Charlene Brigham, As Personal Representative of the Estate of James Brigham; Carl Hall; Gary Millsap; Donald E. Reed, Plaintiffs-appellants, v. Eugene Water & Electric Board, Defendant-appellee, 357 F.3d 931 (9th Cir. 2004)

We must decide whether employees of an electric utility who reside on their employer's remote premises were wrongfully denied overtime pay in violation of the federal Fair Labor Standards Act and Oregon state law.

* James Brigham, Carl Hall, Gary Millsap, and Donald Reed ("the employees") are current or former employees stationed at the Carmen Smith Hydroelectric Project ("project" or "site"), a power generation facility straddling the upper McKenzie River some 70 miles east of Eugene, Oregon, on lands located predominantly within the Willamette National Forest.¹  Owned by the City of Eugene, Oregon through its Eugene Water & Electric Board ("EWEB"), the site was constructed in the early 1960's and is comprised of three sizeable dams and two powerhouses. Although the site is partially monitored at a central facility in Eugene, four EWEB employees work and are required to live (along with their families) on-site in housing provided to them by EWEB.² 

These employees³  worked four-day weeks, which usually were comprised of three "maintenance" shifts and one "duty" shift.⁴  On maintenance shifts, the employees worked from 6:30 a.m. to 5:00 p.m. Accounting for breaks, the employees performed ten hours' work during the course of a maintenance shift and accordingly were paid ten hours' wages. Any work performed beyond ten hours was paid at a double-time rate.

In contrast, duty shifts lasted a full 24 hours. During that time, a designated employee was responsible for the operation and safety of the entire project. Between 6:30 a.m. and noon, he was charged with monitoring, inspecting, and logging the status of the two generating plants and performing any necessary maintenance. At noon, he usually returned to his house. In the evening, he was required to inspect and again to log the status of powerhouses, a task which took about an hour. Thereafter — and indeed, for the entirety of his shift — the on-duty employee was required to remain at Carmen Smith, available for emergency phone or radio contact with the central dispatcher in Eugene.⁵  Each house on the site was also equipped with a system that would alert the employee to any automated monitoring alarms, to which (along with any calls from the central dispatcher) the duty employee was required to respond "immediately."⁶  Subject to these restrictions — as well as the requirement that they be "fit" — on-duty employees were free to sleep, to eat, and to spend time with their families.

Although the employees performed only about 6 hours of scheduled work during the course of a duty shift, they were paid ten hours' wages. On-duty employees also were compensated at a double-time rate for any call-out time lasting beyond a call's first 15 minutes. And, in addition to these wages, EWEB provided the employees with free housing, electricity, water, garbage service, and satellite television, along with a bus driver and the cost of fuel and maintenance for a school bus to transport the employees' children to school.

Between their maintenance and duty shifts, the employees were often on some form of duty status — either performing actual maintenance or on standby — for as much as 60 hours per week.

II

The employees filed suit in the circuit court of Lane County, Oregon, on August 14, 2000, alleging that their duty shift on-call time was uncompensated work⁷  and, accordingly, seeking compensation for unpaid overtime under the Fair Labor Standards Act ("FLSA"), 29 U.S.C. § 207(a) (1),⁸  and under two provisions of Oregon law, O.R.S. 279.340⁹  and O.R.S. 652.020.¹⁰  EWEB removed the case to federal court on September 15, 2000, and filed a motion for summary judgment on July 16, 2001. The employees opposed EWEB's motion, and argued alternatively that, if the district court granted summary judgment on the FLSA claim, it should decline to exercise supplemental jurisdiction over their state-law claims.

On August 30, 2001, the district court heard oral argument on the motion and that same day issued an order granting summary judgment to EWEB, denying the employees' request that it decline supplemental jurisdiction, and dismissing their state-law causes of action with prejudice. Judgment was entered on August 31, and seven days later the employees filed a motion to amend the judgment on the grounds that the district court should not have decided their state-law claims. Subsequently, the employees also objected to defense counsel's bill of costs.

The district court denied the motion to amend on October 31, 2001, and two days later issued an order awarding EWEB costs in the amount of $1,437.93. The employees timely filed an amended notice of appeal.

III

We first consider whether the (formally) uncompensated 14 hours of each 24-hour duty shift constituted compensable working time within the meaning of the FLSA.¹¹ 

* As the Supreme Court long ago recognized, the inquiry into whether "on-call" or "waiting" time constitutes compensable "working" time for purposes of FLSA § 207(a) (1) is particularly challenging. Although "no principle of law found either in the statute or in Court decisions precludes waiting time from also being working time [,] we cannot ... lay down a legal formula to resolve cases so varied in their facts as are the many situations in which employment involves waiting time." Skidmore v. Swift & Co., 323 U.S. 134, 136, 65 S. Ct. 161, 89 L. Ed. 124 (1944). " [F]acts may show that the employee was `engaged to wait,' which is compensable, or they may show that the employee `waited to be engaged,' which is not compensable." Owens v. Local No. 169, Ass'n of W. Pulp & Paper Workers, 971 F.2d 347, 350 (9th Cir. 1992) (quoting Skidmore, 323 U.S. at 137, 65 S. Ct. 161).

In determining whether the employees spent the uncompensated 14 hours of each duty shift "engaged to wait," and therefore working within the meaning of § 207(a) (1), or simply "waiting to be engaged," and therefore not working within the meaning of the statute, the Court has directed us to "scrutin [ize] and constru [e] the agreements between the particular parties, apprais [e] their practical construction of the working agreement by conduct, consider [] the nature of the service, and its relation to the waiting time, and all of the surrounding circumstances." Skidmore, 323 U.S. at 137, 65 S. Ct. 161. Our most recent cases emphasize that "the two predominant factors in determining whether an employee's on-call waiting time is compensable overtime are `(1) the degree to which the employee is free to engage in personal activities; and (2) the agreements between the parties.'" Berry, 30 F.3d at 1180 (9th Cir. 1994) (quoting Owens, 971 F.2d at 350).¹²  We address these considerations in turn.

B

In Owens, we enumerated an illustrative list of factors to consider in gauging the extent to which employees could pursue personal activities during the course of their on-call shifts:

(1) whether there was an on-premises living requirement; (2) whether there were excessive geographical restrictions on employee's movements; (3) whether the frequency of calls was unduly restrictive; (4) whether a fixed time limit for response was unduly restrictive; (5) whether the on-call employee could easily trade on-call responsibilities; (6) whether the use of a pager could ease restrictions; and (7) whether the employee had actually engaged in personal activities during call-in time.

Owens, 971 F.2d at 351 (footnotes omitted). "Because ` [n]o one factor is dispositive,' a court should balance the factors permitting personal pursuits against the factors restricting personal pursuits to determine whether the employee is so restricted that he is effectively engaged to wait." Berry, 30 F.3d at 1183 (quoting Owens, 971 F.2d at 351). Of course, an employee need not "have substantially the same flexibility or freedom as he would if not on call, else all or almost all on-call time would be working time, a proposition that the settled case law and the administrative guidelines clearly reject." Owens, 971 F.2d at 350-51 (quoting Bright v. Houston Northwest Med. Ctr. Survivor, Inc., 934 F.2d 671, 677 (5th Cir. 1991) (en banc)).

In this case, the Owens factors are closely divided. As previously noted, EWEB did require the employees to live on premises (factor 1). Indeed, it forbade them from maintaining an off-site primary residence. While on-call, the employees were subject to strict geographic constraints (factor 2): They had to remain within earshot of their home phones and alarm systems because they were required to respond instantaneously to any alerts (factor 4). Use of a pager (or, for that matter, a wireless phone or two-way radio) would do little to lessen these burdens (factor 6), as the plant was readily accessible only by foot from the employees' homes.

At the same time, the employees acknowledge that each was called out, on average, only about once or twice a month (factor 3). When an employee was sick, on vacation, or otherwise occupied by personal needs, he usually was able to trade duty shifts with his colleagues (factor 5). And, perhaps most notable in this regard, the employees routinely engaged in personal activities while they were on-call (factor 7). As the district court highlighted, they were able to use portions of their duty shifts to: sleep, eat, read, study, exercise, watch television, help their children with homework, play games, maintain their homes and yards, work on their motorcycles, and entertain guests.

This especially narrow division of the Owens factors reveals this case to be far closer than our previous FLSA waiting time cases (with the arguable exception of Service Employees International Union, Local 102 v. County of San Diego [SEIU], 60 F.3d 1346 (9th Cir. 1994), which evaluated fewer factors than we do here). Cf. SEIU, 60 F.3d at 1355 ("During the shift, there is an on-premises living requirement and strict geographical restrictions on employees' movements. However, the record shows the actual calls are infrequent, the on-call employee can trade on-call responsibilities, and employees do engage in personal activities."); Berry, 30 F.3d at 1183 (" [A]ll of the relevant factors in this case, except perhaps one, weigh in favor of concluding the coroners are free to engage in personal activities while on-call."); Owens, 971 F.2d at 353 ("The mechanics here were not required to remain on the premises; they were not required to remain at home to receive calls; they were not required to respond to all calls; they were not subject to a fixed acceptance rate; they received an average of only six calls a year; they were not required to be reachable by beeper; and, those who did not request beepers were not expected to respond to a call within a fixed time.") (footnotes omitted).

Nevertheless, we must find a way to balance these factors, see Berry, 30 F.3d at 1183, and comparatively evaluating them within the specific contours of this case, we conclude that they weigh — at least narrowly — in favor of the employees. First, the geographic and response-time restrictions associated with their duty shifts were especially restrictive. Because the employees had to be able to hear their phones ring at all times and were required to respond instantaneously to alerts and calls while on their duty shifts, they were effectively tethered to their homes. Perhaps in good weather they could enjoy use of their yards; at all other times, they would have to remain inside. In the event of a call, the employees had to be able to reach the powerhouse and accompanying dam — which was half a mile away and most expeditiously reached on foot — as soon as humanly possible.¹³  These limitations are more severe than those at issue in our prior cases, as well as those of our sister circuits. Cf. id. at 1178, 1184 (employees limited to remaining within the county, given fifteen minutes to respond by phone or two-way radio rather than in person); Owens, 971 F.2d at 349 (mechanics limited at most to pager area, could leave a forwarding number, and were required only to reply within ten minutes of receiving a call or page); see also, e.g., Ingram v. County of Bucks, 144 F.3d 265, 268-69 (3d Cir. 1998) (sheriff's deputies could leave word where they were reachable, were not required to report to office within a fixed amount of time, and often took 15-45 minutes before leaving their location in response to a call); Martin v. Ohio Turnpike Comm'n, 968 F.2d 606, 612 (6th Cir. 1992) (employees were free to leave a forwarding number, policy imposed no significant travel restrictions, no fixed time to respond in person); Renfro v. City of Emporia, Kan., 948 F.2d 1529, 1532 (10th Cir. 1991) (twenty minute response time supports the holding that on-call time is compensable).

Second, we believe that at least two of the Owens factors weighing against the employees have a lesser significance in this case. That the average number of call-outs and emergency alarms was low seems less important here than it might in the context of the coroners in Berry or the manufacturing plant mechanics in Owens. While on call, these particular employees were responsible for the safety of thousands of people and, accordingly, had to be absolutely prepared to respond at all times (i.e., rested, sober, clothed, and otherwise able to race immediately to the trouble source if needed), without regard to how often they were actually called out. Such constant pressures simply did not exist in Berry (where the coroners' dead bodies were not going to go anywhere) or Owens (where the mechanics freely could decline to respond to calls, and where, at most, a short period of plant downtime was at stake).

And although it is true that the employees could arrange to exchange their duty shifts with each other in advance — thereby enabling them to engage in personal activities on special occasions or to deal with illnesses and other family emergencies — we are inclined to see this particular Owens factor as somewhat less important than its peers. After all, our task ultimately is to determine how to characterize the time the employees spent on their duty shifts, and the employees' ability occasionally to avoid those shifts altogether does little to shape our view of the restrictions imposed on them during such shifts.

We therefore conclude that the Owens factors weigh narrowly in favor of the employees.

C

Our analysis of the issue, however, does not end there. The degree to which the employees were free to engage in personal activities is but one of the two factors that our caselaw directs us to consider in characterizing their duty shift on-call time. See Berry, 30 F.3d at 1180; Owens, 971 F.2d at 350. We now turn to an analysis of the second factor, the parties' agreement and its significance.

* We first observe that to the extent the parties dispute whether they had an agreement,¹⁴  we are quite certain that there was one. Our caselaw clearly recognizes that an agreement cognizable for purposes of the FLSA overtime inquiry may arise by conduct. In Owens, for instance, we explained that "the Plaintiff mechanics in the present case may not have liked the company's formal call-in system, but by continuing to work, they constructively accepted the new terms." Id. at 355. And in Berry, we reiterated that a "constructive agreement may arise if employees have been informed of the overtime compensation policy and continue to work under the disclosed terms of the policy." Berry, 30 F.3d at 1180; see also Gen. Elec. Co. v. Porter, 208 F.2d 805, 813 (9th Cir. 1953) ("We hold that as a matter of law the unilateral action of the employer was impliedly accepted by the firemen and that a new contract was created whereby the employees agreed to work on a two-platoon system at a fixed monthly wage.").

Our sister circuits likewise have recognized the force of constructive agreements in the FLSA overtime compensation context. See, e.g., Braziel v. Tobosa Dev. Servs., 166 F.3d 1061, 1063 (10th Cir. 1999) (" [A]n agreement to exempt sleep time from paid work under the FLSA can be implied.... Although it is clear from the record and appellants' pleadings that they became unhappy with the policy [,] it is equally clear that appellants understood and acquiesced to the policy when they were hired."); Bodie v. City of Columbia, 934 F.2d 561, 564-65 (4th Cir. 1991) (" [C]ontinuance in the job and acceptance under the new plan of payment was sufficient to create a valid agreement, even though the agreement was implied and not in writing."); Rousseau v. Teledyne Movible Offshore, Inc., 805 F.2d 1245, 1248 (5th Cir. 1986) ("Of course, it is clear that the employees did not like the no leave rule. But their dislike does not negate the existence of an agreement. As the district court pointed out, continuance of employment can be evidence of an implied agreement to the terms of that employment."); Ariens v. Olin Mathieson Chem. Corp., 382 F.2d 192, 197 (6th Cir. 1967) (" [W]e are of the opinion there was a meeting of minds resulting in a valid agreement that plaintiffs would not be paid for sleeping time. The work schedule was explained in the pamphlet given to each man before he commenced work. The men `found out' about their work schedule on their first day of duty. The work schedules were posted and plaintiffs continued throughout the time in question to accept paychecks which excluded sleeping time from hours worked. Certainly, this was sufficient to constitute an implied agreement between the parties....").

In this case, the record demonstrates (and the employees do not contest) that each of them was aware of EWEB's duty shift compensation policy when he began working such shifts at Carmen Smith, and that each continued to work under the relevant policy during the time he worked duty shifts. Whether the parties' agreement is delineated "express" — as arising through the employees' acceptance of duty shifts with a prior understanding of how they were to be compensated for those shifts, see Berry, 30 F.3d at 1180 — or "constructive" — as arising from the employees' decision to continue working under the policy, see id. — we arrive at a single conclusion: The parties had an agreement.

2

The far more interesting question is what we ought to make of the parties' agreement in this case. Berry explains the meaning of an agreement in the FLSA overtime context as follows:

The significance and importance of evaluating the agreements between the parties is that the existence of such agreements assists the trier of fact in determining whether the parties characterized the time spent waiting on-call as actual work. An agreement between the parties which provides at least some type of compensation for on-call waiting time may suggest the parties characterize waiting time as work. Conversely, an agreement pursuant to which the employees are to be paid only for time spent actually working, and not merely waiting to work, may suggest the parties do not characterize waiting time as work.

Id. at 1180-81. From that language, the employees argue that — because they were compensated for 10 hours of work on each duty shift when they performed only 6 hours of regularly scheduled work — their agreement with EWEB demonstrates that the parties characterized duty shift on-call time as time "worked" within the meaning of FLSA. Thus, they claim, the employees are owed overtime for the formally uncompensated 14 hours of each duty shift. For its part, EWEB counters that the four hours additional compensation associated with the employees' duty shifts merely recognized the inconvenience of being on-call in a remote location — quite in contrast to the portion of the parties' agreement providing double compensation for any actual call-out time, which recognized such time (and only that time) as actual additional work.¹⁵ 

We think both parties err in their attempts to draw conclusions from the agreement. In addressing this issue, we begin with a recognition that the United States Department of Labor (DOL) has promulgated substantial interpretive guidance designed to assist in assessing the compensability of waiting time under FLSA. See 29 C.F.R. §§ 785.14-23. Although these interpretive rules are non-binding, see United States v. Mead Corp., 533 U.S. 218, 232, 121 S. Ct. 2164, 150 L. Ed. 2d 292 (2001) (" [I]nterpretive rules may sometimes function as precedents, [but] they enjoy no Chevron status as a class."), we have nonetheless — along with our sister circuits — turned to these longstanding DOL regulations in resolving FLSA waiting time disputes.¹⁶  Most pertinent to the present case is 29 C.F.R. § 785.23, which provides:

An employee who resides on his employer's premises on a permanent basis or for extended periods of time is not considered as working all the time he is on the premises. Ordinarily, he may engage in normal private pursuits and thus have enough time for eating, sleeping, entertaining, and other periods of complete freedom from all duties when he may leave the premises for purposes of his own. It is, of course, difficult to determine the exact hours worked under these circumstances and any reasonable agreement of the parties which takes into consideration all of the pertinent facts will be accepted.

29 C.F.R. § 785.23 (2003) (emphasis added).¹⁷ 

The crucial insight of this regulation is that — in these specific circumstances, where an employee not only works but also resides on his or her employer's premises — the characterization of time for FLSA purposes need not be an all-or-nothing proposition, as it otherwise would be in the off-premises residency context. Rather, as the guidance perceptively notes, " [i]t is... difficult to determine the exact hours worked under these circumstances," id., and the parties' agreement provides a particularly important benchmark for assessing how many hours the employees actually labored.

Thus, the agreement between these employees and EWEB signifies neither that the entirety of the employees' duty shifts was time worked, nor that living on premises in a remote rocky canyon was inconvenient. It suggests instead that the parties agreed that duty shift call time was equivalent to about four hours' actual work on a maintenance shift.

3

Of course, not even § 785.23 treats such an agreement as dispositive. Rather, that guidance suggests that the parties' agreement should be accepted only if it is "reasonable" in light of the "pertinent facts." Having never directly confronted a case arising out of these particular circumstances, we have not previously elaborated on the substance of this inquiry. But that does not mean that we are rudderless: the Owens factors provide a worthy lens through which to gauge the reasonableness of the parties' mutual assessment of the work equivalence of the time spent by the employees on duty shifts in their on-premises homes.

Four of the Owens factors — the existence of an on-premises living requirement, the severity of geographic restrictions, the response time associated with emergency alarms and calls, and the viability of pager use — seem somewhat duplicative in the context of on-site residence. Nonetheless, they together speak to the pressures¹⁸  and other constraints imposed on employees by virtue of their on-call duties, and we see no reason why these factors cannot be accounted for as work time equivalents given the ways in which they diminished the otherwise unfettered enjoyment of the employees' time while on-call.

At the same time, all parties agreed that emergency call-outs were infrequent — averaging once or twice a month per employee — thereby suggesting that the employees were able to enjoy long periods of uninterrupted personal time. And the employees each testified to having actually enjoyed a wide variety of activities during their duty shift call time: sleeping, eating, reading, studying, exercising, watching television, helping their children with homework, playing games, maintaining their homes and yards, working on their motorcycles, and entertaining guests. Analyzed in this light, the parties' agreement treating the otherwise formally uncompensated duty shift call time as equivalent to four hours' actual work is eminently reasonable.

4

Unfortunately, this does not end our analysis. The district court viewed the parties' agreement somewhat differently, apparently holding that the mere existence of an agreement between the parties ended the FLSA inquiry. Quoting Owens, it concluded that "by continuing to work, plaintiffs constructively accepted the policy terms." District Court order at 19 (quoting Owens, 971 F.2d at 355) (alterations omitted).

Properly understood, however, the parties' agreement is just the starting point of the FLSA overtime analysis in this context — not its conclusion. Rather than providing employers with an exception to the FLSA overtime pay requirements, § 785.23 simply offers a sound methodology for calculating how many hours the employees actually worked within the meaning of FLSA. If that number exceeds 40 in a given workweek, the additional hours must be paid at the time-and-a-half rate demanded by 29 U.S.C. § 207(a) (1).

Although the record suggests that the employees usually worked three maintenance shifts and one duty shift in a given workweek — a schedule that would, under the analysis we set forth today, entail exactly 40 hours of compensable working time and thus no overtime — it is not clear that the employees always worked such a schedule. Indeed, the district court noted that the employees "often ... were on duty for 60 or more hours in a work week." Such a statement can only make sense if the employees sometimes worked at least four maintenance shifts and one duty shift in a workweek (40 maintenance hours plus 24 duty hours, or 64 total hours constituting 50 compensable hours under the formula we announce here), or three maintenance shifts and two duty shifts (30 maintenance hours plus 48 duty hours, or 78 total hours also constituting 50 compensable work hours).

Because the record does not contain the employees' time sheets for the relevant periods, and because the district court is, in the first instance, better suited to make such a determination, we must remand this case for a calculation of how much overtime each of the employees is owed under the formula we elaborate today.¹⁹  We imagine that such an inquiry will entail scrutiny of the employees' time cards and perhaps their employment records, but we leave it to the district court to craft a workable mechanism for determining how many hours each employee worked in each week he was employed at Carmen Smith, and, correspondingly, how much overtime compensation he is entitled to receive.

IV

For the foregoing reasons, the district court's judgment is VACATED and REMANDED for proceedings consistent with this opinion.