Kristen Silloway v. City and County of San FranciscoKristen Silloway v. City and County of San Francisco
FOR PUBLICATION
OPINION
Argued and Submitted February 12, 2024 San Francisco, California
Filed September 11, 2024
Before: Carlos T. Bea, David F. Hamilton,* and Morgan Christen, Circuit Judges.
Opinion by Judge Hamilton; Partial Concurrence and Partial Dissent by Judge Bea
SUMMARY**
Fair Labor Standards Act
The panel reversed the district court‘s summary judgment for the City and County of San Francisco, and remanded, in two cases in which staff nurses employed by the City allege that the City violated the Fair Labor Standards Act (FLSA) by not paying them time-and-a-half for overtime work.
The FLSA provides that employees should generally receive time-and-a-half pay for working overtime, but one of the Act‘s exemptions from that requirement applies to employees working in a bona fide professional capacity. The City claims that staff nurses fall into that exemption.
The dispute over whether the professional-capacity exemption applies to staff nurses depends on whether the City has shown that staff nurses were paid on a “salary basis” during the relevant time. The City claims that staff nurses were compensated on a salary basis because their annual compensation figures were documented at the start of every year through employment agreements and published salary ordinances. The plaintiffs contend that the City compensated them on an hourly basis because it divided those annual figures into hourly rates and paid staff nurses only for each hour worked.
The district court concluded that the annual pay figures published in the salary ordinance provided definitive evidence that the staff nurses were compensated on a salary basis.
The panel held that the district court erred. To determine whether employees are compensated on a salary basis, courts must look beyond conclusory language in contracts and similar documents such as the salary ordinance. Courts must instead analyze how employees are actually paid. The proper focus for the salary basis test is whether an employee receives a predetermined amount of compensation on a weekly or less frequent basis, irrespective of any promises made in an employment contract.
The panel held that material factual questions remain in dispute regarding whether the City satisfied the salary basis test as a matter of practice. Plaintiffs offered evidence showing that the City did not record them as working hours consistent with their full-time equivalencies in a significant number of pay periods. Those discrepancies raise material factual questions as to whether the staff nurses received their predetermined amounts of compensation in those pay periods. The panel remanded for those factual issues to be resolved.
Judge Bea concurred in part and dissented in part. He agreed that summary judgment in favor of the City should be reversed. But rather than remand for further discovery on whether the plaintiffs are salaried under the FLSA, he would hold that there is no genuine issue of disputed fact as to that question. The plaintiffs are not salaried under that statute because the City does not pay them a predetermined amount of compensation each week that is independent of the number of hours they work. He would remand with instructions to grant the plaintiffs’ cross-motion for summary judgment on their claim for overtime compensation under the FLSA.
COUNSEL
Caitlin E. Gray (argued), Maximillian D. Casillas, and Winnie G. Vien, Weinberg Roger & Rosenfeld, Los Angeles, California; Eduardo G. Roy, Prometheus Partners LLP, San Francisco, California; for Plaintiffs-Appellants.
Spencer J. Wilson (argued), Anastasia Bondarchuk, Ryan P. McGinley-Stempel, and Linda M. Ross, Renne Public Law Group, San Francisco, California, for Defendant-Appellee.
HAMILTON, Circuit Judge:
In these appeals, we address whether staff nurses for the City and County of San Francisco are entitled to time-and-a-half overtime, or whether the method of compensating the nurses satisfies the “salary basis test” in the Fair Labor Standards Act so that the nurses are exempt from the overtime requirement as bona fide professional employees.
The City employs staff nurses in its hospitals, jails, and clinics. Many work more than 40 hours in a week. The Fair Labor Standards Act provides that employees should generally receive time-and-a-half pay for working overtime, but one of the Act‘s exemptions from that requirement applies to employees working in a bona fide professional capacity. The City claims that staff nurses fall into that exemption. The plaintiffs disagree.
The dispute over whether the professional-capacity exemption applies to staff nurses depends on only one issue: whether the City has shown that staff nurses were paid on a “salary basis” during the relevant time. The City claims that staff nurses were compensated on a salary basis because their annual compensation figures were documented at the start of every year through employment agreements and published salary ordinances. In response, plaintiff nurses contend that the City compensated them on an hourly basis because it divided those annual figures into hourly rates and paid staff nurses only for each hour worked.
The district court granted summary judgment for the City. It concluded that the annual pay figures published in the salary ordinance provided definitive evidence that the staff nurses were compensated on a salary basis. That was error. To determine whether employees are compensated on a salary basis, courts must look beyond conclusory language in contracts and similar documents such as the salary ordinance. Courts must instead analyze how employees are actually paid. The proper focus for the salary basis test is whether an employee receives a predetermined amount of compensation on a weekly or less frequent basis, irrespective of any promises made in an employment contract.
We must reverse the grant of summary judgment. The City‘s compensation system does not necessarily flunk the salary basis test, but material factual questions remain in dispute regarding whether the City satisfied the test as a matter of practice. As we explain, plaintiffs offered evidence showing that the City did not record them as working hours consistent with their full-time equivalencies in a significant number of pay periods. Those discrepancies raise material factual questions as to whether the staff nurses received their predetermined amounts of compensation in those pay periods. We reverse and remand this case for those factual issues to be resolved.
I. Factual Background
Because plaintiffs lost on summary judgment in the district court, we take the facts in the light most favorable to them as the nonmoving parties, giving them the benefit of factual disputes and reasonable inferences from the evidence. See Tuuamalemalo v. Greene, 946 F.3d 471, 476 (9th Cir. 2019).
Staff nurses’ compensation is determined by many factors. The base salary for each staff nurse is the starting point. Base salaries are established through negotiations between the City and the nurses’ union. The agreed-upon amounts are subject to approval by the San Francisco Board of Supervisors. If the negotiated amounts are approved, then the salary figures are recorded in a Memorandum of
The City‘s payroll department translates each nurse‘s annual salary into an hourly rate by dividing the annual amount by 2,080, the number of hours a full-time nurse working 40 hours per week would expect to work in a year. A nurse who works 40 hours every week (or uses accrued time off as discussed below) would receive the full amount published in the salary ordinance.
A staff nurse can choose to work fewer hours than a full-time nurse. For example, a staff nurse could choose to work only 30 hours per week. In accounting jargon, the nurse working 40 hours per week would be referred to as a 1.00 full-time equivalent, or 1.00 FTE for short, while the nurse working 30 hours per week would be referred to as a 0.75 FTE. The staff nurse working three-fourths as much as the full-time nurse would in turn receive three-fourths as much in base salary compensation.
Staff nurses can also earn additional pay to supplement their base pay. One way is by working particular shifts, like evening or night shifts, which earn premium pay on top of normal hourly rates. Another way is by working overtime shifts as a staff nurse. Staff nurses who work overtime earn time-and-a-half (150% of their normal hourly rates) during those shifts.
A third way that staff nurses can earn additional pay is by working so-called “per diem” shifts. Nurses working these shifts are referred to as “per diem nurses.” The City offers per diem shifts on an as-needed basis to staff nurses employed by the City as well as to other nurses not already employed by the City. Staff nurses are never required to work these per diem shifts, but if they choose to do so, they earn 125% of their normal hourly rate regardless of whether they have worked more than, less than, or exactly 40 hours of regular shifts in the applicable week.
Staff nurses also accrue time in designated leave banks while working shifts, accumulating paid time off for vacations, illnesses, and holidays. So long as a staff nurse does not take off more time than the nurse has accrued in a particular leave bank, the nurse will not suffer any reduction in base compensation. However, if a staff nurse takes off more time than he or she has accrued, the City will deduct the amount of compensation equal to the amount of time missed. The City will also deduct pay if a nurse arrives late to a shift without permission from a supervisor.
All of these factors are taken into account when the City runs its payroll every two weeks. The payroll process begins with supervisors reviewing each nurse‘s work schedule and making adjustments to reflect additional hours worked or time taken off. The supervisors then submit the revised schedules to the payroll department.
Employees in the payroll department manually enter the timesheets into an accounting software system. The hours are entered under payroll codes that reflect the time spent performing different activities. Specific codes designate the amount of time devoted to working regular shifts, shifts earning differential pay, overtime shifts earning time-and-a-half pay, or per diem shifts earning time-and-a-quarter pay. Other payroll codes indicate the amounts of time each nurse allocated to vacation or illness, as well as the amounts of time consumed by unexcused absences. After all this information is entered into the system, the payroll department runs the software‘s final accounting process, which aims to catch any discrepancies between the time reported and payments allowed under the Memorandum of Understanding. This accounting process also
Sometimes errors slip through. A nurse might realize that she was not paid for all the hours she worked or that her differential pay was not paid properly. When that happens, nurses work with their supervisors and the payroll department to figure out what happened and correct the problem.
II. Statutory Background
In 1938, Congress enacted the Fair Labor Standards Act (FLSA) “to eliminate both substandard wages and oppressive working hours.” Helix Energy Solutions Group, Inc. v. Hewitt, 598 U.S. 39, 44 (2023) (internal quotation marks and citation omitted). The FLSA curbs extra-long working hours by, among other things, requiring employers to pay employees overtime pay. Id. Generally, employers must pay covered employees time-and-a-half when they work more than forty hours in a week.
Many employees, however, are exempt from the overtime requirement. As relevant here, an employer need not pay overtime to “any employee employed in a bona fide executive, administrative, or professional capacity.”
For a staff nurse to qualify as a bona fide professional, the City must show that the staff nurse‘s employment satisfies three tests: a duties test, a salary level test, and a salary basis test. See
A. The Salary Basis Test
The regulations establish two paths for satisfying the salary basis test. Helix, 598 U.S. at 55. Employees who are compensated on a “weekly[] or less frequent basis” are governed by the test in
Section 602(a) requires an employer to show that the employee at issue receives “on a weekly, or less frequent basis, a predetermined amount constituting all or part of the employee‘s compensation, which amount is not subject to reduction because of variations in the quality or quantity of the work performed.” If an employee performs any work in a given week, the employee must be paid full compensation for that week.
Section 604(b) provides an alternative path for an employer to show that an employee is paid on a salary basis. Under § 604(b), an employer may compensate employees “on an hourly, a daily or a shift basis” without running afoul of the salary basis test so long as two requirements are met: (1) the employment arrangement must include “a guarantee of at least the minimum weekly required amount paid on a salary basis regardless of the number of hours, days or shifts worked,” and (2) there must be a “reasonable relationship” between the employee‘s guaranteed amount of money and the money actually earned.
As the regulations show, both § 602(a) and § 604(b) seek to ensure that exempt professional employees receive a fixed minimum amount of money in their paychecks. Implicit in that promise, and made explicit in
In all these circumstances, the FLSA permits only full-day deductions. Partial-day deductions are off-limits for private employers. Public employers are allowed to make partial-day deductions, as we discuss below, but that flexibility comes not from § 602(b) but a separate FLSA regulation,
Supplemental compensation is a different story under the FLSA. While the FLSA strictly regulates deductions from pay, it permits employers to provide additional compensation on any basis—flat sum, straight-time hourly, or time-and-a-half hourly—without losing the benefit of the exemption, so long as “the employment arrangement ... includes a guarantee of at least the minimum weekly-required amount paid on a salary basis.”
B. The Public Accountability Principle
Only in the limited circumstances spelled out in § 602(b) may a private employer deduct money from a bona fide professional employee‘s compensation. But the FLSA gives public employers much more leeway. Most significantly, public employers like the City of San Francisco
C. Improper Deductions
Aside from the permissible deductions mentioned in § 602(b) and, for public employers the public accountability principle codified in § 710, an employer may not deduct pay from a bona fide professional employee‘s paycheck without losing the exemption.
All that said, the FLSA‘s regulations also offer an escape hatch to employers who make only “isolated or inadvertent” improper deductions. Under what is sometimes referred to as the “window of correction,” employers can retain the professional-capacity exemption if they reimburse employees for any improper deductions. See
D. The Burden-Shifting Framework
Courts have developed a burden-shifting framework for applying the FLSA in many contexts. At the outset, when an employee alleges that her employer is violating the FLSA, the employee bears the burden of proving that she performed work for which she was not properly compensated. Brock v. Seto, 790 F.2d 1446, 1447-48 (9th Cir. 1986). If an employer invokes the professional-capacity exemption to the FLSA‘s overtime requirement, then the employer bears the burden of showing that the employee falls within the exemption. Klem v. County of Santa Clara, 208 F.3d 1085, 1089 (9th Cir. 2000).2
III. Procedural Background
On March 8, 2018, Tatyana Litvinova filed a putative collective-action complaint against the City and County of San Francisco alleging that the City violated the Fair Labor Standards Act by not paying staff nurses time-and-a-half for overtime work, including per diem shifts. Litvinova v. City and County of San Francisco, No. 3:18-cv-1494-RS (N.D. Cal.). She moved to certify a collective action under
On October 22, 2020, Kristen Silloway, Christa Duran, and Brigitta van Ewijk filed a similar complaint on behalf of themselves and “similarly situated dual-status registered nurses.” Silloway v. City and County of San Francisco, No. 3:20-cv-7400-RS (N.D. Cal.). Given the factual similarity between the two cases, the district court issued an order treating them as related. Between the two separate collective actions, a total of about 353 plaintiffs opted in.
On cross-motions for summary judgment, the district court granted summary judgment in favor of the City, concluding that the staff nurses were paid on a “salary basis” and therefore exempt from the FLSA overtime requirements. Litvinova v. City and County of San Francisco, 615 F. Supp. 3d 1061, 1069 (N.D. Cal. 2022). The district court treated the published salary ordinance, which referred to staff nurses as salaried employees, as “dispositive evidence” that the nurses were compensated on a salary basis. Id. at 1066. The district court found the nurses’ hourly pay rates to be a mere “accounting fiction” used for administrative purposes, id. at 1066-67, and it rejected plaintiffs’ allegations of improper pay deductions by finding that the City‘s expert report provided adequate explanations for those discrepancies, id. at 1069.
Silloway timely appealed the district court‘s decision. Litvinova filed a Motion for Reconsideration under
IV. Standard of Review
On appeal, we review “both the granting of summary judgment and rulings regarding exemptions to the FLSA de novo.” Haro v. City of Los Angeles, 745 F.3d 1249, 1256 (9th Cir. 2014). Summary judgment is not appropriate unless, viewing the evidence in the light most favorable to the nonmoving parties and drawing all reasonable inferences in their favor, no genuine issues of material fact remain in dispute.
V. Analysis
The district court erred in granting summary judgment to the City. The salary ordinance, which the district court found to be dispositive evidence that the staff nurses were paid on a salary basis, is neither the starting point nor the ending point for that inquiry. Rather, the salary basis test asks whether an employee actually receives a predetermined amount of compensation on a weekly or less frequent basis as a matter of practice.
In this case, the parties dispute several factual issues that are material to answering that question. The most significant is whether staff nurses are guaranteed the opportunity to work the hours corresponding to their full-time equivalency every week. According to an expert report submitted by the City itself, the City recorded staff nurses as working or being credited
Additionally, the FLSA‘s “actual practice” and “window of correction” provisions offer the City no refuge, at least on summary judgment. Assuming that the 72 abnormal pay periods represent improper deductions—as we must in reviewing a grant of summary judgment against the plaintiffs—the City made improper deductions much more frequently than in cases where courts have found that no “actual practice” existed. Questions about the propriety of these 72 deductions leave material factual issues in dispute as to whether the City maintained an “actual practice” of making improper deductions. As for the “window of correction” defense, the City has not provided evidence showing that the staff nurses were reimbursed for any of these possibly improper deductions, so summary judgment cannot be granted or affirmed on that ground, either.
We address these points in more detail below, but the takeaway is this: the plaintiffs identified evidence that creates a material dispute of fact as to whether staff nurses actually received a predetermined amount of compensation on a weekly or less frequent basis. If they did not, they are not exempt from the overtime requirement. Summary judgment was not appropriate.
A. The Ordinance and the Memorandum of Understanding
We start with the district court‘s reasoning, which centered on the City‘s published salary ordinance and the Memorandum of Understanding. The salary ordinance lists the low and high ends of the range of biweekly compensation a full-time nurse could expect to receive. Echoing the salary ordinance, the Memorandum of Understanding explained that compensation rates were based on a full-time employee working “on a biweekly basis for a normal work schedule of five days per week, eight hours per day.” Throughout these documents, staff nurses were described as receiving salaries. The district court concluded that these public and contractual statements were “dispositive evidence” that the City paid the plaintiff staff nurses on a salaried basis. Litvinova, 615 F. Supp. 3d at 1066.
The district court‘s analysis centered on the wrong evidence. In 2004, the Department of Labor revised the FLSA regulations, shifting the focus of the salary basis test from the “employment agreement” to the pay an employee actually receives. See Orton v. Johnny‘s Lunch Franchise, LLC, 668 F.3d 843, 847-48 (6th Cir. 2012) (discussing revised regulations). The district court should have determined whether, as a matter of practice, staff nurses received predetermined amounts of compensation on a weekly or less frequent basis. See
B. Section 602(a) and the Public Accountability Principle
The question of law at the heart of this case is whether the City‘s compensation scheme, which assigns each staff nurse an hourly rate and computes paychecks based on the number of hours worked, satisfies the salary basis test. More than twenty years ago, we held that municipalities could use an hourly accounting system without offending the salary basis test. McGuire v. City of Portland, 159 F.3d 460, 464 (9th Cir. 1998). In light of revisions to the regulations governing the salary basis
We start by examining the interplay between two provisions in the FLSA‘s implementing regulations, § 602(a) and § 710. Section 602(a) requires an employer to pay a professional employee “a predetermined amount constituting all or part of the employee‘s compensation” on a weekly or less frequent basis. Other than a few limited exceptions listed in § 602(b), that predetermined amount cannot be subject to reduction due to the quantity or quality
of work performed. However, § 710 adds another permissible deduction to that list for public employers: partial-day deductions.
The City‘s ability to make partial-day deductions allows it to reduce staff nurses’ compensation in direct correlation to the amount of time worked. The FLSA permits private employers to deduct salaried pay only in full-day increments, but public employers can make minute-by-minute pay deductions for unexcused absences. That is the purpose of the public accountability principle: to prevent public employers from spending taxpayers’ money on employees who are not working. Thus, for public employers, § 710 qualifies § 602(a)‘s mandate that an employee‘s predetermined amount of compensation shall not be “subject to reduction because of variations in the ... quantity of the work performed.”
Section 710 does not, however, give public employers free rein to make pay deductions. To keep the benefit of the professional-capacity exemption, public employers must “otherwise meet[] the salary basis requirements of § 541.602.” § 710. So while public employers can make partial-day deductions for unexcused absences, they still cannot cause employees to miss work and suffer resulting pay deductions. See
Section 710‘s modification means that, as a practical matter, the salary basis test applies differently to private and public employers. Whereas a private employer must pay its employees predetermined amounts on a weekly or less frequent basis, a public employer must give its employees the opportunity to earn predetermined amounts on a weekly or less frequent basis, a prospect that will be fulfilled so long as employees do not miss work for unexcused reasons. In
both situations, neither private nor public employers can cause employees to receive less than the predetermined amounts of compensation. Any deduction must be due to an employee‘s own actions.
A hypothetical example shows how the salary basis test plays out differently for private and public employers. Imagine that a hospital employs a 1.00 FTE staff nurse and pays her $124,800 per year. That would mean that the staff nurse would have an hourly rate of $60 and earn $2,400 in a typical week. One week, a scheduling error occurs, resulting in the staff nurse being scheduled to work only 38 hours.
If the hospital is a private employer, it must pay the staff nurse $2,400 for that week. The private employer may not deduct any compensation for the two hours the employer caused the employee to miss work. Nor may the employer deduct compensation for the employee showing up late to work because that would be an impermissible partial-day deduction.
If the hospital is a public employer, however, it must pay the staff nurse only $2,340 for working 39 hours that week. Due to the public accountability principle, the employer is not obliged to pay the staff nurse for the hour that she was late. But it must still pay the staff nurse for the two hours that the employer caused the staff nurse not to work—it did not give the staff nurse the opportunity to work those hours.
From an accounting standpoint, the public employer could determine the compensation owed to the staff nurse by using a top-down approach, starting with a $2,400 weekly amount and then making any necessary deductions—in this example, subtracting $60. Or it could use a bottom-up approach that counts the number of hours actually worked, multiplies them by a $60 hourly rate, and makes any necessary adjustments—in this example, adding $120 because the employer caused the staff nurse to miss two hours. Through both adjustments, the staff nurse would be paid according to the number of hours she worked and not paid for hours she missed due to unexcused absences.
The only substantive disagreement we have with Judge Bea concerns our application of the public accountability principles embodied in
In 1966, Congress amended the FLSA to extend application of the Act to employees of public transit companies, hospitals, schools, and similar entities. That amendment was challenged immediately on constitutional grounds. The Supreme Court upheld application of the FLSA to those state and local employees in Maryland v. Wirtz, 392 U.S. 183 (1968). Congress then amended the FLSA in 1974 to broaden its application to almost all categories of state and local government employees, including police, fire, sanitation, public health, and parks employees. In a challenge to those new amendments, the Court overruled Maryland v. Wirtz and held that the constitutional commerce power did not authorize Congress to apply the FLSA to employees working in “areas of traditional governmental functions.” National League of Cities v. Usery, 426 U.S. at 852. And nine years after that, the Court overruled National League of Cities and its “traditional governmental functions” test in Garcia v. San Antonio Metro. Transit Authority, 469 U.S. 528, 557 (1985), which allowed broad application of the FLSA to state and local government employees.
In the wake of Garcia v. San Antonio Transit, “few public employers compensated employees in a manner that would satisfy the ‘on a salary basis’ provision” due to public accountability laws, and many local governments feared financial ruin due to the possibility of retroactive overtime payments. See Exemptions from Minimum
But those statutory amendments did not address one important concern of state and local governments. Many of those governments operate under constitutional, statutory, and/or regulatory provisions that bar governments from paying employees for time not actually worked or covered by accrued leave. Such prohibitions on “ghost employment” are discussed in terms of the “public accountability” principle. The Department of Labor has tried to accommodate that principle while also protecting employees through the revised rule that is now codified as
C. Applying the Salary Basis Test
The City failed to show beyond reasonable dispute that it guaranteed staff nurses the opportunity to work the number of hours corresponding to their full-time equivalencies during the relevant time period. The expert report submitted by the City revealed at least 72 employee pay periods in which the City recorded staff nurses as working fewer hours than their full-time equivalencies. These 72 discrepancies create factual questions as to whether the staff nurses received their predetermined amounts of compensation in each of these 72 pay periods.6
The City offers two reasons why the discrepancies should not prevent staff nurses from being considered salaried employees. First, the City argues that even if a few errant deductions were made, plaintiffs have not shown that the City maintains an actual practice of paying staff nurses less than their predetermined amounts. Second, the City argues that staff nurses have a window of correction available to them to correct any improper deductions. For the reasons we discuss below, both of these arguments fail, at least as a matter of law on summary judgment.
The City retained Dr. Piling Fan and Dr. Hossein Borhani to analyze the payroll data of staff nurses and to opine on whether the City “fulfills its obligation to provide the opportunity to staff nurses to work and be paid based on fixed schedules.”7 Dr. Fan and Dr. Borhani selected a sample of 26 plaintiff-nurses and retrieved about four years of payroll data for each nurse. The experts then graphed the sampled payroll records in a horizontal bar chart. Each horizontal bar in the graph represented one pay period for a single nurse. Within each bar, the experts color-coded the number of hours recorded for specific payroll codes. The color coding allows readers to discern how many hours each nurse spent doing certain activities. For example, by looking at Brigitta Van Ewijk‘s chart, we can determine that she worked or was credited with 80 hours in the two-week pay period ending on November 17, 2017. She reported 48 of those hours as “Regular Hours – Worked,” 16 hours as “Sick Leave Pay,” 8 hours as “Educational Leave w/ Pay,” and 8 hours as “Holiday OT Pay (1.5 times).”
Dr. Fan and Dr. Borhani concluded that “staff nurses work based on fixed schedules and are consistently paid for all the work and non-work hours on their regular work schedules.” For the most part, the graphed data supported the experts’ conclusion. The staff nurses appear to have worked the hours associated with their full-time equivalencies in over 2,000 employee pay periods across the four years of sample data. However, the graphs show at least 72 employee pay periods in which the sampled staff nurses appear to have worked fewer hours than their full-time equivalencies. Because staff nurses are paid according to the number of hours worked, these discrepancies raise unanswered factual questions as to whether the staff nurses earned their predetermined amounts of compensation in these pay periods.
Consider, for instance, plaintiff Kristen Silloway. During the two-week pay period ending May 31, 2019, Silloway worked 48 hours, which was 24 fewer hours than she had worked in other pay periods spanning from November 2017 to May 2021. The expert report acknowledged this discrepancy but dismissed it, concluding that “it appears Ms. Silloway‘s schedule [] changed from an FTE of [0].9 to [0].6” for this one pay period. When asked about this conclusion during her deposition, however, Dr. Fan said that she did not verify whether Silloway had reduced her full-time equivalency for that pay period. Dr. Fan had just assumed that Silloway reduced her full-time equivalency because her hours were lower than normal. After being pressed further on the issue, Dr. Fan noted that Silloway had also worked 47 hours in per diem shifts during this pay period. The additional hours that Silloway worked as a per diem nurse during this pay period are irrelevant for the salary basis test because, as noted above in footnote 4, only the hours that staff nurses work as staff nurses count toward their base compensation. Per diem shifts do not count. And the City does not dispute that Silloway worked only 48 hours as a staff nurse during this pay period. Because staff nurses’ compensation is determined by the number of hours worked, there is a factual dispute as to whether Silloway received her full, predetermined amount of compensation during this pay period. The City has not provided any other evidence showing that she received her full compensation
In fact, the City failed to provide definitive proof that staff nurses received their predetermined amounts of compensation during any of the pay periods in which these 72 discrepancies occurred. Instead, in the district court and on appeal, the City relies upon the unsubstantiated explanations provided by Dr. Fan and Dr. Borhani. We highlight a few of the experts’ deficient explanations to demonstrate why they do not satisfactorily resolve the factual disputes lingering around these discrepancies.
First, during the pay period ending on November 3, 2017, Silloway was recorded as working only 48 hours as a staff nurse, again 24 hours shy of her normal 72 hours as a 0.9 FTE. In their expert report, Dr. Fan and Dr. Borhani tried to explain away this discrepancy by saying that Silloway reduced her FTE from 0.9 to 0.6 for that pay period. The experts did not cite any documentation supporting that assertion, and the City has not provided any, either. Instead, when asked about this discrepancy during her deposition, Dr. Fan shifted her explanation, saying that the discrepancy may have been due to incomplete data. Again, though, Dr. Fan did not research further to confirm her hypothesis. And, more pertinent to the issue of summary judgment, the City has not provided evidence showing beyond reasonable dispute that Silloway received her predetermined amount of compensation for this pay period.
Second, during the pay period ending on January 12, 2018, Silloway was recorded as working 62 hours, of which 38 hours were recorded as “Regular Hours – Worked,” 12 hours were recorded as “Holiday OT Pay (1.5 times),” 2.9 hours were recorded as “Sick Leave Pay,” and 9.1 hours were recorded as “Sick Leave (Unpaid).” Dr. Fan and Dr. Borhani acknowledged in the expert report that Silloway was credited with 10 fewer hours during this pay period than her normal full-time equivalency. They asserted that this discrepancy was because Silloway “took paid sick leave[] and had insufficient sick leave in her bank to cover the rest of her FTE schedule.”
That explanation sounds like a proper invocation of the public accountability principle, but it does not reconcile with the payroll data. As noted, the graph shows that Silloway took 9.1 hours of “Sick Leave (Unpaid)“. So if the graphed payroll data show 9.1 hours of unpaid sick leave, it is puzzling why there would be an additional 10 hours of unpaid sick leave unrecorded in the accounting system. Dr. Fan and Dr. Borhani did not investigate this discrepancy any further, leaving it in dispute as to whether this discrepancy was an improper deduction of Silloway‘s compensation.
The experts did not investigate other discrepancies at all. Plaintiff Analisa Ruiz had nine pay periods between December 2017 and October 2021 in which the City‘s records credit her with fewer hours than her full-time equivalency. The expert report did not acknowledge these discrepancies. In its response brief, the City chalked them up to “incomplete payroll data that the experts had not fully investigated.” These nine unexplained discrepancies create disputed factual issues as to whether Analisa Ruiz received her predetermined compensation in these pay periods.
The experts’ uncorroborated and speculative explanations leave the propriety of the 72 discrepancies unknown. Perhaps there are permissible reasons for each discrepancy, but the City has not provided evidence proving them. Without such evidence, factual questions remain as to whether the staff nurses were provided the opportunity to work their full-time equivalencies in these pay periods and, consequently, whether the staff nurses were
2. Actual Practice
The City argues that even if the 72 discrepancies were improper deductions from staff nurses’ compensation, the City did not intend to make improper deductions, so it should not lose the benefit of the professional-capacity exemption. This argument is premised on another FLSA provision,
Pursuant to
The City‘s own expert report precludes it from meeting its burden as the moving party to show that no material facts remain in dispute as to whether the City maintained an actual practice of making improper deductions. See In re Oracle Corp. Sec. Litig., 627 F.3d 376, 387 (9th Cir. 2010) (“The moving party initially bears the burden of proving the absence of a genuine issue of material fact.“). Dr. Fan and Dr. Borhani analyzed payroll records for 26 nurses across 140 two-week pay periods. Among all the sampled nurses, the experts analyzed data from 2,251 employee pay periods.8 Plaintiffs assert that the expert report shows that staff nurses worked or were credited with fewer hours than their full-time equivalencies in 72 of these 2,251 employee pay periods. Because staff nurses are paid according to the number of hours they report, plaintiffs have presented factual disputes as to whether the staff nurses received their predetermined amounts of compensation in these 72 employee pay periods. And, as discussed above, the City has not provided evidence that would conclusively resolve those factual disputes. In reviewing the grant of summary judgment, we must view these disputed facts in the light most favorable to the plaintiffs and assume that the City paid these staff nurses less than their predetermined amounts of compensation in these pay periods.
Such a high number of improper deductions could support a finding that the City maintains an actual practice of making improper deductions. Plaintiffs identified evidence showing that the City made improper deductions in about 3.2% of employee pay periods. That rate is higher than in other cases where isolated errors did not indicate an actual practice. See Auer v. Robbins, 519 U.S. 452, 462 (1997) (just one improper deduction occurring under “unusual circumstances” did not show an actual practice); Childers v. City of Eugene, 120 F.3d 944, 947 (9th Cir. 1997) (one improper deduction over ten years was not an actual practice); Ellis v. J.R.‘s Country Stores, Inc., 779 F.3d 1184, 1189-90, 1194–96 (10th Cir. 2015) (one improper deduction during employee‘s more than four years of employment did not show an actual practice); Kennedy v. Commonwealth Edison Co., 410 F.3d 365, 372 (7th Cir. 2005) (three improper deductions over 470,000 employee work weeks did not show an actual practice); DiGiore v. Ryan, 172 F.3d 454, 457-58, 464-65 (7th Cir. 1999) (five improper deductions over about three years did not show an actual practice), overruled on other grounds by Whetsel v. Network Property Servs., LLC, 246 F.3d 897, 904 (7th Cir. 2001); Aiken v. City of Memphis, 190 F.3d 753, 762 (6th Cir. 1999) (one improper deduction was not an actual practice); Carpenter v. City & County of Denver, 115 F.3d 765, 767 (10th Cir. 1997) (two allegedly improper deductions under unusual circumstances did not show an actual practice); Ahern v. County of Nassau, 118 F.3d 118, 120–21 (2d Cir. 1997) (one instance of pay docked for violating employment rules was not an actual practice); Rebischke v. Tile Shop, LLC, 229 F. Supp. 3d 840, 852 (D. Minn. 2017) (concluding that improper deductions in 0.5% of paychecks over about three years was an “isolated” practice); Martinez v. Hilton Hotels Corp., 930 F. Supp. 2d 508, 522 (S.D.N.Y. 2013) (three improper deductions among five employees over four years did not show an actual practice); Crabtree v. Volkert, Inc., 2012 WL 6093802, at *9 (S.D. Ala. Dec. 7, 2012) (improper deductions from 1% of checks issued to certain employees over about four years showed an isolated practice).
The 3.2% error rate precludes summary judgment. Not only is it higher than in other cases where summary judgment was granted, it also suggests that a flaw in the City‘s accounting process resulted in recurring improper deductions. The data bear out this possibility. In the first 40 pay periods analyzed by the experts—from the period ending June 3, 2016 to the one ending December 1, 2017—17.5% of pay periods included a discrepancy.9 But those pay periods analyzed data from no more than seven staff nurses. Such a small sample size does not inspire confidence in the statistics drawn from it. Once the sample size was expanded to at least twelve nurses, the error rate jumped to 46%. In other words, in the 100 pay periods from the period ending December 15, 2017 to the one ending October 1, 2021, the City recorded at least one staff nurse as working or being credited with fewer hours than her full-time equivalency in nearly half the pay periods. And that statistic takes into account only a fraction of the staff nurses employed by the City. A pattern reaching that level of consistency tends to show a gap in the City‘s accounting process.
Factoring these statistics into the
None of the other
Finally, regarding the fifth factor, the parties dispute whether the City had a policy that prohibited improper deductions. The City contends that it did have such a policy and cites provisions from the Memorandum of Understanding that it claims guaranteed staff nurses the opportunity to work at least the hours associated with their full-time equivalencies. The plaintiffs disagree, arguing that far from prohibiting improper deductions, the Memorandum of Understanding explicitly permitted them. Plaintiffs cite provisions that purportedly authorized the City to cancel shifts without pay for reasons of “inclement weather conditions, shortage of supplies, traffic conditions, or other unusual circumstances.” Plaintiffs argue that these provisions violate
In response, the City contends that these provisions were merely “boilerplate language” lifted from other contracts the City has made. The City also asserts that these provisions were never invoked against staff nurses. In support of that assertion, the City cites an email from Steven Ponder, Classification and Compensation Director, to the Department of Public Health‘s Human Resources Director and Payroll Manager in December 2019, saying that the provisions permitting the City to cancel shifts did not apply to staff nurses. Plaintiffs attack the credibility of this email, noting that it was sent almost a year after Litvinova filed suit. Aside from convenient timing, plaintiffs also argue that the content of the email was never communicated to staff nurses or other relevant employees as required by the fifth element. See
As one last piece of evidence in support of its argument, the City cites deposition testimony from multiple staff nurses saying that the City has never exercised these provisions against them. All of the deposed staff nurses testified that they could not recall an instance in which the City denied them the opportunity to work a shift as a staff nurse. Plaintiffs respond by arguing that even if the deposed staff nurses could not recall an instance in which they had a shift cancelled, the City does not audit its payroll data to ensure that staff nurses actually received their guaranteed hours. And, as the 72 discrepancies show, plaintiffs have evidence that staff nurses sometimes worked or were credited with fewer
Suffice it to say, each side has marshaled evidence in support of its interpretation of the Memorandum of Understanding. The plaintiffs cite two provisions that seem to allow the City to make improper deductions. The City cites deposition testimony indicating that, as a matter of practice, those provisions have never been applied to staff nurses. This factor does not weigh in favor of either side.
In sum, the district court erred in concluding that no material factual questions remain in dispute as to whether the City maintained an actual practice of making improper deductions. In considering summary judgment, the number of discrepancies in the payroll data cannot be dismissed as mere isolated incidents. Nor can they be swept under the rug as the misdeeds of a single rogue manager. The Memorandum of Understanding includes provisions from which a reasonable person could conclude that the City retained the ability to cancel shifts. The possibility that those provisions could lead the City to make improper deductions caused the City‘s Compensation Director to clarify that those provisions cannot be exercised against staff nurses.
We offer no definitive answer as to whether the 72 discrepancies showed actual improper deductions from staff nurses’ predetermined amounts of compensation. All we decide is that the answer depends on disputed factual questions.
3. Window of Correction
The City also contends that even if the 72 discrepancies were improper deductions, they were merely “isolated or inadvertent” errors fixable through a correction process. This argument relies on
Regarding the second part of that provision, the City has a process in place for making corrections. Once an error is discovered, the payroll department works with nurses and their supervisors to rectify the issue. The City presented evidence showing that this process works in practice, at least some of the time. For example, it was discovered that Kristen Silloway was not “paid up to [her] FTE due to short education credits” during the pay period ending on October 16, 2020. The payroll team worked with Silloway to create a Problem Description form to track the issue and resolve it. In that instance, Silloway chose to use 10 hours of accrued vacation leave and ended up receiving her full predetermined amount of compensation.
The resolution of that issue, however, raises questions as to why the error correction process was not used to remedy the other 72 discrepancies. The expert report shows that after the correction was made, Silloway‘s reported hours for the pay period ending October 16, 2020 matched her full-time equivalency. In other words, this pay period was not one of the 72 discrepancies identified by the plaintiffs, but it would have resembled those discrepancies if the correction had not been made.
The City failed to show that corrections were made for the other 72 discrepancies. Again, because these appeals come to us from a grant of summary judgment, we construe the facts in the light most favorable to the plaintiffs and assume that each of the 72 instances of underreported time resulted in an improper deduction of compensation. Just as the City did not provide
Facts also remain in dispute concerning the first element of
a. “Isolated” Improper Deductions
The Department of Labor explained that the same factors for determining whether an employer maintains an “actual practice” determine whether improper deductions were “isolated.” Defining and Delimiting the Exemptions for Executive, Administrative, Professional, Outside Sales and Computer Employees, 69 Fed. Reg. 22122, 22181 (Apr. 23, 2004) (commentary on revisions to
b. “Inadvertent” Improper Deductions
Material facts are also in dispute as to whether the improper deductions were “inadvertent.” The same Department of Labor commentary defines “inadvertent deductions” as “those taken unintentionally, for example, as a result of a clerical or time-keeping error.” 69 Fed. Reg. at 22181. Perhaps that is true of the apparent errors here. A supervisor reviewing time records might have forgotten to make a necessary correction before sending a timesheet to payroll. Or, while manually entering the timesheets into the City‘s accounting system, a payroll employee could have accidentally entered the wrong time in the payroll software, causing the employee‘s records to show a lower-than-normal number of hours. However, the City has not provided evidence explaining what caused hours to be underreported in these pay periods. Without that evidence, whether these deductions were inadvertent remains a disputed factual issue.
VI. Conclusion
We reverse the judgment of the district court. Material factual questions remain in dispute as to whether the plaintiff staff nurses received predetermined amounts of compensation on a weekly or less frequent basis during the relevant time. We remand for those factual issues to be resolved consistent with this opinion.
REVERSED AND REMANDED.
BEA, Circuit Judge, concurring in part and dissenting in part:
The district court‘s grant of summary judgment in favor of the City and County of San Francisco (City) should be reversed. On that, the majority and I agree. But rather than remand for further discovery on whether the Plaintiffs-Appellants (Nurses) are salaried under the Fair Labor Standards Act (FLSA), I would hold that there is no genuine dispute of fact as to that question. The Nurses are not salaried under that statute because the City does not pay the Nurses a predetermined amount of compensation each week that is independent of the number of hours they work. I would therefore remand with instructions to grant the Nurses’ cross-motion for summary judgment on their claim for overtime pay under the FLSA.
The FLSA requires public and private employers to pay their workers time-and-a-half “for work over 40 hours a week.” Helix Energy Sols. Grp., Inc. v. Hewitt, 598 U.S. 39, 44 (2023). But the statute also “exempts certain categories of workers” from this overtime compensation requirement. Id. Under
Two “pathways” exist for meeting the salary-basis requirement under the Secretary‘s rules. Hewitt, 598 U.S. at 57. The first is set forth in
While
Under
Finally, while
II
It is undisputed that—on paper and in practice—the City pays the Nurses by the hour with no promise of “a preset and non-reducible” amount of pay. See Hewitt, 598 U.S. at 52. The City‘s own charter prohibits public employees from receiving compensation for hours they do not work. See San Francisco Charter § A8.400(g). The Memorandum of Understanding between the City and the Nurses—which outlines the Nurse‘s compensation scheme—provides that the Nurse‘s “[s]alaries . . . shall be calculated . . . proportionate to the hours actually worked.” The City‘s own expert and Director of Compensation admitted at deposition that the Nurses’ “compensation . . . [is] based on [the] hours they worked.” And to calculate the Nurses’ paychecks, the City‘s payroll system relies solely on the hours that the Nurses work. There is simply no dispute that the Nurses are hourly-rate employees under the City‘s compensation scheme. With facts like these, the City does not pay the Nurses on a salary basis under a plain English reading of either
In holding that the Nurses can nevertheless be considered salaried under the FLSA, the majority revises key parts of the Secretary‘s rules and practically writes out others altogether. For example,
To justify this remodeling of the regulatory text, the majority relies on
The problem with the majority‘s “public vs. private” employer distinction is that it does not hold up under closer scrutiny. As my colleagues concede, public employers
Taking the majority‘s reasoning to its logical conclusion, the answer would be $480. The private hospital “is not obliged to pay the staff nurse for the” four workdays that he missed because those days count as “full-day absences” under
To get around this problem, my colleagues explain that their reasoning is limited to
Thus, the limiting principle that supposedly reins in the majority‘s public-private distinction does not exist. If we follow my colleagues’ reading of “the interplay” between
What is left of
“The broader regulatory structure” at play also suffers a blow with the majority‘s holding today. Id. at 55. As mentioned earlier, my colleagues all but read out
The City is free to pay the Nurses solely by the hour, but that does not satisfy the salary-basis test under the Secretary‘s rules. Until the City guarantees them a fixed amount of pay that does not depend on the days or hours they work, the Nurses are not salaried under the FLSA, and the City must pay them overtime under that statute. For these reasons, the district court entered summary judgment for the wrong party. I would reverse and remand with instructions to grant summary judgment in favor of the Nurses on their claim for overtime compensation under the FLSA.