See’s Candy Shops, Inc. v. Superior Court

Facts

Two policies, deliberately distinct. See’s Candy recorded work hours through the Kronos timekeeping system; a punch “shows the actual time (to the minute) when the employee punched into the system.” Pay was then computed from those punches subject to two separate adjustments. (See’s Candy Shops, Inc. v. Superior Court (2012) 210 Cal.App.4th 889, 892.)

Under the nearest-tenth rounding policy, in and out punches were rounded up or down to the nearest tenth of an hour — “every six minutes beginning with the hour mark,” so punches move to the nearest three-minute mark. A 7:58 a.m. punch becomes 8:00 a.m.; an 8:02 a.m. punch also becomes 8:00 a.m. (See’s Candy, supra, 210 Cal.App.4th at p. 892.)

Under the separate grace period policy, employees with schedules programmed into Kronos could voluntarily punch in up to 10 minutes early and out up to 10 minutes late; they were not permitted to work during that window, and were paid from scheduled start and stop times rather than punch times. As the court explained, grace-period punches “accurately show when the employee punched in or out, but they do not show the beginning or end of the employee’s work shift, i.e., compensable time,” and if work was performed the manager had to adjust the record. Where the grace period applied, rounding generally became irrelevant because the compensable time began exactly at the scheduled time. (See’s Candy, supra, 210 Cal.App.4th at pp. 892–893.) Keeping these two policies apart is the analytical key to the whole opinion.

The action. Pamela Silva, a nonexempt hourly employee from about 1993 to 2010, filed a class action in October 2009 alleging failure to pay for all work performed, failure to pay overtime, unlawful meal and rest period policies, failure to pay premiums, and inaccurate wage statements, plus derivative claims under Business and Professions Code section 17200 and the Private Attorneys General Act. (See’s Candy, supra, 210 Cal.App.4th at p. 893.) The trial court certified a class on two separate issues: whether class members lost compensation from the rounding policy, and whether they lost compensation from the grace period policy. See’s Candy answered with 62 affirmative defenses, including that unpaid amounts were de minimis, that the rounding policy complied with state and federal law, and that the grace period policy was lawful. (Ibid.)

The motion, and what it did not reach. Silva moved for summary adjudication on four defenses — two de minimis defenses (the 10th and 41st) and the two rounding defenses (the 39th and 40th). She “did not move for summary adjudication on See’s Candy’s affirmative defense that its grace period policy is ‘lawful under both federal and California law.’ ” (See’s Candy, supra, 210 Cal.App.4th at p. 894.) That omission drives the outcome. Her legal theory was that no California authority permits rounding and that the policy violates section 204, which requires payment of “All wages” twice monthly, and section 510, which requires premium pay for “Any work” beyond eight hours a day or 40 a week. (Ibid.)

The dueling experts. Silva’s showing rested largely on three paragraphs of a 2010 report by See’s Candy’s own expert, Dr. Ali Saad, who had found that rounding produced “a net surplus of rounded over actual shifts of 2,230 employee work hours” that “resulted in a net economic benefit to the employees as a group,” averaging seven seconds per employee per shift — although Silva herself showed an “aggregate shortfall” of .47 hours. (See’s Candy, supra, 210 Cal.App.4th at p. 894.) She also offered charts (exhibit N) purporting to show a $725 loss, without foundational or authentication evidence. (Id. at p. 895.)

See’s Candy opposed by urging the federal standard and supplementing Dr. Saad’s analysis through April 2011. In the 2011 report Dr. Saad applied the grace period first and then compared actual to rounded punches, concluding that the policy “is both mathematically and empirically unbiased,” produced a total gain of 2,749 hours for the class, and “did not negatively impact employees’ overtime compensation.” Breaking it down: 59.1 percent of the class gained, 33 percent lost, and 7.9 percent were unaffected. Dr. Saad also withdrew the one data point Silva had relied on personally: he “modified his earlier conclusion with respect to plaintiff Silva” and “found that Silva was fully compensated for all of her worktime.” (See’s Candy, supra, 210 Cal.App.4th at p. 896.) Dr. Saad added that “[f]rom a mathematical perspective . . . the methodology of rounding to the nearest tenth of an hour for pairs of punches is exactly neutral,” and that the employee-favorable result “is meaningless — the extremely small excess amount could have been a minutely small shortfall with a different sample of data.” (See’s Candy, supra, 210 Cal.App.4th at p. 896.)

Procedural history

Silva’s reply shifted ground. Relying on See’s Candy’s response to her undisputed fact No. 18 — “Disputed. Clockings made during the grace period accurately show when the employee punched in or out, but do not show the beginning and end of the employee’s shift, i.e., compensable time. Hours worked are not necessarily reflected in the time between two punches.” — she argued that See’s Candy had admitted maintaining “inaccurate” time records and had therefore admitted violating California law. (See’s Candy, supra, 210 Cal.App.4th at p. 897.)

The trial court first issued a tentative ruling for Silva on that theory, then ruled the other way, finding that Silva had not met her burden and, alternatively, that See’s Candy had raised triable issues. (See’s Candy, supra, 210 Cal.App.4th at pp. 897–898.) Silva moved for reconsideration on “New Law” (Sullivan v. Oracle Corp. (2011) 51 Cal.4th 1191 and Securitas Security Services USA, Inc. v. Superior Court (2011) 197 Cal.App.4th 115) and “New Facts” — a report by Dr. Thomas Thompson calculating class losses of $1,411,595.54. Critically, Dr. Thompson “did not separate out the grace period time adjustments, and instead apparently assumed each and every employee was working during the grace period and was not paid for that time.” (Id. at p. 898.)

At the reconsideration hearing Silva’s counsel argued that rounding is lawful only if the employer performs “a mini actuarial process” every two weeks, and — decisively — “acknowledged that the grace period rules raised separate issues, and See’s Candy’s grace period policy was not before the court on the summary adjudication motion.” (See’s Candy, supra, 210 Cal.App.4th at p. 899.) The court then granted reconsideration and granted summary adjudication for Silva on all four defenses, reasoning that “based on See’s admission that its time records are inaccurate, See’s has violated California law”; that See’s Candy “failed to sufficiently address the plain language of Labor Code section 204”; and that See’s Candy “does not (and cannot) dispute that the federal rounding standard requires payment for all work time.” (Ibid.)

See’s Candy petitioned for a writ of mandate as to the 39th and 40th defenses. After the writ was summarily denied, the Supreme Court granted review and directed Division One to vacate that order and issue an order to show cause; extensive writ briefing and several amicus briefs followed. (See’s Candy, supra, 210 Cal.App.4th at p. 892.) The standard of review is de novo, with the reviewing court’s obligation being “ ’ “ ‘to determine whether issues of fact exist, not to decide the merits of the issues themselves’ “ ’ “ and with all doubts resolved in favor of the party opposing the motion. (Id. at pp. 899–900.)

Issue

Two questions, one legal and one evidentiary. First, what is “the appropriate legal standard” for evaluating a rounding claim under California law, where “[a]lthough California employers have long engaged in employee time-rounding, there is no California statute or case law specifically authorizing or prohibiting this practice”? (See’s Candy, supra, 210 Cal.App.4th at p. 901.) Second, applying that standard, did Silva carry her summary adjudication burden of completely disposing of the 39th and 40th affirmative defenses? (Id. at pp. 899–900, 907.)

Holding

Reasoning

1. A gap in California law, filled from federal practice. The court began by acknowledging the vacuum: rounding is long-standing California practice with “no California statute or case law specifically authorizing or prohibiting” it. (See’s Candy, supra, 210 Cal.App.4th at p. 901.) The federal rule it borrowed is 29 C.F.R. § 785.48(b), adopted about 50 years earlier under the Fair Labor Standards Act, which recites the practice of recording start and stop times to the nearest five minutes, tenth or quarter hour, presumes that “this arrangement averages out so that the employees are fully compensated for all the time they actually work,” and accepts it “provided that it is used in such a manner that it will not result, over a period of time, in failure to compensate the employees properly for all the time they have actually worked.” (Ibid.) Federal courts have read that to permit rounding so long as the policy “does not ‘consistently result[] in a failure to pay employees for time worked’ “ and, “on average, favors neither overpayment nor underpayment,” while condemning policies that “systematically undercompensate[] employees,” such as one that “encompasses only rounding down.” (Id. at pp. 901–902, quoting Alonzo v. Maximus, Inc. (C.D.Cal. 2011) 832 F.Supp.2d 1122, 1126–1127, and Eyles v. Uline, Inc. (N.D.Tex., Sept. 4, 2009, No. 4:08-CV-577-A) 2009 WL 2868447.)

2. The DLSE Manual as persuasive, not binding. The court’s California hook is the Division of Labor Standards Enforcement’s adoption of the federal language in its Enforcement Policies and Interpretations Manual, sections 47.1 and 47.2. (See’s Candy, supra, 210 Cal.App.4th at p. 902.) It was careful about the Manual’s status: statements in it “are not binding on the courts because the rules were not adopted under the Administrative Procedure Act” (Tidewater Marine Western, Inc. v. Bradshaw (1996) 14 Cal.4th 557, 568–577), but “may be considered for their persuasive value.” (Id. at pp. 902–903.) The doctrinal weight therefore rests on the federal regulation plus a non-binding agency interpretation — a point worth remembering when the rule is described as settled California law.

3. Why the federal standard fits California’s employee-protective scheme. Agreeing with Alonzo, the court reasoned that “[i]n the absence of controlling or conflicting California law, California courts generally look to federal regulations under the FLSA for guidance,” and that the policies underlying the regulation — “recognizing that time-rounding is a practical method for calculating worktime and can be a neutral calculation tool for providing full payment to employees — apply equally to the employee-protective policies embodied in California labor law.” Then the crucial qualifier: “Assuming a rounding-over-time policy is neutral, both facially and as applied, the practice is proper under California law because its net effect is to permit employers to efficiently calculate hours worked without imposing any burden on employees.” (See’s Candy, supra, 210 Cal.App.4th at p. 903.) The permission is conditional, and both conditions are the employer’s to satisfy.

4. Section 204 is a timing statute. Silva’s strongest textual argument was that section 204’s command that “All wages” be paid twice monthly requires the employer to “unround” every pay period. The court read the phrase in context: “the reference to ‘All wages’ in section 204, subdivision (a) pertains to the timing of wage payments and not to the manner in which an employer ascertains each employee’s worktime,” noting the observation that “the sole purpose of [section 204] is to require an employer of labor who comes within its terms to maintain two regular pay days each month.” (See’s Candy, supra, 210 Cal.App.4th at pp. 904–905, citing In re Moffett (1937) 19 Cal.App.2d 7, 13.) [Moffett is a Court of Appeal decision; the See’s Candy opinion describes it as coming from the California Supreme Court, which appears to be an error in the opinion.] It also identified the argument’s “false premise — that using unrounded figures within a finite time period is the only way to measure ‘All’ earned wages,” explaining that “the question whether all wages have been paid is different from the issue of how an employer calculates the number of hours worked and thus what wages are owed. Section 204 does not address the measurement issue.” (Id. at p. 905.) A legislative-acquiescence point followed: the Legislature has amended section 204 since the DLSE adopted the federal rule “and has never indicated that the state agency’s adoption of the federal rounding rule is inconsistent with its statutory provision.” (Ibid.)

5. Section 510 sets a rate, not a measurement method. The same treatment disposed of the overtime statute: “as with section 204, this code section has nothing to do with rounding or calculating time. Rather this provision sets the multiplier for the rate at which ‘Any’ overtime work must be paid.” (See’s Candy, supra, 210 Cal.App.4th at p. 905.)

6. The daily-overtime objection is converted into a fact question. Silva’s most sophisticated argument was structural: because California pays premium rates after eight hours in a day, “a gain of 3 minutes of Regular Time is valued less than a loss of 3 minutes of Overtime,” so nearest-tenth rounding can never be truly neutral in California. (See’s Candy, supra, 210 Cal.App.4th at pp. 905–906.) The court did not reject the premise. It answered that the difference between daily and weekly overtime “does not show that rounding under the DOL rounding regulation will always burden the employee under California law,” that “[t]here is no analytical difference between rounding in the context of daily overtime and rounding in the context of weekly overtime,” and that “the issue whether California’s overtime rules mean a rounding rule is biased against employees is a factual issue and not a legal one.” (Id. at p. 906.) Later it went further, conceding that “California’s overtime rules may mean that under a nearest-tenth rounding policy, an employee will not be fully compensated for the premium time if an employee works more than eight hours in one day. However, the issue whether this will result in undercompensation over time is a factual one.” (Id. at p. 912.) That concession is the doorway plaintiffs have used ever since — and it is the question now pending in Camp.

7. Sullivan, Kirby and Brinker distinguished, not disputed. The court accepted that under Sullivan v. Oracle Corp. a California employer generally must pay all employees working in California state overtime wages, but held that Sullivan “did not address” whether sections 204 and 510 prohibit rounding absent biweekly reconciliation, invoking the rule that “an opinion is not authority for a proposition not therein considered.” (See’s Candy, supra, 210 Cal.App.4th at pp. 906–907, quoting Ginns v. Savage (1964) 61 Cal.2d 520, 524, fn. 2.) As to Kirby v. Immoos Fire Protection, Inc. and Brinker Restaurant Corp. v. Superior Court — cited for plain-meaning construction, employee-protective interpretation, deference to wage orders, and the employer’s duty to keep accurate records — the court said simply: “We agree with the principles expressed in Kirby and Brinker, and to the extent they are relevant here, we have applied them in this case.” (Id. at p. 907.)

8. The “inaccurate records” admission was about the grace period, not rounding. The court’s answer to the trial court’s central rationale is a lesson in reading a discovery response. See’s Candy’s “Disputed” answer described its grace period policy, under which scheduled time rather than punch time sets pay because the employer assumes the employee “is not working and not under its control” then. (See’s Candy, supra, 210 Cal.App.4th at p. 909.) That difference “does not show See’s Candy’s time records are inaccurate for purposes of the summary adjudication motion,” because the parties agreed that a grace period is lawful “if the employee is not working or is not under the employer’s control” — which raises factual questions Silva never joined. (Ibid.) She “did not produce any evidence showing the class members who clocked in during the grace period were working or were under the employer’s control,” See’s Candy produced contrary evidence, and her counsel confirmed she was not challenging the grace period policy in the motion. (Id. at pp. 909–910.) Morillion v. Royal Packing Co. (2000) 22 Cal.4th 575 did not help, because Silva “presented no evidence on whether the employees were prevented from using ’ “the time effectively for [their] own purposes.” ’ ” (Id. at p. 910.) The court expressly preserved the claim for later: if the evidence shows employees were working or under control during the grace period and unpaid, “they may be entitled to recover those amounts in the litigation and any applicable penalties.” (Id. at pp. 910–911.)

9. Burden allocation decided the case. A plaintiff moving on an affirmative defense must “completely” dispose of it. (See’s Candy, supra, 210 Cal.App.4th at pp. 899–900, citing Code Civ. Proc., § 437c, subd. (f)(1).) Silva’s evidence consisted of the employer’s own expert finding neutrality, unauthenticated charts, and a rebuttal expert whose model assumed the disputed fact. Dr. Thompson’s conclusions were “invalid to show whether the nearest-tenth rounding policy resulted in a loss to employees” because “Silva presented no evidence to support this assumption.” (Id. at pp. 907–908.) And even had the burden shifted, Dr. Saad’s 2011 report raised triable issues. (Id. at p. 908.) A footnote adds a further structural point: even though, on the court’s own description, “a minority of the employees (33 percent) had a net loss of a minimal amount during the class period,” “a court cannot grant summary adjudication on part of a defense.” (Id. at p. 908, fn. 7.)

10. The court also narrowed what its holding did. Distinguishing Eyles, the court stressed that “the employer here is not asking the court to find its policy was lawful as a matter of law; instead it is merely opposing plaintiff’s motion for summary adjudication and asking the court to permit it to litigate its affirmative defense at trial.” (See’s Candy, supra, 210 Cal.App.4th at p. 913.) It also declined the invitation to treat the grace period as an “automatic round down,” observing that “the grace period policy is not a rounding policy per se; it is a policy under which an employer seeks to accurately pay employees from the time they begin and end work,” and that the reciprocity question “is not before us.” (Ibid.) On the discipline theory, it noted there was “no evidence that See’s Candy uses the rounded time (as opposed to the clocked-in time) when applying its tardiness policy.” (Id. at p. 911.)

Significance

See’s Candy is the foundational California authority on time rounding, and for more than a decade it has been cited for the proposition that neutral rounding is lawful. Two qualifications belong with every such citation.

First, the decision is a writ vacating summary adjudication on affirmative defenses. It holds that See’s Candy is entitled to litigate its defenses, not that its policy is lawful; the court said so twice, at pages 892 and 913. Second, the rule it announces is conditional and fact-intensive. Facial neutrality is not enough — the policy must be neutral “both facially and as applied” (See’s Candy, supra, 210 Cal.App.4th at p. 903), and whether it undercompensates over time is “a factual issue and not a legal one” (id. at p. 906). Properly read, the case supplies employers a defense that must be proved with data, not a safe harbor.

Currency — read this before relying on the rule. The rule announced here has already been narrowed twice, and its survival is now before the California Supreme Court. Three layers, in order.

Layer one: the meal-period context is already carved out. In Donohue v. AMN Services, LLC (2021) 11 Cal.5th 58 — the only California Supreme Court treatment of this case, which discusses it under the name See’s Candy I — the court held that an employer may not round time punches in the meal-period context, because the neutrality premise fails there. It pointedly declined to bless the rule generally: “This court has never decided the validity of the rounding standard articulated in See’s Candy I, and we are not asked to do so here.” (Donohue, supra, 11 Cal.5th at p. 72.) The meal-period holding is framed as an alternative that does not depend on that question — even assuming See’s Candy is valid, a meal-period rounding policy fails its own neutrality standard, because rounding up never creates a premium-pay obligation that is not owed while rounding down can defeat one that is. Donohue also cast doubt on the efficiency rationale that See’s Candy rests on: rounding “was developed as a means of ‘efficiently calculat[ing] hours worked’ . . . and is useful ‘in some industries, particularly where time clocks are used[,]’ . . . [b]ut technological advances may help employers to track time more precisely, and ‘employers are in a better position than employees to devise alternatives.’” (Id. at p. 73, quoting See’s Candy, supra, 210 Cal.App.4th at p. 903, 29 C.F.R. § 785.48(b) (2020), and Troester v. Starbucks Corp. (2018) 5 Cal.5th 829, 848.) And: “As technology continues to evolve, the practical advantages of rounding policies may diminish further.” (Donohue, supra, 11 Cal.5th at pp. 73–74.)

Layer two: rounding is already unlawful where actual minutes were captured. That is a live limit today, not a pending question. In Camp v. Home Depot U.S.A., Inc. (2022) 84 Cal.App.5th 638, the Sixth District held that “if an employer, as in this case, can capture and has captured the exact amount of time an employee has worked during a shift, the employer must pay the employee for ‘all the time’ worked.” (Id. at p. 660.) The court then “respectfully invite[d] the California Supreme Court to review the issue of neutral time rounding by employers.” (Id. at p. 661.)

Layer three: whether neutral rounding survives at all is pending. The Supreme Court granted review in Camp, S277518, on the question: “Under California law, are employers permitted to use neutral time-rounding practices to calculate employees’ work time for payroll purposes?” That matter remains pending — the Supreme Court’s official Issues Pending — Civil list current as of September 4, 2026 still lists S277518 on that identical issue, and no Supreme Court merits opinion in Camp appears in CourtListener’s index as of September 2026 (only the Court of Appeal’s 2022 decision, H049033). The Court of Appeal’s own reasoning here identifies the pressure point: See’s Candy conceded that California’s daily overtime rule “may mean” an employee is not fully compensated for premium time, treating the consequences as a fact question rather than a legal one. (See’s Candy, supra, 210 Cal.App.4th at p. 912.) Confirm the status of S277518, and read Donohue and Camp, before citing this case.

The opinion’s other durable contribution is its insistence on separating a rounding policy from a grace period policy. Practitioners routinely blend them, as Dr. Thompson’s model did, and the court treated that conflation as fatal to the plaintiff’s showing. Grace period claims live or die on whether the employee was working or under the employer’s control — a Morillion question — while rounding claims live or die on neutrality over time.

Key quotes

“[T]he rule in California is that an employer is entitled to use the nearest-tenth rounding policy if the rounding policy is fair and neutral on its face and ‘it is used in such a manner that it will not result, over a period of time, in failure to compensate the employees properly for all the time they have actually worked.’ ” (See’s Candy Shops, Inc. v. Superior Court (2012) 210 Cal.App.4th 889, 907.)

“Assuming a rounding-over-time policy is neutral, both facially and as applied, the practice is proper under California law because its net effect is to permit employers to efficiently calculate hours worked without imposing any burden on employees.” (Id. at p. 903.)

“Our ruling leaves open the issue whether the parties will prevail in proving their various claims and defenses relating to See’s Candy’s nearest-tenth rounding policy and a related grace period policy.” (Id. at p. 892.)

“We agree that California’s overtime rules may mean that under a nearest-tenth rounding policy, an employee will not be fully compensated for the premium time if an employee works more than eight hours in one day. However, the issue whether this will result in undercompensation over time is a factual one.” (Id. at p. 912.)

Read the full opinion (Justia) →

Practice pointer

Open questions

Whether neutral rounding survives at all. The question is squarely before the California Supreme Court in Camp v. Home Depot U.S.A., Inc., S277518. See’s Candy rests on a federal regulation and a non-binding DLSE Manual provision (See’s Candy, supra, 210 Cal.App.4th at pp. 902–903), neither of which binds the Supreme Court.

The daily-overtime asymmetry. The court accepted that California’s eight-hour rule “may mean” an employee is undercompensated for premium time, and made the consequences a fact question. (See’s Candy, supra, 210 Cal.App.4th at p. 912.) No case has yet resolved how a rounding policy must be designed or audited to remain neutral in a daily-overtime jurisdiction.

Grace period reciprocity. “The issue whether reciprocal rules are required to ensure fairness under a grace period policy is not before us.” (See’s Candy, supra, 210 Cal.App.4th at p. 913.) Whether an employer may deem early punches non-compensable without an offsetting rule is unresolved.

The disadvantaged minority. The court flagged but declined to decide “whether those employees would be entitled to a recovery for these wages if See’s Candy establishes that over time the rounding policy is neutral.” (See’s Candy, supra, 210 Cal.App.4th at p. 908, fn. 7.) Whether class-wide neutrality defeats an individual employee’s shortfall claim is an open and consequential question.

Rounding technology. The court rejected “for purposes of this summary adjudication motion” Silva’s suggestion that the federal regulation applies only to “ ‘time clocks’ “ and not to software such as Kronos. (See’s Candy, supra, 210 Cal.App.4th at p. 913.) The qualifier leaves room for argument where the timekeeping system’s capabilities are themselves in evidence.