Cortina v. North Am. Title Co.

• Court: California Court of Appeal
• Decided: May 29, 2026
• Docket: F085389
• Status: Published

Opinion

Filed 5/29/26

CERTIFIED FOR PARTIAL PUBLICATION *

IN THE COURT OF APPEAL OF THE STATE OF CALIFORNIA

FIFTH APPELLATE DISTRICT

CAROLYN CORTINA et al., F085389 Plaintiffs and Respondents, (Super. Ct. No. 07CECG01169) v.

NORTH AMERICAN TITLE COMPANY, OPINION Defendant and Appellant.

APPEAL from a judgment of the Superior Court of Fresno County. Jeffrey Y. Hamilton, Jr., Judge. O’Melveny & Myers, David Marroso, Adam J. Karr, Anton Metlitsky, Jenya Godina; McCormick, Barstow, Sheppard, Wayte & Carruth, Scott M. Reddie; Morgan, Lewis & Bockius, Barbara J. Miller and John D. Hayashi for Defendant and Appellant. Wagner, Jones, Kopfman & Artenian, Andrew B. Jones, Lawrence M. Artenian; Cornwell & Sample, Stephen R. Cornwell, René Turner Sample; Wanger Jones Helsley and Patrick D. Toole for Plaintiffs and Respondents. -ooOoo- In Duran v. U.S. Bank National Assn. (2014) 59 Cal.4th 1 (Duran), Justice Carol Corrigan used the phrase “exceedingly rare beast” to describe “a wage and hour class

* Pursuant to California Rules of Court, rules 8.1105(b) and 8.1110, this opinion is certified for publication with the exception of part III of the Discussion. action that proceeded through trial to verdict.” (Id. at p. 12.) The plaintiffs in Duran had alleged their employer misclassified them as exempt from the laws requiring premium pay for overtime and mandatory meal breaks and rest periods during the workday. Such lawsuits are notoriously difficult to manage on a large scale, are often deemed unsuitable for class treatment, and usually settle if class certification is granted. The Duran action, which involved a class of 260 people, lasted more than 16 years and ultimately ended in decertification due to the unmanageability of issues arising from the employer’s defenses. We dare say the present case is even rarer and more beastly than Duran. The original complaint was filed 19 years ago. There have been numerous interlocutory appeals and writ proceedings, the most recent of which involved review by the California Supreme Court. (North American Title Co. v. Superior Court (2024) 17 Cal.5th 155.) At long last, an appeal of the underlying judgment is presented for review on the merits. Unfortunately, the end is still nowhere in sight. The wage and hour claims of approximately 700 plaintiffs, divided into two classes based on whether they had been designated as exempt or nonexempt employees, were litigated in a bifurcated bench trial. Upon conclusion of the first phase, the “Nonexempt” class was ordered decertified. However, the trial court ruled against the defendant employer on certain defenses asserted as to the claims of the “Exempt” class. The court then appointed a referee to conduct the second phase of trial, doing so without the parties’ consent and over defendant’s strenuous objections. The reference proceedings lasted years and included live testimony from over 230 class members, all leading to the entry of a $43 million judgment. A nonconsensual reference of the scope and magnitude ordered in this case is not merely a rare occurrence. It appears unprecedented in our state jurisprudence. A trial court’s authority to delegate matters to a referee without the parties’ consent is strictly circumscribed by the California Constitution and Code of Civil Procedure. The reference

2. proceedings below were entirely unauthorized. This error alone compels reversal of the judgment. Prejudicial errors were also committed in the first phase of trial, which became known as the “liability phase” even though it did not truly resolve all issues of liability. Critical aspects of the trial plan and format contravened the holdings of Duran. The employer’s affirmative defenses were also rejected out of hand, on a classwide basis, due to misinterpretations of the applicable law. The employer relied on two different Labor Code exemptions, both of which largely depend on how employees spend their time during the workday. To provide a very simplified primer, certain types of employees may be exempt from California’s overtime laws if they spend over half of their time performing duties and tasks meeting the legal definitions of exempt work. If an employee spends 49 percent of their time on exempt work, such as managerial tasks, but the other 51 percent of their time on nonexempt work, they do not qualify for the exemption. “Given California’s uniquely quantitative approach … [citation], some proof about how individual employees use their time will often be necessary to accurately determine an employer’s overtime liability [on a classwide basis].” (Duran, supra, 59 Cal.4th at p. 27.) It was alleged that all members of the roughly 400-person “Exempt” class, despite having a variety of different job titles and working in different offices throughout the state over a 10-year period, all performed the same allegedly nonexempt job the same way, all the time. One problem with this theory was that several cohorts within the class, e.g., “Branch Managers,” indisputably did perform exempt work on a regular basis. The issue then became how to make a classwide showing that, despite variation among and between class members, none met the 51 percent threshold. In the Branch Manager example, plaintiffs elicited testimony from roughly 15 percent of the cohort (about 24 out of 156 people) and argued the work habits and experiences of those witnesses could be relied upon as representative of the entire group.

3. The Duran opinion rejects such reasoning as scientifically invalid. “If sampling is used to estimate the extent of a party’s liability, care must be taken to ensure that the methodology produces reliable results.” (Duran, supra, 59 Cal.4th at p. 42.) A valid extrapolation model requires expert input to determine the appropriate sample size and margin of error, and the sample group of witnesses must be randomly selected. (Id. at pp. 42–43.) None of those requirements were met. Without common proof of employee misclassification, the trier of fact “may have sufficient evidence to make judgments as to particular individuals, but lacks a basis to extrapolate from those findings to a class-wide judgment.” (Marlo v. United Parcel Service, Inc. (C.D.Cal. 2008) 251 F.R.D. 476, 485 (Marlo).) The absence of common proof leaves the alternative of hearing individual testimony from hundreds of class members, which turns the class action device on its head. In the “liability phase,” the trial court imposed limits on how many witnesses could testify. The number was arbitrarily set at approximately 100 witnesses for each side to prove or refute the claims of all 700 people comprising both classes. The court later authorized individual testimony from all members of the “Exempt” class in the second phase of trial (such authorization being the main reason for appointing a referee) but prohibited the employer from questioning the witnesses on matters related to its exemption defenses. The exemption defenses were deemed to have failed on a classwide basis in the first phase of trial. As held in Duran, “decertification must be ordered whenever a trial plan proves unworkable.” (Duran, supra, 59 Cal.4th at p. 32.) Here, the obligation to decertify both classes was apparent during the initial phase of trial. Failure to decertify the “Exempt” class, in combination with the other prejudicial errors, warrants reversal and a decertification order by this court. (Id. at pp. 24, 50.) The cause will be remanded for retrial of the named plaintiffs’ individual claims. “The trial court is of course free to entertain a new certification motion on remand [as to the “Exempt” class], but if it

4. decides to proceed with a class action it must apply the guidelines set out here.” (Id. at p. 33.) FACTUAL AND PROCEDURAL BACKGROUND Parties Defendant below and appellant herein, and respondent to the cross-appeal, is identified in the operative complaint (amended posttrial and shortly before entry of judgment) as “North American Title Company, (formerly known as North American Title Company, Inc. and CalAtlantic Title, Inc., and now known as Lennar Title, Inc.).” (Bolding and some capitalization omitted.) In the appellate briefing, this party identifies itself as “formerly known as North American Title Company and now called Lennar Title, Inc.” We will refer to this party, in its capacity as both appellant and respondent, as defendant or NATC. 1 Plaintiffs below and Respondents herein, and appellants in the cross-appeal, include the following 11 individuals hereinafter collectively referred to as the “named plaintiffs” or “class representatives”: Kimberly Baker, Judith Bates, Catherine Bell, Melodie Benton, Carolyn Cortina, Martha Dominguez, Cheryl Fuller, Laurel Johnstone, Teresa Spencer, Tina Texeira, and Mary Weidmark. Regarding both the appeal and cross-appeal, the term “plaintiffs” will collectively refer to the same 11 individuals and, where applicable, all unnamed plaintiffs classified as exempt employees while employed by NATC during the class period, i.e., the “Exempt” class. The terms defined in the previous paragraph do not apply to Janet Doran. Her unique circumstances were the subject of a prior appeal, Cortina v. North American Title Co. (May 30, 2019, F074938) [nonpub. opn.]. There are no claims in the appeal or

1 The California Style Manual states, “The terms ‘cross-appellant’ and ‘cross-respondent’ are never used in appellate titles and should be avoided in the body of opinions.” (Cal. Style Manual (4th ed. 2000) § 6:3, p. 212.)

5. cross-appeal regarding the trial court’s entry of judgment against Doran. Accordingly, the part of the judgment concerning Doran will be affirmed. A codefendant, North American Services, LLC (“NAS”), prevailed at trial against both classes and all named plaintiffs. The term “defendants” will collectively refer to both NATC and NAS. There are no claims in the appeal or cross-appeal regarding the trial court’s entry of judgment in favor of NAS. Accordingly, the part of the judgment concerning NAS will be affirmed. Factual Background Overview The class period for all claims against NATC was April 17, 2003, through December 31, 2012. During that time, NATC employed several hundred people as escrow officers and in other escrow-related positions. Those employees collectively worked in over 80 offices located in 23 counties throughout California. As best we can discern from the record, NATC and its employees used the terms “branch” and “office” synonymously. What do escrow officers do? Defendant’s opening brief offers this summary: “Escrow officers employed by [NATC] serve as fiduciaries who play an integral part in real estate transactions by receiving, holding, and transferring funds; reviewing contracts, title reports, insurance policies, and loan documents; preparing and following escrow instructions; and checking for and taking affirmative actions to clear encumbrances such as liens or unpaid taxes, among other varying duties. [Record citations.] Escrow officers must make important decisions that can expose their employer to liability, including the critical determination that all requisite conditions for escrow closure have been satisfied.” The summary in NATC’s official job description identifies these responsibilities:

“Preparation of all documents associated with the processing and closing of escrow, conveyance of title, recording of documents, clearing of title requirements and issuance of title insurance. Interviews escrow principals to obtain pertinent facts, directs the preparation of escrow

6. instructions, drawing of documents, issuance of title policies and the disbursement of funds, and otherwise assists escrow principals, their lenders, and/or agents in the consummation of escrows involving the financing or sale of single family and small multi-family residences. Reconciles escrow instructions with commitment and/or preliminary title report. Develops and maintains business by the extensive application of basic marketing and customer service techniques.” Plaintiffs generally contend that an escrow officer’s job is routine and standardized. At trial they introduced into evidence a flow chart summarizing what they referred to as the “Escrow Process.” (A copy of that exhibit, which was partially created by class member Christine English, is attached to this opinion as Appendix A, post, at page 138.) One class member described escrow officers as “glorified data-processing people.” The more common sentiment was that escrow work is “complicated” and presents “unique” challenges from one transaction to another. Many witnesses agreed the “Escrow Process” chart was accurate, though perhaps overly “basic” in its synopsis. Other witnesses were more critical of it, contending it was incomplete. By all accounts, the profession was very demanding but also lucrative. Escrow officers progress in their career by building a “book of business.” This involves cultivating relationships with realtors, banking professionals, and others in the real estate industry. Escrow officers with a reputation for meeting their closing deadlines and error-free transactions develop a client base that refers work specifically to them. Companies such as NATC compete for and hire escrow officers based on their books of business, which gives such individuals leverage in a thriving real estate market. In the early part of the class period (circa 2003 to 2006), experienced escrow officers with an established book of business could easily command base salaries ranging from $50,000 to $70,000 per year, and higher. Those figures did not include other performance-based compensation.

7. NATC had “[d]esk [i]ncentive” and “[b]ranch [i]ncentive” policies under which escrow officers and other employees received a percentage of the revenue they generated, individually and as a branch, above certain threshold amounts. For example, class member Dawna Monroe’s base salary gradually increased from $51,600 to $69,600 as she was promoted from escrow officer to “Unit Manager,” and later to “Branch Manager.” She testified that in addition to her salary, “I think average after taxes, my commission would run about [$]6[,000] to [$]8,000” per month. The most striking example among the higher-earning class members was Gina Oldham, who held the position of Branch Manager. In the year 2003, she earned $103,500 in base salary plus $184,906 in incentive pay, which defendant argued constituted “commissions” for purposes of a separate exemption defense. Oldham’s incentive compensation also exceeded her base salary in 2004, for combined earnings of over $250,000. In 2005 her base salary increased to $15,000 per month ($180,000 annually), but the appellate record does not show her incentive compensation for that year. During the second phase of trial, NATC submitted evidence showing that most people in the Exempt class had at least some pay periods where their incentive compensation exceeded the amount they received in base salary. A notable example was Christine English, who at different times held the positions of “Advisory Escrow Officer” and “Senior Escrow Officer.” Her incentive compensation exceeded her base salary in over 40 percent of her pay periods during eight years of employment with NATC. Job Titles and Classifications In 2003, most NATC employees involved in escrow and escrow-related work were internally classified and treated as exempt employees. In other words, they were not paid extra for working overtime. Employees in this category included those with the job titles of Senior Escrow Officer, Special Projects Escrow Officer, and Escrow Officer Supervisor.

8. In addition to the job titles noted above, some evidence indicates that two different “Escrow Officer” positions, one exempt and one nonexempt, existed prior to 2004. The record is murky on this point, and defendant should have clarified it at trial. The apparently nonexempt position was Escrow Officer (job code 1169), and the exempt position was called Escrow Officer California (job code 1032). The former was an entry-level job. The latter, i.e., Escrow Officer California, required “[o]ne year [of] Escrow Officer experience … or two years general escrow experience.” Defendant’s discovery responses show these as two distinct positions, and some evidence in the second phase of trial reportedly showed employees being paid overtime as “Escrow Officers” in 2003. Nevertheless, since they appear to have all been lumped into the “Exempt” class below, we use the term “2003 Escrow Officer” in reference to plaintiffs classified under either or both job codes (1032 and 1169) prior to January 2004. NATC conducted a “Staffing Study” in 2003 to determine how many hours its employees were working, including those employees it classified as exempt. The Staffing Study preceded a corporate decision to reclassify certain positions and create a new exempt position called Escrow Unit Manager (Unit Manager). Internal company documents state, in pertinent part: “The Escrow Officer Job Title became overly broad and needed to be narrowed in scope. The Escrow Officer Job Title is being split into two Job Titles, one is Exempt and one is Non-Exempt. Incumbents fulfilling the 80% Exempt Duties Test and who manage two or more individuals, may be moved to the new Exempt Job Title,” i.e., designated as an exempt Unit Manager. 2

2 Other parts of the same documents refer to The Fair Labor Standards Act of 1938 (29 U.S.C. § 201 et seq.), thus indicating NATC’s reliance on federal law. The reference to an “80% Exempt Duties Test” alludes to older versions of federal regulations that were revised in 2004. (See fn. 3, post.) Under those older regulations, the overtime exemptions for managerial and administrative employees required that 80 percent of the employee’s time be devoted to exempt work. (29 C.F.R. former §§ 541.1(e) (2003), 541.2(d) (2003), 541.112(a) (2003), 541.209(a) (2003).) In short, NATC ostensibly imposed upon itself a more stringent “duties test” than was required under California law.

9. Effective January 2004, defendant reclassified the positions of 2003 Escrow Officer, Senior Escrow Officer, Special Projects Escrow Officer, and Escrow Officer Supervisor as nonexempt and thus eligible for overtime compensation. Some employees who previously held those positions were moved into the exempt position of Unit Manager. “Unit” referred to a group consisting of one escrow officer and at least two assistants over whom the escrow officer had supervisory authority. Four additional job titles are of particular importance to this case: Advisory Escrow Officer, Branch Manager, Escrow Manager, and Escrow Operations Manager. NATC classified these positions as exempt both before and after 2004. Procedural History A detailed summary of the trial court proceedings, from the point of class certification through entry of judgment, is set forth in the Discussion, post. We provide here an abbreviated procedural history. In April 2007, Carolyn Cortina filed a putative class action complaint against NATC alleging various Labor Code violations for failing to pay overtime, and failing to provide meal breaks and rest periods, to all of its California employees “in the position of Escrow Officer or [an] equivalent position, however titled ….” In addition to the statutory claims, a cause of action was pleaded under Business and Professions Code section 17200 et seq., commonly known as the unfair competition law (UCL). Cortina was originally hired as a nonexempt Escrow Assistant. She was promoted to Unit Manager in 2004, bypassing the nonexempt Escrow Officer position, and worked as a Unit Manager until leaving the company in 2005. In a declaration filed in support of class certification, Cortina alleged the job titles of Unit Manager and Escrow Officer were interchangeable. She further declared, “I did not manage anyone.” The complaint was amended in 2009, adding four more proposed class representatives. Those parties had previously worked in positions including the nonexempt job titles of Escrow Assistant and Junior Escrow Officer. In 2010, the trial

10. court certified two classes. A second amended complaint was filed shortly thereafter, adding more class representatives and naming NAS as a codefendant. NAS was sued pursuant to a theory of liability as a joint employer. A class of current and former employees whom NATC had classified as nonexempt was labeled the “Non-Exempt Class” (Nonexempt class). It was defined to include the job titles of Junior Escrow Officer, Escrow Officer, Senior Escrow Officer, Special Projects Escrow Officer, and Escrow Supervisor (also known as an Escrow Officer Supervisor). The position of Escrow Assistant was specifically excluded from the class definition. A class of current and former employees whom NATC had classified as exempt was labeled the “Exempt Class” (Exempt class.) It was defined to include the following job titles: “[2003] Escrow Officer, Senior Escrow Officer, Special Projects Escrow Officer, Escrow Supervisor, Advisory Escrow Officer, Escrow Unit Manager, Escrow Manager, Branch Manager, and Escrow Operations Manager.” The position of Escrow Assistant was specifically excluded from the class definition. The claims of the Nonexempt class were different from those of the Exempt class. The Nonexempt class alleged NATC had policies and practices prohibiting them from reporting overtime on their timesheets and/or pressuring them not to report overtime or to underreport overtime, and requiring or pressuring them to skip meal breaks and rest periods. In other words, policies or practices requiring class members to work “off the clock.” The claims of the Exempt class were based on NATC’s alleged misclassification of them as exempt employees when, according to their allegations, they were in fact nonexempt. Put differently, the Exempt class alleged entitlement to overtime compensation and meal breaks and rest periods, which they did not receive. Plaintiffs later withdrew their claims for missed rest periods.

11. Prior to trial, both classes dismissed their statutory causes of action under the Labor Code and elected to proceed solely on the UCL claim. Trial began in September 2015. Witness testimony in the first phase of trial ended in December 2015. Written submissions followed, including closing briefs. In October 2016, the trial court issued a statement of decision. The court ruled against NATC on its exemption defenses. It further ordered the appointment of a referee to make findings and recommendations on the claims of the Exempt class for restitution under the UCL. The Nonexempt class was ordered decertified and the individual claims of its class representatives were also referred to the referee. In March 2021, the trial court issued an order adopting nearly all of the referee’s findings and recommendations. Later that year, due to what defendant labels “corporate reorganizations,” plaintiffs amended their complaint two more times. Issues surrounding the “corporate reorganizations” caused delays in the entry of judgment. On August 31, 2022, the trial court entered judgment in favor of the named plaintiffs, except Kimberly Baker and Carolyn Cortina, on their individual claims. Judgment was likewise entered for the Exempt class against NATC. NATC’s financial liability was calculated at $43,547,946, with prejudgment interest accounting for over half of the sum. Judgment was entered in favor of NATC as to the individual claims of Baker and Cortina. Defendant filed a motion for new trial, as did plaintiffs Baker and Cortina. While the motions for new trial were pending, defendant sought writ relief from this court on matters unrelated to the present appeal. Further background on the writ proceedings can be found in North American Title Co. v. Superior Court, supra, 17 Cal.5th 155. The writ proceedings resulted in a stay of the underlying action and denial of the motions for new trial by operation of law. Meanwhile, both sides filed timely notices of appeal. The remittitur in the writ matter issued on July 14, 2025.

12. DISCUSSION The issues in this case are numerous, complex, and require familiarity with several areas of law. Part I of the Discussion provides a broad legal overview before addressing defendant’s claims regarding the first phase of the trial. Part II addresses defendant’s claims regarding the second phase of trial, i.e., the nonconsensual reference proceedings. Part III addresses plaintiffs’ cross-appeal. Many of the issues present questions of law, which are reviewed de novo. (Leider v. Lewis (2017) 2 Cal.5th 1121, 1127.) “We review class action trial management decisions for abuse of discretion.” (Duran, supra, 59 Cal.4th at p. 49.) A trial court abuses its discretion when it makes decisions based on erroneous legal assumptions or reasoning. (American Honda Motor Co., Inc. v. Superior Court (2011) 199 Cal.App.4th 1367, 1372; Mejia v. City of Los Angeles (2007) 156 Cal.App.4th 151, 158.) I. First Phase of Trial Legal Overview A. California Wage and Hour Laws In California, claims for unpaid wages are governed by the Labor Code and a series of wage orders promulgated by the Industrial Welfare Commission (IWC). (Brinker Restaurant Corp. v. Superior Court (2012) 53 Cal.4th 1004, 1026 (Brinker).) The IWC is a state agency vested with the quasi-legislative power to develop and issue wage orders, i.e., regulations “specifying minimum requirements with respect to wages, hours, and working conditions” for all industries and occupations. (Ibid.) Although it was defunded by the Legislature in 2004, the IWC’s wage orders continue to have the same force of law as when the agency was operational. (Mendiola v. CPS Security Solutions, Inc. (2015) 60 Cal.4th 833, 838, fn. 6; see Brinker, at p. 1027 [“The IWC’s wage orders are to be accorded the same dignity as statutes”].)

13. “Of the 18 wage orders in effect today, ‘16 cover[] specific industries and occupations, one cover[s] all employees not covered by an industry or occupation order, and a general minimum wage order amend[s] all others to conform to the amount of the minimum wage currently set by statute.’ ” (Mendiola v. CPS Security Solutions, Inc., supra, 60 Cal.4th at pp. 838–839.) IWC wage order No. 4-2001 (Cal. Code Regs., tit. 8, § 11040 (Wage Order 4)) governs wages, hours, and working conditions for all persons employed in “professional, technical, clerical, mechanical, and similar occupations.” (Id., subd. 1.) We accept the parties’ shared contention that Wage Order 4 applies to the claims and defenses in this case. (See id., subd. 2(P) [providing a nonexhaustive list of included occupations].) The Fair Labor Standards Act of 1938 (29 U.S.C. § 201 et seq. (FLSA)), along with regulations promulgated by the United States Department of Labor, is the “federal counterpart” to California’s state wage and hour laws. (United Parcel Service Wage & Hour Cases (2010) 190 Cal.App.4th 1001, 1009 (UPS Wage & Hour Cases); see Montano v. Montrose Restaurant Associates (5th Cir. 2015) 800 F.3d 186, 190 [noting the Department of Labor “is authorized to promulgate rules interpreting and clarifying the FLSA”].) The FLSA “does not preempt state law in this area, and therefore state law is controlling to the extent it is more protective of workers than federal law.” (Alvarado v. Dart Container Corp. of California (2018) 4 Cal.5th 542, 554.) “In the absence of controlling or conflicting California law, California courts generally look to federal regulations under the FLSA for guidance.” (See’s Candy Shops, Inc. v. Superior Court (2012) 210 Cal.App.4th 889, 903.) Our state has a public policy “favoring an eight-hour workday and a six-day 40-hour workweek, and discouraging employers from imposing work in excess of those limits.” (Alvarado v. Dart Container Corp. of California, supra, 4 Cal.5th at p. 561.) The concept of “overtime” is thus defined as “[a]ny work in excess of eight hours in one workday and any work in excess of 40 hours in any one workweek and the

14. first eight hours worked on the seventh day of work in any one workweek.” (Lab. Code, § 510, subd. (a); accord, Wage Order 4, subd. 3(A)(1).) Unless subject to an exemption, overtime must be compensated at a premium rate of at least 150 percent of the employee’s regular rate of pay (commonly known as “time and a half”). Any work in excess of 12 hours in one day, or in excess of eight hours on any seventh day of a workweek, must be compensated by no less than double the employee’s regular rate of pay. (Lab. Code, § 510, subd. (a); Wage Order 4, subd. 3(A)(1)(a), (b).) Labor Code section 1194 provides for a cause of action against an employer to recover unpaid overtime compensation. Nonexempt employees are further entitled to meal breaks and rest periods during the workday (id., § 512, subd. (a); Wage Order 4, subds. 11(A), 12(A)), and those rights are enforceable under Labor Code section 226.7. Missed breaks are compensable by “one additional hour of pay at the employee’s regular rate of compensation for each workday that the meal or rest or recovery period is not provided.” (Lab. Code, § 226.7, subd. (c); accord, Wage Order 4, subds. 11(B), 12(B).) B. Relevant Exemptions The Labor Code authorizes exemptions from the requirement of premium overtime pay for “executive” and “administrative” employees. (Lab. Code, § 515, subd. (a).) The applicable wage orders establish such exemptions and detail the prerequisites. Three elements are common to both exemptions. The employee must (1) be “primarily engaged in the duties that meet the test of the exemption”; (2) “regularly exercise[] discretion and independent judgment in performing those duties”; and (3) earn “a monthly salary equivalent to no less than two times the state minimum wage for full-time employment.” (Ibid.) Employees who qualify for either exemption will also be exempt from the meal and rest period requirements of Labor Code section 512. (E.g., Wage Order 4, subd. 1(A); see Lab. Code, § 226.7, subd. (e).) “ ‘[T]he assertion of an exemption from the overtime laws is considered to be an affirmative defense, and therefore the employer bears the burden of proving the

15. employee’s exemption.’ ” (Negri v. Koning & Associates (2013) 216 Cal.App.4th 392, 397 (Negri). “[W]e narrowly construe exemptions against the employer, ‘and their application is limited to those employees plainly and unmistakably within their terms.’ ” (Peabody v. Time Warner Cable, Inc. (2014) 59 Cal.4th 662, 667.) Nevertheless, “[d]etermining whether or not all of the elements of the exemption have been established is a fact-intensive inquiry.” (UPS Wage & Hour Cases, supra, 190 Cal.App.4th at p. 1014.) The applicability of an exemption is a mixed question of law and fact, and the analysis can be complicated. (Ramirez v. Yosemite Water Co. (1999) 20 Cal.4th 785, 794 (Ramirez).) The requirement of “ ‘primarily’ ” engaging in duties meeting the tests for the exemptions means spending “more than one-half of [one’s] worktime” on such duties. (Lab. Code, § 515, subd. (e); accord, Wage Order 4, subd. 2(O).) The trier of fact must, “first and foremost,” determine what work is “actually performed by the employee during the course of the workweek.” The amount of time the employee actually “spends on such work” must then be considered “together with the employer’s realistic expectations and the realistic requirements of the job.” (Wage Order 4, subds. 1(A)(1)(e), (2)(f).) In other words, this element alone contemplates a four-part assessment of (1) the tasks actually performed by the employee; (2) the amount of time devoted to each task; (3) the employer’s realistic expectations of how the employee will or should be spending their time; and (4) how those facts align with the realistic requirements of the job. (Judicial Council of California Civil Jury Instructions (CACI) Nos. 2720, 2721; see Safeway Wage & Hour Cases (2019) 43 Cal.App.5th 665, 671, 679; Martinez v. Landry’s Restaurants, Inc. (2018) 26 Cal.App.5th 783, 797.) In conjunction with that assessment, the trier of fact must also evaluate the degree to which the employee’s job involves the exercise of discretion and independent judgment. (Lab. Code, § 515, subd. (a).) “Whether an employee satisfies the elements of the exemption is a question of fact,” but the “appropriate manner of evaluating the employee’s duties is a question of

16. law.” (Heyen v. Safeway Inc. (2013) 216 Cal.App.4th 795, 817.) In a jury trial, the trier of fact will be instructed on which tasks may be characterized as exempt or nonexempt. (See Directions for Use to CACI Nos. 2720, 2721.) In a bench trial, as was conducted in this case, the trial court must perform the legal analysis itself. In determining whether a given task qualifies as exempt or nonexempt work, the employee’s duties “shall be construed in the same manner” as under certain federal regulations “effective as of the date of [the wage] order.” (Wage Order 4, subds. 1(A)(1)(e), (2)(f).) Due to the Legislature’s defunding of the IWC in 2004, the applicable law is frozen in time, to a certain extent, circa the early part of this century. (See Harris v. Superior Court (2011) 53 Cal.4th 170, 190 (Harris) [“The IWC Statement issued in connection with Wage Order [4] clearly states that ‘only those federal regulations specifically cited in its wage orders, and in effect at the time of promulgation’ shall be applied in defining exempt duties under California law”].) Wage Order 4 was last revised effective January 1, 2001, and it incorporates several federal regulations that no longer exist. The California Supreme Court has arguably suggested, however, that current federal regulations have some degree of “instructive” value since they are “ ‘intended to be consistent with the old regulations.’ ” (Gallardo v. AIG Domestic Claims, Inc. (9th Cir. 2015) 629 Fed.Appx. 783, 784, fn. 2, citing and quoting Harris, supra, 53 Cal.4th at p. 189, fn. 8.) At a minimum, the legal analysis requires cross-referencing and parsing older versions of multiple federal regulations as they existed in January 2001. 3 (E.g., Harris, at pp. 180–182.)

3 The IWC was defunded, and its operations ceased, effective July 1, 2004. (Huntington Memorial Hospital v. Superior Court (2005) 131 Cal.App.4th 893, 902, fn. 2; accord, IWC Web site [as of May 29, 2026] [announcing the end of IWC operations], archived at .) Approximately seven weeks later, in August 2004, the United States Department of Labor “substantially revised the FLSA overtime exemption regulations.” (Robinson-Smith v. Government Employees Ins. Co. (D.C. Cir. 2010) 590 F.3d 886, 892, fn. 6; accord, Combs v. Skyriver Communications, Inc. (2008) 159

17. Another potentially complicated issue is the possibility of both exempt and nonexempt periods during a course of employment. (See, e.g., UPS Wage & Hour Cases, supra, 190 Cal.App.4th at p. 1013 [employee qualified for exemptions in three different positions held during employment period of over 10 years]). Here, roughly 25 percent of the Exempt class (around 100 members) held multiple allegedly exempt job positions during the nearly 10-year class period ending on December 31, 2012. Furthermore, there is precedent for examining and applying exemptions on a week-to-week basis. (Dunbar v. Albertson’s, Inc. (2006) 141 Cal.App.4th 1422, 1426–1427, 1430 (Dunbar).) The trier of fact may, however, “make reasonable inferences about a party’s activities during the relevant period based on his or her activities in earlier and later periods, particularly where there is nothing to suggest the employee’s duties and responsibilities changed significantly.” (Batze v. Safeway, Inc. (2017) 10 Cal.App.5th 440, 479.) Under federal law, employees who perform a sufficient combination of managerial and administrative duties may be classified as exempt. (29 C.F.R. § 541.708.) “In other words, work that is exempt under one section of the [overtime regulations] will not defeat the exemption under any other section.” (Ibid.) This approach implicitly recognizes that some jobs “may not fit comfortably” into only one exemption category, i.e., “executive, administrative, or professional.” 4 (Anderson v. City of Cleveland (E.D.Tenn. 2000) 90 F.Supp.2d 906, 915–916 [discussing a predecessor statute, former 29 C.F.R. § 541.600]; accord, Campanelli v. Hershey Co. (N.D.Cal. 2011) 765 F.Supp.2d 1185, 1197 [discussing 29 C.F.R. § 541.708].)

Cal.App.4th 1242, 1255, fn. 5.) Due to the timing of those events, none of the IWC wage orders incorporate the August 2004 revisions to, or current version of, part 541 of title 29 of the Code of Federal Regulations. 4 A “[p]rofessional [e]xemption” also exists under California law, but it has no relevance to this case. (Wage Order 4, subd. 1(A)(3); see Lab. Code, § 515, subd. (a).)

18. In a 2003 opinion letter, the Department of Industrial Relations, Division of Labor Standards Enforcement (DLSE), wrote that it “construes state law, like federal law [citation], to permit the so-called ‘tacking’ of one type of exempt work with another type, so that, for example, an employee who spends more than 50% of his worktime performing exempt managerial and exempt administrative work would meet the ‘primarily engaged’ test.” (Dept. Industrial Relations, DLSE Opn. Letter No. 2003.05.23 (May 23, 2003) p. 5 [citing former 29 C.F.R. § 541.600].) As the state agency “ ‘ “empowered to enforce California’s labor laws,” ’ ” the DLSE’s opinion letters, “ ‘ “ ‘ “while not controlling … do constitute a body of experience and informed judgment to which courts and litigants may properly resort for guidance.” ’ ” ’ ” (Brinker, supra, 53 Cal.4th at p. 1029, fn. 11; e.g., Stone v. Alameda Health System (2024) 16 Cal.5th 1040, 1072; Hill v. R+L Carriers, Inc. (N.D.Cal. 2010) 690 F.Supp.2d 1001, 1008 [citing DLSE Opn. Letter No. 2003.05.23 (May 23, 2003) as an indication “California appears to recognize a … ‘combination’ exemption” similar to 29 C.F.R. § 541.708]; Campanelli v. Hershey Co., supra, 765 F.Supp.2d at p. 1197, fn. 22 [same].) We have found no published decisions analyzing the concept of “ ‘tacking’ ” under California law, but some cases impliedly accept the premise that overtime exemptions are not mutually exclusive. (See, e.g., UPS Wage & Hour Cases, supra, 190 Cal.App.4th at p. 1013 [affirming trial court’s determination the appellant was both “an exempt executive employee and an exempt administrative employee”].) In the present case, defendant has at times relied on a combination of two or even three different exemptions (discussed momentarily under separate subheadings) to dispute the claims of numerous class members. Plaintiffs have never denied the availability of such “ ‘tacking’ ” defenses under California law. They argue defendant simply did not prove any theories of exemption.

19. 1. The Executive Exemption The executive exemption applies to those in managerial and/or supervisory positions, i.e., any employee “[w]hose duties and responsibilities involve the management of the enterprise in which they are employed or of a customarily recognized department or subdivision thereof.” (Wage Order 4, subd. 1(A)(1)(a).) Such individuals must “customarily and regularly” direct the work of two or more people. (Id., subd. 1(A)(1)(b).) They must also have the authority to hire and fire personnel or, alternatively, be someone “whose suggestions and recommendations as to the hiring or firing and as to the advancement and promotion or any other change of status of other employees will be given particular weight.” (Id., subd. 1(A)(1)(c).) The remaining elements, as earlier discussed, consist of regularly exercising discretion and independent judgment; being “primarily engaged” in duties satisfying “the test of the exemption”; and earning “a monthly salary equivalent to no less than two (2) times the state minimum wage for full-time employment.” (Wage Order 4, subd. 1(A)(1)(d)–(f).) Duties meeting “the test of the exemption” are to be construed in accordance with title 29 of the Code of Federal Regulations, sections “541.102, 541.104– 111, and 541.115–116,” as those regulations existed on January 1, 2001. (Wage Order 4, subd. 1(A)(1)(e).) “[I]t is generally clear that work such as the following is exempt work when it is performed by an employee in [a managerial or supervisory context]: Interviewing, selecting, and training of employees; … directing their work; maintaining their production or sales records for use in supervision or control; appraising their productivity and efficiency for the purpose of recommending promotions or other changes in their status; handling their complaints and grievances and disciplining them when necessary; planning the work; determining the techniques to be used; apportioning the work among the workers; [and] determining the type of materials, supplies, machinery or tools to be used .…” (29 C.F.R. § 541.102 (2001); see Wage Order 4, subd. 1(A)(1)(e).)

20. The list in the previous paragraph is non-exhaustive. The delegation of tasks to subordinate employees may qualify as exempt work that also constitutes the exercise of discretion and independent judgment. (See Bargas v. Rite Aid Corp. (C.D.Cal. 2017) 245 F.Supp.3d 1191, 1212 [applying California law in finding store manager “had discretion in prioritizing and delegating tasks to her employees on a daily basis”]; Thomas v. Speedway SuperAmerica, LLC (6th Cir. 2007) 506 F.3d 496, 507 [gas station manager’s discretionary exempt duties included delegating work among her employees and determining the weekly work schedule].) Various forms of problem-solving may also qualify as exempt work. (See UPS Wage & Hour Cases, supra, 190 Cal.App.4th at pp. 1019–1020; see generally Demos v. City of Indianapolis (S.D.Ind. 2000) 126 F.Supp.2d 548, 562 [“Problem-solving generally requires the exercise of discretion and independent judgment”].) “Moreover, ‘exempt work includes not only the tasks necessary for the actual management of a department and the supervision of its employees, but also tasks that are “closely associated with the performance of the duties involved in such managerial and supervisory functions or responsibilities.” ’ ” (Batze v. Safeway, Inc., supra, 10 Cal.App.5th at p. 473; see Wage Order 4, subd. 1(A)(1)(e).) Put differently, activities that are not inherently managerial or supervisory may sometimes be classified as exempt depending on the surrounding circumstances. This includes “work which is properly viewed as a means for carrying out exempt functions.” (Wage Order 4, subd. 1(A)(1)(e).) Defendant maintains that a majority of the Exempt class, including Escrow Operations Managers, Branch Managers, Escrow Managers, and Unit Managers, qualified for the executive exemption. 2. The Administrative Exemption Wage Order 4 “provides a five-part test to determine whether the administrative employee exemption applies. The employee must (1) perform ‘office or non-manual work directly related to management policies or general business operations’ of the

21. employer or its customers, (2) ‘customarily and regularly exercise[] discretion and independent judgment,’ (3) ‘perform[] under only general supervision work along specialized or technical lines requiring special training[, experience, or knowledge]’ or ‘execute[] under only general supervision special assignments and tasks,’ (4) be engaged in the activities meeting the test for the exemption at least [51] percent of the time, and (5) earn twice the state’s minimum wage [on a salaried basis].” (Eicher v. Advanced Business Integrators, Inc. (2007) 151 Cal.App.4th 1363, 1371, quoting Wage Order 4, subd. 1(A)(2).) Some elements of the exemption are construed in accordance with title 29 of the Code of Federal Regulations, sections “541.201–205, 541.207–208, 541.210, and 541.215,” as in effect on January 1, 2001. (Wage Order 4, subd. 1(A)(2)(f).) To satisfy those elements, the employee’s work must be both “qualitatively” and “quantitatively” administrative. (Harris, supra, 53 Cal.4th at p. 181, italics omitted.) The quantitative requirement means the work “must be of substantial importance to the management or operations of the business.” (Ibid.) Qualitatively, “the work must be administrative in nature.” (Id. at p. 182.) “ ‘The administrative operations of the business include the work performed by so-called white-collar employees [including executive assistants] engaged in “servicing” a business as, for example, advising the management, planning, negotiating, representing the company, purchasing, promoting sales, and business research and control.’ ” (Harris, supra, 53 Cal.4th at p. 180, fn. 6.) Examples of workers who do not qualify for the exemption include “ ‘bookkeepers, secretaries, and clerks of various kinds [who] hold the run-of-the-[mill] positions in any ordinary business,’ ” i.e., positions devoted to routine clerical tasks. (Ibid.) “ ‘Employees whose work is “directly related” … to general business operations include those … whose responsibility it is to execute or carry it out. The phrase [“directly related”] also includes a wide variety of persons who … carry out major assignments in conducting the operations of the business .…’ ” (Ibid.)

22. The element of discretion and independent judgment, while the same for both the executive and administrative exemptions, was a particularly contested issue at trial in connection with defendant’s reliance on the administrative exemption. “[T]he phrase ‘exercise of discretion and independent judgment’ is defined as generally involving ‘the comparison and the evaluation of possible courses of conduct and acting or making a decision after the various possibilities have been considered. The term … implies that the person has the authority or power to make an independent choice, free from immediate direction or supervision and with respect to matters of significance.’ ” (UPS Wage & Hour Cases, supra, 190 Cal.App.4th at p. 1024.) Plaintiffs argued that all escrow officers are constrained “by the requirements of a highly systematic process while handling an escrow account.” Defendant argued the so- called “escrow process” allows for, and often requires, the exercise of discretion and judgment to successfully perform the job. (See generally UPS Wage & Hour Cases, supra, 190 Cal.App.4th at p. 1026 [“merely because an employer requires adherence to regulations, guidelines or procedures does not mean an [employee] does not exercise discretion or judgment”]; id. at p. 1030 [concluding discretion element met for administrative and executive exemptions]; cf. Thomas v. Speedway SuperAmerica, LLC, supra, 506 F.3d at p. 507 [finding element met even though discretion “was somewhat circumscribed by [a] district manager’s supervision and [the employer’s] standardized operating procedures”].) Defendant maintains that Advisory Escrow Officers and 2003 Escrow Officers were properly classified as exempt administrative employees. Defendant further contends most class members with managerial titles, including Branch Managers and Unit Managers, also qualified for the administrative exemption or at least performed exempt administrative tasks in addition to their managerial duties. 3. The Outside Salesperson Exemption Labor Code section 1171 and Wage Order 4 both provide an exemption for outside salespersons. An “ ‘outside salesperson’ ” is defined as any adult employee “who

23. customarily and regularly works more than half the working time away from the employer’s place of business selling tangible or intangible items or obtaining orders or contracts for products, services or use of facilities.” (Wage Order 4, subd. 2(N).) The applicability of this exemption depends on “how the employee actually spends his or her time as well as the employer’s realistic expectations of the job and the extent to which the employee diverges from them.” (Walsh v. IKON Office Solutions, Inc. (2007) 148 Cal.App.4th 1440, 1454.) Defendant did not rely on the outside salesperson exemption during the liability phase of trial. During the restitution phase and in its motion for a new trial, defendant identified three class members to whom the exemption allegedly applied. Our main reason for noting this exemption is because it was the defense at issue in Duran, supra, 59 Cal.4th 1. 4. The Commissioned Employee Exemption Only two wage orders permit an exemption based on commission earnings: Wage Order 4 and IWC wage order No. 7-2001 (Cal. Code Regs., tit. 8, § 11070 (Wage Order 7)). The exemption applies “to any employee whose earnings exceed one and one-half ([1.5]) times the minimum wage if more than half of that employee’s compensation represents commissions.” (Wage Order 4, subd. 3(D); Wage Order 7, subd. 3(D).) Despite the facial simplicity of this language, the exemption has been construed as having additional prerequisite elements. (See, e.g., Peabody v. Time Warner Cable, Inc., supra, 59 Cal.4th at pp. 668–670; Ramirez, supra, 20 Cal.4th at pp. 803–804.) Defendant relied on the commission exemption as a tertiary defense in the liability phase of trial. The defense was asserted with greater emphasis in the restitution phase, but it is only cursorily mentioned in defendant’s appellate briefs. The inadequate briefing results in forfeiture of issues regarding the merits of that defense. (See Aton Center, Inc. v. United Healthcare Ins. Co. (2023) 93 Cal.App.5th 1214, 1241 [a reviewing court need not address undeveloped arguments]; Lewis v. County of Sacramento (2001) 93

24. Cal.App.4th 107, 116 [appellate review is limited to issues “adequately raised and briefed”].) C. The Unfair Competition Law The common law tort of unfair competition “developed as an equitable remedy against the wrongful exploitation of trade names and common law trademarks that were not otherwise entitled to legal protection.” (Bank of the West v. Superior Court (1992) 2 Cal.4th 1254, 1263.) The UCL’s predecessor statute, former section 3369 of the Civil Code, was amended in 1933 to provide for injunctive relief against any “unfair or fraudulent business practice” and/or “unfair, untrue, or misleading advertising.” (Stats. 1933, ch. 953, § 1, p. 2482; Nationwide Biweekly Administration, Inc. v. Superior Court (2020) 9 Cal.5th 279, 297–298 (Nationwide Biweekly Administration).) The UCL thus began “as a Depression-era enactment that for 30 years did little more than reinforce common law proscriptions against trade name infringement.” (Stern, Cal. Practice Guide: Bus. & Prof. Code, § 17200 Practice (The Rutter Group 2025) ¶ 1:2.) The availability of monetary restitution as a UCL remedy was established by case law in 1973 and codified in 1976. (Korea Supply Co. v. Lockheed Martin Corp. (2003) 29 Cal.4th 1134, 1147.) “In 1977, the Legislature moved the relevant sections of the Civil Code embodying the UCL into the Business and Professions Code at sections 17200 to 17206.” (Nationwide Biweekly Administration, supra, 9 Cal.5th at p. 299.) The UCL was amended in 1992 to expansively define unfair competition as “any unlawful, unfair or fraudulent business act or practice[,] and unfair, deceptive, untrue or misleading advertising[,] and any act prohibited by [the false advertising law (Bus. & Prof. Code, § 17500 et seq.)].” (Bus. & Prof. Code, § 17200; Stats. 1992, ch. 430, § 2; see Kasky v. Nike, Inc. (2002) 27 Cal.4th 939, 949.) The modern UCL is intended to protect business competitors and consumers “by promoting fair competition in commercial markets for goods and services.” (Kasky v.

25. Nike, Inc., supra, 27 Cal.4th at p. 949.) By broadly proscribing “any” unlawful business “act” or “practice” (Bus. & Prof. Code, § 17200), “the UCL ‘ “borrows” ’ rules set out in other laws and makes violations of those rules independently actionable.” (Zhang v. Superior Court (2013) 57 Cal.4th 364, 370.) For that reason, UCL claims are often derivative of other separately pleaded causes of action. “Virtually any law—federal, state or local—can serve as a predicate for an action under [the UCL].” (Smith v. State Farm Mutual Automobile Ins. Co. (2001) 93 Cal.App.4th 700, 718.) “However, a practice may violate the UCL even if it is not prohibited by another statute. Unfair and fraudulent practices are alternate grounds for relief.” (Zhang, at p. 370.) As recognized in Cortez v. Purolator Air Filtration Products Co. (2000) 23 Cal.4th 163 (Cortez), unlawfully withheld overtime pay may be recovered as restitution in a UCL action. (Id. at pp. 173–177.) Moreover, the UCL’s four-year statute of limitations (Bus. & Prof. Code, § 17208) effectively supersedes the three-year statute of limitations otherwise applicable to wage and hour claims under the Labor Code (see Code Civ. Proc., § 338, subd. (a)). (Cortez, at pp. 167, 178–179.) “Thus, the general rule is that a UCL cause of action borrows the substantive portion of the borrowed statute to prove the ‘unlawful’ prong of that statute, but not the limitations procedural part of the borrowed statute.” (Blanks v. Seyfarth Shaw LLP (2009) 171 Cal.App.4th 336, 364.) Because of the UCL’s four-year limitations period, plaintiffs were able to include the 2003 Escrow Officers in the certified classes and extend the class period back to April 2003 instead of April 2004. “While the scope of conduct covered by the UCL is broad, its remedies are limited.” (Korea Supply Co. v. Lockheed Martin Corp., supra, 29 Cal.4th at p. 1144.) The equitable nature of a UCL action precludes the recovery of damages or financial penalties by private plaintiffs. (Ibid.; see Bus. & Prof. Code, § 17206 [authorizing civil penalties only in actions brought by the government].) Injunctions are “ ‘the primary form of relief available under the UCL,’ while restitution is merely ‘ancillary.’ ”

26. (Clayworth v. Pfizer, Inc. (2010) 49 Cal.4th 758, 790.) “The object of restitution is to restore the status quo by returning to the plaintiff funds in which he or she has an ownership interest.” (Korea Supply, at p. 1149.) “Where the remedies invoked are purely and exclusively equitable, the right to a jury trial does not exist.” (Tibbitts v. Fife (1958) 162 Cal.App.2d 568, 572; accord, Hodge v. Superior Court (2006) 145 Cal.App.4th 278, 284–285.) Claims brought exclusively under the UCL must therefore be tried “by the court rather than by a jury.” (Nationwide Biweekly Administration, supra, 9 Cal.5th at p. 305.) This explains why a bench trial was conducted in this case following plaintiffs’ dismissal of their statutory causes of action under the Labor Code. D. Class Treatment of Employee Misclassification Claims 1. General Principles The class action is “a procedural device for aggregating claims ‘when the parties are numerous, and it is impracticable to bring them all before the court.’ ” (Kim v. Reins International California, Inc. (2020) 9 Cal.5th 73, 86.) The device is authorized by Code of Civil Procedure section 382, but the statute “provides scant legislative guidance regarding the multitude of procedural questions that arise in such complex actions.” 5 (Estrada v. Royalty Carpet Mills, Inc. (2024) 15 Cal.5th 582, 612.) The requirements for class treatment have developed judicially and are generally modeled upon rule 23 of the Federal Rules of Civil Procedure. (See Estrada, at pp. 613–614; In re Tobacco II Cases (2009) 46 Cal.4th 298, 318 (Tobacco II).) Where state law authority is lacking, California courts “may look to the procedures governing federal class actions … for

5 Code of Civil Procedure section 382 states, “If the consent of any one who should have been joined as plaintiff cannot be obtained, he may be made a defendant, the reason thereof being stated in the complaint; and when the question is one of a common or general interest, of many persons, or when the parties are numerous, and it is impracticable to bring them all before the court, one or more may sue or defend for the benefit of all.”

27. guidance.” (Washington Mutual Bank v. Superior Court (2001) 24 Cal.4th 906, 922; see Tobacco II, at p. 318.) The party advocating for class treatment must show “the existence of an ascertainable and sufficiently numerous class, a well-defined community of interest [among class members], and substantial benefits from certification that render proceeding as a class superior to the alternatives.” (Brinker, supra, 53 Cal.4th at p. 1021; accord, Richmond v. Dart Industries, Inc. (1981) 29 Cal.3d 462, 470.) “The community of interest requirement embodies three factors: (1) predominant common questions of law or fact; (2) class representatives with claims or defenses typical of the class; and (3) class representatives who can adequately represent the class.” (Richmond, at p. 470.) Class action proponents must also demonstrate manageability, i.e., that the “ ‘litigation of individual issues, including those arising from affirmative defenses, can be managed fairly and efficiently.’ ” (Cobos v. National General Ins. Co. (2025) 112 Cal.App.5th 1272, 1281, quoting Duran, supra, 59 Cal.4th at p. 29.) Predominance “hinges on ‘whether the theory of recovery advanced by the proponents of certification is, as an analytical matter, likely to prove amenable to class treatment.’ ” (Brinker, supra, 53 Cal.4th at p. 1021.) Generally, “ ‘if the defendant’s liability can be determined by facts common to all members of the class, a class will be certified even if the members must individually prove their damages.’ ” (Id. at p. 1022, italics added.) But in cases alleging the misclassification of workers as exempt from overtime laws, “some proof about how individual employees use their time will often be necessary to accurately determine an employer’s overtime liability.” (Duran, supra, 59 Cal.4th at p. 27.) Since many exemptions depend on whether an employee spends at least 51 percent of their time doing certain types of work, “a misclassification claim has the potential to raise numerous individual questions that may be difficult, or even impossible, to litigate on a classwide basis.” (Ibid.)

28. Class treatment may be appropriate, “even if the case involves an exemption that typically entails fact-specific individual inquiries,” when the employees’ job duties are highly standardized in a manner that requires most time to be spent on nonexempt tasks. (Duran, supra, 59 Cal.4th at p. 31.) Corporate policies may similarly have the resulting effect of “employees uniformly spending most of their time on nonexempt work.” (Ibid.) Outside of those two scenarios, “California courts have been reluctant to certify class actions alleging misclassification.” (Id. at pp. 30–31.) The certification ruling is often the most consequential decision in a class action lawsuit. (3 Newberg and Rubenstein on Class Actions (6th ed. 2026) § 7:18.) If the motion to certify is granted, the case will likely settle. (Duran, supra, 59 Cal.4th at p. 27, fn. 27.) This is especially true of certified employee misclassification cases, which almost never proceed to trial. (Id. at p. 25.) “Settlement should never be treated as a foregone conclusion, however.” (Id. at p. 27.) Trial courts have a continuing obligation to “manage individual issues” after a class has been certified, and to “decertify a class action if individual issues prove unmanageable.” (Id. at p. 29.) 2. Notable Pre-Duran Cases The first California Supreme Court decision to address class treatment of employee misclassification claims was Sav-On Drug Stores, Inc. v. Superior Court (2004) 34 Cal.4th 319 (Sav-On). There, two plaintiffs asserted causes of action under the Labor Code and the UCL alleging their employer (“a drugstore chain”) had misclassified them and approximately 600 to 1400 putative class members as exempt managerial employees. (Id. at pp. 325–326.) The superior court’s decision to grant class certification was upheld by the California Supreme Court. The Sav-On plaintiffs supported their motion for class certification with “evidence that deliberate misclassification was [the employer’s] policy and practice,” and “that, owing in part to operational standardization and perhaps contrary to what [the employer] expected, classification based on job descriptions alone resulted in widespread de facto

29. misclassification.” (Sav-On, supra, 34 Cal.4th at p. 329.) The plaintiffs’ evidence was disputed and arguably less robust than the employer’s opposing evidence, the latter of which included “declarations from 51 current employees …, each [holding one of the two managerial positions in question], describing their work.” (Ibid.) Those declarations were offered to show that the “actual activities performed by [the putative class members], and the amount of time spent by each [putative class member] on exempt activities, … varied significantly from store to store and individual to individual, based on multiple factors .…” (Id. at p. 325.) The Sav-On opinion discusses how trial courts are “ ‘afforded great discretion in granting or denying certification’ ” and need only find that the requirements for class treatment are supported by a preponderance of the evidence. (Sav-On, supra, 34 Cal.4th at pp. 326–327.) “[A] reviewing court is not authorized to overturn a certification order merely because it finds the record evidence of predominance less than determinative or conclusive.” (Id. at p. 338.) The intermediate appellate court in Sav-On erred “to the extent it engaged in any reweighing of that evidence,” and “to the extent it stated or implied that the community of interest requirement for certification mandates that class members’ claims be uniform or identical.” (Ibid.) Further, “neither variation in the mix of actual work activities undertaken … [by putative class members], nor differences in the total unpaid overtime compensation owed each class member, bar class certification as a matter of law.” (Sav-On, supra, 34 Cal.4th at p. 335.) “Individual issues do not render class certification inappropriate so long as such issues may effectively be managed.” (Id. at p. 334.) In theory, widespread employee misclassification may be proven through “statistical evidence, sampling evidence, [and/or] expert testimony” that accounts for individual variation among class members. (Id. at p. 333 & fn. 6; see Brinker, supra, 53 Cal.4th at p. 1054 (conc. opn. of Werdegar, J.).) The Sav-On opinion does not, however, address the practical challenges of implementing those methods. (See Sav-On, at p. 339, fn. 12 [“It is not our role at this

30. stage either to devise or to dictate the methods by which a trial court conducting a particular class action may choose to manage it”].) Following the Sav-On decision, several appellate courts published opinions affirming the denial of class certification in similar wage and hour lawsuits. (E.g., Dailey v. Sears, Roebuck & Co. (2013) 214 Cal.App.4th 974, 977–978; Soderstedt v. CBIZ Southern California, LLC (2011) 197 Cal.App.4th 133, 138 (Soderstedt); Mora v. Big Lots Stores, Inc. (2011) 194 Cal.App.4th 496, 498–499; Arenas v. El Torito Restaurants, Inc. (2010) 183 Cal.App.4th 723, 726; Dunbar, supra, 141 Cal.App.4th at p. 1424.) The trial courts in those cases determined that individual issues arising from exemption defenses weighed against a finding of predominance. In Dailey, for example, “highly individualized inquiries” regarding how much time managers spent on exempt and nonexempt activities would have “dominate[d] resolution of the key issues in [the] case.” (Dailey, at p. 992.) In Dunbar, the evidence showed “significant variation in the work performed by grocery managers from store to store and week to week.” (Dunbar, at p. 1431.) In Soderstedt, there was variation in the individual circumstances of approximately 150 workers classified as administrative employees. (Soderstedt, at pp. 138–139, 148–149.) Appellate courts have also affirmed decertification orders issued at various stages when “individual issues related to an exemption defense threaten[ed] to overwhelm the litigation.” (Duran, supra, 59 Cal.4th at p. 30.) “For example, in Walsh v. IKON Office Solutions, Inc., supra, 148 Cal.App.4th at pages 1445–1448, the court certified an overtime class action involving the outside salesperson exemption. It later decertified the class when discovery revealed that the circumstances of each class member’s employment differed significantly. The Court of Appeal affirmed this ruling, noting that differences in time spent on sales activities and work outside the office meant that adjudication of the exemption would require individual hearings on liability and damages.” (Duran, at p. 30.) “Similarly, Keller v. Tuesday Morning, Inc. (2009) 179

31. Cal.App.4th 1389 upheld the decertification of an overtime class action brought by retail store managers. Although the trial court initially certified a class based on [the Sav-On decision], two years later it determined that individual inquiries concerning how managers spent their time would overwhelm the issues susceptible to classwide proof.” (Duran, at p. 30, citing Keller, at p. 1399.) 3. Duran v. U.S. Bank National Assn., supra, 59 Cal.4th 1 By 2014, nearly a decade after Sav-On, only two class actions predicated on theories of employee misclassification had ever been tried to verdict in a California state court. (Duran, supra, 59 Cal.4th at p. 25.) The first was Bell v. Farmers Ins. Exchange (2004) 115 Cal.App.4th 715 (Bell), which went to trial only on the issue of damages (the employer’s liability was determined on summary adjudication). (Duran, at pp. 16, 25, fn. 23; Bell, at p. 720.) The second was Duran, where allegations of classwide misclassification were litigated at trial. The present case went to trial approximately 15 months after Duran was published and was likely only the second time—and possibly last time—such issues were tried to verdict in this state. Given the scarcity of precedent, we examine Duran and its holdings in detail. The Duran plaintiffs were loan officers employed by a national bank with over 130 branches in California. (Duran, supra, 59 Cal.4th at pp. 12–13.) The plaintiffs “sued for unpaid overtime, claiming they had been misclassified as exempt employees under the outside salesperson exemption.” (Id. at p. 12.) As previously noted, that exemption “applies to employees who spend more than 50 percent of the workday engaged in sales activities outside the office.” (Ibid.) A putative class action complaint alleged three causes of action based on the employer’s failure to pay overtime: (1) a violation of Labor Code section 1194; (2) a violation of the UCL; and (3) the tort of conversion. (Duran, supra, 59 Cal.4th at p. 14 & fn. 3.) The trial court certified a class of 260 current and former employees who held a particular sales position between December 1997 and September 2005. (Duran, at

32. pp. 13–15.) The job title is abbreviated as “BBO” (for business banking officer) throughout the Duran opinion. (Id. at p. 13 & fn. 1.) The propriety of class treatment was disputed from the outset. (Duran, supra, 59 Cal.4th at p. 14.) “Relying on Sav-On …, [the trial court] found common questions of law and fact predominated over individual issues based on evidence that: (1) the BBO position was ‘standardized’; (2) [the employer] classified all BBOs as exempt without examining each employee’s duties or work habits; and (3) [the employer] failed to train or monitor BBOs to ensure that exemption requirements were satisfied.” (Id. at pp. 14– 15.) Following class certification, the parties spent months developing and arguing over “competing trial management plans.” (Duran, supra, 59 Cal.4th at p. 15.) The defendant employer “proposed to divide the class into 20 or 30 groups and have special masters conduct individualized evidentiary hearings on liability and damages.” (Ibid.) The plaintiffs objected, arguing the employer “had no due process right to assert its affirmative defenses against each individual class member.” (Ibid.) “As an alternative, plaintiffs proposed the use of surveys and random sampling, as described in a declaration from [their] statistics expert .… First, the parties would identify all tasks performed by BBOs and classify which were sales related. Next, the amount of time class members typically spent on outside activities would be assessed using a classwide survey. The parties’ experts would then jointly design a random sample of surveyed class members to proceed through focused discovery and a phase one trial. Finally, aggregate, classwide damages would be determined at a phase two trial. Once an aggregate damages figure was established, the parties would agree upon a claims procedure to distribute damages to individual class members.” (Duran, supra, 59 Cal.4th at p. 15.) The defense “strenuously objected to the use of representative sampling.” (Duran, supra, 59 Cal.4th at p. 15.) “There was no precedent for using random sampling to

33. establish liability in a class action involving the outside sales exemption.” (Id. at p. 16, italics omitted.) Moreover, the law was unsettled as to whether statistical sampling could “legitimately be used to prove a defendant’s liability to absent class members” in any type of case. (Id. at p. 39.) The trial court in Duran rejected the various proposals and adopted a much simpler trial plan of its own creation. Under the court’s plan, a “representative witness group” would be created by randomly selecting 20 people from the 260-person class. (Duran, supra, 59 Cal.4th at p. 16.) With exception of the named plaintiffs, no class members other than the selected individuals would testify at trial. “Any findings on liability and damages for this sample would then be extrapolated to the remainder of the class.” (Id. at pp. 15–16.) The defendant’s objections to the plan were overruled, and its subsequent requests for decertification were denied. (Id. at pp. 16–17, 20.) “[A]round the same time the trial court finalized the trial management plan and selected the [representative witness group], plaintiffs moved to dismiss their claims under the Labor Code and proceed solely on a claim for equitable relief under the [UCL].” (Duran, supra, 59 Cal.4th at p. 16.) The motion was granted and resulted in a few litigants opting out of the case, including four people originally chosen for the 20-person group. Over the defendant’s objections, the opt-outs were removed from the group of testifying witnesses and replaced by other class members. (Id. at pp. 16–17.) A total of 21 class members testified in a bifurcated liability phase of trial: the two named plaintiffs and 19 of the randomly selected witnesses. One randomly selected witness did not respond to a trial subpoena. (Duran, supra, 59 Cal.4th at pp. 17–18 & fn. 11.) The 21 class members “all testified that they generally spent more than half of their workday inside bank offices,” thus negating the main element of the disputed exemption. (Id. at p. 18, italics added.) Although the employer was prepared to counter this evidence with sworn declarations from dozens of other class members, the trial court “refused to hear any testimony about the work habits of BBOs not included in the

34. [representative witness group].” (Ibid.) The defense could only attempt to impeach the testifying class members and present evidence of the job requirements and employer expectations through non-class-member testimony from upper management. 6 (Id. at p. 19.) Although some witnesses in the sample group were impeached with their own prior declarations, the trial court “concluded [the employer] did not carry its burden of proof on the outside salesperson exemption.” (Duran, supra, 59 Cal.4th at p. 21.) “Based primarily on testimony from [only 21 class members], the court ruled that the entire class … was misclassified as exempt, and all class members were owed overtime in amounts to be determined in phase two of the trial.” (Ibid.) The court also granted a motion to prohibit the employer “from introducing any evidence pertaining to liability” in the restitution phase. “The court thus barred any challenge to its phase one decision that all class members were misclassified as exempt and all were entitled to overtime compensation.” (Ibid.) “After the second phase of trial, which focused on testimony from statisticians, the court extrapolated the average amount of overtime reported by the sample group [during the liability phase] to the class as a whole, resulting in a verdict of approximately $15 million and an average recovery of over $57,000 per person.” (Duran, at p. 12.)

6 As with the executive and administrative exemptions, the outside salesperson defense presents “[a]ncillary questions [that] include ‘whether the employee’s practice diverges from the employer’s realistic expectations, whether there was any concrete expression of employer displeasure over an employee’s substandard performance, and whether these expressions were themselves realistic given the actual overall requirements of the job.’ ” (Duran, supra, 59 Cal.4th at p. 26, quoting Ramirez, supra, 20 Cal.4th at p. 802.) Those inquiries serve to prevent employers from skirting the law by merely “fashioning an idealized job description that [has] little basis in reality,” and employees from evading a valid exemption by intentionally or negligently spending disproportionate amounts of time on nonexempt work. (Ramirez, at p. 802.)

35. The employer successfully appealed. The appellate court found several reversible errors stemming from “the trial plan’s reliance on representative sampling.” (Duran, supra, 59 Cal.4th at p. 24.) “In addition to reversing the trial court’s judgment, the Court of Appeal ordered the class decertified.” (Ibid.) The California Supreme Court granted review and upheld the appellate court’s disposition “in its entirety.” (Id. at p. 50.) The California Supreme Court identified two general categories of prejudicial error: (1) the methodology used to extrapolate classwide findings and (2) the wholesale exclusion of relevant defense evidence. (Duran, supra, 59 Cal.4th at p. 50.) We will focus on the first category. Certain principles explained by Duran regarding inferential statistics are covered elsewhere in this opinion, but for now it suffices to highlight the following points. “A sample must be randomly selected for its results to be fairly extrapolated to the entire class.” (Duran, supra, 59 Cal.4th at p. 43.) For genuine randomization, each member of the class must have an equal probability of being selected for inclusion in the sample. (Ibid.) Selection bias, which diminishes the reliability of any conclusions extrapolated from the sample, “occurs when members of the [class] are chosen based on a nonrandom criterion or are selectively included or excluded from the sample group.” (Ibid.) The randomness of the 21-person sample in Duran was tainted by the inclusion of two named plaintiffs. “These plaintiffs were selected, not randomly, but by class counsel.” (Duran, supra, 59 Cal.4th at p. 45.) This is not to say class representatives should be excluded from testifying as to liability (id. at p. 43), but “data derived from the named plaintiffs’ testimony would not normally be included in a statistical analysis of a random sample” (id. at p. 45). Randomness was further undermined by the removal of four opt-out parties from the 20-person sample, which also constituted selection bias. (Duran, supra, 59 Cal.4th at p. 45.) A similar error was made in the removal of a fifth person from the original

36. 20-person group (id. at p. 16, fn. 8) “because his work activities differed from those of other BBOs and the court did not consider him to be a ‘true’ BBO” (id. at p. 44). “As a randomly selected member of the class, [the witness] may in fact have been representative of a sizeable, if unknown, group of BBOs whose job duties differed from the norm.… By characterizing [him] as atypical and excluding his testimony, the court imposed a greater uniformity on the [group of testifying witnesses] than its ‘random’ sampling methodology warranted.” (Duran, at p. 44.) In addition to being randomized, a sample “must be sufficiently large to provide reliable information about the larger group.” (Duran, supra, 59 Cal.4th at p. 42.) “It is impossible to determine an appropriate sample size without first learning about the variability in the population,” i.e., variability in the group the sample is intended to represent. (Ibid.; see id. at p. 38, fn. 32 [defining “ ‘population’ ” as used in statistical analysis].) Determining statistical variability and the necessary sample size requires expertise in the field of statistics. (See id. at pp. 33, 42.) Therefore, any trial plan that relies on statistical sampling “must be developed with expert input and must afford the defendant an opportunity to impeach the model or otherwise show its liability is reduced.” (Id. at p. 13.) “With input from the parties’ experts, the court must determine that a chosen sample size is statistically appropriate and capable of producing valid results within a reasonable margin of error.” (Duran, supra, 59 Cal.4th at p. 42.) The margin of error “is a statistical measurement of the reliability of an estimate produced by sampling. It reflects the amount by which the estimate may be wrong given a certain confidence interval.” (Id. at p. 46; see ibid. & fn. 36 [explaining “confidence interval”].) The sampling methods used in Duran were “profoundly flawed” (id. at p. 13) because the sample was not random, the extrapolation model was designed without expert input, the sample size was too small, and the margin of error was “intolerably high” for both the

37. restitution findings (id. at p. 46) and the determination of classwide liability (see id. at pp. 48–49). In summary, one relatively straightforward exemption defense “raised a host of individual issues.” (Duran, supra, 59 Cal.4th at p. 32.) “While common issues among class members may have been sufficient to satisfy the predominance prong for certification, the trial court also had to determine that these individual issues could be effectively managed in the ensuing litigation.” (Ibid.) Manageability “is just as important as the existence of common questions uniting the proposed class” (id. at p. 29), and “a statistical plan for managing individual issues must be conducted with sufficient rigor” (id. at p. 31). “With no sensitivity to variability in the class, the court forced the case through trial with a flawed statistical plan that did not manage but instead ignored individual issues.” (Id. at p. 33.) The Duran opinion instructs that “[i]f statistical evidence will comprise part of the proof on class action claims, the court should consider at the certification stage whether a trial plan has been developed to address its use.” (Duran, supra, 59 Cal.4th at p. 31.) “Rather than accepting assurances that a statistical plan will eventually be developed, trial courts would be well advised to obtain such a plan before deciding to certify a class action. In any event, decertification must be ordered whenever a trial plan proves unworkable.” (Id. at p. 32.) 4. Notable Post-Duran Cases On remand from the California Supreme Court, the Duran case “was reassigned to a new judicial officer” who ruled on competing motions to recertify it as a class action. (Duran v. U.S. Bank National Assn. (2018) 19 Cal.App.5th 630, 634.) To demonstrate manageability, the plaintiffs conducted a survey of putative class members and developed a new trial plan. The survey was intended to show “ ‘margins of errors for various possible numbers of witnesses who could testify at trial,’ ” and statistical proof on issues of liability and restitution. (Id. at pp. 634–635.)

38. The plaintiffs’ new trial plan “proposed using representative testimony with sample sizes ranging from 15 to 50 BBOs, having associated respective margins of error ranging from 11 percent to 5.53 percent.” (Duran v. U.S. Bank National Assn., supra, 19 Cal.App.5th at p. 635.) In other words, if the numbers were accurate, findings of exempt or nonexempt status could be reliably extrapolated to approximately 89 to 94 percent of the entire class (but be wrong for the remaining 6 to 11 percent). (See Duran, supra, 59 Cal.4th at pp. 23 & fn. 19 [converting margin of error percentage to number of class members], 46 [explaining margin of error principles].) The trial plan was largely based on the plaintiffs’ new survey data, which a defense expert opined was tainted by selection bias. (Duran v. U.S. Bank National Assn., supra, 19 Cal.App.5th at pp. 640, 648.) The defense expert further opined that “ ‘far larger sample sizes than those proposed would be required’ ” to achieve a margin of error at or below 11 percent. (Ibid.) On remand, the trial court appointed an independent expert “to advise it on the survey science underpinning the parties’ motions.” (Duran v. U.S. Bank National Assn., supra, 19 Cal.App.5th at p. 636.) Ultimately, the plaintiffs’ survey was deemed “unreliable as evidence of uniformity in how BBOs spent their time, and unreliable as statistical support for selecting a representative witness group to testify as to liability or restitution without causing the inquiry to devolve into a multiplicity of individual mini trials .…” (Id. at p. 650.) This was especially true “in light of the [defendant’s] right to call witnesses outside the sample to establish its affirmative defense.” (Ibid.) The denial of class certification was unanimously affirmed on appeal. (Id. at p. 651.) Other courts have heeded Duran’s admonitions regarding trial plans that rely on statistical evidence. The case of Martinez v. Joe’s Crab Shack Holdings (2014) 231 Cal.App.4th 362 involved a putative class of restaurant employees and the employer’s reliance on the executive exemption. (Id. at pp. 367–368.) The trial court’s denial of class certification was reversed on appeal with instructions to reconsider issues concerning predominance and manageability. Despite faulting the trial court for being

39. overly focused on the likelihood of individual issues, the appellate court noted “[t]he statistical sampling plan proposed by plaintiffs below lacked the specificity contemplated in Duran.” (Martinez, at p. 383.) The plaintiffs were thus advised “to submit a revised plan upon remand that more adequately addresses the concerns expressed by the Supreme Court.” (Ibid.) In Mies v. Sephora U.S.A., Inc. (2015) 234 Cal.App.4th 967, the denial of class certification in an overtime lawsuit involving both the executive and administrative exemptions was affirmed on appeal. The plaintiff had argued the possibility of using “statistical evidence to determine liability on a classwide basis,” but “did not propose or justify any particular statistical method” of proof. (Id. at p. 973.) The appellate court held there was no obligation “to accept her unjustified assurances regarding what she could prove, especially given the declaration of [a defense] expert stating the use of sampling to determine liability with respect to absent class members would not work.” (Id. at p. 985; accord, Dailey v. Sears, Roebuck & Co., supra, 214 Cal.App.4th at pp. 999–1000 [trial courts may reject generic statistical sampling proposals]; see Dunbar, supra, 141 Cal.App.4th at p. 1432 [“the party seeking class certification must explain how [such methods] will effectively manage the issues in question”].) The Parties’ Respective Burdens A major dispute in this appeal concerns which burdens are borne by the respective parties in this type of class action. Plaintiffs raise the issue in response to defendant’s reliance on Duran. They accuse defendant of arguing “that Duran somehow reversed the burden of proof for a defendant employer’s affirmative defenses in a class action trial.” The issue is presented in terms of whether plaintiffs had a burden to prove all class members were misclassified as exempt employees. The subtext of plaintiffs’ argument is that statistical sampling errors in their trial plan do not matter because it was defendant’s burden to prove the exemption defenses.

40. Defendant contends it was plaintiffs’ burden to “demonstrate a policy or practice that amounted to deliberate misclassification or widespread de facto misclassification.” Defendant also contends plaintiffs “were required to demonstrate, via common proof, that ‘misclassification was the rule rather than the exception.’ ” The arguments are derived from Duran and Sav-On, but neither case specifically characterizes theories of misclassification as a burden of proof at trial. (See Duran, supra, 59 Cal.4th at p. 37; Sav-On, supra, 34 Cal.4th at pp. 329–330.) They are theories recognized as being “amenable to class treatment.” (Sav-On, at p. 329.) It is settled that exemptions are affirmative defenses, “and therefore the employer bears the burden of proving the employee’s exemption.” (Ramirez, supra, 20 Cal.4th at pp. 794–795.) Class treatment is proper if exemption defenses can be litigated without excessive individual inquiries into the class members’ responsibilities and apportionment of their time. (See Modaraei v. Action Property Management, Inc. (2019) 40 Cal.App.5th 632, 635–636, 643; Duran v. U.S. Bank National Assn., supra, 19 Cal.App.5th at p. 650; Soderstedt, supra, 197 Cal.App.4th at p. 157.) Employee plaintiffs have the burden to show they worked overtime without receiving premium compensation. (Maldonado v. Epsilon Plastics, Inc. (2018) 22 Cal.App.5th 1308, 1328; Kizer v. Tristar Risk Management (2017) 13 Cal.App.5th 830, 839.) Meeting that requirement is generally sufficient to shift the burden to the defendant employer to prove the applicability of an exemption. But if the same type of claim is litigated in a class action trial, the plaintiffs have likely already promised to provide common proof of widespread misclassification. As we have discussed, class certification is rarely granted in an overtime exemption case without such assurances being made. 7

7 This discussion assumes the typical scenario of class certification being requested by the plaintiff(s) and opposed by the defendant(s). (See Van v. LLR, Inc. (9th Cir. 2023) 61 F.4th 1053, 1066, fn. 10 [using same assumption]; Lessin v. Ford Motor Co. (S.D.Cal.

41. “In wage and hour cases where a party seeks class certification based on allegations that the employer consistently imposed a uniform policy or de facto practice on class members, the party must still demonstrate that the illegal effects of this conduct can be proven efficiently and manageably within a class setting.” (Duran, supra, 59 Cal.4th at p. 29.) The real question, therefore, is which party has the burden of demonstrating the manageability of individual issues arising from exemption defenses at the time of trial. In other words, does plaintiffs’ success at the class certification stage put the onus on the defendant to develop a trial plan that fulfills the plaintiffs’ promise of manageability by statistical proof or other common evidence? The answer is no. The same question was addressed in Marlo, supra, 251 F.R.D. 476, a federal district court ruling later affirmed by the Ninth Circuit Court of Appeals (Marlo v. United Parcel Service, Inc. (9th Cir. 2011) 639 F.3d 942). Michael Marlo, referred to in the decision as “Plaintiff,” filed a putative class action complaint against United Parcel Service, Inc. (UPS) alleging a failure to pay overtime compensation, and failure to provide meal breaks and rest periods, in violation of California law. (Marlo, supra, 251 F.R.D. at p. 479.) UPS relied on the executive exemption under the applicable IWC wage order. (Id. at p. 481 & fn. 3.) The Marlo court certified a class of approximately 1,200 workers employed by UPS in three different supervisory positions. (Marlo, supra, 251 F.R.D. at pp. 479–480.) “Notably, the Court accepted Plaintiff’s representation that common proof of misclassification would be offered to determine the class-wide applicability of the exemption.” (Id. at p. 480.) When the case proceeded beyond motions for summary judgment, and triable issues of fact were found to exist on matters including the exercise of discretion and independent judgment required by the exemption, “the Court became

2024) 756 F.Supp.3d 885, 928 [noting the proponent of class treatment is “usually the plaintiff”].)

42. increasingly concerned that individualized issues may predominate over class-wide issues.” (Id. at p. 479.) “[T]he Court met with the parties on several occasions to discuss various substantive and management issues in preparation for trial .…” (Marlo, supra, 251 F.R.D. at p. 479.) During those conferences, the court repeatedly “inquired of Plaintiff’s counsel how they intended to try this case,” but “received unsatisfactory answers.” (Id. at p. 487.) Beyond the fact of a “uniform classification policy” and certain “standardized procedures,” the plaintiffs intended to rely on individual witness testimony and the results of a survey conducted of 160 class members (roughly 13 percent of the entire class). (Id. at p. 483.) The survey, however, was found unreliable due to “methodological and design problems.” (Id. at p. 486.) “In light of the individualized nature of some aspects of the exemption test,” the court’s manageability concerns “ripened into doubt regarding the continuing efficacy of a class action.” (Id. at p. 480.) Facing decertification, the Marlo plaintiffs argued their only burden was to show UPS classified them as exempt employees and did not pay overtime—“a prima facie case to which UPS was willing to stipulate.” (Marlo, supra, 251 F.R.D. at p. 482.) At that point, according to the plaintiffs, “UPS would have the burden of proving that it properly applied the exemption on a class-wide basis.” (Ibid.) UPS disagreed. It argued plaintiffs’ decision to litigate the matter as a class action resulted in a self-imposed “burden of ultimately proving misclassification on a class-wide basis.” (Ibid.) The court had to decide “whether there is any difference between an individual case, where it is well-established that an employer has the burden of proving that the overtime exemption was appropriate, and a class action challenging an employer’s decision to classify as exempt a group of employees.” (Ibid.) The Marlo court stated its understanding of California law based on the Sav-On decision: “[I]n order to maintain a class action challenging the overtime exemption, a plaintiff must have common evidence to support a legal theory of misclassification, either

43. ‘that deliberate misclassification was defendant’s policy or practice’ or similarly, that ‘classification based on job descriptions alone resulted in widespread de facto misclassification.’ [Citation.] A class action is appropriate if ‘plaintiffs are able to demonstrate pursuant to either scenario that misclassification was the rule rather than the exception .…’ ” (Marlo, supra, 251 F.R.D. at pp. 481–482, fn. omitted, citing and quoting Sav-On, supra, 34 Cal.4th at pp. 329–330.) The Marlo court went on to rule as follows:

“The Court finds that Plaintiff has the ultimate burden of proving misclassification at a class action trial. A plaintiff bringing a class action normally has the ultimate burden of proving class-wide liability at trial. [Citation.] Because Plaintiff has brought a class action challenging UPS’s exemption [determinations] as a policy of misclassification, Plaintiff must be ‘able to demonstrate pursuant to either scenario that misclassification was the rule rather than the exception .…’ [quoting Sav-On], supra, 34 Cal.4th at p. 330 .… Where, as here, Plaintiff has sought the benefit of class treatment, he too must accept its burden. This burden may be met with common proof of misclassification. [Citation.]

“The Court finds this approach consistent with the rule that an employer has the burden of proving the exemption as an affirmative defense in an individual case. As to any individual, UPS would have the burden of proving the exemption was proper. However, as to a class-wide finding of misclassification which is the result Plaintiff seeks here, Plaintiff would have the ultimate burden of showing misclassification on a class-wide basis. This does not mean to succeed at trial, or to otherwise maintain a class action, that Plaintiff is required to show that all or substantially all [class members] were misclassified. Rather, Plaintiff must show that it is more likely than not that UPS’s exemption as applied to [the class members] was a policy or practice of misclassification. In any event, a plaintiff must provide common evidence of misclassification to maintain class certification and proceed with a class action trial.” (Marlo, supra, 251 F.R.D. at pp. 482–483.) It is unnecessary to decide whether Marlo provides an accurate statement of the law on all points. The holding is generally correct insofar as where class plaintiffs have given assurances regarding common proof of misclassification at the certification stage, they will, as a practical matter, have a burden of production on that issue at trial. “Class

44. actions do not create a ‘requirement of common evidence.’ Instead, class litigation may be appropriate if the circumstances of a particular case demonstrate that there is common evidence.” (Duran, supra, 59 Cal.4th at p. 36.) If evidence that was promised to satisfy the manageability requirement does not materialize, there is a strong likelihood of decertification. (See id. at p. 29 [“Trial courts … have the obligation to decertify a class action if individual issues prove unmanageable”]; Cobos v. National General Ins. Co., supra, 112 Cal.App.5th at p. 1281 [noting the proponent of class certification has the burden to demonstrate manageability].) Both Duran and Sav-On indicate that a pledge to provide common proof, when made in support of class certification, is expected to be honored at trial. The Sav-On opinion includes the following language about the underlying certification ruling: “A reasonable court, even allowing for individualized damage determinations, could conclude that, to the extent plaintiffs are able to demonstrate pursuant to either scenario that misclassification was the rule rather than the exception, a class action would be the most efficient means of resolving class members’ overtime claims.” (Sav-On, supra, 34 Cal.4th at p. 330, italics added.) The Duran opinion contains remarks contemplating the production by class plaintiffs of “common proof [that] would be sufficient to call for the employer to defend its claimed exemption.” (Duran, supra, 59 Cal.4th at pp. 37–38, italics added.) Returning to the issue we framed above, Duran cautions that “a misclassification claim has the potential to raise numerous individual questions that may be difficult, or even impossible, to litigate on a classwide basis.” (Duran, supra, 59 Cal.4th at p. 27.) “If statistical methods are ultimately incompatible with the nature of the plaintiffs’ claims or the defendant’s defenses, resort to statistical proof may not be appropriate.” (Id. at p. 40.) Here, defendants steadily maintained that classwide findings of exempt or nonexempt status could not be reliably extrapolated from a fractional sample of class

45. member testimony. Plaintiffs, on the other hand, represented “that every single requirement of Duran” would be “scrupulously adhered to” at trial. The class certification order is “necessarily provisional in that it [is] subject to development of a trial plan that [will] manage the individual issues surrounding [the exemption defenses].” (Duran, supra, 59 Cal.4th at p. 33.) The inability of class plaintiffs to develop such a trial plan, after warranting it could be done using statistical methodology, does not permit them to simply shrug their shoulders and say manageability is the defendant’s problem. It may be that a feasible plan cannot be devised. That is precisely why Duran advises trial courts “to obtain such a plan [from the plaintiffs] before deciding to certify a class action.” (Id. at p. 32.) Plaintiffs’ UCL Arguments Plaintiffs assert that “under the terms of the UCL, proof of liability as to any exempt class representative was sufficient to entitle absent class members to recover restitution for wages they lost as a result of Defendant[’]s unfair business practices.” The phrase “exempt class representative” is used in reference to the named plaintiffs in the relevant certified class. The argument relates to the parties’ dispute over whether common or “classwide” proof of misclassification was necessary. As with the arguments discussed in the previous section, plaintiffs’ contentions are made in the hope of avoiding reversal under Duran. Despite some ambiguity, plaintiffs appear to contend a finding of misclassification for only one class representative, or perhaps each of the six class representatives who testified at trial, triggered a conclusive presumption that all 400 class members were in fact misclassified as exempt employees. The trial court expressed a similar view by declaring “no proof of classwide injury” was required to determine the “liability of NATC.” 8 (Bolding and capitalization omitted.)

8 Plaintiffs’ UCL arguments are set forth in the respondents’ brief under the following subheading: “In a Class Action Trial of a UCL Claim, Once a Named Plaintiff

46. Preliminarily, it is important to remember the purpose and limitations of the class action device. The main objective is “litigative efficiency and economy” (American Pipe & Construction Co. v. Utah (1974) 414 U.S. 538, 556), but “convenience alone cannot justify procedures that substantially curtail the parties’ ability to litigate their case” (Duran, supra, 59 Cal.4th at p. 42). “Class action rules and due process principles do not permit [class] certification … unless it permits the defendant to fairly litigate individual liability questions.” (Kight v. CashCall, Inc. (2014) 231 Cal.App.4th 112, 131; accord, Duran, at p. 35 [“ ‘a class cannot be certified on the premise that [the defendant] will not be entitled to litigate its statutory defenses to individual claims’ ”].) Plaintiffs argue that litigating wage and hour claims under the UCL somehow eliminates the normal requirements for class treatment. They purport to believe the only relevant individual liability questions are those concerning the class representatives, because a finding that one or more named plaintiffs were misclassified will conclusively establish all class members were misclassified. We will explain why plaintiffs’ argument fails, and why the trial court’s related analysis was misguided. Additional Background In its tentative statement of decision for the first phase of trial, the trial court referenced Tobacco II, supra, 46 Cal.4th 298. During its brief remarks about Tobacco II, the court opaquely asserted that defendant had “urged a burden of proof that applied to a cause of action for failure to pay overtime, not one for a breach of the UCL.” The court went on to cite Cortez, supra, 23 Cal.4th at page 177, for the proposition “that restitution can be ordered without proof of individual harm” in a UCL action.

Shows That He or She Suffered Injury in Fact and Lost Money or Property as a Result of the Unfair Competition, No Further Individualized Proof of Injury or Causation is Required to Impose Restitution Liability Against the Defendant in Favor of Absent Class Members.” (Bolding omitted.) As previously noted, the size of the relevant class is approximated.

47. The trial court later issued a final written statement of decision. A discussion labeled “Liability of NATC” began with the subheading, “No Proof of Classwide Injury Required.” (Bolding omitted.) Two paragraphs followed, which are reproduced here verbatim:

“In 2004, Proposition 64 amended the standing requirement to bring an action for relief, by amending Business and Professions Code section 17204. Same now states a claim must be brought by: ‘by a person who has suffered injury in fact and has lost money or property as a result of the unfair competition.’ [Sic.]

“In 2009, the Supreme Court determined that ‘the effect of Proposition 64 is to prevent uninjured private persons from suing for restitution on behalf of others.’ [Tobacco II, supra,] 46 Cal.4th 298, 314. It also found that the new standing requirement applied solely to the class representative, not to the absent class members. ‘Restitution under the UCL and FAL may be ordered without individualized proof of harm.’ People v. Sarpas (2014) 225 Cal. App.4th 1539, 1548.” No explanation was given regarding the perceived relevance of Tobacco II and the issue of standing. The only hint came toward the end of the second phase of trial, when the trial court issued its “Order Adopting and Rejecting Referee’s Recommendations” (later incorporated by reference into the final judgment). (Capitalization omitted.) Defendant had objected to being precluded from asserting any exemption defenses as to individual class members in the restitution phase. The court responded to the objection, in relevant part, as follows:

“The sole cause of action on which this case was tried was for breach of [the UCL], which permits restitution where an unlawful, fraudulent, or unfair practice is proven. Where such is brought as a class action, only the named plaintiffs must prove they lost money or property as a result of the practice. [Citations to Bus. & Prof. Code, § 17204 and Tobacco II, supra, 46 Cal.4th at p. 306.] For absent class members, the proof required is that they were exposed to the wrongful conduct.” Law and Analysis

48. The types of prohibited business practices listed in the UCL—unlawful, unfair, and fraudulent—are sometimes referred to as different “prong[s]” of the statute. (E.g., Tobacco II, supra, 46 Cal.4th at pp. 311–312.) The fraud prong incorporates Business and Professions Code section 17500, which is commonly known as the false advertising law (FAL). (Nationwide Biweekly Administration, supra, 9 Cal.5th at pp. 292, 305.) “The procedures set forth in the FAL and the UCL are in many respects parallel to one another, and the UCL specifically provides that any practice that violates the FAL is also prohibited by the UCL.” (Id. at p. 305.) The “restitutionary remedies” authorized by the UCL and FAL are “identical and are construed in the same manner.” (Cortez, supra, 23 Cal.4th at p. 177, fn. 10.) The “without individualized proof” language relied upon by plaintiffs and the trial court traces back to Fletcher v. Security Pacific National Bank (1979) 23 Cal.3d 442. The case involved a FAL claim based on a bank’s fraudulent/unfair practice of computing “ ‘per annum’ interest rates on the basis of a 360-day year.” (Id. at p. 445.) Fletcher holds that under the FAL, “once an unfair trade practice has been established, a trial court has discretion to order restitution without requiring proof of each class member’s lack of knowledge as to that unfair trade practice.” (Id. at p. 449, italics omitted.) Subsequent California Supreme Court decisions, all involving false or deceptive advertising claims under the UCL and FAL, have rephrased Fletcher’s holding by using the words, “without individualized proof of deception, reliance, and injury.” (E.g., Bank of the West v. Superior Court, supra, 2 Cal.4th at p. 1267; Committee on Children’s Television, Inc. v. General Foods Corp. (1983) 35 Cal.3d 197, 211.) To appreciate the contextual meaning of this language, one must understand the distinct nature of claims brought under the fraud prong of the UCL. “A claim based upon the fraudulent business practice prong of the UCL is ‘distinct from common law fraud. “A [common law] fraudulent deception must be actually false, known to be false by the perpetrator and reasonably relied upon by a victim who incurs

49. damages. None of these elements are required to state a claim for … relief” under the UCL.’ ” (Morgan v. AT&T Wireless Services, Inc. (2009) 177 Cal.App.4th 1235, 1255, quoting Tobacco II, supra, 46 Cal.4th at p. 312.) Under the UCL, “a fraudulent business practice is one that is likely to deceive members of the public.” (Morgan v. AT&T Wireless Services, Inc., supra, 177 Cal.App.4th at p. 1255, italics added.) “An advertisement or promotional practice is likely to deceive if it includes assertions that are (1) untrue, or (2) ‘ “true[, but are] either actually misleading or which [have the] capacity, likelihood or tendency to deceive or confuse the public.” [Citation].’ [Citations.] By focusing on whether ‘members of the public’ are likely to be deceived, the [UCL] views the challenged ad or promotional practice through the eyes of the ‘reasonable consumer’—that is, the ‘ordinary consumer acting reasonably under the circumstances’—unless the advertisement or practice is ‘aimed at a particularly susceptible audience.’ [Citations.] This focus on the ‘reasonable consumer’—rather than any particular consumer—means that an ad or practice may be ‘fraudulent’ even without any ‘ “individualized proof of deception, reliance and injury.” [Citation.]’ ” (Shaeffer v. Califia Farms, LLC (2020) 44 Cal.App.5th 1125, 1135–1136.) In Tobacco II, the California Supreme Court addressed changes to the UCL enacted by voter initiative in 2004. (Tobacco II, supra, 46 Cal.4th at pp. 305–306.) Previously, “any person acting for the interests of itself, its members or the general public” had standing to bring a UCL action. (Bus. & Prof. Code, former § 17204, as amended by Stats. 1993, ch. 926, § 2.) The 2004 amendment restricted private standing to “a person who has suffered injury in fact and has lost money or property as a result of the unfair competition.” (Bus. & Prof. Code, § 17204, as amended by Prop. 64, as approved by voters, Gen. Elec. (Nov. 2, 2004) § 3.) “Tobacco II held that, for purposes of standing in context of the class certification issue in a ‘false advertising’ case involving the UCL, the class members need not be assessed for the element of reliance. Or, in other words, class certification may not be

50. defeated on the ground of lack of standing upon a showing that class members did not rely on false advertising. In short, Tobacco II essentially ruled that, for purposes of standing, as long as a single plaintiff is able to establish that he or she relied on a defendant’s false advertising, a multitude of class members will also have standing, regardless of whether any of those class members have in any way relied upon the defendant’s allegedly improper conduct.” (Cohen v. DIRECTV, Inc. (2009) 178 Cal.App.4th 966, 981.) The Tobacco II opinion notes, with regard to UCL restitution in a false advertising context, “that relief under the UCL is available without individualized proof of deception, reliance and injury.” (Tobacco II, supra, 46 Cal.4th at p. 320.) However, as explained in numerous subsequent cases, the quoted language does not concern or alter the main requirements for class treatment, including the elements of predominance and manageability. (See Downey v. Public Storage, Inc. (2020) 44 Cal.App.5th 1103, 1108– 1109, 1120–1121; Davis-Miller v. Automobile Club of Southern California (2011) 201 Cal.App.4th 106, 123–124; Cohen v. DIRECTV, Inc., supra, 178 Cal.App.4th at pp. 979– 981; Kaldenbach v. Mutual of Omaha Life Ins. Co. (2009) 178 Cal.App.4th 830, 847– 850.) Depending on the plaintiffs’ theories of liability and monetary recovery, the predominance of individual issues may preclude class certification even in cases brought under the UCL’s fraud prong. (E.g., Downey, at p. 1119, [holding trial court “had ample basis to fear it would be ‘required to litigate numerous and substantial questions determining … individual [class members’] right to recover’ ”]; Tucker v. Pacific Bell Mobile Services (2012) 208 Cal.App.4th 201, 228 [“ ‘[I]f the issue of materiality or reliance is a matter that would vary from consumer to consumer, the issue is not subject to common proof, and the action is properly not certified as a class action.’ ”].) “The statement in Tobacco II and prior cases that relief is available under the [UCL] and [FAL] ‘without individualized proof of deception, reliance and injury’ does no more than reaffirm that the focus of both the [UCL] and [FAL] is the ‘reasonable

51. consumer’ and whether the content of the challenged advertisement is ‘likely to deceive’ that ‘reasonable consumer’ [citations], such that there is no need for plaintiffs to prove that ‘individual’ class members were actually, subjectively deceived or actually, subjectively relied on a deceptive advertisement.” (Downey v. Public Storage, Inc., supra, 44 Cal.App.5th at p. 1121.) Moreover, the “without individualized proof” language has never been applied in a wage and hour context. While liability for false advertising may be provable regardless of the individual circumstances of class members, “an employer’s liability for misclassification under most Labor Code exemptions will depend on employees’ individual circumstances.” (Duran, supra, 59 Cal.4th at pp. 36–37, italics added.) “Liability to one employee is in no way excused or established by the employer’s classification of other employees.” (Id. at p. 37.) The ability to recover unpaid wages as a form of UCL restitution was established in Cortez, supra, 23 Cal.4th 163. Contrary to what the trial court suggested below, Cortez does not hold that classwide entitlement to restitution may be found without common proof of liability. The Cortez opinion discusses two of the earliest UCL restitution cases, including Fletcher, and notes the remedy in those cases “was literally restoration of money, the return of money acquired from an individual to that individual.” (Cortez, at p. 177, citing Fletcher v. Security Pacific National Bank, supra, 23 Cal.3d 442 & People v. Superior Court (Jayhill) (1973) 9 Cal.3d 283.) Summarizing Fletcher, the high court explained “that under [the FAL], the trial court had the power to order restitution of money, collected through excess interest charges, to the persons from whom it had been collected even absent individualized proof that the claimant lacked knowledge of the overcharge when the transaction occurred.” (Cortez, at p. 177.) The rationale for allowing overtime restitution is that earned but unpaid wages “are as much the property of the employee who has given his or her labor to the employer in exchange for that property as is property a person surrenders through an unfair

52. business practice.” (Cortez, supra, 23 Cal.4th at p. 178.) “An order that earned wages be paid is therefore a restitutionary remedy authorized by the UCL.” (Ibid.) Unearned wages, on the other hand, are not recoverable as restitution because the requisite property interest is missing. (Lee v. Luxottica Retail North America, Inc. (2021) 65 Cal.App.5th 793, 803.) Therefore, individual class members in an overtime class action are not entitled to UCL restitution if they were properly classified as exempt employees. “[A]ny compensation awarded to the class must be based solely on overtime hours worked by nonexempt employees. Overtime hours worked by exempt employees are irrelevant.” (Duran, supra, 59 Cal.4th at p. 41, citing King & Muraco, Classwide Determinations of Overtime Exemptions: The False Dichotomy Posed by Sav-On and a Suggested Solution (2006) 21 Lab.Law. 257, 268–269.) In Duran, the California Supreme Court acknowledged, but declined to resolve, the following issue:

“The parties disagree about whether standards of proof are relaxed or heightened when a wage and hour class action proceeds solely under the UCL. Citing several false advertising cases, [the] plaintiffs assert that relief under the UCL is generally available without individualized proof of injury. [Citation.] In contrast, [the defendant] argues the UCL imposes a stricter proof requirement for restitution than is otherwise required for damages, because restitution is only available for property the defendant wrongfully obtained through an unfair business practice. [Citation.] Thus, they argue, [the] plaintiffs must show they were actually misclassified and are owed unpaid wages for overtime.” (Duran, supra, 59 Cal.4th at p. 41, fn. 35.) The Duran court did not decide whether misclassification must be shown for each class member as a prerequisite to individual restitution because, inter alia, “the restitution award [there] was subject to [an intolerably] high margin of error.” (Duran, supra, 59 Cal.4th at p. 41, fn. 35.) Under the statistical model used to extrapolate the number of overtime hours worked by all 260 class members, the margin of error was 43.3 percent. As such, findings that each class member worked an average of 11.87 hours of overtime each week may have been 5.14 hours too high (id. at pp. 19–20), i.e., “nearly double the

53. true extent of [the employer’s restitutionary obligation]” (id. at p. 46). The court also noted that in Bell, supra, 115 Cal.App.4th 715, it was determined “a 32 percent margin of error was so large that the resulting damages award violated due process.” (Duran, at p. 46, citing Bell, at pp. 756–757.) “[B]roadly speaking, the law tolerates more uncertainty with respect to damages than to the existence of liability.” (Duran, supra, 59 Cal.4th at p. 40.) In Duran, experts for both sides opined that the statistical model used to determine liability would have had a 13 to 14 percent margin of error even if the sampling methodology had otherwise been perfect. (Id. at pp. 22–23.) The California Supreme Court refrained from deciding whether a 13 percent margin of error would be acceptable (id. at pp. 48–49), but the discussion necessarily implies a conclusion relevant to plaintiffs’ UCL argument in this case. If greater certainty is required for liability findings, and a 32 percent margin of error would be too high for a damages finding, then any statistical model used to conclusively establish classwide liability would have to reliably show, at a minimum, that a significant majority of the class members were misclassified. And that is assuming such an all-or-nothing determination is even proper. 9

9 The Duran opinion does not necessarily condone the type of trial procedure employed in this case, where the results of sampling are used to establish universal liability, regardless of the margin of error or variation among witness testimony, and the defendant is prohibited from further contesting the exemption status of individual class members. (See Duran, supra, 59 Cal.4th at pp. 36–38, 49.) “If the trial proceeds with a statistical model of proof, a defendant accused of misclassification must be given a chance to impeach that model or otherwise show that its liability is reduced because some plaintiffs were properly classified as exempt.” (Id. at p. 49, italics added.) Where sampling evidence does not compel a classwide liability finding for one side or the other, further proceedings or alternative dispositions may be warranted. (See id. at pp. 56–57 (conc. opn. of Liu, J.) For example, “the court might notice patterns that suggest unmanageable variation in particular subgroups, resulting in partial decertification. Or the court might conclude that the individualized evidence shows only a few outliers that can be handled through minitrials without disrupting the class proceeding.” (Id. at p. 56.)

54. We need not reach the issue left undecided in Duran to reject plaintiffs’ UCL argument. The Duran plaintiffs, like plaintiffs here, elected to proceed at trial “solely on a claim for equitable relief under the [UCL].” (Duran, supra, 59 Cal.4th at p. 16.) There were two named plaintiffs in Duran, and both testified along with 19 class members from the so-called representative witness group in a liability phase of trial. (Id. at pp. 18, 45.) The employer defendant was prevented from introducing evidence regarding the exempt status of other class members, but it was afforded a full opportunity to defend against the claims of the named plaintiffs and the sample group. (See id. at pp. 18–19.) If proving liability to the named plaintiffs were sufficient to establish a conclusive presumption of misclassification as to all class members, the California Supreme Court would not have issued a detailed opinion explaining why both phases of the trial were “seriously flawed.” (Id. at p. 33.) “[W]ithout following a valid statistical model developed by experts, the court improperly extrapolated liability findings from a small, skewed sample group to the entire class.” (Ibid., italics added.) The 260 class members in Duran all held the same position, and it was generally agreed they spent most of their time on sales activity. (Duran, supra, 59 Cal.4th at pp. 12–13, 26.) The main dispute was over how much time the class members spent working inside versus outside of the office. (Id. at p. 26.) Here, in contrast, there were approximately 400 class members, nine different positions, and two disputed exemptions hinging on the amount of time spent on administrative and/or managerial tasks. There were eight named plaintiffs in the relevant class, but only six testified at trial. The named plaintiffs had collectively held only four of the nine job titles at issue. To extrapolate liability findings to the entire class based on the individual experiences of the named plaintiffs would be patently unfair and a violation of due process. 10

10 The breakdown of positions held by any of the named plaintiffs who testified at trial was as follows: 2003 Escrow Officer, four; Senior Escrow Officer, none; Special Projects Escrow Officer, none; Escrow Officer Supervisor, none; Advisory Escrow

55. The Trial Plan Was Materially Flawed Additional Background A. 2010 (Class Certification) In 2010, after a few years of discovery, the parties filed competing motions regarding class certification. At the time, the putative representatives for what was eventually designated the “Exempt” class consisted of two former Unit Managers and one former 2003 Escrow Officer. Plaintiffs estimated a total class size of 470 people, including those who were classified as nonexempt while employed by defendant. Plaintiffs’ reliance on certain discovery responses implied that the proposed class definition(s) would not include the positions of Branch Manager, Escrow Manager, or Escrow Operations Manager. As to the claims of employee misclassification, plaintiffs relied on the discovery responses to argue all putative class members “perform[ed] non-exempt work more than 50[ percent] of their work time.” The argument presumed an ultimate finding that all work performed by escrow officers was nonexempt. Plaintiffs had propounded a written interrogatory asking defendant to identify the positions in which employees spent “the majority of their time performing Escrow Officer job duties,” and defendant’s response included 2003 Escrow Officers, Advisory Escrow Officers, and Unit Managers. The response did not include Branch Managers, Escrow Managers, or Escrow Operations Managers. Defendant’s motion papers suggested the size of the putative class was twice as large as plaintiffs were estimating. Based on a declaration from its human resources (HR) director, Diana Bloxham, defendant claimed “approximately 356 of about 974 putative class members were at some point [classified as] exempt employees.”

Officer, one; Unit Manager, two; Branch Manager, one; Escrow Manager, none; Escrow Operations Manager, none.

56. Bloxham’s estimates notably included current and former employees in the positions of Escrow Manager and Branch Manager. Defendant argued the “misclassification claims” were not suitable for class treatment due to the individual issues associated with its exemption defenses. Defendant also complained about the lack of a trial plan to deal with those issues. Plaintiffs argued a trial plan was unnecessary at the certification stage. On the issue of manageability, plaintiffs’ counsel proposed to utilize, inter alia, “pattern and practice evidence, statistical evidence, sampling evidence, [and] expert testimony.” The trial court eventually certified the “Exempt” and “Nonexempt” classes of current and former employees. Noting its uncertainty about the proposed composition of the Exempt class, the court allowed plaintiffs to unilaterally decide which positions would be included in the class definition. Plaintiffs elected to include nine different job titles, including Branch Manager, Escrow Manager, and Escrow Operations Manager. B. 2011–2012 (Early Trial Plan Disputes) By late 2011, defendant had moved to compel the submission of a trial plan. The parties were ordered to meet and confer, and a schedule was set for filing a proposed plan and responses to the proposal. Plaintiffs outlined, in general terms, the anticipated use of “representative testimony to prove liability and damages.” The plan included, in relevant part, “random sampling evidence” and testimony from a “statistician regarding results of sampling analysis concerning job duties and tasks of Exempt Class Members ….” Defendant argued the “planned use of unexplained sampling and representative testimony to establish liability and damages” would be unfair and violate its due process rights. It correctly observed that samples yield only probabilities, “each with a margin of error.” Referencing the parties’ meet and confer efforts, defendant alleged plaintiffs could not explain “how large the proposed sample(s) would be, what data would be used by their ‘expert(s)’ and what margins of error their expert believe[d] [could] be achieved

57. through the use of sampling.” Defendant further observed that “many sizeable samples would need to be taken to address each variable, including nine different job titles over a period of [many] years in approximately 90 offices located throughout the State of California.” In December 2011, the trial court heard arguments on the trial plan. Over defendant’s objections, and “in light of not having any other feasible alternative,” the court stated it was “inclined to allow statistical sampling … subject to defense experts and proper objections and things like that.” Describing the trial plan as being “at an infant stage,” the court instructed plaintiffs’ counsel to “[b]egin formulating your statistical plan and experts.” Plaintiffs committed to providing their statistical methodology, including sample sizes, by the end of January 2012. The parties mutually agreed to a February 2013 trial date. In early 2012, the First Appellate District, Division One, published the decision that was later affirmed by Duran, supra, 59 Cal.4th 1. On April 18, 2012, a status conference was held in this case regarding, inter alia, plaintiffs’ proposed sampling methodology. We have no transcript of the hearing, but the appellate record indicates plaintiffs conceded they had not determined a sampling size or margin of error for any statistical evidence. The record further indicates the court set a briefing schedule for a motion to decertify both classes. In May 2012, defendant filed a decertification motion based on the appellate court decision in Duran and plaintiffs’ failure to develop a statistical methodology. The California Supreme Court granted review in Duran while the motion was pending. Under former rule 8.1105(e)(1) of the California Rules of Court, the grant of review resulted in automatic depublication of the underlying opinion. Plaintiffs’ opposition to the decertification motion argued against any consideration of the (then) pending Duran case. Plaintiffs also informed the trial court they “no longer intend[ed] to engage in a random sampling of [c]lass [m]embers as a tool

58. to establish liability and damages.” Their new proposal was to rely on nonrandom “representative testimony.” Regarding the misclassification claims, plaintiffs argued statistical evidence was unnecessary because all class members “performed essentially the same job processing escrows.” The quoted assertion was, of course, a disputed issue for which common proof was needed in lieu of testimony from hundreds of individual class members. At the motion hearing, defense counsel argued the revised trial plan was “even worse” than before. The defense objected to plaintiffs’ proposal to use a nonrandom sample of roughly 90 class members to extrapolate liability findings for both classes. Defense counsel also referenced a supporting declaration from its statistics expert, Ali Saad, Ph.D., who reportedly opined that a nonrandom sample “is not a representative sample.” 11 Plaintiffs’ counsel stated their intention to use some form of “survey” evidence, explaining, “[W]e’re not getting rid of all statistical evidence, but the random sampling we have backed away from.” Plaintiffs’ counsel and the trial court both noted the lack of an alternative defense plan, which led to defense counsel arguing, “We do not have the burden of coming up with a trial plan in a class action case.” The trial court replied, “I agree. I agree.” Defense counsel later said, “Your Honor mentioned that defendant had chosen not to offer some representative statistical sampling or some survey proposal. That’s because we don’t believe they are possible under the facts … of this case and under the law applicable to the claims in this case.”

11 Defendant’s decertification motion had voluminous attachments, including the declaration of Dr. Saad, which were not designated for inclusion in the record on appeal. During the motion hearing, however, plaintiffs’ counsel acknowledged the expert declaration and noted Saad had criticized the methodology of the proposed plan. Counsel further explained that plaintiffs’ decision to abandon a random sampling plan was partially in response to Saad’s opinions and the (then) pending Duran case.

59. The decertification motion was denied by written order in August 2012. The trial court found plaintiffs had proposed an appropriate “mix of representative testimony and expert testimony, the [latter] to involve sampling and survey evidence.” The following comments were made regarding the lack of randomization: “Defendants are highly unlikely to randomly pick employees and put them on the stand with no idea of what they will say either. Plaintiffs will have their representative witnesses—those representing plaintiffs’ theory. And defendants will have theirs.” Citing Teamsters v. United States (1977) 431 U.S. 324, the trial court described the anticipated evidence as “exactly the type of proof commonly offered to show a wrongful pattern and practice.” 12 Defendant challenged the denial of its motion by petitioning this court for a writ of mandate. An order to show cause was issued in late 2012, which resulted in the February 2013 trial date being vacated. C. 2013–2014 (Pre-Duran and Post-Duran Proceedings) Defendant’s writ petition remained pending throughout 2013. In March 2014, the petition was denied. (North American Title Company v. Superior Court (Mar. 17, 2014, F065900) [nonpub. opn.].) Ten weeks later, the California Supreme Court issued its opinion in Duran. (Duran, supra, 59 Cal.4th 1.) Defendant unsuccessfully sought review by the high court of the denial of its writ petition, and the parties stipulated to a rescheduled trial date of September 21, 2015. D. 2015 (Pretrial Proceedings) The class period was originally defined as “four years prior to the filing of the original [c]omplaint up to the time final judgment is entered.” It was later redefined as running from April 17, 2003, through December 31, 2012. Prior to trial, the combined

12 The Teamsters case, which involved claims of discriminatory employment practices in violation of federal law, is distinguished as inapposite in the Duran opinion. (Duran, supra, 59 Cal.4th at pp. 35–37, citing Teamsters v. United States, supra, 431 U.S. at pp. 328, 335, 360–362 & fn. 50.)

60. size of the Exempt and Nonexempt classes was estimated to be 717 people, including 18 opt-outs. In August 2015, plaintiffs served a witness list identifying approximately 425 class members (inclusive of both classes), plus dozens of additional people, as trial witnesses. Two weeks later, defendant filed an ex parte application for an order shortening the time to file another decertification motion (the fully drafted motion was attached). Defendant argued the witness list (1) showed plaintiffs did not have a workable trial plan and (2) vindicated defendant’s longstanding position on manageability. Defendant further noted/alleged that more than half of the class members on the list had not been deposed, and plaintiffs had previously “fought tooth and nail to prevent [defendant] from taking additional class member depositions .…” The application was briefly heard on September 3, 2015, and continued to the following day. In opening remarks, the trial court noted the trial plan had previously “[consisted] of a hundred witnesses and some nefarious survey that I have still never seen or heard or know of.” The court continued: “Now we’re talking about 488 witnesses. I can tell you right now, without reading any[ more], not even another word, that I would decertify.” Plaintiffs’ attorneys were chastised for not providing an updated trial plan, which the court had apparently been waiting to receive. They were ordered to file the updated plan the next morning and be prepared to show compliance with Duran. On September 4, 2015, the hearing resumed with plaintiffs’ counsel stating, “It would be our intention to show on a class-wide basis from several different sources that all of the people who are classified as exempt in fact do virtually the same job duties.” Counsel further stated, “We will have statistical evidence that will show that people who are classified as managers in an exempt position spend the majority of their time doing the same duties as escrow officers, because we have survey data to show that. [¶] … [¶] We also have answers to interrogatories from the defendants admitting that all job descriptions performed the majority of their time doing the duties of escrow officers.”

61. The defense attorneys claimed they were familiar with the survey evidence and knew it did not constitute valid proof of universal misclassification. Accordingly, and for other reasons stated during the lengthy hearing, the defense argued the trial plan was “completely inconsistent” with Duran. Plaintiffs’ counsel offered further assurances: “We have represented to the Court, and we do this in the trial plan and we did it today, that we will show on a class-wide basis that these people, regardless of job title, all do the same thing the majority of the time.” “Duran has a series of very, very specific directions of what proper survey and statistical evidence should have … [and we] believe that we will demonstrate that every single requirement of Duran has been scrupulously adhered to.” Plaintiffs’ revised trial plan described a case-in-chief that would include “approximately 55 to 70 witnesses.” This estimate was acceptable to the trial court if the total number of trial witnesses, including defendant’s witnesses, did not exceed 200. Based on those numbers, and giving weight to the promise of valid survey evidence, it denied defendant’s application to shorten time and did not reach the merits of the decertification motion. Defendant proposed an alternative trial plan. The written proposal asserted objections under Duran and reiterated defendant’s “position that, because there is no classwide evidence of liability in this case, [the necessity of individual class member testimony] results in this case being unmanageable as a class action.” Subject to its objections, defendant proposed a five-phase trial. “Phase One” would involve a hearing under Evidence Code section 402 regarding the admissibility of plaintiffs’ survey evidence and expert witness testimony. Assuming class treatment remained proper after the first phase, “Phase Two” would focus on plaintiffs’ claims of unlawful policies and practices affecting both classes. “Phase Three,” if necessary, would determine the claims against codefendant NAS. In “Phase Four,” which would occur only if “uniform, classwide” policies/practices and unlawful effects were shown in the earlier phases,

62. defendant “would have the opportunity to establish the absence of liability to specific class members.” The final phase, if necessary, “would involve an individualized assessment of damages.” (Bolding omitted.) Another pretrial conference was held on September 14, 2015, which parts of the record label as the first day of trial. Either way, the trial plan remained unsettled. Plaintiffs’ counsel struggled to articulate various aspects of their anticipated case-in-chief, and the trial court made the following statement: “I have to say I’m a little concerned that Plaintiffs are either unable, which is more concerning, or unwilling, … to tell me exactly what it is, the number of witnesses that they intend to prove and how they intend to prove it.” The parties continued to debate the necessity of “ ‘class-wide proof,’ ” and questions arose over what plaintiffs’ long-promised survey evidence was expected to show. The hearing ended without a clear indication of whether the survey evidence would be used to demonstrate misclassification. E. 2015–2021 (Trial) Defendant’s trial plan was rejected in favor of a bifurcated trial in which the claims of both classes were heard together in a liability phase. Plaintiffs elicited live testimony from 53 class members, 44 of whom were members of the Exempt class or both classes. The rest were exclusively members of the Nonexempt class. The 44 witnesses from the Exempt class were not randomly chosen; each person was selected by counsel. Of those 44, eight were members of the class only by virtue of having been employed as an escrow officer in 2003. The remaining 36 Exempt class members included Branch Managers (24 witnesses), Unit Managers (11 witnesses), and Advisory Escrow Officers (seven witnesses). (All parenthetical numbers are approximated). Two of the 36 witnesses had worked as an Escrow Operations Manager (Douglas Aunkst and Dana Forysthe), and one had worked as an Escrow Manager

63. (Lori Baker). None of the witnesses had ever been an Escrow Supervisor or Special Projects Escrow Officer (two very uncommon positions held by only a few individuals). Plaintiffs also called seven current and former NATC executives who held positions higher than those in the Exempt class, plus HR director Bloxham. During Bloxham’s testimony, plaintiffs introduced defendant’s discovery responses identifying the positions in which employees spent “the majority of their time performing Escrow Officer job duties.” Subject to an objection, defendant had identified the following job titles: “Escrow Officer, Senior Escrow Officer, Junior Escrow Officer, Advisory Escrow Officer, Escrow Unit Manager, Special Projects Escrow Officer, [and] Escrow Officer Supervisor.” Although viewed by plaintiffs as a proverbial smoking gun, the trial court denied their motion to treat the discovery responses as conclusive proof of misclassification. On October 21, 2015, after 17 days of witness testimony, plaintiffs conceded that their survey evidence was only tangentially relevant to liability issues. Moreover, such relevance did not concern the question of misclassification. The survey evidence was found to be unreliable for any purpose, and the related expert witness testimony was ordered stricken. Trial continued with another 13 days of testimony from plaintiffs’ chosen witnesses, concluding on November 18, 2015. A two-week recess followed, during which time defendant filed a detailed and voluminous motion for judgment (Code Civ. Proc., § 631.8). As relevant here, the motion noted (1) plaintiffs’ failure to provide valid statistical data and (2) the methodological errors in the “ ‘representative’ testimony” of select class member witnesses. Defendant highlighted the ratios of testifying witnesses to the sizes of various cohorts within the Exempt class: “For example, … the court heard

64. testimony from only … [six] out of 46 Advisory Escrow Officers, 11 out of 85 Escrow Unit Managers, and 21 out of 158 Branch Managers.” 13 (Fns. omitted.) Trial resumed on December 3, 2015. After hearing argument on defendant’s motion for judgment, the trial court announced its ruling would be deferred. No ruling was ever made. The trial court requested “an offer of proof” regarding the anticipated defense case, and in doing so made the following statements: “[W]hat I don’t want is essentially a tit for tat. [For example,] Christine English testified X, so we’re going to have Sallie Mae testify Y. And then we’re going to go to Ms. Yanos, and then we’re going to have Cindy testify Y. I’m not going to engage in that.” In response to the inquiry, defense counsel discussed, inter alia, calling Branch Managers “falling kind of in two categories; one, folks who just didn’t do escrows at all ….” The trial court interrupted, “Let me stop you there. That’s I think now getting more to what I’m a little concerned about. Rather than an affirmative defense—and clearly that’s fine, and you’re able to do it, I just don’t want there to be another 60 of those [witnesses].” The defense case began with expert testimony on matters related to the claims of the Nonexempt class, which lasted an entire day. On December 7, 2015, following a three-day recess, defense counsel advised the trial court as follows: “We were discussing the comments the Court made last week about having the tit for tat evidence issue, and given that we feel we’ve tried a lot of our case within [plaintiffs’] case, and the fact that the depositions took place a lot closer in time to the events, and also the fact that we can’t

13 Defendant made this argument prior to additional Branch Manager and Advisory Escrow Officer testimony on December 8, 2012. Defendant also overlooked the earlier testimony of Michelle Byrnes Bryant. By our count, the trial witnesses included 24 of 156 total Branch Managers, i.e., roughly 15 percent of the cohort. Defendant’s figures with respect to Advisory Escrow Officers and Unit Managers were within a low single-digit variance of our own estimates from a close examination of the appellate record.

65. talk to these folks [class members], obviously, and they [plaintiffs’ counsel] clearly have, and can, we have decided that we’re going to submit most of the class member testimony through their depositions, so that’s going to shorten the trial considerably.” Relevant to the misclassification issue, the defense case included live testimony from two Escrow Managers (Anita Rubeck and Virginia Williamson) and one Escrow Operations Manager (Pearl Grace). Defendant also submitted transcripts of 30 class member depositions: 13 from the Nonexempt class and 17 from the Exempt class. The latter included approximately six 2003 Escrow Officers, three Unit Managers, three Advisory Escrow Officers, nine Branch Managers, three Escrow Managers (there are discrepancies in the record as to the Escrow Manager classifications), and one Escrow Operations Manager. Plaintiffs submitted five deposition transcripts of Exempt class members, consisting of one 2003 Escrow Officer, two Unit Managers, and two Branch Managers. The first phase of trial concluded with the submission of closing briefs. Defendant’s brief reasserted its objections under Duran. Notwithstanding those objections, defendant argued plaintiffs failed to show the existence of any policies resulting in classwide misclassification. Defendant further argued the testimony of plaintiffs’ own “cherry-picked” witnesses demonstrated the applicability of the executive and administrative exemptions. The trial court ultimately decertified the Nonexempt class for lack of common proof of unlawful policies and Labor Code violations. As to the Exempt class, however, it found NATC did not meet its burden on the exemption defenses (for erroneous reasons discussed later in this opinion). Consequently, all members of the Exempt class were found to have been misclassified and thus eligible to seek restitution. The restitution phase began in 2017 and concluded in 2021. Over 230 individual class members gave live testimony. Defendant was prohibited from asking the witnesses any questions relating to its exemption defenses since those issues were deemed to have

66. been fully resolved in the first phase of trial. At the conclusion of the proceedings, the Exempt class was determined to consist of approximately 400 people (including 14 opt-outs). Analysis Plaintiffs argue Duran involved “the use of a statistical analysis of survey evidence,” which they contend is distinguishable from their reliance upon class member testimony in this case. The word “survey” is used with conspicuous intentionality, e.g., “[n]either the Plaintiffs’ liability case nor the trial court’s liability finding involved extrapolation of anything from survey evidence.” Brushing aside all prior assurances that misclassification would be shown through survey evidence, plaintiffs contend their failure to produce such evidence renders Duran inapplicable. They misread Duran. The Duran plaintiffs “proposed the use of surveys and random sampling” as part of their trial plan. (Duran, supra, 59 Cal.4th at p. 15, italics added.) The proposal was rejected in favor of “representative testimony” from a small group of class members from which liability findings were extrapolated to the entire class. (Id. at p. 16.) The extrapolation model was fundamentally flawed because it was not designed with expert input to determine the necessary sample size and margin of error, there was no assessment of variability among class members, and the sample was not random. (Id. at pp. 42–43.) Here, plaintiffs relied on testimony from roughly 13 percent of the Exempt class to show all class members “performed essentially the same job of processing escrows” and uniformly “[did] the same thing the majority of the time.” Whatever label plaintiffs wish to use, the method of proof was a form of statistical sampling. “Sampling is a methodology based on inferential statistics and probability theory. ‘The essence of the science of inferential statistics is that one may confidently draw inferences about the whole from a representative sample of the whole.’ [Citation.] Whether such inferences

67. are supportable, however, depends on how representative the sample is.” (Duran, supra, 59 Cal.4th at p. 38.) “In general, when a trial plan incorporates representative testimony and random sampling, a preliminary assessment should be done to determine the level of variability in the class.” (Duran, supra, 59 Cal.4th at p. 33.) As in Duran, “[n]o such assessment was done here.” (Ibid.) The trial court seemed frustrated with plaintiffs’ vague and ever-changing trial plan, and it expressed a primary concern for the total number of trial witnesses. There was no inquiry into how many people had ever held the nine positions at issue, nor the ratios of trial witnesses to the sizes of those cohorts. When defendant called attention to the problem, it was ignored. Plaintiffs even criticized defendant for “attempting to invent subclasses.” Plaintiffs and the court both eschewed any consideration of subclassing in favor of an all-or-nothing approach to liability. (See Sav-On, supra, 34 Cal.4th at p. 344 [discussing subclassing as a tool to manage variability among class members]; Cal Rules of Court, rule 3.765(b) [a class action may be “limited to particular issues” and the class “may be divided into subclasses”].) Plaintiffs’ statistical sampling violated every basic requirement of Duran. The sample group of witnesses were not randomly chosen to testify. (Duran, supra, 59 Cal.4th at p. 43 [“A sample must be randomly selected for its results to be fairly extrapolated to the entire class”].) There was no expert input to determine the proper sample size and margin of error (id. at pp. 13, 42), and “the variability of the population” was disregarded (id. at p. 42). Even where valid statistical methods are utilized, such methods “cannot entirely substitute for common proof.” (Duran, supra, 59 Cal.4th at p. 31.) “There must be some glue that binds class members together apart from statistical evidence.” (Ibid.) “Assume that the court permitted proof through random sampling of class members, and that the data, in fact, indicated that one out of every [10 class members was] exempt. How would the finder of fact accurately separate the one exempt [class member] from the

68. nine nonexempt [class members] without resorting to individual mini-trials?” (In re Wells Fargo Home Mortgage Overtime Pay Litigation (N.D.Cal. 2010) 268 F.R.D. 604, 612.) Plaintiffs continue to rely on defendant’s discovery responses as proof “that all job descriptions revolved around the processing of escrows[,] and that all employees in all job titles spent the majority of their time processing escrows.” However, as explained above, the discovery responses did not cover the positions of Branch Manager, Escrow Manager, and Escrow Operations Manager. The appellate record shows approximately 156 class members were Branch Managers during the class period, and two-thirds of them would not have been part of the lawsuit but for their employment in that position. There were also 23 Escrow Managers and nine Escrow Operations Managers (these estimates include a small number of opt-outs and a few discrepancies in the Branch Manager and Escrow Manager designations). The three groups thus comprised nearly half of the Exempt class. Even assuming the discovery responses cured all statistical errors for the cohorts to which they pertained, common proof was lacking for a sizeable portion of the class. Plaintiffs also overstate the degree to which defendant was “[b]ound” by its discovery responses concerning the other groups. (Bolding omitted.) Answers to written interrogatories “do not assume the status of judicial admissions.” (Weiss v. Baba (1963) 218 Cal.App.2d 45, 50; see Weil & Brown, Cal. Practice Guide: Civil Procedure Before Trial (The Rutter Group 2025) ¶ 8:1247 [“Interrogatory answers are not preclusive”].) They may be used as evidentiary admissions to help prove material facts, but a “contradiction between an answer to an interrogatory and a witness’s testimony at trial does not in itself affect the latter’s admissibility.” (Phillips v. Cooper Laboratories (1989) 215 Cal.App.3d 1648, 1661; see Code Civ. Proc., § 2030.410; compare with Code Civ. Proc., § 2033.410, subd. (a) [affirmative responses to requests for admissions are conclusively binding].) “Absent significant prejudice to the opponent, a party may

69. present evidence at variance to prior answers to interrogatories.” (Phillips, at p. 1661.) And if plaintiffs themselves testified to spending none of their time processing escrows or less than half of their time on such work, defendant’s interrogatory responses would not preclude the trier of fact from accepting the witnesses’ testimony as true. Defendant identified errors in the statistical methodology before, during, and after the liability phase. In other words, defendant impeached the extrapolation model. (See Duran, supra, 59 Cal.4th at p. 49 [“If the trial proceeds with a statistical model of proof, a defendant accused of misclassification must be given a chance to impeach that model .…”].) As shown by a closer examination of different cohorts within the Exempt class, defendant also demonstrated significant variation among and between those cohorts. (See id. at p. 32 [“Wide variation among class members is a factor informing whether the exemption question can be resolved by a simple ‘yes’ or ‘no’ answer for the entire class”].) 1. Escrow Managers The title of Escrow Manager was held by approximately 23 class members in 12 different counties during the nearly 10-year class period. We are including three opt-outs and a few people for whom there are record discrepancies concerning their title of Branch Manager versus Escrow Manager, most notably Richard Drumm. Despite the small number of people in this cohort, their inclusion in the universal finding of misclassification accounted for near or above $1 million of defendant’s financial liability (inclusive of the $491,023 awarded to Drumm alone). Plaintiffs’ only witness for this position, Lori Baker, testified to processing escrows during her time as both a Branch Manager and Escrow Manager. She acknowledged a distinction between Branch Manager and Escrow Manager, but never really explained the difference. Baker further acknowledged that her managerial positions involved “jobs that a normal escrow officer would not have done.” She supervised and directed the work of “[seven] to 14” people, and she also made hiring

70. decisions. Baker estimated spending 80 percent of her time on “escrow officer related” tasks and 20 percent on managerial tasks. Defendant’s trial witnesses were Anita Rubeck and Virginia (“Ginny”) Williamson. Rubeck was hired toward the end of the class period, but Williamson was an Escrow Manager from 2004 to 2012. Both answered “no” when asked if they ran an “escrow desk” during their time as managers. “Running a desk” was the parlance used by counsel and witnesses in broad reference to processing escrows and/or performing the duties of an escrow officer. Williamson testified to spending most of her time managing employees in offices throughout several counties in Southern California. Her duties included hiring, firing, and making other staffing decisions for various branches; training personnel; “handling claims”; attending management meetings; and “being available for the staff” to answer questions and resolve issues. The people who reported to her included Branch Managers, “some of the advisory floaters, and some escrow officers.” Defendant also proffered the deposition testimony of Richard Drumm and Patricia Hopper, but the transcripts are not included in the appellate record. Drumm reportedly testified to supervising nine escrow officers and 10 escrow assistants, and spending “well in excess of 50% of his time on managerial and supervisory duties.” Another class member (Ingrid Sheikh) who worked with Drumm testified in the restitution phase that Drumm did not process escrows. 2. Escrow Operations Managers Approximately nine class members worked as Escrow Operations Managers during the class period: Douglas Aunkst, Amy Chadim, Susan Coffman, Dana Forsythe, Susanna Juarez, Janelle Martinsen, Betty Sorensen, Meloney Turner, and Pearl Grace. Only five were awarded restitution, but the collective sum was a high six-figure amount. Plaintiffs elicited testimony from Forsythe and Aunkst, the latter of whom was called as an adverse witness (Evid. Code, § 776).

71. Forsythe worked in Fresno County throughout most of the class period. During her time at the Clovis branch, she held the positions of 2003 Escrow Officer, Unit Manager, and Branch Manager. She processed escrows in all three positions. When her job title changed to Escrow Operations Manager, she was transferred to the Fresno branch and prohibited from running an escrow desk. She testified in deposition to being promoted “for the purpose of training and for the purpose of managing” escrow officers and assistants. Her trial testimony was essentially the same, describing her duties as “training and auditing, making sure that everyone was following the company’s procedures.” Although she still occasionally processed escrows, Forsythe described her main role as managerial. (See 29 C.F.R. § 541.102 (2001); Wage Order 4, subd. 1(A)(1)(e); e.g., UPS Wage & Hour Cases, supra, 190 Cal.App.4th at pp. 1018– 1019 [characterizing training and auditing as exempt management-related duties].) Forsythe oversaw the branches in Clovis and Fresno, as well as the Hanford branch in Kings County, and she traveled to the Concord branch in Contra Costa County for monthly meetings. Plaintiffs’ examination of Aunkst was intended to address other topics, but it touched upon his time as an Escrow Operations Manager in 2003 and 2004. While holding that position, the Branch Managers from six offices in Southern California reported directly to him. Aunkst did not personally perform any escrow work. His duties included recruiting, staffing decisions, and being “the advisory person that escrow officers [would] contact or the branch manager or the next line up the chain to discuss a transaction, to discuss how … to handle a transaction.” Juarez was an Escrow Operations Manager in Northern California during the entire class period, overseeing more than 20 branches. Grace was hired in December 2012 to replace Juarez when she retired. Defendant called Grace as a trial witness and submitted a transcript of Juarez’s deposition. Juarez testified to the primary duties of advising escrow officers and other managers “on anything pertaining to the

72. escrow process[, i.e.,] [t]rouble shooting, questions, [and] problems that arise.” She gave further examples: “Difficulties with clients, difficulties with title issues, difficulties with escrow issues, difficulties with other questions pertaining to the statutes that pertain to [or] impact escrow in any way, shape, or form; with lenders, interacting, et cetera.” 3. Branch Managers Approximately 156 class members worked as Branch Managers in 20 different counties and over 60 cities throughout California. Plaintiffs called 24 witnesses from this group at trial, with the geographic representation skewed toward the counties of Alameda, Santa Clara, and Solano. Deposition testimony from 11 additional Branch Managers (approximated) was collectively submitted by both sides. Some of plaintiffs’ own witnesses gave testimony favorable to the defense, e.g., Debra Alvarez and Shirley Hamilton. A substantial amount of testimony refuted plaintiffs’ theory that all class members spent most of their time processing escrows despite allegedly nominal managerial titles. Defendant proffered the deposition testimony of Carol Bisordi, Mark Bisordi, John Collins, Linda Evans, and Roderick Santa Elena as examples of Branch Managers who never, or generally did not, process escrows as part of their job. Mark Bisordi testified he was not an escrow officer, did not run an escrow desk, and spent only 20 percent of his time on “minimal escrow work.” Mark Bisordi was an opt-out, and plaintiffs abandoned their claims as to Carol Bisordi during the restitution phase. Evans, despite testifying she was not an escrow officer and escrow work was “not applicable” to her job, was awarded $112,173. Santa Elena, who reportedly testified in deposition to spending most of his time on managerial duties, was awarded $103,111. Additional outliers were revealed during the restitution phase, even despite NATC being prohibited from questioning witnesses on topics concerning its exemption defenses. Dottie Agueda, Robert Kirschke, Gary Kramer, Martin Puga, and Pamela Thompson all

73. admitted they did not run an escrow desk and/or performed no escrow work as Branch Managers. Agueda, Kramer, and Pamela Thompson were denied restitution because of their admissions. Puga and Kirschke were awarded $83,850 and $47,248, respectively. Linda Moore testified during the restitution phase to overseeing “approximately 12-plus employees” as Branch Manager of the Brentwood and Oakley offices in Contra Costa County during the early part of the class period. She described her job duties as “[a]ll managerial” and gave examples of “scheduling,” writing “evaluation reports,” “the daily running of the office,” “reporting to [upper] management what was going on in the office,” and attending monthly “manager meetings” in Walnut Creek. When Moore remarked that she also had an escrow desk, plaintiffs’ counsel asked if she spent more time on “escrow work or administrative duties.” Moore replied, “Probably more administrative duties. Probably a 50/50 at that point” (the time frame was unclear). She was awarded $107,296. Office size was a variable that may have affected how Branch Managers apportioned their time. Some offices had only one or two escrow officers, and there were instances where the Branch Manager was the only person running an escrow desk, e.g., Cheryl Fuller and Heather Walker. Judith Engels, who worked in small offices in Solano County, switched back and forth from being classified as a nonexempt escrow officer and an exempt Branch Manager when other people transferred in and out of her branch. Engels testified to spending “[m]aybe 10 percent” of her time on “anything managerial.” A few of plaintiffs’ other trial witnesses also testified to a 90/10 split between escrow work and managerial tasks. They included Gayle Reid, who was impeached with deposition testimony wherein she had estimated it was “50/50.” At the other end of the spectrum were people like Alvarez and Hamilton. Alvarez was a Branch Manager in Placer County during 2009 and 2010, with supervisory authority over escrow officers and assistants in the Roseville and Auburn offices. She did not run an escrow desk in 2009 (an approximate four-month period), and only processed

74. escrows if someone was out sick or on vacation. Alvarez did run an escrow desk during her last five months of employment. However, “nothing stopped on the [managerial] side.” Alvarez testified about her job duties: “If any of the escrow associates had any escrow issues or needed help with drawing files or just had questions on policies and procedures, I was their advisor. If they needed additional help on preparing documents, I did that as well.… I handled customer complaints. I handled escrow claims with the Department of Insurance. I helped in the recruiting efforts. I helped in the sales efforts with trying to get new [sales] associates in the door and put business on the escrow officer’s desk. I handled day-to-day operations such as office supplies and vacations and sick leave. Just whatever came up, that was my job.” Alvarez also confirmed her involvement in the hiring and firing of personnel, being someone who could make “an effective recommendation in that regard.” Hamilton supervised approximately 15 people as the Branch Manager in San Francisco between 2003 and 2009. She testified to having many exempt duties and estimated spending 25 to 30 percent of her time on what plaintiffs’ counsel labeled “administrative” work. However, regarding the 70 to 75 percent of time spent on “escrow” work, a significant portion was devoted to answering questions and resolving problems that arose in the transactions of her subordinate escrow officers and assistants. For example: “I spent a lot of time on phones with [disgruntled or difficult] clients, trying to—in most cases straightening out a situation so that the client was happy and we closed that escrow accordingly.” She testified to spending “half [the] time” dealing with those kinds of issues because there were “so many people” working under her. Based on this and other testimony, defendant argued she clearly met the 51 percent threshold for exempt work. Hamilton’s annual base salary was $108,000 before the class period began in 2003, and it increased to $120,000 (well over $180,000 in today’s dollars at an average inflation rate of 2.5%). She was awarded $349,296.

75. Catherine Berggren was a manager in San Mateo County. Defendant’s records showed her as an Escrow Manager from 2008 to 2009 and Branch Manager thereafter. She was still a Branch Manager when deposed in late 2012. Berggren supervised nine employees across two offices (San Mateo and Half Moon Bay), including escrow officers, assistants, and a sales representative. She testified to delegating work to her subordinates, monitoring their workloads and performance, issuing “directive[s]” as needed regarding time management, reviewing and approving timesheets, recruiting, making hiring recommendations and decisions (and having authority to terminate employment), approving expenses, writing reports, and attending management meetings. Berggren further testified that her subordinates came to her with “a wide variety of questions” on a regular basis: “[S]ometimes they’re very technical questions … [and] sometimes they’re super-simple questions. I mean, there are questions all day long. Feels like every minute.” Berggren unequivocally viewed herself as a manager, but she also had an escrow desk. When asked if she spent more or less than 50 percent of her time on managerial duties, she said it was “a hard question.” Despite being pressed to quantify it, her response was, “[I]t’s probably a pretty good balance, but I don’t keep track.” Although she did not formally opt-out, Berggren testified she did not consider herself an appropriate class member and did not want to be part of the class. She gave deposition testimony in response to a subpoena but did not participate in any phases of the trial. Berggren was awarded $222,290. In a cursory argument, plaintiffs suggest the testimony of “various corporate managers” provided common proof of misclassification vis-à-vis the Branch Manager cohort. They specifically refer to Aunkst, Frank Catomer, Steven Miller, and John Thompson. The argument fails. Citing pages 1299 to 1300 of the reporter’s transcript, plaintiffs allege Aunkst testified “that during the time he was an escrow operations manager, six branch managers

76. reported to him, and they all worked a full-time escrow desk.” First, the record citations are wrong. Second, the paraphrasing of Aunkst’s testimony is incomplete. Aunkst answered “yes” when asked if six unidentified Branch Managers “had an escrow desk.” He then explained, “They had a segment of their job that was the same as an escrow officer that was underneath them.” Aunkst did not purport to know how the six individuals divided their time between processing escrows and performing other duties. He testified as follows:

“[The Branch Managers’] tasks would include the scheduling of associates. If you called in sick for the day, you need vacation time, that’s your contact person. They worked with the building management for their leases [and when] they needed repairs, they handled safety meetings, they conducted the office meetings, they were the go-to person when the escrow officers have problems on a file that they’re trying to make a judgment call on and they want to run that judgment call by another. They’re your advisory person for the branch. So they could spend a majority of their time working with the other units within the branch.… [For example,] helping them with their staffing, maybe hiring, interviewing.” Catomer held the non-class position of Division Manager for the counties of San Joaquin and Stanislaus late in the class period, and he also oversaw San Mateo County (the time period for San Mateo was not entirely clear). When asked if Branch Managers “also operate, typically, as escrow officers,” Catomer said it was true in most, “but not all” cases. In other words, some Branch Managers did not also function as escrow officers. Catomer gave no testimony regarding how Branch Managers divided their time among various tasks. Plaintiffs parenthetically cite to the testimony of Miller at “RT5091–50092” (sic), which merely indicates the witness was a County Manager for Sacramento County during an unspecified period. Elsewhere Miller testified to being the Division Manager for the counties of Contra Costa, Placer, and Sacramento between 2003 and 2008. Relevant to plaintiffs’ argument, Miller answered “yes” when asked if Branch Managers “were also escrow officers” and he agreed “they had similar responsibilities.” He testified to lacking

77. sufficient personal knowledge to estimate how much time Branch Managers spent on managerial tasks versus, in counsel’s words, “regular escrow duties and performing the escrow process.” John Thompson was the Division Manager for the counties of Solano and Yolo between 2001 and 2005. He testified that the Branch Managers in those counties were also escrow officers. Based on his personal observations, Thompson estimated they “probably spent 95 percent of their time processing escrows.” The testimony of John Thompson is not as helpful to plaintiffs as it may appear. First, Thompson verified he and other people in upper management believed the Branch Managers were properly classified as exempt. His testimony was thus consistent with defendant’s position that escrow officers performed exempt tasks. Second, even if “processing escrows” were found to constitute nonexempt work—a finding never actually made by the trial court and which cannot be implied (for reasons explained post)—Thompson admitted he had no knowledge “about the goings-on in offices outside of the eight offices [in] the two counties” he oversaw during part of the class period, i.e., from 2003 to 2005. That limits the scope of his testimony to approximately 12 class members, four of whom were among the select group of Branch Managers chosen by counsel to testify at trial (Engels, Fuller, Rebecca Larson, and Judith Scaife). 14 Plaintiffs called other “corporate managers” as witnesses during the liability phase, including Michelle Byrnes Bryant. Bryant was a Branch Manager and also held non-class positions during the class period. She testified that Branch Managers in the counties of Fresno and Kings were expected to spend the majority of their time on managerial duties.

14 Depending on when John Thompson left the company in 2005, his testimony may have been applicable to the following Branch Managers: Kristi Adnan, Irma Dell, Fuller, Keyna Lagardo, Mark Dobrzensky, Engels, Linda Gabel, Mary Jeffcoat, Larson, Scaife, Leslie Stewart, Louise Stills, and Gennifer Treat.

78. 4. Other Positions The rest of the Exempt class mostly consisted of 2003 Escrow Officers, Unit Managers, and Advisory Escrow Officers. Approximately 207 class members were 2003 Escrow Officers, and more than 60 percent of them never held other class positions. The total number of 2003 Escrow Officers who testified at trial was 21, i.e., a nonrandom sample size of 10 percent with an unknown margin of error. The Unit Manager position was more prevalent in Southern California than in other parts of the state. The 84 people in this group (estimated) collectively worked in 14 different counties, but mainly the counties of Los Angeles (17), Riverside (14), San Diego (12), Fresno (11) and Orange (10). Plaintiffs’ allegedly representative sample of 13 people, which did not include any class members from Riverside County, was unreliable due to selection bias and other statistical errors. 15 The position of Advisory Escrow Officer was held by around 45 class members in 13 counties. Plaintiffs elicited testimony from seven of them during the liability phase: Paulette Anderson, Christine English, Sally Menne, Dawna Monroe, Jennifer Rodriguez (also known as Jennifer Bull Mizrahi), Teresa Spencer, and Adrienne Swanson. The variation among this group was significant, again demonstrating unreliability given the numerous Duran errors. Also, six of the seven witnesses had worked in Los Angeles County and/or Orange County. The sample excluded all class members from central California and other notable areas like Riverside County (four class members), San

15 The ratios by county of plaintiffs’ Unit Manager witnesses, including deposition testimony, are as follows: Alameda, 0 of 2; Contra Costa, 1 of 7; Fresno, 1 of 11; Kings, 1 of 2; Napa, 1 of 1; Sacramento, 1 of 2; San Mateo, 0 of 1; Sonoma, 1 of 1; Stanislaus, 1 of 4; Los Angeles, 4 of 17; Orange, 1 of 10; Riverside, 0 of 14; San Bernardino, 0 of 3; San Diego, 2 of 12. Trial witness Adrienne Swanson worked in both Los Angeles County and Orange County during the class period. Two non-testifying class members also worked in multiple counties. All numbers are approximated.

79. Bernardino County (six class members), and San Diego County (nine class members). (All numbers approximated.) According to defendant’s official job description, Advisory Escrow Officers reported to Escrow Operations Managers or Division Managers (a non-class position in upper management). The job duties were summarized as follows:

“Develops, evaluates, approves or disapproves technical escrow phraseology and instructions. Assists in the investigation of claims and legal actions against the company. Advises customers and other internal company departments in resolving technical problems. Personally participates in resolving customer complaints and disputes. Develops, revises, and maintains good escrow practices and procedures. Provides functional supervision to Escrow Officers.” Other listed duties included training, auditing, and assisting “in the actual processing of escrows when needs arise ….” Menne was an experienced escrow officer who worked in the Santa Clarita office (Los Angeles County) between 2003 and 2005. Defendant’s records showed her classified as an Advisory Escrow Officer during the final 17 months of her employment. When deposed in 2010, Menne testified she had never heard of the job title and did not know she was ever classified as such. At trial, she was questioned only about her duties as an escrow officer. Rodriguez and Swanson similarly denied any memory of having the Advisory Escrow Officer job title. Like Menne, Swanson denied knowing the position even existed. Another witness, Spencer, knew she held the title in the final months of her employment but did not know why. Spencer’s testimony vaguely suggested it was due to her experience handling “sub-escrow” transactions. Monroe, who was classified as an Advisory Escrow Officer for only a few months, remembered the title but did not testify about what the position entailed. English was NATC’s highest revenue-generating escrow officer in the state. English acknowledged her position was “unique among escrow officers” due to the size

80. and complexity of the transactions she handled. She knew her title was Advisory Escrow Officer but felt it was a misnomer because she “never did any advising to anyone.” When questioned about the duties listed in the job description, she denied performing most of them. Anderson, in contrast, testified the job description was generally accurate. From 2003 through 2005, she divided her time between the Fremont and Pleasanton offices in Alameda County. While in Fremont, which was “off and on” based on staffing needs, Anderson performed a mix of advisory and “escrow officer duties.” For example, “[i]f an escrow officer had left the company, [she] would go in and help close out the remaining orders on that desk.” She also briefly served as Branch Manager of the Fremont office. When Anderson was in Pleasanton, she did not have an escrow desk. She explained, “I would be available to assist the branches when they would call in needing help. Sometimes they would need help with the software, they couldn’t get a file to balance or they weren’t sure how to prepare an instruction. Some of the other duties covered following up on our outstanding checks or determine why funds are still being held in a transaction, and then if we’re not able to make contact with those people, then we would start working on the escheatment process.” Anderson estimated the apportionment of her time on a weekly basis: “40 percent was spent doing advisory, helping the branches, and 60 percent would be the other duties of the outstanding checks and the funds held.” The 60 percent was further broken down by time devoted to supervising two assistants to whom she delegated escheatment work (“[p]robably 40 percent”) and time spent “working on the claims administration” (“20 percent”). From December 2005 through October 2006, Anderson performed the same type of advisory functions in Pleasanton and for the Tracy office in San Joaquin County. She did not process escrows during that time. Anderson further testified to knowing that an Advisory Escrow Officer in Contra Costa County, Carolyn Knoll, had “very similar”

81. advisory duties and did not have an escrow desk. 16 The only Advisory Escrow Officer whom Anderson knew to have their own escrow desk was Robin Isaacs. In 2007, Anderson moved to the Pleasant Hill office in Contra Costa County to lead a team of employees responsible for the company-wide implementation of new computer software (“STARS”). Although her new job title was “STARS specialist,” she continued to be classified as an Advisory Escrow Officer on her employment records. Anderson testified she did not perform her previous advisory functions after 2006. Her new duties were focused on training employees to use STARS and answering questions about the software (troubleshooting). After the training period ended, she “found a position in the applications department.” Anderson’s annual salary was over $100,000. Based on the universal finding of misclassification, entitlement to restitution was presumed for the period of April 2003 through January 2012. Anderson was awarded $298,155. There was evidence that another Advisory Escrow Officer, James Pierce, did not process escrows. Defendant’s records showed Pierce held the position in San Bernadino County from 2009 to 2011, and in San Diego County during the latter part of 2011 and early 2012. Trial witness Susan Schafer worked under Pierce during his time in San Bernardino County. Schafer identified him as her supervisor and testified, “He was the advisory escrow officer, but he didn’t work the desk, he ran the management reports. He oversaw our system, our procedure that we did.” Counsel asked, “At any time while you were there in the Redlands office, did Mr. Pierce ever do any actual escrow work that you

16 Knoll was an Advisory Escrow Officer from early 2003 until March 2007. Knoll testified during the restitution phase of trial and explained that she processed escrows as needed if someone “was on vacation or out ill.” The testimony generally indicated that more of her time was spent on advisory duties than on processing escrows. In 2003, she “very seldom needed to work overtime [to fulfill her] own responsibilities as an escrow advisory,” and any excess time was attributable to “processing escrows.” By 2005 she was “doing more training” and spending far more time on her advisory duties than on escrow work.

82. could tell?” Schafer answered “no” and confirmed, in response to a follow up question, that Pierce was “more of an administrator.” Additional Errors Given the trial court’s express familiarity with Duran, it appears to have concluded the statistical sampling errors were mooted by evidence that allegedly precluded NATC’s exemption defenses under any circumstances. Two reasons were given for that conclusion. The first was a perceived technicality related to the salary requirement of Labor Code section 515 and Wage Order 4. The second was an ambiguously worded statement finding NATC’s classification of employees as exempt or nonexempt was done on a “speculative basis” and “based on title alone.” The court’s legal analysis was incorrect. The Trial Court Misconstrued the Salary Requirement 1. Overview The executive and administrative exemptions require that qualifying employees be paid “a monthly salary equivalent to no less than two times the state minimum wage for full-time employment.” (Lab. Code, § 515, subd. (a); accord, Wage Order 4, 1(A)(1)(f), (2)(g).) The requirement has two components. There is an earnings threshold, which is governed by state law. In addition, the pay structure must be salary-based. Because “salary” is not defined in the Labor Code or IWC wage orders, California courts have relied on federal law to determine an employer’s compliance with the second component. (Semprini v. Wedbush Securities, Inc. (2020) 57 Cal.App.5th 246, 251–252.) Subject to certain exceptions, employees must receive, “ ‘on a weekly, or less frequent basis, a predetermined amount … [of] compensation … [that] is not subject to reduction because of variations in the quality or quantity of the work performed.’ ” (Id. at p. 252, quoting 29 C.F.R. § 541.602(a) (2019).) Regarding the earnings threshold, California’s minimum hourly wage changed twice during the class period. From 2003 through 2006, it was $6.75. It increased to

83. $7.50 in 2007, and to $8.00 in 2008. (Dept. Industrial Relations, History of California Minimum Wage [as of May 29, 2026], archived at .) The corresponding monthly earnings requirements were $2,340, $2,600, and $2,773.33. Put differently, the earnings threshold was between $28,080 and $33,280 per year depending on when the class members were employed. 17 There are no disputes over the earnings threshold or any aspect of the salary requirement in terms of pay periods and minimum predetermined amounts of compensation. Defendant’s written policies stated that exempt employees would “receive a minimum pay rate of twice the California State minimum wage rate” and be classified according to “the wage and hour laws of the states where the Company conducts business.” Vicki Whited, a former regional manager and senior vice president at NATC, estimated the base salary for an escrow officer ranged from $4,000 to $8,000 per month (annual salaries of $48,000 to $96,000) between 2003 and 2005. To put those numbers in perspective, an annual salary of $50,000 in 2005 would be more than $80,000 in today’s dollars, and a salary of $100,000 would be over $160,000. (See U.S. Bureau of Labor Statistics, CPI Inflation Calculator, [as of May 29, 2026], archived at .) Most of plaintiffs’ trial witnesses from the Exempt class testified about their salaries. The testimony was consistent with Whited’s estimates. The low end of pay for a 2003 Escrow Officer was around $4,000 per month in base salary, but most witnesses earned more. Salaries were also generally higher for managerial positions. Monroe’s annual salary, for example, went from $51,600 to $57,600 when she became a Unit Manager. Melodie Benton earned $5,000 per month ($60,000 annually) as a Unit

17 We have calculated the earnings thresholds using the applicable minimum wage x 2 x 40 (hours per week) x 52 (weeks per year) ÷ 12 (months). (Accord, Kao v. Holiday (2017) 12 Cal.App.5th 947, 958.)

84. Manager. Stephanie Strickland was recruited in 2003 for her experience and large book of business; she negotiated a starting salary of $8,500 per month ($102,000 annually). Nearly all Branch Managers who testified earned at least $70,000 per year. Many had annual salaries near or above $100,000. Some Branch Managers testified to monthly salaries of $10,000. 2. Additional Background Throughout the class period, defendant distributed employee handbooks containing official policies relevant to the salary-basis issue. The policies covered timekeeping, compensation, and pay periods, as well as sick leave, vacation, and personal leave. Each handbook had a signature page for acknowledgment of receipt of the policies and the obligation to read them. At trial, many class members testified to receiving the handbooks and/or to the authenticity of their signature on the acknowledgement forms. Employees classified as exempt were not required to document the hours they worked, but they were required to report absences. In 2006, the phrase “in whole date increments” was added to the reporting policy. From 2007 onward, the policies clarified that such reporting was necessary only for absences “in whole day increments.” The written policies explained that full-time employees could accrue up to 30 days of paid sick leave (in some years it was 32 days). The stated purpose of the sick leave benefit was to “protect … against loss of pay” for absences “due to personal illness or injury.” Full-time employees also received paid vacation benefits and, during most of the class period, two “paid personal days” each year. During the trial testimony of Aunkst, in his capacity as NATC’s controller for the Southern California region from 2004 onward, plaintiffs’ counsel asked him if exempt escrow officers were subject to a reduction in pay for working “less than 40 hours in a week.” He answered “no.” When asked why not, Aunkst replied, “They’re an exempt associate.” He then added, “We don’t reduce their pay if they’re out for any hours less than eight.”

85. Defendant’s HR director, Bloxham, was questioned on the same topic. She was asked, in very general terms, whether “escrow officers” would be paid if they were absent and had exhausted their sick leave and vacation time. Bloxham answered, “[I]f they run out of sick leave and they wanted to take a sick day off, they could take the day because they’re sick, and it would be unpaid.” When asked if the policy applied to both exempt and nonexempt employees, she testified, “The difference is that an exempt associate would report any [absences] in whole day increments, and then a whole-day increment would be deducted from their sick leave balance or their vacation balance, and if they didn’t have a balance, they still would be paid for that, for that [absence].” (Italics added.) Plaintiffs’ counsel posed a hypothetical scenario in which “an exempt employee … missed two weeks and they’d run out of vacation pay and sick leave.” Bloxham testified those would be unpaid absences. She then specified, “They were off two weeks and they had no vacation or sick balance, they would not be paid for those workweeks. [¶] … [¶] Exempt associates would be paid for the whole workweek if they took an hour or two off in a day ….” She further explained, “If you need to take time off from work and you’re exempt and you have burned through your sick and vacation balances, if it is a whole workweek that you start the two weeks on, then you would not be paid for those times.” In a written opposition to defendant’s motion for judgment, plaintiffs alleged Aunkst had testified “that if an exempt employee is out for more than eight hours, his or her pay will be reduced.” As explained above, Aunkst did not say that. Plaintiffs further alleged such employees could “avoid a reduced paycheck only by cashing in some unused vacation, sick, or personal time.” Based on those alleged circumstances, plaintiffs argued “no class member earns a true ‘salary’ in the sense that it is ‘not subject to reduction based on the quantity or quality of the work,’ and therefore, no class member can qualify for either the administrative or executive exemption.”

86. When the motion for judgment was heard, defense counsel stated, “[I]t appears that … plaintiffs [are] making some kind of argument that the salary basis for exempt folks was not satisfied because there was the potential for deductions for absences of greater than a day. That’s not the law.” Counsel further explained the flaws in plaintiffs’ argument and filed a supplemental trial brief on the issue. As previously noted, the motion for judgment was taken under submission but never ruled upon. The parties revisited the issue in their closing briefs. Plaintiffs copied and pasted the argument they had written in opposition to defendant’s motion for judgment. Defendant’s closing brief did not discuss the issue as extensively as in its prior supplemental brief, but defendant did cite key state and federal authority. The trial court ruled against defendant on the salary-basis issue. The statement of decision contains the following analysis and conclusion:

“[I]f pay is docked for an absence for sickness or disability where there is no leave time left, the pay cannot be characterized as a ‘salary.’ Negri[, supra,] 216 Cal.App.4th 392 .… ‘Deductions from pay may be made when an exempt employee is absent from work for one or more full days for personal reasons, other than sickness or disability.’ Cited is 29 C.F.R. 541.602(b)(1) (emp. Added). The evidence here established that class members were subject to a loss in pay amount if they did not have adequate leave balances for missed days, and for furlough days. The claimed exception for exempt employees fails on this basis alone.” It should be noted that the reference to Negri was without citation to any pages of the opinion. A pinpoint citation was not possible because Negri does not discuss unpaid absences for sickness or disability. The facts of Negri did not involve employee absences or any issues of pay docking. “The question presented … [was] whether a compensation scheme based solely upon the number of hours worked, with no guaranteed minimum, can be considered a ‘salary’ within the meaning of the pertinent wage and hour laws.” (Negri, supra, 216 Cal.App.4th at p. 395.) The opinion holds “that such a payment schedule is not a salary and, therefore, does not qualify the employee as exempt.” (Ibid.)

87. “A salary is generally understood to be a fixed rate of pay as distinguished from an hourly wage.” (Id. at p. 397.) The trial court’s use of the phrase, “Cited is 29 C.F.R. 541.602(b)(1)” seems to imply the emphasized quote regarding absences for reasons other than sickness or disability appears in Negri. It does not. The Negri opinion quotes the general rule, set forth in subpart (a) of part 541.602 of title 29 of the Code of Federal Regulations (part 541.602), that salaried employees must be paid a “ ‘predetermined amount … not subject to reduction because of variations in the quality or quantity of work performed.’ ” (Negri, supra, 216 Cal.App.4th at p. 398, italics omitted.) The trial court’s partial quotation of subpart (b) of part 541.602 is accurate, but the sentence cannot be read in isolation. The very next provision explains, “Deductions from pay may be made for absences of one or more full days occasioned by sickness or disability (including work-related accidents) if the deduction is made in accordance with a bona fide plan, policy or practice of providing compensation for loss of salary occasioned by such sickness or disability.” (29 C.F.R. § 541.602(b)(2).) “The employer is not required to pay any portion of the employee’s salary for full-day absences for which the employee receives compensation under the plan, policy or practice. Deductions for such full-day absences also may be made before the employee has qualified under the plan, policy or practice, and after the employee has exhausted the leave allowance thereunder.” (Ibid., italics added.) 3. Law and Analysis The provisions of part 541.602 are collectively known as the “salary basis test” applicable to FLSA exemptions for executive and administrative employees. (Semprini v. Wedbush Securities, Inc., supra, 57 Cal.App.5th at p. 252.) Part 541.603 of title 29 of the Code of Federal Regulations (part 541.603) further explains the test by identifying circumstances under which improper reductions in pay will or will not preclude an exemption defense. “California statutes and regulations contain no corresponding

88. provisions.” (Rhea v. General Atomics (2014) 227 Cal.App.4th 1560, 1567.) Pursuant to the “general approach” of relying on the FLSA and federal regulations in the absence of conflicting or controlling state law, California courts have applied parts 541.602 and 541.603 to determine whether employees are paid a salary for purposes of Labor Code section 515 and the IWC wage orders. (Rhea, at pp. 1567–1568.) Prior to August 23, 2004, the salary basis test was set forth in part 541.118 of title 29 of the Code of Federal Regulations (former part 541.118). (Baden-Winterwood v. Life Time Fitness, Inc. (6th Cir. 2009) 566 F.3d 618, 627; Negri, supra, 216 Cal.App.4th at p. 398, fn. 3.) All further citations to former part 541.118 are to the version in effect from 2003 through August 22, 2004. Unless otherwise specified, all further citations to parts 541.602 and 541.603 are to the versions effective August 23, 2004, which remained the same throughout all subsequent years of the class period. 18 The pre-2004 and post-2004 salary basis tests are not identical, but for our purposes the differences between former part 541.118 and its recodification as part 541.602 are inconsequential. (See, e.g., Sumuel v. ADVO, Inc. (2007) 155 Cal.App.4th 1099, 1102–1103 & fn. 4 (Sumuel).) There is a material difference between part 541.603 and the pre-August 2004 law, which will be explained, post. The applicable version of part 541.602 provides, in relevant part:

“(a) General rule. An employee will be considered to be paid on a ‘salary basis’ within the meaning of these regulations if the employee regularly receives each pay period 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. Subject to the exceptions provided in paragraph (b) of this section, an exempt employee must receive the full salary for any week in which the employee performs any work without regard to the number of days or hours worked. Exempt

18 Part 541.602 has been amended since the class period ended, but the amendments did not substantively change the relevant provisions concerning unpaid absences. (See 29 C.F.R. § 541.602(a)(1)–(2), (b)(1)–(2) (2025).)

89. employees need not be paid for any workweek in which they perform no work.…

“(b) Exceptions. The prohibition against deductions from pay in the salary basis requirement is subject to the following exceptions:

(1) Deductions from pay may be made when an exempt employee is absent from work for one or more full days for personal reasons, other than sickness or disability. Thus, if an employee is absent for two full days to handle personal affairs, the employee’s salaried status will not be affected if deductions are made from the salary for two full-day absences. However, if an exempt employee is absent for one and a half days for personal reasons, the employer can deduct only for the one full-day absence.

(2) Deductions from pay may be made for absences of one or more full days occasioned by sickness or disability (including work-related accidents) if the deduction is made in accordance with a bona fide plan, policy or practice of providing compensation for loss of salary occasioned by such sickness or disability. The employer is not required to pay any portion of the employee’s salary for full-day absences for which the employee receives compensation under the plan, policy or practice. Deductions for such full-day absences also may be made before the employee has qualified under the plan, policy or practice, and after the employee has exhausted the leave allowance thereunder.…” (29 C.F.R. § 541.602(a) (eff. Aug. 23, 2004).) The quoted language means what it says. “[E]mployers can deduct pay when an employee takes off one or more full days for personal reasons (not including illness or disability). [Citation.] But deductions can be made only for full days missed: ‘if an exempt employee is absent for one and a half days for personal reasons, the employer can deduct only for the one full-day absence.’ [Citation.] Similarly, if an employee takes off one or more full days due to sickness or disability, the employer may deduct compensation for those full days (and only those full days) if ‘the deduction is made in accordance with a bona fide plan, policy or practice’ that compensates for the loss of salary.” (Silloway v. City and County of San Francisco (9th Cir. 2024) 117 F.4th 1070, 1075; accord, Rhea v. General Atomics, supra, 227 Cal.App.4th at p. 1567

90. [acknowledging the regulations permit reductions in pay “for full day absences for personal reasons, sickness or disability”].) The trial court focused on the theoretical possibility of unpaid absences if, in the court’s words, employees “did not have adequate leave balances for missed days.” However, an employer is not required to pay a salaried employee “ ‘for any workweek in which they perform no work.’ ” (Negri, supra, 216 Cal.App.4th at p. 398, quoting 29 C.F.R. § 541.602(a) (2012).) Therefore, HR director Bloxham’s testimony in response to the hypothetical scenario involving a two-week absence due to illness was consistent with part 541.602 and reflected a lawful policy for exempt employees. Bloxam further testified that exempt employees would not have their pay docked for short-term absences of less than a full workweek, even if they had exhausted their leave benefits. Aunkst testified similarly by answering “no” when plaintiffs’ counsel asked if “exempt escrow officers” would be docked pay “if they work[ed] less than 40 hours a week.” Plaintiffs note Aunkst also said, “We don’t reduce their pay if they’re out for any hours less than eight.” (Italics omitted.) Plaintiffs contend Aunkst’s statement arguably implied that “full-day deductions could be made during a week in which work was performed .…” Such deductions are not categorically prohibited. In most instances, such deductions are lawful and have no preclusive effect on an exemption defense. Employers need not pay salaried employees for full-day absences taken for personal reasons “other than sickness or disability.” (29 C.F.R. § 541.602(b)(1).) In addition, “[d]eductions from pay may be made for absences of one or more full days occasioned by sickness or disability … if the deduction is made in accordance with a bona fide plan, policy or practice of providing compensation for loss of salary occasioned by such sickness or disability.” (Id., § 541.602(b)(2), italics added.) Under the latter exception, a reduction in pay for full-day absences “after the employee has exhausted the leave allowance” under the plan or policy is expressly permitted. (Ibid.; accord, Holden v. Cenpatico Behavioral Health, LLC (D.Mass. 2017) 347 F.Supp.3d 77, 87 [“[T]he

91. regulations allow for the deduction of full day absences once sick leave is exhausted”].) “Thus, when either of the two exceptions described in [part] 541.602(b)(1)–(2) are satisfied, an employer is allowed to make full day deductions from pay without the employee losing his or her exempt status.” (Dept. of Industrial Relations, DLSE Opn. Letter No. 2009.11.23 (Nov. 23, 2009) p. 4.) In Sumuel, supra, 155 Cal.App.4th 1099, the appellate court noted the lack of “any state or federal case law” interpreting the phrase “ ‘bona fide plan, policy or practice’ ” as used in part 541.602(b)(2). (Sumuel, at pp. 1108–1109.) The Sumuel court relied upon, and essentially adopted, a test outlined by the United States Department of Labor (DOL) in a 2005 opinion letter: “ ‘[A] plan that has defined sick leave benefits which have been communicated to eligible employees, and that operates as described in the plan, will in general qualify as bona fide. In addition, … the plan must be administered impartially, and its design should not reflect an effort to evade the requirement that exempt employees be paid on a salary basis.’ ” (Ibid., quoting U.S. Dept. of Labor, Wage & Hour Div., Opn. Letter (Jan. 7, 2005); see Sumuel, at p. 1109 [concluding the respondent’s plan was “bona fide” because it “was communicated to employees, operated as described, was administered impartially, and was not designed to evade the overtime pay requirement”].) Given the trial court’s stated belief that unpaid absences “for sickness or disability” were categorically prohibited, it does not appear the court determined or considered whether defendant had a bona fide plan or policy with respect to the leave benefits it provided to class members. Defendant requested specific findings about its plan/policies, but pertinent findings were not made. (See Code Civ. Proc., §§ 632, 634.) In any event, even if adverse findings could have been made, the court was mistaken about the preclusive effect of theoretical unpaid absences. “The salary basis test is not intended as an overtime trap for employers who offer [leave benefits] for a legitimate business reason, just because [the plan’s] design or implementation leaves something to be desired.” (Sumuel, supra, 155 Cal.App.4th at

92. p. 1111.) Part 541.603(a) requires an “actual practice of making improper deductions” rather than a mere possibility of improper deductions. (Accord, Ellis v. J.R.’s Country Stores, Inc. (10th Cir. 2015) 779 F.3d 1184, 1188.) Sporadic violations are generally insufficient to nullify an asserted exemption on salary-basis grounds. (See 29 C.F.R. § 541.603(a) [listing “factors to consider when determining whether an employer has an actual practice of making improper deductions”]; Silloway v. City and County of San Francisco, supra, 117 F.4th at pp. 1084–1085 [citing “cases where isolated errors did not indicate an actual practice”].) The regulation declares that its provisions “shall not be construed in an unduly technical manner so as to defeat [an] exemption” (29 C.F.R. § 541.603(e)), which “further supports the conclusion that the DOL intended [the regulation] to offer independent bases by which an employer might preserve [an] FLSA exemption despite some improper salary deductions” (Rebischke v. Tile Shop, LLC (D.Minn. 2017) 229 F.Supp.3d 840, 855). Even when an “actual practice” is shown, the preclusive effect is limited. “If the facts demonstrate that the employer has an actual practice of making improper deductions, the exemption is lost during the time period in which the improper deductions were made for employees in the same job classification working for the same managers responsible for the actual improper deductions.” (29 C.F.R. § 541.603(b), italics added.) “Employees in different job classifications or who work for different managers do not lose their status as exempt employees.” (Ibid.) Here there was no evidence, or even allegations, of class members not being paid for absences occasioned by sickness or disability. Therefore, no actual practice of improper deductions for such absences could have been found. The statement of decision also mentioned “furlough days,” which alluded to the testimony of Laurel Johnstone. Plaintiffs’ counsel introduced timesheets and pay records showing Johnstone had six unpaid absences while classified as an exempt Unit Manager. Johnstone characterized those absences, which occurred between November 27, 2006,

93. and February 15, 2007, as mandatory “furloughs.” She was the only witness to allege such unpaid absences or the existence of a furlough program. In its briefing to the trial court, defendant noted that for each alleged furlough day, Johnstone had signed documents confirming the absences were voluntary, at her own request, and “ ‘for … personal reasons other than for illness.’ ” Defendant also cited a DLSE opinion letter (Dept. Industrial Relations, DLSE Opn. Letter No. 2009.08.19 (Aug. 19, 2009)) as authority for its position that “a furlough day does not undermine the salary basis test.” Defendant repeats those arguments on appeal, but we need not address the furlough issue on the merits. Defendant was/is correct that even if Johnstone’s six unpaid absences were improper, the preclusive effect on any exemption defenses would be limited to “the time period in which the improper deductions were made for employees in the same job classification [(Unit Managers)] working for the same managers responsible for the actual improper deductions.” (29 C.F.R. § 541.603(b); accord, Beauperthuy v. 24 Hour Fitness USA, Inc. (N.D.Cal. 2011) 772 F.Supp.2d 1111, 1133 [“Under Department of Labor regulations, even if an employer incorrectly docks an exempt employee’s salary, it loses the exemption only for that workweek, not for all time”].) Prior to August 23, 2004, an employer could fail the salary basis test upon a showing of “either an actual practice of making [improper] deductions or an employment policy that create[d] a ‘significant likelihood’ of such deductions.” (Auer v. Robbins (1997) 519 U.S. 452, 461; see Escribano v. Travis County (5th Cir. 2020) 947 F.3d 265, 274 [explaining the different standards applicable before and after August 2004]; Baden-Winterwood v. Life Time Fitness, Inc., supra, 566 F.3d at pp. 627–629 [same].) As under the current regulations, deductions for full-day sickness and disability absences were allowed if the employer had a bona fide plan or policy to provide paid leave for those reasons. (former 29 C.F.R. § 541.118.) Any hypothetical application of the old standard would be limited to the initial 16 months of the nearly 10-year class period, i.e.,

94. April 17, 2003, through August 22, 2004. Defendant argues it passed the salary basis test under the old standard as well, but it is unnecessary to reach that question. The trial court’s erroneous finding of a universally preclusive effect during the entire class period is enough to demonstrate prejudice. Plaintiffs no longer stand behind the arguments they presented to the trial court. They now attempt to frame the issue as one concerning the parties’ evidentiary burdens. Plaintiffs contend defendant “never offered any evidence in support of its burden to prove its exempt classifications satisfied the threshold salary basis test.” The argument is not persuasive. Defendant’s policies relevant to the salary requirement were admitted into evidence, as were the pay records of numerous class members. The documentary evidence, combined with testimony from Bloxham, Aunkst, and many class members who acknowledged being paid above the earnings threshold on a salaried basis, as well as the complete lack of evidence or allegations of improper deductions for absences occasioned by sickness or disability, was sufficient to show defendant’s compliance with the salary requirement. (See, e.g., Archuleta v. Wal-Mart Stores, Inc. (10th Cir. 2008) 543 F.3d 1226, 1233–1234 [evidence held sufficient to meet salary basis test included employer’s “compensation policy for full-time pharmacists, as well as the compensation policy statements that each of its pharmacists signed”].) In summary, neither the theoretical possibility of improper pay docking nor Laurel Johnstone’s six unpaid absences justified the wholesale rejection of NATC’s exemption defenses as to all class members throughout the entire class period. A Uniform Classification Policy Is Not Common Proof of Misclassification Plaintiffs seeking class certification in wage and hour cases often rely on the employer’s “policy” of treating groups of employees alike for exemption classification purposes. Typical phrases include “blanket exemption policy,” “uniform classification policy,” and classification “based on job descriptions [or job title] alone.” (E.g., Marlo v. UPS, Inc., supra, 639 F.3d at p. 948; Myers v. Hertz Corp. (2nd Cir. 2010) 624 F.3d 537,

95. 549; Sav-On, supra, 34 Cal.4th at p. 329; see Beauperthuy v. 24 Hour Fitness USA, Inc., supra, 772 F.Supp.2d at p. 1131; Jimenez v. Domino’s Pizza, Inc. (C.D.Cal. 2006) 238 F.R.D. 241, 251.) Earlier cases gave weight to such evidence at the certification stage (e.g., Sav-On, at pp. 329–330), but many courts now view “uniform exemption policies [as] merely a minor factor in the predominance analysis” (In re Wells Fargo Home Mortgage Overtime Pay Litigation (9th Cir. 2009) 571 F.3d 953, 958). In Duran, the California Supreme Court agreed with federal decisions holding “that the uniform application of an exemption, standing alone, ‘does nothing to facilitate common proof on the otherwise individualized issues.’ ” (Duran, supra, 59 Cal.4th at p. 32, fn. 29, citing and quoting In re Wells Fargo Home Mortgage Overtime Pay Litigation, supra, 571 F.3d at p. 959.) In other words, “a blanket exemption policy ‘ “does not eliminate the need to make a factual determination as to whether class members are actually performing similar duties.” ’ ” (Marlo v. UPS, Inc., supra, 639 F.3d at p. 948.) “[A]n individualized inquiry is necessary … because the policy may properly classify some employees as exempt but not others.” (Soderstedt, supra, 197 Cal.App.4th at p. 153; see Duran, supra, 59 Cal.4th at p. 37 [“Liability to one employee is in no way excused or established by the employer’s classification of other employees”]; Sullivan v. Oracle Corp. (2011) 51 Cal.4th 1191, 1208 [stating that “abstract classification decision[s]” are not unlawful under the UCL].) The trial court, in discussing its universal rejection of the exemption defenses, referred to evidence of employees being classified “based on title alone” and “on a speculative basis, without proper assessment of what true managerial or administrative tasks the various job classifications were charged with doing ….” Similar reasoning was held erroneous in Walsh v. IKON Office Solutions, Inc., supra, 148 Cal.App.4th 1440. An employer is not liable for unpaid overtime, even if it “classifies employees without regard to the law or investigating what work they do, … if the employees were, in fact, subject to the exemption.” (Id. at p. 1461.) Without common proof of actual misclassification,

96. the employer’s liability will necessarily depend on “individualized proof,” which defeats the purpose of a class action. (Id. at p. 1462; see Payton v. CSI Electrical Contractors, Inc. (2018) 27 Cal.App.5th 832, 843 [“The existence of any common policy is not sufficient .… The policy in question must be a means to establish liability on a classwide basis”].) The Doctrine of Implied Findings is Not Applicable “[I]t is a fundamental principle of appellate procedure that a trial court judgment is ordinarily presumed to be correct and the burden is on an appellant to demonstrate, on the basis of the record presented to the appellate court, that the trial court committed an error that justifies reversal of the judgment.” (Jameson v. Desta (2018) 5 Cal.5th 594, 608– 609.) Under the correlative doctrine of implied findings, “the reviewing court must infer, following a bench trial, that the trial court impliedly made every factual finding necessary to support its decision.” (Fladeboe v. American Isuzu Motors, Inc. (2007) 150 Cal.App.4th 42, 48.) However, if the judgment is based on a statement of decision that “does not resolve a controverted issue, or … is ambiguous, and the record shows that the omission or ambiguity was brought to the attention of the trial court [in a procedurally proper manner], … it shall not be inferred on appeal … that the trial court decided in favor of the prevailing party as to those facts or on that issue.” (Code Civ. Proc., § 634.) Furthermore, the doctrine of implied findings does not apply to legal conclusions or otherwise preclude reversal for errors of law. (Orange County Water Dist. v. Alcoa Global Fasteners, Inc. (2017) 12 Cal.App.5th 252, 312.) “[E]ven in the absence of a statement of decision, we are not compelled to resort to a presumption if the record adequately demonstrates the legal theory the court applied.” (Border Business Park, Inc. v. City of San Diego (2006) 142 Cal.App.4th 1538, 1550.) When errors of law affect the trial court’s factfinding and discretionary decisions, the judgment is subject to reversal. (See Ramirez, supra, 20 Cal.4th at pp. 802–803, citing Engalla v. Permanente Medical Group, Inc. (1997) 15 Cal.4th 951, 972–973 [“when trial court applies incorrect standard

97. that keeps it from resolving factual disputes, remand for further factfinding [is] appropriate]; Knapp v. AT&T Wireless Services, Inc. (2011) 195 Cal.App.4th 932, 939 [“A trial court’s decision that rests on an error of law is an abuse of discretion”]; Affan v. Portofino Cove Homeowners Assn. (2010) 189 Cal.App.4th 930, 944–945.) Plaintiffs and defendant both made objections to the trial court’s statement of decision. (See Code Civ. Proc., § 632; Cal. Rules of Court, rule 3.1590.) Defendant requested specific findings regarding its exemption defenses as to the groups of class members who held positions identified in the class definition. Defendant made detailed objections to the court’s lack of findings regarding common proof of misclassification; the expectations and actual duties of each position; which duties/tasks constituted exempt or nonexempt work; how class members apportioned their time performing duties and tasks in each position; and other required elements of the administrative and executive exemptions, e.g., “Whether testifying Unit Managers customarily and regularly exercised discretion and independent judgment.” The trial court’s only real discussion of the misclassification issue was as follows:

“While there was evidence that [NATC] attempted to divide employees into exempt and non-exempt categories after 2003, there was abundant evidence to support the inference that such was done on a speculative basis, without proper assessment of what true managerial or administrative tasks the various job classifications were charged with doing, and that a smattering of managerial duties or title might be bestowed, but without meeting the requirements of the law for an exempt position. There was also convincing evidence that prior to 2004, NATC mischaracterized of employees as exempt based on title alone, such as ‘escrow officer.’ ” (Sic.) The trial court’s explanation is ambiguous. It is unclear whether defendant’s “speculative” classification process was treated as having a preclusive effect on the exemption defenses and/or as common proof of misclassification (both of which would be legally flawed conclusions), or whether the court found actual misclassification based on actual job duties, apportionment of time, or other reasons. Emphasis on the changes in

98. job classifications between 2003 and 2004 indicated a narrow focus on 2003 Escrow Officers and Unit Managers. Also, defendant’s classification decisions “prior to 2004” cover only the first nine months of the class period. The remark about “a smattering of managerial duties” may indicate the court’s general impression about how some class members apportioned their time, but it omits essential findings regarding, inter alia, actual job duties and task classification. Exemption defenses require “a fact-intensive inquiry.” (UPS Wage & Hour Cases, supra, 190 Cal.App.4th at p. 1014.) The applicability of an exemption is a mixed question of law and fact, but the legal analysis cannot be performed until the factual findings are made. (See Sav-On, supra, 34 Cal.4th at p. 330 [“task classification is a mixed question of law and fact appropriate for a court to address separately from calculating the amount of time specific employees actually spend on specific tasks”].) “The primary consideration in a misclassification case pertains to ‘the realistic requirements of the job,” which is a question of fact. (Duran, supra, 59 Cal.4th at p. 32, italics omitted; see Heyen v. Safeway Inc., supra, 216 Cal.App.4th at p. 828.) “The nature of an employee’s activities and the time an employee spends on each activity are questions of fact.” (Stickles v. Atria Senior Living, Inc. (N.D.Cal. 2022) 642 F.Supp.3d 1104, 1110; accord, Rodriguez v. Parivar, Inc. (2022) 83 Cal.App.5th 739, 751–752.) Whether an employee regularly exercises discretion and independent judgment, which is required for both the executive and administrative exemptions, is also a question of fact. (See Bothell v. Phase Metrics, Inc. (9th Cir. 2002) 299 F.3d 1120, 1129.) Independent of defendant’s objections to the ambiguities and omissions, the statement of decision is beset by legal error. As previously explained, the analysis of defendant’s liability to the Exempt class begins with a discussion of the UCL under the heading, “No Proof of Classwide Injury Required.” (Bolding omitted.) The discussion has no logical relevance, thus implying the trial court agreed with plaintiffs’ argument that the UCL allows classwide rejection of exemption defenses without common proof of

99. misclassification. It further suggests the court may have adopted plaintiffs’ erroneous position that misclassification of all “absent class members” is established upon a mere showing of liability to the named plaintiffs. Next, the trial court determined the exemption defenses were categorically precluded by the possibility of unpaid absences “for sickness or disability.” Having misconstrued both the UCL and the salary basis test in ways that substituted for common proof of misclassification, it is reasonably probable the court did not otherwise conduct a full and proper analysis of the exemption defenses. But even assuming it did make unstated findings on all the factual and legal issues raised by those defenses, such findings were undoubtedly affected by the Duran errors. In other words, the findings were made (and extrapolated to the entire class) based on fundamentally flawed and unreliable methods of statistical sampling. In Ramirez, supra, 20 Cal.4th 785, a trial court’s erroneous legal analysis prevented the California Supreme Court from implying findings and using the ordinary substantial evidence test to review the rulings on an employer’s exemption defenses. (Id. at pp. 802–804; see id. at p. 803 [noting it was “unclear from the record that the [trial] court ever saw the need to resolve inconsistent testimony presented at trial”].) A footnote in the opinion states: “On remand, the trial court should, as [the plaintiff] requested, itemize the types of activities that it considers to be sales related, and the approximate average times that it finds the employee spent on each of these activities. Because the question whether a particular activity is sales related is a mixed one of law and fact, this itemization will enable an appellate court to review whether the trial court’s legal classifications are correct, and whether its factual findings are supported by substantial evidence.” (Id. at p. 803, fn. 5.) Plaintiffs’ Additional Arguments In addition to the discovery responses we have discussed, plaintiffs rely on defendant’s answer to “Special Interrogatory No. 21” propounded by named plaintiff

100. Cortina prior to class certification. The interrogatory instructed defendant to identify all facts supporting its “contention that Escrow Officers were exempt from the payment of overtime wages during the COVERED PERIOD.” The appellate record does not disclose how “COVERED PERIOD” was defined. Defendant’s final amended response to Special Interrogatory No. 21, as verified by HR director Bloxham in November 2007, was as follows: “Defendant considers Escrow Officers to be non-exempt employees, and Escrow Officers have been so classified since January 1, 2004.” Plaintiffs have repeatedly argued this statement, standing alone and in conjunction with the previously discussed response to “Special Interrogatory No. 29” identifying job positions in which employees spent most of their time performing “Escrow Officer job duties” (a term also not defined in the record), constituted an admission that all class members were nonexempt employees. Plaintiffs’ argument was asserted in their trial brief, in a motion in limine, in their opposition to defendant’s motion for judgment, and in their closing brief. The trial court denied plaintiffs’ motion in limine to have the interrogatory responses treated as dispositive proof of misclassification. At the end of the liability phase, plaintiffs requested that the statement of decision include language deeming the response to Special Interrogatory No. 21 “binding” upon defendant. The statement of decision omitted any reference to the discovery responses. It is apparent the court was not persuaded by plaintiffs’ argument. Defendant’s position, below and on appeal, is that the answer to Special Interrogatory No. 21 was meant to indicate it “treated Escrow Officers as non-exempt” at the time of the response and since January 2004. (Italics added.) If that was its intention, the answer was poorly worded. However, the explanation is consistent with defendant’s other discovery responses and the testimony of Bloxham, both in deposition and at trial. Defendant’s response also makes sense when considered in light of the “off-the-clock” claims of the Nonexempt class, which were always the main focus of the case until the

101. Nonexempt class was decertified. Defendant never denied its obligation to pay overtime to members of the Nonexempt class based on its internal reclassification of their jobs. Plaintiffs’ “Special Interrogatory No. 37” asked, “If you contend that any of the ESCROW OFFICER JOB POSITIONS identified in Interrogatory No. 29 do not require payment of overtime compensation under California law, please state all facts that support this contention.” (Bolding and underlining omitted.) Defendant’s response identified all positions listed in “Interrogatory No. 29,” including “Escrow Officer,” and essentially recited the summary of duties for each position as found in NATC’s official job descriptions. (Bolding omitted.) When read together, the responses to Special Interrogatory Nos. 21 and 37 show, at most, ambiguity and arguably inconsistent responses. HR director Bloxham testified that the reclassification of various escrow officer positions in 2004 was not because NATC believed those positions were previously misclassified. Bloxham was specifically asked, “So [NATC] reclassified them as nonexempt, but they believed that they were exempt?” Her answer was “yes.” She alleged the reclassification was a strategic business decision to financially incentivize escrow officers to remain with NATC instead of working for other companies. Her testimony about the market demand for escrow officers in the early 2000s (a historic boom period for the real estate industry) was consistent with other witnesses’ testimony indicating that talented escrow officers were “highly recruited.” Some class members notably testified to receiving five-figure retention bonuses (aka “signing bonuses”) to stay with NATC when other companies made efforts to lure them away. We have already explained that interrogatory responses “do not assume the status of judicial admissions.” (Weiss v. Baba, supra, 218 Cal.App.2d at p. 50; see Weil & Brown, Cal. Practice Guide: Civil Procedure Before Trial, supra, ¶ 8:1247.) Generally, “a party may present evidence at variance to prior answers to interrogatories.” (Phillips v. Cooper Laboratories, supra, 215 Cal.App.3d at p. 1661.) Also, relevant to Bloxham’s

102. explanation for the 2004 reclassification, an employer’s decision to pay employees premium compensation for overtime is no more determinative of the employees’ legal status than is a decision to internally classify a group of workers as exempt. (See Arenas v. El Torito Restaurants, Inc., supra, 183 Cal.App.4th at p. 735 “[T]here is no estoppel effect given to an employer’s decision to classify a particular class of employees as exempt—whether right or wrong .…”]; 29 C.F.R. § 541.604(a) (2004) [“An employer may provide an exempt employee with additional compensation without losing the exemption or violating the salary basis requirement”].) Bloxham also testified to the job duties of 2003 Escrow Officers being broader than those of post-2004 escrow officers and including a supervisory component, hence the creation of the Unit Manager position and the subsequent “narrow[ing]” of escrow officers’ duties. In other testimony she stated her understanding/belief that “ ‘escrow work’ ” includes marketing, “directing others,” and “making independent decisions.” Defendant has always maintained, and extensively argued, that 2003 Escrow Officers were properly classified as exempt. “The appropriateness of any employee’s classification as exempt must be based on a review of the actual job duties performed by that employee.” (UPS Wage & Hour Cases, supra, 190 Cal.App.4th at pp. 1014–1015.) Put differently, “whether an employee is exempt depends not only upon factors related to the job, itself (e.g., ‘employer’s realistic expectations’ and ‘realistic requirements of the job’), but also ‘first and foremost’ upon what an employee actually does on the job .…” (Mies v. Sephora U.S.A., Inc., supra, 234 Cal.App.4th at p. 978.) A reviewing court does not make those factual determinations in the first instance on direct appeal, and we express no views in that regard. However, as previously explained, a determination that “Escrow Officer job duties” mostly entailed nonexempt tasks would not establish classwide liability. The interrogatory responses did not pertain to Branch Managers, Escrow Managers, and Escrow Operations Managers. The trial evidence also showed some

103. Advisory Escrow Officers did not spend most of their time processing escrows and may have been properly classified as exempt even if the 2003 Escrow Officers and Unit Managers were not. There are no certified subclasses within the Exempt class, and all trial errors affect the class as a whole. Prejudice and Remedy Prejudice Trial errors are prejudicial when “a different result would have been probable if such error … had not occurred or existed.” (Code Civ. Proc., § 475; see Cal. Const., art. VI, § 13; Cassim v. Allstate Ins. Co. (2004) 33 Cal.4th 780, 800–802.) The errors in this case were clearly prejudicial. As in Duran, “there can be little question that the trial court’s findings on liability and damages would have been different absent its erroneous … reliance on faulty statistical methodology” and the other mistakes discussed herein. (Duran, supra, 59 Cal.4th at p. 50.) Plaintiffs argue Duran is distinguishable because, unlike the defendant employer there, NATC “was not denied an opportunity to present its affirmative defenses” in the liability phase of trial. This case is factually distinguishable in that regard, but plaintiffs’ argument ignores the trial court’s restriction on the number of witnesses for each side and, more significantly, its directive regarding “tit for tat” evidence. In essence, the court informed defendant it considered evidence of variation in the class members’ jobs and work experiences to be nonprobative and did not want such evidence presented. The court later used the term “non-cumulative” in referencing an off-the-record discussion about the deposition transcripts defendant would be submitting in lieu of live witness testimony. However, demonstrating variability among class members goes toward impeaching an opposing party’s statistical sampling model (Duran, supra, 59 Cal.4th at pp. 32–34, 42) and showing a lack of “operational standardization” (Sav-On, supra, 34 Cal.4th at p. 329). As such, the limitations imposed by the court were unwarranted.

104. NATC’s trial counsel stated that they altered the defense plan based on “the comments the Court made … [regarding] the tit for tat evidence issue.” It must also be remembered that NATC was defending against two distinct sets of claims: those of the Exempt class and those of the Nonexempt class. In addition, the court mistakenly believed NATC’s exemption defenses were categorically barred for reasons concerning the salary requirement. Although the defense case was not restricted in the same manner and degree as in Duran, those differences do not make the errors harmless. The prejudice associated with Duran error is the inherent unreliability of flawed statistical sampling. “If sampling is used to estimate the extent of a party’s liability, care must be taken to ensure that the methodology produces reliable results.” (Duran, supra, 59 Cal.4th at p. 42.) “To be reliable, the sample must be sufficiently large and free from bias caused by various sampling errors.” (Id. at p. 22.) The facial persuasiveness of sampling data, whether tending to favor plaintiffs or the defense, does not justify drawing inferences “ ‘from the part to the whole’ ” if the methodology is unsound. (Id. at p. 38.) Trial courts may not disregard “the internal rules of a scientific specialty, such as statistics, and then rely on that specialty’s established reliability as if the rules had been followed.” (Id. at p. 45.) Remedy Defendant proposes three alternative remedies. The first is reversal with instructions to enter a judgment on the merits in its favor and against the Exempt and Nonexempt classes, including all named plaintiffs. The second proposal is to have the judgment reversed and the Exempt class decertified, with orders barring future recertification and vacating the restitution awards to all named plaintiffs on all claims. The third alternative is reversal, decertification, and “vacatur of the named plaintiffs’ restitution awards.” The request for a judgment on the merits is based on plaintiffs’ “failure to adduce classwide proof of misclassification.” Defendant does not provide legal authority for

105. such a disposition. Defendant also fails to identify precedent or authority for the type of order it seeks to prevent future recertification. The absence of common proof of misclassification is fatal to class treatment, but it does not mean the named plaintiffs failed to show they worked uncompensated overtime, i.e., the showing necessary to require an employer to demonstrate the applicability of an exemption. (See Marlo, supra, 251 F.R.D. at p. 485 [“In the absence of common proof, the jury may have sufficient evidence to make judgments as to particular individuals, but lacks a basis to extrapolate from those findings to a class-wide judgment”].) Defendant elsewhere argues the trial court did not “evaluate the application of exemptions as to any particular class member,” but does not address whether any class member met “the initial burden to show he or she worked overtime hours without receiving overtime pay.” (Kizer v. Tristar Risk Management, supra, 13 Cal.App.5th at p. 839.) As such, the argument for a judgment against the entire Exempt class fails to persuade. Defendant’s argument for a judgment against the Nonexempt class is based on plaintiffs’ failure to present classwide proof of unlawful “off-the-clock” policies and resulting wage and hour violations. The lack of such proof is why the trial court decertified the Nonexempt class prior to the restitution phase. A class “can be decertified at any time, even during trial,” when it becomes apparent that “individual issues dominate the case.” (MacManus v. A.E. Realty Partners (1987) 195 Cal.App.3d 1106, 1117.) Moreover, as held in Duran, “decertification must be ordered whenever a trial plan proves unworkable,” i.e., when “individual issues prove unmanageable.” (Duran, supra, 59 Cal.4th at pp. 32, 29, italics added.) Defendant does not convincingly explain why it believes the court was obligated to enter a defense verdict against the entire Nonexempt class instead of decertifying the class. Defendant is correct, however, about reversing the restitution awards for the named plaintiffs in both classes. As explained in the next part of the opinion, the nonconsensual reference proceedings were unauthorized and prejudicial. The trial court

106. delegated its judicial duties not only with respect to issues of restitution (an error in and of itself) but also questions of defendant’s liability to named plaintiffs in the Nonexempt class. The appropriate remedy is the one authorized and applied by the California Supreme Court in Duran: reversal and decertification of the Exempt class. (Duran, supra, 59 Cal.4th at pp. 24, 50.) “Even if certification had once appeared appropriate, it should have been apparent … that individual issues predominated to such an extent that they rendered class treatment impossible.” (Id. at p. 24.) Once plaintiffs conceded they had no survey evidence to support their claims of misclassification, and no plan to meet the basic requirements for statistical sampling by “representative testimony,” the manageability problems could no longer be ignored. “On remand, a new trial will be required for both liability and restitution, and the trial court may entertain a new class certification motion.” (Id. at p. 50.) II. Second Phase of Trial Defendant’s claims regarding the nonconsensual reference proceedings, aka the restitution phase of trial, are not mooted by the reversal for errors in the earlier phase. Defendant’s notice of appeal challenges the entire judgment against it, which includes findings of liability and restitution awards to named plaintiffs who were members of the Nonexempt class. In its appellate briefing, defendant argues for reversal of the named plaintiffs’ restitution awards for their “Non-Exempt Class claims.” Independent of those reasons, the following discussion serves to ensure the same judicial delegation error is not repeated in the event of recertification of the Exempt class on remand. Additional Background During the initial phase of trial, plaintiffs’ counsel asked the trial court “whether the case was bifurcated on the issue of liability and damages.” Counsel explained, “[I]t’s not entirely clear to us … whether it is bifurcated or whether we just put on our whole case all at one time.” The trial court’s response included these statements: “We’re

107. getting a little bit ahead of the curve there, but I’ve thought that out, and I have an idea of how I’m going to do it that probably will—were you to get to that point, and based on the number of different individuals who would have different amounts of damages, it would probably be something like a special master that you would pay for, because the Court’s not going to take the next four months after this case to hear individual damage amounts.” In the trial court’s statement of decision, the section concerning decertification of the Nonexempt class includes a three-paragraph discussion under the subheading “Judgment on Individual Class Member Claims.” (Bolding omitted.) The first paragraph cites authority for adjudicating the claims of named plaintiffs, i.e., class representatives, despite decertification of the class. The second paragraph explains that a judgment will be entered against plaintiffs Carolyn Cortina and Kimberly Baker. The third paragraph identifies the remaining named plaintiffs and states, “Their claims for restitution of overtime pay for time employed as a non-exempt worker are to be considered by the referee, discussed below.” 19 The trial court’s further discussion of its nonconsensual reference order begins on the penultimate page of the statement of decision. The relevant text is as follows:

“The Court orders restitution of unpaid overtime wages to the [E]xempt class members earned during the class period. It will refrain from entering final judgment until an amount is determined. [¶] … Federal Rules of Civil Procedure, Rule 53, permits special masters to be appointed without the consent of the parties to resolve questions involving ‘difficult computation of damages.’ Use of same in class action [sic] was approved

19 The statement of decision contains errors in the identification of certain class representatives. Those errors are corrected in the final judgment. The named plaintiffs awarded restitution as members of the Nonexempt Class or both classes are Judith Bates, Catherine Bell, Melodie Benton, Cheryl Fuller, Laurel Johnstone, Teresa Spencer, Tina Texiera, and Mary Weidmark. For reasons not explained, judgment was also entered in favor of Nonexempt class member Martha Dominguez despite an ultimate finding that she was not entitled to restitution.

108. in Levya v. Medline Industries, Inc. (9th Cir. 2013) 716 F.3d 510, 515. [Citations.]

“Pursuant to Code of Civil Procedure section 639[, subdivision] (a)(3), due to the consumption of time that would be required to assess the amounts of individual restitution, such issue of fact is to be considered by a referee, on the basis of his review of individual testimony and any other evidence as the referee finds appropriate. The referee is to report his findings to the Court..… [¶] No party has established an economic inability to pay for the reference.… [Citations.] For these reasons, the Court orders that NATC shall bear the cost of the reference.” Defendant objected to the reference proceedings on multiple grounds, including the absence of legal authorization and the “[c]onstitutionally improper delegation of judicial power.” Defendant’s objections were overruled. The trial court appointed a retired superior court judge to serve as the referee and approved an hourly rate of $500 for his services. In addition to the referee’s fees and expenses, defendant was ordered to pay all court reporter fees and costs, and all expenses associated with obtaining the testimony of class members, e.g., airfare, lodging, and ground transportation. The reference proceedings lasted approximately three years and included live testimony from over 230 individual class members. The referee’s final report included a declaration that he “spent a total of 811 hours and charged a total of $390,864.51 in fees and $6,647.01 in costs in fulfilling the duties prescribed by the [trial court’s] appointment.” The trial court adopted all of the referee’s findings and recommendations except for those concerning four individuals (Agueda, Cortina, Kramer, and Pamela Thompson). Law and Analysis “The starting point in any discussion of the reference power is the constitutional prohibition against delegation of judicial power.” (De Guere v. Universal City Studios, Inc. (1997) 56 Cal.App.4th 482, 496.) Our state Constitution vests judicial power in the California Supreme Court, the intermediate appellate courts, and the superior courts of each county. (Cal. Const., art. VI, § 1.) “To avoid an unconstitutional delegation of

109. judicial authority, the Constitution requires the stipulation of the parties before a trial court may refer a cause to be tried by a referee.” (People v. Superior Court (Laff) (2001) 25 Cal.4th 703, 721 (Laff), citing Cal. Const., art. VI, § 21.) The California Constitution permits the Legislature to authorize the delegation of “subordinate judicial duties” to “officers such as commissioners,” which has been interpreted to include special masters and referees. (Laff, supra, 25 Cal.4th at p. 721, citing and quoting Cal. Const., art. VI, § 22.) “Masters and referees perform subordinate judicial duties only if their findings and recommendations are advisory and not binding until adopted by the court.” (Laff, at p. 721.) “If authorized by law, a superior court’s referral of a particular aspect of a judicial proceeding to a special master or referee to render advisory findings and recommendations may be made with or without the consent of the parties and is known as a special reference.” (Ibid., italics added.) The issue presented is whether the reference proceedings below were authorized by law. The trial court relied on Code of Civil Procedure section 639, which allows nonconsensual references only in five limited circumstances. (Id., subd. (a)(1)–(5); Laff, supra, 25 Cal.4th at p. 734 [“only those issues particularly described in the statute may be referred without the consent of the parties”]; Jovine v. FHP, Inc. (1998) 64 Cal.App.4th 1506, 1522 [emphasizing nonconsensual references are authorized “only in certain specified circumstances”].) “The procedure is most commonly employed where complicated accounts can more conveniently be examined or taken outside of court, and to resolve discovery disputes or certain types of family law issues.” (Jovine, at p. 1521.) Although the trial court relied on a particular provision (Code Civ. Proc., § 639, subd. (a)(3)), we list here the five permissible grounds for a special reference:

“(1) When the trial of an issue of fact requires the examination of a long account on either side; in which case the referees may be directed to hear and decide the whole issue, or report upon any specific question of fact involved therein.

110. “(2) When the taking of an account is necessary for the information of the court before judgment, or for carrying a judgment or order into effect.

“(3) When a question of fact, other than upon the pleadings, arises upon motion or otherwise, in any stage of the action.

“(4) When it is necessary for the information of the court in a special proceeding.

“(5) When the court in any pending action determines that it is necessary for the court to appoint a referee to hear and determine any and all discovery motions and disputes relevant to discovery in the action and to report findings and make a recommendation thereon.” (Code Civ. Proc., § 639, subd. (a)(1)–(5).) Subdivision (a)(1) and (2) of Code of Civil Procedure section 639 refers to financial accounting. (See In re Marriage of Galis (1983) 149 Cal.App.3d 147, 151 [holding neither provision applied because the “case did not involve an accounting”]; Dunner v. Hoover (1941) 43 Cal.App.2d 753, 759.) “The reference is allowed because a trained accountant is generally better able to efficiently and inexpensively examine a ‘long account’ than a trial court judge is able to do through adversarial court proceedings.” (De Guere v. Universal City Studios, Inc., supra, 56 Cal.App.4th at p. 499.) “To do a proper examination of a long account, a referee needs to know two things .… First, the referee must know the proper accounting methodology to be applied. Second, the referee must have the accounting records at issue.” (Ibid.) The trial court did not rely on the accounting provisions, nor were they applicable. There was no “special proceeding” (Code Civ. Proc., § 639, subd. (a)(4)), which is a term that excludes ordinary civil actions in law or equity. (See id., §§ 21, 22; Tide Water Assoc. Oil Co. v. Superior Court (1955) 43 Cal.2d 815, 822; Prahl v. Allstate Northbrook Indemnity Co. (2025) 110 Cal.App.5th 118, 125; Ruisi v. Thieriot (1997) 53 Cal.App.4th 1197, 1209–1210 (Ruisi).) Nor was the referee appointed to hear and resolve discovery disputes. (Code Civ. Proc., § 639, subd. (a)(5).)

111. In short, this was not a “right for the wrong reasons” scenario. (Estate of Beard (1999) 71 Cal.App.4th 753, 776–777 (Beard).) “The trial court has no power to refer issues other than those explicitly specified by statute.” (Ruisi, supra, 53 Cal.App.4th at p. 1209.) The legality of the nonconsensual reference thus depends on the meaning of Code of Civil Procedure section 639, subdivision (a)(3). “We begin by considering the statute’s language and structure, bearing in mind that ‘our primary goal is to determine and give effect to the underlying purpose of the law.’ ” (Gund v. County of Trinity (2020) 10 Cal.5th 503, 511.) “ ‘If the plain, commonsense meaning of a statute’s words is unambiguous, the plain meaning controls.’ [Citation.] If, however, the statute is susceptible to more than one interpretation, we ‘may consider various extrinsic aids, including the purpose of the statute, the evils to be remedied, the legislative history, public policy, and the statutory scheme encompassing the statute.’ ” (Holland v. Assessment Appeals Bd. No. 1 (2014) 58 Cal.4th 482, 490.) Again, the provision to be interpreted reads, “When a question of fact, other than upon the pleadings, arises upon motion or otherwise, in any stage of the action.” (Code Civ. Proc., § 639, subd. (a)(3).) Textually unchanged since the year 1851, this provision was recodified in 1872 from section 183 of the Practice Act. (Stats. 1851, ch. 5, § 183; Laff, supra, 25 Cal.4th at pp. 729–730; Yu v. Superior Court (2020) 56 Cal.App.5th 636, 648–649.) “The Practice Act is the precursor to the Code of Civil Procedure.” 20 (Yu, at p. 648.) Since its earliest enactment, the provision has been understood to prohibit a nonconsensual “reference of an issue arising upon the pleadings.” (Hastings v. Cunningham (1868) 35 Cal. 549, 552.) In Geeseka v. Brannan (1852) 2 Cal. 517, the amount of a plaintiff’s damages was impliedly construed as an issue arising upon the

20 Prior to 1851, a predecessor statute used sightly different language. (Stats. 1850, ch. 142, § 164; Seaman v. Mariani (1850) 1 Cal. 336, 336–337.)

112. pleadings. There, a partially completed bench trial preceded the plaintiff’s motion to have the cause referred “in order to ascertain ‘the damages sustained by the plaintiff.’ ” (Id. at p. 519.) The motion was granted, and a judgment was subsequently entered “upon the facts found by the referee.” (Ibid.) The California Supreme Court reversed the judgment because the reference was ordered “without [the] consent of both parties.” (Ibid.) “The case having been submitted to the Court, it was the duty of the Court to find upon facts adduced by the parties, and not the facts presented in a referee’s report.” (Ibid.) Under the Code of Civil Procedure, pleadings are defined as “the formal allegations by the parties of their respective claims and defenses, for the judgment of the court.” (Code Civ. Proc., § 420.) “The pleadings allowed in civil actions are complaints, demurrers, answers, and cross-complaints.” (Id., § 422.10.) A UCL lawsuit is a civil action. (Nationwide Biweekly Administration, supra, 9 Cal.5th at p. 322.) A civil complaint must contain factual allegations “constituting the cause of action,” and a “demand for judgment for the relief to which the pleader claims to be entitled.” (Code Civ. Proc., § 425.10, subd. (a)(1), (2).) “Issues arise upon the pleadings when a fact or a conclusion of law is maintained by the one party and is controverted by the other.” (Id., § 588.) Therefore, a nonconsensual reference under Code of Civil Procedure section 639, subdivision (a)(3), is limited in scope to questions of fact arising “outside the pleadings.” (In re Marriage of Galis, supra, 149 Cal.App.3d at p. 151, italics added; see, e.g., Beard, supra, 71 Cal.App.4th at pp. 761–763, 777–778 & fn. 11 [nonconsensual reference permissibly ordered regarding issues arising from litigant’s alleged disobedience of a court order issued during probate proceedings].) Factual issues raised in and by the pleadings, and thus arising from the pleadings, cannot reasonably be construed as “question[s] of fact, other than upon the pleadings, aris[ing] upon motion or otherwise, in any stage of the action.” (Code Civ. Proc., § 639, subd. (a)(3), italics added.) To conclude differently would render the qualifying language meaningless and

113. effectively remove all restrictions on the referral of factual issues without the parties’ consent. The complaint in this lawsuit alleged UCL violations based on defendant’s failure to pay the named plaintiffs and other class members “employment compensation required by the California law [cited therein].” The named plaintiffs and other class members expressly sought, in the body of the complaint, “full restitution … of all employment compensation wrongfully withheld, as necessary and according to proof, … by means of the unfair and unlawful practices complained of [t]herein.” In the prayer for relief, a request was made for restitution “of all funds unlawfully acquired by Defendants by means of any acts or practices declared by [the trial court] to be violative of the California Labor Code, [IWG wages orders], and [the UCL].” Restitution was “the relief to which” the named plaintiffs, i.e., “the pleader,” on behalf of themselves and all class members, “claim[ed] to be entitled” in the complaint. (Code Civ. Proc., § 425.10, subd. (a)(2).) It follows that the amounts of restitution to which each named plaintiff and absent class member were allegedly entitled were questions of fact “aris[ing] upon the pleadings.” (Id., § 588.) Those questions were also inherent to the pleaded allegations of “employment compensation wrongfully withheld” and “unlawfully acquired” by NATC. Such questions could not be referred to a referee without defendant’s consent because they were not “question[s] of fact, other than upon the pleadings,” and did not arise outside of the pleadings. (Id., § 639, subd. (a)(3).) Therefore, the reference proceedings were not authorized by law. Furthermore, the reference concerned questions of liability arising upon the pleadings. For both the Exempt and Nonexempt classes, determining “ ‘eligibility for recovery’ ” and the amount of recoverable restitution “depended on the same factual showing—unpaid overtime hours worked.” (Bell, supra, 115 Cal.App.4th at pp. 743– 744.) “[T]he failure to pay legally required overtime compensation falls within the UCL’s definition of an ‘unlawful … business act or practice .…’ ” (Sullivan v. Oracle

114. Corp., supra, 51 Cal.4th at p. 1206.) “But for an employer to adopt an erroneous classification policy is not unlawful in the abstract.” (Id. at p. 1208.) “[E]mployees must carry the initial burden of proving they have performed work that was not compensated in accordance with applicable labor laws.” (Duran, supra, 59 Cal.4th at p. 41; see Hernandez v. Mendoza (1988) 199 Cal.App.3d 721, 726 [“Once an employee shows that he performed work for which he was not paid, the fact of damage is certain; the only uncertainty is the amount of damage”].) The statement of decision contained no specific findings that any named plaintiff or other member of the Exempt class worked overtime for which they were not compensated. The trial court merely “order[ed] restitution of unpaid overtime wages to the [E]xempt class members earned during the class period” and tasked the referee with “assess[ing] the amounts of individual restitution.” Notably, in explaining its reasons for decertifying the Nonexempt class (many of whom were also members of the Exempt class), it observed “there was substantial evidence that some employees worked unreported overtime,” but “also a wealth of evidence that many employees did not work any overtime, worked overtime and were paid for it, or were disciplined for failing to report overtime they did work ([but were still] paid for it).” The court further noted that “[s]everal witnesses who claimed a large volume of unreported and unpaid overtime were impeached,” and later reiterated that the “individual representative testimony” was “often impeached.” With specific regard to the named plaintiffs, the trial court wrote, “Their claims for restitution of overtime pay for time employed as a non-exempt [class member] are to be considered by the referee .…” Defendant objected to this ambiguous statement by, inter alia, noting the lack of findings that any named plaintiff “worked time that she did not report during the period the named [p]laintiff was classified as nonexempt,” or that defendant “knew or had constructive knowledge” of such class members working “off-the-clock” and/or failing to report overtime. (See Brinker, supra, 53 Cal.4th at

115. p. 1051 [“liability is contingent on proof [the employer] knew or should have known off-the-clock work was occurring”].) The referee found that several class members, including a class representative for the Nonexempt class (Martha Dominguez), did not meet their evidentiary burden to show they worked uncompensated overtime hours. The trial court adopted all such findings, as well as the referee’s findings of the number of overtime hours worked by all other class members and his corresponding “restitution recommendations.” The court’s only deviations from the referee’s findings and recommendations were as to certain “conditional awards” proposed for Cortina and three people whom the court concluded were not actually “members of the [E]xempt class.” 21 The trial court’s nonconsensual reference order “was in excess of its jurisdiction and therefore voidable.” (Jovine v. FHP, Inc., supra, 64 Cal.App.4th at p. 1532; see Ruisi, supra, 53 Cal.App.4th at p. 1208 [“An invalid reference constitutes jurisdictional error which cannot be waived”].) As such, “all the proceedings founded thereon must fall.” (Hastings v. Cunningham, supra, 35 Cal. at p. 552.) The error was not only prejudicial in terms of defendant having to pay nearly $400,000 for the referee’s fees and expenses (plus substantial additional costs), but it also deprived NATC of the “right of a party to have the trier of fact observe his demeanor, and that of his adversary and other witnesses, during examination and cross-examination.” (Linsk v. Linsk (1969) 70 Cal.2d 272, 278–279.) The reference likewise infringed upon NATC’s “right ‘to a decision upon the controverted facts from the judge who heard the evidence.’ ” (Badgley v. Van Upp (1993) 20 Cal.App.4th 218, 224, quoting Linsk, at p. 279; cf. European Beverage, Inc. v. Superior Court (1996) 43 Cal.App.4th 1211, 1213 [“a party is entitled to have the same

21 The trial court also purported to reject “conditional awards of restitution” to Doran and Dobrzensky. The referee, however, did not recommend conditional awards for those people. The referee expressly made “no findings or award” with respect to Doran, and found that Dobrzensky was “entitled to zero overtime hours” and no restitution.

116. judge try all portions of a bifurcated trial that depend on weighing evidence and issues of credibility”].) The trial court impermissibly delegated to the referee the duties of hearing the individual testimony of approximately 243 witnesses, assessing their credibility, weighing the evidence, and resolving conflicts in the evidence—including inconsistencies in the testimony of various witnesses as compared to their prior depositions and/or testimony in the first phase of trial. While the referee’s findings and recommendations were deemed “advisory,” the court deferred to the referee’s credibility determinations and expressly invoked the principle that a “referee’s recommendations are entitled to great weight.” (Beard, supra, 71 Cal.App.4th at p. 777; see generally Abbott v. Mandiola (1999) 70 Cal.App.4th 676, 683 [“credibility determinations require a personal presence that a cold transcript cannot convey”].) “As fact finders, referees necessarily make credibility determinations” and weigh evidence. (In re Marriage of Petropoulos (2001) 91 Cal.App.4th 161, 176.) However, those functions are strictly limited under Code of Civil Procedure section 639, subdivision (a)(3) to questions of fact arising outside of the pleadings. In other words, to ancillary factual questions not dispositive of the ultimate issues in the case. The trial court erroneously gave weight to, and adopted, the referee’s findings on ultimate issues of NATC’s legal and financial liability. Lastly, it bears mentioning that rule 53 of the Federal Rules of Civil Procedure (rule 53) was both irrelevant and misconstrued by the trial court. “As a general rule, California courts are not bound by the federal rules of procedure but may look to them and to the federal cases interpreting them for guidance or where California precedent is lacking.” (Wershba v. Apple Computer, Inc. (2001) 91 Cal.App.4th 224, 239–240, disapproved on another ground in Hernandez v. Restoration Hardware, Inc. (2018) 4 Cal.5th 260, 269–270.) As explained, reference proceedings are exclusively governed by the California Constitution and state statutory law. Moreover, rule 53 does not authorize

117. the type of nonconsensual reference ordered in this case. “ ‘The courts have tended to read [rule 53] narrowly, closely circumscribing the range of circumstances in which reference to a master is appropriate.’ ” (Burlington Northern R. Co. v. Department of Revenue of State of Wash. (9th Cir. 1991) 934 F.2d 1064, 1072.) Rule 53 permits, in relevant part, the nonconsensual appointment of a special master to “hold trial proceedings and make or recommend findings of fact on issues to be decided without a jury if appointment is warranted by: [¶] (i) some exceptional condition; or [¶] (ii) the need to perform an accounting or resolve a difficult computation of damages[.]” (Fed. Rules Civ. Proc., rule 53(a)(1)(B)(i)–(ii), 28 U.S.C.) “Calendar congestion, complexity of the case, and great length of time required for trial are not ‘exceptional’ circumstances warranting appointment of a special master under Rule 53(a)(1)(B)(i).” (Thakur v. Cofiroute United States (C.D.Cal., Aug. 5, 2020) No. 8:19-cv-02233-ODW (JDEx), 2020 U.S. Dist. LEXIS 259215, *9 [2020 WL 10731939, *3], citing La Buy v. Howes Leather Co. (1957) 352 U.S. 249, 258–259; accord, Sierra Club v. Browner (5th Cir. 2001) 257 F.3d 444, 446–447; Burlington Northern R. Co. v. Department of Revenue of State of Wash., supra, 934 F.2d at p. 1072.) Rule 53(a)(1)(B)(ii) is analogous to the accounting provisions in Code of Civil Procedure section 639, subdivision (a)(1) and (2). “Accountings and other damages computations may be referred without the parties’ consent because they generally do not call for any peculiar judicial talent or insight.” (Beazer East, Inc. v. Mead Corp. (3rd Cir. 2005) 412 F.3d 429, 441.) However, when the issues referred are “not akin to a complicated accounting or difficult damages calculation” but are instead “foremost among the ‘basic issues’ to be tried,” a nonconsensual appointment will be viewed as “ ‘an abdication of the judicial function.’ ” (Id. at p. 442.) The Advisory Committee Notes to Rule 53 (2003 amend.) explain that a nonconsensual appointment of a special master for accounting purposes is justified under subdivision (a)(1)(B)(ii) “only as to essentially ministerial determinations that require mastery of much detailed information

118. but that do not require extensive determinations of credibility. Evaluations of witness credibility should only be assigned to a trial master when justified by an exceptional condition [under subdivision (a)(1)(B)(i)].” 22 (Italics added.) Here, the complicated mathematical aspects of the reference proceedings were not handled by the referee. The referee’s primary function, as explained in his final report, was listening to class members testify and deciding whether to accept, reject, or adjust the estimated number of overtime hours they claimed to have worked. For the Nonexempt class members, credibility determinations were of critical importance because their claims were based on disavowing the accuracy of their own time sheets. The referee needed to decide whether they had worked overtime they did not report, worked more overtime than they reported, or accurately filled out their timecards and had unreliable memories or were being less than truthful. Based on the referee’s evaluation of witness testimony and pertinent documents, “findings were made of the number of uncompensated overtime hours worked by each individual.” The difficult computational task of converting those hours to dollar amounts, accounting for multiple variables including those unique to each class member (e.g., individual salaries and changes in salary over time), was performed by plaintiffs’ retained expert, Edward Garcia. Plaintiffs superficially contend the reference proceedings were appropriate but offer no relevant authority on the issue. They argue the California Supreme Court has “endorsed innovative management practices by trial judges,” which is true. (E.g., Duran,

22 The Ninth Circuit case cited by the trial court in its statement of decision, Levya v. Medline Industries, Inc., supra, 716 F.3d 510, noted the potential applicability of rule 53 while discussing possible ways to manage the calculation of damages if the case went to trial and the defendant was found liable. The evidence at the class certification stage indicated the employer’s “computerized payroll and time-keeping database would enable the court [or a special master] to accurately calculate damages and related penalties for each claim.” (Leyva, at p. 514.) The discussion did not contemplate the need for any live testimony by class members or witness credibility determinations.

119. supra, 59 Cal.4th at p. 33 [“We have encouraged trial courts to be ‘procedurally innovative’ in managing class actions”].) The trial court’s reference order was both innovative and unprecedented, but neither made the reference proceedings lawful. “Procedural innovation must conform to the substantive rights of the parties.” (Id. at p. 40.) III. Plaintiffs’ Cross-Appeal * A significant portion of plaintiffs’ appeal is mooted by reversal of the judgment against defendant. The remaining claims concern the trial court’s entry of judgment in favor of defendant and against plaintiffs Kimberly Baker and Carolyn Cortina. For the following reasons, we conclude that aspect of the judgment must also be reversed. Baker and Cortina may therefore pursue their respective claims on retrial. Background Cortina is the original named plaintiff in this lawsuit. She conducted discovery on behalf of herself and as a putative class representative in early stages of the case, and she gave sworn deposition testimony in February 2009. Baker was added as a named plaintiff and putative class representative in the first amended complaint, filed in June 2009. Baker gave sworn deposition testimony in January 2010. Both litigants were class representatives at the time of trial. On September 3, 2015, defendant served plaintiffs’ counsel with a notice in lieu of subpoena for the trial appearances of all class representatives. (Code Civ. Proc., § 1987, subd. (b).) Trial began approximately two weeks later. The parties apparently had unreported discussions about whether Cortina and Baker would appear at trial, but neither side has ever candidly disclosed the substance of those conversations. On October 26, 2015, during a colloquy between the trial court and the attorneys, plaintiffs’ counsel generally indicated they were having scheduling difficulties with their

* See footnote, ante, page 1.

120. witnesses: “We’re finding that as we get towards the end of the month, … [it’s] a pretty impossible time for escrow officers, so prying them away from their office at this point is not easy.” As the exchange continued, one of the defense attorneys remarked, “Ms. Cortina didn’t want to come.” The comment was ignored, and the record does not indicate whether the trial court knew of any issues regarding Cortina’s ability or willingness to appear at trial. Two days later, on October 28, 2015, plaintiffs’ counsel sent an email to the defense attorneys for NATC and NAS. The message said, “Attached are the transcripts of Caroline Cortina [sic] who is out of state and unavailable. We highlighted testimony we expected both sides would want, but please let us know if you have any objections or suggestions. [¶] The best way to go may to have someone sit in the stand and read responses, and then objections can be made as we go.” (Sic.) Defense counsel was previously informed that Cortina lived in Texas. She testified about it in her deposition, and a declaration she filed in support of plaintiffs’ motion for class certification was executed in Texas. The trial judge was not the same judicial officer who granted class certification, so it is unclear whether the trial judge had the same knowledge or was aware of the parties’ recent discussions about Cortina’s availability. What the record shows is that Cortina ultimately did not appear, her deposition transcripts were not read aloud in court, and her testimony was not proffered in any form during the first phase of trial. If the parties had any pertinent discussions about Cortina after the October 2015 email, there is no evidence of it in the appellate record. The parties’ unreported communications about Kimberly Baker occurred close in time to those regarding Cortina. This is evident from the trial court’s Register of Actions, which denotes a filing on November 3, 2015, of “Defendants’ Counter Designations of Deposition Transcript of Kimberly Baker.” The filed document, dated October 30, 2015, attached transcript excerpts that were also later included with the parties’ moving and

121. opposition papers for Baker’s motion for new trial. The “Counter Designations” state, in relevant part:

“Defendants [NATC] and [NAS] provide the following counter-designations to Plaintiffs’ designations of the deposition transcript of Kimberly Baker taken on January 7, 2010. Attached hereto are the pages and exhibits designated by Defendants, as well as those pages designated by Plaintiffs, with the relevant designated portions by both parties highlighted.” (Italics added.) “Counter Designations” would seem to indicate a separate filing by plaintiffs designating the portions of Baker’s deposition transcript they wished to have admitted in lieu of live testimony. However, no such filing was made. Plaintiffs have never directly explained this circumstance. On the same day defendant filed the counter-designations, plaintiffs introduced into evidence a 43-page document marked as “Exhibit 1372.” Plaintiffs’ counsel accurately described the exhibit: “This is a chart presented by … defendants in response to an … interrogatory. It has a formal title, if you will, of Interrogatory [Nos.] 422 to 424, Class Members Employed in Class Positions Between 4/17/2003 and 12/31/2012 .… [¶] … [¶] It is a chart which shows all of the class members and shows the dates of their employment, the office they worked in and their job title and the job title number and showing the date that they held those jobs.” Exhibit 1372 is significant because, with the exception of defendant’s counter-designations regarding Kimberly Baker, it essentially constituted the only evidence in the first phase of trial showing the class positions held by Baker and Cortina during the class period. Defendant’s records showed Cortina was employed as a Unit Manager from July 2004 to August 2005 and held no other class positions. In other words, the evidence showed Cortina was exclusively a member and representative of the Exempt class.

122. Defendant’s records showed Baker was employed as a nonexempt “Escrow Officer, Jr.” (also known as a Junior Escrow Officer) between May 2005 and November 2005. No other positions or dates were listed, thus indicating Baker’s exclusive membership in, and representation of, the Nonexempt class. Baker’s deposition transcripts contained testimony that she also worked as an escrow officer in late 2004 and/or early 2005 before being demoted to the junior position, but NATC classified both jobs as nonexempt after 2003. Baker began her employment with NATC in March 2004 as an Escrow Assistant, which was both a nonexempt position and one expressly excluded from both class definitions. On March 11, 2016, the parties filed a joint stipulation agreeing that the presentation of evidence for the initial phase of trial had concluded. The sparsely worded document included the following statement: “[T]he parties have submitted deposition designations in lieu of live testimony and agree and request that the transcripts submitted by each of the parties be admitted into evidence as to those portions so designated .…” Although plaintiffs claim the joint stipulation shows a mutual agreement to designate transcripts from the depositions of Baker and Cortina, it is only supportive of their claims as to Baker. Contrary to the assertions in plaintiffs’ briefs, the record contains no evidence of an agreement to have Cortina’s deposition transcripts admitted in lieu of trial testimony. The October 2015 email discussed above shows a proposal to have portions of Cortina’s deposition transcripts read into the trial record, which never occurred. 23

23 The plaintiffs’ attorney who sent the email dated October 28, 2015, was not in court as often as plaintiffs’ other trial attorneys. The trial court had previously made clear, during a pretrial conference on September 4, 2015, that it did not want to have the deposition testimony of absent witnesses read into the record and preferred written transcripts even over video recordings of the depositions. On September 14, 2015, the court reiterated its preference not to have counsel “play attorney and witness and actually read [from the transcripts] like that.” These facts stand in contrast to plaintiffs’ current position that the email somehow memorialized an agreement regarding the admission of

123. In August 2016, the trial court orally announced a tentative statement of decision. The tentative rulings as to the Exempt class were discussed first. The court stated, in relevant part, “The Court in this matter orders restitution of unpaid overtime wages to the exempt class members during the class period. The amounts of individual restitution are to be recommended by a referee to be appointed by the Court ….” Next, the trial court discussed the claims of the Nonexempt class and explained its intention to decertify the class. The tentative statement of decision concluded as follows:

“[Case law] allows a trial court judge to enter judgment for named plaintiffs after the decertification of the class. Several of the named plaintiffs testified, one of whom, Cortina, did not testify. Judgment against Cortina will be entered on her individual claim.

“Judgment entered for the other named plaintiffs are Lori Baker, Judith Bates, Catherine Bell, Melody Bates, Martha Dominguez, Cheryl Fuller, Laurel Johnstone, Teresa Spencer, Tina Texeira and Mary Weidmark. [Sic.] Their claims for restitution of overtime pay remain an issue, however, to be considered by the referee.” The above-quoted statements contain several mistakes. First, Cortina was not part of the Nonexempt class and the trial court had no basis to believe or conclude otherwise. Second, although Lori Baker was a class member, she was not a named plaintiff. Kimberly Baker was a named plaintiff. Third, there was no person named “Melody Bates” in either of the classes. The name missing from the court’s listing of class representatives, which it otherwise provided in alphabetical order by last name, was Melodie Benton. We further observe that since Kimberly Baker was not mentioned during the oral tentative ruling, and considering the court was purporting to identify “named plaintiffs” but referred to a nonnamed plaintiff with the same last name, the

Cortina’s deposition transcripts into evidence. The message reads as an initial proposal, which the defense and plaintiffs’ other lawyers would have likely viewed as ill-informed, and it furnishes no insight into any subsequent discussions of the matter.

124. circumstances strongly indicate the court’s original intention to rule in favor of Kimberly Baker and have her to participate in the next phase of trial. The trial court further ordered “[d]efendants to draft the Court’s tentative statement of decision.” When defendants did so, they included prefatory explanations in the “Notice of Proposed Statement of Decision,” calling attention to some of the errors described in the previous paragraph. As relevant here, the notice read:

“[T]he Court stated it would enter judgment against named plaintiff Carolyn Cortina on her individual claims because she did not testify at trial, and that it would enter judgment in favor of the remaining named plaintiffs. Named plaintiff Kimberly Baker, however, also did not testify at trial. A different class member, Lori Baker, did testify. But Lori Baker is not a named plaintiff. Defendants drafted the [Proposed] Statement of Decision assuming that, consistent with the Court’s decision regarding named plaintiff Carolyn Cortina, the Court will enter judgment against named plaintiff Kimberly Baker for the same reason.” Plaintiffs filed objections to the proposed statement of decision, but those objections did not address the rulings against Kimberly Baker and Cortina. The trial court later revised its tentative statement of decision in various ways and adopted defendants’ proposed revision concerning Kimberly Baker. Inexplicably, it refrained from correcting the erroneous references to Lori Baker and “Melodie Bates.” The final statement of decision separately addresses the claims of the two classes under the headings, “DECERTIFICATION OF THE NON-EXEMPT CLASS” and “EXEMPT CLASS,” in that order. (Bolding omitted.) The section dedicated to the Nonexempt class states, in relevant part: “There are several named plaintiffs, two of whom (Carolyn Cortina and Kimberly Baker) did not testify. Judgment against Ms. Cortina and Ms. Baker is to be entered on their individual claims.… [¶] The other named plaintiffs are: Lori Baker, Judith Bates, Catherine Bell, Melodie Bates .…” Whereas the oral tentative ruling stated that a judgment would be entered for the other named plaintiffs, the final statement of decision merely reads, “Their claims for

125. restitution of overtime pay for time employed as a non-exempt worker are to be considered by the referee .…” The section dedicated to the Exempt class “orders restitution of unpaid overtime wages to the exempt class members earned during the class period” without any mention of Cortina. “[D]ue to the consumption of time that would be required to assess the amounts of individual restitution, such issue[s] [were] … to be considered by a referee .…” During the reference proceedings, the parties disagreed about whether Cortina and Kimberly Baker were allowed to testify. The referee agreed to hear testimony from Cortina but not Baker, apparently basing the ruling on Cortina’s membership in the Exempt class. In May 2019, plaintiffs filed a motion with the trial court “for an order to clarify the scope of the reference” as to certain individuals, including Baker. Plaintiffs did not seek clarification as to Cortina. Plaintiffs’ motion for clarification argued that the trial court’s ruling as to Kimberly Baker was unclear because it ignored or overlooked the filed designations of her deposition testimony. The motion was supported, in pertinent part, by evidence that Baker resided in Arizona at the time of trial. The trial court issued a written ruling denying the motion as to Baker for the following reasons:

“The clarification sought … as to Ms. Baker is not really a request for clarification, as the Statement of Decision clearly states that judgment is to be entered against her, since she did not appear at trial. The proper procedure for raising a claim that a judgment or part thereof is error is a motion for new trial or to vacate a judgment, under Code of Civil Procedure sections 657 or 663. [Citation.] As no final judgment has yet been entered, such a request is premature.” In the referee’s final report, he recommended that Cortina be awarded $50,418 in restitution. The trial court summarily rejected the recommendation. On August 31, 2022, a final judgment was entered “[i]n favor of all defendants and against Class Representatives Carolyn Cortina [and] Kimberly Baker.”

126. On September 8, 2022, the trial court signed a stipulated order establishing the briefing schedule for all motions for new trial. The deadlines notably required all opposition papers to be filed by October 21, 2022, and all reply papers to be filed by October 28, 2022. The order further stated, “The hearing on all Motions for New Trial shall be set for November 4, 2022 .… [¶] Based on the August 31, 2022 date of the Clerk’s mailing of a copy of the judgment …, the last day on which the Court may render rulings on said Motions For New Trial, is November 14, 2022.” On September 9, 2022, i.e., the day after the stipulated order was issued, defendant petitioned this court for a writ of mandate regarding issues unrelated to those in the present appeal. Meanwhile, Cortina and Baker proceeded to file timely motions for a new trial. Defendant filed timely oppositions to both motions. On October 21, 2022, i.e., the day defendant’s opposition papers were filed, this court issued an order to show cause on defendant’s writ petition. Among other directives, the underlying action was ordered “stayed pending determination of the writ proceeding … or until further ruling of this court.” Due in part to the California Supreme Court’s review of the matter in North American Title Co. v. Superior Court, supra, 17 Cal.5th 155, the stay remained in effect until July 2025. The parties agree that the motions for new trial were denied by operation of law. (Code Civ. Proc., § 660, subd. (c).) Law and Analysis Plaintiffs’ claims are reviewed under the ordinary standards for a direct appeal of a judgment in a civil action. “[T]he merits of a motion for a new trial denied by operation of law may be reviewed upon appeal in the same manner as if expressly denied by the court.” (Estate of Shepard (1963) 221 Cal.App.2d 70, 73.) We are nevertheless mindful of the reality that the trial court was prevented from ruling on the motions for new trial by our issuance of a stay. Carolyn Cortina

127. The only explanation given for entry of judgment against Carolyn Cortina was her failure to testify during the initial phase of trial. In her motion for a new trial, Cortina challenged this ruling on grounds of prejudicial irregularity in the court proceedings, insufficient evidence, and/or legal error. (Code Civ. Proc., § 657, subds. 1, 6, 7.) As in other contexts, questions of legal error are reviewed de novo. (Smith v. Magic Mountain LLC (2024) 106 Cal.App.5th 1128, 1135; see Hoffman-Haag v. Transamerica Ins. Co. (1991) 1 Cal.App.4th 10, 15–16.) Before addressing issues of legal error, we will discuss an erroneous factual premise of the trial court’s ruling. The court obviously perceived a material connection between its decertification of the Nonexempt class and Cortina’s failure to testify during the initial phase of trial. However, the evidence showed Cortina represented the Exempt class. This fact was never in dispute. In a prior declaration filed in support of class certification, Cortina declared she was hired in 2002 as an Escrow Assistant and held that position until she was promoted directly to Unit Manager in July 2004. Attached to the declaration was an employment record (“Status Change Form”) confirming the new job title of “Escrow Unit Manager.” Plaintiffs’ motion for class certification also included responses by defendant to written discovery propounded by Cortina. In those responses, defendant repeatedly asserted variations of the following statement: “Plaintiff was hired by Defendant and began working as an Escrow Assistant on July 9, 2002. Plaintiff worked as an Escrow Assistant until she was promoted to Escrow Unit Manager, effective July 1, 2004. Plaintiff remained employed as an Escrow Unit Manager until the effective date of her resignation, August 1, 2005.” As earlier noted, the definitions of the Exempt and Nonexempt classes both exclude the position of Escrow Assistant. Deposition transcripts further show that despite personally considering the terms “escrow officer” and Unit Manager to be interchangeable, Cortina acknowledged her official job title and the fact defendant classified her as an exempt employee when she

128. was promoted in 2004. Cortina testified to the same essential facts when she appeared before the referee in 2019. 24 Having established that decertification of the Nonexempt class was irrelevant to Cortina’s claims, we consider the legal consequences of a class representative’s failure to testify at trial. There is no specific requirement that all class representatives testify in a class action trial or at every stage of a multiphase trial. (See, e.g., Espejo v. The Copley Press, Inc. (2017) 13 Cal.App.5th 329, 354–355 [rejecting claims of error asserted because “five of the six class representatives did not testify at trial”].) Once again, the key principle is common proof. Where common proof exists to prove an issue in the case, individualized proof is not necessarily required. (E.g., Spence v. American Airlines, Inc. (N.D.Tex., May 22, 2024, No. 4:23-cv-00552-O) 2024 U.S. Dist. LEXIS 139810, *27 [2024 WL 3635309, *9] [“Indeed, the liability and causation elements under Plaintiff’s ERISA claims are all subject to class-wide common proof using Defendants’ documents and record. To win at trial, no testimony or evidence is needed [from] any member of the class since the entire focus is on Defendants’ course of conduct pertaining to the entire Plan”].) The Duran opinion suggests that phases of trial devoted to statistical sampling may very well not include testimony from all class representatives if the sample group or groups of witnesses are properly randomized. (Duran, supra, 59 Cal.4th at p. 45.) Here, the initial phase of trial was supposed to determine all issues of liability based on representative testimony, but it was prejudicially affected by erroneous methodologies and other consequential mistakes. Moreover, the so-called liability phase resulted only in classwide findings on defendant’s exemption defenses. The trial court did not actually address or make findings on plaintiffs’ threshold burden of establishing they and the

24 In the appellate briefs, plaintiffs allege for the first time that Cortina was a member of both classes. This perplexing contention is made without any supporting record citations. Regardless, the contention does not affect the outcome of the appeal.

129. classes they represented “performed work that was not compensated in accordance with applicable labor laws.” (Id. at p. 41.) In light of the fundamentally flawed and procedurally defective trial plan, there is no justification for the challenged ruling against Cortina. Although plaintiffs bear significant responsibility for the procedural defects, bifurcation was ordered by the trial court sua sponte. Were we to assume the court impliedly found classwide proof of uncompensated overtime as to the Exempt class in the first phase of trial, denying Cortina recovery for not personally appearing was erroneous because her individual testimony would not have been necessary to establish the threshold element of liability. Alternatively, based on what actually happened (delegation of the threshold overtime determinations to the referee), the disparate treatment of Cortina was erroneous because (1) class representatives are not necessarily required to testify in all phases of trial, (2) Cortina testified in the second phase, and (3) hundreds of other Exempt class members were awarded restitution without testifying in the initial phase. Defendant, purporting to rely on Code of Civil Procedure section 581, subdivision (d), argues the trial court appropriately found Cortina abandoned her claims. The cited provision states: “Except as otherwise provided in subdivision (e), the court shall dismiss the complaint, or any cause of action asserted in it, in its entirety or as to any defendant, with prejudice, when upon the trial and before the final submission of the case, the plaintiff abandons it.” The abandonment argument is not persuasive. Generally, “a finding of abandonment must be predicated on a clear, unequivocal, and express intent to abandon an action” or “conduct on the part of the plaintiff that prevents the successful prosecution of the action.” (1A Cal.Jur.3d (2022) Actions, § 389, p. 523.) The statutory provision relied upon by defendant is understood to concern dismissals based on a plaintiff’s clearly shown intention to no longer pursue his or her claims. (See D & J, Inc. v. Ferro Corp. (1986) 176 Cal.App.3d 1191, 1195; Kaufman & Broad Bldg. Co. v. City & Suburban

130. Mortg. Co. (1970) 10 Cal.App.3d 206, 212–213.) A plaintiff’s mere failure to appear at trial is addressed in other parts of the statute that authorize a dismissal only “without prejudice.” (Code Civ. Proc., § 581, subd. (b)(3), (5).) Defendant also cites the distinguishable case of Shaw v. County of Santa Cruz (2008) 170 Cal.App.4th 229. The plaintiffs in Shaw appeared in court and litigated only some of their numerous claims against the defendants. (Id. at pp. 236–237.) Despite being notified of the trial court’s intention to dismiss the unprosecuted claims, the Shaw plaintiffs “did not thereafter bring to the court’s attention that they wished a trial on any of their remaining claims. Nor did they make any posttrial motions, including any request that the court vacate the dismissal .…” (Id. at p. 237.) Consequently, the dismissal was affirmed on appeal. (Id. at p. 285.) The other cases cited in defendant’s briefing are also distinguishable. (E.g., Stewart v. Stewart (1909) 156 Cal. 651, 656 [affirming dismissal where plaintiff and her counsel appeared for trial but “refus[ed] to offer any testimony in support of her complaint”].) The abandonment theory is further refuted by the absence of a dismissal order. A finding of abandonment results in the dismissal of a plaintiff’s entire complaint or particular causes of action therein, either with or without prejudice. (Code Civ. Proc., § 581.) The trial court did not dismiss the complaint or Cortina’s UCL claim; it entered a judgment against Cortina and in favor of defendant. Therefore, the record does not support an implied finding of abandonment. Even assuming such a finding was made, entry of judgment on that ground was improper for all the reasons discussed. Kimberly Baker Defendant contends Kimberly Baker was also found to have abandoned her claims. For the reasons stated above, we conclude no such finding was made and reject the abandonment theory. The propriety of the trial court’s ruling against Baker is, however, a closer issue.

131. As with Cortina, the only stated reason for entry of judgment against Baker was because she “did not testify” during the initial phase of trial. Baker failed to object to the statement of decision in accordance with Code of Civil Procedure section 632, which implicates the doctrine of implied findings (see id., § 634). However, the trial court responded to Baker’s subsequent motion for clarification by indicating it considered the ruling to be sufficiently clear: a judgment would be entered against her because “she did not appear at trial.” This does not override the doctrine of implied findings, but it may be considered as part of our overall analysis. A plaintiff’s absence from trial is not inherently fatal to the plaintiff’s causes of action. Not testifying at trial could, of course, have the practical consequence of not meeting applicable evidentiary burdens. But the Code of Civil Procedure provides for alternative methods of proof. Subject to certain requirements, “[a]ny party may use for any purpose the deposition of any person or organization, including that of any party to the action,” in lieu of live testimony. (Code Civ. Proc., § 2025.620, subd. (c).) Large portions of Baker’s deposition testimony were designated for use in lieu of in-person testimony during the initial phase of trial. The deposition transcript excerpts were filed with the trial court and accepted without comment or dispute. Although the trial court was clearly informed of which parts were relied upon by the respective parties, defendant contends the court was constrained to consider only the portions designated by the defense. The reason being that plaintiffs’ counsel did not file a separate designation. Defendant’s argument is not persuasive as a stand-alone proposition or as applied to the circumstances of this case. Presuming the general admissibility of the deposition transcripts, Code of Civil Procedure section 2025.620 (section 2025.620) does not specifically impose any restrictions on which parts the court may review and consider. The more pertinent issue, which is not addressed by either side, is the foundational requirements of section 2025.620. An adverse party may use the deposition of another party to the action “for any purpose,” regardless of whether “the deponent is available to

132. testify, has testified, or will testify at the trial.” (§ 2025.620, subd. (b).) But when a party seeks to rely on their own deposition testimony, admissibility is contingent upon certain findings by the court. The first of several qualifying grounds is that “[t]he deponent resides more than 150 miles from the place of the trial.” (Id., subd. (c)(1).) Alternatively, the court may find “[e]xceptional circumstances exist that make it desirable to allow the use of any deposition in the interests of justice and with due regard to the importance of presenting the testimony of witnesses orally in open court.” (Id., subd. (c)(3).) The distance provision is rarely an obstacle in cases where it may legitimately apply. “Unavailability need not be shown. Hearsay can be used to provide the foundation to establish that a deponent resides 150 miles from the courthouse.” (Monroy v. City of Los Angeles (2008) 164 Cal.App.4th 248, 264.) In Monroy, a trial court was found to have abused its discretion by refusing to admit proffered deposition testimony when “[t]here were no facts to suggest that the information [regarding the deponent’s place of residence] was unreliable.” (Id. at p. 265.) Defendant does not dispute that Baker resided outside of California at the time of trial. Plaintiffs’ motion for clarification, filed during the reference proceedings and prior to entry of judgment, was supported by evidence of her residency in Arizona. During the prior phase of trial, however, there was no evidence before the trial court regarding where she lived. In theory, the court’s ruling against Baker could have initially been based on a technical failure to comply with section 2025.620, subdivision (c)(1). To be clear, defendant does not argue for affirmance based on the distance provision of section 2025.620. We note the issue because “[o]rdinarily, when the court gives an incorrect legal reason for its ruling, we look for any other correct legal basis on which to sustain the order.” (Kemp Bros. Construction, Inc. v. Titan Electric Corp. (2007) 146 Cal.App.4th 1474, 1477.) But here it is obvious that the only reason the trial court ruled against Baker was her physical absence from the courtroom. We do not reach the conclusion lightly, given the presumption that a court’s duties are regularly performed

133. (see Evid. Code, § 664), but it is the only conclusion reasonably inferable from the record. “When the record clearly demonstrates what the trial court did, we will not presume it did something different.” (Lafayette Morehouse, Inc. v. Chronicle Publishing Co. (1995) 39 Cal.App.4th 1379, 1384.) The trial court was very open to, and accommodating of, the use of deposition transcripts in lieu of live testimony if the parties stipulated to it. On December 3, 2015, during a discussion about designating transcripts, the court itself noted, “I think we talked about we weren’t going to really be critical on the distance,” i.e., the 150-mile requirement in section 2025.620. The parties made no offers of proof, nor did the court make any inquiries or findings, as to the residency of the numerous witnesses whose deposition testimony was admitted in lieu of courtroom testimony. To recap the chronology of events, the trial court tentatively found in favor of Kimberly Baker but mistakenly referred to her as Lori Baker. When defense counsel drafted the written statement of decision, they proposed entry of judgment against Kimberly Baker because she did not testify at trial. Plaintiffs’ counsel did not object, and the court made the proposed revision. The court did not change its mind because it suddenly realized there was no evidence of where Baker lived or went back and reassessed her deposition testimony. When Baker moved for clarification of the ruling, the court responded that the basis was already clear: “she did not appear at trial.” The court declined to reconsider its ruling at that time, stating Baker’s arguments should be raised in a motion for new trial. She filed a motion for new trial, but the stay issued by this court prevented the motion from being heard and ruled upon. Defendant argues the judgment against Kimberly Baker must be affirmed because there is substantial evidence to support it. The argument is partially based on the abandonment theory, and partially on her deposition transcript. More specifically, it is based on the parts of the transcript defendant designated for the trial court’s consideration. Defendant’s position would be more persuasive if the evidence

134. conclusively disproved Baker’s claims, but it leaves open questions that include the truth or falsity of her testimony about working as an escrow officer in late 2004 and early 2005, and whether she worked uncompensated overtime during that period. The issues presented by Baker’s appeal are not strictly limited to whether the record would support a judgment against her on the merits. Defendant wants us to ignore the grave procedural errors upon which its own appeal is based. Doing so is neither required nor warranted. As to the Nonexempt class, the initial phase of trial focused on whether NATC had “a company-wide policy of pressuring class members to work unreported overtime as well as through meal periods .…” The quoted language comes directly from the statement of decision. The trial court found a lack of common proof in that regard. The decision notes there was substantial evidence of uncompensated overtime, but “also a wealth of evidence that many employees did not work overtime [or] worked overtime and were paid for it .…” Accordingly, the court decertified the class but delegated to a referee the task of determining whether, and to what extent, the named plaintiffs in the Nonexempt class were entitled to restitution. The delegation was not quite so explicitly worded, but the lack of findings on the threshold issue of liability is clear from the record. The most glaring example involved Martha Dominguez. Four named plaintiffs were exclusively members of the Nonexempt class: Kimberly Baker, Catherine Bell, Martha Dominguez, and Tina Texeira. Dominguez’s testimony on direct examination at trial contained general allegations of working unreported and uncompensated overtime. On cross-examination, Dominguez had trouble providing details to substantiate her claims. Defendant asserted objections to the statement of decision based on insufficient evidence “that [Dominguez] worked off-the-clock in any workweek,” noting “she denied any specific knowledge of a workweek where she failed to correctly report her time.” The trial court disregarded the

135. objections and referred Dominguez’s claim to the referee along with those of the other named plaintiffs (but not Kimberly Baker). Dominguez testified again in the reference proceedings. It did not go well for her. Based in part on Dominguez’s prior deposition and trial testimony, the referee concluded “the evidence simply doesn’t support an award of overtime” and “there is no basis for an award of overtime.” The trial court adopted the referee’s findings and recommendations. The referee likewise made credibility determinations and weighed conflicting evidence in deciding the claims of Texeira. Defendant argued Texeira’s time sheets showed that she was paid for all claimed overtime. The referee concluded Texeira’s testimony was “more reliable” and also found in her favor on the issue of whether defendant knew or should have known she was working unreported overtime. The trial court adopted the referee’s findings and recommendations. To summarize, the trial court did not actually determine the threshold liability issues prior to the (unauthorized) reference proceedings. It deferred making those findings until after the second phase of trial, and it placed “ ‘great weight’ ” on the referee’s recommendations. As such, there was no justification for barring Kimberly Baker from participating in the reference proceedings and ruling against her simply because she did not appear in court during the initial phase of trial. “As class actions are originally creatures of equity [citation], so the rules for administering them must be equitable.” (Fireside Bank v. Superior Court (2007) 40 Cal.4th 1069, 1084.) By the same token, “any remedy to cure procedural missteps in handling a class action must itself be equitable.” (Id. at p. 1087.) The challenged ruling was made in conjunction with a decertification order issued in the midst of a procedurally defective class action trial. Despite plaintiffs’ own procedural missteps, the interests of justice call for reversal of the judgment against Kimberly Baker and allowing her claims to be decided on the merits in the retrial.

136. DISPOSITION The judgment is affirmed in part and reversed in part. It is affirmed as to defendant North American Services, LLC, against all plaintiffs, including Janet Doran. It is affirmed as to defendant North American Title Company against plaintiff Janet Doran. In all other respects, the judgment is reversed. The case is ordered decertified as a class action and is remanded for retrial of the named plaintiffs’ individual claims. The trial court may entertain a new class certification motion as to the decertified “Exempt” class. North American Title Company is awarded the costs associated with its appeal. Each party shall bear their own costs for the cross-appeal.

PEÑA, J. WE CONCUR:

LEVY, Acting P. J.

DESANTOS, J.

137. Appendix A