9th Circuit Upholds Statistical Sampling to Determine Liability in Off the Clock Overtime Class Action

Allstate re-classified its adjusters to be non-exempt some years back.  Rather than require employees to keep their work time on time sheets or use a time clock, the employees were paid a standard eight hours per day / 40 hours per week.  However,

the manager of each local office has the power to file a timekeeping “exception” or “deviation” from the default expectation of 8 hours per day and 40 hours per week. This adjustment takes place when a claims adjuster’s request for overtime or early leave is approved. Managers do not adjust time cards based on either their own observations of work habits or on the technological records contained in computer and telephone systems. Each local office has a nonnegotiable compensation budget, which creates a functional limit on the amount of overtime a manager may approve.

Auto-punching, overtime pay only upon request, and a budget restricting overtime... What could go wrong?  

Right.  Jiminez, an adjuster, filed a class action. He claimed Allstate had an "unofficial policy" of discouraging employees from reporting overtime.  As a result, he and the class  members worked "off-the-clock" overtime for which they were not compensated.

Of note, the panel approved a district court's formulation of the elements of an off the clock work claim as follows:

Under California law, there are three elements of an off-the-clock claim of the type raised by the class here: “[A] plaintiff may establish liability for an off-the-clock claim by proving that (1) he performed work for which he did not receive compensation; (2) that defendants knew or should have known that plaintiff did so; but that (3) the defendants stood idly by.” Adoma v. Univ. of Phoenix, Inc., 270 F.R.D. 543, 548 (E.D. Cal. 2010) (internal quotation marks omitted).

Unfortunately, the court did not also cite Jong v. Kaiser Found. Hospital, a California decision (prior post here).

Anyway, the Court of Appeals here agreed with the district court that the class action should be certified.  The district court found these common questions predominated over individual ones:

(i) whether class members generally worked  overtime without receiving compensation as a result of Defendant’s unofficial policy of discouraging reporting of such overtime, Defendant’s failure to reduce class members’ workload after the reclassification, and Defendant’s policy of treating their pay as salaries for which overtime was an “exception”; (ii) whether Defendant knew or should have known that class members did so; and (iii) whether Defendant stood idly by without compensating class members for such overtime.

The Court of Appeals decided that these common questions would resolve the "common issue" of whether Allstate could be liable for off-the-clock work.  You may ask how a class can prove that its employees worked under the "unofficial" policy or the "official" policy requiring payment for all overtime?  

With statistics, that's how.  The Ninth Circuit panel held that the statistical models proposed by the plaintiff, and approved by the district court, could be used to prove liability:

the district court carefully analyzed the specific statistical methods proposed by plaintiffs. It struck some of the expert testimony offered by plaintiffs as insufficiently empirically supported and took pains to ensure that the statistical analysis it did accept conformed to the legal questions to which the analysis was being applied. Unlike the putative class in Comcast, 133 S.Ct. at 1434, which relied on
statistical analysis that was not closely tied to the relevant legal questions, or in Duran, 325 P.3d at 940, which used a sample of 20 names drawn from a hat without evidence showing that the number of names chosen or the method of selection would produce a result that could be “fairly extrapolated to the entire class,” the district court has accepted a form of statistical analysis that is capable of leading to a fair determination of Allstate’s liability, and preserved the rights of Allstate to present its damages defenses on an individual basis.

Allstate argued that the "unofficial policy" did not exist and that it had strong policies against off-the-clock work. But the court held that this argument was properly made at trial rather than certification:

Allstate argues that its formal policies which call for employees to be  paid for all overtime worked are lawful, and that the alleged informal “policy-to-violate-the-policy” does not exist. This argument is appropriately made at trial or at the summary judgment stage, as it goes to the merits of the plaintiffs’ claim. See In re Whirlpool Corp. Front-Loading Washer Products Liab. Litig. , 722 F.3d 838, 857 (6th Cir. 2013) (noting that if a defendant has a strong argument against classwide liability, it “should welcome class certification” as that allows it the opportunity to resolve claims of all class members at once). Whether any of these common questions are ultimately resolved in favor of either side is immaterial at this class certification stage, where we determine whether any answer that the questions could produce will drive resolution of the class’ claims.

So, take-aways: 

- "auto-clocking" is not a good practice if you want to avoid off-the-clock class actions.  

-  courts are continuing to certify now, ask about liability later.  

- statistical sampling can be used to determine liability without violating due process, at least for now. The U.S. Supreme Court has yet to rule on this issue.

This case is Jimenez v. Allstate Ins. Corporation and the opinion is here.

Court of Appeal: Employer Not Responsible for Off-the-Clock Work

Kaiser won summary judgment against an employee who claimed he was working off the clock, contrary to Kaiser policy and procedures for reporting time worked.

To recover on a claim for unpaid work time, the employee must prove that the employer knew or should  have known (via the exercise of reasonable care) that the employee was working off the clock. That is called "actual" or "constructive" knowledge.  This court distinguished between "should" have known and "could" have known.  That is, it is not enough for the employee to argue that the employer somewhere had enough data to infer that an employee was working off the clock. Quoting analogous federal authorities interpreting the federal Fair Labor Standards Act, the Court of Appeal explained:

In Forrester v. Roth’s I.G.A. Foodliner, Inc. (9th Cir. 1981) 646 F.2d 413, the court held that “where an employer has no knowledge that an employee is engaging in overtime work and that employee fails to notify the employer or deliberately prevents the employer from acquiring knowledge of the overtime work, the employer’s failure to pay for the overtime hours is not a violation of § 207.” (646 F.2d at p. 414.) “An employer must have an opportunity to comply with the provisions of the FLSA. This is not to say that an employer may escape responsibility by negligently maintaining records required by the FLSA, or by deliberately turning its back on a situation. However, where the acts of an employee prevent an employer from acquiring knowledge, here of alleged uncompensated overtime hours, the employer cannot be said to have suffered or permitted the employee to work in violation of § 207(a).” (Id. at pp. 414-415; see also, e.g., Kellar v. Summit Seating, Inc. (7th Cir. 2011) 664 F.3d 169, 176-178; cf. Brown v. Scriptpro, LLC (10th Cir. 2012) 700 F.3d 1222, 1230-1231.) This qualification was implicitly endorsed by our Supreme Court in a different context (Morillion v. Royal Packing Co. (2000) 22 Cal.4th 575, 585 [“ ‘ “[T]he words ‘suffer’ and ‘permit’ as used in the statute mean ‘with the knowledge of the employer.’ ” ’ ”]

Here is what the plaintiff-employee testified to:

Jong acknowledged that he “knew of Kaiser’s written policy that OPMs should be clocked in whenever they were working,” that he was always paid for time he recorded on Kaiser’s recording system, including overtime hours, that he was instructed he was eligible to work and be paid for overtime hours, that there was never an occasion when he requested approval to work overtime that was denied and there were occasions when he worked and was paid overtime even though he did not seek pre-approval, that he was not told by any of his managers or supervisors or any other Kaiser management personnel that he should perform work before he clocked in or after he clocked out or otherwise work off-the-clock,1 and that he signed the attestation form and understood it was an attestation that he would not work off-the clock.

The plaintiff tried to get around that testimony by arguing that Kaiser knew he was working off the clock - or should have known - because other employees dong the same job testified that they were working off the clock; sort of "me-too" wage hour evidence.

Agreeing with the trial court, the Court of Appeal held that "me too" evidence was irrelevant:

Under these admitted circumstances, evidence that Kaiser was aware that many OPMs worked more than 40 hours a week before being reclassified would not support a finding that after the reclassification Kaiser knew or should have known that Jong was not correctly reporting his hours.

The Court of Appeal also rejected the plaintiff's argument that Kaiser "should have known" the plaintiff was working off the clock because the "alarm code" records showed he was reporting to work before his shift started, and activated the alarm code long after he punched out.

We again find the trial court’s reason for rejecting the sufficiency of this evidence to be persuasive: “[E]ven assuming that the availability of such data could otherwise meet the ‘should have known’ requirement, as opposed to ‘could have known’ [citation], the alarm data does not show what Jong was doing during the time between disarming the alarm and clocking in, or between checking out and arming the alarm.” While the summary judgment papers may have contained evidence that Jong was working whenever the alarm was off, that information was not before Kaiser when paying Jong and Kaiser could reasonably believe that he did not begin or end work except as he reported.

So, although this case involves an individual employee's summary judgment loss, this case is going to help in class action cases involving off the clock work.  The court will not accept the premise that other employees' testimony that a job took them more than 40 hours to perform proves that it takes more than 40 hours to perform ALL class members' job duties. 

There are good take-aways for employers prevention efforts here, as well.  It is key to have a written policy prohibiting off-the-clock work.  Employers also should have employees acknowledge, in writing, and perhaps under penalty of perjury, that they accurately report hours worked.  It is helpful, too, to have proof of paying overtime when work is reported. 

This case is Jong v. Kaiser Foundation Health Plan, Inc. and the opinion is here. 

Court of Appeal Reverses Order Decertifying Class

Allstate employs auto insurance field adjusters. They track work  time via a computerized system. The system "assumes" that the arrival at the first job site for the day is the beginning of the work day. Therefore, there is the potential that adjusters performing work for the company before the arrival is "work off the clock."

Among the overtime tasks those adjusters declared they performed outside their eight-hour shifts were (1) logging onto their work computers, (2) downloading their assignments, (3) making courtesy calls to auto repair shops and car owners to confirm appointments, (4) checking their voice mail, and (5) traveling to and from their first and last appointments of the day.

Allstate claimed it had a policy prohibiting work off the clock.  It had a policy requiring approval for overtime.  If an adjuster worked before the start of the day, there was a means to claim the work time.

The Allstate workers filed a class action, in part alleging that Allstate's timekeeping system was illegal because Allstate permitted off the clock work.  The trial court initially granted certification.  After Wal-Mart v. Dukes came out, though, the company filed a motion to "decertify" the class.  The trial court granted decertification, which prompted the plaintiff to appeal.

The Court of Appeal here reversed the trial court and decided that the class should have been certified.  Here are the key points:

-  A motion to "decertify" a previously certified class action can be brought only when there has been a significant change - newly discovered facts or new law.

- On review of a motion to decertify, the court of appeal evaluates the trial court's stated rationale.  If the trial court's stated rationale is wrong, the appellate court will reverse.

- The court will ignore individual issues regarding how to calculate damages for individual employees if there is a common question applicable to all class members regarding liability:

Damage calculations have little, if any, relevance at the certification stage before the trial court and parties have reached the merits of the class claims. At the certification stage, the concern is whether class members have raised a justiciable question applicable to all class members. Although Allstate may have presented evidence that its official policies are lawful, “this showing does not end the inquiry.” (Jimenez, supra, 2012 WL 1366052, *8.) Here, the question is whether Allstate had a practice of not paying adjusters for off-the-clock time. (Ibid.) The answer to that question will apply to the entire class of adjusters. If the answer to that question is “yes” – which is the answer the trial court initially assumed when it first certified the Off-the-Clock class, and is the answer we must presume in reviewing decertification (Brinker, supra, 53 Cal.4th at p. 1023) – then, in Duke’s phrase, that answer is the “glue” that binds all the class members. (Dukes, supra, 131 S.Ct. at p. 2552 [a class requires the “glue” of a single answer for a question applicable to all class members].) If some adjusters had more uncompensated time off the clock than other adjusters, that difference goes to damages.

- The court's analysis of whether common questions predominate - usually the central issue on a motion for class certification is notable because it is part of a recent trend of holding that the absence of commonality does not preclude a finding of commonality:

Commonality exists when the class claim poses a question for which the answer advances the litigation. As Dukes explained, class “claims must depend upon a common contention . . . . That common contention, moreover, must be of such a nature that it is capable of classwide resolution – which means that determination of its truth or falsity will resolve an issue that is central to the validity of each one of the claims in one stroke.” (Dukes, supra, 131 S.Ct. at p. 2551.) * * *

* * * *
Allstate disputes whether a company-wide practice existed of adjusters working off the clock. According to Allstate, it instructs adjusters not to begin work before they arrive at their first appointment. Allstate asserts that at most “the evidence shows that reactions differed from manager to manager and from employee to employee, purportedly leading some adjusters to work off-the-clock, while others did not.” Allstate also asserts its policy is to pay for all overtime that adjusters work, and indeed, appellant concedes he received overtime pay 70 times.

But the Court of Appeal wasn't hearing it:

We need not, however, address the accuracy of Allstate’s assertions because doing so goes to the merits of the class claims. As our Supreme Court said in Brinker, supra, 53 Cal.4th at page 1024, inquiries into the merits as part of a certification motion are “closely circumscribed.” We instead assume based on the evidence appellant and other adjusters put to the trial court that Allstate had a company-wide practice of adjusters working off-the-clock. (Id. at p. 1023 [court assumes claims have merit].) An unlawful practice may create commonality even if the practice affects class members differently. “[C]lass treatment does not require that all class members have been equally affected by the challenged practices—it suffices that the issue of whether the practice itself was unlawful is common to all.”

So, this means that to defeat class certification, it is necessary to establish through evidence the absence of an unlawful practice. Yet, the court says that the employer's evidence that a practice is not unlawful as to all potential class members is part of the "merits" and, therefore, not part of the certification inquiry. That seems like a rather one-sided ruling, says Captain Obvious.

Anyway, this case is an important warning to employers with timekeeping systems that "assume" that hours worked start at a given time.  Employers should ensure timekeeping systems are not based on automatic punching and allow a worker to clock in or out based on when the work day (or meal breaks) actually starts and stops.

This case is Williams v. Superior Court and the opinion is here.