Dissenting opinion filed by Circuit Judge MILLETT.

Heartland Plymouth Court MI, LLC (“Heartland”) successfully petitioned this Court to review an Order of the National Labor Relations Board (“the Board” or “NLRB”). The Order found Heartland violated its collective-bargaining agreement by failing to bargain over the effects of reducing employee hours. In granting the petition, we also denied the Board’s cross-application to enforce its Order. Neither outcome was a surprise. As we explained in our Judgment, and as this Court had explained over a decade earlier, we possess a “fundamental and long-running disagreement” with the Board over “whether an employer has violated section 8(a)(5) of the National Labor Relations Act [NLRA] when it refuses to bargain with its union over a subject allegedly contained in a colleetive[-]bargaining agreement.” See Enloe Med. Ctr. v. NLRB, 433 F.3d 834, 835 (D.C. Cir. 2005). Facts may be stubborn things, but the Board’s longstanding “nonacquiescence” towards the law of any circuit diverging from the Board’s preferred national labor policy takes obduracy to a new level. As this case shows, what the Board proffers as a sophisticated tool towards national uniformity can just as easily be an instrument of oppression, allowing the government to tell its citizens: “We don’t care what the law says, if you want to beat us, you will have to fight us.”

Emphasizing the real-world consequences of forcing parties to waste time and resources litigating, Heartland moves here for an award of attorney fees. In response, the Board provided a sweeping—and startling—defense of its nonac-quiescence policy. The Board said it would be justified in refusing to apply the law of any circuit. The Board’s logic makes no exception for the scenario in Heartland’s case, where the Board knew that it would end up in a circuit with adverse law. Nor does the Board reject nonacquiescence when any presentation would be a putsch—ie., when no circuit at all supports the Board’s legal position. See NLRB Atty Fee Resp. Br. at 13 & n.8. Because the Board’s actions go beyond whatever limited justification nonacquies-cence may have, we agree with Heartland that the Board is guilty of bad faith, grant Heartland’s motion for attorney fees, and award it $17,649.00.

I.

Factual Summary

Our Judgment already details the facts giving rise to the Board’s NLRA suit *19against Heartland, and we need not repeat them here. See Dkt. No. 1611466 (hereinafter “Judgment”). For purposes of nomenclature, however, it is worth noting the Board’s suit was predicated upon its view that the employer’s refusal to bargain on a matter allegedly within a collective-bargaining agreement requires a “clear and unmistakable” waiver. Our precedent, in contrast, consistently rejects that view; considering the contents of a collective-bargaining agreement is a question of “contract coverage.” This difference will manifest itself in the Board’s conduct before our Court, which informs Heartland’s motion for attorney fees.

Heartland first appealed the Board’s adverse Order to our Court in 2018. See Case No. 13-1227. Due to the Supreme Court’s pending decision in NLRB v. Noel Canning, — U.S. -, 134 S.Ct. 2550, 189 L.Ed.2d 538 (2014), Heartland’s appeal was-held in abeyance. When the Supreme Court found the recess appointments . of two Board members unconstitutional, the Board set aside its Order against Heartland, and moved to dismiss Heartland’s appeal. .We granted the Board’s motion; the Board reassigned Heartland’s case to a new panel—now properly comprised of Senate-confirmed Board members—and readopted its prior Order. See JA 533-34. Unsurprisingly, -Heartland appealed the Order here again. The Board, too, knew that this was Heartland’s second appeal to the D.C. Circuit. See NLRB Merits Br. Cert, as to Parties, Rulings, and Related Cases (“The ruling under review has previously been before the Court.”); NLRB Atty Fee Resp. Br. at 4 (“On January 29, 2015, a panel of the reconstituted Board issued a new Decision and Order incorporating its earlier decision by reference.”) (emphasis added).

Given our well-established “contract coverage” precedent, Heartland’s second appeal was pre-ordained.1 Accordingly, Heartland’s petition was granted, and the Board’s cross-petition to enforce its Order denied, in an unpublished Judgment without, oral argument. See Fed. R. App. *2034(a)(2); D.C. Cir. R. 84(j); D.C. Cir. R. 36(d). As we said, “[t]he Board’s refusal to adhere to our precedent dooms its decision before this court.” Judgment at>2i: While our Court previously recognized the Board’s right of nonacquiescence, see Enloe, 433 F.3d at 838, we did so with a certain end in mind. See Judgment at 2. Namely, we presumed the Board would recognize a stalemate with our case law, one resolvable by seeking certiorari to the Supreme Court. See Enloe, 433 F.3d at 838.

In .this case, the Board neither confessed the error of the Order against Heartland under our law, nor sought to preserve its argument against our precedent for certio-rari (or even en banc reconsideration). The Board did not seek a transfer to'the Sixth Circuit either. The- Sixth Circuit embraces the Board’s “clear and unmistakable” waiver-policy. See, e.g.] Beverly Health and Rehab. Servs., Inc. v. NLRB, 297 F.3d 468, 480 (6th Cir. 2002). Further, Michigan, covered by the Sixth Circuit, is where Heartland’s operations exist and where the conduct underlying the Board’s dispute occurred. See Judgment at 1-2. It is thus the only other jurisdiction in which the NLRA permits an appeal on these facts. See 29 U.S.C. § 160(f) (permitting petitions to review the Board’s decisions to be filed “in the circuit wherein the unfair labor practice in question was alleged to have been engaged in or wherein [any aggrieved] person resides or transacts business, or in the United States Court of Appeals for the District of Columbia”).2

In lieu of its legitimate options, the Board chose obstinacy. The Board cross-petitioned our Court to enforce its Order. In its responsive brief, the Board spent several pages asking us to uphold its “clear and unmistakable” waiver policy here. See NLRB Merits Br. at 17-20. Our adverse precedent made only a cameo appearance, where the Board spent a few sentences on an illusory distinction. See id. at 21-22 (stating Enloe does not apply “[b]ecause the effects of the change in hours are not matters that were covered by the parties’ agreement,” so, to the Board, “the contract coverage doctrine does not play a role”). The Board’s tactics forced Heartland to waste resources in replying. See Heartland Merits Reply Br. at 2-3, 8-10.

Given the Board’s behavior, it is little wonder that when Heartland moved for attorney fees, it sought fées under both the “not-substantially-justified” and “bad faith” provisions of the Equal Access to Justice Act. See 28 U.S.C. § 2412(b) (allowing “bad faith” attorney fee awards against the United States government); § 2412(d)(1)(A) (allowing attorney fee awards against the United States government “unless the court finds ... the position of the United States was substantially justified.... ”).3 Though Heartland also argues for attorney fees related to the Board’s conduct at the administrative level, our award applies only to the Board’s conduct before our Court.

Replying to' Heartland’s motion, the Board referenced its general policy of flouting any circuit’s NLRA interpretation *21with which the Board disagrees—a policy described colloquially as “nonacquies-cence.” The Board’s rationale for nonac-quiescence is two-fold: (1) the NLRA’s multi-venue provision, see 29 U.S.C. § 160(f), renders the Board clueless as to what circuit will govern the enforcement of its orders on appeal; and (2) the Board’s “uniform and nationwide” jurisdiction over labor policy gives it the right to disagree with any circuit, whenever it wants. See NLRB Atty Fee Resp. Br. at 13-14. The Board ignores the fact that these two rationales invoke different forms of nonac-quiescence. But, the breadth of the Board’s argument reveals the first reason is largely delusory. The second reason—a species of .nonacquiescence known as “in-tracircuit nonacquiescence”—provides the Board’s overarching rationale. The Board thinks its right to disagree extends beyond preferring one circuit’s position to another in a split, but also includes “staking] out its own position contrary” to any circuit. See id. at 13. The Board identifies no limit to its nonacquiescence. Neither the Board’s abusive tactics nor the extremism asserted in opposition to Heartland’s motion for attorney fees are justified.

II.

The Propriety of Nonacquiescence.,

We begin first with the goal of nonac-quiescence, as stated by the Board itself over sixty years ago: to “achieve[]” “a uniform and orderly administration of a national act, such as the [NLRA].” See Ins. Agents Int'l Union, 119 NLRB 768, 773 (1957). By “determining]” “whether to acquiesce in the contrary views of a circuit court of appeals or whether, with due deference to the court’s opinion, to adhere to its previous holding until the Supreme Court ... has ruled otherwise,” id. (emphasis added), the Board claims to ensure a nationally uniform labor policy. Understood in the most charitable light, not acquiescing to a given circuit’s diverging legal interpretation until the Supreme Court has the last word puts two roles in harmony—the Board’s role of national say in what labor law should be, and “the judicial department^]” “emphatic[ ]” “province and duty ... to say what the law is.” Marbury v. Madison, 5 U.S. (1 Cranch) 137, 177, 2 L.Ed. 60 (1803).

Our approval' of nonacquiescence presumed its stated virtue: opposing adverse circuit decisions permits the Board to bring national labor law questions to Supreme Court resolution. See, e.g., Enloe, 433 F.3d at 838 (“The Board is, of course, always free to seek certiorari.”); Yellow Taxi Co. v. NLRB, 721 F.2d 366, 385 (D.C. Cir. 1983) (Wright, J., concurring) (observing, in our circuit’s first embrace of nonac-quiescence, it would be “unwise” to oppose it, “particularly in light of the instances in which positions taken by the Board were first repeatedly rejected by a large number of circuits, then accepted by others, and later accepted by the Supreme Court”).. Indeed, when .our Court discussed different forms of agency nonacquiescence in Johnson v. United States Railroad Retirement Board., 969 F.2d 1082 (D.C. Cir. 1992), it predicated the method’s acceptability upon the agency redressing a circuit’s conflicting interpretation, not defying it ad infinitum. See id. at 1092 (“When an agency honestly believes a circuit court has misinterpreted the law, there are two places it can go to correct the error: Congress or the. Supreme Court”).

To that end, nonacquiescence allows for an issue’s “percolation” among the circuits; generating a circuit split that can improve the likelihood of certiorari being granted. See id. at 1093; see also id. at 1097 (Buckley, J., concurring in part and dissenting in part) (“Catching Congress’s ear ... is more easily said than done; and *22given the huge volume of petitions for certiorari that flood the Supreme Court, it is often [more] necessary to establish a split among the circuits before the Court will examine [the] issue.”); see also SuPREME Ct. R. 10(a) (Noting circuit splits as indicative of “the reasons the Court considers” to grant certiorari). But, nonaequies-eence is justifiable only as a means to judicial finality, not agency "aggrandizement. As we said in, Johnson, nonacquies-cence is divorced from its purpose when an agency asserts it with no stated intention of seeking certiorari.4 See 969 F.2d at 1092.

Achieving judicial finality through national uniformity requires nonacquies-cence to rest on certain conditions. First, as explained above, any nonacquiescence depends upon the agency actually seeking Supreme Court review of adverse decisions.5 Second, nonacquiescence requires candor in its application. See Estreicher & Revesz, Nonacquiescence, supra "n.4, at 756. The agency should clearly assert its nonacquieseence, specifying its arguments against adverse precedent to preserve them for Supreme Court review. These two conditions characterize proper nonac-quiescence.

In cases where the appeal implicates a statute’s multi-venue provision, the reviewing Court must assess a third condition: venue uncertainty. When an agency’s assertion of .venue uncertainty is plausible on the facts and proper nonacquiescence is otherwise pursued, the agency acts in good faith. But, when an agency’s assertion of venue uncertainty is implausible on the facts, the situation is no different than intracircuit nonacquiescence—where the agency’s conduct would constitute bad faith if its nonacquiescence is- not clearly asserted and accompanied by a preservation of arguments for Supreme Court or en banc review. Cf. Johnson, 969 F.2d at 1091-92 (rejecting the agency’s assertion of nonacquiescence when “[t]here [was], of course, some venue uncertainty under the ... statute.... But the Board has never attempted to invoke venue uncertainty to justify its actions, and it seems to be asserting a right of nonacquiescence in its most sweeping sense.”). Given the facts here, this third condition requires some elaboration.

Intracircuit nonacquiescence is not the same as refusing to apply an adverse circuit’s law due to the underlying statute’s multi-venue provision. For example, when a party appeals an adverse NLRB order under the NLRA, the statute provides the appealing party with multiple venue options. See 29 U.S.C. § 160(f). This uncer*23tainty means, in some circumstances, the Board may have issued its order “without knowing which circuit court ultimately will review its actions.” Johnson, 969 F.2d at 1091. In those circumstances, the Board’s nonacquiescence to an adverse circuit’s law is a function of ignorance, not defiance.

There are, however, multiple instances when an agency’s assertion of venue uncertainty is implausible, ie., it knows that its order will be subjected to an adverse circuit’s law on appeal. Estreicher & Revesz point out two examples: (1) when “all courts of proper venue have adopted positions contrary to the agency’s policy”; and (2) when an order has been issued by an agency on remand from an adverse circuit court which retained jurisdiction over the action. See Estreicher & Revesz, Nonac-qwiescence, supra n.4, at 687 & n.34. In these cases, any nonacquiescence is necessarily intentional and, thus, of the intracir-cuit variety. These are just “example[s],” however, see id. at 687, and there are others. When a case’s facts result in only two venue choices for the party appealing the adverse order, and one circuit’s precedent is in agreement with the agency’s legal interpretation while the other is adverse to it, the agency knows any appeal will be to the adverse circuit. See Ithaca Coll. v. NLRB, 623 F.2d 224, 227 (2d Cir. 1980) (“Certainly the College was not going to seek review in the D.C. Circuit when it had a favorable precedent in the Second Circuit.”). Furthermore, “for [NLRB] purposes, which circuit’s law should apply is readily ascertainable” when it cross-petitions to enforce its order before an adverse court, instead of invoking its transfer rights to enforce the order in a favorable venue. Cf. Donald L. Dotson & Charles M. Williamson, NLRB v. The Courts: The Need for an Acquiescence Policy at the NLRB, 22 Wake FOREST L. Rev. 739, 739 n.1 (1987) (noting the “Board’s [historic] policy [was] to seek enforcement of its orders in the circuit in which the unfair labor practice arose. Therefore, for Board purposes, which circuit’s law ... is readily ascertainable”). Under any of these scenarios, the multi-venue provision provides no plausible stumbling block to the agency knowing where it will have to defend its order.

In any event, venue uncertainty cannot license improper nonacquiescence. Nothing about venue uncertainty excuses: (1) a less-than-candid representation of the agency’s disagreement with adverse circuit law; (2) the failure to indicate the preservation of opposing arguments for Supreme Court review; or (3) the failure to seek certiorari of adverse decisions to achieve a national resolution. Letting the mere possibility of venue uncertainty excuse those conditions not only makes nonacquiescence unbounded—it also would be a failure. Distinguishing, case-by-case, plausible venue uncertainty from intracircuit nonacquies-cence is critical to avoid “nonacquiescence in its most sweeping sense,” ie., a form divorced from the end of judicial finality and the requirement of candor. See Johnson, 969 F.2d at 1091-92; see also NLRB v. Ashkenazy Prop. Mgmt. Corp., 817 F.2d 74, 76 (9th Cir. 1987) (“Any future act of ‘nonacquiescence’ should be dealt with by this court in the specific context in which it occurs so that we may address the agency’s particular violation of the rule of law and fashion a remedy that is appropriate in light of all of the relevant circumstances.”).

Unfortunately, the NLRB’s history with nonacquiescence reveals “its primary goal is ... to see its interpretation of the federal labor laws prevail in as many cases as possible, rather than to change contrary law in particular circuits or ... serve as a percolator for the Supreme Court.” See Ross E. Davies, Remedial Nonacquiescence, 89 Iowa L. Rev, 65, 100 (2003); cf. *24NLRB v. Gibson Prods. Co., 494 F.2d 762, 766 (6th Cir. 1974) (“It is apparent from the foregoing chronology of this case that the Board, disagreeing with [the Supreme Court’s] requirement of contemporary necessity for a bargaining order in second category cases, has simply sought to avoid it_”). Indeed, in the only instances we can find where the NLRB ever addressed the “contract coverage”—“clear and unmistakable” circuit split before the Supreme Court, the Board was opposing cer-tiorari. None of the reasons the Board set forth in these briefs would prohibit seeking certiorari in an appropriate case.6 Moreover, we are unmoved by the coincidence of the Board opposing certiorari in cases where certiorari was denied. See Davies, Remedial Nonacquiescence, 89 Iowa L. Rev. at 78 & n.43 (citing a 1997 letter from the acting NLRB solicitor to the clerk of the Fourth Circuit, which described the Board’s “enviable record in the Supreme Court” as “persuasive evidence that the Board has exercised good judgment in deciding when it is appropriate to continue to insist that intermediate courts have overstepped their authority” in disagreeing with the Board). After all, there is a difference between theory and practice. See id. at n.45 (noting that, as of the article’s 2003 publication, “[t]he Board has not been the prevailing party on the merits in a case before the Supreme Court since 1996.”). It is difficult to see the Board’s steadfast refusal to seek certiorari on the “contract coverage” question as something other than an evasion of finality in the name of hegemony.

As a former NLRB Chairman and Chief Counsel, respectively, explained;

In fact, rather than promoting uniformity, the Board’s policy of nonacquiescence has fostered a bifurcated system in which litigants willing to pursue their case to the appellate level are able to avoid [the] Board[’s] orders. Thus, the Board’s policy has had the effect of needlessly protracting litigation, establishing a two-tiered system of labor law in the same jurisdiction, encouraging disrespect for [the] Board[’s] orders, and antagonizing the courts ... Even worse, it compels litigants to expend resources in litigating cases in which it is clear that the appropriate circuit will not enforce the Board’s order.

Dotson & Williamson, NLRB v. The Courts, 22 Wake Forest L. Rev. at 745 (emphasis added). Our Court shares these concerns. We noted in Johnson that nonac-quiescence allows agencies to work their will on not only the courts, but on the American people too. See 969 F.2d at 1092 (“The Board, in the end, can hardly defend *25its policy of selective nonacquiescence by invoking national uniformity. The policy has precisely the opposite effect, since it results in very different treatment for those who seek and who do not seek judicial review.”).

For these reasons, and others, our sister circuits have spilled much ink admonishing the NLRB’s nonacquiescence. See id. at 1091 (“Intracircuit nonacquiescence has been condemned by almost every circuit court of appeals that has confronted it.”); Dotson & Williamson, NLRB v. The Courts, 22 Wake Forest L. Rev. at 739-40 n.3 (finding instances of circuit courts rejecting the Board’s nonacquiescence dating back as early as 1953). We also think “the Board should reconsider its single-minded pursuit of its policy goals without regard for the supervisory role of the Third Branch.” See, e.g., Glenmark Assocs. Inc. v. NLRB, 147 F.3d 333, 339 n.8 (4th Cir. 1998).

In Yelloio Taxi, we warned the NLRB that sweeping nonacquiescence “may ... require[ ] [us]' to secure adherence to the rule of law by measures more direct than refusing to enforce its orders.” 721 F.2d at 383. At least one other circuit has already awarded attorney fees against the NLRB for relitigating, via nonacquiescence, an issue the Court already decided. See Enerhaul, Inc. v. NLRB, 710 F.2d 748, 751 (11th Cir. 1983). More than a decade ago, we told the NLRB that our positions on the “contract coverage” analysis were “stalemated” absent the Board seeking certiorari. See Enloe, 433 F.3d at 838. Not only has the Board refused to do so over the ensuing decade, its theory in support of nonacquiescence grows even more sweeping. In short, as we said of the Rail Road Retirement Board in Johnson-, “After ten years of percolation, it is time for the Board to smell the coffee.” 969 F.2d at 1093.-

III.

The. Board’s Nonacquiesoence Against Heartland Amounts To Bad Faith

The legal dispute in Heartland’s case demonstrates persistent nonacquies-cence without either candor or the pursuit of judicial finality. As we mentioned, our “contract coverage” case law has diverged from the Board’s “clear and unmistakable” waiver policy for almost a quarter century. Now, seven of the twelve geographic circuits take a side in That debate.7 With a split engulfing most circuits, there is no serious' argument for nonacquiescence in the name of percolation. Cf. Johnson, 969 F.2d at 1093 (“But now that three circuits have rejected the Board’s position, and not one has accepted it, further resistance would show contempt for the rule of law.”); id. at 1097 .(Buckley, J., concurring in part and dissenting in part) (“[G]iven the huge volume of petitions for certiorari that flood the Supreme Court, it is often necessary to establish a split among the circuits before the Court will examine an issue”) (emphasis added). And yet here, the Board gave us no indication at' all that it intends to seek certiorari of any adverse ruling, or en banc reconsideration of our precedent. Indeed, the Board did not even invoke nonac-*26quiescence by name until it replied to Heartland’s motion for attorney fees.

Worse still, the Board’s lack of candor is evident in its handling of our “contract coverage” precedent. Rather than confess the error of its Order against Heartland under our law, the Board’s merits brief, in relevant part, urges us to embrace the “reasonableness” of its “clear and unmistakable” waiver analysis. See NLRB Merits Br. at 17-20. Then, as a brief aside, it pretends there is no conflict between its Order and our law. See id. at 21 (“[Because the effects in the change in hours are not matters that were covered by the parties’ agreement, the contract coverage doctrine does not play a role”). The Board’s reasoning is nonsensical because, if a subject is not covered by a contract, then the contract certainly does not clearly and unmistakably waive bargaining over that matter. “[D]isguis[ing] its disagreement by means of a disingenuous distinction of adverse circuit precedent” is yet another indication of improper nonacquies-cence. See Estreicher & Revesz, Nonac-quiescence, supra n.4, at 755.

On these facts, nothing about the NLRA’s multi-venue provision sanitizes the Board’s eleventh-hour nonacquiescence plea. The Board knew ruling against Heartland would prompt an appeal to our circuit. Why? It already did. Recall that Heartland previously appealed the same ruling to our Court before the case was held in abeyance due to Noel Canning. See NLRB Merits Br. Cert, as to Parties, Rulings, and Related Cases (“The ruling under review has previously been before the Court.”). When the Board readopted its prior Order against Heartland—with the only material difference being that the Board panel was now comprised of Senate-confirmed members—it had every reason to think Heartland would appeal here again. For another matter, Heartland’s appellate options were twofold: (1) our circuit, to which it previously appealed the same substantive Order and which has favorable law; or (2) the Sixth Circuit, which embraces the Board’s “clear and unmistakable” waiver policy. There is no reason to think Heartland would seek appellate review in a circuit where it would almost certainly lose. See Ithaca Coll., 623 F.2d at 227 (“Certainly the College was not going to seek review in the D.C. Circuit when it had a favorable precedent in the Second Circuit.”). On these facts, it requires a willful suspension of disbelief to think: (1) Heartland would not appeal again; and (2) would not appeal again here.

If the Board did not want to sacrifice its Order against Heartland or defend nonacquiescence before us, it still had a viable option: transfer the case to the Sixth Circuit. As we noted above, the facts favored a transfer, and the Board’s Order would have almost assuredly been enforced in that jurisdiction. The Sixth Circuit accepts the Board’s “clear and unmistakable” waiver position; the NLRA allows the Sixth Circuit jurisdiction over Heartland’s appeal; Heartland’s operations are within the Sixth Circuit; and the underlying conduct took place within the Sixth Circuit.8 Instead, the Board cross-peti*27tioned for enforcement here. This was punitive. The Board chose to put its Order on a suicide mission with our precedent simply to lock horns with Heartland. The Board was the perpetrator here, not venue uncertainty.9

There is one other indication that venue uncertainty is not the real reason behind the Board’s behavior. The Board’s response to Heartland’s attorney fee motion offers an extreme and unbounded view of nonacquiescence. This position, combined with the Board’s conduct on the merits, embraces the following nonacquiescence standard: the Board can employ nonac-quiescence: (1) without ever saying so to the Court until after judgment is entered; (2) without ever seeking certiorari to resolve the disputed issue; (3) even when it knows what law will apply in advance of the appeal; and (4) even when every circuit in the country disagrees with it. See NLRB Atty Fee Resp. Br. at 13-14. In sum, the Board’s candor-free approach to nonacquiescence asks this Court to let the Board do what no private litigant ever could: make legal contentions not warranted by existing law and supported by no argument for modifying, reversing, or establishing new law. This is intolerable. See, e.g., Fed. R. Civ. P. 11(b)(2). We are under no obligation to bless the desire of “federal agencies [to] be subject to no law at all— as, indeed, it appears [the NLRB] believe[s] to be the case.” See U.S. Dep’t of Energy v. FLRA, 106 F.3d 1158, 1164-67 (4th Cir. 1997) (Luttig, J., concurring). Had Heartland’s case been one where the Board carefully applied nonacquiescence towards national uniformity, it would have proceeded differently. Where, as here, the Board “assert[s] a right of nonacquies-cence in its most sweeping sense,” and where its “sincerity” towards national uniformity is doubtful on the case’s facts, the theoretical possibility of “some venue uncertainty” is rendered an implausible justification. See Johnson, 969 F.2d at 1091-92.

Taken together, the Board’s conduct before our Court makes out a clear case of bad faith litigation. The standard for an award of attorney fees for bad faith is met “where the party receiving thé award has been the victim of unwarranted, oppressive, or vexatious conduct on the *28part of his opponent and has been forced to sue to enforce a plain legal right.” Am. Hosp. Ass’n v. Sullivan, 938 F.2d 216, 222 (D.C. Cir. 1991). Contrary to the out-of-circuit-cases the Board cites, “[t]his principle is no less applicable” to conduct occurring within litigation itself. Id. To be sure, “[b]ad faith by a litigant is serious business, and the standard for finding it is, appropriately, ‘stringent.’ ” Id. at 223 (D.H. Ginsburg, J., dissenting). But the Board’s conduct before us manifests a stubborn refusal to recognize any law.

The Board’s' obstinacy forced Heartland to waste time and resources fighting for a freedom the Board knew our precedent would provide. The Board did nothing to employ permissible nonacquiescence; it just saved the concept as a post-hoc rationalization in case Heartland had the temerity to ask us not to make it pay for the Board’s hubris. And worse, when it did finally mention nonacquiescence in response to Heartland’s attorney fee motion, the Board proposed an exasperatingly expansive rationale.

It is clear enough that the Board’s conduct was intended to send a chilling message to Heartland, as well as others caught in the Board’s crosshairs: “Even if we think you will win, we will still make you pay.” This roguish form of nonacquies-cence assures the Board’s gambit is virtually cost-free—the Board either enjoys the fruits of a settlement, or it dares a party to employ “the money and power [needed] to pay-for and survive the process of fighting with an agency through its administrative processes and into the federal courts of appeals.” Davies, Remedial Nonacquies-cence, 89 Iowa L. Rev. at 79. With seeking certiorari or en banc reconsideration in its hands, the Board can decide it is worth losing a few battles to still win the war. The Board can thus continue its adherence to the “clear and unmistakable” waiver policy without the Supreme Court ever telling it to stop, even with the occasional defeat in an adverse circuit. This bald attempt at a litigation advantage is bad faith. See Sullivan, 938.F.2d at 222; cf. id. at 223-24 (D.H. Ginsburg, J., dissenting) (arguing against a finding- of bad faith because-, unlike here, “I am aware of no reason for believing that the Secretary thought or could reasonably have thought he would gain any advantage” from perpetuating confusion about the law and “chilling” private parties “in the assertion of their rights”).

A few words in response to our dissenting colleague. The dissent acknowledges the propriety of awarding Heartland fees based on the Board’s “failure to candidly acknowledge binding circuit precedent in its answering brief and for pressing only a gossamer-thin argument for distinguishing Enloe.” Dissent Op. 33. We also agree that “an agency’s persistent defiance of uniform and settled circuit precedent could ignite a separation-of-powers firestorm.” See id. at 29. The Board should take note of these conclusions.

We are at a loss to understand, however, how either of these’ conclusions is consistent with the rest of the dissent. If the Board’s reply brief merits a fee award, was it not “thumbing its nose at settled decisional law?” But see id. at 29. If “Heartland had to file a petition for judicial review in this circuit,” id. at 31, where else could the Board expect to be? But see id. As the Board cross-petitioned to enforce its own Order here—asking us to bless its’ “clear and unmistakable” waiver policy in the process—did it not do more than simply “litigat[e] [Heartland’s] appeal?” But see id. at 30. Is the Board’s refusal to seek certiorari on the “contract coverage” issue, even after it has percolated among the circuits, something other than “persistent defiance” of judicial finali*29ty? But see id. at 29. The Board’s- entire litigation conduct before us consisted of: (1) a reply brief that every member of this Panel finds susceptible to the bad faith label; (2) a cross-petition the Board knew our precedent would not permit, but would force Heartland to respond; and (3) labeling all of this “nonacquiescenee” only after the fact, and with the most sweeping logic. The bad faith speaks for itself.

Granting Heartland’s motion for attorney fees “serve[s] the dual purpose of vindicating judicial authority ... and making the prevailing party whole for expenses caused by his opponent’s obstinacy.” See Chambers v. NASCO, Inc., 501 U.S. 32, 46, 111 S.Ct. 2123, 115 L.Ed.2d 27 (1991). We recognize the Board’s unimpeded access to the public fisc means these modest fees can be dismissed as chump change. But money does not explain the Board’s bad faith; “the pleasure' of being above the rest” does. See C.S. Lewis, Mere' Christianity 122 (Harper Collins 2001). Let -the word go forth: for however much the judiciary has emboldened the administrative state, we “say what the law is.” Marbury, 5 U.S. (1 Craneh) at 177. In other words, administrative hubris does not get the last word under our Constitution. And citizens can count on it.

IV.

For the foregoing reasons, .we grant Heartland’s motion for attorney fees and award it $17,649.00 for the Board’s bad faith litigation.

So ordered.

dissenting:

I certainly understand my colleagues’ concern that an agency’s persistent defiance of uniform-and settled circuit precedent could ignite a separatión-of-powers firestorm. But this case is nothing like that, and I strongly disagree that a bad-faith award of all the fees that Heartland incurred in this appeal is warranted.

Awarding fees for bad faith is an exceptional sanction that should only be employed “when extraordinary circumstances or dominating reasons of fairness so demand.” Nepera Chem., Inc. v. Sea-Land Serv., Inc., 794 F.2d 688, 702 (D.C. Cir. 1986). The standards for bad faith “are necessarily stringent,” Lipsig v. National Student Mktg. Corp., 663 F.2d 178, 180 (D.C. Cir. 1980) (quotation marks and citation omitted), requiring a factual finding that “the losing party has acted in bad faith, vexatiously, wantonly, or for oppressive reasons.” Alyeska Pipeline Serv. Co. v. Wilderness Soc’y, 421 U.S. 240, 258, 95 S.Ct. 1612, 44 L.Ed.2d 141 (1975) (internal quotation marks, omitted). Moreover, “[b]e-cause inherent powers” like an attorneys’ fees sanction for bad faith “are shielded from direct democratic concerns, they must be exercised with restraint and discretion.” Roadway Exp., Inc. v. Piper, 447 U.S. 752, 764, 100 S.Ct. 2455, 65 L.Ed.2d 488 (1980). That especially demanding standard is not met in this case, for four reasons.

■First, for all of the majority opinion’s concerns about an agency thumbing its nose at settled decisional law, this case involves an issue on which there is an inter-circuit conflict and on which the Board’s position accords with the majority view. Compare Local Union 36, Int'l Bhd. of Elec. Workers, AFL-CIO v. NLRB, 706 F.3d 73, 81-82 (2d Cir. 2013) (adopting the Board’s “clear and unmistakable waiver” test); Local Joint Exec. Bd. of Las Vegas v. NLRB, 540 F.3d 1072, 1079-1080 & n.11 (9th Cir. 2008) (same); Beverly Health & Rehab. Servs., Inc. v. NLRB, 297 F.3d 468, 481-482 (6th Cir. 2002) (same); Capitol Steel & Iron Co. v. NLRB, 89 F.3d 692, 697 (10th Cir. 1996) (same), with Bath *30Marine Draftsmen’s Ass’n v. NLRB, 475 F.3d 14, 25 (1st Cir. 2007) (adopting contract-coverage rule); NLRB v. United States Postal Serv., 8 F.3d 832, 836 (D.C. Cir. 1993) (same); Chicago Tribune Co. v. NLRB, 974 F.2d 933, 937 (7th Cir. 1992) (same). See also Mississippi Power Co. v. NLRB, 284 F.3d 605, 612-613 (5th Cir. 2002) (describing the competing standards).

So there has been no “putsch” here (Majority Op. 18). This case, by its terms, does not implicate at all the majority opinion’s concerns about a Board refusal to acquiesce in the face of .uniformly adverse circuit precedent. To be sure; the Board discussed a potentially sweeping realm for ■nonacquiescence in its brief. See NLRB Opp’n to Mot. for Att’y Fees at 13. But the bad faith for which we can authorize fees must have occurred in the Board’s actual conduct of its appellate litigation in the case at hand, not in a later overstatement in its opposition to attorneys’ fees concerning hypothetical facts not before us.

Second, the last time the Board was before this court 'on this very same issue, this court unanimously assured the Board that it had “every right” to “refuse[] to acquiesce in our analysis” of when and under what circumstances the terms of a collective bargaining agreement may discharge an employer’s collective-bargaining duties. Enloe Med. Ctr. v. NLRB, 433 F.3d 834, 838 (D.C. Cir. 2005). See generally, e.g., Independent Petroleum Ass’n v. Babbitt, 92 F.3d 1248, 1261 (D.C. Cir. 1996) (“[Ijntercircuit nonaequiescence is permissible, especially when the law is unsettled.”); American Tel. & Tel. Co. v. FCC, 978 F.2d 727, 737 (D.C. Cir. 1992) (acknowledging the agency’s'- “right to refuse to acquiesce in one (or more) court of appeals’ interpretation of its statute”); Johnson v. United States R.R. Ret. Bd., 969 F.2d 1082, 1093 (D.C. Cir. 1992) (noting the general right of an agency to engage in inter-circuit nonacquiescence, at least where its position has not been rejected. by' every circuit to address the question). The Board should not be labeled a “bad faith” actor for taking this court at its word and litigating the appeal at all, which is what the comprehensive award of attorneys’ fees for the entire appeal does.

In particular, I see nothing remotely approaching bad faith in requiring Heartland to file its petition for review and to prosecute its appeal by filing either an opening brief or, easier still, a motion for summary reversal, see D.C. Cir. Handbook of Practice and Internal Procedures VII.G.1 That is because Heartland is located within the jurisdiction of .the Sixth Circuit, and the law of that circuit is on all fours with the Board’s “clear and unmistakable waiver” rule. See, e.g., Beverly Health, 297 F.3d at 480 (“A management-rights clause is a waiver of the union’s right to bargain over [mandatory subjects].”); id. (“A union can waive its statutory right to bargain [in a collective bargaining agreement], but such a waiver must be ‘clear and unmistakable.’ ”) (quoting Metropolitan Edison Co. v. NLRB, 460 U.S. 693, 708 (1983)); Uforma/Shelby Bus. Forms, Inc. v. NLRB, 111 F.3d 1284, 1290 (6th Cir. 1997) (similar).

Accordingly, as the majority opinion acknowledges (at 19-20, 26-27), there was nothing remotely bad faith about the Board’s application and enforcement of its “clear and unmistakable waiver” rule in the agency proceedings. And given the *31Board’s decision, Heartland was destined to lose unless and until it sought judicial review- in this circuit rather than the Sixth Circuit. Had the Board filed first in the Sixth Circuit, Heartland’s petition for review would have been doomed. In short, having lost before the Board in a proceeding that quite properly applied the “clear and unmistakable waiver” rule, Heartland had to file a petition for judicial review in this circuit and had to affirmatively prosecute its appeal by filing an opening brief -or motion for summary disposition raising the contract-coverage issue to have a legal leg to stand on. I do not understand how it could be bad faith for the Board to require that Heartland do so.

The majority opinion says (at 26) that the Board should have known the case was destined for this circuit after remand, and thus apparently should have given up before Heartland even filed its petition. But as the circuit conflict attests, plenty of losing litigants before the Board have chosen to litigate in their home jurisdictions long after this court first adopted the “contract coverage” rule in 1993, see United States Postal Service, supra, and even after our reaffirmation of that rule in Enloe in 2005, see Bath Marine, supra, Local Union 86, supra, and Local Joint Exec. Bd., supra. Moreover, this court did not retain jurisdiction after granting the Board’s motion to dismiss the case in the wake of NLRB v. Noel Canning, 134 S. Ct. 2550 (2014). See Heartland Plymouth Court MI, LLC v. NLRB, No. 13-1227, 2014 WL 4627817 (D.C. Cir. Aug. 26, 2014). There thus was no guarantee that the second round of review would land here just because the first one did. Compare Starbucks Corp. v. NLRB, No. 09-1273 (D.C. Cir. Aug. 19, 2010) (dismissing petition for review on Board motion to reconsider in light of New Process Steel v. NLRB, 560 U.S. 674, 130 S.Ct. 2635, 177 L.Ed.2d 162 (2010)), with NLRB v. Starbucks Corp., 679 F.3d 70 (2d Cir. 2012) (second petition for review filed in and adjudicated by the Second Circuit).-

To be sure, the Board could have beaten Heartland to the punch by petitioning the Sixth Circuit for enforcement or moving to transfer the case to the Sixth Circuit. But the Board’s failure to deprive an employer of its chosen forum for review or to forgo imposing on the employer the additional costs of litigating a transfer motion cannot by itself meet the “stringent” requirement for bad faith, Nepera Chem., 794 F.2d at 702.

Third, the majority opinion (at 25-26) decries the Board’s failure to have sought certiorari to resolve the circuit conflict in an earlier case. But, again, the question is whether the Board litigated this appeal in bad faith, not whether it should have taken an additional procedural step in some other case. Sanctioning the Board for failing to seek certiorari is 'doubly inappropriate because the questions of whether and when-Supreme Court review should be sought to eliminate the conflict and establish a single, uniform federal rule rest exclusively with the Solicitor General in the Department of Justice and not with the Board. 28 U.S.C. § 518(a); see also 28 C.F.R. § 0.20 (Solicitor General is assigned duty of “[cjonducting, or assigning and supervising, all Supreme Court cases, including * * * petitions for and in opposition to certiorari”). Surely we cannot sanction as “bad faith” the Board’s failure to make a decision Congress has said it cannot make.

It also bears noting that cases in which the Board ends up at loggerheads with this court’s contract-coverage rule do not appear to arise with significant frequency. Since we first adopted the contract-coverage rule for Board cases in 1993 in United States Postal Serv., only Enloe and this *32case have arisen in which the Board found itself directly at odds with circuit precedent. That is only two cases in 23 years. The Board, moreover, has won more than it has lost in circuit court decisions generally, and in this circuit has argued in other cases that its order can be sustained under either standard. See BP Amoco Corp. v. NLRB, 217 F.3d 869, 873 (D.C. Cir. 2000) (“Here, the Board acknowledges the force of the ‘covered by’ principle but contends it does not apply because the Board’s decision expressly found that the collective bargaining agreement did not incorporate the reservation of rights clauses.”). The frequency with which a conflict is joined and whether a Supreme Court decision in the particular case would have any practical effect on the outcome of the case— whether the dispute over the standard of review is outcome determinative—are among the traditional factors that the Solicitor General. could reasonably consider in selecting .the issues it chooses to present to the Supreme Court each year for certio-rari review. See Johnsonf 969 F.2d at 1097 (Buckley, J., concurring in part and dissenting in part) (discussing legitimate governmental considerations that may result in agency non-acquiescence in conflicting circuit decisions enduring for some time); see generally■ Margaret Meriweather Cor-dray & Richard Cordray, The Solicitor General’s Changing Role in Supreme Court Litigation, 61 B.C. L. Rev. 1323, 1328-1330 (2010) (discussing certiorari factors considered by Solicitors General).

Fourth, the award of fees for bad faith is an equitable exercise of the court’s inherent power to control litigation before it. See, e.g., Copeland v. Martinez, 603 F.2d 981, 984 (D.C. Cir. 1979) (award of fees seiwes. to “protect[ ] the integrity of the judicial process”). And in this case, Heartland bears responsibility for a not insignificant amount of the fees it incurred.

To begin with, given the clarity of our precedent, Heartland could have short-circuited this litigation by moving for summary reversal. To be sure, a party seeking summary disposition bears “the heavy burden of establishing that the merits of his case are so clear that expedited action is justified.” Taxpayers Watchdog, Inc. v. Stanley, 819 F.2d 294, 297 (D.C. Cir. 1987) (per curiam). But for many of the reasons the majority opinion discusses (at 19-20 & n.l), the law in this circuit was just that clear and plainly adverse to the Board’s position, making this a signature case for such summary disposition.

Contrary to the majority opinion’s suggestion (at 27 n.9), an opposition by the Board preserving its arguments for review en banc or by the Supreme Court would not have altered the straightforward task of panel disposition since the law of the circuit would have controlled. See, e.g., La-Shawn A. v. Barry, 87 F.3d 1389, 1393 (D.C. Cir, 1996) (en banc) (“[T]he same issue presented in' a later case in the’same court should lead to the same result.”) (emphasis in original).

Heartland chose instead to initiate the ordinary briefing process and to then file a full-throated opening brief that raised additional issues for our review beyond the contract-coverage dispute. Heartland’s failure to reasonably mitigate the fees it incurred should factor into the court’s decision to award fees for bad faith. See Wright v. Jackson, 522 F.2d 955, 958 (4th Cir. 1975) (“An award [of fees] for obstinacy, although a penalty, is only for the unnecessary efforts occasioned by the obstinacy.”); cf. Leffler v. Meer, 936 F.2d 981, 987 (7th Cir. 1991) (noting “the duty to mitigate legal fees by promptly, where possible, disposing of baseless claims through summary procedures”); Thomas v. Capital Sec. Servs., Inc., 836 F.2d 866, 879 (5th Cir. 1988) (factoring into fee award *33“the extent to which the nonviolating party’s expenses and fees could have been avoided or were self-imposed”).

Worse still, Heartland itself filed a vastly overblown application for fees that unjustifiably included the agency litigation that the Board had every right to pursue under the Sixth Circuit’s “clear and unmistakable waiver” precedent. Heartland thus has not exhibited the care and calibration that equity desires in those who themselves seek equity.

Having said that, the majority opinion (at 25-26) quite fairly calls the Board out for its failure to candidly acknowledge binding circuit precedent in its answering brief and for pressing only a gossamer-thin argument for distinguishing Enloe. Indeed, I might well have been persuaded that a small amount of fees should be awarded only for the portion of Heartland’s reply brief that was dedicated to rebutting the Board’s frail argument. But that is not the course that the majority opinion takes or that Heartland sought.

For the foregoing reasons, I respectfully dissent.