A group of retired City of San Francisco employees challenged the City’s disability retirement benefit calculations under the San Francisco Employees’ Retirement System (SFERS). The lawsuit originated as a class action filed by Joyce Carroll in 2017. The parties agreed to a stipulated class certification order which included the requirement that members of the class retired because they were “incapacitated for performance of duty because of disability determined by the retirement board to be of extended and uncertain duration.”
The plaintiffs, all of whom met the disability requirement, and were at least 40 years old when hired and had between 10 and 22.222 years of credited service at retirement. The operative complaint asserted FEHA claims based on disparate treatment (intentional discrimination) and disparate impact (facially neutral policy with unequal effects), along with related claims for declaratory relief, breach of contract, and equal protection violations under the California Constitution.
They focused on “Formula 2,” a provision in the City’s Charter that imputes additional years of service up to age 60 for employees whose actual service yields less than 40% of their average final compensation under Formula 1 (which multiplies average final compensation by 1.8% times years of service). Plaintiffs claimed this formula disproportionately disadvantaged older hires by capping their benefits below the 40% maximum that younger entrants could more easily achieve.
At a four-day bench trial, plaintiffs’ expert, Jeffrey Petersen, Ph.D., presented hypothetical arithmetic calculations showing that employees entering SFERS at age 40 or older could never reach the 40% benefit cap under Formula 2, assuming continuous service, while those entering at 37 or younger could. He did not use actual employee data or perform statistical analysis.
The City’s expert, Dubravka Tosic, Ph.D., criticized this approach, emphasizing the need for real-world data on the entire SFERS population, including factors like service breaks, reciprocity, and purchased credits, which could alter outcomes. Tosic also noted that considering average final compensation (which rises with age) and extending analysis to Formula 1 scenarios showed no overall age-based disadvantage.
The trial court ruled for the City. On disparate treatment, it found no adverse employment action (as benefits were calculated per a fixed formula) and no discriminatory animus, concluding Formula 2 was motivated by pension eligibility (credited service years), not age, akin to the U.S. Supreme Court’s reasoning in Kentucky Retirement Systems v. EEOC, 554 U.S. 135 (2008). For disparate impact, the court deemed plaintiffs’ hypothetical evidence insufficient, as it lacked actual statistical disparities across the protected group. Other claims failed derivatively, with no breach of contract (benefits matched Charter promises) and no equal protection violation (rational basis satisfied).
The Court of Appeal affirmed the trial court in the published case of Carroll v. City & County of S.F. -A169408 (November 2025).
Reviewing factual findings for substantial evidence and legal conclusions de novo, it upheld the disparate treatment ruling, agreeing pension status – not age – drove Formula 2, supported by parallels to Kentucky Retirement Systems (e.g., tracking normal retirement rules like the 60/10 provision, uniform ex ante terms, non-stereotypical assumptions, and scenarios benefiting older workers).
Plaintiffs’ “inexorable zero” argument failed, as hypotheticals ignored real variables like service breaks. For disparate impact, the court confirmed the need for actual statistical proof of disproportionate effects, which plaintiffs lacked. The amendment denial was not reversible error, as the trial court’s decision was not solely based on the pleading variance but on broader evidentiary shortcomings.
The judgment was affirmed, with no costs awarded, reinforcing that retirement formulas tied to pension eligibility, even if correlated with age, do not inherently violate FEHA absent proof of discriminatory motive or actual adverse impact.