Whereas sex and race discrimination rely on an individual’s membership in a particular class, age discrimination is relative. The proper inquiry is not whether the other recruiters are outside the protected class, but whether they are significantly younger than Earl.” (Earl, supra, 658 F.3d at p. 1116.)
“As much as it sounds reasonable to terminate him without a PIP, it would not be consistent with our procedure.” (Earl, supra, 658 F.3d at p. 1114 [internal Nielsen HR e-mail regarding a younger comparator].)
“We conclude that the more lenient disciplinary process Nielsen used for younger recruiters thus raises a triable issue of pretext, regardless of the company’s formal policy or practice.” (Earl, supra, 658 F.3d at p. 1118.)
Practice pointer
Earl is a discovery plan disguised as an opinion. To build the comparator record: (1) serve interrogatories and document requests keyed to the policy’s purpose, not the precise infraction — every employee disciplined in the relevant period for violations of the same family of rules (here, all “proper collection and verification of household information” policies) — and resist employer attempts to narrow to the identical violation; (2) demand the employer’s internal communications about those disciplinary decisions, because e-mails like Burns’s (“not . . . consistent with our procedure”; “consistency is of utmost importance”) are the difference between argument and evidence; (3) capture how the employer itself described and grouped the violations when it was not litigating — HR searches, memoranda, warning letters — and hold it to those characterizations; (4) chart every comparator’s age, violations, discipline sequence (DIP/PIP-equivalents), and outcome, and frame similarity as the fact question Hawn and Earl say it is; and (5) do not concede comparators over 40 — the test is “significantly younger,” and a 17-to-23-year gap carried the day here. Pair the comparator showing with procedural deviation: identify the step the employer skipped for your client, and if the employer denies the step was “policy,” pivot to Earl’s fallback — selective leniency for younger employees suffices “regardless of the company’s formal policy or practice.” (Earl, supra, 658 F.3d at pp. 1114–1118.) For employers: discipline sequencing is evidence; if HR articulates a consistency norm for one protected employee, expect it to be enforced — in litigation — for all of them.
Open questions
Earl decided sufficiency at summary judgment; whether Nielsen’s explanation was actually pretextual, and whether the comparators were in fact similarly situated, remained jury questions. (Earl, supra, 658 F.3d at pp. 1114, 1116.) The panel expressly declined to reach Earl’s other two pretext theories — shifting explanations and statistical evidence of age-skewed hiring — leaving undecided what quantum of either would independently defeat summary judgment on such a record. (Id. at p. 1113.) It quantified nothing about “significantly younger”: the 36-to-42-year-old comparators were 17 to 23 years Earl’s junior, and the opinion offers no minimum gap — a question that continues to generate line-drawing disputes. It left the “specific and substantial” standard’s post-Desert Palace status exactly where it found it — recited, questioned via the Cornwell cross-reference, and “tempered,” but not resolved. (Id. at p. 1113.) And because the disability claim was waived on appeal, the opinion says nothing about how Earl’s progressive neuropathy — disclosed to her supervisor months before termination — would have fared under the FEHA’s disability provisions, or how overlapping age-and-disability theories should be sequenced at summary judgment.
