joins, dissenting:
In this case, a majority of the Court accepts the premise that the Court has *206appellate jurisdiction over a Board decision that refers a matter to an RO for an initial decision. Although this order of the en banc court is well meaning, it is legally incorrect, misguided in practice, and works a substantial injustice on veterans, taxpayers, and survivors.
I. INTRODUCTION
The infirmity of the majority opinion is simply breathtaking. The order announces a purportedly limited rule without providing analysis sufficient to support its conclusion. Rather than identify any statutory basis for its holding, the majority first cites Clemons v. Shinseki, 23 Vet.App. 1 (2009), for the proposition that a claim for a benefit generally encompasses all possible diagnoses. This statement is true but irrelevant. It says nothing about how to draw a jurisdictional line once the Secretary explicitly bifurcates a claim and considers multiple diagnoses separately. The majority then states its desired conclusion but provides no directly applicable supporting authority. Finally, the order asserts that the dissent has failed to appreciate a number of decisions that the majority itself fails to analyze. In short, rather than coming to a reasoned conclusion based on analysis of the Court’s precedent, the majority starts with a predetermined outcome and works backward from it.
As discussed below, the Court, by its decision, exercises direct appellate review over an interlocutory decision and, in doing so, ignores two key facts: (1) Interlocutory decisions by the Board are properly considered only through the Court’s authority under the All Writs Act (AWA), 28 U.S.C. § 1651(a), and (2) veterans are better served if such errors are corrected promptly though petitions for extraordinary relief in the nature of writs of mandamus. In so doing, the majority loses sight of the injury that Mr. Young sought to remedy by appealing to the Court, i.e., the Board’s failure to accord him expedited processing before VA. By requiring Mr. Young and all future appellants to seek this relief through an appeal to the Court rather than through a petition, the majority creates an untenable framework whereby an appellant must needlessly endure months of avoidable processing at the Court to ensure expedited processing below. Neither the law nor common sense supports such an incongruous result.
Although the majority is clearly motivated by what it perceives to be the just result in this case, i.e., providing Mr. Young with the expedited processing of his claim to which he is statutorily entitled, it need not cast aside the venerable principles of finality to obtain that result. Rather, under my view, Mr. Young would have been immediately free to challenge the Board’s referral decision through a petition, which would have provided him with expedited processing more quickly than a successful appeal to the Court. By focusing only on its desired result without fully considering its consequences, the majority overlooks a legally sound method for obtaining the same remedy that would ultimately be more advantageous to veterans and their survivors. Absent a more thorough discussion, it is entirely unclear why the majority chooses the slowest available method for ensuring “expedited” claims processing.
Before addressing the merits, I note that the issue of the Court’s jurisdiction over the underlying subject matter arises no more or less merely because the parties agreed to settle the portion of the case pertaining to the application for attorney fees and expenses. The issue of jurisdiction can be raised at any point in a proceeding by any party or by the Court itself. Fugere v. Derwinski, 972 F.2d 331, *207834 n. 5 (Fed.Cir.1992); Clemons, 23 Vet.App. at 2-3. Here, for example, the issue of jurisdiction over the subject of the underlying appeal was first raised by the Court in connection with Mr. Young’s application for attorney fees and expenses pursuant to the Equal Access to Justice Act, 28 U.S.C. § 2412(d) (EAJA). Once raised, the issue cannot be ignored and the Court cannot rest its decision on the mere agreement of the parties that it has jurisdiction. Clemons, 23 Vet.App. at 2-3 (2009). Therefore, to the extent the Court lacks jurisdiction over the merits, it also lacks the authority to ratify a payment of attorney fees and expenses from the U.S. Treasury. See Heath v. West, 11 Vet.App. 400, 403-04 (1998) (dismissing the petitioner’s EAJA application because of a lack of jurisdiction where the Court concluded that it lacked jurisdiction over the underlying petition).
Moreover, in granting the parties’ joint motion to dismiss Mr. Young’s EAJA application pursuant to the parties’ stipulated agreement, the majority has essentially ignored the U.S. Supreme Court’s clear prohibition against exercising jurisdiction by agreement of the parties. As the Supreme Court has succinctly stated: “[N]o action of the parties can confer subject matter jurisdiction upon a federal court.” Ins. Corp. of Ir., Ltd. v. Compagnie des Bauxites de Guinee, 456 U.S. 694, 702, 102 S.Ct. 2099, 72 L.Ed.2d 492 (1982) (emphasis added); see also Mitchell v. Maurer, 293 U.S. 237, 244, 55 S.Ct. 162, 79 L.Ed. 338 (1934) (jurisdiction cannot be enlarged or conferred by agreement of the parties). Although the majority cites Bond v. Derwinski, 2 Vet.App. 376, 377 (1992), and Dofflemyer v. Brown, 4 Vet.App. 339, 339 (1993), for the proposition that the Secretary’s decision to enter into a settlement or stipulated agreement with the appellant “effectively moots the case or controversy” before the Court and therefore deprives the Court of jurisdiction, the majority nevertheless exercises jurisdiction over Mr. Young’s EAJA application by granting the parties’ joint motion to dismiss. In so doing, the majority is allowing the agreement of the parties on the EAJA matter to supersede the predicate question whether the Court had jurisdiction to review the Board’s referral decision in the underlying appeal. In other words, the Court cannot review the parties’ agreement with respect to the EAJA matter until it satisfies itself that it had jurisdiction over the underlying appeal. Consequently, the specific rule that the Court cannot exercise jurisdiction simply by agreement of the parties — if only to award fees — trumps the general rule that an agreement of the parties may moot the case or controversy before the Court.
Insofar as the majority concludes that the Court properly exercised jurisdiction over the referral decision, in light of Bond and Dofflemyer it nevertheless errs in granting the parties’ joint motion to dismiss. As explained above, the majority cannot, on the one hand, exercise jurisdiction over the EAJA matter and grant the parties’ joint motion to dismiss pursuant to a stipulated agreement, while, on the other hand, hold that such an agreement moots any case or controversy before the Court. Ante at 201-02. In the cases cited by the majority to support that disposition, the Court neither granted nor denied the parties’ joint motions to dismiss pursuant to the parties’ agreements, but rather ordered the Clerk of the Court to enter a voluntary dismissal of the matters pursuant to Rule 42 of the Court’s Rules of Practice and Procedure. See Dofflemyer, 4 Vet.App. at 339; Bond, 2 Vet.App. at 377; see also U.S. Vet.App. R. 42 (“On motion of the appellant or petitioner for dismissal, the Clerk may dismiss an appeal, petition, or application for attorney fees and expenses on terms requested by *208the appellant or petitioner, agreed upon by the parties, or previously fixed by the Court.”). Indeed, if the Court lacks jurisdiction over a matter, it necessarily lacks the authority to rule on a motion related to that matter. Therefore, if the majority wishes to place the Court’s imprimatur on the parties’ stipulated agreement by granting the joint motion to dismiss, it must necessarily overrule Bond and Dofflemyer. However, if the majority wishes to rule in accordance with the precedent established in those cases and order the Clerk to enter a voluntary dismissal of the EAJA matter, then any discussion of the Court’s jurisdiction to review the Board’s referral decision is necessarily dicta and not binding in future appeals. See Jama v. Immigration & Customs Enforcement, 543 U.S. 335, 352 n. 12, 125 S.Ct. 694, 160 L.Ed.2d 708 (2005) (“Dictum settles nothing, even in the Court that utters it.”); see, e.g., Freeman v. Shinseki, 24 Vet.App. 404, 412 (2011) (holding that dicta in Willis v. Brown, 6 Vet.App. 433 (1994), was not controlling); Parrish v. Shinseki, 24 Vet.App. 391, 396 (2011) (holding that dicta in Hilkert v. West, 12 Vet.App. 145 (1999) (en banc), was not controlling).
II. REFERRALS ARE NOT “FINAL” DECISIONS OF THE BOARD
A. Referral Decisions are Generally not Final
As with all inferior Federal courts, this Court is a creature of statute and its authority is limited to that granted by Congress. See Mayer v. Brown, 37 F.3d 618, 619-20 (Fed.Cir.1994) (holding that the Court’s “jurisdiction is limited by statute to review of ‘decisions of the Board of Veterans’ Appeals’ ” and that 38 U.S.C. § 7261 “does not independently grant jurisdiction over [actions of the Board Chairman]”). Therefore, unlike the majority, I find it necessary to look to the Court’s jurisdictional statute to resolve a question as to the extent of its authority. Pursuant to 38 U.S.C. § 7266(a), the Court is authorized to conduct appellate review only of “final decision[s]” of the Board. This Court has long held that it cannot hear interlocutory appeals or otherwise conduct appellate review of Board decisions that are not final, such as remands, see, e.g., Breeden v. Principi 17 Vet.App. 475, 478 (2004), and, relevant to this appeal, referrals, see Link v. West, 12 Vet.App. 39, 47 (1998) (“Claims that have been referred by the Board to the RO are not ripe for review by the Court.”). In Jarrell v. Nicholson, 20 Vet.App. 326 (2006) (en banc), the Court reiterated that it does not have jurisdiction to review a claim absent a final decision by the Board on the claim in dispute.
The U.S. Court of Appeals for the Federal Circuit (Federal Circuit) has also been clear in its holdings that this Court’s appellate authority is limited to those Board decisions that meet the definition of “final” under 38 U.S.C. § 7266(a). In Mayer v. Brown, supra, the Federal Circuit held that this Court does not have jurisdiction to review the Board Chairman’s decision to deny reconsideration. In Howard v. Gober, the Federal Circuit held that this Court does not have jurisdiction to consider a motion to revise a final decision based on clear and unmistakable error (CUE) that was not first decided by the Board. In short, the finality requirement is well established by and firmly entrenched in the Court’s jurisprudence.
The question presented in this case is whether a Board decision referring an issue for adjudication by an RO is a “final” decision or is better characterized as an interlocutory decision. Black’s Law Dictionary defines a “final” judgment as the “the last action that settles the rights of the parties and disposes of all issues in *209controversy, except for the award of costs” and notes that under “the final-judgment rule” “a party may appeal only from a ... final decision that ends the litigation on the merits.” Black’s Law Dictionary 705, 919 (9th ed. 2009) (emphasis added). In contrast, it defines “interlocutory” as “not constituting a final resolution of the whole controversy.” Id. at 889. It further defines an “interlocutory appeal” as “[a]n appeal that occurs before the trial court’s final ruling on the entire case.” Id. at 113 (emphasis added).
1. Distinction Between Final and Interlocutory Decisions
The distinction between final and interlocutory Board decisions has two important implications beyond the existence of the Court’s appellate authority. First, the Federal Circuit has held en banc that “[principles of finality and res judicata apply to agency decisions that have not been appealed and become final.” Cook v. Principi 318 F.3d 1334, 1337 (Fed.Cir.2002) (en banc). Thus, if a type of Board decision is categorized as “final,” then it is binding on future proceedings if it is not appealed. Therefore, if the Board refers a matter to the RO and the appellant does not challenge that action immediately, the Court will be unable to correct any error on direct appeal of a later effective-date determination because the Board’s determination of when the claim was first filed will be beyond the Court’s jurisdiction.
Despite the deep unfairness of this result, the majority fails to offer a coherent argument as to how this outcome is not compelled by its decision. Initially, the majority appears to assert that the determination of when a claim was first filed for purposes of determining whether to refer or remand the claim is independent from the same determination for any other purpose. However, collateral estoppel generally prevents any adjudication system from issuing decisions that reach inconsistent conclusions about the same factual issue. Hence, this Court has repeatedly modified decisions to avoid collateral estoppel issues when the Board has made unnecessary findings of finality unfavorable to a claimant. See Juarez v. Peake, 21 Vet.App. 537, 544 (2008); Seri v. Nicholson, 21 Vet.App. 441, 444-45 (2007). Rather than address Cook and the cases cited therein, the majority cites Federal Circuit precedent outside of veterans law pertaining to court litigation to assert that a different test would apply and then fails to explain its application.
Second, where the Court has jurisdiction over a final Board decision, its jurisdiction is exclusive. As the Court noted in Wachter v. Brown, “[a] party cannot be in ‘two places at the same time.’ ” 7 Vet.App. 396, 397 (1995) (quoting Bellsouth Corp. v. FCC, 17 F.3d 1487, 1489 (D.C.Cir.1994)). This principle prevents VA from making determinations that may conflict with the Court’s rulings in the case and vice versa. As a result, the Court has held that the Board lacks jurisdiction to grant a motion for reconsideration as long as the Court has jurisdiction over the decision. Cerullo v. Derwinski, 1 Vet.App. 195 (1991). Similarly, the Court held in Pulac v. Brown, 10 Vet.App. 11, 12 (1997), that the Court cannot take jurisdiction over an appeal if the appellant filed a motion for reconsideration before filing his Notice of Appeal (NOA). See also Losh v. Brown, 6 Vet.App. 87 (1993) (holding that, if an NOA and motion for reconsideration are filed simultaneously, jurisdiction remains with the Board). Therefore, under Cerullo and Pulac, if the Court has jurisdiction over the Board’s decision to “refer” a claim, then the RO may not take any action on that claim until all appeals of the Board decision have been completed and jurisdiction is returned to VA. To hold otherwise allows VA to poten*210tially waste time and resources operating under a mistaken premise as to when a claim was first filed, when the Court can reach a different conclusion.
2. VA’s Treatment of Referred Matters
Whether the Court has jurisdiction over a Board decision to refer a matter for adjudication by an RO must turn on the undisputed facts as to how VA treats such decisions. First, VA does not consider such decisions to have any res judicata effect on the merits of the referred matter. Second, when a matter is referred to an RO, it commences work on the matter immediately regardless of whether the Board decision ordering the referral has been appealed to this Court. Moreover, neither party argues that VA should treat referred matters differently than it actually does. Nonetheless, the parties argue that the referral decision is a “final” decision. The essence of the parties’ argument is that the Board’s characterization of its action as a referral instead of a remand has binding effect on how the claim will be processed because it governs the priority that the matter receives in further Agency proceedings.
Unfortunately, the parties’ desire to ignore the plain characteristics of the referral decision cannot transform an interlocutory decision into a final one. The parties agree that the only effect of the Court’s decision would be on the priority assigned to the processing of the claim by VA. The first flaw in the parties’ argument is that the authorities cited above amply demonstrate that a final decision is one that resolves a claim on the merits, while a decision that merely governs how a claim will be processed is an interlocutory one. A Board decision sending a claim back to an RO for additional action on a claim simply cannot be parsed into discrete pieces where every specific determination is classified as a separate decision that can be deemed final if an erroneous determination could have an adverse effect on the processing of the claim. Such a semantics game ignores the plain definitions of “final” and “interlocutory” and essentially swallows the concept of interlocutory actions. This attempt to separate the referral action from the merits would also plainly contradict Federal Circuit precedent interpreting this Court’s jurisdiction:
A “decision” of the Board, for purposes of the Veterans Court’s jurisdiction under section 7252, is the decision with respect to the benefit sought by the veteran: those benefits are either granted (in which case the Secretary of Veterans Affairs (Secretary) is bound by the decision and, under section 7252, may not appeal to the Veterans Court), or they are denied.
Maggitt v. West, 202 F.3d 1370, 1375 (Fed.Cir.2000).
3. Majority’s Misapplication of Prior Precedent
The majority and the parties cite to Manlincon v. West, 12 Vet.App. 238, 240-41 (1999), as an example of a case where the Court reviewed the propriety of a Board decision referring a matter to an RO. However, the question of the Court’s jurisdiction to conduct such a review was simply never raised or considered in Manlincon. Therefore, Manlincon is not binding precedent on the issue presently before the Court. A situation analogous to the present case occurred in Harms v. Nicholson, 20 Vet.App. 238 (2006) (en banc), where the appellant argued that a prior ruling on the substance of a matter was precedent that the Court had jurisdiction over the issue. However, the en banc Court flatly rejected this notion and held the decision was not precedent on the issue because the cited decision assumed jurisdiction without explaining why juris*211diction would exist. Id. at 245 (citing United States v. L.A. Tucker Truck Lines, Inc., 344 U.S. 33, 38, 73 S.Ct. 67, 97 L.Ed. 54 (1952) (holding that an issue not “raised in briefs or argument nor discussed in the opinion of the Court” cannot be taken as “a binding precedent on this point”); Webster v. Fall, 266 U.S. 507, 511, 45 S.Ct. 148, 69 L.Ed. 411 (1925) (stating that “[questions which merely lurk in the record, neither brought to the attention of the court nor ruled upon, are not considered as having been so decided as to constitute precedents”)). Hence, the Court’s en banc decision in Harms squarely disposes of the argument that Manlincon or any similar case purporting to exercise jurisdiction without any express consideration of the issue has precedential value. Therefore, the second half of the majority’s first footnote is empty verbiage because it merely asserts that the majority can rely on a case unsupported by any jurisdictional analysis to justify the majority’s outcome.
The majority attempts to support its reasoning by citing King v. Nicholson, 19 Vet.App. 406 (2006), for the proposition that the Court has jurisdiction to determine whether the Board had jurisdiction. Ante at 203. However, King actually holds that the Court has jurisdiction to review the Board’s determination that it did not have subject-matter jurisdiction. 19 Vet.App. at 409. The issue here is not subject-matter jurisdiction, but finality. In this case, the Board did not find that it lacked subject-matter jurisdiction over the referred matter and did not make a final decision refusing to act on that matter. Rather, it made an interlocutory decision to take a particular type of action on the matter. It is inherent in the referral that the Board would not refuse to review the matter, as it did in King, if the RO denied the claim. Accordingly, King and the other eases where the Court has reviewed a final decision of the Board refusing to act based upon a lack of statutory authority are clearly distinguishable from this decision where the Board made a factual determination as to what action to take. However, by relying on King, the majority eviscerates its claim that it is not holding that all referral decisions are final and subject to direct appeal. Ante at 202.
4. Adverse Impact on Veterans and Survivors
Ultimately, the Court’s review of the Board’s decision to refer a matter not only ignores the important distinctions between “final” and “interlocutory” actions but also increases the inefficiency of the VA claims process and adversely affects claimants. First, under Cerullo and Pulac, both supra, it is clear that, if the Court asserts jurisdiction over the propriety of a referral decision, VA is barred from acting on the claim until this Court’s review is concluded. As explained in more detail below, the delay caused by such review would negate any priority treatment that might ultimately be obtained and, as such, achieve exactly the opposite result than that sought by the parties. Second, determining the scope of a claim prior to its full development may well result in unfavorable decisions where a fuller record would better show that a particular matter is best viewed as part of a larger claim. Third, under Cook, a referral decision not appealed to this Court would have to be treated as a binding determination that precludes a contrary determination in a later VA decision absent new and material evidence or a successful collateral attack.
The parties and the majority wish to simply ignore these very real consequences, all of which work to the disadvantage of the veteran in terms of the speed with which the claim is resolved. However, no authority is offered nor exists for the proposition that the Court can strip *212the defining characteristics of finality from a decision and still call it final. See Kuzma v. Principi, 16 Vet.App. 140,145 (2002) (en banc order) (“An epigram widely attributed to Abraham Lincoln is appropriate in this case: How many legs does a dog have if you count his tail as a leg? Four. You can call a tail a leg if you want to, but it doesn’t make it a leg.”). Indeed, the majority does not even attempt to address the proper meaning of either “final” or “interlocutory.” In addition, it offers no authority for the proposition that the Court can ignore the clear statement by the Federal Circuit in Maggitt that a Board “decision” for the purposes of appealing to this Court is the decision on the benefit sought, and the majority recasts every procedural action taken on a case as a “final” decision independent of the merits of the appeal. Such a semantics game would not only obliterate a clearly expressed limitation on the Court’s jurisdiction but would also require the Board to issue a full decision supported by reasons or bases and a notice of appellate rights not only each time it refers a matter to an RO, but also each time it assigns a docket number to an appeal. See 38 U.S.C. § 7104(d) (defining the statutory requirements “[e]ach decision of the Board shall include”). Again, the majority’s interpretation would place unnecessary burdens on the already overwhelmed VA ROs and the Board and would have an adverse effect on the timeliness of claims resolution. See Board of Veterans’ Appeals, Report of the Chairman 15 (2012) (stating that 41,005 appeals were pending at the Board at the end of fiscal year 2011); An Examination of Poorly Performing U.S. Department of Veterans Affairs Regional Offices: Hearing Before the Subcomm. on Disability Assistance and Memorial Affairs of the H. Comm, on Veterans’ Affairs, 112th Cong. 1 (2011) (statement of Rep. Runyan, Sub-comm. Chairman) (noting that 809,000 claims were pending at VA ROs in 2011). There is simply no basis for believing that Congress intended this Court’s review of the Board’s operations to routinely extend to micromanaging the processing of appeals in this manner.
Accordingly, the Court cannot divorce the referral decision from the merits of the claim, and the Court should conclude that it lacks jurisdiction to conduct appellate review of that matter until presented with a proper appeal of a final Board decision resolving the merits of the claim. See Ingram v. Nicholson, 21 Vet.App. 232, 254 (2007) (the issue of when a claim was first raised can be addressed when, and if, it ever “actually becomes relevant to an award of benefits”). Consequently, because the Court lacked jurisdiction to consider a direct appeal of the Board’s referral decision, I conclude that it lacks jurisdiction to consider the EAJA application filed in connection with that appeal and reject the attempt to use the authority of this Court to enforce an EAJA award. See, e.g., Heath, 11 Vet.App. at 403-04.
B. The Referral of a Bifurcated Matter
Although it is absolutely clear that the Court generally does not have jurisdiction to review a referral decision, the majority appears willing to make an exception where a claim has been bifurcated and the Court has jurisdiction over the theory of entitlement that was not referred. However, such an exception cannot be justified under the Court’s caselaw. In Tyrues v. Shinseki, 23 Vet.App. 166 (2009) (en banc), vacated and remanded for reconsideration, 462 Fed.Appx. 956 (Fed.Cir.2012),5 *213this Court granted the Secretary broad discretion to dismember a claim and adjudicate the pieces in jurisdictionally separate proceedings. See also Locklear v. Shinseki, 24 Vet.App. 311, 315 (2011) (“Bifurcation of a claim generally is within the Secretary’s discretion.”). Tyrues explicitly held that “this Court’s jurisdiction is controlled by whether the Board issued a ‘final decision’ — i.e., denied relief by either denying a claim or a specific theory in support of a claim." 23 Vet.App. at 178 (emphasis added). Thus, Tyrues states clearly that the Court’s jurisdiction over a dismembered claim extends only to the specific theories that are subject to a final decision.
In applying this holding, Tyrues held that the Board decision at issue “was final concerning the issue of ... direct service connection,” but was “preliminary” — i.e., not final — concerning the issue of presumptive service connection, which was remanded by the Board. Id. at 180-81. This language makes clear that when the Board bifurcates a claim, the Court has jurisdiction over only the theory or theories that have been finally denied. However, in this case, the majority bootstraps the Court’s jurisdiction over a theory denied by the Board and unquestionably before the Court to review a portion of the Board decision that expressly returned part of the claim to the RO for initial consideration, thus keeping that matter within the administrative adjudication process. This bootstrapping violates not only the express language and logic of Tyrues but also the common sense notion that, once a matter is bifurcated by the Board, the Court’s jurisdiction over the finally decided portion does not give it authority to entertain an interlocutory attack on the portion that is still under consideration by the Agency.
The majority’s citation to Clemons cannot demonstrate that the Court has jurisdiction over the referral decision. Clemons merely stands for the proposition that the default scope of a claim is controlled by the nature of the disability, not any particular diagnosis. 23 Vet.App. at 5. Clemons addresses only the scope of the Court’s jurisdiction where a claim has not been broken apart. Nothing in Clemons provides jurisdiction over the nonfinal portions of a claim that have been bifurcated and sent to the RO for further proceedings. Nevertheless, the majority holds that Clemons enables the Court to bestow on itself jurisdiction over those non-final portions of a bifurcated claim by reassembling the claim for the limited purpose of reviewing a referral decision, while simultaneously allowing VA to adjudicate the merits of the nonfinal portions of that claim.
Simply put, in Tyrues, the full Court, divided though it was, made it crystal clear that once the Secretary breaks up a claim, the pieces are distinct and separately appealable. In such a situation, the Court’s jurisdiction derives not from an appeal of the original claim as a whole, but from an appeal of each theory that is finally decided by the Board. Even when it is readily apparent that the Board’s instructions to the RO on the nonfinal portion of a claim are erroneous, the Court simply has no authority to conduct interlocutory review merely because the Court happens to have jurisdiction over a different portion of that claim.
The majority’s treatment of Tyrues, however, misstates this precedent. The majority cites it as evidence of “a longstanding practice of exercising jurisdiction over theories or parts of a claim." Ante at 205 (emphasis added). However, the majority offers in its citation to Tyrues no *214support for adding the emphasized language to the citation. In other words, the majority fails to explain how Tyrues supports its assertion that the Board’s final denial of a theory offered to support a finding of service connection confers jurisdiction on the Court to review a separate theory that the Board referred. There is a substantial difference between a bright-line rule that divides the Court’s jurisdiction by theory and a fuzzy assertion that the Court can review whatever “parts” of a Board decision it wishes when a matter has been bifurcated. The net effect of the majority’s holding is to grant the Court the ability to extract from a Board decision whatever matters the Court wishes to review, regardless of whether those matters were finally decided by the Board. Ante at 203-05. Thus, it appears that the Court can now review notice, duty-to-assist, and other procedural challenges as to referred or remanded matters so long as it has jurisdiction over a “part” of the claim.
III. REFERRAL DECISIONS SHOULD BE REVIEWED THROUGH WRITS
To the extent that the Court possesses a limited ability to conduct an interlocutory intervention into a claim being processed by VA, that power exists under the AWA, 28 U.S.C. § 1651(a). See Cox v. West, 149 F.3d 1360, 1363-64 (Fed.Cir.1998). Under that authority, the Court may “compel action of the Secretary unlawfully withheld or unreasonably delayed.” Ramsey v. Nicholson, 20 Vet.App. 16, 21 (2006) (quoting 38 U.S.C. § 7261(a)(2)). However, the Court’s power under the AWA is limited to extraordinary situations and, of most concern, the Federal Circuit has emphasized that “ ‘extraordinary writs cannot be used as substitutes for appeals, even though hardship may result from delay and perhaps unnecessary trial.’ ” Lamb v. Principi, 284 F.3d 1378, 1384 (Fed.Cir.2002) (quoting Bankers Life & Cas. Co. v. Holland, 346 U.S. 379, 383, 74 S.Ct. 145, 98 L.Ed. 106 (1953)). By determining that the Court has jurisdiction to determine whether a remand or referral is appropriate, the majority necessarily removes the remedy of a writ of mandamus from the hands of the appellant.
Even setting aside the binding precedent of the Federal Circuit, the difference between this case and review of a final Board decision is illustrated by the Court’s decisions in Ribaudo v. Nicholson, 20 Vet.App. 552 (2007) (en banc) and Ramsey, supra. Both cases dealt with arguments that the appeals involved were statutorily entitled to more expeditious treatment. In each case, the Court held that the Chairman of the Board could not unilaterally stay the processing of selected appeals while the Secretary pursued an appeal of a decision of this Court. Ribaudo, 20 Vet.App. at 559; Ramsey, 20 Vet.App. at 37. However, neither Ribaudo nor Ramsey was an exercise of direct appellate review. Rather, both cases were decided pursuant to petitions seeking extraordinary relief under the AWA. It is clear from these cases that, if the appellant disputes the priority assigned to the processing of his appeal, the proper vehicle for such a challenge is a petition for extraordinary relief, not a direct appeal.
Using petitions for extraordinary relief is not only the proper tool under the jurisdiction provided to this Court by Congress, it is also the device that provides the best relief for veterans. The risk of harm created by a Board decision to refer a matter instead of remanding it is that the matter will not be processed promptly. Petitions are far superior for addressing time-sensitive matters, as a review of the Court’s processing statistics readily and strikingly illustrates. Last year, the median time from the filing of a Notice of Appeal at this *215Court to disposition of the appeal was 328 days or 10.9 months, while the median time from the filing of a petition to disposition was 54 days or 1.8 months.6 By concluding that a veteran may challenge a referral decision only through an appeal and not through a petition, the majority forces the veteran to needlessly wait, on average, an additional nine months to obtain expeditious processing of a claim. In essence, the majority would impose, as a matter of law, a necessary and substantial delay to remedy a complaint that a matter was not being processed with the alacrity required by law. This result is particularly inapposite to the legal basis advanced by Mr. Young and relied on by the majority for exercising jurisdiction over the Board’s referral decision; namely, that an appeal to the Court is necessary to ensure that VA processes the matter expeditiously. In short, I fail to see how an appeal to the Court, which takes, on average, six times as long to process as a petition, is a better method for ensuring expedited processing.7
It is true that the Court rarely grants a petition for extraordinary relief. However, it should not be assumed from this fact that petitions are an ineffective tool for obtaining relief. The reality is that the Court regularly orders the Secretary to respond to a petition that sets forth a well-pleaded complaint that the processing of a claim has been improperly delayed. When the Court issues such an order, the great majority of the time the Secretary responds by correcting the problem within the short time allotted for a response, and the petition is dismissed as moot because the relief sought has been obtained. Thus, the petition, in practice, is a tool far better than the direct appeal for obtaining expeditious processing of the underlying issue.
In rejecting the petition as the proper tool, the majority compares the 88 petitions denied to the one petition granted while glossing over the 61 petitions that were dismissed either voluntarily or by the Court for reasons other than default. See United States Court of Appeals for Veterans Claims, 2011 Annual Report, http:// www.uscourts.cavc.gov/annual_report/. A cursory search on Westlaw for “petition & moot & dismissed & da(aft 9/30/2010 & bef 10/1/2011)” produces 54 results, which shows that these dismissals were almost exclusively based upon mootness because the Secretary responded to the petition by remedying the problem without requiring a Court order. The fact that the Court *216had to formally order the Secretary to act in only one petition actually demonstrates how much more efficient a tool the petition provides when it is clear that a mistake has been made.
The majority obscures this fact by citing Costanza v. West, 12 Vet.App. 133, 134 (1999), for the proposition that an 11-month delay does not amount to an arbitrary refusal to act sufficient to justify mandamus. Ante at 203-04 n. 3. However, when the Board refers a matter, an individual need not wait until there has been delay sufficient for the Court to find that VA has arbitrarily refused to act on that matter. Rather, the referral decision can be challenged immediately — just as the stay orders in Ribaudo and Ramsey were — because the Secretary has announced his intention as to how the matter will be processed, thereby eliminating any danger of the Court imputing a determination that does not actually exist.
Although the majority maintains that it is “a red herring” to assert that petitions would provide for faster relief because “none of those dismissals involved a petition to amend a Board decision referring rather than remanding part of a claim,” ante at 204 n. 3, there is no doubt that claimants and their representatives would use the proper tool once the Court guides them in the right direction. See, e.g., Ingram, supra (explaining the proper time and procedures for asserting a pendingunadjudicated-claim argument); DiCarlo v. Nicholson, 20 Vet.App. 52, 57 (2006) (explaining that there is no freestanding “finality claim”). Unfortunately, this case not only chooses the wrong path, but also leaves other claimants who have received improper referrals to wonder how to proceed. May a claimant file a petition if no related theory was denied? If so, why are claimants who have a related theory denied treated differently? May a claimant who does not receive a final Board decision as to any benefit file a direct appeal as to a referral action? If so, on what basis would we have jurisdiction if Clemons clearly cannot be applied? Are we without any type of jurisdiction if a referral action cannot be linked to a final decision by the Board? If so, how can it be that some referral errors may be corrected, but not others? This opinion leaves future recipients of referrals wondering whether to file a petition, a direct appeal, or nothing at all. When these future cases arise, the Court will be forced to either review all referral actions through the time-consuming direct appeal process or to treat some claimants differently than others.
Although the issue is not presently before the Court, it appears that this case may well have merited relief through a petition if the Secretary had taken the unusual step of refusing to remedy a well-supported assertion that the claim was not being properly processed. The memorandum decision addressing the Board referral decision is short on details and fails to state what standard of review it applied. Young v. Shinseki, No. 09-1621, 2010 WL 2640592 (Vet.App. June 30, 2010). However, it appears that the decision was a straightforward application of Clemons to undisputed facts nearly identical to the facts of that case. Hence, it appears likely that the appellant would have demonstrated a clear and indisputable right to the writ if the Secretary had contested the petition. See Cheney v. U.S. Disk Court, 542 U.S. 367, 380-81, 124 S.Ct. 2576, 159 L.Ed.2d 459 (2004).
The majority maintains that its decision does not foreclose the use of petitions to review these types of errors. Ante at 203-04 n. 3. This assertion, however, is wholly inaccurate. The Supreme Court has said in no uncertain terms that “it is established that the extraordinary writs cannot *217be used as substitutes for appeals ... and whatever may be done without the writ may not be done with it.” Bankers Life, 346 U.S. at 383, 74 S.Ct. 145 (emphasis added). Thus, it is indisputable that by-holding that such matters can be reviewed on direct appeal, the majority is also holding that they cannot be reviewed through a petition.
The majority asserts that it is unclear in my analysis whether “a claimant would be forced to seek mandamus to either correct an error in a Board decision that reflects referral instead of remand or to expedite the decision-making process.” Ante at 203 n. 3. The implicit premise in this assertion is that a Board decision may only be corrected through direct appellate review. However, there is simply no support for this premise. The first step in determining the Court’s jurisdiction does not involve looking at the form of the VA action that allegedly contains error. Rather, the first step is to look to the harm asserted and the relief requested by the party seeking review. In this case, the only harm asserted by the appellant in the Board’s referral decision was a loss of priority in the claims adjudication process, and the only remedy sought was more expeditious processing. As demonstrated above, the Court’s authority to grant such a relief is through its mandamus power and is exclusive to that power because there cannot be any overlap with its direct appellate authority. In exercising its mandamus power, the Court may be required to correct an error in the Board decision in order to provide the remedy requested. In doing so, however, the Court would be correcting an error in a nonfinal Board decision.
A proper interpretation of the line that divides the Court’s jurisdiction is simple. If a claimant wishes the Court to review whether or not he or she is entitled to a benefit, then that review must wait until there is a Board decision that denies that benefit because such arguments can only be considered on direct appellate review. On the other hand, if the claimant wishes the Court to intervene in the adjudication process to address a matter other than entitlement to a benefit, then the proper method for seeking review is through a petition for mandamus because speeding the process of adjudication aids the Court’s direct appellate authority by ensuring that the Secretary does not avoid it through unlawful delay. As the claimant in this case was seeking a Court order for a speedier decision rather than a determination of entitlement, the only proper vehicle is a petition for mandamus.
IV. ATTORNEY FEE ISSUES
Ultimately, only attorneys can benefit from this decision, and they will do so at the expense of veterans, survivors, and taxpayers. In particular, by finding that the Court has jurisdiction to consider interlocutory matters on direct appeal, rather than through petitions, the majority grants an unearned financial benefit to attorneys filing such appeals in the form of EAJA fees. By endorsing a remedy that takes six times longer on average to resolve than a petition, the majority creates an opportunity for unscrupulous attorneys to draw out litigation to their personal gain. The likelihood of such an abuse is compounded by the fact that the Court routinely awards EAJA fees in direct appeals, when — as here — the Secretary does not contest the application, even when more efficient alternatives exist. Even for ethical attorneys, a direct appeal by its very nature involves a larger investment of time, thereby generating greater costs to taxpayers through EAJA fees than a petition would generate. In short, the majority’s decision unnecessarily increases the burden on the public fisc.
*218This boon to attorneys and increased cost to taxpayers is especially troublesome in light of the directly proportional detriment suffered by claimants. Although attorneys will earn more EAJA fees in a direct appeal than a petition, their clients will lose the benefit of the remedy sought — expedited processing of their claims. Indeed, in some cases, the Court’s decision will be moot by the time the full appellate process has been completed. In my view, the Court should not structure the system to pit the financial interests of unscrupulous attorneys against the best interest of their clients, nor should it create a situation where taxpayers are charged more for fewer benefits to claimants. See Massie v. Shinseki, 25 Vet.App. 123, 135 (2011) (“Any monetary incentive for attorneys who represent veterans before VA should be oil for the gears of the system, not sand in the works.”). Unfortunately, the majority’s decision produces just such a result, requiring attorneys to take the slow, costly route to relief.
V. CONCLUSION
A Board decision referring a matter to an RO for an initial decision is unquestionably an interlocutory action. The Court’s jurisdiction to intervene in interlocutory matters is authorized only by the AWA. Moreover, the Court’s AWA power is the more effective way to provide prompt relief when the Board errs in referring rather than remanding a matter. Not only is a petition faster, but the Court’s jurisdiction over a petition is not dependant on the happenstance of the Board finally deciding a portion of a bifurcated claim in the same decision that it refers another portion of that claim. Indeed, the only beneficiaries of the majority’s decision are the attorneys who now have every incentive to forego a motion for reconsideration — the avenue that would be most beneficial to veterans and survivors — and file an appeal with the Court in hopes of recovering EAJA fees for minimal effort.
The memorandum decision in this case erred in exercising direct appellate review on the basis of an NOA filed as to a Board decision that was not “final” as to the matter that the appellant wished to dispute. If relief in such cases is needed, it may be obtained most efficiently through filing a motion for reconsideration at the Board or a petition for a writ of mandamus if the authority of the Court is needed. However, because the Board decision on appeal does not give the Court jurisdiction over the underlying matter in dispute, it lacks the authority either to address the merits of the referred matter or to order the payment of fees for the litigation of a matter that was not properly before the Court.
Nevertheless, by concluding that the Court has jurisdiction to review a Board referral decision, the majority has established the jurisdictional predicate necessary to award EAJA fees in appeals of such decisions. In so doing, the majority has expanded the Court’s appellate jurisdiction beyond that envisioned by Congress to include interlocutory matters that have only a tangential connection to a final denial of benefits. See Briley v. Shinseki, 25 Vet.App. 196, 197 (2012) (explaining that “federal courts ‘possess only that power authorized by Constitution and statute, which is not to be expanded by judicial decree’ ” (quoting Kokkonen v. Guardian Life Ins. Co. of Am., 511 U.S. 375, 377,114 S.Ct. 1673, 128 L.Ed.2d 391 (1994))). This result is all the more egregious because it is claimants that will ultimately suffer from delayed justice while litigation costs increase. Accordingly, I must dissent.
. The continuing viability of Tyrues was not briefed in this case, and the Federal Circuit’s decision vacating the Court's decision for reconsideration is not yet final. Therefore, I *213accept Tyrues as controlling for purposes of this decision.
. By comparison, the average time it takes for a RO to process a claim is 188 days. Department of Veterans Affairs, Fiscal Year 2011 Performance and Accountability Report 1-69 (2011). Thus, if the Court were to routinely use direct appellate review to address erroneous interlocutory decisions by the Board referring a matter instead of remanding it, it is very likely that the matter will have been decided by the RO prior to the Court’s decision.
. Although not pursued in this case, the quickest method for correcting the Board's error likely would have been to immediately seek reconsideration at the Board. See 38 C.F.R. § 20.1000(a) (2011) (stating that "[r]e-consideration of an appellate decision may be accorded at any time by the Board ... on motion of the appellant or his or her representative” to correct, inter alia, an “obvious error of fact or law”); see also 38 U.S.C. § 7103. Where, as here, the Board’s error is clear, the Board would have every incentive to correct that error quickly, as it would be in the best interest of the veteran and less costly to the Secretary in terms of manpower, litigation costs, and potential EAJA fees. Reconsideration is available regardless of whether the Court determines that it has jurisdiction to review a Board referral decision; however, once the appellant files a Notice of Appeal with the Court, the appellant may not return to the Board to seek reconsideration. See Cerullo, 1 Vet.App. at 197 ("Once an appellate body takes jurisdiction over a claim, the lower tribunal may not consider the same issues.”).