Sean A. Ravin v. Robert L. Wilkie

• Court: United States Court of Appeals for Veterans Claims
• Decided: March 20, 2019
• Docket: 16-2057
• Status: Published

Opinion

UNITED STATES COURT OF APPEALS FOR VETERANS CLAIMS

No. 16-2057

SEAN A. RAVIN, APPELLANT,

V.

ROBERT L. WILKIE, SECRETARY OF VETERANS AFFAIRS, APPELLEE.

On Appeal from the Board of Veterans' Appeals

(Argued September 27, 2018 Decided March 20, 2019)

Kenneth M. Carpenter, of Topeka, Kansas, for the appellant.

James R. Drysdale, with whom James M. Byrne, General Counsel; Mary Ann Flynn, Chief Counsel; and Richard A. Daley, Deputy Chief Counsel, were on the brief, all of Washington, D.C., for the appellee.

Thomas E. Andrews, III, of Columbia, South Carolina, for National Organization of Veterans Advocates, Inc. (NOVA) as amicus curiae.

Before DAVIS, Chief Judge, and SCHOELEN, PIETSCH, BARTLEY, GREENBERG, ALLEN, TOTH, and FALVEY, Judges.1

TOTH, Judge, filed the opinion of the Court. PIETSCH, Judge, filed a dissenting opinion in which FALVEY, Judge, joined. FALVEY, Judge, filed a dissenting opinion in which PIETSCH, Judge, joined.

TOTH, Judge: An attorney representing a VA benefits claimant may charge a fee contingent on an award of past-due benefits. An attorney may also seek fees under the Equal Access to Justice Act (EAJA) for successful representation before this Court. Federal law requires an attorney receiving both contingency and EAJA fees to refund to the claimant the smaller of the two awards if they are received for the "same work." We conclude that representation before VA and representation before this Court are not the same work for purposes of this refund requirement. Thus, EAJA fees awarded by this Court need not be offset against contingency fees based on past- due benefits that result from a decision by VA.

1 Judge Meredith did not participate in the consideration or decision of this appeal. Although we reached the opposite conclusion in Carpenter v. Principi, 15 Vet.App. 64 (2001) (en banc), that decision cannot be sustained, as many of its central premises falter under closer scrutiny. Rather than engaging in a textual analysis of the relevant statutory language, the Court employed an interpretive methodology in a novel way that cannot square with EAJA. The policy judgments on which the Court grounded its decision appear less than certain in the light of two decades of experience. There are good reasons to believe that Carpenter's rule has unnecessarily complicated the relationship between VA claimants and their attorneys. And, it has not, to the best of our knowledge, encouraged any reliance interests that might overcome these problems. Thus, to the extent that it is inconsistent with today's decision, Carpenter is overruled.

I. BACKGROUND Before recounting the specific facts of the present case, we provide a brief overview of the statutory provisions that govern compensation for the representatives of VA claimants. A. Under 38 U.S.C. § 5904, an attorney or other authorized agent representing a claimant (for the sake of simplicity we'll refer to attorneys for the remainder of this opinion) in proceedings before VA may charge a fee for services subject to certain limitations. Services for which a fee is charged cannot precede the date on which the Notice of Disagreement is filed in the case. 2 38 U.S.C. § 5904(c)(1). Attorneys must file a copy of any fee agreement for such services with the Secretary, who may reduce the fee if it is unreasonable or excessive. 38 U.S.C. § 5904(c)(2), (3)(A). Also, an attorney representing a claimant in this Court must file with the Court a copy of any fee agreement between them. 38 U.S.C. § 7263(c). The Court can independently review that agreement—or a Board decision regarding an agreement filed with the Secretary—and "may order a reduction in the fee called for in the agreement if it finds that the fee is excessive or unreasonable." 38 U.S.C. § 7263(c), (d).

2 Over the years, Congress has permitted attorneys to begin charging for services at progressively earlier points in the claims process. Before 2006, a fee could not be charged for service predating the first Board decision in the case. See Veterans Benefits, Health Care, and Information Technology Act of 2006, Pub. L. No. 109-461, § 101(c)(1), 120 Stat. 3403, 3407. In a 2017 amendment due to take effect in February 2019, Congress allowed fees to be charged even earlier than the filing of the Notice of Disagreement—namely, the date notice of an initial decision by the agency of original jurisdiction is issued. Veterans Appeals Improvement and Modernization Act of 2017, Pub. L. No. 115-55, § 2(n), 131 Stat. 1105, 1110.

2 A fee "may be based on a fixed fee, hourly rate, a percentage of benefits recovered, or a combination of such bases." 38 C.F.R. § 14.636(e) (2018). One of the most prevalent arrangements is a contingency agreement that permits the Secretary to pay a fee capped at 20% of "the total amount of any past-due benefits awarded on the basis of the claim" directly to the attorney out of a claimant's past-due benefits award. 38 U.S.C. § 5904(d)(1)-(2). This applies to past-due benefits awarded based on proceedings before VA or before this Court. 38 U.S.C. § 5904(d)(3). Not long after this Court was established, Congress authorized another method of compensating attorneys by allowing litigants before the Court to seek EAJA awards. Federal Courts Administration Act of 1992 (FCAA), Pub. L. No. 102-572, § 506(a), 106 Stat. 4506, 4513. EAJA's purpose is to ensure that those trying to vindicate their rights against wrongful government action can obtain adequate representation. Parrott v. Shulkin, 851 F.3d 1242, 1249 (Fed. Cir. 2017). To achieve this goal, EAJA gives a court the authority to award a prevailing party in a lawsuit against the United States reasonable representation costs when the government's position was not substantially justified. 28 U.S.C. § 2412(d)(1)(A). But in making EAJA available to VA claimants in this Court, Congress included the following offset provision: Section 5904(d) of title 38, United States Code, shall not prevent an award of fees and other expenses under section 2412(d) of title 28, United States Code. Section 5904(d) of title 38, United States Code, shall not apply with respect to any such award but only if, where the claimant's attorney receives fees for the same work under both section 5904 of title 38, United States Code, and section 2412(d) of title 28, United States Code, the claimant's attorney refunds to the claimant the amount of the smaller fee.

§ 506(c), 106 Stat. at 4513.3 We'll refer to this statutory language as section 506(c) or the EAJA offset provision. B. Attorney Sean A. Ravin began representing veteran Ira L. Easterling after the Board issued a 2009 decision denying entitlement to a total disability rating based on individual unemployability (TDIU). He entered his appearance in this Court and filed a copy of a pro bono representation agreement signed by both client and attorney. The contract specified that it covered "only" representation before the Court, that the veteran would not be charged, and that Mr. Ravin would

3 This language hasn't been codified in a specific section of the U.S. Code. It is sometimes referred to as a "note" to 28 U.S.C. § 2412. This doesn't change the fact that it bears the full weight of federal law. See Conyers v. MSPB, 388 F.3d 1380, 1382 n.2 (Fed. Cir. 2004).

3 seek compensation for fees and expenses under EAJA. By signing, Mr. Easterling also accepted that any EAJA award would be paid by VA, "separate and apart from any recovery to which I am entitled as a result of my claim and that such award will not affect the amount of any recovery to which I may be entitled." R. at 279. In May 2011, the Court vacated the Board decision for inadequate reasons or bases and remanded the matter. One week later, attorney and client signed a new agreement regarding representation before VA. In this contract, Mr. Easterling agreed to pay Mr. Ravin a contingent fee of 20% of any past-due benefits awarded by the Secretary and authorized VA to withhold any such amount and make direct payment to Mr. Ravin. The agreement also contained a clause in which Mr. Easterling acknowledged that representation before VA would not be considered the "same work" as the prior representation before this Court and that no potential offset against EAJA fees would be required unless this Court ordered the award of past-due benefits. R. at 209-12. Meanwhile, Mr. Ravin sought on the veteran's behalf a $5,787 EAJA award, which the Court granted. In December 2011, the Board determined that the evidence of record was in equipoise and granted TDIU. The following month, the VA regional office (RO) assigned a 2006 effective date resulting in a past-due benefits award of $115,632. VA informed Mr. Ravin that he would be paid $17,239.40, which was equal to 20% of the past-due benefits ($23,126.40) less $5,787 to offset the EAJA fees previously awarded and a $100 administrative fee charged by VA. R. at 195-97. Mr. Ravin challenged the EAJA offset, but the Board concluded that it was mandated by Carpenter. The Board sent copies of its decision to both Mr. Ravin and Mr. Easterling. This appeal followed. The case was referred to a panel to resolve Mr. Ravin's argument that congressional action had abrogated Carpenter. We permitted two interested amici curiae to file briefs and participate in oral argument. Together, Mr. Ravin and his supporting amici argued that 2006 amendments to section 5904 had effectively nullified Carpenter. Alternatively, they argued that Carpenter should be overruled. The panel sua sponte requested a vote on whether the full Court should consider that latter argument, and a majority of the Court voted for en banc consideration. The parties and one

4 amicus were permitted to submit supplemental memoranda of law on the question and new oral argument was held.4

II. ANALYSIS Mr. Ravin and the supporting amicus contend that the meaning of the phrase "same work" in the EAJA offset provision is plain and only obliges an attorney to refund the smaller of section 5904 and EAJA fees when section 5904 fees are awarded by the Court—that is, when a decision of the Court leads directly to the award of past-due benefits—rather than by VA after a remand from the Court. This conclusion, they reason, is buttressed by the fact that this is the rule in the Social Security context. Finally, they contend that Carpenter, which reached a different interpretation of section 506(c), should be overruled owing to its questionable analysis and deleterious effects. The Secretary does not mount heavy opposition to the appellant's statutory- analysis arguments but contends that Mr. Ravin hasn't offered persuasive reasons for overruling Carpenter. A. Statutory analysis always begins with the language of the statute itself. O'Brien v. Wilkie, 30 Vet.App. 21, 25 (2018). The EAJA offset provision mandates that an attorney refund to the client the smaller of an EAJA fee and a section 5904 fee if they are received for the "same work." § 506(c), 106 Stat. at 4513. It's a basic canon of statutory construction that words generally take "their ordinary, contemporary, common meaning," which may be derived from dictionaries from the era of the statutory provision's enactment. Sandifer v. U.S. Steel Corp., 571 U.S. 220, 227 (2014). At the time this provision was enacted in 1988, "same" meant "identical"; "similar in kind, quality, quantity, or degree"; "conforming in every detail." AMERICAN HERITAGE DICTIONARY, SECOND COLLEGE EDITION 1088 (1982); BLACK'S LAW DICTIONARY 1340 (6th ed. 1990) ("Identical, equal, equivalent"); accord WEBSTER'S NEW WORLD DICTIONARY, THIRD COLLEGE

4 At oral argument before the panel, Mr. Ravin's counsel said Mr. Easterling was informed that Mr. Ravin had filed the present appeal, but counsel couldn't tell us to what extent the veteran understood the practical consequences of Mr. Ravin's litigating position. Prior to en banc oral argument, the Court issued an order advising Mr. Easterling of his right to intervene in this litigation and instructed the Secretary to serve a copy of the order on the veteran. Four days later, the Secretary filed notice of this service. To date, Mr. Easterling has not sought to intervene or otherwise responded to the Court's order. The Court is satisfied that Mr. Easterling has been provided with sufficient notice and opportunity to be heard in this matter. We conclude that his absence from this litigation does not prevent the Court from resolving Mr. Ravin's appeal.

5 EDITION 1186 (1988) ("Being the very one, identical; alike in kind, quality, amount, or degree; unchanged; not different"). The noun it modifies—"work"—was defined as: "Physical or mental effort or activity directed toward the production or accomplishment of something." AMERICAN HERITAGE DICTIONARY at 1390; see BLACK'S at 1605. Congress's use of the word "work" is telling: it focuses on the discrete actions of the attorney rather than the broader context of those actions—such as a claim or case. If Congress wished to require an offset that covered all representation during a single claim or case, "it could have said so explicitly, but it did not"; we must presume this choice was deliberate. Saunders v. Wilkie, 886 F.3d 1356, 1366 (Fed. Cir. 2018). Although a statute can depart "from the natural and popular acceptation of language," Sandifer, 571 U.S. at 228, nothing in the text or context of section 506(c) signals that Congress intended to deviate from the ordinary meaning of "same work." In its basic formulation, "same work" means a single, identical effort or activity rather than an extended series of related activities that remain distinct despite sharing a common goal. As a matter of plain language, therefore, "same work" does not cover work done before separate tribunals such as a court and an agency. Although similar, these activities lack the unitary, identical, or undifferentiated qualities necessary to be regarded as the "same." Instead, they are better viewed as actions that are distinct in nature even though they may share salient features, contain some substantive overlap, or are generally oriented towards the same purpose.5 The prohibition on duplicative fees for the "same work" at issue here, then, does not apply when one set of fees are court-awarded EAJA fees and the other set are agency-awarded fees based on its grant of past-due benefits. Conversely, when a decision of this Court results in the award of past- due benefits—such as a reversal of the Board's denial of benefits—the "same work" prohibition applies and an offset is required. This plain reading of "same work" is supported by the meaning this phrase bears in the analogous Social Security provision on which section 506(c) was modeled. See Mount Lemmon Fire Dist. v. Guido, 139 S. Ct. 22, 26-27 (2018) (consulting earlier statute "on which many aspects of the [Age Discrimination in Employment Act] are based"). When Congress passed EAJA in

5 While not directly relevant to our analysis, we acknowledge the practical differences in attorneys' work before VA and work before the Court. In general terms, an attorney's work before the agency involves procuring evidence and forwarding arguments to persuade the agency to award benefits. By contrast, work before this Court is usually limited to developing legal arguments showing discrete error in a Board ruling, as this Court (sparing few exceptions) lacks the authority to make factual findings on the merits of a veteran's claim.

6 1980, it made it available to Social Security claimants seeking judicial review of agency determinations.6 Five years later it added a savings provision "to prevent the windfall of an unauthorized double recovery of fees." Astrue v. Ratliff, 560 U.S. 586, 595 (2010). That provision reads in full: Section 206(b) of the Social Security Act [42 U.S.C. § 406(b)(1)] shall not prevent an award of fees and other expenses under section 2412(d) of title 28, United States Code. Section 206(b)(2) of the Social Security Act [§ 406(b)(2)] shall not apply with respect to any such award but only if, where the claimant's attorney receives fees for the same work under both section [406(b)] of that Act and section 2412(d) of title 28, United States Code, the claimant's attorney refunds to the claimant the amount of the smaller fee.

Pub. L. No. 99-80, § 3, 99 Stat. 183, 186 (1985). Except for statutory citations, this language is identical to the provision at issue here. There can be little doubt that our EAJA offset provision is modeled on the Social Security's EAJA savings provision. See, e.g., Carpenter, 15 Vet.App. at 94 (Kramer, C.J., dissenting) (noting that the language is "exactly the same" and analogizing to the Social Security provision). Section 406(b) governs fees in Social Security cases only for representation before a court; fees for representation in administrative proceedings are controlled by section 406(a). Gisbrecht v. Barnhart, 535 U.S. 789, 794 (2002). Subsection (b) says that a "court may determine and allow as part of its judgment a reasonable fee for such representation, not in excess of 25 percent of the total of the past-due benefits to which the claimant is entitled by reason of such judgment." 42 U.S.C. § 406(b)(1). Thus, the Social Security EAJA provision makes clear that the "same work" language pertains only to section 2412 EAJA fees awarded by a court and section 406(b) fees based on past-due benefits awarded by a court. Consequently, the Social Security savings clause does not require offset of court-awarded EAJA fees against any fees based on past-due benefits granted by the Social Security Commissioner because the work underlying the separate fees is not the "same work." See Rice v. Astrue, 609 F.3d 831, 839 (5th Cir. 2010).

6 This court has repeatedly declined to consult regulations from the Social Security Administration when evaluating similar regulations promulgated by VA. See, e.g., Withers v. Wilkie, 30 Vet.App. 139, 148-49 (2018). That makes sense as such regulations are drafted by distinct agencies subject to distinct statutory mandates and have no legal import outside of their respective regulatory schemes. This is different from consulting statutory language used by Congress across several statutes that courts have interpreted as having a specific, uniform meaning that Congress intended to retain across the various statutes.

7 Where a term is "obviously transplanted from another legal source, whether the common law or other legislation, it brings the old soil with it." Sekhar v. United States, 570 U.S. 729, 733 (2013). When Congress imported the phrase "same work" from the Social Security EAJA savings provision, it also imported the context that gave it meaning. This context excludes administrative proceedings from the "same work" inquiry. In other words, consulting the meaning of "same work" in the parallel Social Security EAJA provision confirms what the plain language indicates: representation before a court is not the "same work" as representation before an agency. B. Based on this analysis, Mr. Ravin would not be obliged to offset any part of the $5,787 in EAJA fees awarded by the Court against the $23,126.40 in contingency fees he received because of VA's award of past-due benefits. This result, of course, cannot be reconciled with our earlier rationale in Carpenter. And although stare decisis normally counsels that we stand by our prior decisions, several reasons convince us that is not the proper course here.7 1. The first "important factor in determining whether a precedent should be overruled is the quality of its reasoning." Janus v. AFSCME, Council 31, 138 S. Ct. 2448, 2479 (2018). On review, Carpenter's analytical approach contained distinct shortcomings: it presented only a cursory analysis of the language of section 506(c) before deeming it ambiguous, and it invoked the pro- veteran canon of construction in a sweeping manner despite that canon's inapplicability in this EAJA context. Before Carpenter, we held through a line of cases that a lawyer's work before the Court and before VA wasn't the "same work" under section 506(c) when the remedy sought or the relief obtained at each stage was different—for example, a judicial remand versus an administrative reversal. See Fritz v. West, 13 Vet.App. 190 (1999); Shaw v. Gober, 10 Vet.App. 498 (1997). Carpenter changed course and overruled these decisions almost entirely on policy grounds. Specifically, it rejected the earlier decisions for interpreting the concept of "work" too narrowly and "leaving open an opportunity for the subversion of the very purpose of EAJA." 15 Vet.App. at 74. EAJA, we noted, was enacted to remove financial deterrents to individuals attempting to defend themselves against unjustified government action and not to provide an attorney "a greater

7 Carpenter also addressed the reasonableness of the 30% contingency fee at issue in that case. 15 Vet.App. at 76-79. This decision does not disturb that part of Carpenter's analysis.

8 fee than contemplated by his fee agreement." Id. at 74-75. The Court then reasoned that to regard work before the Court and the Board as distinct would allow the EAJA fee to enhance the attorney's fee, rather than reimburse the veteran for the cost of representation. Id. at 76. Having laid out these various policy concerns, Carpenter concluded its analysis by invoking, in summary fashion, the presumption spelled out in Brown v. Gardner, 513 U.S. 115 (1994), that textual ambiguities in veterans benefit statutes should be read so as to benefit veterans. Carpenter, 15 Vet.App. at 76 ("If there is any room for interpretive doubt as to what constitutes the 'same work' for the purposes of EAJA, such doubt must be resolved in the veteran['s] favor."). Based on these considerations and the pro-claimant nature of the overall VA benefits claims process, the Court concluded that "representation of a claimant in pursuit of a claim at all stages of the adjudication process is the 'same work,' regardless of the tribunal before which it is performed," with the result that a fee including "both an EAJA award plus a contingency fee for work performed before the Court, Board, and VA on the same claim such that the fee is enhanced by an EAJA award is unreasonable" under section 5904(c). Id. This essentially read "same work" as "same claim" or "same case." As we've said, the "starting point in every case involving construction of a statute is the language itself." Kelly v. Robinson, 479 U.S. 36, 43 (1986). Curiously, though, the Court in Carpenter made no attempt to ascertain the meaning of "same work," even though the text is of primary importance when construing a statute like the EAJA offset provision. Likewise, the Court made no attempt to analyze the statute's structure, context, or analogy to related legal provisions to determine the meaning of "same work" as used in § 506(c). See, e.g., Mount Lemmon Fire Dist., 139 S. Ct. at 26-27; Cook v. Snyder, 28 Vet.App. 330, 338 (2017). Instead, we merely pronounced the statutory language to be ambiguous and proceeded to invoke the Gardner doctrine that ambiguity in veterans benefit statutes should be resolved in the favor of veterans. Carpenter, 15 Vet.App. at 76. But there are flaws in this rationale. As already discussed, the phrase "same work" is not ambiguous. A veteran "cannot rely upon the generous spirit that suffuses the law generally to override the clear meaning of a particular provision." Disabled Am. Veterans v. Gober, 234 F.3d 682, 692 (Fed. Cir. 2000) (quotes omitted). Where a court can discern a clear reading of a provision, it applies that reading, and a canon of deference such as Gardner "leaves the stage," as it exists only to resolve ambiguity. Arangure v. Whitaker, 911 F.3d 333, 340 (6th Cir. 2018)

9 (discussing Chevron U.S.A. Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837 (1984)). It seems an open question whether the pro-veteran Gardner canon is properly invoked in the EAJA context, since "EAJA is not a veterans benefit statute. It is a statute of general applicability" that "allows claimants to recover fees in all sorts of nonveterans cases." Parrott, 851 F.3d at 1251. We need not resolve the question here because, in any event, Gardner is of little utility when it doesn't help resolve ambiguity in a clearly "pro-veteran" manner. See, e.g., Burden v. Shinseki, 727 F.3d 1161, 1169 (Fed. Cir. 2013) (Gardner is of no assistance in deciding whether a statutory interpretation benefitting a veteran's children is more veteran-friendly than one favoring his common-law spouse). Carpenter presumed that an interpretation maximizing the refund to a veteran in a specific case was the most favorable reading of section 506(c). But if (as discussed below) such an interpretation discourages attorneys from representing veterans, thereby depressing the likelihood of their obtaining past-due benefits, that interpretation is not necessarily the most favorable to veterans. Instead of statutory analysis, Carpenter grounded its interpretation of the "same work" language in policy considerations—a practice the Supreme Court has expressly disclaimed: "[W]e cannot overrule Congress's judgment based on our own policy views." SCA Hygiene Prods. Aktiebolag v. First Quality Baby Prods., LLC, 137 S. Ct. 954, 967 (2017). Congress's judgment is ascertained through the statutory text it passes. By relying on policy considerations rather than principles of statutory construction, Carpenter critically undermined its own analysis.8 2. Additionally, legal developments since Carpenter have undermined its reasoning. Cf. Janus, 138 S. Ct. at 2482-83. Carpenter premised its conclusion on the idea that the Court had to arbitrate fee agreements between claimants and attorneys. A "narrow" reading of "same work" was necessary to maximize the fees retained by the client since, in the "uniquely pro-claimant adjudicatory system, protecting the interests of the veteran"—and other VA claimants—"is paramount." 15 Vet.App. at 75. In other words, the Court believed its interpretation was necessary

8 Relatedly, Carpenter's departure from existing precedent was abrupt. It overruled Shaw and Fritz without oral argument, without an en banc conference, without supplemental briefing, and without any notice to the parties to the litigation or the broader veterans law community. 15 Vet.App. at 80 (Steinberg, J., concurring in part and dissenting in part). These procedural shortcuts deprived the Carpenter Court of the sort of input that is critical to informed judicial decision making. And we have little doubt that the lack of input and discussion severely weakened Carpenter's analytical foundation.

10 "to protect the veteran from the drafting of a fee agreement which might unintentionally, or intentionally, deprive a veteran of his rights under the law." Id. This concern is understandable. Congress long curtailed attorneys' involvement in veterans benefits by restricting the scope of their representation and the fees they could seek. In 1864, for example, Congress prohibited attorneys from charging claimants more than $10 to help "establish a claim." Act of July 4, 1864, ch. 247, sec. 12, 13 Stat. 387, 389. Demanding or receiving a higher fee (including a fee based on a percentage of the claim) was a "high misdemeanor" punishable by fine or imprisonment "at hard labor" or both. Id., sec. 13. "The limitation was designed to protect the veteran from extortion or improvident bargains with unscrupulous lawyers." Walters v. Nat'l Ass'n of Radiation Survivors, 473 U.S. 305, 360 (1985). The $10 cap remained in place for more than 120 years. S. Rpt. No. 109-297, at 6-7 (2006).9 In 2001, when Carpenter was decided, attorneys could not charge for services until after issuance of the first final Board decision on a claim. 38 U.S.C. § 5904(c)(2) (2000). In practical terms, this meant that attorneys did not become involved in claims until an appeal to this Court had been filed. See S. Rpt. No. 109-297, at 7 ("This allowed claimants to have the benefit of legal counsel when pursuing their cases before the [Court], but did not permit claimants to hire attorneys until after VA had completed its administrative proceedings.") In 2006, however, Congress changed this situation by permitting attorneys to begin charging fees at the time a Notice of Disagreement is filed. See supra note 2. This amendment to section 5904, endorsed by the Senate Committee on Veterans' Affairs, reflected a consensus among veterans service organizations, VA officials, academics, and other commentators that the then- current prohibition on attorney fees for representation before the Board reflected an overly paternalized restriction on the rights of veterans and other VA claimants to choose legal counsel. See S. Rpt. No. 109-297, at 8-20; see also In re Mason, 12 Vet.App. 135, 137 (1999) (Nebeker, C.J., concurring) (criticizing the restriction, to the extent that it was based on fear of attorney misconduct, as "an unfounded indictment based on mistrust"). Rather than continuing to bar attorneys from representing claimants before the Board, Congress thought it better to allow claimants to decide for themselves whether to pay a fee to retain a lawyer. The better course, it determined, was to authorize the Secretary to prescribe minimum standards for attorney

9 Available at https://congress.gov/109/crpt/srpt297/CRPT-109srpt297.pdf.

11 experience, training, and ethics and to formulate reasonable restrictions on charges to clients. See generally § 101(a), (b), (e)-(i), 120 Stat. 3405-09. Carpenter reflects the sort of excessive concern that Congress moved away from through its 2006 amendments. As already noted, Congress has enacted legislation that will permit attorney involvement at an even earlier point in the claims process. See Pub. L. No. 115-55, § 2(n), 131 Stat. 1105, 1110. To the extent that the Court in Carpenter may have thought its interpretation of "same work" was required by a paternalistic legal obligation "to protect the veteran" from predatory fee agreements, 15 Vet.App. at 75, such considerations no longer obtain with the same force. 3. Another circumstance when precedent should give way is when "experience has pointed up the precedent's shortcomings." Pearson v. Callahan, 555 U.S. 223, 233 (2009). Carpenter's offset rule has had unintended consequences. As amicus curiae observes, that case "caused an immediate and dramatic decline in attorney representation at the agency level." Supp. Memorandum of NOVA at 11. During the fiscal year when Carpenter was handed down, a claimant was represented by an attorney in 10% of Board decisions. REPORT OF THE CHAIRMAN, BOARD OF VETERANS' APPEALS, FISCAL YEAR 2002, at 14 (2003). The attorney representation rate steadily declined to 2%. REPORT OF THE CHAIRMAN, BOARD OF VETERANS' APPEALS, FISCAL YEAR 2007, at 20 (2008). It recovered only in the year after Congress's 2006 amendments to section 5904, which allowed lawyers to charge a fee for representation at the Board prior to a remand from the Court. REPORT OF THE CHAIRMAN, BOARD OF VETERANS' APPEALS, FISCAL YEAR 2008, at 23 (2009) (7.9% attorney-representation rate).10 Of course, these statistics don't necessarily establish a causative link between Carpenter and declining attorney-representation rates, but the strong correlation cannot be ignored. In its brief, amicus curiae NOVA also argues that, because of Carpenter's offset rule, "it is difficult for an attorney to build an economically sustainable practice around continuing to represent most veterans after a remand from" the Court. Supp. Memorandum of NOVA at 10. To support this argument, it attempts to show that the average EAJA award granted by the Court will usually require a complete offset of the average contingency fee received based on an award of past-due benefits by VA. We don't have enough evidence before us to confirm whether this is true.

10 These reports can be accessed at https://www.bva.va.gov/Chairman_Annual_Rpts.asp.

12 What is difficult to deny, however, is that Carpenter's rule creates a disincentive against continued representation. An attorney who receives a contingency fee award from VA after having received an EAJA award following a remand from this Court must offset the two awards against each other. But if one attorney obtains an EAJA award at the Court, and a different attorney takes up the representation before VA and obtains a contingency fee award there, neither attorney would be required to offset because each received only one award. Cf. § 506(c). Carpenter, we conclude, has introduced needless uncertainty and complication to the attorney-client relationship in the VA benefits context. The consequences and potential harms support overruling that erroneous decision. 4. Finally, we discern no reliance interests counseling against overturning Carpenter. The reliance interests typically attendant in contract cases mean that stare decisis considerations should generally be at their highest there. See Payne v. Tennessee, 501 U.S. 808, 828 (1991). But fee agreements in the VA benefits context are not typical in that regard. Whatever expectations Carpenter may have settled for attorneys, we cannot see a viable argument that its EAJA offset rule has come to be relied upon by claimants. Certainly, none of the parties before us—including the Secretary—has made such an assertion. Nor do we believe that invalidating Carpenter's rule will dissuade a claimant otherwise interested in retaining an attorney from doing so.

III. CONCLUSION We conclude that the phrase "same work" in section 506(c) refers solely to work performed at the Court. Thus, the EAJA offset provision requires a refund to a claimant only when an attorney receives an EAJA award from the Court and an award of fees under section 5904 by virtue of a Court decision. An EAJA offset is not mandated against a contingent, or other, fee awarded by VA. Having erroneously come to a contrary conclusion, Carpenter is overruled in relevant part. Accordingly, we REVERSE the Board's determination that an EAJA offset was required here, VACATE the June 6, 2016, decision, and REMAND this matter for further proceedings consistent with this decision.

PIETSCH, Judge, with whom FALVEY, Judge, joins, dissenting: I join the dissent of my colleague Judge Falvey and offer a separate statement addressing the majority's misplaced reliance

13 on the "same work" language of the EAJA savings provision, applicable to the Social Security Administration's dual attorney fee structure, to support their definition of "same work" in the EAJA offset provision, applicable to the single statutory fee structure for attorneys representing veterans. Although Congress used virtually identical language in both EAJA provisions, the majority ignores the difference in the statutory fee structure of the two benefits systems, the history leading to the EAJA savings provision, and the beneficent underpinnings of the veterans' system. The majority's conclusion that the "same work" in the "analogous" Social Security savings clause confirms the plain meaning of the "same work" phrase in the EAJA offset provision is flawed. The statutory structure for awarding fees for attorneys representing social security claimants at the administrative level and before the court is distinct because it is set out in separate statutory subsections. Congress authorized the payment of attorney fees under a fee agreement capped at the lesser of 25% of past-due benefits or $6,000 for representation of claimants before the Commissioner of Social Security. 42 U.S.C. § 406(a); Culbertson v. Berryhill, 139 S. Ct. 517, 520 (2019) (noting the current fee cap amount as $6,000). If no fee agreement is in place, the agency may prescribe the maximum fees that may be charged for the services performed with any claim. Id. In 1965, Congress amended the Social Security statute, adding a separate statutory section that allowed attorneys representing Social Security claimants to collect fees for their work before a court. 42 U.S.C. § 406(b)(1). Congress limited the fees for work before the court to no more than 25% of past-due benefits awarded to a claimant. To ensure that attorneys were paid for their work and at the same time to ensure that attorneys did not take more money than authorized by sections 406(a) and (b), Congress enacted a provision that made it a criminal act, punishable by a fine of $500, imprisonment for a year, or both, for attorneys who "collected or even demanded from the client anything more than the authorized allocation of past-due benefits. . . " Gisbrecht v. Barnhart, 535 U.S. 789, 797 (2002) (citing 42 U.S.C. §§ 406(a)(5), (b)(2)). In 1980, Congress passed EAJA and afforded another avenue for attorneys representing Social Security claimants in court to recover fees. To allow attorneys to receive attorney fees under EAJA, without violating section 406(b)(2), Congress amended the EAJA law to include a savings clause in 1985. See Astrue v. Ratliff, 560 U.S. 586, 595 (2010); Parrish v. Comm'r of Soc. Sec. Admin., 698 F.3d 1215, 1218 (9th Cir. 2012). The savings clause allowed an attorney to seek fees under both section 406(b) and EAJA, but the attorney could not be paid twice for the same work.

14 If the court awarded attorney fees under section 406(b) and EAJA, the savings clause required the attorney to refund the amount of the smaller fee to the claimant. See Pub. L. No. 99-80, § 3, 99 Stat. 183, 186 (1985). "[A]n EAJA award offsets an award under Section 406(b), so that the [amount of the total past-due benefits the claimant actually receives] will be increased by the . . . EAJA award up to the point the claimant receives 100 percent of the past-due benefits." Gisbrecht, 535 U.S. at 796 (alterations in original; internal quotations omitted). The history leading to the EAJA savings clause reveals that Congress wanted to maximize the past-due benefits received by claimants. The statutory attorney fee scheme in veterans law does not and has never distinguished between work before the agency and work before the Court. See 38 U.S.C. § 5904. This Court possesses no statutory authority to award attorney fees, except under EAJA. In contrast, a district court may compensate the attorney representing a Social Security claimant under either EAJA or section 406(b)(1). And, unlike the Social Security administrative system, there is no monetary cap on attorney fees awarded at the VA administrative level. Moreover, at our Court, the attorney- veteran direct fee representation agreement is presumed reasonable if the fee is no more than 20% of past-due benefits. 38 U.S.C. § 5904(d); In re Fee Agreement of Vernon, 8 Vet.App. 457, 459 (1996) (holding that a fee agreement that complies with section 5904(d) is reasonable absent evidence to the contrary). Fee agreements, other than direct fee representation agreements, must simply be reasonable. 38 U.S.C. § 5904(c). Given the completely different statutory attorney fee structure between Social Security law and veterans law, it is not reasonable to assume that Congress intended the definition of "same work" in the EAJA savings provision to have the same meaning in the EAJA offset provision. Rather, it is more likely that it was the context of maximizing the award of past-due benefits to claimants behind the EAJA savings provision that was the soil transplanted to the EAJA offset provision. As Justice Frankfurter explained, "the significance of an enactment, its antecedents as well as its later history, its relation to other enactments, all may be relevant to the construction of words for one purpose and in one setting but not for another. Some words are confined to their history." Some Reflections on the Reading of Statutes, 47 COLUM. L. REV. 527, 537 (1947). Moreover, our country's relationship with its veterans is unique. Congress recognized the sacrifices veterans make for their country and "created a paternalistic veterans' benefits system to care for those who served their country in uniform." Jaquay v. Principi, 304 F.3d 1276, 1280 (Fed. Cir. 2002). The entire veterans benefits scheme "is thus imbued with special beneficence from a

15 grateful nation." Sneed v. Shinseki, 737 F.3d 719, 728 (Fed. Cir. 2013) (quoting Bailey v. West, 160 F.3d 1360, 1370 (Fed. Cir. 1998) (Michel, J., concurring). In Hodge v. West, 155 F.3d 1356, 1362 (Fed. Cir. 1998), the Federal Circuit recognized the pro-claimant environment created by the general VA statutory scheme and determined that the application of concepts involved in the Social Security adjudication process–a system not designed to be "strongly and uniquely pro-claimant"– was inappropriate. Here, the majority allows an attorney to collect both an EAJA fee from the Court and 20% of a veteran's past-due benefits under a direct fee agreement at the agency level, where there is no monetary cap. The attorney collects two sets of legal fees for successfully representing a veteran in the pursuit of a single claim for TDIU benefits. Given the veterans' benefits scheme Congress created, including the singular statutory fee structure, it is difficult to understand the majority's interpretation of the "same work" in the EAJA offset provision. In the end, the majority's decision reallocates a portion of a grateful nation's payments to veterans and awards them to attorneys. Therefore, I respectfully dissent.

FALVEY, Judge, with whom PIETSCH, Judge, joins, dissenting: I respectfully dissent. Rather than disregard stare decisis based on a false premise of a negative impact of our Carpenter decision, I would allow Congress to fix any purported error we may have made while ensuring that neither veterans nor the fisc becomes a trough to be raided by counsel in 6-minute increments. Today the majority tosses aside one flawed statutory interpretation, only to replace it with their own, equally flawed interpretation. The shift from one extreme—requiring a full offset of all EAJA fees—to the other—pretending that reviewing the same record is not the "same work" because it is at the agency—has the consequence of taking money out of veterans' pockets. A reading of a provision requiring an offset for the "same work" to allow an attorney to earn twice for work done only before the Court is no more faithful to the text of the statute than a reading requiring a full offset of all funds. A. The majority elects to disregard an almost 20-year old precedential decision of this Court without providing a special justification for doing so. See Dickerson v. United States, 530 U.S. 428, 443 (2012). Stare decisis—"the idea that today's Court should stand by yesterday's

16 decisions"—is "a foundation stone of the rule of law." Kimble v. Marvel Entertainment, 135 S. Ct. 2401, 2409 (2015) (quoting Michigan v. Bay Mills Indian Community, 572 U.S.782, 798 (2014)). When examining Carpenter, the Court should consider whether (1) the rule established has proven unworkable; (2) the rule is subject to reliance interests, which are particularly strong in property or contract matters; (3) subsequent legal developments have eroded the rule over time; or (4) Congress has had an opportunity to weigh in on the matter. See Kimble, 135 S. Ct. at 2409-2412. The majority does not state that the Carpenter EAJA offset rule defies workability or that the law has developed so as erode the scope and effect of the rule. Nor do they suggest that Congress has weighed in on the matter, despite having revised when attorneys may charge veterans fees. Rather, the majority appears to depart from precedent because they conclude that the reasoning of Carpenter is faulty; however, the majority does not point to how the purported faulty reasoning in Carpenter has led to erroneous results, see Arizona v. Gant, 556 U.S. 332, 352 (2009) (Scalia, J., concurring) (abandoning prior precedent when the precedent was badly reasoned and produced erroneous results). Following stare decisis means "sticking to some wrong decisions" and not "scrapping settled precedent" simply because the current Court would decide a case differently than the previous Court. Kimble, 135 S. Ct. at 2409. When the settled precedent involves the interpretation of a statute, as it does here, any contentions as to the faulty reasoning or error of the previous Court can be corrected by Congress. See generally, Patterson v. McLean Credit Union, 491 U.S. 164, 171-73 (1989). The legislature's ability to correct statutory interpretation errors is precisely why I would hold that stare decisis counsels that we leave Carpenter in place. "'Stare decisis is usually the wise policy, because in most matters it is more important that the applicable rule of law be settled than that it be settled right.'" Payne v. Tennessee, 501 U.S. 808, 827 (1991) (quoting Burnet v. Coronado Oil & Gas Co., 285 U.S. 393, 406 (1932) (Brandeis J., dissenting)). And in statutory interpretation cases, stare decisis has "special force" because "Congress remains free to alter what we have done." Patterson, 491 U.S. at 172-73. Although, five years after the Carpenter decision, Congress amended section 5904, allowing veterans to hire attorneys to represent them at an earlier point in the claims process, Congress did not amend the unitary statutory attorney fee structure in section 5904 or enact legislation to abrogate the Carpenter decision. See Johnson v. First Nat'l Bank of Montevideo, Minn., 719 F.2d 270, 277 (8th Cir. 1983) (stating that Congress acts with knowledge of existing law, and that "absent a clear manifestation of contrary intent, a newly-enacted or revised

17 statute is presumed to be harmonious with existing law and its judicial construction"); see also Midlantic Nat'l Bank v. N.J. Dep't of Envtl. Prot., 474 U.S. 494, 501(1986) ("The normal rule of statutory construction is that if Congress intends for legislation to change the interpretation of a judicially created concept, it makes that intent specific."). As the majority observes, Congress has legislated in the area of attorney fees in veteran cases. See ante at 2, note 2. And Congress has not been shy about telling us when we erred in our interpretation of the EAJA statute. See Pentecost v. Principi, 17 Vet.App. 257, 259-60 (2003) (Congress enacted the Veterans Benefits Act of 2002 to override our decision and clarify that non- attorney practitioners could collect EAJA); Cook v. Brown, 6 Vet.App. 226, 235, 237 (1994) (noting that Congress passed the FCAA which includes the "same work" language at issue because the Court held that it lacked the power to award EAJA fees). And "[c]onsiderations in favor of stare decisis are at their acme in cases involving property and contract rights, where reliance interests are involved." Payne, 501 U.S. at 828. The majority "discern[s] no reliance interests counseling against overturning Carpenter" and "cannot see a viable argument that [our] EAJA offset rule has come to be relied upon by claimants." Ante at 13. To the contrary, this case implicates agreements that assign rights and duties between veterans and their attorneys. We should not be so quick to think that veterans are unaware of their legal rights when entering representation agreements. It is not difficult to find practitioners who entice veterans by trumpeting the offset requirement on their website.11 Even so, the Court summarily takes money out of veterans' pockets and hands it to their attorneys. I cannot agree with the majority's blindness towards veterans' reliance interests. Similarly, I cannot agree with the idea that stare decisis should not apply because Carpenter led to an "immediate and dramatic decline in attorney representation at the agency level." Ante at

11 See Gang & Associates, LLC, What About Attorney Fees for Representation at CAVC?, VETERANS DISABILITY INFO, https://www.veterans disabilityinfo.com /what-about-attorney-fees-representation-cavc.php (last visited Feb. 25, 2019) ("[I]f an attorney represents a veteran at the Board or Regional Office level as well, and attorney fees were collected under the EAJA, then there is usually an offset of the amount already paid by the EAJA."); Attig | Steel, Is it a Good Strategy to Use a VA Benefits Lawyer at the Veterans Court?, VETERANS LAW BLOG, https://www.veterans lawblog.org/va-benefits-lawyer/ last visited Feb 25., 2019) ("[T]he veterans disability attorney has to reimburse the lower of the 2 fees to the Veteran under the EAJA offset rule (so long as the work performed at the CAVC and before the VARO/BVA are on the same claim"); Goodman Allen Donnelly, Three Major Misunderstandings That Keep Veterans From Getting Quality Representation Before the VA, THE VETERAN & MILITARY LAW GROUP, http://veteranmilitarylaw.com/three-major-misunderstandings-keep-veterans-getting-quality- representation-va/ last visited Feb 25, 2019) ("Other rules require an attorney who is paid for representing you at the VA to offset any fees you owe by the EAJA fees paid for the same work at the Court.").

18 12. This conclusion is wrong for at least two reasons. First, in support of their alarmist pronouncement, the majority pulls data from the Board Chairman's annual report showing the percentage of matters at the Board handled by attorneys. Id. But, looking only at the percentage of matters, rather than the actual number of matters handled by attorneys, can be misleading. This is because the percentage reflects a veteran's representational choice rather than the availability of attorneys practicing before the Board. To better understand attorney representation before the Board, it helps to look at a bigger picture of how many cases were handled by attorneys over time and how those attorneys performed. The two graphs below summarize data compiled from the Chairman's Reports from 1992 through 2017, the same reports cited by the majority, ante at 12, and which are available for review at https://www.bva.va.gov/Chairman_Annual_Rpts.asp. Graph 1 details the number of Board cases from 1992 to 2017 in which an attorney represented the veteran. Examining the number of cases handled by attorneys before our Carpenter decision in 2001, we can see that attorney representation before the Board was consistently growing. But in 2002, attorney representation dropped by nearly a thousand cases. Of course, we cannot blame this on Carpenter, because in 2002 the Board already had a multi-year delay. Thus, cases decided in 2002 were filed several years before Carpenter. And the Board would not adjudicate cases filed after Carpenter for many years to come. See James D. Ridgway & David S. Ames, Misunderstanding Chenery and the Problem of Reasons-or-Bases Review, 68 SYRACUSE L. REV. 303,328 (2018) (showing average pending time for a Board appeal of over 1,000 days in 2002). Moreover, the number of cases handled by attorneys increased in 2003 and 2004, showing that any purported negative impact from Carpenter was quickly resolved. Although there appears to have been another decline in attorney representation beginning with 2005 and ending in 2007— more likely something related to Carpenter—this decline also appears to have quickly abated with a then-record number of attorneys representing veterans in 2008. If there was any negative impact from Carpenter, attorneys and veterans quickly adjusted. Despite the grim picture painted in the majority decision, more and more veterans have been able to find an attorney able and willing to represent them at the Board. In 2018, despite any impact from Carpenter, cases at the Board with attorney representation exceeded 15,000. See REPORT OF THE CHAIRMAN, BOARD OF VETERANS' APPEALS, FISCAL YEAR 2018, at 31 (2019). Because

19 attorney representation at the Board continues to grow rapidly, I see no reason to throw away our precedent.

Graph 1. Number of Veterans Represented by Attorneys 9000 8,056 8000 7,451 BVA Cases Handled by Attorneys

7,421 7000 6,036 6000

5000 4,286 4,354 3,289 3,467 4000 2,808 4,048 2636 2,570 2131 4,296 3000 2,241 1651 3,759 2000 1160 763 861 1718 1000 1914 806 786 873 1366 0 2004 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Year

Second, there is another explanation for why veterans may not have sought attorney representation at the Board—they may have favored VSO representation. To illustrate, from 1992 to 2017, Graph 2 details the "Allowance" percentage—the percentage of cases in which the Board granted benefits—for matters handled by Disabled American Veterans (DAV) and by attorneys. As the graph reflects, until 2011, DAV was more successful on average than attorneys. This difference in success rate may better explain the decrease in the percentage of cases handled by attorneys that so worries the majority. Perhaps more veterans selected VSO representation, even though attorneys were able and willing to represent them. When attorneys began to outperform one of the largest VSOs, the number of cases handled by attorneys skyrocketed. Compare Graph 1, with Graph 2. In my view, it is not unreasonable to conclude that veterans made informed decisions not to give up 20% of their past-due benefits to attorneys who were less successful than VSOs.

20 Graph 2. Allowance of Benefits by Representation 45.00% Percentage of Cases Where Claim is Allowed/Granted

40.00%

35.00%

30.00%

25.00%

20.00%

15.00% 2004

2011 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

2005 2006 2007 2008 2009 2010

2012 2013 2014 2015 2016 2017 Year

Attorney Allowance Percentage DAV Allowance Percentage

Even if we assume that Carpenter may have had some impact on attorney behavior, it is misguided to think that this requires overruling our en banc decision. This appeal concerns an offset provision governing section 5904(d) direct-pay fee agreements. We have recognized that there is nothing stopping an attorney from having a representation agreement outside section 5904(d) where VA must collect the fee on behalf of the attorney. See Gumpenberger v. Wilkie, __Vet.App.__, 2019 U.S. App. Vet. Claims LEXIS 184, at *19 (Feb. 7, 2019) ("It is only when an agent or attorney desires VA assistance in enforcing a fee agreement via withholding that he or she is limited to entering into a direct-pay fee agreement with a client."); see also Ravin v. Wilkie, 30 Vet.App. 310, 314 (2018). Presumably, if a veteran wanted to keep the attorney who represented her or him at the Court, but the attorney did not want to give up the EAJA offset, the parties could execute an agreement for agency representation where the veteran pays the attorney 20% of past due benefits plus the amount of EAJA offset. I will not join the majority in rejecting an en banc decision simply to make VA a more convenient bill-collector for the private bar.

21 B. Likewise, I will not join the majority in judicially amending a statute to make it less favorable to veterans. The majority's new rule is unfavorable because the majority holds that "'same work' in section 506(c) refers solely to work performed at the Court. Thus, the EAJA offset provision requires a refund to a claimant only when an attorney receives an EAJA award from the Court and an award of fees under section 5904 by virtue of a Court decision." Ante at 13. Though I agree with my colleagues' focus on the text of the statute, I part ways with them when they arrive at an interpretation of an offset for "same work" that allows an attorney to collect twice for work only done once. At issue is the offset provision for fee agreements under section 5904. This provision requires that "where the claimant's attorney receives fees for the same work under both section 5904 of title 38, United States Code, and section 2412(d) of title 28, United States Code, the claimant's attorney refunds to the claimant the amount of the smaller fee." § 506(c), 106 Stat. at 4513 (emphasis added). Inexplicably, the majority reads this phrase to focus not on the work done by the attorney but on where this work is done. See ante at 6, 13 (requiring an offset only when work is done before this Court). The majority properly begins with the text by addressing the meaning of "same" in the statutory phrase "same work" at the time of enactment. This definition shows that at the time of enactment "same" meant "identical"; "similar in kind, quality, quantity, or degree"; "conforming in every detail." AMERICAN HERITAGE DICTIONARY, SECOND COLLEGE EDITION 1088 (1982); BLACK'S LAW DICTIONARY 1340 (6th ed. 1990). But under this meaning of the word "same," one would expect that when an attorney performs legal research or reviews files as part of representation before the Court, that attorney would not claim that reviewing those same files or doing identical or similar research for representation before the agency is not the same work. After all, work such as reading the same document many times is as close to an example of "same work" as we are likely to see. Even so, the majority does very little with their definition of "same" and even less so with the definition of "work." After plucking the definition of work out of the dictionary, the majority argues that "if Congress wished to require an offset that covered all representation during a single claim or case, it could have said so explicitly, but it did not; we must presume this choice was deliberate." Ante at 6. Of course, Congress could have required an offset based only on the tribunal

22 where the work is performed, but Congress did not do so. Perhaps this choice too was deliberate. Undeterred by this possibility, the majority simply concludes that "'same work' does not cover work done before separate tribunals such as a court and an agency." Ante at 6. Yet the majority ignores that about half of the work that Mr. Ravin and the non-attorney practitioner did to receive EAJA fees in the Easterling case consisted of reviewing the veteran's file, reviewing the Court's remand, and researching legal theories related to TDIU—the same issue presented to the Board on remand. See Easterling v. Shinseki, 09-2797, Appellant's Petition for Attorney Fees and Expenses (U.S. Vet.App. August 23, 2011). How is it that reviewing the same record and researching the same legal issue does not constitute the "same," "identical," or "similar in kind" work? The majority does not say. Instead, they say that Mr. Ravin can keep the entirety of his EAJA award and the 20% of past due benefits for Board representation after the Court's remand. Researching the legal issue in dispute and reviewing the file are the kinds of tasks that any attorney representing a client before an adjudicative body is expected to perform. So an attorney who successfully represents a client before the Board and is paid 20% of past-due benefits gets paid for work that necessarily includes reviewing the veteran's file and legal research. Likewise, an attorney who secures a remand from our Court is paid under EAJA for work that includes reviewing the veteran's file and researching the claims. We can see that there is more "same work" being done than the majority would have you believe when they require an offset only when an attorney obtains an award of benefits because of a Court decision. Congress required an offset for "same work." But under the majority's rule, rather than offsetting for the same work, attorneys are guaranteed to be paid twice for it. Worse still, the majority's rule will allow attorneys who continue to represent the veteran after a Court remand to receive a fee of 20% of the veteran's past-due benefits no matter how little work they perform to represent the client before the agency. The attorney need only enter an appearance after a remand from this Court without doing any work to retain both the full EAJA fees and be engorged by 20% of the veteran's past-due benefits. Here, nearly half of the EAJA award for representation at the Court is compensation for document review and research that any attorney would have done to represent the veteran before the Board. And, under the majority's rule, Mr. Ravin gets paid for reviewing the record and researching TDIU before the Court and then gets paid for performing the very same work before the Board. Rather than considering what specific work done before the Board and before the Court

23 constitutes the "same work" the Court doubles down on a purported mistake from Carpenter by again picking an atextual one-size-fits-all solution, albeit this time one that redistributes money from veterans to attorneys. See Carpenter, 15 Vet.App. at 81 (Steinberg J., concurring in part and dissenting in part) (expressing concern that the majority opinion in Carpenter is "more an exercise in policymaking, that is, in judicial legislation, than judicial review of policies established by the popularly-elected and therefore publicly accountable legislative and executive branches. I reach this conclusion because the majority appears to be intent, . . . on reaching a one-size-fits-all holding as to EAJA offset."). The majority seems to ignore those portions of the attorney fee statutes that focus on the "reasonableness" of the fee. See, e.g., 38 U.S.C. § 7263(d) (allowing the Court to order a reduction of a section 5904 fee award if it is "unreasonable”); see also 38 U.S.C. § 5904(c)(3)(A) (permitting the Secretary to reduce a fee if it is excessive or unreasonable). But it is a basic principle of statutory interpretation that "a section of a statute should not be read in isolation . . . and that in fulfilling our responsibility in interpreting legislation, we must not be guided by a single sentence or member of a sentence, but [should] look to the provisions of the whole law." Ramsey v. Nicholson, 20 Vet.App. 16, 34 (2006) (quoting Moreau v. Brown, 9 Vet.App. 389, 396 (1996)). "Additionally, the Court is mindful that statutes must be considered as a whole and in the context of the surrounding statutory scheme." Gazelle v. McDonald, 27 Vet.App. 461, 464 (2016), aff'd sub nom. Gazelle v. Shulkin, 868 F.3d 1006 (Fed. Cir. 2017). A proper reading of the offset provision, alongside those provisions in which Congress directed this Court and VA to ensure that the fee charged to a veteran is reasonable, should lead to an examination of the work done by the attorney to determine whether the fees are reasonable. If there is an overlap of work performed by an attorney at each separate venue, a proper interpretation of the offset provision would be to reduce the attorney fees accordingly. Yet, after today's decision, an attorney who obtains a remand from this Court and then signs up for the ride down to the agency will be paid twice no matter what additional work, if any, is done at the agency. This interpretation effectively eviscerates the "same work" language in the EAJA offset provision by limiting the offset provision to work performed before the Court. Unfortunately, the notion of reasonableness is absent in the majority's decision. "[A] reasonable fee 'means not the reasonableness of the agreed-upon contingent fee, but a reasonable recompense for the work actually done.'" Gisbrecht v. Barnhart, 535 U.S. 789, 811 (2002) (Scalia J., dissenting). Determining if a fee is reasonable or

24 if work done before the agency and the court is the "same" requires reviewing the charged fee and the work done by the attorney. Comparing the work done at the agency and the Court against the results achieved would allow the Court (or the agency) to confirm if the charged fee is reasonable and discount fees for work that is identical or similar in kind—work such as record review or duplicative research—i.e., the same work. Here, even a cursory review of Mr. Ravin's work done before the Board after the Court remand should give us pause. First, he received $5,787 in EAJA fees for obtaining a reasons-or- bases remand at the Court. Then before the agency, Mr. Ravin got $17,239.40, which represented 20% of the veteran's past-due benefits. ($23,126.40 less $5,787 to offset the EAJA fees, and less a $100 administrative fee charged by VA under a section 5904(d) representation agreement). And now he wants $5,787, that would otherwise be paid to the veteran. At oral argument, it was difficult for Mr. Ravin to articulate what work he performed at the Board to justify his fee. The Secretary's supplemental submission reflected that Mr. Ravin submitted a three-page letter that informed VA of his new office address, summarized the Court's remand, and regurgitated an argument from his brief submitted to the Court. This letter also included the same block quote of a medical opinion submitted in his earlier brief (and quoted in the Court's single-judge decision), which he presumably copied and pasted into the "new" submission to the Board. Compare Supplemental Record of Proceedings (ROP), with Easterling, 09-2797, Appellant's Brief at 2-6. At most, Mr. Ravin's letter to the Board contained three and a half lines of new words not submitted before to VA—about 31 original words. Compare R. at 233- 34 in the supplemental ROP, with R. at 283-86 (Court's May 4, 2011, Memorandum Decision remanding Mr. Easterling's claim for TDIU). Thus, the Court's new interpretation blesses payment of more than $23,000 for fewer than three full pages of text. This is particularly troubling because when Mr. Ravin wrote those three pages, he had already been paid $5,787 under EAJA for reviewing the entire record, researching the relevant legal issues, and drafting most of his argument as part of work before the Court. He already got paid for typing the recycled block quote. So the Court not only allows a payment of almost $8,000 per page, or about $742.78 per word, but does so while pretending that most of those pages are not work that is the same or similar in kind, quality, quantity, or degree, as work done before this Court. I do not call for a fresh review of the reasonableness of the fee for Board

25 representation.12 The Board did not make findings on this issue. But interpreting "same work" within the VA attorney fee statutory scheme should not lead to a definition that results in an unreasonable fee. See Gazelle, 27 Vet.App. at 464. Even without considering whether the fee is unreasonable, the majority's statutory interpretation is plainly misguided. The majority creates a rule that omits any consideration of whether the "work" compensated is the "same". Instead, we are left with a new, Court-created rule that would demand section 506(c) to require an offset "where the claimant's attorney receives fees for [work done at the same venue] under both section 5904 of title 38, United States Code, and section 2412(d) of title 28." "Courts are not authorized to rewrite a statute because they might deem its effects susceptible of improvements." Badaracco v. Commissioner, 464 U.S. 386, 398 (1984). "If a statute needs repair, there's a constitutionally prescribed way to do it. It's called legislation." Perry v. Merit Sys. Prot. Bd., 137 S. Ct. 1975, 1990 (2017) (Gorsuch J., dissenting). If the statute needs changing to limit the offset to one venue, the majority should let Congress do its job. Our amicus here tells its dues-paying attorney members that it "is active on a variety of fronts in Washington, DC, protecting [their] interests." Why Join NOVA?, NOVA (2018), https://www.vetadvocates.org/cpages/why-join-nova (last visited Feb. 25, 2019). And that it "regularly monitors pending legislation and regulatory changes" and regularly meet[s] with VA officials and [has] been successful in addressing members' issues in dealing with" VA. Id. It appears that attorneys would be fine without my colleagues' judicial amendment. "Besides, the law of unintended consequences being what it is, judicial tinkering with legislation is sure only to invite trouble." Perry, 137 S. Ct. at 1990 (Gorsuch J., dissenting). And here, this tinkering is particularly unwise as it would impair the Secretary's independent review of fee awards for reasonableness or to determine when the attorney would be paid for the same work, as the Court settles the issue once and for all. It erects an artificial barrier between work before the Court and the agency. Even if the Secretary sought to prevent unjust enrichment of an attorney who only copied their Court pleadings or worse yet filed nothing, our decision today ensures that such an attorney leaves richer and the veteran poorer.

12 The Board did not address which prong of § 5904(c)(3)(A) it was affirming the fee reduction under— "excessive" or "unreasonable." Instead, the Board broadly cited to section 5904 and primarily relied on Carpenter. But our decision there turned on finding that "a fee agreement allowing an attorney to collect and retain both an EAJA fee as well as a fee from the client for the same work is 'unreasonable' pursuant to 38 U.S.C. §§ 5904 and 7263." Carpenter, 15 Vet.App. at 73. Thus, the issue of reasonableness is before us; however, with limited discussion from the Board, I see no reason to address the issue beyond how it informs the proper reading of "same work."

26 Now, it is true that there may be good reason for not devoting more adjudicative resources to reviewing attorney billing. It is also true that we as a Court are poorly equipped to make policy determinations that juggle VA resources used on reviewing fees on top of a significant claim backlog. See Walters v. Nat'l Ass'n of Radiation Survivors, 473 U.S. 305, 331, n. 12 (1985). ("Congress is an institution better equipped to amass and evaluate the vast amounts of data bearing on such an issue.") But that is precisely why I would leave this issue to Congress to address, if a problem exists. C. In conclusion, I would affirm the Board's decision. I would do so because, despite ample opportunity to correct any perceived flaws in our Carpenter decision, Congress has let the decision stand. In the 18 years since Carpenter, the number of cases handled by attorneys before the agency has risen to an all-time high. If our Carpenter decision had such a negative impact, all parties quickly shrugged it off. Even so, the majority adopts a new rule that refuses to look at the work performed when considering what constitutes the "same work" and so potentially unjustly enriches attorneys at the expense of veterans. Accordingly, I respectfully dissent.