Cunningham v. Electronic Data Systems Corp.
This case was before the Court on Defendant, Electronic Data System’s (“EDS”) second Motion for Summary Judgment, seeking a finding that Plaintiff was exempt from the FLSA under the so-called “Air Carrier” Exemption. After a lengthy discussion of the history and application of the “Air Carrier” Exemption, which exempts any employee of a company covered by the Railway Labor Act (“RLA”) from FLSA coverage, the Court found the exemption inapplicable to the Plaintiff, because Defendant failed to prove that it was “directly or indirectly owned or controlled by, or under common control with, a carrier or carriers.”
Reaching its decision, the Court reasoned:
“The RLA defines the scope of its coverage of carrier affiliates with the following language:
The term ‘carrier’ includes … any company which is directly or indirectly owned or controlled by or under common control with any carrier by railroad [or air] and which operates any equipment or facilities or performs any service (other than trucking service) in connection with the transportation, receipt, delivery, elevation, transfer in transit, refrigeration or icing, storage, and handling of property transported by railroad [or air].
The National Mediation Board (“NMB”), the agency charged with administering the RLA, formulates this statutory language into the following test:
When an employer is not a rail or air carrier in the transportation of freight or passengers, the NMB applies a two-part test in determining whether the employer and its employees are subject to the RLA. First, the NMB determines whether the nature of the work is that traditionally performed by employees of rail or air carriers-the function test. Second, the NMB determines whether the employer is directly or indirectly owned or controlled by, or under common control with, a carrier or carriers-the control test. Both parts of the test must be satisfied for the NMB to assert jurisdiction.
In re Int’l Cargo, 31 NMB at 406 (citations omitted). This test is now well-established in the agency’s case law and, as the Court noted in its prior opinion, courts defer to it as the proper construction of the statute. Cunningham, 579 F.Supp.2d at 542;District 6, 139 F.Supp.2d at 561.
Though the NMB’s interpretation of the RLA governs this case, the FLSA context remains relevant in one crucial sense. It is well-settled that “exemptions to the FLSA are narrowly construed against [ ] employers.” Bilyou v. Dutchess Beer Distribs., Inc., 300 F.3d 217, 222 (2d Cir.2002) (internal quotations omitted). As a matter of substantive law, it is not apparent how that doctrine should fit into the Court’s analysis, because by the statute’s plain language the FLSA’s air carrier exemption and the RLA are coextensive. 29 U.S.C. § 213(b)(3). It would be strange to apply the RLA’s definition of “carrier” differently in FLSA cases than in RLA cases, and other courts have not done so. See Verrett, 70 F.Supp.2d at 1281-83;Slavens v. Scenic Aviation, Inc., 221 F.3d 1353, at *1 (10th Cir.2000). However, though the FLSA context would not seem to alter the substance of applicable RLA law, the narrow standard of construction for FLSA exemptions does have an important procedural impact-it charges the employer with the burden of proving that an exemption applies. Cunningham, 579 F.Supp.2d at 540;Martin v. Malcolm Pirnie, Inc., 949 F.2d 611, 614 (2d Cir.1991) (“[A]n employer bears the burden of proving that its employees fall within an exempted category of the [FLSA].”). Thus, summary judgment is only appropriate here if EDS has submitted enough undisputed evidence to prove that it meets both the “function” and “control” prongs of the NMB’s carrier-affiliate test. See Martin, 949 F.2d at 616. EDS has met this burden with respect to the function prong but not with respect to the control prong.
EDS plainly satisfies the function prong. The controlling inquiry is “whether the nature of the work is that traditionally performed by employees of rail or air carriers.” In re Int’l Cargo, 31 NMB at 406. As the agency applies the test, any work that proximately advances the carrier’s commercial transportation services, even if that work is not inherently aviation-related, satisfies the standard. Thus, employers that provide sky-cap, janitorial, shuttle-bus, and cargo-handling services have all been found to satisfy the function prong. Id. at 406 (cargo-handling); In re Kanonn, 31 NMB 409, 417 (June 18, 2004) (janitorial and sky-cap services); In re Air Serv Corp., 35 NMB at 210 (shuttle-bus). The focus of the test is whether the carrier affiliate’s services are sufficiently connected to the carrier’s commercial transportation operations that a work stoppage at the carrier affiliate would impede those operations. See Verrett, 70 F.Supp.2d at 1281 (finding that employer that provided IT services to American Airlines met the function prong) (“When the activities of carrier affiliates are necessary to the operations of an air carrier, and a labor dispute at the affiliate could cripple airline operations, those affiliates must be subject to the RLA because such disruption is the very type of interruption to air commerce the RLA was designed to prevent.”); In re Milepost Indus ., 27 NMB 362, 366 (May 9, 2000) (function prong satisfied where employer’s work was “an integral part of the carrier’s business.”).
Like the employer in Verrett, which also provided IT services to American, EDS provides American with services that are crucial to the airline’s commercial operations. (Shanks Decl. ¶ 3.) In fact, EDS’s services to American are more or less the same as the services found to pass the “function” prong in Verrett: EDS acquired SABRE, the IT company at issue in Verrett, soon after that case was decided, and EDS continued thereafter to offer the airline the “same scope of IT services” that SABRE provided before the acquisition. (Arnold Decl. ¶ 4-5.) The Court agrees with Verrett ‘s reasoning that because EDS’s services, like the services SABRE provided previously, are “absolutely integral” to American’s operations, those services pass the function prong. Verrett, 70 F.Supp.2d at 1283. This conclusion finds further support in the agency opinions holding that services of more peripheral significance to air transport operations pass the function test. See, e.g., In re Dobbs Int’l Services, 34 NMB 97, at *7 (March 2, 2007) (in-flight catering); In re John Menzies, 31 NMB 490, 504 (Aug. 26, 2004) (cabin-cleaning).
The control prong requires more analysis. The relevant inquiry is whether “the employer is directly or indirectly owned or controlled by, or under common control with, a carrier or carriers.” In re Int’l Cargo, 31 NMB at 406. EDS does not claim to be owned by or under common ownership with a carrier, so it can only satisfy this prong of the test by showing that it is “controlled by” American. The Court summarized the NMB’s framework for applying this statutory language in its opinion denying defendant’s first summary judgment motion:
[I]n determining whether an entity is controlled by an air carrier, the NMB considers factors including “the extent of the carrier control over the manner in which the company conducts its business; access to the company’s operations and records; the carrier’s role in personnel decisions; the carrier’s degree of supervision over the company’s employees; the carrier’s control over employee training; [and] whether company employees are held out to the public as employees of the carrier,” John Menzies, 31 N.M.B. at 504-05, “the carrier’s role in the entity’s daily operations,” “the entity’s employees’ performance of services for the carriers,” and “the degree to which the carriers affect other conditions of employment,” Int’l Total Servs., 26 N.M.B. 72, 75.
Cunningham, 579 F.Supp.2d at 542 (alterations omitted). Not addressed in this summary is the threshold issue of whether the NMB’s control analysis should apply to EDS as a single corporate entity or instead to a discrete area of EDS’s operations relating to air travel. The Court considers this question, which has received surprisingly little direct attention from the NMB, infra at pages 14-16.
The NMB’s long list of control factors, though now an established feature of the agency’s precedent, does not provide precise guidance on questions of carrier control. Rather, with so many considerations in play, the line between “control” and “lack of control” blurs. The agency has determined, for example, that a sky-cap, baggage handling, cabin cleaning, and food services contractor, which gave air carriers substantial control over its personnel decisions and allowed the carriers to train its employees directly on a range of procedures, was not “controlled by” its air carrier clients. In re Ogden Aviation Serv., 23 NMB 98 (Feb. 5, 1996). On the other hand, the NMB also held that a contractor providing janitorial and sky-cap services, which accepted only limited input from the carrier as to personnel matters, but whose employees worked under close air carrier supervision, and which received its equipment and office space from the air carriers, was “controlled by” the air carriers. In re Kanonn Serv. Enter. Corp., 31 NMB 409 (June 18, 2004). Though some distinctions do exist between the cases-in Kanonn, the carriers provided equipment and office space and supervised affiliate employees more closely-it is difficult to draw a principled line between the outcomes, particularly under the NMB’s amorphous list of factors. See Ogden, 23 NMB at 107; Kanonn, 31 NMB at 417; compare also In re Huntleigh USA Corp., 29 NMB 121 (Dec. 17, 2001) (sky-cap and baggage handling contractor “controlled by” carrier).
Assessing the relative weight the NMB places on each factor brings the test into sharper focus. The factors certainly vary in importance. One in particular-the extent of the carrier’s role in the subject company’s daily operations-appears to determine the outcome in many cases. Where the carrier controls the details of the day-to-day process by which the contractor provides its services-for example, the number of employees assigned to particular tasks, the employees’ attire, the length of their shifts, and the methods they use in their work-the control prong is likely satisfied. On certain occasions, the NMB has acknowledged that this factor bears heightened significance. Ogden, 23 NMB at 104 (“[The Board] focuses on the carriers’ role in the entity’s daily operations and its effect on the manner in which employees perform their jobs.”); Int’l Total Servs., 26 N.M.B. at 75 (emphasizing the “carriers’ role in the entity’s daily operations”). Further, in every case the parties cite holding that a carrier affiliate satisfied the control prong, the air carrier exerted substantial control, on a day-to-day basis, over how affiliate employees did their jobs. See Kanonn, 31 NMB at 417 (“Delta determines how many employees work each shift and at what locations. Delta also must authorize the use of overtime and dictate the maximum hours Kanonn employees can work each month. While Delta does not directly supervise Kanonn employees, Delta managers meet with [a Kanonn supervisor] on a daily basis to review Kanonn employee performance.”).The day-to-day operational control factor is not independently dispositive; in some of the cases finding a lack of requisite control, carriers oversaw aspects of the contractor’s daily operations. See, e.g., In re Signature Flight Support, 32 NMB 214, 220 (Aug. 31, 2005) (carrier imposed “stringent demands” on employer’s fueling and towing procedures and met with employer throughout the day to discuss scheduling needs and discuss employee performance). But there is little doubt that the agency places more weight on this factor than the others on its list (such as “records access” and whether the subject employees are “held out to the public” as the carrier’s own), and it seems that, where the carrier’s control over daily operations is particularly extensive, the control prong will always be satisfied. See In re Int’l Cargo, 31 NMB 396 (cargo services employer “controlled by” carrier where carrier dictated storage and handling techniques, set requirements for performance of administrative functions, required that the contractor use a specific computer system, and determined the priority of assignments).
The carrier’s power over the subject employer’s personnel decisions is another consideration of heightened import. Cases finding a lack of requisite control usually emphasize that the carrier affiliate “hires, trains, pays, promotes, transfers, evaluates, disciplines, and if necessary, discharges” its own employees. See Ogden Aviation, 23 NMB at 106. On the other hand, carrier authority over hiring decisions and employee termination is a solid indication that the carrier “controls” the employer. Huntleigh, 29 NMB at 126 (“Southwest is entitled to request the removal of any Huntleigh employee that Southwest believes displays improper conduct.”); In re AirServ Corp., 35 NMB at 211 (“FedEx has final authority to approve or reject a[job] applicant…. [Air Serv] has [ ] never refused a request by FedEx to discipline or terminate an employee.”). This factor is not a perfect bellwether either-in Kanonn, the carrier was not heavily involved in personnel matters-but a carrier’s right to terminate employees or reverse hiring decisions seems to argue strongly for satisfaction of the control prong. See id.
Thus, though all of the NMB’s factors remain relevant to the analysis, day-to-day operational control and influence over personnel decisions do much to determine outcomes under the control prong.
The Court now turns to application of the control factors to this case. As a threshold matter, the parties disagree over the proper scope of that application. Plaintiffs argue that the control prong should apply to EDS generally, as a multi-national IT services provider with clients in a wide array of industries. Under that framework, there is little doubt that American does not “control” EDS, because EDS provides services to hundreds of non-carrier clients. Defendants, on the other hand, argue that the test should apply “solely to that portion of an employer’s business that relates to an air carrier.” (Def. Mem. at 19.) Under that view, EDS’s satisfaction of the prong would turn on whether American controls the manner in which EDS provides services to American, specifically, regardless of the extent of EDS’s non-carrier operations.
Defendants have the better argument. NMB precedent demonstrates that the agency limits application of its control-prong factors to, at the very least, the segment of an employer’s operations that relate to the carrier or carriers in question. In Argenbright Security, 29 NMB 332 (June 13, 2002), the carrier affiliate provided “security and ancillary services” through two distinct operating divisions-one that served airlines exclusively, and another that served only non-carriers. The particular set of carrier-affiliate employees at issue in the case-sky-cap workers whose union sought a ruling that the NLRA governed disputes between it and the carrier-affiliate-worked for Continental airlines at Newark airport. In finding the control prong satisfied, the NMB considered only the employer’s relationship with Continental at Newark, not its relationship with other carriers, and not its non-carrier business. Id. at 338-39. Other agency decisions show a similarly limited control analysis. See, e.g., Milepost Indus., 27 NMB at 366-67 (narrowing control analysis to relationship between employer and particular carriers served by employees involved in representation dispute). The rule from these cases-that the control analysis should consider the specific relationship between an employer and a commercial air carrier that a particular dispute implicates-stands on solid ground. If the test applied to employers like EDS as a whole, entire units of workers providing carrier-support services would not be subject to the RLA, regardless of the integrality of their work to air transportation, simply because they fall under the umbrella of large corporations with additional divisions serving other industries. Such a result would not advance the statute’s purpose of protecting the continuity of commercial air transportation. See Verrett, 70 F.Supp.2d at 1283 (“The fact that a company provides services to carriers and to some non-carriers does not detract from RLA jurisdiction, as long as there is still an essential element of transportation-related service.”). Thus, EDS meets the control test if its relationship with American-the client for whom the Cunninghams worked-evinces the requisite control.
A further issue, which the parties have not addressed, is whether the control test may apply to an even narrower segment of the employer’s operations-defined either geographically or functionally-than its relationship with a specific carrier client. The NMB is apparently willing to focus the control analysis upon a specific geographical unit of a carrier affiliate’s operations. In re Huntleigh USA Corp., 29 NMB 121 (Dec. 17, 2001) (limiting analysis to operations at one airport). If the same scope limitation were appropriate for a particular category of employees working for a particular client, then EDS might satisfy the control prong simply by showing that Staff Augs at American-as a distinct unit of employees-are “controlled by” American. This framework would require applying the “craft or class” test used to define bargaining units under the RLA to determinations of whether specific subsets of carrier-affiliate employees fall under NMB or NLRB jurisdiction. See, e.g., In re Merger of Grand Trunk Western Railroad Co., 17 NMB 282, 288 (June 18, 1990). The NMB’s carrier-affiliate case law, however, does not appear to address the issue. Because the parties have not briefed the question, and because the current record does not permit a determination as to whether Staff Augs constitute a “craft or class,” the Court declines to consider the issue now, though trial of this action may well bring it to the fore.
While EDS’s formulation of the control prong (that it requires analysis only of the IT services EDS provides to American) is correct, at least as far as it goes, defendant has not met its burden of proving that the prong is satisfied. To establish that its relevant operations are “controlled by” American, EDS relies mainly upon its contract with the airline. Admittedly, portions of the contract do support defendant’s case, at least to some extent. For example, the Agreement requires that EDS house three American software applications on its computers, (Agrmnt. at 35); and obtain American’s approval before using any third-party software. (Id. at 31-32). The Agreement also affords American the right to reject any particular software update EDS wishes to implement, (Id. at 15), and sets conditions under which EDS may use American office space. (Id. at 32-33.) Further, the subsidiary “service level” agreements impose performance standards on EDS, covering such issues as “average response time to outages” and “the number of outages per month on network resources.” (Shanks Decl. ¶¶ 7 .) These contractual provisions sketch the parameters of a relationship that might approach the level of control present in the cases defendant cites, see supra n. 4 (collecting cases), but nothing in the record informs the Court as to the actual nature of American’s relationship with EDS, or the actual level of control that one entity exerts over the other. Do all EDS employees in fact work from American office space? Does American actually exercise tight oversight of the software EDS uses? The record does not provide any answers. In the same vein, does American actually monitor and enforce the service levels in a way that approximates the level of daily supervision seen in the NMB cases? See, e.g., Kanonn, 31 NMB at 417; In re Air Serv, 35 NMB at 211. The little evidence defendant has submitted on this question suggests that American, though it can access compliance data on an EDS website, in fact monitors compliance only “periodically,” mainly through monthly and quarterly reports from EDS. (Shanks Decl. ¶¶ 6-11.) That comparatively sparse level of operational supervision argues against satisfaction of the control prong. See Kanonn, 31 NMB at 417.
Indeed, the most obvious gap in the evidence concerns the two most significant NMB factors: American’s day-to-day involvement in EDS’s operations, and its influence over EDS’s personnel decisions. The record does not indicate where core IT employees assigned to the American account work, let alone whether American controls their schedules, work procedures, attire, or anything else about their workday. And while the contract provides American with some power over whom EDS hires to fill the roles of account manager and five other “key” positions, nothing in the record indicates whether American actually exerts any such influence. Moreover, the NMB has found that control over only top managers does not necessarily demonstrate sufficient power over personnel decisions to satisfy the control prong. In re Bombardier, 32 NMB 131, 139-40 (Jan. 31, 2005) (lack of requisite control even where carrier retained “right to approve [employer’s] selection of the General Manager ….”). The record is silent as to whether American controls personnel decisions regarding other core IT workers, beyond the small group mentioned in the Agreement.
Aside from the contractual provisions, most of the remainder of EDS’s control evidence pertains exclusively to Staff Augs. According to the record, Staff Augs work from American office space, among American employees, and under the supervision of American managers, who control Staff Aug assignments, work schedules, and dress code. (Def. 56.1 ¶¶ 95-96, 105; Cuccrese Decl. ¶¶ 7-15.) American establishes Staff Aug hiring criteria, interviews candidates, and makes final hiring decisions. (Def. 56.1 ¶ 93.) When American is dissatisfied with a Staff Aug’s performance, it can demand that EDS discipline or remove the individual, and EDS habitually complies with those demands. (Id. at ¶¶ 109-11.) This evidence certainly shows that Staff Augs like the Cunninghams are “controlled by” American, but it only works to emphasize the inadequacy of the record as a whole, under the interpretation of the control prong EDS has proposed. American certainly controls personnel decisions regarding Staff Augs, but nothing shows similar control over EDS’s other IT workers-the employees who provide American with “midrange hosting (i.e., servers), mainframe hosting, network, applications development and maintenance, helpdesk and desktop [services].” (Id. at 9.) Similarly, with respect to operational control, it is evident that American supervises Staff Augs on a daily basis, but the record does not indicate whether American controls any details of daily operations for core IT workers. EDS has submitted ample declarations about Staff Aug life, but no evidence addresses the circumstances in which core IT employees work. And the record does not say what portion of the EDS-American relationship Staff Augs represent. For all the Court knows, Staff Augs are a minor aspect of the relationship. Thus, under EDS’s own formulation of the control prong-that it implicates the “portion of [EDS’s] business that relates to [American]”-the Staff Aug evidence does not satisfy EDS’s burden to prove that it is “controlled by” American. Rather, that evidence highlights the insufficiency of the record concerning American’s overall relationship with EDS.
Because EDS has not met its burden with respect to the control prong, summary judgment is denied. See Cunningham, 579 F.Supp.2d at 540.”
