Browning-Ferris Industries of California, Inc., d/b/a BFI Newby Island Recyclery, and FRP-II, LLC.

National Labor Relations Board - Board Decisions

Aug 27, 2015

362 N.L.R.B. 186 (N.L.R.B. 2015)

362 NLRB No. 186 NOTICE: This opinion is subject to formal revision before publication in the bound volumes of NLRB decisions. Readers are requested to notify the Ex- ecutive Secretary, National Labor Relations Board, Washington, D.C. 20570, of any typographical or other formal errors so that corrections can be included in the bound volumes. Browning-Ferris Industries of California, Inc., d/b/a BFI Newby Island Recyclery, and FPR-II, LLC, d/b/a Leadpoint Business Services, and Sanitary Truck Drivers and Helpers Local 350, Interna- tional Brotherhood of Teamsters, Petitioner. Case 32-RC-109684 August 27, 2015 DECISION ON REVIEW AND DIRECTION BY CHAIRMAN PEARCE AND MEMBERS MISCIMARRA, HIROZAWA, JOHNSON, AND MCFERRAN In this case, we consider whether the Board should ad- here to its current standard for assessing joint-employer status under the National Labor Relations Act or whether that standard should be revised to better effectuate the purposes of the Act, in the current economic landscape. The issue in this case is whether BFI Newby Island Recyclery (BFI), and Leadpoint Business Services (Leadpoint) are joint employers of the sorters, screen cleaners, and housekeepers whom the Union petitioned to represent. The Regional Director issued a Decision and Direction of Election finding that Leadpoint is the sole employer of the petitioned-for employees.1 The Un- ion filed a timely request for review of that decision, contending that (a) the Regional Director ignored signifi- cant evidence and reached the incorrect conclusion under current Board precedent; and (b) in the alternative, the Board should reconsider its standard for evaluating joint- employer relationships. In granting the Union’s request for review, we invited the parties and interested amici to file briefs addressing the following questions: 1. Under the Board’s current joint-employer standard, as articulated in TLI, Inc., 271 NLRB 798 (1984), enfd. mem. 772 F.2d 894 (3d Cir. 1985), and Laerco Trans- portation, 269 NLRB 324 (1984), is Leadpoint Busi- ness Services the sole employer of the petitioned-for employees? 2. Should the Board adhere to its existing joint- employer standard or adopt a new standard? What considerations should influence the Board’s decision in this regard? 3. If the Board adopts a new standard for determining joint-employer status, what should that standard be? If it involves the application of a multifactor test, what 1 An election was conducted on April 25, 2014, after which the bal- lots were impounded. factors should be examined? What should be the basis or rationale for such a standard? In response, the General Counsel, a group of labor and em- ployment law professors, and several labor organizations, as well as other amici, have urged the Board to adopt a new standard. Employer groups, in contrast, argue that the Board should adhere to its current standard. The current standard, as reflected in Board decisions such as TLI and Laerco, supra, is ostensibly based on a decision of the United States Court of Appeals for the Third Circuit, NLRB v. Browning-Ferris Industries of Pennsylvania, Inc., 691 F.2d 1117 (3d Cir. 1982), enfg. 259 NLRB 148 (1981), which endorsed the Board’s then-longstanding standard. But, as we will explain, the Board, without explanation, has since imposed additional requirements for finding joint-employer status, which have no clear basis in the Third Circuit’s decision, in the common law, or in the text or policies of the Act. The Board has never articulated how these additional re- quirements are compelled by the Act or by the common- law definition of the employment relationship. They appear inconsistent with prior caselaw that has not been expressly overruled. Moreover, these additional requirements-which serve to significantly and unjustifiably narrow the circum- stances where a joint-employment relationship can be found-leave the Board’s joint-employment jurispru- dence increasingly out of step with changing economic circumstances, particularly the recent dramatic growth in contingent employment relationships. This disconnect potentially undermines the core protections of the Act for the employees impacted by these economic changes. In the Supreme Court’s words, federal regulatory agencies “are supposed, within the limits of the law and of fair and prudent administration, to adapt their rules and practices to the Nation’s needs in a volatile, chang- ing economy.”2 Having carefully considered the record and the briefs,3 we have decided to revisit and to revise 2 American Trucking Assns. v. Atchison, T. & S.F. Ry. Co., 387 U.S. 397, 416 (1967). See, e.g., UGL-UNICCO Service Co., 357 NLRB No. 76, slip op. at 1 (2011) (quoting American Trucking Assns., supra, and revising Board’s successor-bar doctrine). 3 The Union, BFI and Leadpoint each filed an initial brief and a brief in response to amici’s briefs. Amicus briefs were filed by the American Federation of Labor and Congress of Industrial Organiza- tions; the American Staffing Association; a group of entities consisting of the Coalition for a Democratic Workplace and 15 other amici; the Council on Labor Law Equality; the Driver Employer Council of America; the Equal Opportunity Employment Commission; the General Counsel; the International Alliance of Theatrical Stage Employees, Moving Picture Technicians, Artists and Allied Crafts of the United States, its Territories and Canada; the International Franchise Associa- tion; a group of labor and employment law professors; the Labor Rela- tions and Research Center at the University of Massachusetts, Amherst; 2 DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD the Board’s joint-employer standard. Our aim today is to put the Board’s joint-employer standard on a clearer and stronger analytical foundation, and, within the limits set out by the Act, to best serve the Federal policy of “en- couraging the practice and procedure of collective bar- gaining.”4 Today, we restate the Board’s joint-employer standard to reaffirm the standard articulated by the Third Circuit in Browning-Ferris decision. Under this standard, the Board may find that two or more statutory employers are joint employers of the same statutory employees if they “share or codetermine those matters governing the essen- tial terms and conditions of employment.”5 In determin- ing whether a putative joint employer meets this stand- ard, the initial inquiry is whether there is a common-law employment relationship with the employees in question. If this common-law employment relationship exists, the inquiry then turns to whether the putative joint employer possesses sufficient control over employees’ essential terms and conditions of employment to permit meaning- ful collective bargaining. Central to both of these inquiries is the existence, ex- tent, and object of the putative joint employer’s control. Consistent with earlier Board decisions, as well as the common law, we will examine how control is manifested in a particular employment relationship. We reject those limiting requirements that the Board has imposed- without foundation in the statute or common law-after Browning-Ferris. We will no longer require that a joint employer not only possess the authority to control em- ployees’ terms and conditions of employment, but also exercise that authority. Reserved authority to control terms and conditions of employment, even if not exer- cised, is clearly relevant to the joint-employment in- quiry.6 As the Supreme Court has observed, the question a group of entities consisting of the National Association of Manufac- turers and two other amici; a group of entities consisting of the National Council for Occupational Health and Safety and nine other amici; a group of entities consisting of the National Employment Law Project and nine other amici; the Retail Litigation Center; the Service Employ- ees International Union; and the United States Chamber of Commerce. 4 29 U.S.C. §151. 5 Browning-Ferris Industries of Pennsylvania, Inc., supra, 691 F.2d at 1123. As explained below, we will adhere to the Board’s inclusive approach in defining the “essential terms and conditions of employ- ment.” The Board’s current joint-employer standard, articulated in TLI, supra, refers to “matters relating to the employment relationship such as hiring, firing, discipline, supervision, and direction,” a nonexhaustive list of bargaining subjects. TLI, supra, 271 NLRB at 798 (emphasis added). 6 See, e.g., Restatement (Second) of Agency §2(1) (“A master is a principal who employs an agent to perform service in his affairs and who controls or has the right to control the physical conduct of the other in the performance of the service.”) (emphasis added); id., §220(1) (“A servant is a person employed to perform services in the is whether one statutory employer “possesse[s] sufficient control over the work of the employees to qualify as a joint employer with” another employer.7 Nor will we require that, to be relevant to the joint-employer inquiry, a statutory employer’s control must be exercised directly and immediately. If otherwise sufficient, control exer- cised indirectly-such as through an intermediary-may establish joint-employer status.8 The Board’s established presumption in representation cases like this one is to apply a new rule retroactively.9 Applying the restated joint-employer standard here, we reverse the Regional Director and find that the Union established that BFI and Leadpoint are joint employers of the employees in the petitioned-for unit. I. FACTS A. Overview BFI owns and operates the Newby Island recycling fa- cility, which receives approximately 1,200 tons per day of mixed materials, mixed waste, and mixed recyclables. The essential part of its operation is the sorting of these materials into separate commodities that are sold to other businesses at the end of the recycling process. BFI sole- ly employs approximately 60 employees, including load- er operators, equipment operators, forklift operators, and spotters. Most of these BFI employees work outside the facility, where they move materials and prepare them to be sorted inside the facility. These BFI employees are part of an existing separate bargaining unit that is repre- sented by the Union. The interior of the facility houses four conveyor belts, called material streams. Each stream carries a different category of materials into the facility: residential mixed recyclables, commercial mixed recyclables, dry waste process, and wet waste process. Workers provided to BFI by Leadpoint stand on platforms beside the streams and sort through the material as it passes; depending on where they are stationed, workers remove from the stream either recyclable materials or prohibited materials. Other material is automatically sorted when it passes through screens that are positioned near the conveyor belts. affairs of another and who with respect to the physical conduct in the performance of the services is subject to the other’s control or right to control.”) (emphasis added). 7 Boire v. Greyhound Corp., 376 U.S. 473, 481 (1964). To be sure, a joint employer will be required to bargain only with respect to those terms and conditions over which it possesses sufficient control for bargaining to be meaningful. 8 See, e.g., Restatement (Second) of Agency §220, comment d (“[T]he control or right to control needed to establish the relation of master and servant may be very attenuated.”). 9 See, e.g., UGL-UNICCO, 357 NLRB No. 76, slip op. at 8 & fn. 28 (2011). BFI NEWBY ISLAND RECYCLERY 3 As indicated, BFI, the user firm, contracts with Leadpoint, the supplier firm, to provide the workers who manually sort the material on the streams (sorters), clean the screens on the sorting equipment and clear jams (screen cleaners), and clean the facility (housekeepers).10 The Union seeks to represent approximately 240 full- time, part-time, and on-call sorters, screen cleaners, and housekeepers who work at the facility.11 The relationship between BFI and Leadpoint is gov- erned by a temporary labor services agreement (Agree- ment), which took effect in October 2009, and remains effective indefinitely. It can be terminated by either par- ty at will with 30 days’ notice. The Agreement states that Leadpoint is the sole employer of the personnel it sup- plies, and that nothing in the Agreement shall be con- strued as creating an employment relationship between BFI and the personnel that Leadpoint supplies. B. Management Structure BFI and Leadpoint employ separate supervisors and lead workers at the facility. BFI Operations Manager Paul Keck oversees the material recovery facility and supervises the BFI employees. BFI Division Manager Carl Mennie oversees the recycling and compost opera- tions and reports to Keck. Shift Supervisors Augustine Ortiz and John Sutter supervise BFI employees at the site, including the control room operator. They also spend a percentage of each workday in the material stream areas, monitoring the operation and productivity of the streams. Ortiz testified that part of his job is to ensure the productivity of the streams. Leadpoint employs Acting On-Site Manager Vincent Haas, three shift supervisors, and seven line leads who work with the Leadpoint sorters. Haas oversees Leadpoint operations at the facility and reports to the Leadpoint corporate office in Arizona. The shift super- visors, who report to Haas, create the sorters’ schedules, oversee the material streams, and coach the line leads. The line leads work on the floor with the sorters and are Leadpoint’s first-line supervisors.12 Frank Ramirez, Leadpoint’s CEO and President, visits the facility two or 10 Consistent with previous Board decisions, we refer to the compa- ny that supplies employees as a “supplier” firm and the company that uses those employees as a “user” firm. 11 BFI solely employs one sorter who works alongside the Leadpoint employees and performs identical job duties. She is part of the Union’s existing unit of BFI employees and makes approximately $5/hour more in wages than the Leadpoint employees. BFI asserts that she was given sorter duties years ago after her position was eliminated owing to the loss of a municipal contract; she is grandfathered into BFI’s existing contract with the Union, which otherwise exempts sort- ers from that bargaining unit. 12 The parties agreed that Leadpoint’s line leads are statutory super- visors. three times per quarter to evaluate whether Leadpoint is meeting BFI’s expectations and goals; he also meets with BFI and Leadpoint managers, and addresses any prob- lems. BFI and Leadpoint maintain separate human resource departments. BFI does not have an HR manager onsite. Leadpoint has an onsite HR manager who operates in a trailer (marked with the Leadpoint logo) outside the fa- cility. Leadpoint employees use the BFI break rooms, bathrooms, and parking lot. C. Hiring The Agreement between BFI and Leadpoint provides that Leadpoint will recruit, interview, test, select, and hire personnel to perform work for BFI. BFI Managers Keck and Mennie, and Shift Supervisors Ortiz and Sutter testified that they are not involved in Leadpoint’s hiring procedure and have no input into Leadpoint’s hiring de- cisions. However, as to hiring, the Agreement requires Leadpoint to ensure that its personnel “have the appro- priate qualifications (including certification and training) consistent with all applicable laws and instructions from [BFI], to perform the general duties of the assigned posi- tion.” BFI also has the right to request that personnel supplied by Leadpoint “meet or exceed [BFI’s] own standard selection procedures and tests.” The Agreement also requires Leadpoint to make “rea- sonable efforts” not to refer workers who were previous- ly employed by BFI and were deemed ineligible for re- hire. Under the Agreement, Leadpoint must ask workers if they were previously employed by BFI and verify with BFI that all workers provided are eligible to work with BFI. If Leadpoint inadvertently refers an ineligible worker, it must immediately cease referring her, upon notification by BFI. Before it refers a worker to BFI, Leadpoint is also re- quired to ensure, in accordance with the Agreement, that she has passed, at minimum, a five-panel urinalysis drug screen, “or similar testing as agreed to in writing with [BFI’s] safety, legal and commercial group.” Leadpoint is not permitted to refer workers who do not successfully complete the drug screen, and BFI may request written certification of such completion. After Leadpoint has referred workers, it is responsible for ensuring that they remain free from the effects of alcohol and drug use and in condition to perform their job duties for BFI. When an applicant arrives at the Newby Island facility, she reports to Leadpoint’s HR department. Leadpoint tests and evaluates an applicant’s ability to perform the required job tasks at BFI by giving her a try-out on the material stream and assessing whether she has adequate hand-eye coordination. If the applicant passes the test, 4 DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD she returns to the Leadpoint HR department for drug testing and background checks. D. Discipline and Termination Although the Agreement provides that Leadpoint has sole responsibility to counsel, discipline, review, evalu- ate, and terminate personnel who are assigned to BFI, it also grants BFI the authority to “reject any Personnel, and . . . discontinue the use of any personnel for any or no reason.” BFI Managers Keck and Mennie, and Shift Supervi- sors Ortiz and Sutter testified that they have never been involved in any disciplinary decisions for Leadpoint em- ployees. However, the record includes evidence of two incidents where discipline of Leadpoint employees was prompted by BFI action. In a June 2013 email from BFI Operations Manager Keck to Leadpoint CEO Ramirez, Keck stated that he observed two Leadpoint employees passing a pint of whiskey at the jobsite. Keck then con- tacted Leadpoint Manager Haas, who immediately sent the two employees for alcohol and drug screening. Ramirez testified that, in response to Keck’s email “re- quest[ing] [the employees’] immediate dismissal,” Leadpoint investigated the complaint and terminated one employee and reassigned the other. In the same email to Ramirez, Keck indicated that he had observed damage to BFI property, including a pa- perwork drop box that had been destroyed. Keck stated that a surveillance camera recorded a Leadpoint employ- ee punching the box, and that he hoped Ramirez agreed that “this Leadpoint employee should be immediately dismissed.” Haas testified that, pursuant to Keck’s email, he reviewed the video, identified the employee, and Leadpoint terminated the employee after an investiga- tion. Haas stated that BFI was not involved in the inves- tigation of the employee and was not consulted in the decision to terminate him. E. Wages and Benefits The Agreement includes a rate schedule that requires BFI to compensate Leadpoint for each worker’s wage plus a specified percentage mark-up; the mark-up varies based on whether the work is performed during regular hours or as overtime. Although the Agreement provides that Leadpoint “solely determines the pay rates paid to its Personnel,” it may not, without BFI’s approval, “pay a pay rate in excess of the pay rate for full-time employees of [BFI] who perform similar tasks.” Mennie testified that Leadpoint has never made such a request. Leadpoint issues paychecks to employees and maintains their pay- roll records. The record includes a Rate Schedule Addendum be- tween BFI and Leadpoint executed in response to a min- imum wage increase from $8.75 to $10 by the City of San Jose. Pursuant to the Addendum, the parties agreed that BFI would pay a higher hourly rate for the services of Leadpoint employees after the minimum wage in- crease took effect. Leadpoint employees are required to sign a benefits waiver stating they are eligible only for benefits offered by Leadpoint and are not eligible to participate in any benefit plan offered by BFI. Leadpoint provides em- ployees with paid time-off and three paid holidays after they have worked for 2,000 hours, and the option to pur- chase medical, life, and disability insurance. F. Scheduling and Hours BFI establishes the facility’s schedule of working hours. It operates three set shifts on weekdays: 4 a.m.- 1 p.m., 2 p.m.-11:30 p.m., and 10:30 p.m.-7 a.m. Leadpoint is responsible for providing employees to cov- er all three shifts. Although Leadpoint alone schedules which employees will work each shift,13 Leadpoint has no input on shift schedules. Keck testified that any mod- ification in shift times would require modifying the facil- ity’s hours of operation and the work schedules for all BFI employees. BFI will keep a stream running into overtime if it de- termines that the material on a specific stream cannot be processed by the end of a shift. A BFI manager will normally convey this decision to a Leadpoint shift super- visor; Leadpoint, in turn, determines which employees will stay on the stream to complete the overtime work. BFI also dictates when the streams stop running so that Leadpoint employees can take breaks. Keck has in- structed Leadpoint employees to spend 5 minutes gather- ing the debris around their stations before breaking. Alt- hough Keck asserted that this assignment would not af- fect the length of breaks, sorter Andrew Mendez testified that, as a practical matter, the clean-up requirement has cut into employees’ break time. The Agreement requires that Leadpoint employees must, at the end of each week, submit to Leadpoint a summary of their “hours of services rendered.” Employ- ees must obtain the signature of an authorized BFI repre- sentative attesting to the accuracy of the hours on the form. BFI may refuse payment to Leadpoint for any time claimed for which a worker failed to obtain a signa- ture. G. Work Processes BFI determines which material streams will run each day and provides Leadpoint with a target headcount of 13 Leadpoint must also supply housekeepers to work a Saturday shift. BFI NEWBY ISLAND RECYCLERY 5 workers needed. BFI also dictates the number of Leadpoint laborers to be assigned to each material stream, but Leadpoint assigns specific Leadpoint em- ployees to specific posts. The record includes an email from Keck to Haas directing Haas to reduce the number of sorters on a specific line by two per shift. The email detailed what positions sorters should occupy on the stream, what materials should be prioritized, and whether a right-handed or left-handed sorter was preferred.14 The email concluded by stating “[t]his staffing change is ef- fective Monday, August 5, 2013.” Ramirez testified that the sorters occupy set work stations along each stream and that BFI dictates the location of these stations. Dur- ing a shift, BFI might direct Leadpoint supervisors to move employees to another stream in response to pro- cessing demands. Before each shift, BFI’s Shift Supervisors Ortiz and Sutter hold meetings with Leadpoint supervisors-the onsite manager and leads-to present and coordinate the day’s operating plan. During those meetings, BFI’s man- agers dictate which streams will be operating and estab- lish the work priorities for the shift. Ortiz testified that he uses the preshift meeting to advise Leadpoint supervi- sors of the specific tasks that need to be completed dur- ing the shift, i.e. maintenance, quality, and cleaning is- sues. Ortiz indicated that Leadpoint supervisors assign employees so as to accomplish these designated tasks. BFI managers set productivity standards for the mate- rial streams. BFI Division Manager Mennie testified that BFI tracks the tons per hour processed on each stream, the proportion of running time to downtime on each stream, and various quality standards. BFI has sole au- thority to set the speed of the material streams based on its ongoing assessment of the optimal speed at which materials can be sorted most efficiently. If sorters are unable to keep up with the speed of the stream, BFI-but not Leadpoint-can make various adjustments, such as slowing the speed of the stream or changing the angle of the screens. The record indicates that the speed of the streams has been a source of contention between BFI and Leadpoint employees. For instance, former-sorter Clar- ence Harlin described an incident during which BFI Shift Supervisor Sutter stood across the stream from sorters and criticized them for failing to remove a sufficient amount of plastic. Harlin responded that it was not pos- sible to pull that much material unless the stream was slowed down or stopped. Sutter responded by calling the entire line of sorters to the control room, where he di- 14 For instance, the email stated that “[t]wo of your employees should be positioned at the east end of the presorts focusing primarily on glass. Their secondary picks should be plastics into the Recycling Stream drop chute.” rected them to work more efficiently and dismissed their requests to slow down or stop the line. Leadpoint employees are able to stop the streams by hitting an emergency stop switch. Sutter testified that he has instructed Leadpoint supervisors on when it is appro- priate for Leadpoint employees to use the switch. A BFI employee who works in the control room monitors the operating status of the streams and is required to restart a stream after it has been stopped. Sorter Travis Stevens testified that he has been instructed by BFI managers on multiple occasions not to overuse the emergency stop switch. He stated that BFI Operations Manager Keck and BFI Shift Supervisor Ortiz held a meeting with an entire line of Leadpoint employees to call attention to the frequency of their emergency stops and to direct Leadpoint employees to minimize the number of stops to reduce downtime. BFI’s managers testified that when, in the course of monitoring stream operation and productivity, they iden- tify problems, including problems with the job perfor- mance of a Leadpoint employee, they communicate their concerns to a Leadpoint supervisor. The Leadpoint su- pervisor is expected to address those issues with the em- ployees. According to the testimony of Leadpoint em- ployees, BFI managers have, on occasion, addressed them directly regarding job tasks and quality issues. Leadpoint Housekeeper Clarence Harlin testified that he receives work directions from BFI managers and em- ployees at least twice a week. Sorters Mendez and Ste- vens both testified that they have received specific as- signments from BFI managers that took priority over the tasks assigned by their immediate Leadpoint supervisors. Sorter Marivel Mendoza testified that Sutter has directed him to remove more plastic from the stream, and has moved him to other streams where assistance was need- ed. H. Training and Safety When Leadpoint employees begin working at the facil- ity, they receive an orientation and job training from Leadpoint supervisors. Periodically, they also receive substantive training and counseling from BFI managers. For instance, following customer complaints about the quality of BFI’s end product, Keck held two or three educational meetings with Leadpoint employees and su- pervisors who worked on the wet waste stream. During the meetings, Keck highlighted the objectives of the op- eration to make sure that Leadpoint employees under- stood BFI’s goals. He also explained the difference be- tween organic and nonorganic materials and specified which materials should be removed from the line. Keck held a similar meeting with Leadpoint employees who worked on the commercial single stream because he was 6 DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD concerned that sorters were allowing too many materials to pass by on the stream without being sorted.15 With regard to one line, Keck told the sorters that BFI would only be able to cover the labor expenses for the line if the processed material generated revenue for BFI. As noted above, BFI Shift Supervisor Sutter similarly called a meeting with a group of sorters to direct them to work more productively. As to safety, the Agreement mandates that Leadpoint require its employees to comply with BFI’s safety poli- cies, procedures, and training requirements. For all em- ployees working in positions deemed safety-sensitive by BFI, Leadpoint must obtain a written acknowledgement that they have read, understand, and agree to comply with BFI’s safety policy. BFI also “reserves the right to enforce the Safety Policy provided to [Leadpoint] per- sonnel.”16 New Leadpoint employees attend a safety orientation that is presented by Leadpoint managers. The record shows that, on occasion, BFI also provides safety train- ing to Leadpoint employees. I. Other Terms According to the terms of the Agreement, Leadpoint personnel shall not be assigned to BFI for more than 6 months. Ramirez testified that Leadpoint employees have been assigned to BFI for more than 6 months, and BFI has never invoked this provision. The Agreement also allows BFI to examine “[Leadpoint’s] books and records pertaining to the Personnel, [Leadpoint’s] obliga- tions and duties under this Agreement, and all services rendered by [Leadpoint] or the Personnel under this Agreement, at any time for purposes of auditing compli- ance with this Agreement, or otherwise.” Mennie testi- fied that he has never asked to inspect Leadpoint’s per- sonnel files. II. THE REGIONAL DIRECTOR’S FINDINGS The Regional Director, applying TLI, supra, found that BFI is not a joint-employer of the Leadpoint employees because it does not “share or codetermine [with Leadpoint] those matters governing the essential terms and conditions of employment” of the sorters, screen cleaners, or housekeepers. First, the Regional Director found that Leadpoint sets employee pay and is the sole provider of benefits. He acknowledged that, under the Agreement, Leadpoint is prevented from paying employ- 15 Ortiz indicated that he also held educational sessions with Leadpoint employees after he became concerned that sorters were not removing a sufficient amount of contaminants from the stream. 16 Leadpoint employees’ personal protective equipment-a safety vest, a hardhat, safety glasses, ear plugs, and gloves-is provided by Leadpoint and differs from the gear that BFI employees use. ees more than BFI pays employees who perform similar work. But he found that this provision was not indicative of BFI’s control over wages because it limits only em- ployees’ maximum wage rate; it would not prevent Leadpoint from lowering wages or offering more bene- fits. Moreover, he found that the provision only applies to Leadpoint sorters, since BFI does not employ any screen cleaners or housekeepers. Next, the Regional Director found that Leadpoint has sole control over the recruitment, hiring, counseling, dis- cipline, and termination of its employees. He noted that there was no evidence to suggest that BFI participates in any of these decisions. With regard to Keck’s email re- porting the misconduct of Leadpoint employees, the Re- gional Director found that Keck merely requested that the employees be terminated; he did not order or direct Leadpoint to terminate them. He thus concluded that BFI does not possess the authority to terminate Leadpoint employees. Finally, the Regional Director found that BFI does not control or codetermine employees’ daily work. He found that Leadpoint employees were supervised solely by the Leadpoint onsite manager and leads, and that nothing in the record supported the Union’s argument that BFI con- trols employees’ daily work functions. While acknowl- edging BFI’s control over the speed of the material stream, the Regional Director found that BFI does not mandate how many employees work on the line, the speed at which the employees work, where they stand on the stream, or how they pick material off the stream.17 The mere ability to control the speed of the stream, he stated, does not “create a level of control that is suffi- ciently direct or immediate” to warrant a finding of joint control. The Regional Director also stated that if BFI has a problem with a Leadpoint employee, it complains to a Leadpoint supervisor who takes care of the matter using her own discretion. To the extent that BFI has directly instructed Leadpoint employees, he found “the instruc- tion was merely routine in nature and insufficient to war- rant a finding that BFI jointly controls Leadpoint em- ployees’ daily work.” Although BFI sets the work hours and shifts of the facility’s operation, the Regional Direc- tor observed that Leadpoint is solely in control of sched- uling its own employees’ shifts, scheduling employees for overtime, and administering requests for sick leave and vacation. 17 Based on our review of the record, we disagree with the Regional Director’s factual findings that BFI does not mandate how many em- ployees work on the line, the speed at which they work, where they stand, or how they pick material. BFI NEWBY ISLAND RECYCLERY 7 III. POSITIONS OF THE PARTIES AND AMICI A. The Union The Union argues first that, under the Board’s current joint-employer standard, BFI constitutes a joint employer of the Leadpoint employees because it shares or code- termines the following essential terms and conditions of employment: employment qualifications, work hours, breaks, productivity standards, staffing levels, work rules and performance, the speed of the lines, dismissal, and wages. BFI’s direct control over employees is evinced by its regular oversight of the employees and its constant control of their work. BFI, it argues, demands compli- ance with “detailed specifications, including the number of employees on each line, where they stand, what they pick, and at what rate they sort.” BFI also trains and instructs employees as to how to do their jobs, directing them on picking techniques, what to prioritize, how to clear jams, and when to use the emergency stop. Alternatively, the Union contends that the Board should adopt a broader standard to better effectuate the purpose of the Act and respond to industrial realities. The Union states that the Board’s current emphasis on whether an employer exercises direct and immediate control over employees conflicts with the language and purpose of the Act, which is focused on ensuring em- ployees’ bargaining rights to the fullest extent. Further, the Union argues that the Board must consider all indicia of control in its joint-employer analysis, rather than the narrow subset of criteria set forth in TLI, supra, 271 NLRB at 798 (hiring, firing, discipline, supervision, and direction). It observes that “a myriad of other essential terms that are mandatory subjects of bargaining may [] also be pertinent to the employees involved.” Based on these concerns, the Union recommends that the Board find joint-employer status where an employer “possesses sufficient authority over the employees or their employer such that its participation is a requisite to meaningful collective bargaining. Such authority can be either direct or indirect.” Finally, the Union asserts that absent a change in the joint-employer standard, a putative employer, like BFI, that is a necessary party to meaningful collective bar- gaining will continue to insulate itself by the “calculated restructuring of employment and insertion of a contractor to insulate itself from the basic legal obligation to recog- nize and bargain with the employees’ representative.” B. BFI and Leadpoint BFI argues that, under the Board’s current joint- employer test, the Regional Director correctly found that BFI is not a joint employer of Leadpoint’s employees. To this end, BFI contends that the Regional Director properly concluded that Leadpoint has sole authority to hire, fire, discipline, supervise, direct, assign, train, and schedule its employees. It further contends that the Un- ion points to only a handful of instances in which BFI managers gave routine instructions to Leadpoint employ- ees, evidence that falls far short of establishing that BFI exerted any meaningful control over them. Although BFI’s physical plant dictates where Leadpoint employees must work, BFI does not decide where particular em- ployees work. Likewise, despite the fact that BFI man- agers meet with Leadpoint supervisors daily to discuss operations, Leadpoint supervisors are solely responsible for controlling and directing their employees. Finally, contrary to the Union, meaningful control cannot be es- tablished by a contractual right or its occasional exercise; instead the Board properly looks to the actual practice of the parties. BFI also urges the Board not to modify its joint- employer standard. It contends that the Union has not presented any compelling reason to revisit Board policy. Any modification, it argues, would undermine the pre- dictability of the law in this area, which the Board has applied uniformly for over 30 years. The Union’s pro- posed standard, in its view, imposes “no meaningful limit on who could be deemed a joint employer of another’s workers.” Thus, a regional director “would be free to exercise her substantial discretion to determine that com- pletely separate companies constituted a joint employer simply because she believes that bargaining would be more effective if both companies were at the table.” Leadpoint echoes the arguments presented by BFI: that Leadpoint is the sole employer of its employees, and that the Board should not modify its joint-employer standard. In support of the current standard, Leadpoint contends that it is a clear and understandable approach that has not proven overly onerous for parties seeking to establish a joint-employer relationship. Leadpoint argues that the “vague and ambiguous” standard proposed by the Union lacks clarity and provides minimal, if any, guidance as to what factors are significant for evaluating joint-employer status. C. The General Counsel The General Counsel urges the Board to abandon its existing joint-employer standard because it “undermines the fundamental policy of the Act to encourage stable and meaningful collective bargaining.”18 The Board, since TLI, supra, has significantly narrowed its approach by (a) requiring evidence of direct and immediate control over employees; (b) looking only to the actual practice of 18 The General Counsel’s brief takes no position on the merits of this representation proceeding. 8 DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD the parties rather than their contract; and (c) requiring an employer’s control to be substantial and not “limited and routine.” He posits that this approach is not consistent with the Act, which broadly defines the term “employer.” Moreover, the contingent work force has grown signifi- cantly over the past several decades. The General Coun- sel submits that in many contingent arrangements, the user firm only has limited and routine supervision over employees, and indirect or potential control over terms and conditions of employment. Nonetheless, the user firm can influence the supplier firm’s bargaining posture by threatening to terminate its contract with the supplier if wages and benefits rise above a set cost threshold. The General Counsel recommends that the Board find joint-employer status where an employer “wields suffi- cient influence over the working conditions of the other entity’s employees such that meaningful bargaining could not occur in its absence.” Such an approach would make no distinction between direct, indirect, and poten- tial control, and would find joint-employer status where industrial realities make an entity essential for bargain- ing. D. Other Amici Amici in support of the Union uniformly urge the Board to adopt a more inclusive joint-employer standard that would give dispositive weight to more forms of em- ployer control. Specifically, they urge the Board to abandon its recent focus on direct and immediate control and consider instead the totality of a putative employer’s influence over employees’ working conditions, including control that is exercised indirectly or reserved via con- tractual right. They also argue that the Board should evaluate a putative employer’s control over a broad range of terms and conditions of employment rather than the limited set of factors enumerated in TLI, supra. In urging the Board to modify its approach, many amici note that that the number of contingent employment relationships has grown significantly in recent years, and that a sizea- ble proportion of the labor force now works for staffing agencies. They posit that the Board’s current narrow focus on direct control absolves many user employers of bargaining responsibilities under the Act despite the fact that their participation is required for meaningful bar- gaining to occur. Amici in support of BFI uniformly contend that BFI is not a joint-employer of Leadpoint’s employees, and urge the Board not to modify its existing approach. They ar- gue primarily that the Board’s standard-which has been applied consistently for over 30 years-has provided employers with stability and predictability in entering into labor supply arrangements in response to fluctuating market needs. Any change, they contend, would destabi- lize these relationships and undermine the expectations of the contracting parties. A more inclusive standard, they argue, would also widen the scope of labor disputes and force firms to participate in bargaining even where they have no authority to set or control terms and condi- tions of employment. Some amici contend that a broader standard could potentially include-and consequently disrupt-any contractual relationship involving labor. Other amici argue that a broader standard would expose employers to unwarranted liability for unfair labor prac- tices committed by the other firm. Some argue too that the common law of agency prohibits the Board from adopting an open-ended approach that considers all of the economic realities of the parties’ relationship. IV. THE EVOLUTION OF THE BOARD’S JOINT-EMPLOYER STANDARD In analyzing the joint-employer issue, and evaluating the various arguments raised by the parties and amici, it is instructive to review the development of the Board’s law in this area. Three aspects of that development seem clear. First, the Board’s approach has been consistent with the common-law concept of control, within the framework of the National Labor Relations Act. Second, before the current joint-employer standard was adopted, the Board (with judicial approval) generally took a broader approach to the concept of control. Third, the Board has never offered a clear and comprehensive ex- planation for its joint-employer standard, either when it adopted the current restrictive test or in the decades be- fore. The core of the joint-employer standard, which we preserve today, can be traced at least as far back as the Greyhound case, a representation proceeding that in- volved a company operating a bus terminal and its clean- ing contractor. There, the Board in 1965 found two stat- utory employers to be joint employers of certain workers because they “share[d], or codetermine[d], those matters governing essential terms and conditions of employ- ment.”19 Significantly, at an earlier stage of that case, the Supreme Court explained the issue presented-whether Greyhound “possessed sufficient control over the work of the employees to qualify as a joint employer with” the cleaning contractor-was “essentially a factual issue” for the Board to determine.20 19 Greyhound Corp., 153 NLRB 1488, 1495 (1965), enfd. 368 F.2d 778 (5th Cir. 1966). See also Franklin Simon & Co., Inc., 94 NLRB 576, 579 (1951) (finding joint-employer status where “a substantial right of control over matters fundamental to the employment relation- ship [was] retained and exercised” by both department store and com- pany operating shoe department). 20 Boire v. Greyhound Corp., 376 U.S. 473, 481 (1964). The Su- preme Court reversed a district court injunction against the Board pro- BFI NEWBY ISLAND RECYCLERY 9 During the period after Greyhound but before the Third Circuit’s 1982 decision in Browning-Ferris Indus- tries of Pennsylvania, supra, some (though certainly not all) of the Board’s joint-employer decisions used the “share or co-determine” formulation.21 But regardless of the wording used, the Board typically treated the right to control the work of employees and their terms of em- ployment as probative of joint-employer status. The Board did not require that this right be exercised, or that it be exercised in any particular manner. Thus, the Board’s joint-employer decisions found it probative that employers retained the contractual power to reject or terminate workers;22 set wage rates;23 set working hours;24 approve overtime;25 dictate the number of work- ers to be supplied;26 determine “the manner and method of work performance”;27 “inspect and approve work,”28 and terminate the contractual agreement itself at will.29 The Board stressed that “the power to control is present by virtue of the operating agreement.”30 Reviewing courts expressly endorsed this approach.31 ceeding, rejecting Greyhound’s argument that the Board was acting in excess of its powers under the Act, given the exclusion of independent contractors from the statutory definition of “employee.” 21 See, e.g., C.R. Adams Trucking, Inc., 262 NLRB 563, 566 (1982), enfd. 718 F.2d 869 (8th Cir. 1983); Springfield Retirement Residence, 235 NLRB 884, 891 (1978); Greenhoot, Inc., 205 NLRB 250, 251 (1973). 22 See Ref-Chem Co., 169 NLRB 376, 379 (1968), enf. denied on other grounds 418 F.2d 127 (5th Cir. 1969); Jewel Tea Co., 162 NLRB 508, 510 (1966). 23 See Ref-Chem, supra, 169 NLRB at 379; Harvey Aluminum, 147 NLRB 1287, 1289 (1964). 24 See Jewel Tea, supra, 162 NLRB at 510; Mobil Oil Corp., 219 NLRB 511, 516 (1975), enf. denied on other grounds sub nom. Alaska Roughnecks and Drillers Assn. v. NLRB, 555 F.2d 732 (9th Cir. 1977), cert. denied 43 U.S. 1069 (1978). 25 Ref-Chem Co. v. NLRB, 418 F.2d 127, 129 (5th Cir. 1969). 26 See Harvey Aluminum, supra, 147 NLRB at 1289; Mobil Oil, su- pra, 219 NLRB at 516. 27 Value Village, 161 NLRB 603, 607 (1966). 28 Ref-Chem Co. v. NLRB, supra, 418 F.2d at 129. 29 Value Village, supra, 161 NLRB at 607; Mobil Oil, supra, 219 NLRB at 516. 30 Value Village, supra, 161 NLRB at 607. See also Jewel Tea, su- pra, 162 NLRB at 510 (“That the licensor has not exercised such power is not material, for an operative legal predicate for establishing a joint- employer relationship is a reserved right in the licensor to exercise such control”); Lowery Trucking Co., 177 NLRB 13, 15 (1969), enfd. sub nom. Ace-Alkire Freight Lines v. NLRB, 431 F.2d 280 (8th Cir. 1970) (observing that “[w]hile [putative employer] never rejected a driver hired by [supplier], it had the right to do so”). 31 See Ref-Chem Co. v. NLRB, supra, 418 F.2d at 129 (affirming the Board’s joint-employer finding where “[t]he terms of the agreements with these two companies gave [putative employer] the right to approve employees, control the number of employees, have an employee re- moved, inspect and approve work, pass on changes in pay and overtime allowed”). See also Ace-Alkire Freight Lines, Inc. v. NLRB, 431 F.2d 280, 282 (8th Cir. 1970) (same where putative employer “retained the right to reject drivers sent to them”); Carrier Corp. v. NLRB, 768 F.2d In addition to recognizing the right to control as proba- tive, the Board gave weight to a putative joint employer’s “indirect” exercise of control over workers’ terms and conditions of employment.32 In so doing, the Board em- phasized that, in order to exercise significant control, a putative employer need not “hover over [workers], di- recting each turn of their screwdrivers and each connec- tion that they made.”33 Instead, the Board assessed whether a putative employer exercised “ultimate control” over their employment.34 Consistent with this principle, the Board in certain cas- es found evidence of joint-employer status where a puta- tive employer, although not responsible for directly su- pervising another firm’s employees, inspected their work, issued work directives through the other firm’s supervisors, and exercised its authority to open and close the plant based on production needs.35 Likewise, the Board found significant indicia of control where a puta- tive employer, although it “did not exercise direct super- visory authority over” the workers at issue, nonetheless held “day-to-day responsibility for the overall opera- tions” of the worksite and determined the scope and na- ture of the contractors’ work assignments.36 Contractual arrangements under which the user employer reimbursed the supplier for workers’ wages or imposed limits on wages were also viewed as tending to show joint- employer status.37 The Third Circuit’s Browning-Ferris decision did not question, much less reject, any of these lines of Board precedent. That decision, rather, carefully untangled the 778, 781 (6th Cir. 1985) (same where, under parties’ agreement, puta- tive employer “had the authority to reject any driver that did not meet its standards and it could also direct [supplier firm] to remove any driver”). 32 Floyd Epperson, 202 NLRB 23, 23 (1973), enfd. 491 F.2d 1390 (6th Cir. 1974). 33 Sun-Maid Growers of California, 239 NLRB 346, 351 (1978), enfd. 618 F.2d 56 (9th Cir. 1980) (finding joint-employer status). 34 Int’l Trailer Co., 133 NLRB 1527, 1529 (1961), enfd. sub nom. NLRB v. Gibraltar Industries, 307 F.2d 428 (1962) (finding joint- employer status), cert. denied 372 U.S. 911 (1963). 35 Id. See also Hamburg Industries, 193 NLRB 67, 67 (1971) (find- ing joint-employer status where putative employer’s superintendents checked the performance of supplier’s workers and the quality of their work, and communicated work directions via supplier’s supervisors). 36 Clayton B. Metcalf, 223 NLRB 642, 643 (1976). 37 See Hamburg Industries, supra, 193 NLRB at 67-68 (assigning weight to putative employer’s “indirect control over wages” via cost- plus arrangement); Hoskins Ready-Mix, 161 NLRB 1492, 1493 (1966) (same, noting that user employer would be the “ultimate source of any wage increases” for workers); Ref-Chem Co., supra, 169 NLRB at 379 (supplier could not make any wage modification without securing ap- proval of the user). See also Industrial Personnel Corp. v. NLRB, 657 F.2d 226, 229 (8th Cir. 1981) (relying on the Board’s finding that user employer reimbursed supplier for employees’ wages), cert. denied 454 U.S. 1148 (1982). 10 DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD joint-employer doctrine from the distinct single- employer doctrine (which addresses integrated enterpris- es only nominally separate), endorsed the Board’s “share or codetermine” formulation, and enforced the Board’s order finding joint-employer status. The Third Circuit explained: The basis of the [joint employer] finding is simply that one employer while contracting in good faith with an otherwise independent company, has retained for itself sufficient control of the terms and conditions of em- ployment of the employees who are employed by the other employer. . . .Thus, the “joint employer” concept recognizes that the business entities involved are in fact separate but that they share or codetermine those mat- ters governing the essential terms and conditions of employment. 691 F.2d at 1123 (citations omitted; emphasis added). The Board subsequently embraced the Third Circuit’s decision, but simultaneously took Board law in a new and different direction. Laerco and TLI, both decided in 1984, marked the beginning of a 30-year period during which the Board-without any explanation or even acknowledgement and without overruling a single prior decision-imposed additional requirements that effec- tively narrowed the joint-employer standard. Most sig- nificantly, the Board’s decisions have implicitly repudi- ated its earlier reliance on reserved control and indirect control as indicia of joint-employer status. The Board has foreclosed consideration of a putative employer’s right to control workers, and has instead focused exclu- sively on its actual exercise of that control-and required its exercise to be direct, immediate, and not “limited and routine.”38 The Board has thus refused to assign any significance to contractual language expressly giving a putative em- ployer the power to dictate workers’ terms and condi- tions of employment. In TLI, for instance, the parties’ contract provided, among other things, that the user em- ployer “at all times will solely and exclusively be re- sponsible for maintaining operational control, direction and supervision over said drivers”.39 Although prior precedent found this type of contractual authority proba- tive of joint employer status, the TLI Board found it ir- relevant, absent evidence that the putative employer “af- fect[ed] the terms and conditions of employment to such a degree that it may be deemed a joint employer.”40 The 38 AM Property Holding Corp., 350 NLRB 998, 1001 (2007), enfd. in relevant part sub nom. Service Employees Int’l Union, Local 32BJ v. NLRB, 647 F.3d 435 (2d. Cir. 2011) 39 TLI, supra, 271 NLRB at 803. 40 Id. at 799. Board later emphasized this narrowed approach in AM Property Holding Corp., a 2007 decision, supra, where it stated that “[i]n assessing whether a joint employer rela- tionship exists, the Board does not rely merely on the existence of such contractual provisions, but rather looks to the actual practice of the parties.”41 In Airborne Express,42 a 2002 decision, the Board held that “[t]he essential element in [the joint-employer] anal- ysis is whether a putative joint employer’s control over employment matters is direct and immediate.”43 This restrictive approach has resulted in findings that an entity is not a joint employer even where it indirectly exercised control that significantly affected employees’ terms and conditions of employment. For example, the Board re- fused to find that a building management company that utilized employees supplied by a janitorial company was a joint employer notwithstanding evidence that the user dictated the number of workers to be employed, commu- nicated specific work assignments and directives to the supplier’s manager, and exercised ongoing oversight as to whether job tasks were performed properly.44 Like- wise, the Board has found, contrary to its earlier ap- proach, that cost-plus arrangements between the employ- ing parties are not probative of joint-employer status.45 Even where a putative joint employer has exercised di- rect control over employees, the Board has given no weight to various forms of supervision deemed “limited and routine.” In TLI, for instance, the user employer in- structed contract drivers as to which deliveries were to be made on a given day, filed incident reports with the sup- plier when drivers engaged in conduct adverse to its op- eration, received accident reports, and maintained driver logs and records.46 Nonetheless, the Board concluded that “the supervision and direction exercised by [the us- 41 350 NLRB at 1000. The AM Property Board refused to give weight to a contractual provision requiring that the supplier plan, or- ganize, and coordinate its operations “in conjunction with the direc- tions, requests and suggestions” of the user’s management, and that all new hires were subject to the initial approval of the user. Id. at 1019. 42 338 NLRB 597, 597 fn. 1 (2002). 43 The Board in Airborne Express added this element in a footnote without any explanation; it cited only TLI as support. But the TLI Board did not use the phrase “direct and immediate control,” let alone identify that concept as the “essential element” in the Board’s test. The Air- borne Express majority also asserted that the Board in TLI “abandoned its previous test in this area, which had focused on a putative joint employer’s indirect control over matters relating to the employment relationship.” 338 NLRB at 597 fn. 1. But TLI did not, in fact, purport to overrule any precedent or alter the Board’s approach. 44 Southern California Gas Co., 302 NLRB 456, 461-462 (1991). 45 See Goodyear Tire and Rubber Co., 312 NLRB 674, 677-678 (1993) (rejecting the argument that participation in a cost-plus contract represented a form of codetermination). 46 271 NLRB at 799. BFI NEWBY ISLAND RECYCLERY 11 er] on a day-to-day basis is both limited and routine.”47 The Board elaborated on this concept in AM Property, supra, where it stated that “[t]he Board has generally found supervision to be limited and routine where a su- pervisor’s instructions consist primarily of telling em- ployees what work to perform, or where and when to perform the work, but not how to perform the work.”48 There, the Board found that the user’s oversight of a supplier’s cleaning employees was “limited and routine” where the user distributed supplies to workers, prepared their timecards, ensured that their work was done proper- ly, and occasionally assigned work.49 V. REVISITING THE JOINT-EMPLOYER STANDARD As the Board’s view of what constitutes joint employ- ment under the Act has narrowed, the diversity of work- place arrangements in today’s economy has significantly expanded. The procurement of employees through staff- ing and subcontracting arrangements, or contingent em- ployment, has increased steadily since TLI was decided.50 The most recent Bureau of Labor Statistics survey from 2005 indicated that contingent workers accounted for as much as 4.1 percent of all employment, or 5.7 million workers.51 Employment in the temporary help services industry, a subset of contingent work, grew from 1.1 mil- lion to 2.3 million workers from 1990 to 2008.52 As of August 2014, the number of workers employed through temporary agencies had climbed to a new high of 2.87 million, a 2 percent share of the nation’s work force.53 Over the same period, temporary employment also ex- panded into a much wider range of occupations.54 A re- cent report projects that the number of jobs in the em- 47 Id. The Board also discounted the user’s role in influencing bar- gaining where user attended the supplier’s collective bargaining negoti- ations and explained that the contract was in jeopardy if the supplier failed to achieve cost savings. 271 NLRB at 798-799. 48 350 NLRB at 1001. See also Flagstaff Medical Center, 357 NLRB No. 65 slip op. at 9 (2011), enfd. in part 715 F.3d 928 (D.C. Cir. 2013). 49 350 NLRB at 1001. 50 The Board previously recognized the “ongoing changes in the American work force and workplace and the growth of joint employer arrangements, including the increased use of companies that specialize in supplying ‘temporary’ and ‘contract workers’ to augment the work- forces of traditional employers.” M. B. Sturgis, Inc., 331 NLRB 1298, 1298 (2000). 51 Bureau of Labor Statistics, U.S. Department of Labor, “Contin- gent and Alternative Employment Arrangements, February 2005,” (July 27, 2005). 52 See Tian Luo, et al., “The Expanding Role of Temporary Help Services from 1990 to 2008,” Monthly Labor Review, Bureau of Labor Statistics, August 2010 at 12. 53 Steven Greenhouse, “The Changing Face of Temporary Employ- ment,” NY Times website, August, N.Y. TIMES, Aug. 31, 2014, at http://www.nytimes.com/2014/09/01/upshot/the-changing-face-of- temporary-employment.html 54 See Luo et al., supra at 5. ployment services industry, which includes employment placement agencies and temporary help services, will increase to almost 4 million by 2022, making it “one of the largest and fastest growing [industries] in terms of employment.”55 This development is reason enough to revisit the Board’s current joint-employer standard. “[T]he primary function and responsibility of the Board . . . is that ‘of applying the general provisions of the Act to the com- plexities of industrial life.’”56 If the current joint- employer standard is narrower than statutorily necessary, and if joint-employment arrangements are increasing, the risk is increased that the Board is failing in what the Su- preme Court has described as the Board’s “responsibility to adapt the Act to the changing patterns of industrial life.”57 As we have seen, however, the Board has never clearly and comprehensively explained its joint-employer doctrine or, in particular, the shift in approach reflected in the current standard.58 Our decision today is intended to address this shortcoming. For the reasons that follow, we are persuaded that the current joint-employer standard is not mandated by the Act and that it does not best serve the Act’s policies. We begin with the obvious proposition that in order to find that a statutory employer (i.e., an employer subject to the National Labor Relations Act) has a duty to bar- gain with a union representing a particular group of statutory employees, the Act requires the existence of an employment relationship between the employer and the employees. Section 2(3) of the Act provides that the “term ‘employee’ . . . shall not be limited to the employ- ees of a particular employer, unless the Act explicitly states otherwise.”59 Section 9(c) authorizes the Board to process a representation petition when it alleges that “employees . . . wish to be represented for collective bar- gaining . . . and their employer declines to recognize their representative.”60 Section 8(a)(5), in turn, makes it an unfair labor practice for an employer “to refuse to 55 Richard Henderson, “Industry Employment and Output Projec- tions to 2022,” Monthly Labor Review, Bureau of Labor Statistics, December 2013. 56 Ford Motor Co. v. NLRB, 441 U.S. 488, 496 (1979), quoting NLRB v. Insurance Agents, 361 U.S. 477, 499 (1960); NLRB v. Erie Resistor Corp., 373 U.S. 221, 236 (1963); and NLRB v. Steelworkers, 357 U.S. 357, 362-363 (1958). 57 See NLRB v. J. Weingarten, Inc., 420 U.S. 251, 266 (1975). 58 It is well established that even when an agency is creating policies to fill a gap in an ambiguous statute, the agency has a responsibility to explain its failure to follow established precedent. Atchison, T. & S.F. Ry. v. Wichita Bd. of Trade, 412 U.S. 800, 807-809 (1973). 59 29 U.S.C. §152(3) (emphasis added). 60 29 U.S.C. §159(c) (emphasis added). 12 DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD bargain collectively with the representatives of his em- ployees.”61 In determining whether an employment relationship exists for purposes of the Act, the Board must follow the common-law agency test. The Supreme Court has made this clear in connection with Section 2(3) of the Act and its exclusion of “any individual having the status of an independent contractor” from the Act’s otherwise broad definition of statutory employees.62 In determining whether a common-law employment relationship exists in cases arising under Federal statutes like the Act, the Court has regularly looked to the Restatement (Second) of Agency (1958) for guidance.63 Section 220(1) of the Restatement (Second) provides that a “servant is a person employed to perform services in the affairs of another and who with respect to the physical conduct in the per- formance of the services is subject to the other’s control or right to control.” The Board’s joint-employer doctrine is best under- stood as always having incorporated the common-law concept of control-as the Supreme Court’s one decision involving the doctrine confirms. In the Greyhound case, as we have seen, the Court framed the issue presented as whether one statutory employer “possessed sufficient control over the work of the employees to qualify as a joint employer with” another statutory employer.64 Thus, the Board properly considers the existence, extent, and object of the putative joint employer’s control, in the context of examining the factors relevant to determining the existence of an employment relationship.65 Accord- 61 29 U.S.C. §158(a)(5) (emphasis added). 62 See NLRB v. United Insurance Co. of America 390 U.S. 254, 256-258 (1968). See also FedEx Home Delivery, 361 NLRB No. 55, slip op. at 1-2 (2014) (reviewing Supreme Court’s application of com- mon-law test in independent-contractor cases arising under Federal statutes). See also NLRB v. Town & Country Electric, Inc., 516 U.S. 85, 92-95 (1995) (where Congress has used the term “employee” in a statute without clearly defining it, the Court assumes that Congress “intended to describe the conventional master-servant relationship as understood by common-law agency doctrine”); Community for Crea- tive Non-Violence v. Reid, 490 U.S. 730, 739-740 (1989) (same). 63 See, e.g., Nationwide Mutual Insurance Co. v. Darden, 503 U.S. 318, 323-324 (1992) (interpreting Employee Retirement Income Secu- rity Act). See also Restatement (Second) of Agency §220, comment g (“Under the existing regulations and decisions involving the Federal Labor Relations Act, there is little, if any, distinction between employ- ee and servant as here used.”). 64 Boire v. Greyhound Corp., supra, 376 U.S. at 481. 65 See generally Vizcaino v. U.S. District Court of the Western Dis- trict of Washington, 173 F.3d 713, 723 (9th Cir. 1999) (describing Restatement (Second) Sec. 220 factors as “useful” in determining whether common-law employment relationship existed between worker and client firm of temporary employment agency for purposes of ERISA). Section 220(2) of the Restatement (Second) provides that: ingly, mere “service under an agreement to accomplish results or to use care and skill in accomplishing results” is not evidence of an employment, or joint-employment, relationship.66 Deciding the joint-employer issue under common-law principles is not always a simple task, just as distinguish- ing between employees and independent contractors in the common law can be challenging (as the Supreme Court has recognized).67 In cases where the common law would not permit the Board to find joint-employer status, we do not believe the Board is free to do so. Even where the common law does permit the Board to find joint- employer status in a particular case, the Board must de- termine whether it would serve the purposes of the Act to do so, taking into account the Act’s paramount policy to “encourage[] the practice and procedure of collective bargaining” (in the words of Section 1). In other words, the existence of a common-law employment relationship is necessary, but not sufficient, to find joint-employer status.68 As the Supreme Court has explained, “[o]ne of In determining whether one acting for another is a servant or an inde- pendent contractor, the following matters of fact, among others, are considered: (a) the extent of control which, by the agreement, the master may ex- ercise over the details of the work; (b) whether or not the one employed is engaged in a distinct occupa- tion or business; (c) the kind of occupation, with reference to whether, in the locality, the work is usually done under the direction of the employer or by a specialist without supervision; (d) the skill required in the particular occupation; (e) whether the employer or the workman supplies the instrumentali- ties, tools, and the place of work for the person doing the work; (f) the length of time for which the person is employed; (g) the method of payment, whether by the time or by the job; (h) whether or not the work is a part of the regular business of the em- ployer; (i) whether or not the parties believe they are creating the relation of master and servant; and (j) whether the principal is or is not in business. 66 Restatement (Second) of Agency §220, comment e (addressing distinction between employees and independent contractors). 67 United Insurance, supra, 390 U.S. at 258 (noting the “innumera- ble situations which arise in the common law where it is difficult to say whether a particular individual is an employee or an independent con- tractor”). See also Restatement (Second) of Agency §220, comment c (“The relation of master and servant is one not capable of exact defini- tion. . . . [I]t is for the triers of fact to determine whether or not there is a sufficient group of favorable factors to establish the relation.”). 68 The General Counsel urges the Board to find joint-employer sta- tus: where, under the totality of the circumstances, including the way the separate entities have structured their commercial relationship, the pu- tative joint employer wields sufficient influence over the working conditions of the other entity’s employees such that meaningful col- lective bargaining could not occur in its absence. Under this approach, the Board would return to its traditional standard and would make no BFI NEWBY ISLAND RECYCLERY 13 the primary purposes of the Act is to promote the peace- ful settlement of industrial disputes by subjecting labor- management controversies to the mediatory influence of negotiation.”69 To best promote this policy, our joint- employer standard-to the extent permitted by the com- mon law-should encompass the full range of employ- ment relationships wherein meaningful collective bar- gaining is, in fact, possible.70 The core of the Board’s current joint-employer stand- ard-with its focus on whether the putative joint em- ployer “share(s) or codetermine(s) those matters govern- ing the essential terms and conditions of employment”- is firmly grounded in the concept of control that is cen- tral to the common-law definition of an employment re- distinction between direct, indirect, and potential control over working conditions and would find joint employer status where “industrial real- ities” make an entity essential for meaningful bargaining. Amicus Brief of the General Counsel at 17. We decline to adopt this test insofar as it might suggest that the applicable inquiry is based on “industrial realities” rather than the common law. To be sure, however, we agree with the General Counsel that “direct, indirect, and potential control over work- ing conditions”-at least as we have explained those concepts here- are all relevant to the joint-employer inquiry. We also agree with the General Counsel that the “way the separate entities have structured their commercial relationship” is relevant to the joint-employer inquiry. Its relevance depends on whether the entities’ relationship tends to show that the putative joint employer controls, or has the right to control-in the common-law sense-employees’ essen- tial terms and conditions of employment. “Sufficient influence” is not enough, however, if it does not amount to control. As explained, we will not find joint-employer status where a puta- tive joint-employer-despite the existence of a common-law employ- ment relationship-could not engage in meaningful collective bargain- ing. But we reject any suggestion that such status should be found only where meaningful collective bargaining over employees’ terms and conditions could not occur without the participation of the putative joint employer. Where two entities “share or codetermine those matters governing the essential terms and conditions of employment,” they are both joint employers-regardless of whether collective bargaining with one entity alone might still be regarded as meaningful, notwithstanding that certain terms and conditions controlled only by the other entity would be excluded from bargaining. 69 Fibreboard Corp. v. NLRB, 379 U.S. 203, 211 (1964). 70 See Management Training Corp., 317 NLRB 1355, 1357 (1995) (recognizing, with regard to employers with close ties to government entities, that an employer may engage in meaningful bargaining with employees even where it does not exercise control over the full range of economic issues). Our dissenting colleagues cite Management Training for the propo- sition that the bargaining obligation should be limited to the employees’ most proximate employer because “employees and their exclusive bargaining representatives can still engage in meaningful bargaining under the Act even with an employer who lacks control over a substan- tial number of essential terms of employment.” But the Board approved of such limited bargaining in Management Training only because some terms of employment were controlled by a government entity that was outside of the Board’s jurisdiction. No such obstacle to bargaining exists here. Moreover, the thrust of Management Training was that an employer subject to the Act is required to bargain over the significant terms of employment that it does control. lationship. The Act surely permits the Board to adopt that formulation. No federal court has suggested otherwise, and the Third Circuit in Browning-Ferris, of course, has endorsed this aspect of the standard. The Board’s post-Browning-Ferris narrowing of the joint-employer standard, however, has a much weaker footing. The Board has never looked to the common law to justify the requirements that a putative joint employ- er’s control be exercised and that the exercise be direct and immediate, not “limited and routine.” This aspect of the current standard is not, in fact, compelled by the common law-and, indeed, seems inconsistent with common-law principles. Because the Board thus is not obligated to adhere to the current standard, we must ask whether there are compelling policy reasons for doing so. The Board’s prior decisions failed to offer any policy rationale at all, and we are not persuaded that there is a sound one, given the clear goals of the Act. Under common-law principles, the right to control is probative of an employment relationship-whether or not that right is exercised. Sections 2(2) and 220(1) of the Restatement (Second) of Agency make this plain, in referring to a master as someone who “controls or has the right to control” another and to a servant as “subject to the [employer’s] control or right to control” (emphasis added). In setting forth the test for distinguishing be- tween employees and independent contractors, Restate- ment (Second) Section 220(2), considers (among other factors) the “extent of control which, by the agreement, the master may exercise over the details of the work” (emphasis added). The Board’s joint-employer decisions requiring the exercise of control impermissibly ignore this principle. Nothing about the joint-employer context suggests that the principle should not apply in cases like this one. In- deed, the Supreme Court’s decision in Greyhound, supra, was entirely consistent with the Restatement (Second) when it described the issue as whether one firm “pos- sessed [not exercised] sufficient control over the work of the employees to qualify as a joint employer.”71 Where a user employer reserves a contractual right (emphasis added) to set a specific term or condition of employment for a supplier employer’s workers, it retains the ultimate authority to ensure that the term in question is adminis- tered in accordance with its preferences. Even where it appears that the user, in practice, has ceded administra- tion of a term to the supplier, the user can still compel the supplier to conform to its expectations. In such a case, a supplier’s apparently independent control over hiring, discipline, and work direction is actually exer- 71 Boire v. Greyhound Corp., supra, 376 U.S. at 481. 14 DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD cised subject to the user’s control. If the supplier does not exercise its discretion in conformance with the user’s requirements, the user may at any time exercise its con- tractual right and intervene. Where a user has reserved authority, we assume that it has rationally chosen to do so, in its own interest. There is no unfairness, then, in holding that legal consequences may follow from this choice.72 Just as the common law does not require that control must be exercised in order to establish an employment relationship, neither does it require that control (when it is exercised) must be exercised directly and immediately, and not in a limited and routine manner (as the Board’s current joint-employer standard demands). Comment d (“Control or right to control”) to Section 220(1) of the Restatement (Second) observes that “the control or right to control needed to establish the relation of master and servant may be very attenuated.”73 The common law, indeed, recognizes that control may be indirect. For ex- ample, the Restatement of Agency (Second) §220, com- ment l (“Control of the premises”) observes that [i]f the work is done upon the premises of the employer with his machinery by workmen who agree to obey general rules for the regulation of the conduct of em- ployees, the inference is strong that such workmen are the servants of the owner... and illustrates this principle by citing the example of a coal mine owner employing miners who, in turn, supply their own helpers. Both the miners and their helpers are servants of the mine owner.74 As the illustration demonstrates, the common law’s “subservant” doctrine addresses situations in which one employer’s control is or may be exercised indi- rectly, where a second employer directly controls the em- 72 The dissent observes that the Board has assigned probative weight only to evidence of actual authority or control in its assessment of various statutory exclusions, including independent contractors and supervisors. But the guiding policy in those areas, as here, is to ensure that statutory coverage is fully effectuated. See FedEx Home Delivery, 361 NLRB No. 55, slip. op. at 9 (2014), quoting Holly Farms Corp. v. NLRB, 517 U.S. 392, 399 (1996), (“[A]dministrators and reviewing courts must take care to assure that exemptions from NLRA coverage are not so expansively interpreted as to deny protection to workers the Act was designed to reach.”). To recognize the significance of the right to control in the joint employment context, in which two putative em- ployers are involved, both serves that policy and is consistent with the common law. 73 “[I]t is not so much the actual exercise of controls as possession of the right to control which is determinative. In other words, ‘subject to the control of the master’ does not mean that the master must stand over the servant and constantly give directions.” The Law of Agency and Partnership Sec. 50 (2nd ed. 1990). 74 See also Restatement (Second) of Agency, Sec. 5, comments e & f, & illustration 6 (discussing subservant relationship between mine owner and miner’s helper). ployee.75 The Federal courts have applied the “subservant” doctrine in cases under Federal statutes that incorporate the common-law standard for determining an employment rela- tionship76-including the National Labor Relations Act.77 The most recent authoritative effort to restate the common law related to employment is consistent with traditional doctrine and similarly makes clear that direct and immediate control is not required.78 In this respect, too, nothing supports the view that common-law principles can or should be ignored in the Board’s joint-employer doctrine. Board case law sug- gests that in many contingent arrangements, control over employees is bifurcated between employing firms with each exercising authority over a different facet of deci- sion making. Where the user firm owns and controls the premises, dictates the essential nature of the job, and imposes the broad, operational contours of the work, and the supplier firm, pursuant to the user’s guidance, makes specific personnel decisions and administers job perfor- mance on a day-to-day basis, employees’ working condi- tions are a byproduct of two layers of control. The 75 See Restatement (Second) of Agency, Sec. 5 (“Subagents and Subservants”) (1958); Warren A. Seavey, Subagents and Subservants, 68 Harv. L. Rev. 658, 669 (1955) (in subservant situation, the “employ- ing servant . . . is in the position of a master to those whom he employs but they are also in the position of servants to the master in charge of the entire enterprise”). The Restatement (Second) Sec. 5, comment e observes that: Illustrations of the subservant relation include that between the mine owner and the assistant of a miner who furnishes his own tools and assistants, the latter, however, being subject to the general mine discipline; the relation between the owner of a building and an employee of a janitor; the relation between the employees of a branch manager of a corporation where the branch manager is free to control and pay his assistants, but where all are subject to control by the corporation as to their conduct. 76 See, e.g., Schmidt v. Burlington Northern & Santa Fe Railway Co., 605 F. 3d 686, 689-690 (9th Cir. 2010) (applying Federal Em- ployers’ Liability Act and finding evidence sufficient to establish em- ployment relationship between railroad line and employee of railroad- car maintenance and repair company). Cf. Williamson v. Consolidated Rail Corp., 926 F.2d 1344, 1350 (3d. Cir. 1991) (observing that use of subservant doctrine is unnecessary where there is evidence of direct control). See generally Kelley v. Southern Pacific Co., 419 U.S. 318, 325 (1974) (recognizing subservant doctrine for purposes of Federal Employers’ Liability Act). 77 Allbritton Communications Co. v. NLRB, 766 F.2d. 812, 818-819 (3d Cir. 1985) (upholding Board’s determination that newspaper was statutory employer of mailroom employees, although second employer operated mailroom). 78 See Restatement of Employment Law, Section 1.04(b) (June 2015) (“An individual is an employee of two or more joint employers if (i) the individual renders services to at least one of the employers and (ii) that employer and the other joint employers each control or supervise such rendering of services as provided in § 1.01(a)(3).”)(emphasis added). (In relevant part, Sec. 1.01(a)(3) defines an employee as an individual who renders service to an employer who “controls the manner and means by which the individual renders service.”) BFI NEWBY ISLAND RECYCLERY 15 Board’s current focus on only direct and immediate con- trol acknowledges the most proximate level of authority, which is frequently exercised by the supplier firm, but gives no consideration to the substantial control over workers’ terms and conditions of employment of the us- er.79 The common-law definition of an employment rela- tionship establishes the outer limits of a permissible joint-employer standard under the Act. But the Board’s current joint-employer standard is significantly narrower than the common law would permit. The result is that employees covered by the Act may be deprived of their statutory right to bargain effectively over wages, hours, and working conditions, solely because they work pursu- ant to an arrangement involving two or more employing firms, rather than one. Such an outcome seems clearly at odds with the policies of the Act. VI. THE RESTATED JOINT-EMPLOYER STANDARD Having fully considered the issue and all of the argu- ments presented, we have decided to restate the Board’s legal standard for joint-employer determinations and make clear how that standard is to be applied going for- ward. We return to the traditional test used by the Board (and endorsed by the Third Circuit in Browning-Ferris): The Board may find that two or more entities are joint em- ployers of a single work force if they are both employers within the meaning of the common law, and if they share or codetermine those matters governing the essential terms and conditions of employment. In evaluating the allocation and exercise of control in the workplace, we will consider the various ways in which joint employers may “share” control over terms and conditions of em- ployment or “codetermine” them, as the Board and the courts have done in the past.80 79 As noted in several briefs in support of the Union, the Board’s longstanding legal formulation for joint-employer status, even post-TLI, nominally acknowledges this bifurcated dynamic by covering employ- ers that “codetermine” employees’ terms and conditions of employ- ment. But the Board’s restrictive application of the test, which pre- cludes any holistic assessment of the way control is allocated between the contracting parties, undermines this aspect of the joint-employer standard. 80 In some cases (or as to certain issues), employers may engage in genuinely shared decision-making, e.g., they confer or collaborate directly to set a term of employment. See NLRB v. Checker Cab Co., 367 F.2d 692, 698 (6th Cir. 1966) (noting that employers “banded themselves together so as to set up joint machinery for hiring employ- ees, for establishing working rules for employees, for giving operating instructions to employees, for disciplining employees for violation of rules, for disciplining employees for violation of safety regulations”). Alternatively, employers may exercise comprehensive authority over different terms and conditions of employment. For example, one em- ployer sets wages and hours, while another assigns work and supervises employees. See D & F Industries, 339 NLRB 618, 640 (2003). Or We adhere to the Board’s inclusive approach in defin- ing “essential terms and conditions of employment.” The Board’s current joint-employer standard refers to “mat- ters relating to the employment relationship such as hir- ing, firing, discipline, supervision, and direction” a non- exhaustive list of bargaining subjects.81 Essential terms indisputably include wages and hours, as reflected in the Act itself.82 Other examples of control over mandatory terms and conditions of employment found probative by the Board include dictating the number of workers to be supplied;83 controlling scheduling,84 seniority, and over- time;85 and assigning work and determining the manner and method of work performance.86 This approach has generally been endorsed by the Federal courts of ap- peals.87 Also consistent with the Board’s traditional approach, we reaffirm that the common-law concept of control in- forms the Board’s joint-employer standard. But we will no longer require that a joint employer not only possess the authority to control employees’ terms and conditions employers may affect different components of the same term, e.g., one employer defines and assigns work tasks, while the other supervises how those tasks are carried out. See Hamburg Industries, supra, 193 NLRB at 67. Finally, one employer may retain the contractual right to set a term or condition of employment. See Hoskins Ready-Mix Con- crete, supra, 161 NLRB at 1493. 81 TLI, supra, 271 NLRB at 798 (emphasis added). After TLI, the Board has continued to take a broad, inclusive approach to determining the relevant object of a putative joint employer’s control, i.e., which terms and conditions of employment matter to the joint-employer in- quiry. See Aldworth Co., 338 NLRB 137, 139 (2002) (the “relevant facts involved in [the joint-employer] determination extend to nearly every aspect of employees’ terms and conditions of employment and must be given weight commensurate with their significance to employ- ees’ work life”), enfd. sub nom. Dunkin’ Donuts Mid-Atlantic Distribu- tion Center v. NLRB, 363 F.3d 437 (D.C. Cir. 2004). 82 Sec. 8(d), defining an employer’s duty to bargain, specifically re- fers to the obligation to “confer in good faith over wages, hours, and other terms and conditions of employment.” 29 U.S.C. Sec. 158(d) (emphasis added). 83 Mobil Oil, supra, 219 NLRB at 516. 84 Continental Winding Co., 305 NLRB 122, 123 fn. 4 (1991). 85 D&F Industries, supra, 339 NLRB at 649 fn. 77. 86 DiMucci Const. Co. v. NLRB., 24 F.3d 949, 952 (7th Cir. 1994) (“Factors to consider in determining joint employer status are: (1) su- pervision of employees’ day-to-day activities; (2) authority to hire or fire employees; (3) promulgation of work rules and conditions of em- ployment; (4) issuance of work assignments; and (5) issuance of operat- ing instructions”). 87 See, e.g., Tanforan Park Food Purveyors Council v. NLRB, 656 F.2d 1358, 1361 (9th Cir. 1981); Sun-Maid Growers of California v. NLRB, 618 F.2d 56, 59 (9th Cir. 1980) (“A joint employer relationship exists when an employer exercises authority over employment condi- tions which are within the area of mandatory collective bargaining.”); Cabot Corp., 223 NLRB 1388, 1389-1390 (1976), enfd. sub nom. International Chemical Workers Union Local 483 v. NLRB, 561 F.2d 253 (D.C. Cir. 1977) (labor relations policies of the contractor or im- pact over the wages, hours, and working conditions of the contractor’s employees). 16 DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD of employment, but must also exercise that authority, and do so directly, immediately, and not in a “limited and routine” manner. Accordingly, we overrule Laerco, TLI, A&M Property, and Airborne Express, supra, and other Board decisions, to the extent that they are inconsistent with our decision today. The right to control, in the common-law sense, is probative of joint-employer status, as is the actual exercise of control, whether direct or indi- rect. The existence, extent, and object of a putative joint employer’s control, of course, all may present material issues. For example, it is certainly possible that in a par- ticular case, a putative joint employer’s control might extend only to terms and conditions of employment too limited in scope or significance to permit meaningful collective bargaining. Moreover, as a rule, a joint em- ployer will be required to bargain only with respect to such terms and conditions which it possesses the authori- ty to control. The dissent repeatedly criticizes our decision as articu- lating a test under which “there can be no certainty or predictability regarding the identity of the ‘employer.’” But we do not and cannot attempt today to articulate eve- ry fact and circumstance that could define the contours of a joint employment relationship. Issues related to the nature and extent of a putative joint-employer’s control over particular terms and conditions of employment will undoubtedly arise in future cases-just as they do under the current test-and those issues are best examined and resolved in the context of specific factual circumstances. In this area of labor law, as in others, the “‘nature of the problem, as revealed by unfolding variant situations,’ requires ‘an evolutionary process for its rational re- sponse, not a quick, definitive formula as a comprehen- sive answer.”’88 Further, while our dissenting colleagues concede that the common law must form the basis of the Board’s joint-employer test, they seem unwilling to apply its mode of analysis. As the Supreme Court has acknowl- edged, multifactor common-law inquiries are inherently nuanced and indeterminate: “In such a situation as this there is no shorthand formula or magic phrase that can be applied to find the answer, but all of the incidents of the relationship must be assessed and weighed with no one factor being decisive. What is important is that the total factual context is assessed in light of the pertinent com- mon-law agency principles.”89 Accordingly, the nuanced approach that the dissent decries is a longstanding neces- 88 Eastex, Inc. v. NLRB, 437 U.S. 556, 574-575 (1978), quoting Electrical Workers v. NLRB, 366 U.S. 667, 674 (1961). 89 United Insurance, supra, 390 U.S. at 258. sity of our common-law mandate, and not a novel or dis- cretionary feature that we introduce here. Our dissenting colleagues also accuse us of articulating a test “with no limiting principle” that “removes all limi- tations on what kind or degree of control over essential terms and conditions of employment may be sufficient to warrant a joint-employer finding.” This is simply not the case. The dissent ignores the limitations that are inherent to the common law, particularly those set forth in the Restatement provisions enumerated above. Instead, the dissent suggests that, under the revised joint-employer test, a homeowner who hires a plumber or a lender who sets the homeowner’s financing terms may each be deemed a statutory employer. But by any common-law analysis, these parties will not exercise, or have the right to exercise, the requisite control over the details of em- ployees’ work to forge common-law employment rela- tionships. It should therefore come as no surprise that the annals of Board precedent contain no cases that im- plicate the consumer services purchased by unsuspecting homeowners or lenders. The dissent is particularly pointed in its criticism of our assignment of probative weight to a putative employ- er’s indirect control over employees; it contends that “anyone contracting for services, master or not, inevita- bly will exert and/or reserve some measure of indirect control by defining the parameters of the result desired to ensure he or she gets the benefit of his or her bargain.” We do not suggest today that a putative employer’s bare rights to dictate the results of a contracted service or to control or protect its own property constitute probative indicia of employer status. Instead, we will evaluate the evidence to determine whether a user employer affects the means or manner of employees’ work and terms of employment, either directly or through an intermediary. In this case, for instance, BFI communicated precise di- rectives regarding employee work performance through Leadpoint’s supervisors. We see no reason why this ob- vious control of employees by BFI should be discounted merely because it was exercised via the supplier rather than directly. Finally, the dissent asserts that today’s decision gives the Board license to find joint-employer status based on only the slightest, most tangential evidence of control and “any degree of indirect or reserved control over a single term . . . may suffice to establish joint-employer status.” Today’s decision, however, makes clear that “all of the incidents of the relationship must be assessed.”90 Here, for example, our conclusion that BFI is a joint em- ployer is based on a full assessment of the facts (set forth 90 United Insurance, supra, 390 U.S. at 258. BFI NEWBY ISLAND RECYCLERY 17 below) that reveals multiple examples of reserved, direct, and indirect control over Leadpoint employees. VII. RESPONSE TO DISSENT’S ARGUMENTS REGARDING THE COMMON LAW Notwithstanding the strong basis in common law for the standard we adopt, our dissenting colleagues assert repeatedly that the Board is not applying common law but instead reverting to the “economic realities” test that was once applied by the Supreme Court in NLRB v. Hearst Publications, 322 U.S. 111 (1944). In Hearst, the Court interpreted the Act to include “employees (who) are at times brought into an economic relationship with employers who are not their employers”; to “reject con- ventional limitations” in defining an employee or em- ployer; and to intend that those definitions be applied “broadly . . . by underlying economic facts.”91 Our dis- senting colleagues also assert that while the Hearst standard would include indirect control over terms of employment within the definition of joint employer, common law does not. Both of these assertions are incorrect. As we have al- ready made clear, our revised standard considers-as does common law-only an entity’s control over terms of employment, not the wider universe of all “underlying economic facts” that surround an employment relation- ship.92 Moreover, courts applying the “economic reali- ties” test for an employer under the Fair Labor Standards Act and the Agricultural Workers Protection Act (AWPA) have recognized that although that test is signif- icantly more expansive than the common-law test, indi- rect control over terms of employment is clearly a factor in the common-law test.93 91 Id. at 129. 92 Citing Justice Stewart’s concurrence in Fibreboard Corp. v. NLRB, 379 U.S. 203 (1964), the dissent sets up a straw man suggesting that our test encroaches on an employer’s decisions concerning the volume and kind of advertising expenditures, product design, the man- ner of financing, and sales. Here, we are dealing only with subjects that are indisputably bargainable. 93 “[The factor of] ‘degree of supervision by the grower, direct or indirect, of the work’ [regulation citation omitted] . . . like the growers’ control over the workers, has more to do with common-law employ- ment concepts of control than with economic dependence.” Antenor v. D & S Farms, 88 F.3d 925, 934 (11th Cir. 1996) (applying AWPA, emphasis added). “[I]n considering a joint-employment relationship [under the AWPA] . . . our inquiry looks not to the common law defini- tions of employer and employee (for instance, to tests measuring the amount of control an ostensible employer exercised over a putative employee), but rather to the ‘economic reality’ of all the circumstances concerning whether the putative employee is economically dependent upon the alleged employer.” Id. at 933, quoting Aimable v. Long & Scott Farms, 20 F.3d 434, 439 (11th Cir. 1994) (emphasis added). See also Williamson v. Consolidated Rail Corp., supra, 926 F.2d at 1350 (in the common-law test for an employment relationship under FELA, “the The dissent also insists that the “current test is fully consistent with the common law agency principles” and should not be revisited or altered. But it fails to dispute or even acknowledge the extensive legal authority we cite to establish the common-law foundation of our ap- proach.94 factual issue before the jury included direct control, as well as indirect control through sub-agency.”) 94 Even where our dissenting colleagues cite case law, their efforts are wholly unpersuasive. In support of their contention (notwithstand- ing their acknowledgment to the contrary) that the common law re- quires proof of direct and immediate control to substantiate employer status, our colleagues rely on a number of early common-law decisions that merely confirm the traditional legal distinction between an em- ployer’s control over the final product and an employer’s control over the work of employees, which we do not dispute. Our colleagues also cite various independent-contractor decisions to support their proposi- tion that courts have “implicitly limited their analysis to looking for direct and immediate control.” But none of these decisions hold, even implicitly, that the existence of indirect control would not be probative of employer status; they are merely garden-variety independent- contractor cases in which courts found that individuals were not em- ployees based on the totality of the circumstances. The dissent’s at- tempt to glean any kind of general principle disfavoring indirect control as a relevant factor from these decisions-without citing any specific facts-is tenuous at best. Likewise, the comments from Sec. 220 of the Restatement (Second) of Agency on which our colleagues rely do not state or suggest that the consideration of indirect control is proscribed under the common law. As to the more recent circuit court decisions that our colleagues cite, the dissent’s assertions regarding direct control depend largely on the quotation of key phrases taken out of context. In Gulino v. N.Y. State Education Dept., 460 F.3d 361 (2d Cir. 2006), for instance, the court found that the Education Department was not a joint employer (subject to Title VII liability) because it did not hire, promote, or demote teach- ers, or determine their pay, tenure or benefits. Id. at 379. Although the court stated that it was looking for a “level of control [that] is direct, obvious, and concrete, not merely indirect or abstract”, it did so only to emphasize that all of the evidence presented to support a joint-employer finding was attenuated and insubstantial. Id. In Doe I v. Wal-Mart Stores, Inc., 572 F.3d 677 (9th Cir. 2009), the plaintiffs were overseas employees who alleged that Wal-Mart was their joint employer because it contracted with their local employers for production of goods. The court emphasized that Wal-Mart contracted with the factories only regarding prices, the quality of products, and the materials used. Id. at 683. As in Gulino, the court’s statement that Wal-Mart did not have the right to exercise an “immediate level of day to day control” over em- ployees was a reflection of Wal-Mart’s total lack of control over work- ing conditions rather than a specific holding on the probative value of indirect control evidence. Id. Indeed, neither of these cases were close, and the courts’ decisions did not turn on any refusal to assign weight to indirect control; rather, in both decisions, there was little if any relevant evidence of control of any sort. In Patterson v. Domino’s Pizza, LLC, 333 P.3d 723, 740 (Cal. 2014), while the Supreme Court of California stated that its employer standard required “a comprehensive and imme- diate level of day-to-day authority over matters such as hiring, firing, direction, supervision, and discipline of the employee” (internal quota- tions omitted), the court was expressly relying on precedent under the California Fair Employment and Housing Act. That decision also ad- dressed the particularized features of franchisor/franchisee relation- ships, none of which are present here. 18 DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD VIII. APPLICATION OF THE RESTATED TEST With the above principles in mind, we evaluate here whether BFI constitutes a joint employer under the Act. As always, the burden of proving joint-employer status rests with the party asserting that relationship.95 Having assessed all of the relevant record evidence, we conclude that the Union has met its burden of establishing that BFI is a statutory joint employer of the sorters, screen clean- ers, and housekeepers at issue. BFI is an employer under common-law principles,96 and the facts demonstrate that it shares or codetermines those matters governing the essential terms and conditions of employment for the Leadpoint employees. In many relevant respects, its right to control is indisputable. Moreover, it has exercised that control, both directly and indirectly. Finding joint- employer status here is consistent with common-law principles, and it serves the purposes of the National La- bor Relations Act. We rely on the following factors in reaching this conclusion. A. Hiring, Firing, and Discipline BFI possesses significant control over who Leadpoint can hire to work at its facility. By virtue of the parties’ Agreement, which is terminable at will,97 BFI retains the right to require that Leadpoint “meet or exceed [BFI’s] own standard selection procedures and tests,”98 requires that all applicants undergo and pass drug tests, and pro- scribes the hiring of workers deemed by BFI to be ineli- gible for rehire.99 Although BFI does not participate in 95 See, e.g. Flagstaff Medical Center, supra, 357 NLRB No. 65 slip op. at 9. 96 It is clear that Leadpoint employees are, in the words of Restate- ment (Second) of Agency §220(1) “employed to perform services in the affairs of” BFI and “with respect to the physical conduct in the perfor- mance of the services” are “subject to [BFI’s] control or right to con- trol.” The record shows that BFI engages in “de facto close supervi- sion” of the work of Leadpoint employees; that the work of Leadpoint employees “does not require the services of one highly educated or skilled;” that Leadpoint employees have “employment over a consider- able period of time with regular hours;” and that the work of Leadpoint employees “is part of the regular business” of BFI. Restatement (Se- cond) of Agency Sec. 220, comment h (“Factors indicating the relation of master and servant”). As a general matter, this case closely resem- bles the situation addressed in Restatement (Second) Sec. 220, com- ment l, which explains that where “work is done upon the premises of the employer with his machinery by workmen who agree to obey gen- eral rules for the regulation of the conduct of employees, the inference is strong that such workmen are the servants of the owner.” Finally, the record here fairly permits categorizing the Leadpoint employees as subservants of BFI, as well as servants of Leadpoint. 97 See Value Village, supra, 161 NLRB at 607; Mobil Oil, supra, 219 NLRB at 516 (relying on user’s right to terminate contract at will as evidence of control). 98 Applicants are tested on BFI’s equipment and are required to meet specific productivity benchmarks in order to qualify for hire. 99 See K-Mart, 159 NLRB 256, 258 (1966) (relying, in part, on con- tract language stating that contracting parties would not “hire an em- Leadpoint’s day-to-day hiring process, it codetermines the outcome of that process by imposing specific condi- tions on Leadpoint’s ability to make hiring decisions. Moreover, even after Leadpoint has determined that an applicant has the requisite qualifications, BFI retains the right to reject any worker that Leadpoint refers to its fa- cility “for any or no reason.”100 Similarly, BFI possesses the same unqualified right to “discontinue the use of any personnel” that Leadpoint has assigned.101 Although BFI managers testified that they have never discontinued use of a Leadpoint employ- ee or been involved in disciplinary procedures, record evidence includes two specific instances where BFI Op- erations Manager Keck reported employees’ misconduct to Leadpoint and “request[ed] their immediate dismis- sal.” In response to Keck’s directive, Leadpoint officials immediately removed the employees from their line du- ties and dismissed them from the BFI facility shortly thereafter. Though the evidence shows that Leadpoint conducted its own investigation of the alleged miscon- duct, it is also plain that the outcome was preordained by BFI’s ultimate right under the terms of the Agreement to dictate who works at its facility.102 B. Supervision, Direction of Work, and Hours In addition, BFI exercises control over the processes that shape the day-to-day work of the petitioned-for em- ployees. Of particular importance is BFI’s unilateral con- trol over the speed of the streams and specific productivi- ty standards for sorting.103 BFI argues that, although it controls the pace of work, Leadpoint supervisors alone decide how employees will respond to BFI’s adjust- ments. This characterization of the process, however, discounts the clear and direct connection between BFI’s decisions and employee work performance. The evi- ployee or former employee of the other without first checking” with the other party). 100 See Pacemaker Driver Service, 269 NLRB 971, 975 (1984), enfd. 768 F.2d 778 (6th Cir. 1985) (relying on user’s unilateral right to reject any driver referred by contractor); Lowery Trucking, supra, 177 NLRB at 15 (noting that “while [the user] never rejected a driver hired by [the supplier], it had the right to do so.”). 101 See Ref-Chem Co., supra, 169 NLRB at 379 (emphasizing user’s “virtually unqualified right to request the removal of an employee of the contractor.”); Hamburg Industries, supra, 193 NLRB at 67 (relying on user’s right to force supplier to remove employees from its plant). 102 As Keck stated in his e-mail to Leadpoint on this matter, the misconduct Keck witnessed “is all I need to proceed.” See Grand Central Liquors, 155 NLRB 295, 297 (1965) (noting that where the user requested the discharge of employees, the supplier complied). 103 Clayton B. Metcalf, supra, 223 NLRB at 644 (emphasizing that putative employer had “day-to-day responsibility for the overall opera- tion of the [facility] and all . . . operations were performed in accord- ance with [its] . . . plan” and that it “exercised considerable control over the manner and means by which [the subcontractor] performed its operations.”) BFI NEWBY ISLAND RECYCLERY 19 dence reveals that the speed of the line and the resultant productivity issues have been a major source of strife between BFI and the workers. BFI managers have direct- ly implored workers to work faster and smarter; likewise, they have repeatedly counseled workers, in the interest of productivity, against stopping the streams. Tellingly, there is no evidence that Leadpoint has had any say in these decisions. Indeed, given BFI’s “ultimate control” over these matters, it is difficult to see how Leadpoint alone could bargain meaningfully about such fundamen- tal working conditions as break times, safety, the speed of work, and the need for overtime imposed by BFI’s productivity standards.104 BFI managers also assign the specific tasks that need to be completed, specify where Leadpoint workers are to be positioned, and exercise near-constant oversight of employees’ work performance.105 The fact that many of their directives are communicated through Leadpoint supervisors hardly disguises the fact that BFI alone is making these decisions.106 Further, in numerous instanc- es, BFI has dispensed with the middleman altogether. BFI managers have communicated detailed work direc- tions to employees on the stream; held meetings with employees to address customer complaints and business objectives, and to disseminate preferred work practices; and assigned to employees tasks that take precedence over any work assigned by Leadpoint.107 We find that all of these forms of control - both direct and indirect - are indicative of an employer-employee relationship. In addition, BFI specifies the number of workers that it requires,108 dictates the timing of employees’ shifts,109 104 Int’l Trailer, supra, 133 NLRB at 1529. See also Carrier Corp. v. NLRB, supra, 768 F.2d at 781 (finding substantial evidence in support of the Board’s joint-employer finding where putative employer “exer- cised substantial day-to-day control over the drivers’ working condi- tions.”). 105 See Hamburg Industries, supra, 193 NLRB at 67 (finding indicia of control where putative employer instructed supplier on the work to be performed and “constantly check[ed] the performance of the work- ers and the quality of the work.”) 106 See Int’l Trailer, supra, 133 NLRB at 1529 (noting that, although putative employer did not directly supervise employees, it issued or- ders, through the other firm’s supervisor, as to how employees should perform their duties). 107 See Sun-Maid Growers, supra, 239 NLRB at 350 (finding indicia of control where putative employer’s supervisors “occasionally provid- ed specifications and instructions regarding the manner in which the work could be performed” and directly assigned work that took prece- dence over other assignments). 108 See Mobil Oil, supra, 219 NLRB at 516 (relying on user’s ability to dictate the size of the supplier’s crew); Hamburg Industries, supra, 193 NLRB at 67 (same). 109 BFI also affects the length of break periods by requiring employ- ees to clean around their work stations before releasing them on break. and determines when overtime is necessary.110 Although Leadpoint is responsible for selecting the specific em- ployees who will work during a particular shift, it is BFI that makes the core staffing and operational decisions that define all employees’ work days. In turn, Leadpoint employees are required to obtain the signature of an au- thorized BFI representative attesting to their “hours of services rendered” each week; failure to do so permits BFI to refuse payment to Leadpoint for time claimed by a Leadpoint worker. C. Wages We find too that BFI plays a significant role in deter- mining employees’ wages. Under the parties’ contract, Leadpoint determines employees’ pay rates, administers all payments, retains payroll records, and is solely re- sponsible for providing and administering benefits. But BFI specifically prevents Leadpoint from paying em- ployees more than BFI employees performing compara- ble work.111 BFI’s employment of its own sorter at $5 more an hour creates a de facto wage ceiling for Leadpoint workers. In addition, BFI and Leadpoint are parties to a cost-plus contract, under which BFI is re- quired to reimburse Leadpoint for labor costs plus a specified percentage markup.112 Although this arrange- ment, on its own, is not necessarily sufficient to create a joint-employer relationship,113 it is coupled here with the apparent requirement of BFI approval over employee pay increases.114 Thus, after new minimum wage legislation went into effect, BFI and Leadpoint entered into an agreement verifying that BFI would pay a higher rate for the services of Leadpoint employees.115 110 Sun-Maid Growers, supra, 239 NLRB at 351 (finding indicia of control where the user dictated employees’ “basic workweek” and number of overtime hours available based on its production schedule); Floyd Epperson, supra, 202 NLRB at 23 (user established work sched- ules). 111 See K-Mart, 161 NLRB 1127, 1129 (1966) (relying on the fact that putative employer directed other firm to start full-time employees at no less than the rate that it paid to certain categories of its employ- ees). 112 See CNN America, 361 NLRB No. 47 slip op. at 6 (2014) (rely- ing on parties’ cost-plus arrangement as evidence of joint-employer status); Hoskins Ready-Mix Concrete, supra, 161 NLRB at 1493, and the cases cited in footnote 37. 113 See Pulitzer Publishing Co., 242 NLRB 35, 36 (1979), enf. de- nied 618 F.2d 1275 (8th Cir. 1980), cert. denied 499 U.S. 875 (1980) (assessing parties’ cost-plus contract as one factor among many). 114 See Hoskins Ready-Mix Concrete, supra, 161 NLRB at 1493 (re- lying on the fact that supplier was required to consult with user and obtain clearance before changing pay rates or hiring new employees at a rate above a specified level). 115 In addition to the factors stated, we rely on the fact that BFI, by the terms of the Agreement, compels Leadpoint and its employees to comply with BFI’s safety policy, and reserves the right to enforce its safety policy as to the workers. See Hamburg Industries, 193 NLRB at 67 (user requires all employees to follow its own safety rules); Man- 20 DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD We find BFI’s role in sharing and codetermining the terms and conditions of employment establishes that it is a joint employer with Leadpoint.116 Accordingly, we reverse the Regional Director and find that BFI and Leadpoint are joint employers of the sorters, screen cleaners, and housekeepers at issue.117 VIII. THE IMPLICATIONS OF TODAY’S DECISION Today’s decision is grounded firmly in the common law, while advancing the policies of the National Labor Relations Act. In both respects, its approach is superior to prior law, which, as we have explained, imposed re- strictions on the joint-employer standard that have no common-law basis and that foreclosed collective bar- gaining even in situations where it could be productive. Certainly, we have modified the legal landscape for em- ployers with respect to one federal statute, the National Labor Relations Act.118 But “reevaluating doctrines, refining legal rules, and sometimes reversing precedent are familiar parts of the Board’s work-and rightly so.”119 As recognized by the Supreme Court: The use by an administrative agency of the evolutional approach is particularly fitting. To hold that the Board’s earlier decisions froze the development . . . of the na- tional labor law would misconceive the nature of ad- ministrative decisionmaking. power, 164 NLRB 287, 287-288 (1967) (user gives employees safety instruction and conducts periodic safety meetings). We also note that BFI and Leadpoint have jointly determined, also by terms of the Agreement, that employees cannot work at BFI for more than 6 months. We find that these terms are further indicative of BFI’s status as an employer of the employees at issue. 116 See Hamburg Industries, supra, 193 NLRB at 67 (finding user to be joint-employer, in substantially similar factual scenario, where user had “considerable control over [supplier’s] operations in such critical areas as work instructions, quality control and the right to reject fin- ished work, work scheduling, and indirect control over wages”). 117 The dissent, in its brief discussion of the facts in this case, con- tends that “the majority’s evidence amounts to a collection of general contract terms or business practices . . . plus a few extremely limited actions that had some routine impact on Leadpoint employees.” In so doing, however, the dissent cannot avoid setting out a list of nine spe- cific ways in which BFI has exercised or reserved control over Leadpoint employees. In our view, our colleagues’ accounting of these factors makes a persuasive case for BFI’s joint-employer status. None- theless, we note that the dissent’s analysis excludes or downplays sev- eral additional critical factors, including BFI’s control over the speed of the lines, productivity standards, and the use of the stop switches, as well as BFI’s direct and ongoing instruction of Leadpoint employees in the details of job performance. 118 The Board’s joint-employer standard, of course, does not govern joint-employer determinations under the many other statutes, federal and state, that govern the workplace and that use a variety of different standards to determine whether a particular business entity has legal duties with respect to particular workers. 119 UGL-UNICCO Service Co., 357 NLRB No. 76, slip op. at 5 (2011). NLRB v. J. Weingarten, supra, 420 U.S. at 265-266. Our colleagues’ long and hyperbolic dissent persistent- ly mischaracterizes the standard we adopt today and grossly exaggerates its consequences, but makes no real effort to address the difficult issue presented here: how best to “encourag[e] the practice and procedure of collec- tive bargaining” (in the Act’s words) when otherwise bargainable terms and conditions of employment are under the control of more than one statutory employer. Instead, the dissent puts the preservation of the current status quo far ahead of any cognizable statutory policy. Our colleagues never adequately explain why the Board should adhere to an approach that they essentially con- cede is not compelled by the common law and that de- monstrably fails to fully advance the goals of the Act.120 As a practical matter, the criticisms that our colleagues level at our joint-employer standard could be made about the concept of joint employment generally-which has been recognized under the Act for many decades and which has long been a familiar feature of labor and em- ployment law. The law-school-exam hypothetical of doomsday scenarios that they predict will result from today’s decision is likewise based on an exaggeration of the challenges that can sometimes arise when multiple employers are required to engage in collective bargain- ing. The potential for these types of challenges to arise has existed for as long as the Board has recognized the joint-employer concept. Nonetheless, employers and unions have long managed to navigate these challenges, and the predicted disasters have not come to pass.121 120 The dissent is simply wrong when it insists that today’s decision “fundamentally alters the law” with regard to the employment relation- ships that may arise under various legal relationships between different entities: “lessor-lessee, parent-subsidiary, contractor-subcontractor, franchisor-franchisee, predecessor-successor, creditor-debtor, and con- tractor-consumer.” None of those situations are before us today, and we decline the dissent’s implicit invitation to address the facts in every hypothetical situation in which the Board might be called on to make a joint-employer determination. As we have made clear, the common- law test requires us to review, in each case, all of the relevant control factors that are present determining the terms of employment. In this case we are specifically concerned with only two employers: BFI and Leadpoint. Likewise, we need not address the dissent’s assertion that the deci- sion somehow undermines other rules under the Act that are not at issue here, such as the prohibition on secondary boycott activity, other than to emphasize that our decision today does not modify any other legal doctrine, create “different tests” for “other circumstances,” or change the way that the Board’s joint-employer doctrine interacts with other rules or restrictions under the Act. 121 For example, 20 years ago, the Board changed its approach in cases involving government contractors, rejecting the position that the Board should assert jurisdiction only where the contractor controlled economic terms and conditions of employment. Management Training Corp., supra. The dissent insisted that the Board had “radically change[d] extant law,” adopting a “doctrine that ha[d] virtually no BFI NEWBY ISLAND RECYCLERY 21 It is not the goal of joint-employer law to guarantee the freedom of employers to insulate themselves from their legal responsibility to workers, while maintaining control of the workplace. Such an approach has no basis in the Act or in federal labor policy. DIRECTION IT IS DIRECTED that the Regional Director for Region 32 shall, within 14 days of this Decision on Review and Direction, open and count the impounded ballots cast by the employees in the petitioned-for unit, prepare and serve on the parties a tally of ballots, and thereafter issue the appropriate certification. Dated, Washington, D.C. August 27, 2015 ______________________________________ Mark Gaston Pearce, Chairman ______________________________________ Kent Y. Hirozawa, Member ______________________________________ Lauren McFerran, Member (SEAL) NATIONAL LABOR RELATIONS BOARD MEMBERS MISCIMARRA AND JOHNSON, dissenting. The National Labor Relations Act (the Act) establishes a comprehensive set of rules for industrial relations in this country, and a primary function of the Board is to foster compliance with those rules by employees, unions, and employers. To comply with these rules, as they have grown and evolved over the last eight decades, substan- tial planning is required. This is especially true in regard to collective bargaining, a process that is central to the Act. The Act’s bargaining obligations are formidable- as they should be-and violations can result in signifi- cant liability. When it comes to the duty to bargain, the resort to strikes or picketing, and even the basic question of “who is bound by this collective-bargaining agree- ment,” there is no more important issue than correctly identifying the “employer.” Changing the test for identi- fying the “employer,” therefore, has dramatic implica- tions for labor relations policy and its effect on the econ- omy. Today, in the most sweeping of recent major decisions, the Board majority rewrites the decades-old test for de- termining who the “employer” is. More specifically, the limitation” and would “cause more problems than it solve[d].” 317 NLRB at 1360-1362. These dire predictions did not come to pass, and Management Training remains the law today. majority redefines and expands the test that makes two separate and independent entities a “joint employer” of certain employees. This change will subject countless entities to unprecedented new joint-bargaining obliga- tions that most do not even know they have, to potential joint liability for unfair labor practices and breaches of collective-bargaining agreements, and to economic pro- test activity, including what have heretofore been unlaw- ful secondary strikes, boycotts, and picketing. Our colleagues are driven by a desire to ensure that the prospect of collective bargaining is not foreclosed by business relationships that allegedly deny employees’ right to bargain with employers that share control over essential terms and conditions of their employment. However well intentioned they may be, there are five major problems with this objective. First, no bargaining table is big enough to seat all of the entities that will be potential joint employers under the majority’s new standards. In this regard, we believe the majority’s new test impermissibly exceeds our statu- tory authority. From the majority’s perspective, the change in the joint-employer analysis is an allegedly necessary adaptation of Board law to reflect changes in the national economy. In making this change, they pur- port to operate within the limits of traditional common- law principles by restoring and clarifying what they claim to be the law applied by the Board prior to 1984. In actuality, however, our colleagues incorporate theories of “economic realities” and “statutory purpose” that ex- tend the definitions of “employee” and “employer” far beyond the common-law limits of agency principles that Congress and the Supreme Court have stated must ap- ply.1 Their decision represents a further expansion of revisions made in the majority decisions in FedEx,2 which similarly revised the Board’s longstanding defini- tion of independent contractor status in a way that will predictably extend the Act’s coverage to many individu- als previously considered to be excluded as independent contractors, and in CNN,3 which imposed after-the-fact joint-employer obligations contrary to the parties’ 20- year-bargaining history, applicable collective-bargaining agreements (CBAs), relevant services contracts and the Board’s own prior union certifications. Second, the majority’s rationale for overhauling the Act’s “employer” definition-to protect bargaining from limitations resulting from third-party relationships that indirectly control employment issues-relies in substan- 1 The common-law agency principles are also known as “master- servant” principles in the older cases and literature, and these terms are used interchangeably both in the doctrine and here. 2 FedEx Home Delivery, 361 NLRB No. 55 (2014). 3 CNN America, Inc., 361 NLRB No. 47 (2014). 22 DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD tial part on the notion that these relationships are unique in our modern economy and represent a radical departure from simpler times when labor negotiations were unaf- fected by the direct employer’s commercial dealings with other entities. However, such an economy has not exist- ed in this country for more than 200 years.4 Many forms of subcontracting, outsourcing, and temporary or contin- gent employment date back to long before the 1935 pas- sage of the Act. Congress was obviously aware of the existence of third-party intermediary business relation- ships in 1935, when it limited bargaining obligations to the “employer,” in 1947, when it limited the definition of “employee” and “employer” to their common-law agen- cy meaning, and in 1947 and 1959, when Congress strengthened secondary boycott protection afforded to third parties who, notwithstanding their dealings with the “employer,” could not lawfully be subject to picketing and other forms of economic coercion based on their dealings with that “employer.”5 This is not mere conjec- ture; it is the inescapable conclusion that follows from Supreme Court precedent recognizing that the Act did not confer “employer” status on third parties merely be- cause commercial relationships made them interdepend- ent with an “employer” and its employees.6 Third, courts have afforded the Board deference in this context merely as to the Board’s ability to make factual distinctions when applying the common-law agency 4 If our colleagues desired to return to a time when labor- management relations were insulated from third-party business rela- tionships and competitive pressures, they would need to go back to our country’s origins. The work of labor economists John R. Commons and Selig Perlman, who are perhaps the two most authoritative histori- ans of the American labor movement, indicates that unions expanded and contracted for the first several centuries of economic development in the United States, and the transition to national markets, combined with unprecedented business competition, caused extensive labor- management instability. See 1 John R. Commons, HISTORY OF LABOUR IN THE UNITED STATES 25-30 (1918); Selig Perlman, A HISTORY OF TRADE UNIONISM IN THE UNITED STATES 36-41 (1922); see also Philip S. Foner, THE HISTORY OF THE LABOR MOVEMENT IN THE UNITED STATES: FROM COLONIAL TIMES TO THE FOUNDING OF THE AMERICAN FEDERATION OF LABOR 338-340 (1947). 5 See, e.g., Sec. 8(b)(4) and (e). 6 See, e.g., NLRB v. Denver Building Trades Council, 341 U.S. 675, 692 (1951) (holding that construction industry general contractors have no “employer” relationship with the employees of subcontractors, not- withstanding the general contractor’s responsibility for the entire pro- ject). In Fibreboard Corp. v. NLRB, 379 U.S. 203 (1964), an employer contracted out the maintenance work and “merely replaced existing employees with those of an independent contractor,” and even though the subcontractor’s employees continued “to do the same work under similar conditions of employment” and the “maintenance work still had to be performed in the plant,” id. at 213, Fibreboard ceased being the “employer.” Indeed, the premise of Fibreboard and comparable deci- sions is that the outsourcing of work may “quite clearly imperil job security, or indeed terminate employment entirely” for employees of the contracting employer. Id. at 223 (Stewart, J., concurring). standard.7 However, our colleagues mistakenly interpret this as a grant of authority to modify the agency standard itself. This type of change is clearly within the province of Congress, not the Board. Thus, in Yellow Taxi Co. of Minneapolis v. NLRB,8 in which the D.C. Circuit de- nounced the Board majority’s “thinly veiled defiance” of controlling precedent regarding the “common law rules of agency,” the court of appeals stated that “[n]o court can overlook an agency’s defiant refusal to follow well established law,” and it observed: The Board here is acting in an area where it is called upon to apply common law principles that have been established since 1800 and where the application of that law under the National Labor Relations Act has been declared by Congress and settled by the courts, includ- ing the Supreme Court, for some 36 years. In this area, there is no dispute as to the governing principles of law; what is involved is the application of law to facts. “[S]uch a determination of pure agency law involve[s] no special administrative expertise that a court does not possess.” [NLRB v. United Ins. Co. of America, 390 U.S. 254, 260 (1968).] To be specific, we understand the common-law standard as codified by the Act to put a premium on direct control be- fore making an entity the joint employer of certain workers. Our fundamental disagreement with the majority’s test is not just that they view indicia of indirect, and even potential, control to be probative of employer status, they hold such indicia can be dispositive without any evidence of direct control. Under the common law, in our view, evidence of indirect control is probative only to the extent that it sup- plements and reinforces evidence of direct control. Fourth, the majority abandons a longstanding test that provided certainty and predictability, and replaces it with an ambiguous standard that will impose unprecedented bargaining obligations on multiple entities in a wide va- riety of business relationships, even if this is based solely on a never-exercised “right” to exercise “indirect” con- trol over what a Board majority may later characterize as “essential” employment terms. This new test leaves em- ployees, unions, and employers in a position where there can be no certainty or predictability regarding the identi- 7 The Supreme Court’s decision in Boire v. Greyhound Corp., 376 U.S. 473, 481 (1964), speaks directly only to the Board’s ability to make factual distinctions under the common-law agency standard. The determination of whether two entities are joint employers “is essentially a factual issue.” Id. 8 721 F.2d 366 (D.C. Cir. 1983). See also NLRB v. Town & Country Electric, Inc., 516 U.S. 85, 94 (1995) (“In some cases, there may be a question about whether the Board’s departure from the common law of agency with respect to particular questions and in a particular statutory context, renders its interpretation unreasonable.”). BFI NEWBY ISLAND RECYCLERY 23 ty of the “employer.” Just like the test of employee sta- tus rejected by the Supreme Court in Nationwide Mutual Insurance Co. v. Darden, 530 U.S. 318, 326 (1992), the majority’s new joint-employer standard constitutes “an approach infected with circularity and unable to furnish predictable results.” This confusion and disarray threat- ens to cause substantial instability in bargaining relation- ships, and will result in substantial burdens, expense, and liability for innumerable parties, including employees, employers, unions, and countless entities who are now cast into indeterminate legal limbo, with consequent de- lay, risk, and litigation expense. Nor can this type of fundamental uncertainty be positively regarded by the courts.9 Fifth, to the extent the majority seeks to correct a per- ceived inequality of bargaining leverage resulting from complex business relationships, where some entities are currently nonparticipants in bargaining, the “inequality” addressed by the majority is the wrong target, and collec- tive bargaining is the wrong remedy. As noted above, the inequality targeted by the new “joint-employer” test is a fixture of our economy-business entities have di- verse relationships with different interests and leverage that varies in their dealings with one another. There are contractually “more powerful” business entities and “less powerful” business entities, and all pursue their own in- terests. The Board needs a clear congressional com- mand-and none exists here-before undertaking an attempt to reshape this aspect of economic reality. The Act does not redress imbalances of power between em- ployers, even if those imbalances have some derivative effect on employees. As Justice Stewart observed 50 years ago: [I]t surely does not follow that every decision which may affect job security is a subject of compulsory col- lective bargaining. Many decisions made by manage- ment affect the job security of employees. Decisions concerning the volume and kind of advertising expendi- tures, product design, the manner of financing, and sales, all may bear upon the security of the workers’ jobs. Yet it is hardly conceivable that such decisions so involve “conditions of employment” that they must be negotiated with the employees’ bargaining representa- tive. Fibreboard Corp. v. NLRB, 379 U.S. 203 (1964) (Stewart, J., concurring) (emphasis added); see also First National Maintenance Corp. v. NLRB, 452 U.S. at 676 (In adopting 9 See, e.g., First National Maintenance Corp. v. NLRB, 452 U.S. 666, 678-679, 684-686 (1981), and other cases discussed in part V, subpart B of this opinion, emphasizing the need for certainty, predicta- bility, and stability. the NLRA, Congress “had no expectation that the elected union representative would become an equal partner in the running of the business enterprise in which the union’s members are employed.”). Requiring collective bargaining wherever there is some interdependence between or among employers is much more likely to thwart labor peace than advance it. Indeed, on matters of economic power and relative in- equality, the Board is not even vested with “general au- thority to define national labor policy by balancing the competing interests of labor and management.” Ameri- can Ship Building Co. v. NLRB, 380 U.S. 300, 316 (1965). “It is implicit in the entire structure of the Act that the Board acts to oversee and referee the process of collective bargaining, leaving the results of the contest to the bargaining strengths of the parties.” H. K. Porter Co. v. NLRB, 397 U.S. 99, 107-108 (1970). Therefore, we are certainly not vested with general authority to define national economic policy by balancing the competing interests of different business enterprises. The Act encourages collective bargaining, but only by an “employer” in direct relation to its employees. Our colleagues take this purpose way beyond what Congress intended, and the result unavoidably will be too much of a good thing. We believe the majority’s test will actually foster substantial bargaining instability by requiring the nonconsensual presence of too many entities with diverse and conflicting interests on the “employer” side. Indeed, even the commencement of good-faith bargaining may be delayed by disputes over whether the correct “em- ployer” parties are present. This predictable outcome is irreconcilable with the Act’s overriding policy to “elimi- nate the causes of certain substantial obstructions to the free flow of commerce.”10 In sum, today’s majority holding does not represent a “return to the traditional test used by the Board,” as our colleagues claim even while admitting that the Board has never before described or articulated the test they an- nounce today. Contrary to their characterization, the new joint-employer test fundamentally alters the law applica- ble to user-supplier, lessor-lessee, parent-subsidiary, con- tractor-subcontractor, franchisor-franchisee, predecessor- successor, creditor-debtor, and contractor-consumer business relationships under the Act. In addition, be- cause the commerce data applicable to joint employers is combined for jurisdictional purposes,11 the Act’s cover- age will extend to small businesses whose separate oper- ations and employees have until now not been subject to Board jurisdiction. As explained in detail below, we 10 Sec. 1 (emphasis added). 11 Valentine Properties, 319 NLRB 8 (1995). 24 DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD believe the majority impermissibly exceeds our statutory authority, misreads and departs from prior case law, and subverts traditional common-law agency principles. The result is a new test that confuses the definition of a joint employer and will predictably produce broad-based in- stability in bargaining relationships. It will do violence as well to other requirements imposed by the Act, nota- bly including the secondary boycott protection that Con- gress afforded to neutral employers. For all of these rea- sons, we dissent. I. THE CURRENT JOINT-EMPLOYER TEST The Act does not expressly define who is an employer, whether joint or sole. In relevant part, Section 2(2) states only that “[t]he term ‘employer’ includes any person acting as an agent of an employer, directly or indirectly.” In cases decided prior to 1984, both the Board and courts occasionally confused resolution of the issue whether two entities are joint employers by, among other things, blurring the distinction between the test for determining “single employer” and the test for determining “joint- employer” status.12 In two cases decided in 1984- Laerco Transportation13 and TLI, Inc.14-the Board clar- ified the law by expressly adopting the Third Circuit’s joint-employer standard in NLRB v. Browning-Ferris Industries of Pennsylvania, Inc., 691 F.2d 1117, 1124 (3d Cir. 1982): “The basis of the [joint-employer] finding is simply that one employer while contracting in good faith with an otherwise independent company, has re- tained for itself sufficient control of the terms and condi- tions of employment of the employees who are employed by the other employer. Thus, the ‘joint employer’ con- cept recognizes that the business entities involved are in fact separate but that they share or co-determine those matters governing the essential terms and conditions of employment.” Applying this test as to “essential terms” in both Laerco and TLI, the Board stated it would focus on whether an alleged joint employer “meaningfully af- fects matters relating to the employment relationship such as hiring, firing, discipline, supervision, and direc- tion.”15 Both TLI and Laerco were cases applying the joint- employer test to the relationship between a company supplying labor to a company using it, the same business relationship at issue in the present case. The Board found that evidence of the “user” employer’s actual but “limited and routine” supervision and direction would 12 See, e.g., Parklane Hosiery Co., 203 NLRB 597, amended 207 NLRB 991 (1973). 13 269 NLRB 324 (1984). 14 271 NLRB 798 (1984), enfd. mem. 772 F.2d 894 (3d Cir. 1985). 15 Laerco, 269 NLRB at 325; TLI, 271 NLRB at 798. not suffice to establish joint-employer status.16 Subse- quently, in AM Property Holding Corp., 350 NLRB 998, 1001 (2007), the Board further explained that it has “generally found supervision to be limited and routine where a supervisor’s instructions consist primarily of telling employees what work to perform, or where and when to perform the work, but not how to perform the work.” In Airborne Express, 338 NLRB 597, 597 fn. 1 (2002), the Board explained that under the existing joint- employer test, “[t]he essential element in [the joint- employer] analysis is whether a putative joint employer’s control over employment matters is direct and immedi- ate.”17 Consistent with this rationale, in AM Property the Board found that a contractual provision giving the user company (AM) the right to approve hires by the supplier company (PBS) to work at AM’s office building was not, standing alone, sufficient to show AM’s status as a joint employer. Instead, “[i]n assessing whether a joint em- ployer relationship exists, the Board does not rely merely on the existence of such contractual provisions, but ra- ther looks to the actual practice of the parties.”18 The AM Property distinction between potential author- ity and the actual exercise of authority is a commonplace, well-established fixture in Board jurisprudence. For ex- ample, in the Board’s single-employer test, we have re- peatedly required proof that “one of the entities exercises actual or active control [as distinguished from potential control] over the day-to-day operations or labor relations of the other.”19 In other contexts where a party bears the burden of proving that an entity falls within a particular statutory definition, members of today’s majority have endorsed this evidentiary distinction, giving weight only to the actual exercise of authority or control.20 16 Laerco, 269 NLRB at 326; TLI, 271 NLRB at 799. Laerco and TLI were decided by different 3-member panels of a Board then com- prised of four sitting members. As such, they collectively represented the unanimous opinion of the full Board at that time. 17 We note that, although concurring Member Liebman advocated revisiting the joint-employer standard represented by TLI, she expressly agreed with the majority that Board decisions applying this precedent “have required that the joint employer’s control over these matters be direct and immediate.” 338 NLRB 597, 597 fn. 1. The majority here is completely mistaken in asserting that the focus on “direct and immedi- ate control” was a new addition to the Browning-Ferris joint-employer test in Airborne. Further, as we shall later explain, there is ample prec- edent in the common law for this requirement predating 1984. 18 350 NLRB at 1000. 19 Mercy Hospital of Buffalo, 336 NLRB 1282, 1284 (2001). See also, e.g., Dow Chemical Co., 326 NLRB 288 (1998); Gerace Con- struction, Inc., 193 NLRB 645 (1971); Los Angeles Newspaper Guild, Local 69, 185 NLRB 303, 304 (1970). 20 E.g., FedEx Home Delivery, 361 NLRB No. 165, slip op. at 14 (2014) (“The Board has been careful to distinguish between actual opportunities, which allow for the exercise of genuine entrepreneurial BFI NEWBY ISLAND RECYCLERY 25 As discussed in section III below, the current test is fully consistent with the common-law agency principles that the Board must apply in determining joint-employer status. Further, as an administrative law judge has accu- rately summarized, the test reflects a commonsense, practical understanding of the nature of contractual rela- tionships in our modern economy. “An employer receiv- ing contracted labor services will of necessity exercise sufficient control over the operations of the contractor at its facility so that it will be in a position to take action to prevent disruption of its own operations or to see that it is obtaining the services it contracted for. It follows that the existence of such control, is not in and of itself, suffi- cient justification for finding that the customer-employer is a joint employer of its contractor’s employees.”21 II. THE MAJORITY’S NEW JOINT-EMPLOYER TEST The majority today expressly overrules TLI, Laerco, Airborne Express, AM Property, supra and related prece- dent, and purports to return to a joint-employer test that allegedly applied prior to this line of precedent. Their analysis begins in a manner that is consistent with the Board’s modern precedent: “The Board may find that two or more entities are joint employers of a single work force if they are both employers within the meaning of the common law, and if they share or codetermine those matters governing the essential terms and conditions of employment.” The “share or codetermine” language is the general statement of the joint-employer test in Browning-Ferris that was adopted and applied by the Board in both TLI and Laerco. Our colleagues go on to adopt TLI and Laerco’s description of essential terms and conditions of employment as “matters relating to the em- ployment relationship such as hiring, firing, discipline, supervision, and direction.” If this was the extent of the majority’s holding, there would be no need to overrule precedent. However, the majority’s decision makes clear that the new test expands joint-employer status far beyond any- thing that has existed under current precedent and, con- autonomy, and those that are circumscribed or effectively blocked by the employer.”); Pacific Lutheran University, 361 NLRB No. 157, slip op. at 24 (2014) (“In order for decisions in a particular policy area to be attributed to the faculty, the party asserting managerial status must demonstrate that faculty actually exercise control or make effective recommendations.”); and Lucky Cab Co., 360 NLRB No. 43, slip op. at 3 (2014) (“We reject, therefore, the judge’s reliance on ‘paper authori- ty’ set forth in the handbook, in light of the contrary evidence of the road supervisors’ actual practice. Schnurmacher Nursing Home v. NLRB, 214 F.3d 260, 265 (2d Cir. 2000), enfg. in relevant part 327 NLRB 253 (1998) (no authority to discipline, despite statement in job description, where the alleged supervisors did not actually discipline or recommend discipline).”). 21 Southern California Gas Co., 302 NLRB 456, 461 (1991). trary to the majority’s claim, under precedent predating TLI and Laerco. In a two-step progression, the first of which misleadingly depicts the limits of common law, the majority removes all limitations on what kind or de- gree of control over essential terms and conditions of employment may be sufficient to warrant a joint- employer finding: [W]e will no longer require that a joint employer not only possess the authority to control employees’ terms and conditions of employment, but must also exercise that authority, and do so directly, immediately, and not in a “limited and routine” manner. . . . The right to con- trol, in the common-law sense, is probative of joint- employer status, as is the actual exercise of control, whether direct or indirect. Moreover, the new test will evaluate the exercise of control by construing “share or codetermine” broadly: In some cases (or as to certain issues) employers may engage in genuinely shared decision-making, e.g., they confer or collaborate to set a term of employment. . . . Alternatively, employers may exercise comprehensive authority over different terms and conditions of em- ployment. For example, one employer sets wages and hours, while another assigns work and supervises em- ployees. . . . Or employers may affect different compo- nents of the same term, e.g. one employer defines and assigns work tasks, while the other supervises how those tasks are carried out. . . . Finally, one employer may retain the contractual right to set a term or condi- tion of employment. [Emphasis added.] Our colleagues concede “it is certainly possible that in a particular case a putative joint employer’s control might extend only to terms and conditions of employment too limited in scope or significance to permit meaningful collec- tive bargaining.” However, the majority fails to provide any guidance as to what control, under what circumstances, would be insufficient to establish joint-employer status. What do the preceding passages and the overruling of cited precedent indicate? First, in any particular case, the majority may consider evidence about virtually any as- pect of employment and may give dispositive weight to an employer’s control over any essential term and condi- tion of employment in finding a joint-employer relation- ship. Second, there will be no requirement that control over any essential term of employment be “direct and immediate” in order for it to be probative and potentially determinative. Indirect control, even a power reserved by contract but never exercised, will be considered and may suffice, standing alone, to find joint-employer sta- 26 DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD tus. Finally, while the majority purports to base its standard on the common law and “sufficient control . . . to permit meaningful collective bargaining,” it remains to be seen whether even the occasional limited and routine discussion or collaboration about a single essential term of employment may suffice to establish joint-employer status. The majority repeatedly states that almost every aspect of a business relationship may be probative, but it provides no significant guidance as to what may or should be determinative. The majority’s new test represents a major unex- plained departure from precedent. This test promises to effect a sea change in labor relations and business rela- tionships. Our colleagues presumably do not intend that every business relationship necessarily entails the joint employment of every entity’s employees, but there is no limiting principle in their open-ended multifactor stand- ard. It is an analytical grab bag from which any scrap of evidence regarding indirect control or incidental collabo- ration as to any aspect of work may suffice to prove that multiple entities-whether they number two or two doz- en-“share or codetermine essential terms and conditions of employment.” III. THE MAJORITY’S NEW TEST IMPERMISSIBLY DEPARTS FROM THE COMMON-LAW AGENCY TEST AND RESURRECTS THE CONGRESSIONALLY-REJECTED ECONOMIC REALITY AND BARGAINING INEQUALITY THEORIES A. The Majority’s Implicit Reliance on Economic Reality and Statutory Purpose Theory Directly Contravenes Congressional Intent The threshold insurmountable problem with the major- ity’s reformulated joint-employer test is that it far ex- ceeds the limits of our statutory authority.22 In fact, this is the third case decided recently where Board majorities have tested or exceeded those limits when dramatically expanding “employer” and “employee” status. In FedEx Home Delivery, 361 NLRB No. 55 (2014), the majority claimed to be applying the common law when it broadened the Act’s definition of “employee,” which (based on language added in 1947 as part of the Taft-Hartley amendments) explicitly excludes any “inde- pendent contractor.”23 In altering the analysis for distin- 22 The majority cites the following passage from American Trucking Assns. v. Atchison, T. & S.F. Ry. Co., 387 U.S. 397, 416 (1967), pur- porting to justify the change in the joint-employer standard: “[Regula- tory agencies] are supposed, within the limits of the law and of fair and prudent administration, to adapt their rules and practices to the Na- tion’s needs in a volatile, changing economy.” Id. (emphasis supplied). As hereafter discussed, the change in the joint-employer standard is neither within the limits of the law nor representative of fair and pru- dent administration. 23 Sec. 2(3). guishing employees from independent contractors, the majority distorted the common-law test to emphasize the perceived economic dependency of the putative employ- ee on the putative employer. Member Johnson’s dissent explained that the majority’s treatment of “employee” and “independent contractor” status in FedEx was contra- ry to the Act and its legislative history, and the majori- ty’s factual findings were contrary to the record.24 In CNN America, Inc., 361 NLRB No. 47 (2014), the majority concluded that a client (CNN) was a joint em- ployer of technical employees supplied by a contractor (TVS), although CNN undisputedly had no direct role in hiring, firing, disciplining, discharging, promoting, or evaluating TVS’ employees, and CNN’s “employer” status was contrary to the TVS collective-bargaining agreements, the services agreement entered into between CNN and TVS, two decades of bargaining history and CBAs (all identifying the contractor as the only “em- ployer”), and prior union certifications by the Board. The Board majority, though ostensibly applying the tra- ditional joint-employer test, relied on factors similar to those emphasized by the majority here (e.g., finding that CNN’s services agreement gave it “considerable authori- ty” over “staffing levels”). Member Miscimarra’s dissent explained that the Board and the courts had long dealt with situations where contractor employees worked at client locations, with substantial interaction between the client and contracting employer, without conferring “employer” status on the client. CNN America, Inc., slip op. at 28, 31-32 (citing NLRB v. Denver Building Trades Council, supra, 341 U.S. at 692; and Fibreboard Corp. v. NLRB, supra, 379 U.S. at 203 (other citations omitted)).25 In this case, our colleagues abandon extant joint- employer law, which had already been strained beyond its rational breaking point in CNN. Instead, similar to what was done in FedEx for the definition of a statutory employee, they have announced a new test of joint- employer status that, notwithstanding their adamant dis- claimers, effectively resurrects and relies, at least in sub- stantial part, on intertwined theories of “economic reali- ties” and “statutory purpose” endorsed by the Supreme Court in NLRB v. Hearst Publications, 322 U.S. 111 (1944), which Congress expressly rejected in the Taft- Hartley Amendments of 1947. In Hearst, the Court ap- 24 Member Miscimarra did not participate in FedEx, but he agrees with Member Johnson’s criticism of the economic realities test applied by the majority and the analysis of “employee” and “independent con- tractor” issues addressed in Member Johnson’s dissent. 25 Member Johnson did not participate in CNN, but he agrees with the criticism of the majority’s joint-employer finding as expressed in Member Miscimarra’s dissent. BFI NEWBY ISLAND RECYCLERY 27 plied the same rationale for the definitions of employee and employer under the original Wagner Act. To eliminate the causes of labor disputes and industrial strife, Congress thought it necessary to create a balance of forces in certain types of economic relationships. These do not embrace simply employment associations in which controversies could be limited to disputes over proper “physical conduct in the performance of the ser- vice.” On the contrary, Congress recognized those economic relationships cannot be fitted neatly into the containers designated “employee” and “employer” which an earlier law had shaped for different purposes. Its Reports on the bill disclose clearly the understand- ing that “employers and employees not in proximate re- lationship may be drawn into common controversies by economic forces, and that the very disputes sought to be avoided might involve “employees (who) are at times brought into an economic relationship with em- ployers who are not their employers.” In this light, the broad language of the Act’s definitions, which in terms reject conventional limitations on such conceptions as “employee,” “employer,” and “labor dispute,” leaves no doubt that its applicability is to be determined broadly, in doubtful situations, by underlying economic facts rather than technically and exclusively by previ- ously established legal classifications.26 In reaction to Hearst, Congress expressly excluded “independent contractors” from the Act’s definition of a statutory employee in the Taft-Hartley Amendments of 1947. The purpose of this revision was manifest in the legislative history of the Amendments and repeatedly acknowledged thereafter by the Supreme Court, which stated in one case that [in Hearst] the standard was one of economic and poli- cy considerations within the labor field. Congressional reaction to this construction of the Act was adverse and Congress passed an amendment specifically excluding ‘any individual having the status of an independent contractor’ from the definition of ‘employee’ contained in s 2(3) of the Act. The obvious purpose of this amendment was to have the Board and the courts apply general agency principles in distinguishing between employees and independent contractors under the Act. . . . Thus there is no doubt that we should apply the 26 322 U.S. at 128-129. See also United States v. Silk, 331 U.S. 704 (1947), applying the same economic realities and statutory purpose theories to the definition of employee under the Social Security Act. common law agency test here in distinguishing an em- ployee from an independent contractor.27 Our colleagues nevertheless cling to the notion that economic and policy considerations may determine the definition of employee and employer. Even assuming that may be true in some cases not dealing with the right to control under common law,28 the Supreme Court squarely rejected reliance on these considerations in Darden, stating that Hearst and Silk, which interpreted “employee” for pur- poses of the National Labor Relations Act and Social Security Act, respectively, are feeble precedents for unmooring the term from the common law. In each case, the Court read “employee,” which neither statute helpfully defined, to imply something broader than the common-law definition; after each opinion, Congress amended the statute so construed to demonstrate that the usual common-law principles were the keys to meaning. . . . To be sure, Congress did not, strictly speaking, “overrule” our interpretation of those stat- utes, since the Constitution invests the Judiciary, not the Legislature, with the final power to construe the law. But a principle of statutory construction can en- dure just so many legislative revisitations, and Reid’s presumption that Congress means an agency law defi- nition for “employee” unless it clearly indicates other- wise signaled our abandonment of Silk’s emphasis on construing that term “‘in the light of the mischief to be corrected and the end to be attained.’” [503 U.S. at 324-325 (footnote and citations omitted).] Accordingly, the inescapable conclusion to be drawn from the Taft-Hartley legislation repudiating the Hearst opinion is that Congress must have intended that com- mon-law agency principles, rather than the majority’s much more expansive policy-based economic realities and statutory purpose approach, here govern the defini- tion of employer as well as employee under the Act. Even if Congress had not been so clear, “it is . . . well established that ‘[w]here Congress uses terms that have accumulated settled meaning under . . . the common law, a court must infer, unless a statute otherwise dictates, that Congress means to incorporate the established mean- ing of these terms.’” Community for Creative Non- Violence v. Reid, 490 U.S. 730, 739 (1989) (quoting NLRB v. Amax Coal Co., 453 U.S. 322, 329 (1981)). 27 NLRB v. United Insurance Co. of America, 390 U.S. 254, 256 (1968). See also Boire v. Greyhound, supra, 376 U.S. at 481 fn. 10, and Nationwide Mutual Insurance Co. v. Darden, supra, 503 U.S. at 324. 28 See, e.g., Allied Chemical Workers Local Union 1 v. Pittsburgh Plate Glass Co., 404 U.S. 157, 168 (1971). 28 DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD Thus, the majority’s new joint-employer test is invalid if it does not comport with the common-law agency princi- ples. Nevertheless, our colleagues now expand the defini- tion of employer by redefining the joint-employer doc- trine in unstated-but unmistakable-reliance on the rationale of Hearst that was repudiated by Congress.29 Our colleagues are motivated by a policy concern that an imbalance of leverage reflected in commercial dealings between the undisputed employer and third-party entities prevents “meaningful bargaining” over each term and condition of employment and is therefore in conflict with the statutory policy of encouraging collective bargaining. This approach reflects a desire to ensure that third parties that have “deep pockets,” compared to the immediate employer, become participants in existing or new bar- gaining relationships, and that they will also be directly exposed to strikes, boycotts and other economic weap- ons, based on the most limited and indirect signs of po- tential control.30 Whether this is good or bad policy- and we think it is bad for numerous reasons discussed below-this fundamental balancing of interests has al- ready been done by Congress. And the simple fact is that Congress has forbidden the Board from applying an eco- 29 An unacknowledged antecedent for the joint-employer theory adopted here is the concurring opinion of then-Member Liebman in Airborne Express, supra, 338 NLRB at 597-599, who contended that “[g]iven business trends driven by accelerating competition, highlight- ed by this case, the Board’s joint-employer doctrine may no longer fit economic realities.” See also AM Property Holding Co., supra, 350 NLRB at 1012 (Member Liebman, concurring in part and dissenting in part). We note as well that the General Counsel relies on Hearst and eco- nomic reality theory in his amicus brief. The majority expressly rejects the General Counsel’s argument, but implicitly relies on much of it. While we disagree with the General Counsel as to the need and basis for overruling the existing joint-employer test, we respect his efforts to address these important issues, which have broad ramifications that extend well beyond this particular case. We also commend his sub- stantial public outreach efforts regarding these important proposed changes. 30 See Michael Harper, Defining the Economic Relationship Appro- priate for Collective Bargaining, 39 Boston College L. Rev. 329, 348 (1998) (“[I]f workers are to be assured the opportunity to utilize collec- tive bargaining leverage to extract a greater share of the returns from their labor, they must be able to bargain with the firms that provide the capital.”); see also Craig Becker, Labor Law Outside the Employment Relation, 74 Texas L. Rev. 1527 (1996) (“At bottom, my intent is to inquire how the principles of labor law might be freed from the limits of outmoded definitions of the employment relationship. That effort involves questioning the sanctity of the doctrine of privity of contract as well as departing from the common-law paradigm of master-servant as foundations for rights and duties in the workplace. Above all, it requires rethinking the nature of power at stake in labor relations so as to bring legal doctrine in line with contemporary economic realities.”) (Emphasis added). nomic realities or statutory purpose rationale in defining employer and joint-employer status under the Act. B. The Majority’s New Test does not Comport with Common-Law Agency Principles Our colleagues do not acknowledge the Congressional rejection of Hearst’s economic realities theory for defin- ing “employee” and “employer” under the Act. Neither do they acknowledge their implicit reliance on this theo- ry in announcing a new joint-employer test. Instead, they attempt, as they must, to persuade that their test of joint-employer status is consistent with common-law agency’s master-servant doctrine. The attempt fails. The “touchstone” at common law is whether the puta- tive employer sufficiently controls or has the right to control putative employees. See Clackamas Gastroen- terology Associates, P.C. v. Wells, 538 U.S. 440, 448- 449 (2003); Restatement (Second) of Agency §§ 2, 220 (1958). Without attribution, our colleagues state that the common law considers as potentially dispositive not only direct control, but also indirect control and even “re- served” control that has never been exercised. They would accordingly jettison the joint-employer test’s re- quirement of evidence that the putative employer’s con- trol be “direct and immediate.” As explained below, however, “control” under the common-law principles requires some direct-and-immediate control even where indirect control factors are deemed probative. The Act, and its incorporation of the common law, does not allow the Board to broaden the standard to include indirect control or an inchoate right to exercise control, standing alone, as a dispositive factor, which the majority does today. Long before Congress anchored “employer” in the common law, courts applying those principles focused on discerning whether the putative master had control over the details of the work (master) or only the results to be achieved (not master). See, e.g., Singer Mfg. Co. v. Rahn, 132 U.S. 518, 522 (1889) (“[T]he relation of mas- ter and servant exists whenever the employer retains the right to direct the manner in which the business shall be done, as well as the result to be accomplished, or, in oth- er words, ‘not only what shall be done, but how it shall be done.’” (quoting New Orleans, M&CR Co. v. Hanning, 82 U.S. 649, 657 (1872).) Further, the Su- preme Court has for over a century adhered to the propo- sition that “under the common law loaned-servant doc- trine immediate control and supervision is critical in de- termining for whom the servants are performing ser- vices.”31 Lower courts as well implicitly limited their 31 Shenker v. Baltimore & Ohio R. Co., 374 U.S. 1, 6 (1963), citing and applying the analysis in Standard Oil Co. v. Anderson, 212 U.S. BFI NEWBY ISLAND RECYCLERY 29 analysis to looking for direct-and-immediate control. See, e.g., Dimmitt-Rickhoff-Bayer Real Estate Co. v. Finnegan, 179 F.2d 882 (8th Cir. 1950) (not attaching any importance to indirect control in finding real estate agents were not employees), cert. denied 340 U.S. 823 (1950); Glenn v. Standard Oil Co., 148 F.2d 51 (6th Cir. 1945) (not attaching any importance to indirect control in finding operators of Standard Oil’s bulk distribution plants were not employees); Spillson v. Smith, 147 F.2d 727 (7th Cir. 1945) (not attaching any importance to in- direct control in finding the musicians of an orchestra were the employees of its leader and not the restaurant where they played). As courts undoubtedly realized, anyone contracting for services, master or not, inevitably will exert and/or re- serve some measure of indirect control by defining the parameters of the result desired to ensure he or she gets the benefit of his or her bargain. For example, Judge Learned Hand wrote, in a case applying common-law principles to decide a production company was not the employer of the entertainers in vaudeville acts under the Social Security Act, that [i]n the case at bar the plaintiff did intervene to some degree; but so does a general building contractor inter- vene in the work of his subcontractors. He decides how the different parts of the work must be timed, and how they shall be fitted together; if he finds it desirable to cut out this or that from the specifications, he does so. Some such supervision is inherent in any joint un- dertaking, and does not make the contributing contrac- tors employees. By far the greater part of [the putative employer’s] intervention in the ‘acts’ was no more than this. It is true, as we have shown, that to a very limited extent he went further, but these interventions were trivial in amount and in character; certainly not enough to color the whole relation. Radio City Music Hall Corp. v. United States, 135 F.2d 715, 717-718 (2d Cir. 1943). The Supreme Court subsequently addressed the same point in construing the coverage of the Act’s prohibition of coercive secondary activity against neutral construc- tion employers by unions: We agree with the Board also in its conclusion that the fact that the contractor and subcontractor were engaged on the same construction project, and that the contrac- tor had some supervision over the subcontractor’s 215 (1909). See also Kelly v. Southern Pacific Co., 419 U.S. 318, 329- 330 (1974), cited with approval in Community for Creative Non- Violence v. Reid, 490 U.S. at 739-740, and in Nationwide Mutual In- surance Co. v. Darden, 503 U.S. at 323. work, did not eliminate the status of each as an inde- pendent contractor or make the employees of one the employees of the other. The business relationship be- tween independent contractors is too well established in the law to be overridden without clear language doing so.32 To aid in applying this well-established common law for employer-employee relationships, the Supreme Court largely adopted the Restatement (Second) of Agency § 220’s nonexhaustive list of factors to be considered. Community for Creative Non-Violence v. Reid, 490 U.S. at 751-752; see also Nationwide Mutual Insurance Co. v. Darden, 503 U.S. at 323-324. The Reid Court wrote: In determining whether a hired party is an employee under the general common law of agency, we consider the hiring party’s right to control the manner and means by which the product is accomplished. Among the oth- er factors relevant to this inquiry are the skill required; the source of the instrumentalities and tools; the loca- tion of the work; the duration of the relationship be- tween the parties; whether the hiring party has the right to assign additional projects to the hired party; the ex- tent of the hired party’s discretion over when and how long to work; the method of payment; the hired party’s role in hiring and paying assistants; whether the work is part of the regular business of the hiring party; whether the hiring party is in business; the provision of employ- ee benefits; and the tax treatment of the hired party. Reid, 490 U.S. at 751-752. The inquiry remains the same. The factors provide useful indicia of the putative employer’s direct-and-immediate control, or its right to such control. The comments to Section 220 of the Restatement clari- fy that the listed factors are not looking to indirect con- trol. Comment j, on the duration of the relationship, pro- vides: “If the time of employment is short, the worker is less apt to subject himself to control as to details and the job is more likely to be considered his job than the job of the one employing him.”33 Comment k, on the source of the instrumentalities and tools, states it is understandable that the owner would regulate such instrumentalities be- cause “if the worker is using his employer’s tools or in- strumentalities, especially if they are of substantial value, it is normally understood that he will follow the direction of the owner in their use.” The same should hold true where one employer establishes rules for the use of its property. Comment l, on the location of work, informs 32 NLRB v. Denver Building Trades Council, supra, 341 U.S. at 689-690 (emphasis added). 33 We note here that Leadpoint is not supposed to keep its employ- ees assigned long term to the BFI project. 30 DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD that although the putative employer’s controlling the location of work usually raises an inference of employer status, “[i]f, however, the rules are made only for the general policing of the premises, as where a number of separate groups of workmen are employed in erecting a building, mere conformity to such regulations does not indicate that the workmen are” employees. Recently, courts applying the common law have con- tinued to make it unmistakably clear that the employer standard requires sufficient proof of direct-and- immediate control. In finding that the New York State Education Department was not the employer of teachers under Title VII, the United States Court of Appeals for the Second Circuit wrote: “[The common-law standard] focuses largely on the extent to which the alleged master has ‘control’ over the day-to-day activities of the alleged ‘servant.’ The Reid factors countenance a relationship where the level of control is direct, obvious, and con- crete, not merely indirect or abstract. . . . Plaintiffs in this case could not establish a master-servant relationship under the Reid test. [The State Education Department] does have some control over New York City school teachers-e.g., it controls basic curriculum and creden- tialing requirements-but SED does not exercise the workaday supervision necessary to an employment rela- tionship.” Gulino v. N.Y. State Education Department, 460 F.3d 361, 379 (2d Cir. 2006) (emphasis added), cert. denied 554 U.S. 917 (2008). Similarly, the United States Court of Appeals for the Ninth Circuit found, applying common-law principles, that Wal-Mart was not the joint employer of its suppliers’ employees where Wal-Mart did not have the right to an “immediate level of ‘day-to- day’ control.” Doe I v. Wal-Mart Stores, Inc., 572 F.3d 677, 682-683 (9th Cir. 2009) (quoting Vernon v. State, 10 Cal. Rptr. 3d 121 (Cal. Ct. App. 2004)). A few years later, the Supreme Court of California used the same language in finding a franchisor not liable under the Cali- fornia Fair Employment and Housing Act for a franchi- see supervisor’s harassment of an employee: “[T]raditional common law principles of agency and respondeat superior supply the proper analytical frame- work . . . . This standard requires ‘a comprehensive and immediate level of ‘day-to-day’ authority’ over matters such as hiring, firing, direction, supervision, and disci- pline of the employee.” Patterson v. Domino’s Pizza, LLC, 333 P.3d 723, 740 (Cal. 2014) (quoting Vernon, supra).34 34 In TLI, supra, 271 NLRB at 798, the Board stated that “there must be a showing that the employer meaningfully affects matters relating to the employment relationship such as hiring, firing, discipline, supervi- sion, and direction.” We read that passage to provide a nonexclusive list of direct-and-immediate control factors to consider, and hereafter Contrary to our colleagues’ characterization, the above-quoted language from Gulino and Wal-Mart can- not be dismissed as meaningless statements made “in cases where there was little if any relevant evidence of control of any sort.” This begs the question why either court felt the need to specifically mention the absence of immediate control. As for Patterson, the majority states (as do we) that the case was decided under a California statute, but they fail to acknowledge that the court’s opinion is founded on “traditional common law princi- ples of agency and respondeat superior.”35 The salient point is that the cases we cite do indicate that evidence of direct and immediate control is essential to a finding of joint-employer status under the common law. By con- trast, the majority does not and cannot cite a single judi- cial opinion that even implicitly affirms its concededly novel two-step version of an alternative common-law test or the proposition that a finding of a joint employer rela- tionship under the common law can be based solely on indirect control. In re Enterprise Rent-A-Car Wage & Employment Practices Litigation, 683 F.3d 462, 468-469 (3d Cir. 2012), provides a useful contrast between the common- law test of joint-employer status and the economic reali- ties test that Congress expressly authorized by the unique language of the Fair Labor Standards Act (FLSA), but rejected in the Taft-Hartley Amendments of our Act. we discuss cases decided after TLI that did examine factors other than those enumerated in that case. However, evidence of control over the specific factors referred to in TLI is usually most relevant to the joint- employer analysis. It is no coincidence that the Supreme Court of California used a similar list in Patterson, as did the Ninth Circuit in EEOC v. Pacific Maritime Assn., 351 F.3d 1270 (9th Cir. 2003). Dis- cussing the Supreme Court’s Clackamas decision in this Title VII case, the Court stated: The Supreme Court seems to suggest that the sine qua non of deter- mining whether one is an employer is that an “employer can hire and fire employees, can assign tasks to employees and supervise their per- formance.” Logically, before a person or entity can be a joint em- ployer, it must possess the attributes of an employer to some degree. Numerous courts have considered the key to joint employment to be the right to hire, supervise and fire employees. Id. at 1277. The Board’s task is to weigh all of the incidents of the relation- ship to determine the sufficiency of the control, and that analysis necessarily includes qualitative assessments of the general significance of specific fac- tors. The new test discards this safeguard against overinclusion in favor of finding any sporadic evidence or tangential effect on working conditions to be potentially sufficient to prove joint-employer status. 35 The majority also distinguishes Patterson on the ground that it involves “the particularized features of franchisor/franchisee relation- ships, none of which are applicable here.” As we state elsewhere in this opinion, the Board has heretofore maintained a unitary joint- employer test for all types of employer relationships. The suggestion that the test will vary from one type of relationship to another is un- precedented, and certainly has no foundation in the common law. BFI NEWBY ISLAND RECYCLERY 31 With respect to the economic realities test, the Third Cir- cuit stated: When determining whether someone is an employee under the FLSA, “economic reality rather than tech- nical concepts is to be the test of employment.” Gold- berg v. Whitaker House Co-op., Inc., 366 U.S. 28, 33, 81 S.Ct. 933, 6 L.Ed.2d 100 (1961) (internal quotation marks omitted). Under this theory, the FLSA defines employer “expansively,” Nationwide Mut. Ins. Co. v. Darden, 503 U.S. 318, 326, 112 S.Ct. 1344, 117 L.Ed.2d 581 (1992), and with “striking breadth.” Rutherford Food Corp. v. McComb, 331 U.S. 722, 730, 67 S.Ct. 1473, 91 L.Ed. 1772 (1947). The Supreme Court has even gone so far as to acknowledge that the FLSA’s definition of an employer is “the broadest def- inition that has ever been included in any one act.” United States v. Rosenwasser, 323 U.S. 360, 363 n. 3, 65 S.Ct. 295, 89 L.Ed. 301 (1945).36 The issue in Enterprise was whether the district court below erred in granting summary judgment against the plaintiff employees’ claim that the parent company of their wholly owned rental car subsidiary was their joint employer with shared liability for alleged overtime wage violations. The district court had relied on a traditional common-law test developed under the ADEA and Title VII. However, the Third Circuit opined that [b]ecause of the uniqueness of the FLSA, a determina- tion of joint employment “must be based on a consid- eration of the total employment situation and the eco- nomic realities of the work relationship.” A simple ap- plication of the [district court’s] test would only find joint employment where an employer had direct control over the employee, but the FLSA designates those enti- ties with sufficient indirect control as well. We there- fore conclude that while the factors outlined today in [that test] are instructive they cannot, without amplifi- cation, serve as the test for determining joint employ- ment under the FLSA.37 It is readily apparent from the distinctions underscored by the Enterprise court that the new joint-employer test an- nounced by our colleagues is rooted in economic reality and statutory purpose theory, not in the “technical concepts” of common-law agency. Indeed, their new definition of em- ployer equals or exceeds the “striking breadth” of the FLSA 36 Id. at 467-468. 37 Id. at 469. The court nevertheless affirmed the grant of summary judgment, finding insufficient proof that the parent company was a joint employer even under the expansive FLSA standard. It is not clear whether the same evidence considered under the majority’s test here would lead to the same result. standard, and it cannot stand in the face of express Congres- sional disapproval. The majority’s explication of its new joint-employer test erases any doubt that the test is the analytical step- child of Hearst, rather than being founded in common law. Our colleagues posit that as a first step they must determine whether an employment relationship exists at all between the alleged joint employer and an employee. Here, the majority does no more than acknowledge the obvious: an entity with no control whatsoever over a person performing services in that entity’s affairs cannot be that person’s employer. But the majority incorrectly sets this “zero control” state as the outer limit of common law master-servant agency, that is, if there is some con- trol over any aspect of the performance of services, then common law would allegedly permit finding an em- ployment relationship. Of course, if that were true, it would obliterate the common-law concept of an inde- pendent contractor and erase the distinction at common law between servant and nonemployee agent. The ma- jority seems vaguely to recognize this, but as far as de- ciding whether it should find that a separate business is a joint employer with an undisputed employer of an undis- puted employee, the majority nevertheless looks to whether it would serve the purposes of the Act to expand the joint-employer definition to serve the Act’s policy of “encouraging the practice and procedure of collective bargaining” (in the words of Sec. 1). In their view, it is necessary to do so because the current test’s “require- ments-which serve to significantly and unjustifiably narrow the circumstances where a joint employment rela- tionship can be found-leave the Board’s joint employ- ment jurisprudence increasingly out of step with chang- ing economic circumstances, particularly the recent dra- matic growth in contingent employment relationships. This disconnect potentially undermines the core protec- tions of the Act for the employees impacted by these economic changes.” Compare the majority’s reasoning to the following passages from Hearst concerning the test for determining whether newsboys were employees or independent con- tractors under the Wagner Act: Congress had in mind a wider field than the narrow technical legal relation of “master and servant,” as the common law had worked this out in all its variations, and at the same time a narrower one than the entire area of rendering service to others. The question comes down therefore to how much was included of the in- termediate region between what is clearly and unequiv- ocally ‘employment,’ by any appropriate test, and what is as clearly entrepreneurial enterprise and not em- 32 DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD ployment. . . . Myriad forms of service relationship, with infinite and subtle variations in the terms of em- ployment, blanket the nation’s economy. Some are within this Act, others beyond its coverage. Large numbers will fall clearly on one side or on the other, by whatever test may be applied. But intermediate there will be many, the incidents of whose employment par- take in part of the one group, in part of the other, in varying proportions of weight, . . . Unless the common- law tests are to be imported and made exclusively con- trolling, without regard to the statute’s purposes, it can- not be irrelevant that the particular workers in these cases are subject, as a matter of economic fact, to the evils the statute was designed to eradicate and that the remedies it affords are appropriate for preventing them or curing their harmful effects in the special situation. 322 U.S. 124-127 (fns. omitted). The only significant dif- ference between the majority’s reasoning here and the Court’s reasoning in Hearst is that the Court at least candid- ly recognized the “intermediate region” into which it ex- tended the Wagner Act’s definition of covered employees was beyond the scope of common law, while the majority blandly and disingenuously assures that the intermediate region into which they extend the definition of joint em- ployer stays well within the limits of that law. Clearly it does not. Contrary to our colleagues, we believe the Board’s traditional joint-employer test accurately reflects common law, and we disagree with any suggestion that their new test constitutes an appropriate way under common law to advance the statutory goal of promoting collective bar- gaining. Indeed, as we discuss below in section V, we find their test is more likely to destabilize collective bargaining than to promote it. IV. EVEN IF THE NEW TEST WERE PERMISSIBLE, THE MAJORITY FAILS TO IDENTIFY SUFFICIENT REASONS TO OVERRULE PRECEDENT AND ADOPT A NEW JOINT- EMPLOYER TEST A. The Majority’s Alleged Return to the Alleged “Tradi- tional Standard” Relies on a Selective Misreading of Precedent Before and After TLI and Laerco The majority states that the TLI and Laerco decisions “significantly and unjustifiably” narrowed the Board’s “traditional” joint-employer standard. This standard al- legedly encompassed far more factors, including those related to indirect control and reserved contractual con- trol, and more comprehensively analyzed employment relationships to determine whether an entity was a joint employer. However, in selecting only the few cases al- legedly supporting this view of traditional practice, the majority has neglected others where the Board found no joint-employer relationship, despite the presence of the “traditional” or “indirect control” factors that the majori- ty claims justify a finding of such a relationship. Contra- ry to the majority, the Board’s prior cases did not mani- fest an intention to apply a broad analytical framework in which indirect control played a determinative role in joint-employer cases. We agree with the majority that the Board has traditionally carried out a fact-intensive assessment of whether a putative employer exercised sufficient control over, or retained the right to control, the employees at issue. We disagree, however, with the notion that prior to TLI and Laerco the Board, as a rule, gave much probative weight to evidence of “indirect con- trol,” or that such evidence, standing alone, was routinely determinative. 38 We will now turn to a discussion of these factors of “indirect control.” This sentence is emblematic of the majority’s attempt to prove too much by the citation of the older cases: Thus, the Board’s joint-employer decisions found it probative that employers retained the contractual power to reject or terminate workers; set wage rates; set work- ing hours; approve overtime; dictate the number of workers to be supplied; determine “the manner and method of work performance”; “inspect and approve work,” and terminate the contractual agreement itself at will. [Footnotes omitted.] The foregoing statement includes footnote citations to precedent that allegedly shows that “the Board typically treated the right to control the work of employees and their terms of employment as probative of joint-employer status. The Board did not require that this right be exer- cised, or that it be exercised in any particular manner.” The majority fails to mention that in many of the cited cases there was evidence that the contractual rights were exercised, and there was other evidence of direct control over employees’ work. The majority’s statement also fails to account for all the Board cases that reach the con- trary result with similar contractual provisions. Thus, we can paraphrase the majority’s statement, with appropriate citations, that during the period preceding TLI and Laerco, the Board found no joint-employer status where putative “employers retained the contractual power to reject or terminate workers;39 set wage rates;40 set work- 38 Apart from our disagreement with the majority’s characterization of the joint-employer tests that existed prior to 1984, we note that in one major respect TLI and Laerco undisputedly broadened the circum- stances in which a joint-employer relationship could be found. That is, by adopting the Third Circuit’s Browning-Ferris joint-employer test, the Board made clear that the more restrictive single-employer test, requiring a showing of less than an arms-length relationship between employers, did not apply. 39 Cabot Corp., 223 NLRB 1388, 1390 fn. 10 (1976), affd. sub nom. Chemical Workers Local 483 v. NLRB, 561 F.2d 253 (D.C. Cir. 1977); BFI NEWBY ISLAND RECYCLERY 33 ing hours;41 approve overtime;42 determine ‘the manner and method of work performance’;43 ‘inspect and ap- prove work,’44 and terminate the contractual agreement itself at will.”45 Additionally, prior to TLI and Laerco the Board found that employers who conferred over the number of employees needed and the hours to be worked were not joint employers.46 The majority also states that prior to TLI and Laerco “the Board gave weight to a putative joint employer’s ‘indirect’ exercise of control over workers’ terms and conditions of employment,” citing Floyd Epperson, 202 NLRB 23, 23 (1973), enfd. 491 F.2d 1390 (6th Cir. 1974). However, it is readily apparent that, while the Board noted anecdotal evidence of the employer’s indi- rect control over wages and discipline in that case, its joint-employer finding was primarily based on evidence of direct and immediate supervision of the employees involved.47 Accordingly, in Fidelity Maintenance & Construction Co., supra, 173 NLRB at 1037, the Board emphasized direct control, saying that “the determinative factor in an owner contractor situation is whether the owner exercises or has the right to exercise sufficient direct control over the labor relations policies of the con- tractor, or over the wages, hours and working conditions” (emphasis added). Likewise, in The John Breuner Co., supra, 248 NLRB at 989, the Board affirmed without comment the administrative law judge’s observation that in prior truck delivery cases where the Board found joint- employer status, “there have always been supporting findings that the retailer or distributor by its supervisors, directly supervised and controlled the employees of his trucking contractor in the performance of their work” (emphasis added). Thus, contrary to the majority, Ep- person and like precedent support the proposition that findings of joint-employer status in cases prior to TLI and Laerco that mention evidence of indirect control nevertheless turn on sufficient proof of direct control. Hychem Constructors, Inc., 169 NLRB 274, 276 (1968); Westinghouse Electric Corp., 163 NLRB 914 (1967); Space Services International Corp., 156 NLRB 1227, 1232 (1966). 40 Cabot, supra; Hychem, supra at fn. 4; Fidelity Maintenance & Construction Co., 173 NLRB 1032, 1037 (1968). 41 S. G., Tilden, Inc., 172 NLRB 752 (1968). 42 Hychem, supra at 276. 43 S. G., Tilden, Inc., supra. 44 Cabot, supra at 1392; Westinghouse, supra at 915. 45 Space Services, supra at fn. 23. 46 The John Breuner Co., 248 NLRB 983, 989 (1980); Furniture Distribution Center, 234 NLRB 751, 751-752 (1978). 47 Id. (“United establishes the work schedule of the drivers, has the authority to make changes in the drivers’ assignments, selects routes for the drivers, and generally supervises the drivers in the course of their employment.”). The majority also contends that “[c]ontractual ar- rangements under which the user employer reimbursed the supplier for workers’ wages or imposed limits on wages were also viewed as tending to show joint- employer status,” citing Hamburg Industries, 193 NLRB 67 (1971). Hamburg concerned a typical cost-plus con- tract where the user employer reimbursed the supplier employer for wages and then paid an additional fee. The Board has cited this factor in cases where the Board found joint-employer status. However, the Board has also found that this factor did not establish joint- employer status.48 In any event, as explained in a subse- quent case, the facts in Hamburg clearly demonstrated significant direct and immediate control of essential terms was exercised by the disputed employer. Specifi- cally, “one employer, a manpower supplier, furnished another employer’s entire work force, including first- level supervisors. That work force was subject to virtual- ly complete control of the second employer. The second employer determined which tasks were to be performed and how they were to be performed. He also, in practice, set the wage rates.”49 Again, before TLI and Laerco, there was no established rule that cost-plus contracts should be given determinative weight in finding joint- employer status. In sum, the precedent cited by the majority falls well short of showing that prior to TLI and Laerco there was a consistently applied “traditional joint-employer test” remotely equivalent to the one they announce today. The indirect control factors cited by the majority existed in many cases where the Board refused to find joint- employer status and thus were not frequently, much less routinely, determinative of joint-employer status. Evi- dence of direct and immediate control was far more often referenced as determinative in finding such status.50 The interpretive key to different outcomes in this precedent is not due to a markedly different legal test; it is simply that 48 See Hychem, supra at 276 (referring to controls under a cost-plus contract as a “right to police reimbursable expenses under its cost-plus contract and do not warrant the conclusion that [user] has hereby forged an employment relationship”); Westinghouse, supra at 915 (cost-plus contract and no joint-employer finding); Space Services, supra at 1232 (cost-plus and no joint-employer finding); Cabot, supra at 1389 (“[C]ost plus contracts merely insured that Cabot obtain a satisfactory work product at cost and protected it against unnecessary charges being incurred.”); International House, supra at 914 (cost-plus “purely arms length dealing”); John Breuner, supra at 988 (cost-plus insufficient to find joint employer). 49 Cabot, supra, 223 NLRB at 1391 fn. 11. 50 We recognize that dictum in Airborne Freight stated that “ap- proximately 20 years ago, the Board, with court approval, abandoned its previous test in this area, which had focused on a putative joint employer’s indirect control over matters relating to the employment relationship.” 338 NLRB at 597 fn. 1. For the reasons just stated, we find this dictum to be a mistaken characterization of general precedent. 34 DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD “minor differences in the underlying facts might justify different findings on the joint-employer issue.” North American Soccer League v. NLRB (NASL), 613 F.2d 1379, 1382 (5th Cir. 1980), cert. denied 449 U.S. 899 (1980); see also Carrier Corp. v. NLRB, 768 F.2d 778, 781 fn. 1 (6th Cir. 1985) (distinguishing TLI and Laerco by noting that a slight difference between two cases can tilt one toward a joint-employer finding, and the court was not deciding those other cases). B. There Is No Judicial Precedent Adverse to the Board’s Current Joint-Employer Standard or Supportive of the Majority’s New Standard It is reasonable to assume that if TLI, Laerco, and progeny departed abruptly from Board precedent without explanation, reviewing courts would by now have had the opportunity to criticize those decisions and would certainly have done so. After all, the Supreme Court and various appellate courts have warned the Board against such unexplained changes. See Allentown Mack Sales & Services v. NLRB, 522 U.S. 359, 375 (1998) (“The evil of a decision that applies a standard other than the one it enunciates spreads in both directions, preventing both consistent application . . . and effective review of the law by the courts.”); NLRB v. Curtin Matheson Scientific, Inc., 494 U.S. 775, 799 (1990) (Blackmun, J., dissenting) (finding the Board had departed from prior standard “without explanation”); Bath Marine Draftsmen’s Assn. v. NLRB, 475 F.3d 14, 25 (1st Cir. 2007) (stating that when “the Board has not been consistent in its choice of standard, as explained above . . . . the Board is not enti- tled to the normal deference we owe it”); LeMoyne-Owen College v. NLRB, 357 F.3d 55, 61 (D.C. Cir. 2004) (“Re- quiring an adequate explanation of apparent departures from precedent thus not only serves the purpose of ensur- ing like treatment under like circumstances, but also fa- cilitates judicial review of agency action in a manner that protects the agency’s predominant role in applying the authority delegated to it by Congress.”). As LeMoyne noted, courts are duty-bound to strike down Board deci- sions that lack explanation or are otherwise arbitrary and capricious in their exercise of statutory authority. In this context, the Board’s direct and immediate con- trol standard has held up well over the last 30 years. While some courts may vary from the Board as to the particulars of a joint-employer test, others have expressly approved or applied the Board’s test, and none have di- rectly criticized that test or reversed a Board decision based on application of that test. Significantly, two of the four Board decisions express- ly overruled by the majority today were reviewed by a court of appeals, and both decisions were upheld. The decision in TLI was reviewed by a panel of the Third Circuit, the original Browning-Ferris circuit, and sum- marily affirmed in an unpublished decision.51 Likewise, the decision in AM Property was reviewed and affirmed by a panel of the Second Circuit.52 In accord with its own precedents, which date to before the issuance of TLI and Laerco, the court expressly endorsed the Board’s standard requiring that “‘an essential element’ of any joint-employer determination is ‘sufficient evidence of immediate control over the employees.’”53 The court specifically supported the Board’s finding that “limited and routine” supervision is insufficient to establish joint- employer status. The cases the Board relied on broadly support the proposition that ‘limited and routine’ supervision, G. Wes Ltd., 309 NLRB at 226, consisting of ‘directions of where to do a job rather than how to do the job and the manner in which to perform the work,’ Island Creek Coal, 279 NLRB at 864, is typically insufficient to create a joint employer relationship. See also Local 254, Serv. Emps. Intern. Union, AFL-CIO, 324 N.L.R.B. 743, 746-49 (1997) (no joint employer rela- tionship where employer regularly directed mainte- nance employees to perform specific tasks at particular times but did not instruct employees how to perform their work); S. Cal. Gas Co., 302 N.L.R.B. 456, 461- 62 (1991) (employer’s direction of porters and janitors insufficient to establish joint employer relationship where employer did not, inter alia, affect wages or ben- efits, or hire or fire employees). Id. at 443. Thus, the Second Circuit has explicitly endorsed the Board’s joint-employer standard. Further, as noted in an earlier case from the same circuit, other courts of appeals have varying standards for determining joint-employer status, but “[w]e see no need to select among these ap- proaches or to devise an alternative test, because we find that an essential element under any determination of joint-employer status in a sub-contracting case is dis- tinctly lacking in the instant case-some evidence of im- mediate supervision or control of the employees.”54 It is most noteworthy that, in addition to the absence of any circuit court precedent in conflict with the Board’s current legal test of joint-employer status, there also is no circuit court precedent in support of the new two-step 51 Teamsters Local 326 v. NLRB, 772 F.2d 894 (3d Cir. 1985). 52 Service Employees, Local 32BJ v. NLRB, 647 F.3d 435 (2d Cir. 2011), aff. in relevant part, enf. in part and denying in part on other grounds 350 NLRB 998. 53 Id. at 443 (quoting Clinton’s Ditch Co-op Co. v. NLRB, 778 F.2d 132, 138 (2d Cir.1985)). 54 International House v. NLRB, 676 F.2d 906, 913 (2d Cir. 1982) (emphasis added). BFI NEWBY ISLAND RECYCLERY 35 legal test articulated by our colleagues. That test, with- out any requirement that an alleged joint employer’s con- trol over those terms be significant or substantial, much less direct and immediate, most closely resembles a sin- gle Board decision’s bizarre distortion of dictum from an Eighth Circuit opinion in NLRB v. New Madrid Mfg. Co., 215 F.2d 908 (1954). In New Madrid, the court denied enforcement of a Board order to the extent that it relied on finding that a company selling its business to an individual remained a coemployer with him. Finding no substantial evidence to support the Board’s contrary finding, the court reasoned, inter alia, that provisions in the contract of sale did not demonstrate a retention of control over the successor’s operations. In particular, the court stated that the con- tract did not “either expressly or by implication, purport to give New Madrid any voice whatsoever in the select- ing or discharging of Jones’ employees, in the fixing of wages for such employees, or in any other element of labor relations, conditions and policies in the plant pur- chaser’s business.” Id. at 913. Thereafter, in Hoskins Ready-Mix Concrete, 161 NLRB 1492 (1966), a Board panel affirmed an adminis- trative law judge’s finding that a cement company and a company leasing trucks and drivers to it were joint em- ployers. In doing so, the Board focused on the lessee’s controls in the parties’ lease and operating agreements. In a footnote citation to New Madrid, the Board convert- ed the aforementioned dictum from negative to positive, incorrectly claiming that the court’s test of co-ownership was whether a contract gave the disputed employer “any voice whatsoever” over terms and conditions of em- ployment.55 This was not then and is not now the joint- employer test of the Eighth Circuit56 or any other court of appeals. It was not then the Board’s joint-employer test, and has not thereafter been the test. Until now, that is. Of course, the Board is free to go its own way and de- termine its own standards, but only within the statutory framework and with adequate explanation of the reasons for departing from long-established precedent. The ma- jority claims that 30 years ago the Board departed with- out explanation from prior precedent by drastically re- stricting its test in a way that denies many workers their Section 7 rights. However, the absence of any judicial criticism of the legal test consistently applied since then undermines this claim. It is simply impossible that all the courts of appeals would have missed this train wreck. 55 Id. at 1493 fn. 2. 56 The Eighth Circuit uses a four-factor test similar to a single- employer analysis. E.g., Industrial Personnel Corp. v. NLRB, 657 F.2d 226, 229 (8th Cir. 1981). In any event, it remains the majority’s burden to rational- ize its new test. V. THE MAJORITY’S NEW JOINT-EMPLOYER TEST IS IMPERMISSIBLY VAGUE AND OVERBROAD AND WILL HAVE SUBSTANTIAL ADVERSE CONSEQUENCES A. The New Test Is Fatally Ambiguous, Providing No Guidance as to When and How Parties May Contract for the Performance of Work Without Being Viewed as Joint Employers Multifactor tests, like the common-law agency stand- ard that we must apply here, are vulnerable to an analysis that can be impermissibly unpredictable and results- oriented. As then-Judge Roberts remarked about the standard for determining whether college faculty are managerial employees under the Act: The need for an explanation is particularly acute when an agency is applying a multi-factor test through case- by-case adjudication. The open-ended rough-and- tumble of factors on which Yeshiva launched the Board and higher education can lead to predictability and in- telligibility only to the extent the Board explains, in ap- plying the test to varied fact situations, which factors are significant and which less so, and why. . . . In the absence of an explanation, the totality of the circum- stances can become simply a cloak for agency whim- or worse.57 Our colleagues’ new multifactor test, in which any de- gree of indirect or reserved control over a single term is probative and may suffice to establish joint-employer status, is woefully lacking the required explanation of “which factors are significant and which less so, and why.” They provide no meaningful guidelines as to the test’s future application. Further, they acknowledge no legitimate grounds for parties in a business relationship to insulate themselves from joint-employer status under the Act. The new test stands in marked contrast to the current test’s focus on evidence of direct-and-immediate control of essential terms of employment, thereby establishing a discernible and rational line between what does and does not constitute a joint-employer relationship under the Act. The current longstanding test thereby recognizes that “[s]ignificant limits . . . exist upon what actions by an employer count as control over the means and manner of performance. Most important, employer efforts to monitor, evaluate, and improve the results or ends of the worker’s performances do not make the worker an em- 57 LeMoyne-Owen College v. NLRB, supra, 357 F.3d at 61 (citations and quotations omitted). 36 DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD ployee. Such global oversight, as opposed to control over the manner and means of performance (and espe- cially the details of that performance), is fully compatible with the relationship between a company and an inde- pendent contractor.”58 By comparison, our colleagues reference as probative all evidence of indirect control for such factors as the place of work, defining the work and how quickly it will need to be done, prescribing the hours when work will need to be performed, setting minimum qualifications for the individuals that the contractor provides and reserving the right to reject an individual (even though the contrac- tor may assign its employee to a different job), inspecting the contractor’s work, giving results-oriented feedback to the contractor that the contractor’s supervisors use in their directions to the contractor’s employees, agreeing to a price for the services that happens to be in the form of a cost-plus formula, and reserving the right to cancel the arrangement. Under the majority’s test, the homeowner hiring a plumbing company for bathroom renovations could well have all of that indirect control over a com- pany employee! By adopting such an overbroad, all- encompassing and highly variable test, our colleagues extend the Act’s definition of “employer” well beyond its common-law meaning, and beyond its ordinary meaning as well. Cf. Allied Chemical Workers Local 1 v. Pitts- burgh Plate Glass Co., supra, 404 U.S. at 168 (1971) (admonishing the Board for extending “employee” in the Act beyond its ordinary meaning by attempting to in- clude retired employees in its scope). The expansive nature of the new test is demonstrated by the evidence relied upon by the majority to find joint- employer status in this case, which involves a “cost-plus” arrangement that is common in user-supplier contracts between separate employers.59 The sum total of this evi- 58 North American Van Lines, Inc. v. NLRB, 869 F.2d 596, 599 (D.C. Cir. 1989) (citations omitted). 59 The Board and the courts have uniformly concluded that cost-plus arrangements do not automatically render the contracting client an “employer” of the vendor’s employees. Therefore, our colleagues concede (as they must) that a cost-plus “arrangement, on its own, is not necessarily sufficient to create a joint-employer relationship.” Indeed, the Board and the courts have uniformly concluded that nothing in cost- plus arrangements necessarily renders the contracting client an “em- ployer” of the vendor’s employees. In Fibreboard, for example, the contracting client (Fibreboard) arranged for employees of the contractor (Fluor) “to do the same [maintenance] work under similar conditions of employment,” where Fibreboard was committed to pay the “costs of the operation plus a fixed fee.” 379 U.S. at 206-207. As noted previously (see fn. 6, supra), Fibreboard was clearly treated as a distinct “employ- er” (having no employment relationship with the subcontractor’s em- ployees), even though the reasons underlying the subcontracting deci- sion were almost exclusively based on employment-related considera- tions. Indeed, the Supreme Court noted that Fibreboard “was induced to contract out the work by assurances from independent contractors dence is (1) a few contract provisions that indirectly af- fect the otherwise unfettered right of Leadpoint (the sup- plier-employer) to hire its own employees; (2) reports made by BFI representatives to Leadpoint of two inci- dents-one where a Leadpoint employee was observed passing a “pint of whiskey” at the jobsite, and another where a Leadpoint employee “destroyed” a drop box- that understandably resulted in discipline; (3) one con- tractually-established pay rate ceiling restriction for Leadpoint employees (obviously stemming from the cost-plus nature of the contract); (4) BFI’s control of its own facility’s hours and production lines; (5) a record- keeping requirement for Leadpoint employee hours (again, obviously stemming from the cost-plus nature of the contract); (6) a sole preshift meeting to advise Leadpoint supervisors of what lines will be running and what tasks they are supposed to do on those lines; (7) monitoring of productivity; (8) establishment of one type of generally applicable production assignment scheme for Leadpoint; and (9) “on occasion,” addressing Leadpoint employees about productivity directly. That is all there is, and the Regional Director correctly decided under extant law that it was not enough to show BFI was the joint employer of Leadpoint employees.60 The majority’s evidence amounts to a collection of general contract terms or business practices that are common to most contracting employers (discussed be- low), plus a few extremely limited BFI actions that had some routine impact on Leadpoint employees. It would be hard to find any two entities engaged in an arm’s- length contractual relationship involving work performed on the client’s premises that lack this type of interaction. Again, we suppose that our colleagues do not intend that every business relationship necessarily entails joint- employer status, but the facts relied upon here demon- strate the expansive, near-limitless nature of the majori- ty’s new standard. that economies could be derived by reducing the work force, decreasing fringe benefits, and eliminating overtime payments.” Id. at 213 (em- phasis added). The majority nevertheless attempts to distinguish the instant case because there was an “apparent requirement of BFI approval over em- ployee pay increases.” In this respect, the majority potentially confers “employer” status on every client/user company that enters into a cost- plus arrangement, because few, if any, clients will give a blank check to supplier-employers regarding wages when the full cost will be charged to the client. This is but one illustration of the multitude of ways that our colleagues fail to appreciate the “complexities of indus- trial life,” which is one of the Board’s most important functions and responsibilities. NLRB v. Insurance Agents, 361 U.S. 477, 499 (1960). 60 Although we might differ from the Regional Director as to the weight assigned to certain evidence, we find no need to do so where we agree with his ultimate finding. We note that the majority does not argue that the Regional Director erred in making this finding. BFI NEWBY ISLAND RECYCLERY 37 There is a further fundamental problem with the new joint-employer test. The majority states that its goal is to reach a large number of employees that they feel have been left unprotected by Section 7 because they work on a contingent or temporary basis. According to the major- ity, the number of workers so employed has dramatically risen since TLI and Laerco were decided and will pre- dictably continue to rise. Further, the majority asserts that “[t]he Board’s current focus on only direct and im- mediate control acknowledges the most proximate level of authority, which is frequently exercised by the suppli- er firm, but gives no consideration to the substantial con- trol over workers’ terms and conditions of employment of the user.” Thus, not only is the majority’s legal justification for a new joint-employer test impermissibly based on econom- ic reality theory, as previously discussed, but its factual justification is flawed as well. The majority focuses on facts limited to a particular type of business model-the user/supplier relationship involving the use of contingent employees-but they rely on these facts to justify a change in the statutory definition of employer, or joint employer, for all forms of business relationships between two or more entities. The number of contractual relationships now potential- ly encompassed within the majority’s new standard ap- pears to be virtually unlimited: • Insurance companies that require employers to take certain actions with employees in order to comply with policy requirements for safety, security, health, etc.; • Franchisors (see below); • Banks or other lenders whose financing terms may require certain performance measurements; • Any company that negotiates specific quality or product requirements; • Any company that grants access to its facilities for a contractor to perform services there, and then con- tinuously regulates the contractor’s access to the property for the duration of the contract; • Any company that is concerned about the quality of the contracted services; • Consumers or small businesses who dictate times, manner, and some methods of performance of con- tractors. Our point is not that the majority intends to make all players in the economy, no matter how small, necessary parties at the bargaining table (although as discussed below, they may well become targets of economic protest in support of bar- gaining or other union causes), but that the majority’s new standard foreshadows the extension of obligations under the Act to a substantial group of business entities without any reliable limitations.61 This kind of overbroad and ambigu- ous government regulation is necessarily arbitrary and ca- pricious. “In the absence of an explanation, the ‘totality of the circumstances’ can become simply a cloak for agency whim-or worse.” LeMoyne-Owen College v. NLRB, su- pra, 357 F.3d at 61. Our colleagues make this sweeping change in the law without any substantive discussion whatsoever of signifi- cant adverse consequences raised by BFI, Leadpoint, and amici. Indeed, they profess to limit themselves to the issue of joint bargaining obligations in the user-supplier context, with a disclaimer that their decision “does not modify any other legal doctrine or change the way that the Board’s joint-employer doctrine interacts with other rules or restrictions under the Act.” However, such a disclaimer cannot possibly be valid, because applying different tests in other circumstances would mark an un- precedented and unwarranted break from the unitary joint-employer test under our Act that has applied to all types of business relationships, each of which is affected by changing the basic joint-employer test. We therefore believe it is necessary to specifically address these con- sequences, and we do so below. B. The New Test Will Cause Grave Instability in Bar- gaining Relationships, Contrary to One of the Board’s Primary Responsibilities Under the Act Our colleagues greatly expand the joint-employer test without grappling with its practical implications for real- world collective-bargaining relationships. They purport to be following the command in Section 1 of the Act to “encourag[e] the practice and procedure of collective bargaining.” Congress did not mean, however, to blindly expand collective-bargaining obligations whether or not they are appropriate. The Act aims to “achiev[e] indus- trial peace by promoting stable collective-bargaining relationships.” Auciello Iron Works, Inc. v. NLRB, 517 U.S. 781, 790 (1996) (emphasis added). Indeed, one of the Board’s primary responsibilities under the Act is to foster labor relations stability. Colgate-Palmolive-Peet Co. v. NLRB, 338 U.S. 355, 362-363 (1949) (“To achieve stability of labor relations was the primary objec- tive of Congress in enacting the National Labor Relations Act.”); NLRB v. Appleton Electric Co., 296 F.2d 202, 206 (7th Cir. 1961) (“A basic policy of the Act [is] to achieve stability of labor relations.”). And the Supreme Court has stressed the need to provide “certainty before- 61 The majority correctly states that “the annals of Board precedent contain no cases that implicate the consumer services purchased by unsuspecting homeowners or lenders.” We hope that continues to be the case, but there is no guarantee that what is past is prologue under their new and impermissibly expansive test. 38 DECISIONS OF THE NAT hand” to employers and unions alike. Employers must have the ability to “reach decisions without fear of later evaluations labeling . . . conduct an unfair labor pra tice,” and a union similarly must be able to discern limits of its prerogatives, whether and when it could use its economic powers . . . , or whether, in doing so, it would trigger sanctions from the Board. Maintenance Corp. v. NLRB, supra, 452 U.S. at 678 679, 684-686 (emphasis added). Collective bargaining was intended by Congress to be a process that could conceivably produce agreements. One of the key analytical problems in widening the net of “who must bargain” is that, at some p predictably will not be achievable because different pa ties involuntarily thrown together as the under the majority’s new test will predictably have wid ly divergent interests. Today’s marked expansion of bargaining obligations to other business entities threatens to destabilize existing bargaining relationships and co plicate new ones. Even if one takes an extremely si plistic user-supplier scenario, the new standard ral of joint-employer status-making many clients “employer” of contractor employees, while making co tractors an “employer” jointly with the clients duce bargaining relationships and problems unlike any that have existed in the Board’s entire 80 which clearly were never contemplated or intended by Congress. Consider the following diagram, which depicts a single cleaning company named “CleanCo,” which has cleaning contracts with three clients. CleanCo employees work at each client’s facilities in circumstances similar to the instant case, and CleanCo periodically adds future cl ents. Assuming circumstances like those presented here, the majority would find that CleanCo and Client A are a “joint employer” at the Client A location; CleanCo and Client B are a “joint employer” at the Client B location; and CleanCo and Client C are a “joint employer Client C location. Such a situation-involving a single vendor and only three clients, each with only one loc IONAL LABOR RELATIONS BOARD c- “the ” First National - oint, agreements r- “bargainers” e- m- m- ’s confer- an n- -will pro- -year history, i- ” at the a- tion-creates all of the following problems under the majority’s test: 1. Union Organizing Directed at CleanCo. CleanCo employees are currently unrepresented and a union seeks to organize them, this gives rise to the fo lowing issues and problems: • What Bargaining Unit(s)? Although CleanCo direc ly controls all traditional indicia of employer status, the new majority test establishes that three different entities-Clients A, B, and C ployer” relationships with discrete and potentially overlapping groups of different CleanCo employees. It is unclear whether a single bargaining unit consis ing of all CleanCo employees could be considered appropriate, given the distinct role that the new m jority test requires each client to play in bargaining. • What “Employer” Participates in NLRB Election Proceedings? If the union files a representation pet tion with the Board, the Act requires the Board to a ford “due notice” and to conduct an hearing” for the “employer.” ly, the Board has no means of identifying less providing “due notice” and affording the right of participation to-“employer” entities like Clients A, B, and C, even though they would inherit bargaining obligations if CleanCo employees select the union. • Who Does the Bargaining? election involving all CleanCo employees, the m jority test would require participation in bargaining by CleanCo and Clients A, B, and C. Here, the m jority test provides that each party to bargain only with respect to ditions which it possesses the authority to control (emphasis added). However, because the majority standard is so broad-spanning “indirect control” and the “right to control never exercised in fact)-nobody could ever reaso ably know who is responsible for bargaining what. • CleanCo-Client Bargaining Disagreements. jority standard throws into disarray the manner in which “employers” such as CleanCo and Clients A, B, and C can formulate coherent vide meaningful responses to union demands, when they will undoubtedly disagree among themselves regarding many, if not most, matters that are the su ject of negotiation. Here, the majority disregards the fact that CleanCo’s client contr 62 We discuss this aspect of the “authority problem” in more detail below. If l- t- -have distinct “em- t- a- i- f- “appropriate Sec. 9(c)(1). Current- -much If the union wins an a- a- “will be required such terms and con- ” ’s “direct control,” ” (even if n- 62 The ma- proposals and pro- b- act will most often BFI NEWBY ISLAND RECYCLERY 39 have resulted from equally difficult negotiations with Clients A, B, and C. Therefore, the “joint” bar- gaining contemplated by the majority will involve significant disagreements between each of the em- ployer entities (i.e., Clean Co and Clients A, B, and C) with no available process for resolving such dis- putes.63 • CleanCo “Confidential” Information-Forced Dis- closure to Clients. The most contentious issue be- tween CleanCo and Clients A, B, and C is likely to involve the amounts charged by CleanCo, which predictably could vary substantially between Clients A, B, and C, depending on their respective leverage, the need for CleanCo’s services, the duration of their respective client contracts (i.e., whether short-term or long-term), and other factors. If a union success- fully organizes all CleanCo employees, the resulting bargaining-since the majority test requires partici- pation by Clients A, B, and C-will almost certainly require the disclosure of sensitive CleanCo financial information to Clients A, B, and C, which is likely to enmesh the parties in an array of disagreements with one another, separate from the bargaining between the union and the “employer” entities. • We have already found, in many prior cases, that this information is sensitive and is not necessary to employees’ exercise of rights under the Act. See, e.g., Flex Frac Logistics, 360 NLRB No. 120 (2014) (detailing disruption occurring when contractor, which “was particularly concerned to maintain the confidentiality of the rates it charges its clients,” had rates disclosed to clients by employee). The majori- ty’s new standard basically guarantees such econom- ic disruption for no legitimate purpose. • How Many Labor Contracts? If a single union or- ganizes all CleanCo employees, the above problems might be avoided if CleanCo engages in three sepa- rate sets of bargaining-each devoted to Client A, Client B, and Client C, respectively-resulting in three separate labor contracts. However, this would be inconsistent with the CleanCo bargaining unit if it encompassed all CleanCo employees, and CleanCo would violate the Act if it insisted on changing the scope of the bargaining unit, which under well- established Board law is a nonmandatory subject of bargaining. • What Contract Duration(s)? If a union represented all CleanCo employees, and if the Board certified 63 We also discuss this aspect of the “authority problem” in more detail below. each client location as a separate bargaining unit, then there presumably would be separate negotia- tions-and separate resulting CBAs-covering the CleanCo employees assigned to Client A, Client B, and Client C, respectively. In this case, however, the duration of each CBA might vary, depending on each side’s bargaining leverage, and a further com- plication would arise where CBA termination dates differ from the termination dates set forth in the var- ious CleanCo client contracts. • Do Client Contracts Control CBAs, or Do CBAs Control Client Contracts? Regardless of whether the CleanCo CBA(s) have termination dates that coin- cide with the expiration of the CleanCo client con- tracts, the majority’s new test leaves unanswered whether CleanCo and Clients A, B, and C could re- negotiate their client contracts, or whether the “joint” bargaining obligations-and the CBA(s)- would effectively trump any potential client contract renegotiations, even though this would be contrary to the Supreme Court’s indication that Congress, in adopting the NLRA, “had no expectation that the elected union representative would become an equal partner in the running of the business enterprise in which the union’s members are employed.” First National Maintenance, supra, 452 U.S. at 676. Likewise, similar to what the majority held in CNN (see discussion infra), the majority would impose its new joint-employer bargaining obligations on Cli- ents A, B, and C, even where the client contracts ex- plicitly identified CleanCo as the only “employer” and stated that CleanCo had sole and exclusive re- sponsibility for collective bargaining. • New Clients (Possibly With Their Own Union Obli- gations). If a union represented all CleanCo employ- ees, and if (under the majority’s new test) all CleanCo clients were deemed joint employers with CleanCo, what happens when Clean Co obtains new clients that previously had cleaning work performed by in-house employees or a predecessor contractor, and those in-house or contractor employees were un- represented or represented by a different union? If, based on CleanCo’s existing union commitments, CleanCo refused to consider hiring or retaining the employees who formerly did the new client’s clean- ing work, the refusal could constitute antiunion dis- crimination in violation of Sec. 8(a)(3). If CleanCo hired the new client’s former employees (or the for- mer employees of a predecessor contractor), then CleanCo could run afoul of its existing union obliga- tions. See Whitewood Maintenance Co., 292 NLRB 40 DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD 1159, 1168-1169 (1989), enfd. 928 F.2d 1426 (5th Cir. 1991). Alternatively, this situation could re- quire further Board proceedings for resolution.64 • Non-Consensual Multiemployer Bargaining. The Board has held that employees solely employed by a supplier employer combined with employees jointly employed by the supplier employer and a single user employer (e.g., CleanCo and either Clients A, B, or C) must be considered inappropriate as a matter of law, absent the consent of the parties. Oakwood Care Center, 343 NLRB 659, 661-663 (2004). A unit consisting of employees jointly employed by the supplier employer and multiple user employers (e.g., CleanCo and Clients A, B, and C) would likewise be inappropriate absent consent, unless the majority is overruling (sub silentio) the Oakwood consent re- quirement. • Potential Board Jurisdiction Over Some Entities and Not Others. The Board does not have jurisdic- tion over governmental employers and employees, over railways or airlines that are subject to the Rail- way Labor Act, or-in a variety of circumstances- religiously-affiliated educational institutions or cer- tain enterprises operated by Indian tribes. If CleanCo is subject to the NLRA, but Clients A, B, or C fall within one or more of the exempt categories identified above, the majority’s new standard will create complex questions about whether the Board may lack jurisdiction over particular “joint” employ- er(s). 2. Union Organizing Directed at Client(s). If two different unions, rather than targeting CleanCo, engage in organizing directed at Client A and Client B, respec- tively, with Client C remaining nonunion, this gives rise to additional issues and problems: • All of the Above Issues/Problems. If the CleanCo employees at Client A are organized by one union, and if the CleanCo employees at Client B are orga- nized by a different union, then the majority test would make CleanCo and Client A the “joint em- ployer” of the CleanCo/Client A employees, and CleanCo and Client B the “joint employer” of the CleanCo/Client B employees. In both cases, the “joint employer” status would give rise to all of the above problems and issues, in addition to those de- scribed below. 64 Such a resolution might result, for example, from a unit clarifica- tion petition seeking to add the new employees to the bargaining unit without an election under the Board’s accretion doctrine, or jurisdic- tional dispute proceedings pursuant to Sec. 10(k) of the Act. • Employee Interchange and Multilocation Assign- ments. If different unions represent the employees of CleanCo/Client A and CleanCo/Client B, and if CleanCo/Client C employees were nonunion, this would create substantial potential problems and po- tential conflicting liabilities regarding CleanCo em- ployees assigned to work at all three client locations or transferred from one client’s facility to another. This is a common situation, arising, for example, where one CleanCo client simply was unhappy with the productivity or attitude of the assigned employ- ee.65 • Strikes and Picketing-“Neutral” Secondary Boy- cott Protection Eliminated. Sections 8(b)(4) and 8(e) of the Act protect neutral parties from being subject- ed to “secondary” picketing and other threats, coer- cion and restraint that have an object of forcing one employer to cease doing business with another. Therefore, if the CleanCo/Client A and CleanCo/Client B employees were involved in a la- bor dispute, under the Board’s traditional joint- employer standard Clients A and B (as non- employers) would be neutral parties protected from “secondary” union activity. Under the majority’s standard, however, Clients A and B would be em- ployers right along with CleanCo and thus subject to picketing. • Renegotiating or Terminating Client Contracts. It is well established that “an employer does not discrim- inate against employees within the meaning of Sec- tion 8(a)(3) by ceasing to do business with another employer because of the union or nonunion activity of the latter’s employees.”66 However, to the extent that CleanCo and Clients A, B, and C are joint em- ployers, then any client’s termination of CleanCo’s 65 The potential problems caused by multilocation assignments or employee interchange between locations could arise, for example, from CBA provisions restricting such assignments or transfers, from union- security provisions in different CBAs requiring dues payments based on a person’s employment without regard to where they were em- ployed, or from conflicting wage rates and benefits applicable at each location. Although these issues might depend on what particular CBA or other policies were in effect, they would obviously cause significant burdens and potential confusion for the employees and each entity considered a joint employer under the majority’s new standards. 66 Plumbers Local 447 (Malbaff Landscape Construction), 172 NLRB 128, 129 (1968). See also Computer Associates International, Inc., 324 NLRB 285, 286 (1997) (“[F]inding a violation of Section 8(a)(3) on the basis of an employer’s decision to substitute one inde- pendent contractor for another because of the union or nonunion status of the latter’s employees is inconsistent with both the language of Sec- tion 8(a)(3) . . . and with legislative policies underlying Section 8(b) of the Act aimed at protecting the autonomy of employers in their selec- tion of independent contractors with whom to do business.”). BFI NEWBY ISLAND RECYCLERY 41 services based on potential union-related considera- tions would create a risk that the Board would find-as it did in CNN, supra-that the contract ter- mination constituted antiunion discrimination in vio- lation of Sec. 8(a)(3). CNN, supra, slip op. at 40-42 (Member Miscimarra, dissenting). 3. Existing CleanCo-Union and/or Existing Client- Union Relationships. Additional issues and problems result from the impact of the majority’s new joint- employer test on existing union relationships and CBAs: • All of the Above Issues/Problems. It is clear, under the majority’s test, that existing collective- bargaining agreements and union relationships in- volving CleanCo, with no mention of Clients A or B, do not prevent Clients A and B from having joint- employer status with CleanCo, which would give rise to all of the issues and problems described above. Again, in CNN, discussed infra, the Board majority found that the client (CNN) was a joint em- ployer, even though any bargaining between CNN and the unions representing employees of contractor TVS would have departed from applicable labor contracts, prior Board certifications, the services agreements between CNN and its vendor (TVS), and 20 years of bargaining history in which the employ- er-party was always TVS (or its predecessor contrac- tors), and not CNN. • Existing CleanCo CBA: Prospective Four-Party Bargaining. If CleanCo was party to an existing company-wide collective-bargaining agreement, in which CleanCo was identified as the only “employ- er,” the majority’s new test clearly imposes an obli- gation to engage in bargaining on all joint-employer entities-i.e., CleanCo and Clients A, B, and C- even though such bargaining would depart from ex- plicit CBA language and the past practice of CleanCo and the union. • “Mandatory” Arbitration, Yet Never Agreed To? If CleanCo had an existing company-wide CBA, the majority’s imposition of “employer” status on Cli- ents A, B, and C would not necessarily bind them to the terms of the existing CleanCo CBA. This would mean that, even though a particular grievance may pertain to essential employment terms that, in the majority’s view, Clients A, B, and C have the right to “share or codetermine,” the CBA’s grievance ar- bitration procedure would not necessarily bind Cli- ents A, B, and C, since they had never agreed to submit to the procedure.67 • Benefit Fund Contributions and Liabilities-Who Pays? Many existing collective-bargaining agree- ments contain extensive provisions regarding benefit fund contributions and benefit liabilities. If such provisions were contained in the CleanCo CBA, then Clients A, B, and C-when participating in the new four-way bargaining described above-would pre- dictably be confronted with demands to assume lia- bility for such provisions. Although the majority test suggests that Clients A, B, and C “will be re- quired to bargain only with respect to such terms and conditions which it possesses the authority to con- trol,” it appears clear that they would face economic demands and potentially be subject to a strike based on a refusal to agree to such demands. • Joint Bargaining Versus “Add-On” CBAs. If CleanCo employees assigned to Clients A, B, or C were organized for the first time by one or more un- ions, the majority clearly imposes a new mandatory bargaining obligation on all joint employer entities. Although an existing collective-bargaining agree- ment generally suspends a party’s obligation to bar- gain for the agreement’s term, the majority’s new test, as noted above, imposes an independent duty to bargain on every joint employer “with respect to such terms and conditions which it possesses the au- thority to control,” which may result in separate sets of negotiations and potential “add-on” CBAs that deviate from the existing union agreements. The foregoing is only a selection of the complications that may arise. And the example is obviously simplistic because it relates only to one service company, which has only three clients-and in the real world, by compar- ison, (i) many businesses, large and small, rely on ser- vices provided by large numbers of separate vendors, and (ii) many service companies have dozens or hundreds of separate clients. Time will no doubt reveal more as em- ployers and unions attempt to apply the limitless joint- employer standard to even more complicated settings than the above example. The only thing that is clear at present is that the new standard does not promote stable collective-bargaining relationships. There is no way that it could, and simple mathematics shows us why. On its face, the majority’s broad test can find up to 18 “joint” employers per work force. How? The majority 67 AT&T Technologies Inc. v. CWA, 475 U.S. 643, 648 (1986); Steelworkers v. Warrior & Gulf Navigation Co., 363 U.S. at 582; Steelworkers v. American Mfg. Co., 363 U.S. at 570-571; Gateway Coal Co. v. UMW, 414 U.S. 368, 374 (1974). 42 DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD finds that there are at least six essential terms and condi- tions of employment (wages, hours, hiring, firing, disci- pline, and direction of work). According to the majority, an “employer” is an entity that exercises-even on a lim- ited and routine basis-any one of three forms of puta- tive control (direct control, indirect control, or potential control) over any one of these terms. Six times 3 is 18, which leaves us with a model where there could be up to 18 employers for a single workforce. See Appendix A (“Why There Are At Least 18 Potential Employers”). In truth, the test can find more than 18 employers because the majority has not limited itself to the specified 6 sup- posedly essential terms, and the majority has not unquali- fiedly represented that there can be only one controller per category of control, e.g., there could be two “indirect controllers,” for example. We do not know the exact limit to the multiplicity of putative employers arising from the majority’s new joint-employer test. But it is surely common sense that placing 18 different cooks involuntarily in a single kitchen will lead to a terrible meal. That is the recipe for dyspeptic collective bargain- ing that the majority has cooked up. The majority states that “a joint employer will be re- quired to bargain only with respect to such terms and conditions which it possesses the authority to control.” This does not temper the impact of the new standard; it only makes matters worse. The majority assumes these bargaining issues are severable, as if the resolution of one issue is not dependent on the resolution of another. This is not how contract negotiations work. And under- scoring the irrationality of the majority’s rule here, the Board has traditionally denounced this type of segmented issue-by-issue negotiating, when unilaterally undertaken by a party, as unlawful “fragmented bargaining.”68 Moreover, how exactly are joint user and supplier em- ployers to divvy up the bargaining responsibilities for a single term of employment that they will be deemed un- 68 See, e.g., E.I. Dupont de Nemours & Co., 304 NLRB 792, 792 fn. 1 (1991) (“What we find unlawful in the Respondent’s conduct was its adamant insistence throughout the entire course of negotiations that its site service operator and technical assistant proposals were not part of the overall contract negotiations, and, therefore, had to be bargained about totally separately not only from each other but from all the other collective bargaining agreement proposals. We find this evinced frag- mented bargaining in contravention of the Respondents duty to bargain in good faith.”); see also NLRB v. Patent Trader, 415 F.2d 190, 198 (2d Cir. 1969), modified on other grounds 426 F.2d 791 (2d Cir. 1970) (When a party “removes from the area of bargaining . . . [the] most fundamental terms and conditions of employment (wages, hours of work, overtime, severance pay, reporting pay, holidays, vacations, sick leave, welfare and pensions, etc.),” it has “reduced the flexibility of collective bargaining, [and] narrowed the range of possible compromis- es with the result of rigidly and unreasonably fragmenting the negotia- tions.”). der the new standard to codetermine, one by direct con- trol and the other by indirect control? How does one know who has authority at all over a term and condition of employment, under the majority’s vague formulation? What if two putative employer entities get into a dispute over whether one has authority over a certain term or condition of employment? What if the putative employ- ers are competitors? Taking the diagram above, what if Client A and Client B are competitors and have no real economic interest in the other client coming to a good- faith agreement with CleanCo on how much it pays em- ployees working for that other client? Does it make sense for the law to attempt to create such an interest? What if there are too many entities to come to an agree- ment? How does bargaining work in this circumstance? Further, this purported division of bargaining responsi- bility creates conflicts between alleged violations of Sec- tion 8(a)(5), which requires employers to bargain in good faith with a certified or recognized union, and Sec- tion 8(a)(2), which makes such bargaining unlawful if the union lacks majority support among the entity’s employ- ees.69 If multiple entities arguably constitute a “joint employer,” and one entity is alleged to have unlawfully failed to bargain over particular terms of employment, the majority’s standard effectively places the burden of proof on the respondent-employer to establish that it did not control those particular employment terms.70 So questions exist as to (i) which entities are the “employ- er,” (ii) which entities must (or must not) engage in bar- gaining over particular employment terms, and even (iii) what party-the respondent(s) versus the General Coun- sel-bears the burden of proof regarding this assortment of issues. 69 The conflict between Sec. 8(a)(5) and Sec. 8(a)(2) results from the Hobson’s Choice that confronts multiple entities that control differ- ent aspects of employment for one or more different employee groups. Potential joint-employer entities risk violating Sec. 8(a)(5) if they fail or refuse to bargain over certain matters because Sec. 8(a)(5) obliga- tions apply generally to “wages, hours, and other terms and conditions of employment.” See Sec. 8(d) (defining the phrase “to bargain collec- tively,” which is required under Sec. 8(a)(5)). Conversely, potential joint-employer entities risk violating Sec. 8(a)(2), which makes it un- lawful for an employer to bargain with a union that does not validly represent its employees, if the Board determines that the entities en- gaged in bargaining when, in fact, they were not an “employer” as to employment terms not within their control. In other words, not only does the majority’s standard promise to create confusion about who is an “employer,” but the majority’s patchwork allocation making differ- ent entities responsible for different issues creates confusion about which “employer” entity may or must bargain over what particular employment terms. As with other aspects of the majority’s new stand- ard, definitive answers will be available only after years of Board and court litigation. 70 See, e.g., Hobbs & Oberg Mining Co., 297 NLRB 575, 586 (1990) (General Counsel’s burden to prove joint-employer status), enfd. 940 F.2d 1538 (10th Cir. 1991), cert. denied 503 U.S. 959 (1992). BFI NEWBY ISLAND RECYCLERY 43 This scenario is made all the worse by the need for years of Board litigation before third parties will actually learn whether (i) they unlawfully failed to participate in bargaining between another employer and its union(s), or (ii) the third parties unlawfully injected themselves into such bargaining when their commercial relationship was insufficient to make them a joint employer. Nor is the Board permitted to engage in the economic analysis needed to sort out the plethora of arm’s-length company- to-company relationships affected by the majority’s new joint-employer test. The Board’s Division of Economic Research was abolished 75 years ago, and Section 4(a) of the Act-adopted by Congress in 1947-prohibits the Board from having any agency personnel engage in “economic analysis.”71 Additionally, we note that the Board lacks the authority to impose labor contract terms on parties,72 and nothing in the Act authorizes the Board to impose requirements on companies regarding how they must arrange or rearrange themselves. The majority even acknowledges some turmoil will re- sult from its decision, but largely dismisses it as being outweighed by the need to protect contingent workers’ Section 7 rights. Certainly any doctrinal change in this area will modify the legal landscape for employers with respect to the National Labor Relations Act. However, given the centrality of collective bargaining under the Act, we must ensure that the prospect of collective bargaining is not foreclosed by business relationships that effectively deny employees’ right to bargain with employers that share control over essential terms and conditions of their employment. [(Footnote omitted.)] Contrary to our colleagues’ assertion, we are not slavish defenders of the status quo. We would support revisiting any Board doctrine that systemically fails to protect Section 7 rights, but we would not do so without evidence of that failure. The majority cites no evidence, and none has been presented, showing that employees in contingent or any comparable employment situations have been unable to bargain with their undisputed employer. The majority uses 71 Sec. 4(a) states in part: “Nothing in this Act shall be construed to authorize the Board to appoint individuals . . . for economic analysis.” This language was added to the NLRA as part of the Labor Manage- ment Relations Act (LMRA), 61 Stat. 136, Sec. 101 (amending NLRA Sec. 4(a)) (1947). The enactment of Sec. 4(a) occurred after the Board abolished its Division of Economic Research in 1940. See 93 Cong. Rec. 6661, reprinted in 2 LMRA Hist. 1577 (June 6, 1947) (analysis of H.R. 3020). See generally John E. Higgins, Jr., Labor Czars- Commissars-Keeping Women in the Kitchen-The Purpose and Effects of the Administrative Changes Made by Taft-Hartley, 47 CATH. U. L. REV. 941, 951-952 (1998). 72 Sec. 8(d); H.K. Porter Co. v. NLRB, 397 U.S. 99 (1970). the phrase “meaningful bargaining” numerous times, but the majority’s premise is that bargaining fails to be “meaning- ful” whenever the employer’s business relationships influ- ence the matters under negotiation. Our colleagues on this front simply cite the large number of employees whose terms and conditions of employment might be affected in some way by a user employer and Board cases finding no duty to bargain with these user employers, and assert that rights have been denied. How do we know that employees have been unable to engage in “meaningful bargaining” with the supplier employer? Under the majority’s test, it is possible to find that “meaningful bargaining” cannot take place with a supplier employer alone if it lacks meaningful control over even a single “essential” facet of employment. Such a definition of meaningful bargaining has never been the law, and it cannot be reconciled with business practices that have been in existence since before the Act. It is difficult, if not impossible, to reconcile this rea- soning with the Board’s rationale in Management Train- ing, 317 NLRB 1355 (1995), addressing whether to as- sert discretionary jurisdiction over a private employer contracting for business with an exempt governmental entity. The Board there modified prior caselaw and held that it would no longer decline to assert jurisdiction in circumstances where the private employer lacked control of what had been deemed essential terms of employment. It reasoned that “[b]ecause of commercial relationships with other parties, an inability to pay due to financial constraints, and competitive considerations which cir- cumscribe the ability of the employer to grant particular demands, the fact is that employers are frequently con- fronted with demands concerning matters which they cannot control as a practical matter or because they have made a contractual relationship with private par- ties or public entities.” Id. at 1359 (emphasis added). Quite obviously, under Management Training, the Board believes that employees and their exclusive bargaining representative can still engage in meaningful bargaining under the Act even with an employer who lacks control over a substantial number of essential terms of employ- ment. C. The New Test Will Dramatically Change Labor Law Sales and Successorship Principles, and Will Discourage Efforts to Rescue Failing Companies and Preserve Employment Expanding the definition of employer will also alter the landscape of successorship law under the Act. It is well established that successor employers,73 although 73 An employer is a successor of its predecessor under the Act when there is a “substantial continuity between the enterprises,” the successor hired a majority of its predecessor’s employees, and the unit is still 44 DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD they must recognize and bargain with the union repre- senting the predecessor’s employees in certain circum- stances, are not obligated to adopt the preexisting collec- tive-bargaining agreement and have the right to unilater- ally set different initial terms and conditions of employ- ment.74 NLRB v. Burns International Security Services, Inc., 406 U.S. 272, 287-288, 294-295 (1972). This rule “careful[ly] safeguards the rightful prerogative of owners independently to rearrange their businesses.” Fall River Dyeing & Finishing Corp. v. NLRB, 482 U.S. 27, 40 (1987) (internal quotations omitted). But the policy con- cerns behind the rule are even deeper than that: [H]olding either the union or the new employer bound to the substantive terms of an old collective-bargaining contract may result in serious inequities. A potential employer may be willing to take over a moribund busi- ness only if he can make changes in corporate structure, composition of the labor force, work location, task as- signment, and nature of supervision. Saddling such an employer with the terms and conditions of employment contained in the old collective-bargaining contract may make these changes impossible and may discourage and inhibit the transfer of capital. On the other hand, a union may have made concessions to a small or failing employer that it would be unwilling to make to a large or economically successful firm. The congressional policy manifest in the Act is to enable the parties to ne- gotiate for any protection either deems appropriate, but to allow the balance of bargaining advantage to be set by economic power realities. Strife is bound to occur if the concessions that must be honored do not corre- spond to the relative economic strength of the parties. Burns, 406 U.S. at 287-288. Under the majority’s expansive joint-employer stand- ard, many user employers will now be considered joint employers of their supplier employers’ employees. Re- bidding contracts has been a common feature of the us- er-and supplier-employer market. Going forward, it may be less common because deeming the user employer to be a joint employer will make terminating or rebidding the contract with the supplier employer much more diffi- appropriate. Fall River Dyeing & Finishing Corp. v. NLRB, 482 U.S. 27, 43-52 (1987). 74 There is a limited exception to this general rule when “‘it is per- fectly clear that the new employer plans to retain all of the employees in the unit,’” unless the successor “clearly announce[s] its intent to establish a new set of conditions prior to inviting former employees to accept employment.” Spruce Up Corp., 209 NLRB 194, 195 (1974) (quoting Burns, 406 U.S. at 294-295), enfd. 529 F.2d 516 (4th Cir. 1975). However, a so-called “perfectly clear” successor employer is still not bound by the predecessor contract itself. It must only adhere to terms established by the contract while negotiating new terms with the incumbent union. cult. The user employer will often have a duty to bargain the decision to lay off the employees or to subcontract those jobs to another supplier employer. See Fibreboard Corp. v. NLRB, supra, 379 U.S. at 215 (1964); CNN, supra, 361 NLRB No. 47, slip op at 17. Assuming the user employer does contract with a new supplier employ- er that would otherwise be a Burns successor able to set its own terms, the user employer, under the broadened standard, will likely be deemed a joint employer with the new supplier employer as well. That user employer’s ongoing bargaining obligation spanning the two supplier employers prevents the new supplier employer from set- ting different terms and conditions of employment than its predecessor had. See Whitewood Maintenance Co., supra, 292 NLRB at 1168-1169 (contractor that substi- tuted one subcontractor for another jointly employed both the old and new subcontractors’ employees, so the new subcontractor could not set its own initial terms), enfd. 928 F.2d 1426 (5th Cir. 1991). Similarly, when a predecessor’s union-represented employees apply for employment with a successor, the successor cannot lawfully extend recognition unless and until it has hired a “substantial and representative com- plement” of employees and has received a demand for recognition from the predecessor union(s).75 In CNN, supra, two unions already represented employees of CNN’s contractor, TVS, as part of a 20-year history in which unionized contractors supplied technical employ- ees to CNN, where only the contractor-and not CNN- was considered the “employer.” When CNN decided to terminate its use of contractor employees and directly hire its own technical workforce, CNN as a successor would have violated the Act if it engaged in bargaining with the TVS unions before it hired a “substantial and representative complement” of its own employees. However, the majority’s expansive joint-employer find- ing converted CNN into an “employer” before it hired any of its own technical employees. And, based on its expansive joint-employer finding, the Board majority determined that CNN-even before it decided to termi- nate the TVS relationship (and before it notified TVS)- was required to notify the TVS unions and engage in bargaining with them over whether CNN might terminate the TVS relationship and hire its own work force. Member Miscimarra stated, in his CNN dissent, that employer status “does not arise as the result of spontane- ous combustion,” and he explained that the expansive joint-employer finding-applied to CNN before it hired its own workforce-was irreconcilable with the parties’ understandings and existing agreements: 75 Fall River Dyeing Corp. v. NLRB, 482 U.S. at 47-48. BFI NEWBY ISLAND RECYCLERY 45 Nothing in such a scenario would promote stable bar- gaining relationships. Rather, CNN’s actions-taken as an “employer” of the TVS technical personnel- would have directly contradicted the then-existing TVS-NABET collective-bargaining agreements (which identified TVS, not CNN, as the employer). CNN’s ac- tions would have violated the CNN-TVS Agreements, which stated . . . that TVS employees “are not employ- ees of [CNN], and shall not be so treated at any time”. . . . Finally, CNN’s actions would have exhibited a total disregard for the elaborate body of law regarding “successorship” and related business changes that has been the subject of nearly a dozen Supreme Court cases and innumerable Board decisions.76 The inability of user employers to freely terminate or rebid client contracts and of new supplier employers to set different initial terms will inhibit our economy and lead to labor strife. The new standard sends a message to user employers to never contract with unionized firms in the first place to avoid being trapped in “permanent” client contracts that cannot be terminated without bar- gaining to agreement or impasse. On the other side, the supplier-employer market will become uncompetitive as potential bidders for contracts where the incumbent sup- plier employer is unionized will be unable to compete with the incumbent employer on labor costs, as the new supplier employer will likely be beholden to the same terms. The Act is being applied in a manner Congress could not conceivably have intended. D. The New Test Threatens Existing Franchising Ar- rangements in Contravention of Board Precedent and Trademark Law Requirements Of the thousands of business entities with different contracting arrangements that may suddenly find them- selves to be joint employers, franchisors stand out. Ac- cording to amicus International Franchise Association (IFA), “in 2012 there were 750,000 franchise establish- ments in the United States employing 8.1 million work- ers, generating a direct economic output of $769 billion. These businesses account for approximately 3.4 percent of America’s gross domestic product.”77 For many years, the Board has generally not held fran- chisors to be joint employers with franchisees, regardless of the degree of indirect control retained.78 The majority 76 CNN, supra, slip op. at 38-39 (Member Miscimarra, dissenting) (footnote and emphasis omitted). 77 Br. of IFA at 1. 78 See, e.g., Speedee 7-Eleven, 170 NLRB 1332 (1968) (franchisor not a joint employer despite a policy manual that described “in meticu- lous detail virtually every action to be taken by the franchisee in the conduct of his store”), and Tilden, S. G., Inc., 172 NLRB 752 (1968) (franchisor not a joint employer, even though the franchise agreement does not mention, much less discuss, the potential impact of its new standard on franchising relations, but it will almost certainly be momentous and hugely disruptive. Indeed, absent any discussion, we are left to ponder whether the majority even agrees with the statement of the General Counsel in his amicus brief that “[t]he Board should continue to exempt franchisors from joint em- ployer status to the extent that their indirect control over employee working conditions is related to their legiti- mate interest in protecting the quality of their product or brand. See, e.g., Love’s Barbeque Rest., 245 NLRB 78, 120 (1978) (no joint-employer finding where franchisees were required to prepare and cook food a certain way because, inter alia, the franchisor established the re- quirements to ‘keep the quality and good will of [the franchisor’s] name from being eroded’ (internal quota- tions and citations omitted), enforced in rel. part, 640 F.2d 1094 (9th Cir. 1981).” (Amicus Br. at 15-16 fn. 32). Given the breadth of the majority’s test and ra- tionale, we are concerned that the majority effectively finds that a franchisor even with this type of indirect con- trol would be deemed a joint employer. The majority’s new test appears to require specific analysis of whether the franchisor shares or codetermines “the manner and method of performing the work.” How- ever, in many if not most instances, franchisor operation- al control has nothing to do with labor policy but rather compliance with federal statutory requirements to main- tain trademark protections. “It is required that the owner of the mark should set up the standards or conditions which must be met before another is permitted to use the certification mark and the owner should permit the use of the mark by others only when they meet those standards or conditions.” State of Fla. v. Real Juices, Inc., 330 F. Supp. 428, 432 (M.D. Fla. 1971). As one court ex- plained: Without the requirement of control, the right of a trademark owner to license his mark separately from the business in connection with which it has been used would create the danger that products bearing the same trademark might be of diverse qualities. If the licensor is not compelled to take some reasonable steps to pre- vent misuses of his trademark in the hands of others the public will be deprived of its most effective protection against misleading uses of a trademark. The public is hardly in a position to uncover deceptive uses of a trademark before they occur and will be at best slow to detect them after they happen. Thus, unless the licen- dictated “many elements of the business relationship,” because the franchisor did not “exercise direct control over the labor relations of [the franchisee]”). 46 DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD sor exercises supervision and control over the opera- tions of its licensees the risk that the public will be un- wittingly deceived will be increased and this is precise- ly what the Act is in part designed to prevent. Clearly the only effective way to protect the public where a trademark is used by licensees is to place on the licen- sor the affirmative duty of policing in a reasonable manner the activities of his licensees. Stanfield v. Osborne Indus., Inc., 839 F. Supp. 1499, 1504 (D. Kan. 1993), affd. 52 F.3d 867 (10th Cir. 1995), abrogat- ed on other grounds by Lexmark Int’l, Inc. v. Static Control Components, Inc., 134 S.Ct. 1377 (2014). If a franchisor fails to maintain sufficient control over its marks, it is con- sidered to have engaged in “naked franchising” and thereby abandoned the mark.79 “The critical question in determin- ing whether a licensing program is controlled sufficiently by the licensor to protect his mark is whether the licensees’ operations are policed adequately to guarantee the quality of the products sold under the mark.” General Motors Corp. v. Gibson Chem. & Oil Corp., 786 F.2d 105, 110 (2d Cir. 1986). The necessity of the franchisor to police the “man- ner and method” of the franchisee is paramount. “‘The pur- pose of the Lanham Act . . . is to ensure the integrity of reg- istered trademarks, not to create a federal law of agency.’ The scope of a licensor’s duty of supervision of a licensee who has been granted use of a trademark must be commen- surate with this limited goal.” Transgo, Inc. v. Ajac Trans- mission Parts Corp., 768 F.2d 1001, 1018 (9th Cir. 1985) (quoting Oberlin v. Marlin American Corp., 596 F.2d 1322, 1327 (7th Cir. 1979)). These cases demonstrate that one important aspect of the franchising relationship is the franchisee’s ability to reap the benefits of manifesting to the customer the ap- pearance of a seamless enterprise through the use and maintenance of the franchisor’s trademark. Federal fran- chise law recognizes this benefit and requires that the franchisor maintain the mark by maintaining enough control over the franchisee to protect consumers. How- ever, even while franchise law requires some degree of 79 Id.; see 15 U.S.C. § 1064(5)(A). See also Barcamerica Interna- tional USA Trust v. Tyfiled Importers, Inc., 289 F.3d 589, 596 (9th Cir. 2002) (“It is well-established that ‘[a] trademark owner may grant a license and remain protected provided quality control of the goods and services sold under the trademark by the licensee is maintained.’ Moore Bus. Forms, Inc. v. Ryu, 960 F.2d 486, 489 (5th Cir.1992). But ‘[u]ncontrolled or “naked” licensing may result in the trademark ceas- ing to function as a symbol of quality and controlled source.’ McCar- thy on Trademarks and Unfair Competition § 18:48, at 18-79 (4th ed. 2001). Consequently, where the licensor fails to exercise adequate quality control over the licensee, ‘a court may find that the trademark owner has abandoned the trademark, in which case the owner would be estopped from asserting rights to the trademark.’ Moore, 960 F.2d at 489.”). oversight and interaction, it was never the intent of Con- gress, by that interaction, to make a franchisee the agent of its franchisor for any purpose. Thus, the new joint- employer standard portends unintended consequences for a franchisor’s compliance with the requirements of an- other Federal act that are totally unrelated to labor rela- tions. The Board has been repeatedly reminded that it “has not been commissioned to effectuate the policies of the Labor Relations Act so single-mindedly that [we] may wholly ignore other and equally important Congres- sional objectives.” Southern Steamship Co. v. NLRB, 316 U.S. 31, 47 (1942). Rather than providing a “careful accommodation of one statutory scheme to another,” the majority’s new standard places “excessive emphasis up- on [the Board’s] immediate task.” Id. E. The New Test Undermines the Parent-Subsidiary Re- lationship in Contravention of Board Precedent In most areas of the law, it is widely recognized that parent and subsidiary corporations are indeed separate entities. The Board, which has developed whole legal doctrines devoted to detecting ostensibly separate com- panies that are in truth either created to evade obligations under the Act (the alter ego doctrine) or so integrated that they function as one (the single employer doctrine), has recognized this principle repeatedly. For example, in Dow Chemical, 326 NLRB 288 (1998), a bipartisan Board majority reaffirmed the longstanding rule under the single employer doctrine that typical parents and sub- sidiaries are not considered a sole “employer” for bar- gaining purposes. See also, e.g., Western Union, 224 NLRB 274 (1976), affd. sub nom. United Telegraph Workers v. NLRB, 571 F.2d 665 (D.C. Cir. 1978), cert. denied 439 U.S. 827 (1978). Indeed, the presumption of separateness for purposes of the Act is so strong that it extends also to unincorporated divisions that are operated independently from the company as a whole. See, e.g., Los Angeles Newspaper Guild, Local 69 (Hearst Corp.), 185 NLRB 303, 304 (1970), enfd. 443 F.2d 1173 (9th Cir. 1971). And here, the Board’s honoring of corporate separateness occurs even as the Board simultaneously recognizes that a subsidiary is, of course, under the po- tential control of its parent. In other words, potential control is not enough to find that a parent is the same employer with its subsidiary for purposes of labor law: Common ownership by itself indicates only potential control over the subsidiary by the parent entity; a sin- gle-employer relationship will be found only if one of the companies exercises actual or active control over the day-to-day operations or labor relations of the other. BFI NEWBY ISLAND RECYCLERY 47 Dow, 326 NLRB at 288 (emphasis in original). The majori- ty now turns this principle on its head, and its wholesale adoption of the “potential control” standard would treat parents and subsidiaries as joint-employing entities for pur- poses of labor law. To our reckoning, no Board has ever taken this leap before. Indeed, the majority’s new test- which applies to admittedly separate and independent com- panies-applies a more onerous “control” standard than the one that the Board uses to find control where a company is actually integrated with another. This makes no sense. Whatever the contradiction in the majority’s logic, the result is serious. The upshot is that the majority’s new test threatens to automatically sweep every parent or af- filiate company in America into being the “employer” of a subsidiary’s employees, with the concomitant bargain- ing obligations, the loss of secondary-employer protec- tion from union strikes discussed below, and all the other deleterious results mentioned above. If this is the out- come intended, upending decades of precedent of labor law and probably centuries of precedent in corporate law, we need a mandate from Congress before we purport to “find” it in our decisional case law. The majority here identifies no such mandate, and its test should be invali- dated on this basis alone. If Congress had wanted us to turn the world of corporate identity upside down, it would have expressly told us so. VI. THE NEW TEST CONFLICTS WITH CONGRESSIONAL INTENT TO INSULATE NEUTRAL EMPLOYERS FROM SECONDARY ECONOMIC COERCION Not only does the majority’s new test impermissibly expand and confuse bargaining obligations under Sec- tions 8(a)(5) and 8(d), it also does violence to other pro- visions of the Act that depend on the “employer” defini- tion. Chief among them is the Section 8(b)(4)(ii)(B) prohibition on secondary economic protest activity such as strikes, boycotts, and picketing. That section “prohib- its labor organizations from threatening, coercing, or restraining a neutral employer with the object of forcing a cessation of business between the neutral employer and the employer with whom a union has a dispute,” but it does not prohibit striking or picketing the primary em- ployer, i.e., the employer with whom the union has the dispute. Teamsters Local 560 (County Concrete), 360 NLRB No. 125, slip op. at 1 (2014). Congress intended to “preserv[e] the right of labor organizations to bring pressure to bear on offending employers in primary labor disputes and . . . [to] shield[] unoffending employers and others from pressures in controversies not their own.” NLRB v. Denver Building Trades Council, supra, 341 U.S. at 692. An entity that is a joint employer with the employer subject to a labor dispute is equally subject to economic protest. See Teamsters Local 688 (Fair Mercantile), 211 NLRB 496, 496-497 (1974) (union’s picketing of a re- tailer did not violate Section 8(b)(4)(ii)(B) because it was the joint employer of a delivery contractor’s employees). To put this in a practical terms, before today’s decision at least, a union in a labor dispute with a supplier employer typically could not picket a user employer urging clients to cease doing business with that user employer-the object there being that the user employer would in turn cease doing business with the supplier employer.80 Likewise, a union with a labor dispute with one franchi- see typically could not picket the franchisor and all of its other franchisees. Today’s expansion of the joint-employer doctrine will sweep many more entities into primary-employer status as to labor disputes that are not directly their own. Un- ions will be able to freely picket or apply other coercive pressure to either or both of the joint employers as they choose. This limits the Act’s secondary-boycott prohibi- tions in a manner Congress did not intend. The targeted joint employer may not have direct control or even any control over the particular terms or conditions of em- ployment that are the genesis of the labor dispute. Here, the economic consequences are far reaching. For exam- ple, a union could picket all of the user employer’s facili- ties even though the supplier employer only provides services at one. Further, assuming that a franchisor ex- erts similar indirect control over each franchisee, as the majority here may often find to be the case, a union could picket the franchisor and all franchisees even though its dispute only involves the employees of one.81 It does not end there. As previously stated, numerous provisions relied upon by the majority are typically in- cluded in a residential renovation contract- i.e., the con- tractor’s employees cannot start work before a certain hour, they must finish work by a certain hour, they can- not use the bathrooms in the house, they have to park their vehicles in certain locations. Suppose that the annu- al revenues of the company with whom the homeowners contract meet the Board’s discretionary standard for as- serting jurisdiction, not an unlikely possibility. Then 80 Of course, the user- and supplier-employer scenario often raises common situs issues as addressed in Sailors Union (Moore Dry Dock), 92 NLRB 547 (1950), and its progeny, but explicitly targeting the sec- ondary employer is blatantly unlawful. 81 Going back to the CleanCo diagram above for an example, Client A likely has no control over what goes on upon the premises of Client C. More importantly, there is no underlying economic relationship between the two that could supply even a remotely rational foundation for the Act to allow economic weapons like strikes, picketing, etc. at Client A to convince it to use its obviously nonexistent “power” over Client C in a labor dispute involving CleanCo employees posted at Client C. 48 DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD suppose that a union initiates an area standards wage protest against this contractor. One day, the homeowners open their front door to discover pickets patrolling the sidewalk in front of their house. In the new joint- employer world, they are a lawful target for this protest activity. Unions may not have any interest in bringing them into any bargaining process, but they may be more than eager to maximize economic injury to the primary employer by expanding the cease-doing-business pres- sure to as many clients as possible. Congress did not intend that every entity with some degree of economic relationship with the employer-disputant be thrown into its labor dispute. The Act is supposed to encourage labor peace, and to this end Congress enacted Sections 8(b)(4) and 8(e), demonstrating its intent to avoid limitless eco- nomic warfare based on dealings between employers and other persons. The majority’s expansive definition of joint-employer status poses particular questions about its applicability to common situs work in the construction industry. As pre- viously stated, the Supreme Court has expressly held that the fact “the contractor and subcontractor were engaged on the same construction project, and that the contractor had some supervision over the subcontractor’s work, did not eliminate the status of each as an independent con- tractor or make the employees of one the employees of the other.”82 We presume that our colleagues do not in- tend to act in direct contravention of an express holding of the Supreme Court, but the breadth of their test and their emphasis on contractual control as probative of joint-employer status seems to pose a dilemma: either they must articulate an exception to a statutory definition that seems to require uniform treatment of employers in all industries, or they must place limits on their test they obviously wish to avoid.83 VII. CONCLUSION The Board is not Congress. It can only exercise the authority Congress has given it. In this instance, our colleagues have announced a new test of joint-employer status based on policy and economic interests that Con- gress has expressly prohibited the Board from consider- ing. That alone is reason enough why the new test 82 Denver Building Trades, 341 U.S. at 692. 83 There is a further question. Denver Building Trades involved a situation in which a subcontractor was the primary employer target of protest, and the general contractor was the neutral employer. In Mark- well & Hartz, the Board applied the same principles of separateness and neutrality when the general contractor was the primary employer in a labor dispute, thereby finding all subcontractors at the common situs to be neutrals. Building & Construction Trades Council (Markwell & Hartz), 155 NLRB 319 (1965), enfd. 387 F.2d 79 (5th Cir. 1967). The breadth of our colleagues’ test raises a genuine concern that they might use it to undermine this decision. should not stand. Even more troubling from an institu- tional perspective, however, is the nature of the new test. The negative consequences flowing from the majority’s new test are substantial. It creates uncertainty where certainty is needed. It provides no real standard for de- termining in advance when entities in a business relation- ship will be viewed as independent and when they will be viewed as joint employers. Moreover, as noted previously, the resulting confusion will cause damage both ways: (i) too many parties will discover after the fact, following years of litigation, they were unlawfully absent from negotiations in which they were legally required to be participants; and (ii) countless other parties will discover they unlawfully injected them- selves into collective bargaining involving another entity and its union(s), based on a relationship that was insuffi- cient, after all, to result in joint-employer status. The majority essentially says that the Board will look at every aspect of a relationship on a case-by-case basis, in litiga- tion, and then decide the limited issue presented. We owe a greater duty to the public than to launch some massive ship of new design into unsettled waters and tell the nervous passengers only that “we’ll see how it floats.” Accordingly, we here defend a standard that serves la- bor law and collective bargaining well, a standard that is understandable and rooted in the real world. It recogniz- es joint-employer status in circumstances that make sense and would foster stable bargaining relationships. Indeed, in the Board’s history of applying this traditional joint-employer test, there have been many cases where two or more employers were found to exercise sufficient control over a common group of employees to warrant joint bargaining obligations and shared liability for unfair labor practices.84 Our quarrel with the majority stems not 84 Our colleagues fault us for making “no real effort to address” the issues they have asserted. But today’s legal framework for bargaining (which they dismissively refer to as “the current status quo”) already supplies the answer. That is, economic interdependence and indirect influence work both ways. Current law offers unions great flexibility when dealing with employers that happen to be interdependent with another entity. As long as the union respects secondary boycott princi- ples, leverage applied to the immediate “employer” is all the more likely to affect suppliers, vendors, and other parties having closely aligned economic interests, which predictably may lead to meaningful discussions and changes across the various entities. Such discussions are likely to occur even “without the intervention of the Board enforc- ing a statutory requirement to bargain,” and there is an “important difference” between such discussions being “permitted” as opposed to making them “mandatory.” First National Maintenance v. NLRB, 452 U.S. 666, 681 fn. 19, 683 (1981). Here, if the Union organizes Leadpoint, then, depending on its actual bargaining strength, it can engage in activities that lead to modifications in BFI’s contract with Leadpoint to accommodate those Union demands. And the Board’s successorship case law permits the Union to remain on the scene even BFI NEWBY ISLAND RECYCLERY 49 from any disagreement about the concept of joint em- ployment status but rather from their imposition of a test that we firmly believe cannot be reconciled with the common-law agency standard the Board is compelled to apply, based on a statute the Board is duty-bound to en- force. The Supreme Court has recently cautioned that a fed- eral agency must explain itself when departing from in- terpretation of well-established rules that have governed business practices for long periods, even when the rules are of the agency’s own making. In Christopher v. SmithKline Beecham Corp., 132 S.Ct. 2156 (2012), the Court reviewed the Department of Labor’s (DOL) new interpretation that pharmaceutical sales representatives would no longer be considered outside salesmen exempt from the FLSA’s overtime provisions. The Court em- phasized that its usual deference to such an agency action was not warranted because of the “potentially massive” economic implications of the new interpretation “for conduct that occurred well before that interpretation was announced,”85 and because deference “would seriously undermine the principle that agencies should provide regulated parties ‘fair warning of the conduct [a regula- tion] prohibits or requires.’”86 The Court also noted that DOL’s “longstanding practice” of exempting detailers went back to the beginning of the FLSA, and that there were currently 90,000 detailers working for pharmaceuti- cal companies with the understanding that they were ex- empt outside sales reps.87 Because DOL’s new interpretation would be so disrup- tive to the regulated industry, the Court could not simply defer to it: It is one thing to expect regulated parties to conform their conduct to an agency’s interpretations once the agency announces them; it is quite another to require regulated parties to divine the agency’s interpretations in advance or else be held liable when the agency an- nounces its interpretations for the first time in an en- forcement proceeding and demands deference. if BFI attempts to switch contractors. The flaw with our colleagues’ approach is that, regardless of the strength of the union, it gives that union an artificial place at the table where there is any interdependency between the employer and other entities. See H. K. Porter Co., 397 U.S. at 107-108 (“It is implicit in the entire structure of the Act that the Board acts to oversee and referee the process of collective bargaining, leaving the results of the contest to the bargaining strengths of the parties.”) 85 Id. at 2167. 86 Id. (quoting Gates & Fox Co. v. Occupational Safety and Health Review Comm’n, 790 F.2d 154, 156 (D.C. Cir. 1986)). 87 Id. at 2167-2168. Accordingly, whatever the general merits of . . . defer- ence, it is unwarranted here. We instead accord the De- partment’s interpretation a measure of deference pro- portional to the “‘thoroughness evident in its considera- tion, the validity of its reasoning, its consistency with earlier and later pronouncements, and all those factors which give it power to persuade.’” United States v. Mead Corp., 533 U.S. 218, 228, 121 S.Ct. 2164, 150 L.Ed.2d 292 (2001) (quoting Skidmore v. Swift & Co., 323 U.S. 134, 140, 65 S.Ct. 161, 89 L.Ed. 124 (1944)).88 What the majority has done here is far broader in scope than DOL’s invalidated interpretive change. In- stead of overturning one discrete longstanding agency interpretation that affects a statutory exemption for a single category of employer, the Board has substantially altered its interpretation of joint-employer status across the spectrum of private business relationships subject to our jurisdiction. Despite the majority opinion’s descrip- tion, this case is not merely about whether the Board should overturn thirty years of precedent based on the TLI and Laerco decisions. That would be serious enough. Our greater concern is the impact of the majority’s re- formulation on a much broader body of law, affecting multiple doctrines central to the Act that have been de- veloped and refined through decades of work by biparti- san Boards, the courts, and Congress. As in Christopher, the majority here gives insufficient consideration to the “potentially massive” economic implications of its new joint-employer standard, and it requires innumerable parties to “divine the agency’s interpretations in advance or else be held liable when the agency announces its in- terpretations for the first time in an enforcement proceed- ing.” We believe that the Board should adhere to the “joint-employer” test that has existed for 30 years with- out a single note of judicial criticism. In our view, the Regional Director correctly applied that test in conclud- ing that Leadpoint was the sole employer of employees in the petitioned-for unit. Accordingly, we respectfully dissent. Dated, Washington, D.C. August 27, 2015 ______________________________________ Philip A. Miscimarra, Member ______________________________________ Harry I. Johnson, III, Member 88 Id. at 2168-2169. 50 DECISIONS OF THE NATIONAL LABOR RELATIONS BOARD NATIONAL LABOR RELATIONS BOARD