Opinion for the Court filed by Circuit Judge GINSBURG.
GINSBURG, Circuit Judge:
Blue Man Vegas, LLC (BMV) petitions for review of the National Labor Relations Board’s decision that it engaged in unfair labor practices by refusing to bargain with the International Alliance of Theatrical Stage Employees, Moving Picture Technicians, Artists & Allied Crafts of the United States, Its Territories & Canada, AFL-CIO (the Union), elected to represent certain of its employees. BMV argues the Board erred in holding the bargaining unit proposed by the Union was appropriate. We deny Blue Man’s petition and grant the Board’s cross-application for enforcement.
I. Background
BMV manages and produces the Las Vegas production of the
Blue Man Grou/p,
a theatrical show in which men wearing blue grease paint on their faces and heads and dressed entirely in black perform a series of skits and dance routines involving music, props, and videos. On stage with the “Blue Men” are seven musicians. The Blue Men'and the musicians are assisted by a stage crew comprising seven departments: audio; carpentry; electrics; properties (props); video; wardrobe; and musical instrument technicians (MITs), who maintain the musical instruments, many of which are unique to
Blue Man Group
productions. There are also a handful of so-called “swings,” who BMV explains are “trained in numerous departments to provide coverage ... as needed due to vacation or illness.” During a performance, each of the seven stage crews performs its own “cue tracks,” which are series of carefully planned actions. For example, a carpentry crew’s cue tracks might involve placing and moving scenic backdrops at specified times.
From 2000 through most of 2005, BMV performed at the Luxor Hotel and Casino. During that time, BMV employed the MITs directly, but the Luxor employed the members of the other stage crews, as to whom it entered into a collective bargaining agreement with the Union. As a result, there were differences in the terms and conditions of employment of the MITs and of the other crews. The MITs reported to BMV’s Production Manager, John Mclnnis, whereas the other stage crews reported to the Luxor; the MITs were paid a salary whereas the others were paid an hourly wage; and the MITs’ pre-per-formance sign-in sheet was separate from the sign-in sheet for the others.
In September 2005, BMV left the Luxor and reopened a month later at the Venetian Hotel and Casino. Incident to the move, BMV decided to employ the entire stage crew directly. To handle its many new stage crew employees, BMV erected a new management structure. A department head would supervise the employees in each of the six departments that previously reported to the Luxor, and the “technical supervisor” would supervise the six new department heads and report to Mclnnis.
Although the employees in all seven stage crew departments were now employed directly by BMV, several differences between the MITs and the other crews were carried over from the Luxor to the Venetian. First, whereas the others
were separated from Melnnis, the production manager, by two levels of supervision (a department head and the technical supervisor), the MITs continued to report directly to Melnnis. Second, the two MITs who had been with BMV at the Luxor were still paid a salary, whereas the members of the other crews were paid a wage, as they had been at the Luxor. (The four MITs hired after BMV left the Luxor were paid a wage, however.) Finally, the MITs’ sign-in sheet remained separate from the sign-in sheet for the other crews.
In March 2006, the Union petitioned the Board for a representation election in a unit comprising all stage crew employees except the MITs. BMV objected that the MITs should be included in the bargaining unit. After a hearing, the Board’s Regional Director (RD) determined, pursuant to § 9(b) of the National Labor Relations Act, 29 U.S.C. § 159(b), that the unit proposed by the Union was an appropriate unit and ordered a representation election. The RD found significant the differences between the MITs and the other stage crews that stemmed from the prior unit’s bargaining history, namely, those relating to supervision, form of payment, and sign-in sheets. He also found significant a number of differences that cannot be attributed to BMV’s time at the Luxor: The MITs have separate substitutes during days off and vacations, “skills separate from the other stage crew members,” and different cue tracks; they “do not ‘swing’ to other stage crew positions”; and they “work in different areas” and “interact[ ]” primarily “with musicians, not stage crew members.” The Board denied BMV’s petition for review of the RD’s decision.
The Union won the ensuing representation election by a vote of 20-14 and the RD duly certified the Union as the exclusive bargaining representative. About a month later, the RD issued a complaint against BMV alleging it had refused to bargain with the Union, in violation of § 8(a)(1) and (5) of the NLRA, 29 U.S.C. § 158(a)(1)
&
(5). BMV argued it was not required to bargain because the exclusion of the MITs rendered the unit inappropriate. Finding BMV had raised or could have raised all issues relating to representation in the prior unit determination hearing and BMV did not proffer any previously unavailable evidence, the Board granted summary judgment for the General Counsel. BMV then petitioned for review in this court and the Board cross-applied for enforcement of its decision.
II. Analysis
BMV challenges the Board’s decision that its refusal to bargain was an unfair labor practice on the ground that the unit was not appropriate.
See Terrace Gardens Plaza v. NLRB,
91 F.3d 222, 225 (D.C.Cir.1996) (“Judicial review [of an order directing a representation election] is available only if the employer refuses to bargain and is found, in a final order of the Board, to have violated § 8(a)(5)” of the NLRA). “This court will uphold an NLRB bargaining unit determination unless it is arbitrary or not supported by substantial evidence in the record.”
Country Ford Trucks, Inc. v. NLRB,
229 F.3d 1184, 1189 (D.C.Cir.2000).
BMV advances three arguments: The Board applied the wrong standard to determine whether the proposed unit was appropriate; the unit determination was not supported by substantial evidence; and the exclusion of the MITs from the proposed unit created a “disfavored residual unit.” None is persuasive.
A. The Unit Determination Standard
BMV’s primary argument is that the Board applied a standard for the unit de
termination that conflicts with the NLRA and has been, for that reason, rejected by the Fourth Circuit. BMV’s position, although superficially plausible, is based upon a misapprehension of the framework governing unit determinations.
The Board’s principal concern in evaluating a proposed bargaining unit is whether the employees share a “community of interest.”
NLRB v. Action Auto., Inc.,
469 U.S. 490, 494, 105 S.Ct. 984, 83 L.Ed.2d 986 (1985);
see also Agri Processor Co., Inc. v. NLRB,
514 F.3d 1, 8-9 (D.C.Cir.2008). “There is no hard and fast definition or an inclusive or exclusive listing of the factors to consider [under the community-of-interest standard]. Rather, unit determinations must be made only after weighing all relevant factors on a case-by-case basis.”
Country Ford Trucks,
229 F.3d at 1190-91 (quotation marks, citations, and ellipsis omitted). Those factors include whether, in distinction from other employees, the employees in the proposed unit have “different methods of compensation, hours of work, benefits, supervision, training and skills; if their contact with other employees is infrequent; if their work functions are not integrated with those of other employees; and if they have historically been part of a distinct bargaining unit.”
Trident Seafoods, Inc. v. NLRB,
101 F.3d 111, 118 n. 11 (D.C.Cir.1996);
see also Agri Processor,
514 F.3d at 9 (collecting factors);
NLRB v. Lundy Packing Co. (Lundy II),
68 F.3d 1577, 1580 (4th Cir.1995) (listing factors). And, although the NLRA provides “the extent to which the employees have organized shall not be controlling,” 29 U.S.C. § 159(c)(5), the Supreme Court has held that the extent of their organization may be “considered] ... as one factor” in determining whether a proposed unit is appropriate.
NLRB v. Metro. Life Ins. Co.,
380 U.S. 438, 442, 85 S.Ct. 1061, 13 L.Ed.2d 951 (1965).
Decisions of the Board and of the courts in unit determination cases generally conform to a consistent analytic framework. If the employees in the proposed unit share a community of interest, then the unit is
prima facie
appropriate. In order successfully to challenge that unit, the employer must do more than show there is another appropriate unit because “more than one appropriate bargaining unit logically can be defined in any particular factual setting.”
Country Ford Trucks,
229 F.3d at 1189 (quotation marks omitted). Rather, as the Board emphasizes, the employer’s burden is to show the
prima facie
appropriate unit is “truly inappropriate.”
Id.
at 1189;
Dunbar Armored, Inc. v. NLRB,
186 F.3d 844, 847 (7th Cir.1999) (“clearly inappropriate”) (quotation marks omitted);
see also Serramonte Oldsmobile, Inc. v. NLRB,
86 F.3d 227, 236 (D.C.Cir.1996) (the Board “need only select
an
appropriate unit, not
the most
appropriate unit”) (quotation marks omitted).
A unit is truly inappropriate if, for example, there is no legitimate basis upon which to exclude certain employees from it. That the excluded employees share a community of interest with the included employees does not, however, mean there may be no legitimate basis upon which to exclude them; that follows apodictieally from the proposition that there may be more than one appropriate bargaining unit. If, however, the excluded employees share an overwhelming community of interest with the included employees, then there is no legitimate basis upon which to exclude them from the bargaining unit. We held in
Trident Seafoods,
for example, the Board’s unit determination was “irrational” and “unsupported by substantial evidence” because the employer
had adduced unrebutted evidence showing that “the functional integration of and the overwhelming similarities between the [excluded] and [included employees] are such that neither group can be said to have any separate community of interest justifying a separate bargaining unit.” 101 F.3d at 120;
see also Jewish Hosp. Ass’n,
223 N.L.R.B. 614, 617 (1976) (unit limited to service employees inappropriate because of “overwhelming community of interest” with maintenance employees);
Lodgian, Inc.,
332 N.L.R.B. 1246, 1255 (2000) (RD required inclusion in unit of employees who “share an overwhelming community of interest with the employees whom the [union] seeks to represent”).
A Venn diagram may clarify these principles. Each rectangle represents the interests of a group of identically situated employees. The region in which two or more rectangles overlap represents the degree to which those groups have common interests. In
Figure 1,
Rectangles A, B, and C all overlap because all the groups have a community of interest with each other. Consequently, any combination of the groups — AB, AC, BC, or ABC — is a
prima facie
appropriate bargaining unit. Note, however, that Rectangles A and B overlap almost completely; this indicates they have an overwhelming community of interest. Any unit that includes one but excludes the other is “truly inappropriate.” Therefore, the only units that could be deemed appropriate in the face of a challenge are AB and ABC.
[[Image here]]
BMV contends the Board applied the wrong standard in making its unit determination, effectively “according] controlling weight to the Union’s extent of organization,” in violation of § 9(c)(5) of the NLRA. According to BMV, the Board erred in basing its decision upon
Lundy Packing Co. (Lundy I),
314 N.L.R.B. 1042, 1043-44 (1994), in which the Board upheld the unit proposed by the union, thereby “failing] to heed” the Fourth Circuit’s subsequent refusal to enforce that deci
sion, which BMV says rested on the ground that the overwhehning-community-of-interest standard unlawfully gives controlling weight to the union’s extent of organization.
BMV’s reading of
Lundy II
and of the Board’s decision in this case reflect a misapprehension of the governing framework just described, as well as a misreading of the Fourth Circuit’s opinion. In effect, BMV contends that, as long as the MITs had a community of interest to any degree with the other stage crews, they could not be excluded from the bargaining unit. That view is obviously at odds with the principles discussed above.
Lundy II,
on the other hand, is consistent with the framework set out above. The Fourth Circuit there objected to the combination of the overwhelming-community-of-interest standard and the presumption the Board had employed in favor of the proposed unit: “By presuming the union-proposed unit proper unless there is ‘an overwhelming community of interest’ with excluded employees, the Board effectively accorded controlling weight to the extent of union organization.”
Lundy II,
68 F.3d at 1581. As long as the Board applies the overwhelming community-of-interest standard only after the proposed unit has been shown to be
prima facie
appropriate, the Board does not run afoul of the statutory injunction that the extent of the union’s organization not be given controlling weight.
Here, the Board correctly applied the overwhelming-community-of-interest standard; it did not presume the Union’s proposed unit was valid, as it had done in
Lundy I.
Rather, the RD first determined “[t]he record ... establishes that the petitioned-for unit, which excludes MITs, is an appropriate unit for collective bargaining”; indeed, he noted, “the parties have never contended” otherwise. The RD then went on to apply the overwhelming-community-of-interest standard to determine whether BMV had shown the exclusion of the MITs rendered the proposed unit truly inappropriate. As the Board says, the RD cited
Lundy I
to support the generally correct proposition that “a unit need not be an all-inclusive unit in order to be an appropriate unit,” and then looked to that decision for guidance as to the “factors” to be considered in deciding whether the two groups of employees have an overwhelming community of interest. The Board’s use of the overwhelming-community-of-interest standard, therefore, did not give controlling weight to the extent of the Union’s organization.
B. Substantial Evidence
BMV contends the Board’s finding that the proposed bargaining unit was appropriate was not supported by substantial evidence. As discussed above, the Board based its finding upon the many differences between the terms and conditions under which the MITs and the other stage crews worked. In attempting to refute the Board’s finding, BMV contends there are few if any relevant differences between the MITs’ terms and conditions of employment and those of the other crews. BMV also contends the Board’s finding conflicts with precedent. In response, the Board argues the differences between the MITs and the employees included in the bargaining unit were sufficiently substantial that the unit could “constitute a distinct and appropriate unit separate and apart from the MITs,” and that its decision was consistent with precedent. We agree with the Board.
BMV launches its challenge to the evidence upon which the Board relied by isolating the differences that “are holdovers from the Luxor,” namely, the different supervisory structure, separate sign-in sheets, and salary versus wage compensa
tion. BMV characterizes these differences as matters of “bargaining history,” and then ties the bargaining history to the “extent of organization,” thus: “The Regional Director reaehe[d] beyond the parties in this case and relie[d] on an IATSE contract with a completely different employer [i.e., the Luxor], This bargaining history is not relevant to this analysis except to demonstrate the Union’s extent of organization.”
We need not decide whether BMV correctly equates bargaining history with extent of organization in the circumstances of this case because this line of argument still would fail for two reasons. First, the differences between the MITs and the other stage crew employees that are “holdovers from the Luxor” are not merely of historical interest; they are present facts the Board could reasonably conclude differentiate the employment interests of the MITs from those of the other crews. As the Board rather forcefully puts it, “the ... suggestion ... that the Board should have ignored the terms and conditions of employment that [BMV] intentionally carried over from the Luxor is absurd.” Second, in light of the numerous differences that are not “holdovers from the Luxor,” the Board cannot be said to have given controlling weight to bargaining history nor, if it is the same thing on the present facts, to the Union’s extent of organization.
As for those differences that do not stem from the Luxor era, BMV maintains they do not distinguish the MITs from the employees in the other stage crews as a group, but rather distinguish the employees in each crew from the employees in every other crew. For example, BMV observes that, although the MITs have separate substitutes, so do the other stage crews because “[s]ubs do not work for more than one department.” BMV makes a similar point with respect to the MITs’ technical skills, cue tracks, use of swings, work space, and lack of interaction with other stage crew employees during the show. Thus, BMV argues, the Board acted arbitrarily by excluding the MITs from the unit on the basis of certain differences between the MITs and the other stage crews while at the same time ignoring the same types of differences among the various crews that were included in the unit.
We need not decide whether that would be an arbitrary or otherwise unlawful decision because that is not what the Board did. Rather, as discussed above, the Board recognized the MITs also differ from the employees in the other crews in ways that are “holdovers from the Luxor” and are therefore unique to the MITs, namely, in terms of supervision, form of payment, and sign-in sheets. The Board did not act arbitrarily by treating the MITs differently from the other stage crew employees in light of those differences.
Moreover, the Board’s finding that the proposed unit was appropriate without the MITs was certainly reasonable and supported by substantial evidence in view of the analytic framework set out above. A unit comprising all the non-MIT stage crews is
prima facie
appropriate because, notwithstanding the differences among them, those employees share a community of interest. It may well be that a unit comprising all the stage crews, including the MITs, would also be
prima facie
appropriate because the MITs also share a community of interest with the other stage crew employees, but that does not necessarily render the unit comprising only the non-MIT stage crews “truly inappropriate.” Indeed, both the differences that are unique to the MITs and the differences that can be found among all the stage crews stand in BMV’s way: The MITs lack an overwhelming community of interest
with the other stage crews (just as each of the non-MIT crews may lack an overwhelming community of interest with each of the other non-MIT crews).
To illustrate, in
Figure 2
Rectangle M represents the interests of the MITs, while Rectangles X and Y represent the interests of the employees in any two other departments. The shaded regions represent interests relating to subs, technical skills, cue tracks, swings, work space, and interaction with members of other stage crews during the show, that is, factors with respect to which each department has (we assume) different interests. The spotted regions represent interests relating to supervision, sign-in sheets, and form of payment, that is, factors carried over from the Luxor, which distinguish the MITs from the employees in all the other stage crew departments. The Board in effect found Unit XY appropriate. As the diagram shows, the Board was justified in doing so, though it could also have found Unit XYM appropriate because all three rectangles overlap, reflecting a community of interest among them, as represented by the crosshatched region. Unlike Rectangles A and B in
Figure 1,
however, Rectangle M does not have a nearly complete overlap with any other rectangle, reflecting the MITs’ lack of an overwhelming community of interest with any of the other stage crews. Consequently, the exclusion of Rectangle M from a unit comprising Rectangles X and Y — that is, the exclusion of the MITs from the unit comprising the other stage crew employees — does not render that unit “truly inappropriate,” notwithstanding the substantial differences among the stage crew employees, as represented by the shaded and spotted regions.
Turning from the facts to the law, BMV claims the Board’s finding that the MITs do not share an overwhelming community of interest with the other stage crews conflicts with the Fourth Circuit’s analysis in
Lundy II
and with the Board’s analysis in
Studio 54,
260 N.L.R.B. 1200 (1982). As BMV notes, “the Board cannot ignore its own relevant precedent but must explain why it is not controlling.”
LeMoyne-Owen College v. NLRB,
357 F.3d 55, 60 (D.C.Cir.2004) (quotation marks omitted). We find the Board’s decision consistent with both
Lundy II
and
Studio 54
because neither case involved differences as extensive as here.
In
Lundy II,
the excluded employees differed from the included employees “in a few respects: (1) the method for calculating their earnings; (2) supervision; and (3) a lack of interchangeability with” the included employees. 68 F.3d at 1580. Re
jecting the Board’s approval of the proposed unit, the court remarked, “The exclusion of ... employees based on such meager differences is, to say the least, problematic.”
Id.
at 1581. Here, according to BMV, “the MITs were excluded from the bargaining unit based on nearly the same ‘meager differences’ — different second line supervision, partly different pay structure, and separate sign-in sheets.” The Board responds that, “[i]n contrast
[to Lundy II],
here, the Board did not fragment a traditionally appropriate unit.” We think the Board’s decision here was consistent with
Lundy II
for a more basic reason: Even if those differences in supervision, pay structure, and sign-in sheet are too “meager” on their own to justify the exclusion of the MITs from the bargaining unit, they are only a fraction of the differences upon which the Board relied. The sum of those differences was sufficient to justify the Board’s decision that the MITs do not share an overwhelming community of interest with the other stage crew employees.
BMV’s comparison of this case to
Studio 5k
is similarly flawed. Studio 54 strove “to create an ambiance through music, lights, props, scenery, and ... the participation of many employees in an evening’s festivities.”
Studio 54,
260 N.L.R.B. at 1200. The union had proposed a bargaining unit of all employees except “stagehands,” including “disc jockeys, house board light operators, disco light board operators, flymen, and preset men.”
Id.
The employer raised no threshold question whether the proposed unit was
prima facie
appropriate; the issue it raised was whether the exclusion of the stagehands rendered the unit inappropriate. Despite a difference in supervision between the stagehands and the other employees, the Board concluded that, in light of the “interchange of job functions” between them, the stagehands did “not possess a community of interest so separate and distinct from [that of the included] employees as to warrant separate representation.”
Id.
From this decision BMV extracts the rule that “minor supervisory differences should not be determinative.”
Be that as it may, we agree with the Board that
Studio 5k
does not conflict with the Board’s decision here because of the panoply of other differences that separate the MITs from the other stage crew departments. Further, as the Board notes, the functional integration of Studio 54’s employees “far exceeded anything in BMV’s show.” For example, non-stagehands at Studio 54 “occasionally perform[ed] stagehand work,” “[a]t least two stagehands ... occasionally work[ed] on non-stage electrical equipment and perform[ed] general maintenance,” and “[n]on-stagehands and stagehands alike often mingle[d] and/or dance[d] with patrons!,] • • • helping] to create the festive atmosphere [Studio 54] desire[d].”
Studio 54,
260 N.L.R.B. at 1200. The only evidence of functional interchange BMV offers is that, when the company performs at a location other than the Venetian, “[t]he entire crew will work together to pack up the needed equipment and gear, load it, transport it, and set it up at the outside site ... with little differentiation between the segments of the stage crew.” But whether BMV performs at the Venetian or offsite, it appears that each stage crew department remains solely responsible for the technical tasks ordinarily within its domain; nothing suggests the MITs perform tasks ordinarily assigned to, say, the wardrobe crew. Therefore, though certainly relevant to this case,
Studio 5k
is not “so inconsistent with the [RD’s] decision so as to mandate reversal here.”
Int’l Union of Operating Eng’rs v. NLRB,
595 F.2d 844, 850 (D.C.Cir.1979);
see Overnite Transp. Co.,
325 N.L.R.B. 612, 612-13
(1998) (holding unit need not include mechanics in light of their separate work area and supervision, different uniforms, special skills, and lack of significant functional interchange).
In summary, we see no reason to disturb the Board’s finding that the proposed unit was not rendered “truly inappropriate” by the exclusion of the MITs. The Board was justified in considering the ways in which the terms and conditions under which the MITs work differed from those under which the other stage crews work, including the differences that stem from BMV’s time at the Luxor and therefore are unique to the MITs. The Board was also justified in considering the differences that do not stem from the Luxor era but distinguish each crew from every other crew. The Board reasonably concluded that whatever interests the MITs shared with the employees in the unit were not overwhelming in light of those numerous differences.
C. Residual Unit
Finally, BMV contends the Board’s decision is arbitrary and capricious because it creates an allegedly “disfavored residual unit.” According to BMV, a residual unit consists of excluded employees “sharing a community of interest with the [included] employees.” Thus, BMV argues, because the MITs “shar[e] an obvious community of interest” with the other stage crew departments, the Board improperly created a residual unit of MITs by excluding them from the unit.
BMV’s supposed rule against residual units is misconceived. It implies that all employees who share a community of interest must be included in the same unit, which proposition conflicts with the principle that more than one bargaining unit may be appropriate in any particular setting.
See, e.g., Country Ford Trucks, 229
F.3d at 1189-91 (holding that although “broader unit encompassing all parts and service department employees at both facilities” may have been appropriate, Board not
“required
” to include all such employees in unit in light of differences between facilities). In any event, the Board’s residual unit policy has no bearing upon this case because it relates only to whether a proposed residual unit is appropriate, not to whether a proposed initial unit is appropriate.
See Carl Buddig & Co.,
328 N.L.R.B. 929, 930 (1999).
III. Conclusion
In sum, we hold the Board applied the correct legal standard to determine whether the proposed bargaining unit was appropriate. The Board’s determination that the MITs may be excluded from the bargaining unit because they do not share an overwhelming community of interest with the stage crew employees included in the unit is supported by substantial evidence and does not conflict with precedent or the Board’s residual unit policy. We therefore deny BMV’s petition for review and grant the Board’s cross-application for enforcement.
So ordered.