The National Labor Relations Board petitions to enforce its order against Hillard Development Corporation,1 doing business as Provident Nursing Home. The Board found that Provident violated §§ 8(a)(5) and (1) of the National Labor Relations Act, 29 U.S.C. § 158(a)(5), (1), by refusing to bargain with the Union as a representative of a bargaining unit that includes thirteen district and charge nurses. Provident argues that its refusal to bargain was not an unfair labor practice because the district and charge nurses are supervisors, as that term is defined in § 2(11) of the Act, 29 U.S.C. § 152(11), and the nurses as such are precluded from participating in collective bargaining. See 29 U.S.C. § 152(3),(11).

Whether mid-level care providers such as Provident's district and charge nurses are supervisors under § 2(11) is a significant legal issue that has divided the circuits. It is also an issue of some societal significance, affecting increasing numbers of people who will need nursing home care as the post World War II baby boomer generation ages.2 The issue is significant in part because labor costs in the healthcare industry comprise a large portion of overall costs (estimated to be roughly 60% of hospital costs).3 The issue is important both to management, concerned with economic viability, and to employees, concerned about job security and workplace rights.

Historically, the NLRB itself has proven unsympathetic to employers' arguments that such nurses may not be unionized. Indeed, the NLRB earlier adopted a unique, more hostile test than that used for other professions to determine whether nurses were supervisors. Under the Board's "patient care" test, nurses were not considered to be exercising authority "in the interest of the employer," as required under the definition of supervisor in § 2(11),4 if they directed less-skilled employees only "in the exercise of professional judgment incidental to the treatment of patients." NLRB v. Health Care & Retirement Corp. of America, 511 U.S. 571, 574 (1994) ("Health Care & Retirement Corp.") (internal quotation marks omitted). The Supreme Court set aside that test in 1994 as contrary to the Act. See id. at 574-84.

This history leads Provident to suggest that less deference should be given to the Board's interpretation of § 2(11) than would normally be required under Chevron USA Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984). Provident also argues that the Board's conclusions are not, in any event, supported by substantial evidence. While some of the issues are close, we enforce the Board's order. We find no reason not to apply our usual standard of deference to the Board's interpretations of ambiguous portions of the Act, and we determine, upon reviewing the record, that the Board's findings of fact are adequately supported.

We recount the facts, drawing supportable inferences in a manner consistent with the Board's findings. Provident operates a for-profit 112-bed residential nursing home in Brighton, Massachusetts. It provides intermediate care for geriatric residents with major mental illnesses. The home has two floors. The first floor houses fifty-two residents and is divided into two districts, or units, both of which are served by one nursing station. The second floor has sixty patients, and is divided into two districts, each of which has its own nursing station. The nurses provide both nursing care and general assistance with residents' daily activities. The Director of Nursing has overall charge of the Nursing Department. Under her are three registered nurses who serve as Unit Managers. The Unit Managers and the Director of Nursing are concededly supervisors.

Under the Unit Managers are approximately thirteen nurses who are employed as district nurses and charge nurses. These nurses, who are either registered nurses or licensed practical nurses, directly provide patient care. The district nurses work the day shift and have some responsibility for reviewing the work and documentation from the other two shifts. Charge nurses perform the same duties as district nurses, but without any monitoring responsibilities outside of their own shift. Charge nurses work on the evening and night shifts. Under these district and charge nurses are approximately thirty Mental Health Assistants ("MHAs"). MHAs are responsible for assisting residents with their daily activities, including bathing, dressing, eating, and walking. Each district and charge nurse oversees the work of two to three MHAs. Between six at night and six in the morning, the charge nurses are the highest level of authority at the facility.

Unit Managers organize residents into groups and determine the duties to be performed by MHAs during each shift for each group. A district or charge nurse then assigns each MHA to one of these predefined groupings, based on the residents' needs and particular MHA skills. Nurses also determine the order in which MHAs take lunches and breaks, within boundaries established by Provident. Because the resident group assignments are not often changed, the MHAs are generally familiar with their assigned duties and do not need to be directed closely. If an MHA is absent and is not replaced for the day, a nurse redistributes the absent MHA's duties to the remaining MHAs on shift. If an MHA has a complaint about an assignment or other work issue, the MHA can bring the complaint to a nurse. If the MHA and nurse cannot resolve the problem, the two will bring it to a Unit Manager. In emergency situations requiring extra MHA assistance, nurses may ask MHAs to work after their shifts have ended or temporarily transfer MHAs between floors. In either event, the nurse will inform the Unit Manager of the nature of the emergency and the solution implemented.

District or charge nurses who observe MHAs abusing residents or acting insubordinately document their observations on an "Employee Counseling Form" or an "Anecdotal Record." MHAs are also required to document any instances of MHA misconduct on an Anecdotal Record. These forms do not ask for a recommended disciplinary measure. Unit Managers decide whether the reports will be labeled verbal or written warnings within Provident's progressive disciplinary system.

A district or charge nurse who finds that an MHA is inappropriately dressed, intoxicated, fighting with other staff members, or sleeping on the job has the authority to send the MHA home. In this situation, the nurse will call a Unit Manager to suggest that the MHA be sent home or to notify the Unit Manager that he or she plans to do so.

Provident's MHAs receive annual written evaluations of their work. Unit Managers assign the nurse who works most frequently with each MHA to prepare the MHA's evaluation. The nurse comments on different areas of the MHA's performance and gives the MHA a score between one and five in each area. Both the Unit Manager and the Director of Nursing then review the evaluation form. They may question the nurse if they feel that a particular rating is too high or low and may ask the nurse to change it or at least review its accuracy before showing the evaluation to the MHA. An MHA who disagrees with a rating can also ask the nurse to change it.

In 1991, Provident began using the evaluation scores as a basis for an annual merit pay increase for some of the MHAs. Certain MHAs do not receive merit pay tied to their evaluation scores: MHAs who are paid per diem, MHAs already earning maximum wages under Provident's salary structure, and MHAs who receive merit pay based on administrative decisions apart from the evaluation system. In early 1993, for example, the administrator raised the pay range for a substantial number of MHAs and changed the evaluation anniversary date for those MHAs. Accordingly, those MHAs were evaluated in 1993 but did not receive merit increases based on those evaluations.

For those employees who are eligible for merit increases, Provident uses a scale that matches a range of evaluation scores with a certain level of pay increase. Provident's administrator and Director of Nursing change the scale periodically and may eliminate merit increases altogether, depending on the nursing home's financial circumstances. Although the charge and district nurses are aware that the evaluation scores they give may affect an MHA's pay, they are generally unaware at the time they write the evaluation of the precise correlation between any given score and the pay increase to which it corresponds. Once the evaluation is complete, the nurse gives the form to the Director of Nursing, who compiles an average score, determines the increase associated with that score, writes on the form the increase that she recommends, and then sends it to the administrator for review. The administrator may approve or disapprove the increase or give the MHA a higher increase than would otherwise correlate to the MHA's evaluation score.

On October 12, 1993, Local 285 of the Service Employees International Union filed a petition with the Board seeking certification as the representative of a unit composed of Provident employees, including both the MHAs and the district and charge nurses. Provident contended that the Board should exclude the district and charge nurses from the bargaining unit because the nurses are supervisors under § 2(11) and thus ineligible for union representation under the Act. After conducting a hearing, the Board's Regional Director issued a decision on November 24, 1993, in which she directed an election and rejected Provident's contention that the nurses are supervisors.

Provident then sought Board review. While the request for review was pending, the Board's regional office conducted an election and impounded the ballots cast. The Board later agreed to review the Regional Director's decision "solely with respect to the Regional Director's finding that evaluations prepared by the Employer's district nurses and charge nurses . . . are not indicative of supervisory status," denying Provident's request for review "[i]n all other respects." Provident filed a request for reconsideration of the portion of its request denied by the Board.

In June 1996, the Board granted Provident's request for reconsideration and remanded the case to the Regional Director for further consideration in light of the Supreme Court's decision in Health Care & Retirement Corp. and the Board's decisions in Ten Broeck Commons, 320 N.L.R.B. 806 (1996), and Providence Hospital, 320 N.L.R.B. 717 (1996).

The Regional Director's supplemental decision concluded once again that the nurses were not supervisors under the Act. On review, the Board affirmed the Regional Director's conclusion that "the nurses' role in evaluating employees is insufficient to confer supervisory status." Hilliard Dev. Corp., Case 1-RC-20057, at 2 (NLRB Feb. 6, 1997). Accordingly, the regional office counted the ballots and ascertained that the Union had won the election. The Regional Director rejected Provident's objections and the Board denied further review.

In order to obtain judicial review of the bargaining unit, Provident refused to bargain with the Union, prompting the Union to file a charge with the Board under §§ 8(a)(5) and (1) of the Act. Provident's answer to the complaint admitted its refusal to bargain but disputed the propriety of the Board's certification. The Board's Decision and Order reaffirmed its earlier finding that Provident's district and charge nurses are not supervisors under § 2(11) of the Act. See Hilliard Dev. Corp., 324 N.L.R.B. No. 46 (Aug. 20, 1997). This application for enforcement of the Order followed.

This court must accept the Board's factual findings as to which employees are supervisors unless those findings are not supported by substantial evidence in the record as a whole. See 29 U.S.C. § 160(e); Telemundo de Puerto Rico, Inc. v. NLRB, 113 F.3d 270, 274 (1st Cir. 1997) (citing Universal Camera Corp. v. NLRB, 340 U.S. 474, 488 (1951)). Under the substantial evidence standard, "[t]he Board . . . may not distort the fair import of the record by ignoring whole segments of the uncontroverted evidence." Maine Yankee Atomic Power Co. v. NLRB, 624 F.2d 347, 360 (1st Cir. 1980). However, "the possibility of drawing two inconsistent conclusions from the evidence does not prevent an administrative agency's finding from being supported by substantial evidence." American Textile Mfrs. Inst. v. Donovan, 452 U.S. 490, 523 (1981) (internal quotation marks omitted). The ultimate question for the court in reviewing the Board's factual findings is "whether on this record it would have been possible for a reasonable jury to reach the Board's conclusion." Allentown Mack Sales & Serv., Inc. v. NLRB, 522 U.S. 359, 118 S. Ct. 818, 823 (1998).

Statutory interpretation questions are reviewed de novo. See 5 U.S.C. § 706. Where the intent of Congress is clear, that intent must be given effect. See Chevron, 467 U.S. at 842-43; NLRB v. Beverly Enterprises--Massachusetts, Inc., 174 F.3d 13, 22-23 (1st Cir. 1999). If, on the other hand, the statute is ambiguous or silent with respect to a specific issue, we will defer to the Board's interpretation so long as that interpretation is a "permissible" one, Chevron, 467 U.S. at 843, that is, one which is "rational and consistent with the statute," Beverly Enterprises--Massachusetts, 174 F.3d at 22 (internal quotation marks omitted).

Provident suggests that the Board is not entitled to deferential review of its interpretation of ambiguous portions of § 2(11) because, it says, the Board is biased and has a metaphorical thumb on the scale, tilting the balance toward a finding of non-supervisory status. It points out that a number of the circuits have rejected Board orders finding mid-level nurses in nursing homes not to be supervisors5 and that several have attributed to the Board a policy bias in this area.6 As evidence of bias, Provident notes that the Board has found nurses to be supervisors only once since the decision in Health Care & Retirement Corp. See First Healthcare Corp., 323 N.L.R.B. 1171 (1997).

We reject Provident's suggestion. After Health Care & Retirement Corp., the Board held that it would no longer use a special test to evaluate whether nurses are supervisors under § 2(11) but would "treat . . . nurses the same as all other employee classifications and . . . apply to them the same test as [that] appl[ied] to all other employees." Ten Broeck Commons, 320 N.L.R.B. at 810. There is no reason, from the facts of this case, to believe that the Board continues de facto, if not de jure, to apply a special test to nurses. The Board here has carefully reviewed the facts, itself evidence that the outcome of this case was not pre-ordained by the Board. Indeed, Provident does not really argue this point. Instead, Provident's argument is basically that the Board's interpretations of particular statutory phrases establish tests in which matters of judgment tend to be resolved against the employer, when a more impartial and reasoned exercise of judgment would be in favor of the employer. If that is so, then the remedy is in Congress.7 So long as Congress has assigned this interstitial policy-making role to the Board, we adhere to our usual rules of deference.8 Accord NLRB v. GranCare, Inc., 170 F.3d 662, 665-66 (7th Cir. 1999) (en banc); see also Beverly Enterprises, Virginia, Inc. v. NLRB, 165 F.3d 290, 299-302 (4th Cir. 1999) (en banc) (Philips, J., dissenting) (rejecting the argument that the Board's purported policy bias gives the court a "licence . . . to skew normal standards of judicial review").

The NLRA at § 2(11) defines as a supervisor:

[A]ny individual having authority, in the interest of the employer, to hire, transfer, suspend, lay off, recall, promote, discharge, assign, reward, or discipline other employees, or responsibly to direct them, or to adjust their grievances, or effectively to recommend such actions, if in connection with the foregoing the exercise of such authority is not of a merely routine or clerical nature, but requires the use of independent judgment.

29 U.S.C. § 152(11). This definition is disjunctive: any one of the listed aspects of authority may signify supervisory status, as long as the exercise of that authority involves the use of independent judgment. See Telemundo de Puerto Rico, 113 F.3d at 273. The Board has set forth a three-part test to interpret this statutory definition: (1) whether the individual has the authority to perform any one of the twelve defined duties, (2) whether exercising that authority requires the use of independent judgment, and (3) whether the employee exercises the supervisory authority in the interest of his or her employer. See Health Care & Retirement Corp., 511 U.S. at 574.

B. The Board's Interpretation of "Independent Judgment"

The Court in Health Care & Retirement Corp. rejected the Board's "patient care" analysis of the phrase "in the interest of the employer," noting that "[w]ith respect to that particular phrase, we find no ambiguity supporting the Board's position." Id. at 579. However, the Court was careful to point out that the case before it did not involve certain other phrases in § 2(11) which the Court recognized to be ambiguous. See id. (finding it "no doubt true" that "phrases in § 2(11) such as 'independent judgment' and 'responsibly to direct' are ambiguous, so the Board needs to be given ample room to apply them to different categories of employees"). The ambiguous phrase "independent judgment," which was "irrelevant" in Health Care & Retirement Corp., id., is at issue here.

The ambiguity of this phrase stems in part from its similarity to the phrase "consistent exercise of discretion and judgment" in § 2(12) of the Act. See 29 U.S.C. § 152(12). The latter phrase describes the work of professional employees, who are covered under the Act. See 29 U.S.C. § 152(3), (12). The Act's simultaneous exclusion of supervisors and inclusion of professional employees creates an inherent tension which the Board and the courts have tried to resolve since enactment of the Taft-Hartley Act in 1947. See D. Rabban, Distinguishing Excluded Managers from Covered Professionals Under the NLRA, 89 Colum. L. Rev. 1775, 1798 (1989); see also D. Barker, Note, NLRB v. Health Care & Retirement Corp.: Erosion of NLRA Protection for Nurses and Other Professionals?, 1996 Wis. L. Rev. 345, 346 (1996).

In the aftermath of Health Care & Retirement Corp., the Board found there to be a distinction between the "independent judgment" exercised by statutory supervisors and the judgment routinely exercised by professional employees. The Board found that "making decisions requiring expert judgment is the quintessence of professionalism; mere communication of those decisions and coordination of their implementation do not make a professional a supervisor." Providence Hospital, 320 N.L.R.B. at 719. Accordingly, the Board held that where the judgment exercised by nurses in assigning and directing employees is indistinguishable from the judgment that professional nurses routinely exercise, "independent judgment" under § 2(11) has not been established. See id. at 730.

The Regional Director applied the Providence Hospital rule in her supplemental decision, which once again found Provident's district and charge nurses not to be supervisors. On review, the Board affirmed this conclusion. In arguing against enforcement of the Board's order, Provident suggests that the interpretation of "independent judgment" which the Board adopted in Providence Hospital should be rejected as unfounded.

We disagree. The Board's interpretation is not irrational or inconsistent with the statute. To the contrary, it harmonizes the Act's definitions of "supervisor" and "professional employee" in a sensible way, consistent with Congress's intent to exclude as supervisors only those employees with "genuine management prerogatives." NLRB v. Bell Aerospace Co., 416 U.S. 267, 280-81 (1974). Thus, we do not disregard the Board's interpretation of this phrase, as Provident urges. Instead we defer to that interpretation as a "permissible construction" of ambiguous language. Chevron, 467 U.S. at 843.

C. The Board's Application of the § 2(11) Test

In addition to challenging the Board's interpretation of "independent judgment," Provident says the Board disregarded record evidence in its conclusions and ignored substantial evidence concerning the supervisory authority of the district and charge nurses. Because Provident makes a manifold attack, we group the challenges and discuss each separately. Under Board precedent, the burden of proving supervisor status falls on the party making the assertion of supervisory status. See Beverly Enterprises--Ohio, 313 N.L.R.B. 491, 496 n.26 (1993). The question of whether that allocation of burdens is correct is not before us.

1. Evaluations and Effective Recommendation of Reward

Provident's keystone argument is that the district and charge nurses complete annual performance evaluations for the MHAs and that these evaluations directly determine the level of merit wage increases for the MHAs. This, Provident says, means that these nurses effectively recommend the reward for the MHAs, which brings them within the statutory meaning of "supervisor."

The Board found that "the evidence fails to establish that the evaluations performed by these nurses directly affect the employees' job status; therefore, the nurses' role in evaluating employees is insufficient to confer supervisory status." Hilliard Dev. Corp., Case 1-RC-20057, at 2 (NLRB Feb. 6, 1997).

We describe the Board's conclusions and the employer's rejoinders. The Board concedes that there is some relationship between the evaluations by the district and charge nurses and the award of merit pay in some instances, but says that this relationship does not suffice to create supervisory status.

First, the Board notes that the instances of influence over merit pay increases are neither universal nor inevitable. About half of the MHAs are per diem employees who do not receive merit pay increases at all. Even among the MHAs who are not per diems, another group does not receive merit increases because they are at or near the maximum wage or cap. There was evidence that these two groups combined constituted more than half of the MHAs in 1992. In 1993, nine of the nineteen MHAs who were otherwise eligible for merit increases -- that is, who were neither per diem employees nor had reached Provident's wage cap -- were nonetheless removed from the merit increase system by an across-the-board administrative merit increase, without any input from the district and charge nurses.

Provident presents the facts differently. Provident contends that in 1992, the majority of aides received raises based on evaluations by the district and charge nurses. Further, Provident says, in 1993, ten of twenty-three did; of the thirteen remaining, four were not eligible because they were per diem employees, and nine received the administrative merit increase, which was a one-time, atypical event.

Second, the Board found it was not clear how many of the district and charge nurses actually completed evaluations of MHAs or whether they knew what effect the evaluations would have on the aides. Nurses who testified said that they did not know their evaluations would affect the merit increases for the MHAs.

Provident says the subsidiary findings to the Board conclusions are wrong and are not supported by the evidence. As to the point that only a few nurses completed MHA evaluations (and did only one or two), Provident says that in 1992 and 1993 all but one of the nurses completed evaluations; many completed more than one. Further, Provident says the Board committed an error of law in considering the frequency with which evaluations were done. See Maine Yankee, 624 F.2d at 360 (noting that the question under § 2(11) is whether authority exists, not how frequently it is exercised). Regarding the Board's statement that it was unclear whether nurses knew the effect of rankings, Provident says this was contradicted by the nurses' testimony.

Third, and most significantly for our purposes, the Board found there was no consistent or direct correlation between the scores received on the evaluations by the charge and district nurses and the percentage of the merit increases (where awarded). There were several reasons for this. The increases awarded did not always match the evaluation scores given, and the range of merit pay increases awarded varied from time to time even within a year due to budgetary factors. In addition, the determination by the district and charge nurses of the evaluation score was subject to review by others and to independent assessment by the Director of Nursing.

With respect to the variance in range of increases received, the Board found that employees with the same ratings, but who were evaluated at different times of the year, received different percentage pay increases. The Board refers to several instances in which specific merit increases awarded by the administrator had no correlation to the ratings awarded, even under the formula in effect at the time: in 1992, two MHAs who received the same evaluation score were awarded different increases, and two MHAs with different evaluation scores received the same increase, contrary to the formula in effect at the time.

As to the responsibility of other supervisory personnel in the process, the evaluation form itself had a space for the Unit Manager to fill out. The job description for both the district nurse and charge nurse positions state that those nurses write the evaluations "[w]ith the . . . Unit Manager." In fact, both the Unit Manager and the Director of Nursing review the evaluations. The Board relied on the testimony of the Director of Nursing, who said that while she had never changed an individual rating herself, if she disagrees with a rating, she will "ask the unit manager to ask the nurse to redo it again to make sure that's the score they want to put in." There was also evidence that the independent evaluation made by the Director of Nursing resulted in higher pay raises for those who emerged with the same scores from the district and charge nurse evaluations but whom the Director gave a good evaluation.

As to the Board's conclusion that there was no consistent or direct correlation between scores received and percentage of merit increases, Provident responds that this is the inevitable result of a rolling (year by year) relationship between evaluation points and percentage increase. Regarding the potential for intervention in the evaluation process by the nurses' superiors, Provident says that such intervention rarely occurred. It also says this point is irrelevant, because the statute only requires the power to recommend reward. Provident relies on the Sixth Circuit's decision in Caremore to support this argument. See Caremore, 129 F.3d at 369 (concluding that nurses had authority to recommend reward because they evaluated aides, even though the nurses' evaluations were subject to administrative review).

None of the various circuit cases in the wars of the nurses, described above, turned on the phrase in § 2(11) which is the employer's lead argument here - the authority "effectively to recommend" reward of other employees. We accept, arguendo, that the evaluation process requires the use of independent judgment. Whether independent judgment is involved in the translation of that evaluation to a pay increase is a different question. But before that question of independent judgment is reached there is the prior question of whether the nurse's authority in this area is the authority "effectively" to "recommend" a "reward . . . [to] other employees." This combination of statutory phrases is not so precise as to rule out Chevron deference.

The Board, in turn, has appropriately noted that § 2(11) does not list "evaluate" as a supervisory function. The employee effectively recommends reward, the Board says, only when an employee's evaluation leads to pay changes, there is a "direct correlation" between the evaluations and merit increases or bonuses to the evaluated employees, and the employee's supervisors do not independently investigate or change the ratings. See First Healthcare Corp., 323 N.L.R.B. at 1171-72. The Board has thus interpreted "effectively recommend" to require a "direct correlation." This standard is to be given deference because it is neither irrational nor contrary to the plain language of the Act. The question then becomes whether the Board's conclusion that the standard has not been met itself meets the substantial evidence standard.

Two themes undergird the Board's conclusion: that there is no direct link between the evaluations and merit pay and that management independently reviews and can change the evaluations. Our review of the record shows that substantial evidence supports the Board's findings on both points. We focus on those themes and do not reach each individual argument made by Provident.

There is some merit in Provident's reproach that the mere use by management of rolling performance evaluations cannot turn a supervisor into a non-supervisor. If the Board's conclusion rested on this point, we might well agree with Provident. But we understand the Board to have concluded something else - that there really is no direct connection between the evaluation of an MHA given by a charge or district nurse and the merit pay increase, if any, that the MHA receives. Management retained and exercised the power over several intervening factors. The 1993 across-the-board administrative merit increase serves as an example. Provident rightly says a one-year exception should not create a rule. We have no way of knowing, from this record, whether the 1993 administrative merit increase was unique. More importantly, even in 1992 there was adequate evidence to support the Board's conclusion that merit pay is not directly linked to the evaluations. Two MHAs received different merit increases although they received exactly the same scores; two receiving different evaluation scores received the same increase. The evaluation themselves are independently reviewed by the Unit Managers, the Director of Nursing and the home's administrator. At least once a nurse was directly told to change a rating; several other times mere diplomatic suggestions to that effect were made. Management in fact reviewed and signed every evaluation and the Director of Nursing entered her own comments on most of the evaluations based on her own observations. This combination of factors differs from that in the case relied on by Provident, the Sixth Circuit decision in Caremore, Inc., and, in our view, suffices.

2. Assigning and Effectively Recommending Assignment of Work

Provident also argues that the district and charge nurses participate in and effectively recommend the assignment of work. It makes the following arguments in support of its contention that the nurses have authority to assign work9 to the MHAs. It says the job descriptions provide the nurses with authority to "assign tasks" to other staff; that the nurses actually assign MHAs to particular residents (through pre-determined groupings of residents); that if an MHA is absent, the nurses redistribute the work among the present MHAs and can ask an MHA to work beyond a shift or to float; and that the nurses determine when the MHAs take breaks, including lunch breaks. The employer also focuses on the fact that the nurses are the highest level of staff physically present at the nursing home on weekends and at night. Provident relies on several decisions from other circuits to support these arguments. See Glenmark Assocs. v. NLRB, 147 F.3d 333, 340-45 (4th Cir. 1998) (finding nurses had statutory authority to assign work); Caremore, 129 F.3d at 369 (same); Beverly California Corp., 970 F.2d at 1553 (same).

The Board found that while the district and charge nurses possess some assignment power, the power is circumscribed and does not amount to the exercise of independent judgment. The Board placed great weight on the fact that the Unit Managers select groups of residents for assignment to the MHAs; the nurses' only authority is to assign MHAs to those predetermined groups. The Unit Managers also determine the specific duties to be performed for each resident, as well as the shift or floor to which each MHA is assigned. Once MHAs are assigned to a group of residents, the assignments tend to remain the same. As to the occasional reassignment when MHAs are absent, the Board found that, although the nurses consider the needs of individual residents, the matching of skills to requirements was essentially routine. Further, all floating work or beyond shift work is based on predetermined requirements that there be a certain number of MHAs per unit. While nurses can request that an MHA work overtime, they cannot compel the MHA to do so. Once nurses do exercise their limited authority to transfer, they must immediately notify the Unit Manager, who reviews the decision. Finally, the determination of order of lunch and other breaks is essentially clerical.

The Board's conclusion that such hemmed-in, limited authority to assign work -- authority which is confined by predetermined groupings and schedules and is subject to post-action recission by a Unit Manager -- is not the independent judgment required by the Act is both rational and supported. See Northeast Utils. Serv. Corp. v. NLRB, 35 F.3d 621, 625 (1st Cir. 1994) (finding non-supervisory status where employee had no ability to assign work other than filling openings on a shift from a pre-determined list). The mere fact that an action is subject to review and to being countermanded by a higher-up employee does not alone mean that the acting employee is not a supervisor. See NLRB v. Metropolitan Life Ins. Co., 405 F.2d 1169, 1177 (2d Cir. 1968) ("[The] power to review held by the immediate supervisors of the [employees] whose status is at issue does not demonstrate that the latter are not 'supervisors' under the Section 2(11) definition."). But the Board does not rely on that fact alone, and its assessment of what is routine work as opposed to work requiring the exercise of independent judgment is adequately supported by the record.

The Board found that the nurses can resolve minor grievances, such as a complaint by an MHA about an assignment. If the complaint is not resolved, it is brought to the Unit Manager.

The Board found that this level of activity did not amount to supervisory authority because management has not granted the nurses authority to bind management, because implementation of a remedy for a grievance can only take place if all of the affected MHAs agree with it, and because the nurses have no power of enforcement, and, in particular, no power to reassign an MHA in response to another MHA's complaint without the consent of the Unit Manager.

The Board's conclusion that this limited role did not involve independent judgment is more than supported. See Northeast Utils. Servs. Corp., 35 F.3d at 625 (finding employees not to be supervisors where they can moderate disputes but do not have ultimate responsibility for the resolution of disputes); Stop & Shop Cos. v. NLRB, 548 F.2d 17, 20 (1st Cir. 1977) (same).

4. Disciplining And Effectively Recommending Discipline

The district and charge nurses do exercise some limited authority in the area of disciplining and recommending discipline. First, they document any instances of patient abuse or insubordination on forms. Second, the nurses send MHAs home when they observe certain defined infractions such as intoxication, fighting, sleeping on the job, or inappropriate dress.

In the first situation, the nurse will document the incident on one of the forms provided, which contain no designated space for the reporter to recommend discipline. Nonetheless, at times the nurse does include a written recommendation as to discipline or label an incident as a disciplinary infraction. The form is given to the Unit Manager and becomes part of the personnel file of the MHA.

The Board found there was no authority to discipline under § 2(11), only reportorial authority. Nurses can only report; others make the actual discipline decisions. This supportable finding is adequate to sustain the Board's conclusion.

Treating Provident's challenge on a clause by clause basis under § 2(11), we conclude that it has not met its burden, despite an able challenge. Because an atomized analysis may mislead, we step back to review the entire picture. Matching the Board's conclusions against the purposes for the exclusion of supervisors from the Act, we find no mismatch. Testing the Board's conclusions against the structure of the workplace, we again find no mismatch.

"To achieve a balanced picture, it is important to note what [the employees at issue] do not do." Telemundo de Puerto Rico, 113 F.3d at 272 (emphasis added). True supervisors are those vested with "genuine management prerogatives." Bell Aerospace Co., 416 U.S. at 280-81; see also 2 P. Hardin, The Developing Labor Law 1610 (3d ed. 1992) (observing that supervisor status derives from "[t]he power to act as an agent of the employer in relations with other employees"). The charge and district nurses are not vested with these prerogatives. For instance, they do not speak for the organization, set its policies, assert financial control, or resolve services disputes - all the sort of activities engaged in by those given true managerial and supervisory authority.

The claim that the district and charge nurses are supervisors rests on their relationship with the aides, who are a rung below them on the organizational ladder. These charge and district nurses are primarily care givers. The central responsibility of these nurses is to care for their patients, and their supervisory authority over the aides is relatively modest. That there are employees lower on the organizational ladder whom higher level employees evaluate and to whom the higher level employees give some direction surely cannot be enough to make the higher level employees supervisors. The NLRB has generally and correctly said that employees who routinely direct other employees based solely on a higher skill level are not supervisors.

The reasons that supervisors are excluded from coverage under the Act are not offended when employees such as the charge and district nurses are covered. The exclusion of "supervisors" from the Act is thought to be of benefit to both management and labor. It represents the principle "[t]hat an employer is entitled to the undivided loyalty of its representatives." NLRB v. Yeshiva Univ., 444 U.S. 672, 682 (1980). The exclusion may also help, "in part, to protect employees from supervisor influence within the union's organization." Metropolitan Life Ins. Co., 40 F.2d at 1178. These concerns are not implicated in this situation. There is no reason to believe that coverage of the district and charge nurses would lead to the inherent conflict of interest which results when a true supervisor must both represent the employer and bargain against it. Further, there is little reason to believe that involvement in collective bargaining will transform these professional nurses into categorically different employees, disloyal in some sense to their employer and unfit for the original purpose for which the employer hired them. Cf. Rabban, 89 Colum. L. Rev. at 1817 (discussing the possibility that a unionized university professor would become a "transformed employee," who would "sacrific[e] the professionalism for which he was hired to the collective security and adversarial posture of the union movement").

The structure of the organization also supports the Board's determination. This is a small nursing home, providing service at full capacity to 112 patients. The management of the home is primarily handled by four nurses: the Director of Nursing and three Unit Managers. Basic, non-professional services are provided by thirty aides. In between are thirteen district and charge nurses. It would be an odd managerial and supervisory structure to have the care of 112 patients entrusted to thirty non-professionals and have seventeen supervisors or managers, all of whom are nurses and professionals. See Children's Habilitation Ctr., Inc. v. NLRB, 887 F.2d 130, 132 (7th Cir. 1989) (using the ratio of supervisory to non-supervisory employees as a "guiding light[] in charge-nurse cases"); see also Beverly Enterprises--Minnesota, Inc., 148 F.3d at 1047. This case is not in the least like Maine Yankee, where this court reversed a determination of non-supervisory status for six employees who were in overall charge of a nuclear plant's nerve center, accountable for any plant failures and given the responsibility to independently direct the employees in the department. See Maine Yankee, 624 F.2d at 355-65. We do not discount the importance of the role played by district and charge nurses, who care for those in great need. But important roles are played by many people who are not supervisors.

For these reasons, we grant the Board's petition and enforce the Board's order.