Carroll v. City and County of S.F.

• Court: California Court of Appeal
• Decided: January 14, 2026
• Docket: A169408M
• Status: Published

Opinion

Filed 12/5/25 (unmodified opn. attached)

CERTIFIED FOR PUBLICATION

IN THE COURT OF APPEAL OF THE STATE OF CALIFORNIA

FIRST APPELLATE DISTRICT

DIVISION FOUR

JOYCE CARROLL et al., Plaintiffs and Appellants, A169408

v. (San Francisco City and County CITY AND COUNTY OF SAN Super. Ct. No. CGC-17-562580) FRANCISCO, ORDER MODIFYING OPINION; NO Defendant and Respondent. CHANGE IN JUDGMENT

THE COURT:

It is ordered that the opinion filed herein on November 12, 2025, be modified as follows:

1. On page 12, the first sentence under the subheading “Standard of Review” starting, “Plaintiffs state at times in their briefing that the trial court erred in its rulings at summary judgment and after the bench trial on their substantive FEHA claims, but their challenges in this appeal . . . .” is replaced with:

Plaintiffs state at times in their briefing that the trial court erred in its rulings at summary judgment and after the bench trial on their substantive FEHA claims, but the substantive challenges as briefed in this appeal are to the trial court’s ruling after the bench trial.

There is no change in judgment.

1 The petition for rehearing is denied.

(Brown, P.J., Streeter, J., and Clay, J.* participated in the decision.)

Date: _____________________ ________________________________ P. J.

_______________________ * Judge of the Superior Court of California, County of Alameda, assigned by the Chief Justice pursuant to article VI, section 6 of the California Constitution.

2 Filed 11/12/25 (unmodified opinion) CERTIFIED FOR PUBLICATION

JOYCE CARROLL et al., Plaintiffs and Appellants, A169408 v. CITY AND COUNTY OF SAN (San Francisco City and County FRANCISCO, Super. Ct. No. CGC-17-562580) Defendant and Respondent.

The City and County of San Francisco (the City) permits employees who have completed at least 10 years of service and who become disabled to receive disability retirement benefits. (S.F. Charter, § A8.603-3.) These benefits are set by what the parties in this case refer to as “Formula 1” if the result of Formula 1’s application exceeds a certain percentage of the employee’s average final compensation, and these benefits are calculated according to what the parties refer to as “Formula 2” if the application of Formula 1 does not exceed a certain percentage of the employee’s average final compensation. Plaintiffs contend on appeal that the City’s use of Formula 2 to calculate disability retirement benefits discriminates against them based on their age of entry (40 and above) into the City’s retirement system in violation of the California Fair Employment and Housing Act (FEHA) (Gov. Code, § 12900 et seq.). The main question before us is whether the trial court erred by determining after a bench trial that plaintiffs had not prevailed

1 under the FEHA. For the reasons explained herein, we conclude that the trial court did not err, and we affirm the judgment. BACKGROUND The City’s Retirement System The City operates a retirement system known as the San Francisco Employees’ Retirement System (SFERS). (See S.F. Charter, §§ A8.509(a), A8.584-1, A8.587-1, A8.600-1, A8.603-1.) The City’s Charter makes service and disability retirement benefits available to qualified employees. (Id. at §§ A8.509(a), (c), A8.584-2 & A8.584-3, A8.587-2 & A8.587-3, A8.600-2 & A8.600-3, A8.603-2 & A8.603-3.) Employees qualify for service retirement when they: (1) reach age 50 and have at least 20 years of credited service (the 50/20 provision); (2) reach age 60 and have at least 10 years of credited service (the 60/10 provision); or (3) reach age 65. (S.F. Charter, §§ A8.509(b), A8.584-2, A8.587-2, A8.600-2, A8.603-21.) An employee’s service retirement benefit is calculated by multiplying the retiree’s total years of credited service by a set percentage that correlates with the employee’s age as stated in the charter, then multiplying that figure by the employee’s average final compensation. (Ibid.) Employees qualify for disability retirement when they are deemed “incapacitated for performance of duty because of disability” and have at least 10 years of credited service. (S.F. Charter, §§ A8.509(c), A8.584-3, A8.587-3, A8.600-3, A8.603-3.) A disability retirement pension benefit is calculated by one of two formulas. Under most of the charter provisions at issue, Formula 1 multiplies the employee’s average final compensation by 1.8 percent, then multiplies that figure by the employee’s total credited years of service. (S.F.

1 Charter provision A8.603-2 replaced attainment of the age of 50 with

attainment of the age of 53. (S.F. Charter, § A8.603-2.)

2 Charter, §§ A8.509(c), A8.587-3, A8.600-3, A8.603-3.) If Formula 1 results in a benefit percentage that exceeds 40 percent of the employee’s average final compensation, that percentage is the employee’s disability pension benefit percentage.2 (S.F. Charter, §§ A8.509(c), A8.587-3, A8.600-3, A8.603-3.) If Formula 1 does not yield a disability pension benefit percentage that exceeds 40 percent of the employee’s average final compensation, the charter directs the employee’s disability retirement benefit to be calculated by Formula 2. (S.F. Charter, §§ A8.509(c), A8.587-3, A8.600-3, A8.603-3.) Formula 2 calculates an employee’s disability retirement benefit by multiplying the employee’s average final compensation by 1.8 percent and then multiplying the resulting figure by “the number of years of City service which would be credited to [the employee] were such City service to continue until attainment by him or her of age 60” up to a maximum of 40 percent of the employee’s final average compensation. (Ibid.) The City can employ workers for positions that do not qualify for pension credit. (See e.g., S.F. Admin. Code, § 16.42(b)(3) [part-time employees who work 1,040 hours during any 12-month period become SFERS members].) In certain situations, such as when an employee works part-time,

2 Charter provision A8.584-3 applied to persons who became miscellaneous officers and employees after November 1, 1976, until charter provision A8.587-3 became operative as of November 7, 2000. Charter provision A8-584-3 replaced the 1.8 percent benefit factor with 1.5 percent and the 40 percent with one-third (approximately 33.333 percent). (S.F. Charter, § A8.584-3.) Under this provision, if Formula 1 does not yield a result that exceeds one-third of employee’s average final compensation, Formula 2 calculates the employee’s disability retirement benefit by multiplying the average final compensation by 1.5 percent and then multiplying the resulting figure by “the number of years of City service which would be credited to [the employee] were such City service to continue until attainment by him of age 60” up to a maximum of one-third of the employee’s final average compensation. (Ibid.)

3 when he or she takes unpaid leave, or when he or she does not work a sufficient number of hours in a given year, the employee may not earn a full credited year of service toward their pension for each calendar year of employment. In certain situations, an employee may purchase service credit when he or she retires, including for prior service as a temporary City employee, redeposited miscellaneous plan service, unpaid parental leave prior to July 1, 2003, qualifying public service, military service, and union representative service. Purchased credit does not change the employee’s SFERS membership date. The Lawsuit Plaintiff Joyce Carroll filed a class action alleging discrimination in violation of the FEHA. The City demurred, arguing that Carroll failed to timely file an administrative charge with the Department of Fair Employment and Housing. (Carroll v. City and County of San Francisco (2019) 41 Cal.App.5th 805, 811, as mod. on den. of rehg. (Nov. 27, 2019) (Carroll).) The trial court sustained the demurrer, and this court reversed. (Ibid.) Plaintiffs filed the operative third amendment complaint (TAC) on August 18, 2020. In addition to alleging a cause of action under the FEHA based on a disparate treatment theory and a cause of action under the FEHA based on a disparate impact theory (hereafter the FEHA claims) and an injunctive relief claim pursuant to Government Code section 12965, the TAC asserts causes of action for declaratory relief, breach of contract, and violation of the equal protection clause of the California Constitution. The TAC further alleges that the City discriminated against plaintiffs and purported class members who were over 40 years of age at the time of hire.

4 The court entered a conditional class certification order, and the parties subsequently agreed to a stipulated class certification order. The class is defined as persons: “i. Who were employed by [the City]; [¶] ii. Who were at least 40 years of age when hired by that employer; [¶] iii. Who had less than 22.222 years of credited service at retirement; [¶] iv. Whose retirement benefits are governed by San Francisco Charter section A8.509, A8.584, A8.587, A8.600, or A8.603; [¶] v. Whose retirement benefits are administered by [SFERS]; [¶] vi. Who has received any payment of retirement benefits from SFERS on or after Nov. 20, 2017; and [¶] vii. Who 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 parties filed cross-motions for summary judgment. Regarding the FEHA claims, the court denied the motions and found triable issues of fact as to discriminatory animus and causation, whether there was an adverse employment action, whether Formula 2 had an actual disproportionate adverse impact on employees over the age of 40, and whether the City had a meritorious business necessity defense. The Bench Trial The court heard testimony from two experts and certain class members in a four-day bench trial. Plaintiffs’ expert, Jeffrey Petersen, Ph.D., testified that he was tasked with assessing whether Formula 2 could result in disparate impact to workers who were age 40 or older when they entered the system. To answer this question, Petersen performed arithmetic, inputting various years of service and ages at SFERS membership into Formula 2 and calculating the resulting benefits. Petersen concluded that Formula 2 produced a disparate impact, and he produced tables reflecting this impact showing his

5 calculations of final disability retirement benefits based on age. According to Petersen, if a class member entered the SFERS between the ages of 18 and 37, “Formula 2 is going to guarantee you a benefit of 40 percent” of final earnings. But older members who enter the system “cannot get a benefit of 40 percent from Formula 2.” Petersen also performed lost income calculations due to the alleged disparate impact for the class members. Petersen testified that, because he calculated the results of an arithmetic formula, he formed an opinion about Formula 2’s disparate impact before he reviewed any class member data. He did not perform any statistical analysis. He confirmed that his tables demonstrating the disparate impact did not depict any retirement benefit percentages actually awarded to employees at or over 40 at time of SFERS membership, nor did they reflect an analysis of the data from the City for the proposed class members. Petersen was not given, and did not look at, actual data for anyone who entered SFERS at below the age of 40 whose retirement benefits were calculated pursuant to Formula 2. He did not evaluate data related to the benefits of all SFERS members who entered the system at age 40 or older, and he did not consider the total monetary payments given to any retirees. Petersen’s calculations assumed that an employee accrued continuous service credit with no breaks in service from the date of SFERS membership to the date of disability. However, he conceded that, based on the class data, there were probably circumstances where an employee could work for the City and not receive service credit, or where an employee may take a leave of absence and not accrue service credit during the leave. He conceded that, in some circumstances, employees could bring over credited years of service from other employers — a concept called reciprocity. If that occurred, Petersen acknowledged that it would “feed into that final percentage, the

6 percentage of final compensation used for [the] benefit.” Petersen conceded that an employee’s total years of credited service would not exactly match the difference between his or her retirement date and date of SFERS membership if the employee “made a purchase or . . . had a break in service.” He acknowledged that the class data showed some class members may have had breaks in service. The City’s expert, Dubravka Tosic, Ph.D. criticized Petersen for focusing on “a sliver” of SFERS members who were 40 and older with 10 to 22 years of credited service. Tosic explained that a disparate impact analysis should look at the “entire population” and “you want to compare what happens with individuals under 40 versus what happens with individuals 40 and over,” but Petersen did not do that analysis. Tosic testified that Petersen performed hypothetical calculations to generate his report and tables, but a disparate impact analysis should look at actual data. She opined that one cannot deduce that a disparate impact occurred from Petersen’s hypothetical calculations. One problem with Petersen’s analysis was that it assumed continuous service after the date of SFERS membership, but her review of the class members’ data showed that that is not always the case. Tosic explained that, under the City’s retirement formulas, the total years of credited service take into account things like reciprocity, part-time work, purchased credit, and individuals who may not have worked during the whole time period and took unpaid leave during which they did not receive credit. “So, total credited service is not always equal to years between SFERS membership date and date of retirement.” With respect to the hypotheticals used by Petersen, Tosic also testified that if a majority of the employees who joined SFERS before age 40 were 38 and 39 and the number of these employees was significantly higher than the number

7 of employees who joined SFERS at age 40 or over, that could be a situation where the difference in final pension benefit percentages would not be statistically significant. Tosic also extended Petersen’s hypothetical calculations and considered individuals whose benefits were calculated under both Formula 1 and Formula 2 because SFERS is “an interconnected pension system that should be looked at as a whole.” She considered persons who had more than 22 years of credited service and found “as years of credited service increases, the benefit percent increases as well.” Under this expanded analysis, some employees who become SFERS members at age 40 or older can receive a benefit percentage that is higher than those who become SFERS members at younger ages. Tosic criticized Petersen for focusing on benefit percentages and ignoring the effect of average final compensation on pension benefits. “Understanding whether or not there may be a disparate impact of the retirement benefit formulas on all members who are 40 or older at [the] time of SFERS membership requires one to look jointly at both components of the benefit formulas” — benefit percent and average final salary. This is important because of the labor economics concept known as the “age earnings profile,” which holds that wages tend to increase with age since workers gain more experience as they grow older. Finally, Tosic testified that, were the City to give all disability retirees a 40 percent benefit, employees who became SFERS members at or over age 40 would receive more than they would receive with a service retirement benefit had they worked until age 60. The Court’s Ruling The parties submitted proposed statements of decision after the trial, and the court issued a tentative decision in favor of the City. Plaintiffs

8 responded to the tentative decision, and the court subsequently issued its statement of decision finding for the City. Regarding the FEHA disparate treatment claim, the court first ruled that plaintiffs had not established an adverse employment action. The court reasoned that adverse employment actions typically involve changes to the terms of conditions of employment, and plaintiffs did not provide authority showing that the award of disability retirement benefits under a set disability pension formula qualified as an adverse employment action. Turning to discriminatory animus and relying on Kentucky Retirement Systems v. EEOC (2008) 554 U.S. 135 (Kentucky Retirement Systems), a case we discuss in further detail post, the court stated: “[T]he Court examined Formula 2’s language and determined that Formula 2 is not discriminatory on its face as there is nothing in Formula 2 that discriminates based on age. (. . . Ky. Ret. Sys. v. E.E.O.C. (2008) 554 U.S. 135, 147 [‘The above factors all taken together convince us that the Plan does not, on its face, create treatment differences that are “actually motivated” by age’].)” The court further found that the City had not judicially admitted, nor was it judicially estopped to deny, that Formula 2 was facially discriminatory. The court then concluded that “[p]laintiffs failed to present evidence that there was discrimination due to their age.” It continued, “At trial, the City argued that Formula 2 is motivated by credited years of service not Plaintiffs’ age at hire. Even if Formula 2 were motivated by both credited years of service and Plaintiffs’ age at hire, Plaintiffs needed to show that discrimination due to Plaintiffs’ age at hire was a substantial motivating factor in the adverse decision. [Citations.] Plaintiffs alleged that Formula 2 discriminates based on age at hire, but at trial, Plaintiffs’ expert examined age at SFERS membership. [Citations.] There is a distinction between date

9 at hire and date at SFERS membership, which Dr. Petersen acknowledged. [Citations.] The Court recognizes that the experts agree that the percentages of benefits awarded under Formula 2 increases with younger employees being provided with a higher percentage benefit than persons who become [SFERS] members at or after age 40. [Citations.] However, Dr. Petersen’s testimony did not support Plaintiffs’ allegations that age at hire is a substantially motivating factor in Formula 2 as at times, Dr. Petersen testified that age at hire did not play a role. [Citations.] The evidence does not show that age at hire substantially motivates Formula 2. Here, like in Kentucky Retirement Systems, Formula 2 appears to be motivated by credited years of service. (Ky. Ret. Sys., supra, 554 U.S. at p. 146 [‘the Plan at issue[] simply seeks to treat disabled employees as if they had worked until the point at which they would be eligible for a normal pension. The disparity turns upon pension eligibility and nothing more’]. . . .)” Regarding the disparate impact claim, the court found: 1) plaintiffs’ evidence was insufficient because they used age at hire and age at SFERS membership interchangeably; 2) plaintiffs’ evidence was insufficient because it was based on hypothetical calculations rather than actual disparities or statistics; and 3) plaintiffs did not show an adverse impact on the “entire protected group: retirees hired by the City at or after age 40.” The court also observed that the City did not appear to have pursued a business necessity defense at trial, and, in any event, the City failed to establish that defense. Briefly addressing damages and causation, the court found plaintiffs provided evidence of damages if they were discriminated against based on age at hire, but plaintiffs did not show Formula 2 was motivated by age at hire and plaintiffs had not been harmed, so plaintiffs had not established causation.

10 The court then denied plaintiffs’ non-FEHA claims. Plaintiffs had maintained that their breach of contract claim was factually derivative of the FEHA claims and the City breached “by paying [class members] reduced disability retirement benefits in an amount substantially motivated by their being age 40 or older at the time they became members of SFERS in violation of FEHA’s superseding law[.]” The court determined there was no breach. The City provided the disability retirement benefits promised in the charter. The court further stated, “As there is no disparate treatment or disparate impact based on age discrimination, the Court will also find that the City did not pay [c]lass [m]embers reduced disability retirement benefits in an amount substantially motivated by their being age 40 or older at the time that they became members of SFERS.” Finally, the court concluded there was no equal protection violation because Formula 2 passed rational basis review. Plaintiffs timely appealed from the judgment.3 DISCUSSION In this appeal, plaintiffs challenge the trial court’s rulings on their FEHA claims and contend the court abused its discretion in failing to grant their request for leave to amend their complaint after trial. We first address plaintiffs’ FEHA challenges and then turn to their claim of error with respect to the court’s denial of their request for leave to amend.

3 The City filed an opposed request for judicial notice of three lawsuits

filed by counsel for plaintiffs in this appeal. We decline to take judicial notice of matters that are not relevant to the disposition of this appeal. (Doe v. City of Los Angeles (2007) 42 Cal.4th 531, 544, fn. 4.)

11 I. The FEHA Claims A. Standard of Review Plaintiffs state at times in their briefing that the trial court erred in its rulings at summary judgment and after the bench trial on their substantive FEHA claims, but their challenges in this appeal are to the court’s ruling after the bench trial. In reviewing a judgment based upon a statement of decision, we review the trial court’s findings of fact for substantial evidence and its legal conclusions de novo. (Thompson v. Asimos (2016) 6 Cal.App.5th 970, 981.) B. Disparate Treatment Plaintiffs argue that the trial court erred by finding no adverse employment action and in its substantial motivating factor ruling. We find that plaintiffs have not established error on appeal with respect to the court’s latter ruling, so we do not address the court’s adverse employment action ruling. 1. Governing Legal Principles “Age discrimination claims can be advanced under a disparate treatment or disparate impact theory. To establish a disparate treatment claim, a plaintiff must prove the defendant intentionally discriminated.” (Mahler v. Judicial Council of California (2021) 67 Cal.App.5th 82, 112 (Mahler).) The plaintiff must prove that he or she suffered an adverse employment action “because of” age. (Gov. Code, § 12940, subd. (a).) Under the FEHA, the discriminatory intent does not need to be proven as the sole motivation behind the challenged action, but the plaintiff must produce evidence that, taken a whole, permits a rational inference that intentional discrimination was a substantial motivating reason for the adverse action. (Harris v. City of Santa Monica (2013) 56 Cal.4th 203, 232.)

12 A plaintiff can prove a disparate treatment claim by direct or circumstantial evidence. (Guz v. Bechtel National Inc. (2000) 24 Cal.4th 317, 354.) Direct evidence is evidence which, if believed, proves the fact of discriminatory animus without inference or presumption (DeJung v. Superior Court (2008) 169 Cal.App.4th 533, 550), and direct evidence can include an employment policy that is discriminatory on its face (Trans World Airlines, Inc. v. Thurston (1985) 469 U.S. 111, 121). 2. Persuasive Federal Authority Plaintiffs view this as a straightforward case of facial discrimination, pointing to the language in Formula 2 calling for the imputation of service years until the retiree would have attained the age of 60. To refute the claim of facial discrimination, the City relies mainly on Kentucky Retirement Systems, supra, 554 U.S. 135, which addressed a disability retirement benefit formula similar to that at issue under the Age Discrimination in Employment Act of 1967 (29 U.S.C. § 621 et seq. (ADEA))4. The trial court also relied on Kentucky Retirement Systems, and, given their common objectives, California courts have relied upon federal cases interpreting the ADEA and title VII of the Civil Rights Act of 1964. (Mahler, supra, 67 Cal.App.5th at p. 112.) We thus begin our analysis with Kentucky Retirement Systems. In Kentucky Retirement Systems, the high court held that Kentucky’s disability retirement plan for hazardous workers (the Kentucky plan) did not violate the ADEA. (Kentucky Retirement Systems, supra, 554 U.S. at p. 138.) Under the Kentucky plan, hazardous workers were eligible for a “ ‘normal’ ”

4 The ADEA states in pertinent part, “It shall be unlawful for an

employer” “to fail or refuse to hire or to discharge any individual or otherwise discriminate against any individual with respect to his compensation, terms, conditions, or privileges of employment, because of such individual’s age.” (29 U.S.C. § 623(a)(1).)

13 retirement pension after 20 years of service or at the age of 55 after five years of service. (Id. at p. 139.) Kentucky calculated a normal retirement pension by multiplying years of service by 2.5 percent, then multiplying that figure by the worker’s final preretirement pay. (Ibid.) A worker who was not eligible for normal retirement when he or she became disabled was eligible for disability retirement if he or she had five years of service or became disabled in the line of duty. (Ibid.) If the disabled worker had less than 20 years of service, Kentucky imputed to the employee the service years that he or she would have worked until the employee attained the age of 55, but the total imputed service years could not exceed the employee’s actual service years, and the sum of the imputed and actual service years could not exceed 20. (Ibid., citing Ky. Rev. Stat. Ann. § 16.582(5)(a) (Baldwin 2003)5.) A hazardous worker who became disabled at 61 after 18 years of service complained of disparate treatment because he did not receive any imputed service years at retirement. (Kentucky Retirement Systems, supra, 554 U.S. at p. 140.) The Equal Employment Opportunity Commission brought an age discrimination suit, and the district court granted summary judgment for the defendants, finding no prima facie case of age discrimination. (Ibid.) The

5 The Kentucky statute stated, “The disability retirement allowance

shall be determined as provided in KRS 16.576, except if the member’s total service credit on his . . . last day of paid employment in a regular full-time position is less than twenty (20) years, service shall be added beginning with his . . . last date of paid employment and continuing to his . . . fifty-fifth birthday. The maximum service credit added shall not exceed the total service the member had on his . . . last day of paid employment, and the maximum service credit for calculating his . . . retirement allowance, including his . . . total service and service added under this section, shall not exceed twenty (20) years.” (Ky. Rev. Stat. Ann. § 16.582(5)(a) (Baldwin 2003).)

14 Sixth Circuit Court of Appeals ultimately reversed sitting en banc, and the high court accepted review. (Id. at pp. 140–141.) The high court began its analysis with Hazen Paper Co. v. Biggins (1993) 507 U.S. 604, wherein a 62-year-old employee alleged age discrimination under the ADEA because his employer had dismissed him to avoid paying pension benefits that were about to vest at 10 years of service. (Kentucky Retirement Systems, supra, 554 U.S. at p. 142.) The plaintiff had to show that age “ ‘actually motivated the employer’s decision,’ ” (id. at p. 141, italics omitted), but Hazen Paper recognized that pension status and age were analytically distinct, and a dismissal based on pension status is not a dismissal “ ‘because of . . . age.’ ” (Kentucky Retirement Systems, at p. 146.) “When the employer’s decision is wholly motivated by factors other than age, the problem of inaccurate and stigmatizing stereotypes disappears. This is true even if the motivating factor is correlated with age, as pension status typically is.” (Hazen Paper, at p. 611.) But Hazen Paper also recognized that discrimination based on pension status could be unlawful where pension status served as a “ ‘proxy for age,’ ” and it left open “ ‘the special case where an employee is about to vest in pension benefits as a result of his age, rather than years of service.’ ” (Kentucky Retirement Systems, at pp. 142–143, quoting Hazen Paper, at p. 613.) Turning to the Kentucky plan at issue, the high court held that the plan was nondiscriminatory because it was motivated by pension status, not age, although the result was that a worker younger than 55 at retirement received imputed service years and a worker aged 55 or older did not. (Kentucky Retirement Systems, supra, 554 U.S. at p. 143.) The court reasoned that the ADEA allowed employers to condition pension eligibility on age (id. at p. 143, citing 29 U.S.C. § 623(l)(1)(A)(i)), and age and pension status

15 remain “ ‘analytically distinct.’ ” (Kentucky Retirement Systems, at p. 143.) The court then assessed whether pension status, “though analytically distinct from age,” could serve as a proxy for age and concluded it could not, based on the following evidence. (Id. at pp. 143–144.) First, Kentucky’s plan provisions were part of “a set of complex systemwide rules” that “involve, not wages, but pensions—a benefit that the ADEA treats somewhat more flexibly and leniently in respect to age.” (Kentucky Retirement Systems, supra, 554 U.S. at pp. 143–144.) And the specific benefit at issue was offered to “all hazardous position workers on the same nondiscriminatory terms ex ante. That is to say, every such employee, when hired, is promised disability retirement benefits should he become disabled prior to the time that he is eligible for normal retirement benefits.” (Id. at p. 144.) The high court also observed that Congress had otherwise approved of programs that calculate permanent disability benefits using a formula that expressly takes account of age. (Ibid.) Second, there was a clear non-age-related rationale for the disparity at issue: The disability rules tracked Kentucky’s normal retirement rules and imputed only those years needed to bring the disabled worker’s years of service to 20 or to the number of years that the individual would have worked had he worked to age 55. (Kentucky Retirement Systems, supra, 554 U.S. at pp. 145–146.) “It is obvious, then, that the whole purpose of the disability rules is, as Kentucky claims, to treat a disabled worker as though he had become disabled after, rather than before, he had become eligible for normal retirement benefit.” (Id. at p. 145.) Third, in some circumstances, Kentucky’s plan worked to the advantage of older workers. (Kentucky Retirement Systems, supra, 554 U.S. at pp. 145–146.) For example, the plan would impute 10 years of service for a

16 45-year-old employee with 10 years of service, but it would impute only five years of service for a 40-year-old employee with 15 years of service. (Ibid.) “[T]hat fact helps to confirm that the underlying motive is not an effort to discriminate ‘because of . . . age.’ ” (Id. at p. 146.) Fourth, the retirement benefit was not grounded in the stereotypical assumptions that the ADEA sought to eradicate. (Kentucky Retirement Systems, supra, 554 U.S. at p. 146.) “The Plan does assume that all disabled workers would have worked to the point at which they would have become eligible for a pension. It also assumes that no disabled worker would have continued working beyond the point at which he was both (1) disabled and (2) pension eligible. But these ‘assumptions’ do not involve age-related stereotypes, and they apply equally to all workers, regardless of age.” (Id. at p. 147.) Fifth, “[t]he difficulty of finding a remedy that can both correct the disparity and achieve the Plan’s legitimate objective—providing each disabled worker with a sufficient retirement benefit . . . further suggests that this objective and not age ‘actually motivated’ the Plan.” (Kentucky Retirement Systems, supra, 554 U.S. at p. 147.) Given this evidence, the court found the Kentucky plan “does not, on its face, create treatment differences that are ‘actually motivated’ by age.” (Ibid.) Because the Equal Employment Opportunity Commission provided no additional evidence that would permit a factfinder to reach a contrary conclusion, the court reversed the appellate court’s judgment. (Id. at pp. 147–148.) The Eighth Circuit Court of Appeals subsequently reached the same conclusion regarding a similar private retirement plan. (Schultz v. Windstream Communications, Inc. (8th Cir. 2010) 600 F.3d 948.) In Schultz, the employer’s pension plan provided for reduced benefits under an “ ‘Early

17 Retirement-50/25’ ” (25 service years at age 50) provision, or unreduced benefits under an “ ‘Early Retirement-30’ ” (30 service years) provision. (Id. at p. 950.) The employer laid off 15 employees as of April 13, 2007, and eight were ineligible for pension benefits. (Ibid.) The employer amended its pension plan to make the eight employees eligible for the first retirement benefit each would have qualified for had he or she been allowed to work until December 31, 2008. (Ibid.) Seven laid-off employees consequently received the unreduced benefit. (Id. at pp. 950–951.) Other laid-off employees who were eligible for the reduced benefit as of the layoff date and one employee who accelerated under the amendment to the reduced benefit sued, alleging differential treatment (zero or less imputed service time) based on age. (Id. at p. 951.) The court rejected the claim. “Just as in Kentucky Retirement Systems, [the employer] imputed years of service to [employees] to make them eligible for early retirement benefits,” and the remaining laid-off employees did not receive imputed time because they were already eligible for pension benefits. (Id. at p. 953.) “[T]he difference in the retirement benefit provided Appellants, as compared to the other seven employees, is based on pension status.” (Id. at pp. 953–954.) 3. Analysis The trial court below found that “credited years of service,” not age, motivated Formula 2. The record shows that the court used “credited years of service” interchangeably with pension status. Specifically, the court wrote, “Here, like in Kentucky Retirement Systems, Formula 2 appears to be motivated by credited years of service. (Ky. Ret. Sys., supra, 554 U.S. at p. 146 [‘the Plan at issue[] simply seeks to treat disabled employees as if they had worked until the point at which they would be eligible for a normal pension. The disparity turns upon pension eligibility and nothing more’][.]”

18 In this case, both Kentucky Retirement Systems and the record support the trial court’s factual determination that pension status, not age, motivated Formula 2. (Thompson v. Asimos, supra, 6 Cal.App.5th at p. 981 [substantial evidence standard of review applies to court’s factual findings].) First, as in Kentucky Retirement Systems, there was “a clear non-age related rationale for the disparity . . . at issue” because Formula 2 tracks the City’s normal retirement rules. (Kentucky Retirement Systems, supra, 554 U.S. at p. 144.) Plaintiffs do not contest the legality of the City’s use of age as a condition of eligibility for normal retirement benefits and instead argue that the imputation of service years until age 60 bears no relationship to the City’s normal retirement thresholds. This is inaccurate. To receive disability retirement, an employee must have at least 10 years of service credit. (S.F. Charter, §§ A8.509(c), A8.584-3, A8.587-3, A8.600-3, A8.603-3.) While the City’s normal retirement rules include 50/20 and 60/10 provisions, the charter connects the criterion of having at least 10 years of service credit to the criterion of having attained the age of 60. (S.F. Charter, §§ A8.509(b), A8.584-2, A8.587-2, A8.600-2, A8.603-2.) The trial court could thus conclude that, like the Kentucky plan, the purpose of Formula 2’s imputation of service years is to treat disabled workers as though they had become disabled after, rather than before, they became eligible for retirement under the normal 60/10 provision. (Kentucky Retirement Systems, at p. 146.) Second, the specific disability retirement benefit at issue here was offered to “all . . . workers on the same nondiscriminatory terms ex ante. That is to say, every such employee, when hired, is promised disability retirement benefits should he become disabled . . . .” (Kentucky Retirement Systems, supra, 554 U.S. at p. 144.) And the California Legislature has used

19 disability retirement formulas like the one at issue for many years. (Gov. Code, §§ 21423, subd. (b), 31727, subd. (b).) Third, in some circumstances, Formula 2 treats older workers similarly or the same as younger workers. As an example, an employee who becomes disabled at 50 who became an SFERS member at 39 but who had a one-year break in service would receive 10 service years, plus 10 imputed service years and a 36 percent final pension benefit, and an employee who becomes disabled at 50 with ten years of continuous service who became an SFERS member at 40 would receive the same number of imputed service years and the same 36 percent final pension benefit. (See S.F. Charter, §§ A8.509(c), A8.587-3, A8.600-3, A8.603-3.)6 And an employee who became an SFERS member at 40 who becomes disabled at 50 and who is able to purchase credit for a total of 22.222 years of service (for work as a temporary City employee, for example) would receive a 40 percent final pension benefit. (Ibid.) Next, Formula 2, like the Kentucky plan, is not grounded in age-related stereotypes. By its terms, Formula 2 assumes that all disabled workers would have worked to the point at which they would have become eligible for benefits under the 60/10 provision, and it assumes no disabled worker would have continued working beyond the point at which he was disabled and eligible for the 60/10 provision benefit. “But these ‘assumptions’ do not involve age-related stereotypes, and they apply equally to all workers, regardless of age.” (Kentucky Retirement Systems, supra, 554 U.S. at p. 147.) Finally, the court could conclude based on the evidence that “[t]he difficulty of finding a remedy that can both correct the disparity and achieve the Plan’s legitimate objective—providing each disabled worker with a

6 We use the charter provisions that allow for a 40 percent maximum

benefit for our examples in this opinion.

20 sufficient retirement benefit . . . further suggests that this objective and not age ‘actually motivated’ the [Formula 2].” (Ibid.) Plaintiffs contend that Kentucky Retirement Systems is not persuasive because the high court “merely applie[d] the ‘reasonable factor other than age’ ” provision of the ADEA (29 U.S.C. § 623(f)(1) (RFOA)),7 and the FEHA does not include a comparable provision. It is true that FEHA does not have a RFOA provision (Mahler, supra, 67 Cal.App.5th at p. 113), but Kentucky Retirement Systems neither cites to, nor purports to apply, the RFOA. “The question before us is whether Kentucky’s system consequently discriminates against the latter workers ‘because of . . . age.’ Age Discrimination in Employment Act of 1967 (ADEA or Act), § 4(a)(1), 81 Stat. 603, 29 U.S.C. § 623(a)(1). We conclude that it does not.” (Kentucky Retirement Systems, supra, 554 U.S. at p. 138.) Plaintiffs’ attempt to differentiate this case from Kentucky Retirement Systems on this basis is therefore unconvincing. Plaintiffs’ contention that Kentucky Retirement Systems is not persuasive because the ADEA and FEHA have different standards to assess

7 This provision states in pertinent part: “It shall not be unlawful for an

employer, employment agency, or labor organization— [¶] (1) to take any action otherwise prohibited under subsections (a), (b), (c), or (e) of this section where age is a bona fide occupational qualification reasonably necessary to the normal operation of the particular business, or where the differentiation is based on reasonable factors other than age, or where such practices involve an employee in a workplace in a foreign country, and compliance with such subsections would cause such employer, or a corporation controlled by such employer, to violate the laws of the country in which such workplace is located[.]” (29 U.S.C. § 623(f)(1).) The high court has explained that this provision “plays its principal role” in disparate impact cases, where the allegedly prohibited act was not based on age. (Smith v. City of Jackson (2005) 544 U.S. 228, 239.) There, the RFOA “preclud[es] liability if the adverse impact was attributable to a nonage factor that was ‘reasonable.’ ” (Ibid.)

21 the employer’s motive is likewise unconvincing. We do not read Kentucky Retirement Systems to be a case that squarely addressed or turned on the differences between “but for” and “substantial motivating factor” causation. The high court ruled that pension status, even when conditioned in part on age, is not the same thing as age (Kentucky Retirement Systems, supra, 554 U.S. at pp. 143–144; see also id. at pp. 145–148), and a disability retirement plan that lawfully makes age in part a condition of pension eligibility and treats workers differently in light of their pension status does not automatically discriminate because of age (id. at p. 143). The high court “follow[ed]” (ibid.) the rule and approach of Hazen Paper, which was not a mixed motive case (Hazen Paper Co. v. Biggins, supra, 507 U.S. at p. 611 [discussing whether employer’s decision was “wholly motivated by factors other than age”]), and held that, even though pension status was conditioned in part on age, pension status — not age — motivated the disability retirement system at issue. (Kentucky Retirement Systems, at p. 143.) The trial court here likewise found that pension status, although conditioned in part on age, motivated the City’s disability retirement system, and the trial court concluded that age was not a substantial motivating factor. As we have explained, this determination was supported by substantial evidence. Plaintiffs further contend that the court erred in its substantial motivating factor determination because: (1) both experts agreed (in reference to Petersen’s calculations) that “ ‘the percentages of benefits awarded under Formula 2 increases with younger employees being provided with a higher percentage benefit than persons who became SFERS members at or after age 40’ ” — evidence plaintiffs claim shows “ ‘the inexorable zero’ ”; and (2) the City purportedly admitted in appellate briefing in Carroll that Formula 2 facially discriminates based on age when the City argued that

22 plaintiffs were on notice of their claims because Formula 2 was publicly available. Neither argument persuades us. First, plaintiffs have not established the “inexorable zero” — an evidentiary showing indicative of intentional discrimination in situations where, for example, a defendant employed zero members of a protected class. (United States v. New York City Department of Education (S.D.N.Y. 2018) 407 F.Supp.3d 365, 418; see Paradise v. Prescott (11th Cir. 1985) 767 F.2d 1514, 1529.) Petersen’s tables show that employees who join SFERS on or before age 37 who retire with at least 10 continuous but less than 22.222 total service years receive a 40 percent final disability pension benefit, and those who join SFERS on or after age 38 who retire in the same service-year range receive a lower final benefit percentage, but his calculations presume continuous service. As set forth above, Formula 2 can yield a final benefit of less than 40 percent for employees who become SFERS members at or before 40 who have breaks in service, and it can yield a 40 percent final benefit in scenarios where an employee becomes a SFERS member at or after 40 and is able to purchase service credit. Further, even in the continuous service scenario assumed by Petersen, not all employees who join SFERS at or before age 40 receive a 40 percent final benefit. Next, the City’s argument in Carroll — in connection with the statute of limitations issue decided there — that plaintiffs were on notice of their discrimination claim because they had the opportunity to obtain knowledge about their injury from the application of Formula 2 is not a concession that the City was liable because Formula 2 facially discriminates based on age. We thus affirm the trial court’s ruling that pension status, not age, motivated Formula 2 for all the reasons stated above.

23 4. Abuse of Discretion Regarding Judicial Admission Ruling Plaintiffs next argue that the trial court abused its discretion in deciding their disparate treatment claim because it failed to recognize that it had the discretion to determine that the City had judicially admitted in Carroll that Formula 2 was facially discriminatory. Specifically, plaintiffs argue that the court did not understand that a representation in an appellate brief could constitute a judicial admission. We find no error. The record here does not affirmatively show that the trial court believed that factual admissions in an appellate brief could not constitute a judicial admission. (Barriga v. 99 Cents Only Stores LLC (2020) 51 Cal.App.5th 299, 334–335 [reviewing court presumes trial court understood scope of its discretion unless record affirmatively shows otherwise].) The court’s statement of decision provides, in relevant part: “Plaintiffs argue that the City has made a judicial admission that Formula 2 is facially discriminatory whereas the City argues that Plaintiffs implicate judicial estoppel to make this argument. The court determines that neither judicial admission nor judicial estoppel applies. On appeal, the City did not argue that Formula 2 was facially discriminatory, rather, what was at issue was a statute of limitations issue and at most, the City repeated [p]laintiffs’ allegations.” (Fns. omitted.) The trial court included a footnote citing a case that recites that statements in complaints, demurrers, answers and cross- complaints can constitute judicial admissions. But the body of the trial court’s order shows that the court declined to find a judicial admission because it concluded that the City did not represent or concede in its appellate briefing in Carroll that Formula 2 was facially discriminatory, not

24 because the court concluded that a statement in an appellate brief can never be a judicial admission.8 C. Disparate Impact Turning to its disparate impact claim, plaintiffs argue that the trial court erred when it erroneously found that (1) their evidence was insufficient because it “was based on hypothetical calculations rather than actual disparities”; and (2) plaintiffs had to show a disparate impact on the “entire protected group: retirees hired by the City at or after age 40.” We find the court’s former ruling to be accurate and dispositive in the circumstances of this case and need not address the latter. “To establish a disparate impact claim, a plaintiff need not prove intent to discriminate, but must prove that ‘regardless of motive, a facially neutral employer practice or policy, bearing no manifest relationship to job requirements, in fact had a disproportionate adverse effect on members of the protected class.’ ” (Mahler, supra, 67 Cal.App.5th at p. 112.) “ ‘[C]laims that stress “disparate impact” . . . involve employment practices that are facially neutral in their treatment of different groups but that in fact fall more harshly on one group than another . . .’ ” (Smith v. City of Jackson, supra, 544 U.S. at p. 239.) “[I]n a disparate impact case, a plaintiff must ‘ “allege[] and prove[], usually through statistical disparities, that facially neutral employment practices adopted without a deliberately discriminatory motive nevertheless have such significant adverse effects on protected groups that they are ‘in operation . . . functionally equivalent to intentional

8 Citing to part of the court’s summary judgment order that

summarizes that plaintiffs had argued the City made a judicial admission in its briefing, plaintiffs represent on appeal that the trial court at summary judgment ruled that the City had made a judicial admission. Not so. The court ruled at summary judgment that there were triable issues of fact.

25 discrimination.’ ” ’ [Citations.] ‘ “ ‘[S]tatistical disparities must be sufficiently substantial that they raise such an inference of causation.’ ” ’ ” (Mahler, at p. 113.) Plaintiffs contend that their evidence was sufficient because it established “that 100% of persons over 40 who are subject to Formula 2 receive a lesser benefit than 100% of those under 40 who are subject to Formula 2,” but that is not the case. Plaintiffs rely on Petersen’s tables to show the impact of Formula 2, but Petersen did not base the calculations that produced these tables on actual class member or employee data. Petersen also did not review actual data for employees who became SFERS members before age 40 who received Formula 2 benefits, and there is no actual data in the record pertaining to this group. Further, Petersen’s tables cannot show that “100% of persons over 40 who are subject to Formula 2 receive a lesser benefit than 100% of those under 40 who are subject to Formula 2” because he makes assumptions — continuous service between the dates of SFERS membership and retirement, and that no employee who joined SFERS at or after age 40 could ultimately have more than 20 years of service credit — that do not necessarily represent reality. The record indicates that not all employees have continuous service after becoming SFERS members. Simply inputting numbers into Formula 2, it is possible for an employee who became an SFERS member at 38 and disabled at 50 to have ten years of credited service due to service breaks, and this employee would receive the same 36 percent final disability pension benefit under Formula 2 as the hypothetical employee who becomes an SFERS member at 40 who works continuously for 10 years and becomes disabled at 50. Employees can also purchase service credit at retirement in certain instances without changing their SFERS membership date. Thus, an

26 employee who becomes a SFERS member at 40 who becomes disabled at 50 but who is able to purchase up to 22.222 years of service credit can receive a 40 percent final benefit under Formula 2. We have no idea how many employees fall into these categories because plaintiffs do not point to data showing how retirement benefits for the class or any comparative group were actually calculated. Stated simply, the trial court correctly recognized on the record before it that plaintiffs failed to prove their disparate impact theory with evidence that Formula 2 “ ‘in fact had a disproportionate adverse effect on members of the protected class.’ ” (Mahler, supra, 67 Cal.App.5th at p. 112.)9 II. The Denial of Leave To Amend A. Additional Background Plaintiffs moved to amend their complaint to change all references to the City discriminating based on “ ‘age at hiring’ ” to “ ‘age at SFERS membership’ ” to conform to proof at trial.10 At the same time, plaintiffs filed their response to the court’s tentative statement of decision wherein they explained that they had moved to amend because it was unclear whether the court sought to deny relief after the bench trial solely because of the pleading and proof variance.

9 We allowed California’s Public Employees’ Retirement System to file

an amicus curiae brief in this appeal, and the arguments amicus makes therein pertain mainly to questions of statutory interpretation when considering the interaction between California statutes governing state retirement benefits and the FEHA. These arguments are not germane to our disposition of this appeal and they were not raised below, so we do not address them. 10 Plaintiffs alternatively sought to amend their complaint to refer to

“ ‘advanced age generally,’ ” but they do not argue that the court erred by denying this aspect of their motion.

27 The court denied plaintiffs’ motion, noting that plaintiffs sought to change their legal theory, given that they had proceeded at trial under the theory that the City discriminated based on “ ‘age at hire’ ” and treated “ ‘age at hire’ ” and “ ‘age at SFERS membership’ ” as the same thing. The court found that plaintiffs did not provide supporting authority for their request in their moving papers; its tentative statement of decision was “not based solely on the ‘age at hire’ theory”; plaintiffs failed to provide a reasonable excuse for the delay in seeking amendment; the City would be prejudiced because plaintiffs sought amendment after the end of trial and the City had prepared a defense in response to plaintiffs’ actual allegations; and the proposed amendments could conflict with the class definition that was based on age at hire. B. Analysis We review the court’s denial of leave to amend for abuse of discretion. (Melican v. Regents of University of California (2007) 151 Cal.App.4th 168, 175.) Plaintiffs contend that the court abused its discretion for many reasons. The City disagrees, but argues there is no reversible error in any event because the result would be the same. In reply, plaintiffs argue that they established a probability of a more favorable result with respect to their disparate treatment claim because the court found they had failed to establish that Formula 2 was motivated by age at hire solely because plaintiffs’ proof focused on age at SFERS membership. We conclude that plaintiffs fail to show reversible error because the court’s ruling on their disparate treatment claim is not as narrow as they suggest. The trial court ruled that Formula 2 was motivated by pension status, not age. It is true the court found that plaintiffs’ evidence did not match their “age at hire” theory such that they did not show “age at hire”

28 substantially motivated Formula 2. But the court did not stop there. Citing Kentucky Retirement Systems, the court broadly concluded Formula 2 did not facially discriminate “based on age,” and plaintiffs did not show that there was discrimination “due to their age.” The court determined that, “like in Kentucky Retirement Systems, Formula 2 appears to be motivated by credited years of service,” citing the high court’s conclusion that the Kentucky plan’s disparate treatment “ ‘turns upon pension eligibility and nothing more.’ ” The court also concluded with respect to plaintiffs’ breach of contract claim: “As there is no disparate treatment or disparate impact based on age discrimination, the Court will also find that the City did not pay [c]lass [m]embers reduced disability retirement benefits in an amount substantially motivated by their being age 40 or older at the time that they became members of SFERS.” (Italics added.) Finally, in its denial of plaintiffs’ motion to amend, the court confirmed that its conclusions were “not based solely on the ‘age at hire’ theory.” On this record, plaintiffs do not show that any purported error in the court’s ruling with respect to their motion to amend constitutes reversible error. DISPOSITION The judgment is affirmed.

29 _________________________ BROWN, P. J.

WE CONCUR:

_________________________ STREETER, J.

_________________________ CLAY, J.*

A169408; Carroll et al. v. City and County of San Francisco

* Judge of the Superior Court of Alameda County, assigned by the Chief

Justice pursuant to article VI, section 6 of the California Constitution.

30 Trial Court: San Francisco City and County Superior Court

Trial Judge: Hon. Andrew Y.S. Cheng

Counsel:

Aiman-Smith & Marcy, Randall B. Aiman-Smith, Reed W.L. Marcy, Hallie Von Rock, Brent A. Robinson and Lisseth Bayona for Plaintiffs and Appellants.

David Chiu, City Attorney, Rose Darling, Chief Labor Attorney, Peter A. Cownan and Adam M. Shapiro, Deputy City Attorneys for Defendant and Respondent.

Rob Bonta, Attorney General, Chris A. Knudsen, Assistant Attorney General, Kelsey Linnett and John T. McGlothlin, Deputy Attorneys General for Public Employees’ Retirement System as amicus curiae on behalf of Defendant and Respondent.