North American Van Lines, a company that transports goods by truck, petitions to review the determination of the National Labor Relations Board that the company committed an unfair labor practice through its role in creating and supporting an "Advisory Council," comprised of management officials and drivers who carry its loads. See 29 U.S.C. Sec. 158(a)(2) (1982) ("It shall be an unfair labor practice for an employer ... to dominate or interfere with the formation or administration of any labor organization or contribute financial or other support to it"). The Board cross-applies for enforcement of its order.

The Board's jurisdiction, and the scope of sections 8(a)(1) and 8(a)(2) of the National Labor Relations Act, extend only to an employer's acts directed at its "employees." The NLRB's jurisdiction does not encompass "independent contractor[s]." See id. Sec. 152(3). The company claims that the drivers at issue are independent contractors rather than employees, and thus that the Labor Board's order should be set aside as exceeding the NLRB's jurisdiction. We agree.

* North American Van Lines (NAVL) engages in interstate trucking operations through several divisions. The Board's order was limited to drivers who carry loads for the company's Commercial Transport Division. NAVL solicits, bills, and otherwise works with the commercial clients who seek to ship goods. NAVL arranges for those loads to be hauled by calling upon the services of a pool of drivers, each of whom has some degree of equity interest in a truck cab and an ongoing and (for an indefinite period) nearly exclusive relationship of carrying loads for NAVL.

NAVL's efforts to match loads with available drivers create the intricate relationship between the company and the drivers that underlies the dispute in this case. NAVL personnel (specifically, "counselors" and dispatchers) use threats and promises of benefits to convince available drivers to carry contracted loads.* Additionally, the company engages in widespread efforts to maintain the pool of drivers who carry the loads. While the company attracts many experienced drivers who own their own trucks, it recruits, trains, and provides competitive financing (for the purchase of the truck cab) to the overwhelming number of drivers who haul for the Commercial Transport Division.

In the action that gave rise to this case, NAVL created an "Advisory Council," composed of management representatives and interested drivers selected by the company, to discuss the company's policies affecting the drivers. The company supported and shaped the operation of the council. Council members discussed issues and information that indisputably concerned working conditions; in addition, the council made suggestions to NAVL management concerning benefits and working conditions. The company accepted and acted upon many of those suggestions.

One driver, who had unsuccessfully sought to join the council, filed a charge alleging that the company's support of the council constituted an unfair labor practice. Following an extensive hearing, the Administrative Law Judge agreed. The ALJ found that the council was a labor organization and that NAVL, through its active involvement, had unlawfully interfered with the organization's formation and operation. ALJ Decision 26, J.A. at 27. The ALJ also concluded that the drivers were employees rather than independent contractors. Id. The NLRB summarily affirmed the ALJ's decision. North American Van Lines, Inc., 288 NLRB No. 11 at 1-3 (March 10, 1988), J.A. at 1-3.

NAVL petitions for review of the Board's decision, claiming that the drivers are independent contractors, not employees, and thus that the company-driver relationship lies outside the Board's jurisdiction. The NLRB cross-applies for enforcement. For the reasons that follow, we grant the petition and deny the cross-application.

Governing precedent makes clear that the court is not "to extend any great amount of deference" to the Board's determination that workers are employees rather than independent contractors. Local 777, Democratic Union Org. Comm., Seafarers Int'l Union v. NLRB, 603 F.2d 862, 872 (D.C.Cir.), reh'g denied, 603 F.2d 891 (1978) (separate statement of decision). All parties agree (and the cases firmly establish) that Congress intended that traditional agency law principles guide the determination whether workers are employees (and thus fall within the Board's jurisdiction) or independent contractors (and thus are outside that jurisdiction). See, e.g., NLRB v. United Ins. Co., 390 U.S. 254, 256, 88 S. Ct. 988, 989-90, 19 L. Ed. 2d 1083 (1968) ("The obvious purpose of this amendment [to NLRA Sec. 2(3) ] was to have the Board and the courts apply general agency principles in distinguishing between employees and independent contractors under the Act.").

"[S]uch a determination of pure agency law involve[s] no special administrative expertise that a court does not possess." Id. at 260, 88 S.Ct. at 991. Quite apart from the law, the Board's underlying findings of fact are essentially undisputed. As in the case that most directly guides us, "[b]asically the issue ... calls for applying general principles of the law of agency--the distinction between employees and independent contractors--to undisputed facts." Local 777, 603 F.2d at 872. Thus, "we need not accord the Board's decision that special credence which we normally show merely because it represents the agency's considered judgment." Id.

The distinction between employees and independent contractors determines the boundaries of the Board's jurisdiction, and consequently invokes one of this court's principal functions: ensuring that the Board exercises power only within the channels intended by Congress. Indeed, this court and our counterparts elsewhere have repeatedly taken the Board to task for exceeding its proper powers by characterizing independent contractors as employees. See, e.g., Yellow Taxi Co. v. NLRB, 721 F.2d 366, 381-83 (D.C.Cir.1983) (MacKinnon, J.); see also id. at 382 n. 37 (listing 12 Board assertions of jurisdiction, rejected by courts of appeals, over independent contractors); id. at 383 & n. 38 (listing other circuits' criticisms of the Board's assertions of jurisdiction over independent contractors).

That being said, it remains the case that some deference to the Board's conclusion is required in this type of case. The reason is that the jurisdictional boundary at issue here is no bright line. The legal distinction between "employees," on the one hand, and "independent contractors" on the other is permeated at the fringes by conclusions drawn from the factual setting of the particular industrial dispute. As long as the Board's conclusions accord with principles of agency law and the Board's exercise of power does not spill over its jurisdictional boundaries, the court must, in light of this factual predicate to the legal determination, allow some latitude for the Board's judgment. Mindful of these broader principles and the applicable legal standard, a reviewing court will uphold the Board if "it made a choice between two fairly conflicting views." NLRB v. United Ins. Co., 390 U.S. at 260, 88 S. Ct. at 991.

In applying traditional agency law principles, the NLRB and the courts have adopted a right-to-control test. The test requires an evaluation of all the circumstances, but "the extent of the actual supervision exercised by a putative employer over the 'means and manner' of the workers' performance is the most important element to be considered in determining whether or not one is dealing with independent contractors or employees." Local 777, 603 F.2d at 873; see also Yellow Taxi Co., 721 F.2d at 374; Building Material and Dump Truck Drivers, Teamsters Local No. 36 v. NLRB, 669 F.2d 759 (D.C.Cir.1981), aff'd sub nom., Shepard v. NLRB, 459 U.S. 344, 103 S. Ct. 665, 74 L. Ed. 2d 523 (1983); City Cab Co. v. NLRB, 628 F.2d 261, 264 (D.C.Cir.1980); Lodge 1858, Am. Fed'n of Gov't Employees v. Webb, 580 F.2d 496, 504-08 (D.C.Cir.), cert. denied, 439 U.S. 927, 99 S. Ct. 311, 58 L. Ed. 2d 319 (1978); Restatement (Second) of Agency Sec. 220 (1958).

Significant limits, relevant to this case, 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 performance do not make the worker an employee. See, e.g., City Cab Co., 628 F.2d at 264; Local 777, 603 F.2d at 872-73, 872 n. 24. Such global oversight, as opposed to control over the manner and means of performance (and especially the details of that performance), is fully compatible with the relationship between a company and an independent contractor.

Second, restrictions upon a worker's manner and means of performance that spring from government regulation (rather than company initiatives) do not necessarily support a conclusion of employment status. See, e.g., Local 777, 603 F.2d at 875-76. Indeed, employer efforts to ensure the worker's compliance with government regulations, even when those efforts restrict the manner and means of performance, do not weigh in favor of employee status. "[T]he employer cannot evade the law ... and in requiring compliance with the law he is not controlling the driver. It is the law that controls the driver." Id. at 875.

Third, evidence of unequal bargaining power, without more, is not to be taken as evidence of control over the manner and means of the worker's performance. As Local 777 points out, the court will draw no inference of employment status from "merely the economic controls which many corporations are able to exercise over independent contractors with whom they contract." Id. at 874 (citing Carnation Co. v. NLRB, 429 F.2d 1130, 1134 (9th Cir.1970)).

Other factors weigh in the determination. To aid the evaluation of the extent of control, courts examine the extent to which the worker has assumed entrepreneurial risk and stands to gain from risks undertaken; any company control over details of the assigned task's performance; the worker's possession of an ownership interest in the tools of the trade; the nature of the parties' understanding; the method of payment (by job or by time); the local business practices regarding supervision of that activity; the degree of skill the job requires; and the treatment and benefits the company provides to the workers in question, compared to other workers who clearly are employees. See generally Yellow Taxi Co., 721 F.2d at 366; Local 777, 603 F.2d at 862; Restatement (Second) of Agency Sec. 220 (1958). The court may, in limited circumstances, examine whether the goodwill resulting from performance inures to the company or contractor. But see Local 777, 603 F.2d at 898 ("[Goodwill] is only of minimal, if any, importance."). However, these factors are of far less importance than the central inquiry whether the corporation exercises control over the manner and means of the details of the worker's performance; indeed, these factors are probative only to the extent that they bear upon and further that inquiry. See, e.g., Restatement (Second) of Agency Sec. 220 (1958).

On balance, we are persuaded that the Board's conclusion in this case cannot be sustained. The evidence of record strongly, clearly establishes that the drivers in question are, in fact, independent contractors. As we shall now see, the Board's treatment of the evidence reflects deep confusion between NAVL's control of the ends of the drivers' performance (which the evidence does establish) and NAVL's control over the manner and means of the driver's performance of the assigned tasks (which the evidence manifestly does not establish). Only the latter is, as we have indicated, the consideration that establishes employee status. The Board also substitutes evidence of NAVL's efforts to ease or facilitate the drivers' performance of their task for evidence of NAVL's control over the manner and means of that performance.

* We regret to say that the Board's conclusion does not reflect even a fairly conflicting view of the evidence--a view that the Board might be allowed to prefer, even though in the first instance the court might reach a different conclusion.

A dispassionate survey of the evidence establishes beyond cavil that the drivers are independent contractors. Here are the principal reasons why that is so. First and foremost, the drivers exercise a great degree of control over both the broad outlines and the manner and means of how they perform their jobs. The drivers retain nearly absolute control over their performance in the cab, their dress, the course they follow in hauling the load, when they work (while hauling loads and in between hauling loads), and where and when they stop, dine, or rest in the midst of driving. Second, the drivers hold an equity interest in their cabs, and a significant minority owns its cabs outright (for a much smaller minority, each driver owns several cabs and subcontracts work to drivers in his or her employ). Third, the drivers assume significant entrepreneurial risks through the control they exercise over the means and manner of performance; they decide when and how the truck must be maintained or repaired, how they will drive, and how frequently they will drive (and thus the amount of their gross income).

Additionally, the relationship between NAVL and the drivers partakes of the commissioning of an independent contractor. NAVL provides a range of fringe, supplemental, and profit-sharing benefits to its (admitted) employees that it does not extend to the drivers in question. NAVL regulates the details of performance and supervises the daily behavior of employees in ways it does not attempt to apply to drivers. NAVL withholds taxes from its payments to employees, but not from payments to the drivers. NAVL pays the drivers based upon the job, rather than upon the amount of time worked. The company does not require that the driver (as opposed to an employee or associate of the driver) perform the hauling or loading (and unloading). The system that NAVL uses to arrange for drivers to carry the company's loads provides the drivers with control over the most fundamental elements of the job: vacation (amount and timing), gross income, and location of job performance. See infra p. 602.

On balance, the drivers' control over the manner and means of performance, and their overall independence, bear (indeed, surpass) the hallmarks of independent contractor status found sufficient by this circuit to establish that status in other cases. See, e.g., Yellow Taxi Co., 721 F.2d at 374; Building Material and Dump Truck Drivers, 669 F.2d at 759; Local 777, 603 F.2d at 862. The evidence, it appears to us, overwhelmingly weighs on the "independent contractor" side of the balance when we confront what Local 777 deems "the fundamental question":

whether or not the putative employer has the right to control the driver during the course of his operation of the cab in the manner and means in which he earns his income and whether the drivers can be most aptly described as working for themselves or for a wage they receive from companies.

Local 777, 603 F.2d at 874. Each element of that question, applied to the drivers in the case before us, must be answered in favor of independent contractor status. Cf. Precision Bulk Transport, Inc., 279 N.L.R.B. 437 (1986) (truck drivers with roughly similar hauling arrangement found to be independent contractors); Don Bass Trucking, Inc., 275 N.L.R.B. 1172 (1985) (same).

In contrast to this substantial evidence of the drivers' independent contractor status, very modest evidence, at best, supports the Board's conclusion that NAVL exercises more than minimal control over the means and manner of the drivers' work. The ALJ's opinion elaborates a broad system of NAVL oversight of the drivers, and characterizes that relationship in a manner most conducive to the inference of control. However, the ALJ's opinion, with all respect, provides precious little evidence or specific findings of control over the manner and means of the drivers' performance. We shall now describe the import of that evidence and detail the principal shortcomings (and even contrary implications) of the Board's asserted support for its finding of employee status.

As we alluded to before, nearly all the Board's arguments and the evidence it has arrayed reflect a fundamental confusion between NAVL's control over the means and manner of the drivers' work and NAVL's oversight of the ends (that is, the overall performance) of their work. See supra p. 599. This defect infects two broad bodies of evidence, comprising the lion's share of the Board's case: (1) evidence addressing the company's efforts to convince the drivers to carry more loads, and (2) the company's intricate system of rewards and punishments.

* First, much of the Board's evidence, and the crux of the Board's case, concerns the company's efforts to induce the drivers to haul more loads. See ALJ Decision at 14-17, 20-21, J.A. at 18-21, 23-23A. NAVL's dispatcher system, most aspects of NAVL's record-keeping, the anecdotal evidence of threats, and the more formal threats of discharge (the "30-day letter" practice) predominantly function to increase the chances that a driver will take the loads that need to be shipped--and especially to keep drivers from turning down too many loads over too long a period.

The essential nature of the relationship between NAVL and the drivers emerges most graphically in how the drivers can (and routinely do) turn down particular loads. See, e.g., J.A. at 189-91, 535-41, 710-12, 2086-89, 2096-97; see also Brief for Petitioner at 27 ("In only one-third of NAVL's communications with contractors do they agree to accept a load."). NAVL's bonus system, which rewards drivers who take undesirable loads, and indeed the dispatcher system, which exists to convince (rather than order) drivers to take particular loads, see supra n.*, further attest to the discretionary nature of the driver's decision.. The company will discipline (or even terminate the contract with) drivers who do not over long periods haul a sufficient number of loads. But that discipline enforces control over the overall performance of the drivers.

By holding and exercising the power to decline particular loads, however, the driver retains control over the outlines and manner and means of the performance of the assigned work. The driver determines when to take a vacation, and for how long. The driver determines whether the cab requires maintenance or repair, or can be kept in service. The driver determines how many loads he or she will carry over a long period, and thus the amount and timing of the driver's income. In choosing among available loads, the driver determines what part of the country to and through which he or she will travel, and where the driver might pass time (or take a vacation) after hauling the load. This discretion, and how that discretion in this particular context augments the driver's control of the particular nature of the work and the particulars of performance at any given time, surround the drivers with powerful indicia of independent contractor status.

In other situations, the ability to refuse particular loads may not bear so decisively in favor of independent contractor status. Especially in cases where the worker (1) gains little control over the nature or performance of the job by declining particular loads; (2) bears a heavy burden for exercising the option to decline loads; (3) has little incentive in practice to decline loads; or (4) is subject to various restrictions upon the manner and means of performance, employee status might still exist even though the driver enjoys the ability to decline particular loads. All these factors existed in and diminish the importance of the case the Board most relies upon, City Cab Co. v. NLRB, 628 F.2d at 261. In that case, the court noted that "[t]echnically, a driver may refuse to accept the dispatcher's calls" and "[t]echnically, ... could refuse an airport fare," but found that "[t]hat independence is illusory because of the nature of the taxi market in Orlando," which in practice curtailed the ability of drivers to decline particular passengers. Id. at 264-65. In the case before us, however, the drivers' ability to decline particular loads is manifest and the exercise of that ability leads to significant independence in practice. See supra 600-01, 602. Additionally, the company in City Cab Co. imposed significant restrictions upon the taxi cab drivers that, combined with restrictions imposed by the nature of the taxi business, supported the finding of employee status and that are not present in the case before us. See City Cab Co., 628 F.2d at 264-65 (describing, e.g., company measures "significantly regulat[ing] the hours that drivers work" and "prescrib[ing] an extensive dress code for its drivers"). The determination to be made is, of course, a contextual one, producing the result that a particular element of worker discretion (here, whether to decline a load) will yield different consequences for the worker's ability to control the manner and means of the job's performance in different contexts, and thus also yield differing ultimate legal conclusions. Cf. NLRB v. United Ins. Co., 390 U.S. at 258, 88 S.Ct. at 991 ("[T]here 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.").

Confusion between NAVL's control over the worker's overall performance and control over the means and manner of the worker's performance of the task likewise infects another principal body of evidence relied upon by the Board, namely that relating to the company's system of driver rewards and punishments.

The Board unjustifiably treats evidence of the operation of the training system, promotion opportunities and criteria, dispatchers' references to promotion, and use of the "30-day letter" and discharges as evidence of the company's control over the manner and means of the drivers' performance. See ALJ Decision 9-10, 13-17, J.A. at 13-14, 17-21. Although the evidence indisputably establishes the existence of an elaborate incentive system, the evidence fails to establish that the company employs those incentives to control the manner and means of the drivers' performance. At most, the Board's anecdotes and arguments establish that NAVL uses its incentive system to persuade, convince, and jawbone drivers into hauling more loads. In this context, that control, as we have seen, supra 601-02, amounts to influence over the ends or overall character of the drivers' performance--control entirely compatible with the drivers' independent contractor status. The inference of control characteristic of an employment relationship (and the Board's conclusion reflects little more than an inference) is one that the court will enforce only if supported by direct evidence. That is not so here.

To the contrary, the evidence suggests that the company employs its incentive system to ensure that the drivers' overall performance meets the company's standards. The counselor system, for example, largely fulfills this function of overall evaluation. While counselors do answer questions about particular loads, drivers initiate the bulk of those relatively infrequent contacts. See Reply Brief for Petitioner at 10 n. 6 (and material cited). On balance, however, most of the counselors' work involves oversight of the drivers' overall performance; NAVL's evaluations of the counselors' treatment of the drivers, for example, focus primarily upon how the counselors handle terminations, debt load, and amount of work per driver (reflected in the drivers' gross earnings and time out of service). See J.A. at 1644-1795. These functions, including the counselors' use of the "30-day letters," see Brief for Respondent at 14 (and material cited), appear, in the context of the entire company-driver relationship, to be part of the monitoring of overall performance (and ongoing review of whether to continue to contract with the driver) that any company might maintain over workers who clearly are independent contractors.

Similarly, the ALJ complains that in its training program NAVL holds out the possibility of transfer to more desirable divisions if drivers, in essence, keep the company happy; that dispatchers hold out similar promises; and that transfers ultimately follow accordingly. See ALJ Decision 14-16, 19, J.A. at 18-20, 22A. Drivers appear to keep the company happy, however, simply by conforming to very general, ends-oriented standards of overall performance, rather than by submitting to control over the means and manner of performance. To the extent it is probative at all, the evidence suggests that the company rewards drivers who complete their runs on time, have low rates of customer dissatisfaction, do not accumulate excessive debt, and haul a sufficient number of runs. Correspondingly, the company wields the proverbial stick when a driver's overall performance, as evaluated under similar criteria, falls short of expectations. These criteria accord entirely with independent contractor status. In sum, the company's incentive system, quite apart from how the company employs it, creates no employment relationship.

The Board also points to a range of measures through which the company assembles and maintains the pool of drivers to haul its loads. The ALJ's opinion describes (1) NAVL's support to drivers to allow them to finance the purchase of their cabs; (2) NAVL's provision of truck service centers on the highways for the drivers' use; (3) NAVL's system of advancing funds to drivers for operating (and other) expenses; and (4) NAVL's limits on the ability of drivers to haul loads for other companies (dubbed "trip-leasing"). See ALJ Decision 11-13, 17-18, J.A. at 15-17, 21-22. In fairness, this evidence does indeed support the conclusion that these measures are at odds in some respect with the traditional independent contractor relationship. The various ways in which the company supports the drivers' efforts hold the potential for the company to use its significant resulting leverage to render the drivers, in practice, employees. However, the evidence once again neither establishes that conclusion nor supports the inference that the company controls the manner and means of the drivers' performance of their work.

Upon close examination, most of those elements fall far short of establishing employment status. In the main, they reflect NAVL's efforts to support the drivers' efforts to establish their businesses and complete their contracted tasks, as opposed to an effort to assert control over the details or manner of performance. For example, NAVL offers financing at competitive rates and frequently sells the financing agreement to third parties. NAVL-provided service centers are also open to other, non-affiliated drivers on similar terms. In addition, NAVL requires that its drivers repay the advances that it provides, and limits the amount of this debt that drivers may have outstanding.

NAVL's restrictions upon "trip leasing," the ability of drivers to use their own truck cab and NAVL's tractor to haul loads for other shipping companies, does trench upon the drivers' independence. Government regulations account for some of the limits imposed upon trip leasing, reducing somewhat the relevance of this restriction to the determination of the drivers' employment status. See supra 599. Apart from the government restrictions, however, the NAVL-inspired limits on trip leasing (which the parties sharply contest) do not amount to or yield control over the manner and means of performance sufficient, in the context of this case, to render the drivers employees. Rather, the ultimate determination in this case requires a broad examination of all facets of the relationship between company and driver, see NLRB v. United Ins. Co., 390 U.S. at 258, 88 S.Ct. at 990; the great degree of driver control over the nature of and manner in which the job is performed, as we have previously recounted, overwhelms the restrictions upon the manner and means that might accompany the limits upon trip leasing. Accord Precision Bulk Transport, 279 N.L.R.B. at 437-38 (truck drivers found to be independent contractors, despite effective bar on trip leasing, "in light of the record evidence showing that the owner-operators enjoy certain freedoms and bear certain risks consistent with the operation of an independent business").

For the foregoing reasons, we find ourselves obliged by the overwhelming weight of the evidence to conclude that the drivers in question are independent contractors. NAVL, we are persuaded, is justified in its complaints that the ALJ applied the incorrect legal test, miscast the burden of proof, and otherwise unduly narrowed the statutory exclusion of "independent contractor[s]" from the Board's jurisdiction. See Brief for Petitioners at 15-22. However, the ALJ and Board's error was more directly and fatally in the lack of support in the record for their fundamental legal conclusion. The ALJ, as we have now seen, mustered essentially uncontested facts regarding the relationship between NAVL and a subset of its drivers, and then inferred (rather than substantiated) the existence of the company's control over the manner and means of the drivers' performance. For its part, the Board, not for the first time, overstepped the bounds of its jurisdiction by attempting to relabel independent contractors as employees. Cf. Yellow Taxi Co., 721 F.2d at 381-83, 382 n. 37, 383 n. 38 (MacKinnon, J.) (chronicling similar Board errors). It did so by, in effect, rubber-stamping the ALJ's flawed approach and his erroneous conclusions.

For the reasons already stated, we are convinced that the evidence, fairly considered, fails to support the conclusion that the drivers are employees under traditional agency law principles. We therefore grant NAVL's petition for review and deny the Board's cross-application for enforcement.

Judgment accordingly.