969 F.3d 1333

Fed. Cir.

2020

JAMES L. KISOR, Claimant-Appellant v. ROBERT WILKIE, SECRETARY OF VETERANS AFFAIRS, Respondent-Appellee

2016-1929

United States Court of Appeals for the Federal Circuit

August 12, 2020

Appeal from the United States Court of Appeals for Veterans Claims in No. 14-2811, Judge Alan G. Lance, Sr.

KENNETH M. CARPENTER, Law Offices of Carpenter Chartered, Topeka, KS, argued for claimant-appellant.

IGOR HELMAN, Commercial Litigation Branch, Civil Division, United States Department of Justice, Washington, DC, argued for respondent-appellee. Also represented by ETHAN P. DAVIS, ROBERT EDWARD KIRSCHMAN, JR., MARTIN F. HOCKEY, JR.; Y. KEN LEE, SAMANTHA ANN SYVERSON, Office of General Counsel, United States Department of Veterans Affairs, Washington, DC.

Before REYNA, SCHALL, and WALLACH, Circuit Judges.

Opinion for the court filed by Circuit Judge SCHALL.

Dissenting opinion filed by Circuit Judge REYNA.

SCHALL, Circuit Judge.

INTRODUCTION AND DECISION

In Kisor v. Shulkin, 869 F.3d 1360 (Fed. Cir. 2017) (”Kisor I“), we affirmed the decision of the United States Court of Appeals for Veterans Claims (“Veterans Court“) in Kisor v. McDonald, No. 14-2811, 2016 WL 337517 (Vet. App. Jan 27, 2016) (“Veterans Court Decision“). In that decision, the Veterans Court affirmed the April 29, 2014 decision of the Board of Veterans’ Appeals (“Board“) that denied Mr. Kisor an effective date earlier than June 5, 2006, for the grant of service connection for his post-traumatic stress disorder (“PTSD“). Id. at *1.

In its decision, the Board held that Mr. Kisor was not entitled to an earlier effective date under 38 C.F.R. § 3.156(c)(1). J.A. 78-91. That regulation states that the Department of Veterans Affairs (“VA“) will reconsider a claim after a final decision if it receives “relevant official service department records that existed and had not been associated with the claims file when VA first decided the claim.” 38 C.F.R. § 3.156(c)(1). The regulation further states that “[a]n award made based all or in part on the records identified by [§ 3.156(c)(1)] is effective on the date entitlement arose or the date VA received the previously decided claim, whichever is later.” Id. § 3.156(c)(3).

In Mr. Kisor‘s case, the Board concluded that two service department records, which were received in 2006 and 2007, were not “relevant” under the regulation because they did not pertain to the basis of the 1983 denial of Mr. Kisor‘s claim, which was the lack of a diagnosis of PTSD. J.A. 85, 89, 90. Rather, they pertained to whether Mr. Kisor was in combat in “Operation Harvest Moon,” a military operation in Vietnam in 1965. In that regard, when it denied Mr. Kisor‘s claim, the VA Regional Office (“RO“) had before it a VA psychiatric examiner‘s report that recited Mr. Kisor‘s account of his participation in Operation Harvest Moon, see J.A. 19-20, and the RO did not dispute that account. The Board reasoned that the documents would not have changed the “outcome” of the VA‘s 1983 decision, which was based on the lack of “a diagnosis of PTSD,” because they bore on a matter relating to entitlement to service connection for PTSD that was not in dispute: the presence of an in-service stressor. Id. at 90-91. The Board thus denied Mr. Kisor an effective date earlier than June 5, 2006, for a grant of service connection for his PTSD. J.A. 91. June 5, 2006 was the date Mr. Kisor submitted a request to reopen his claim, which the VA granted. J.A. 34. Pursuant to 38 U.S.C. § 5110(a) and 38 C.F.R. § 3.400(q)-(r), as in effect in 2014, the effective date of the grant of service connection for Mr. Kisor‘s reopened claim was the date he submitted his request to reopen.

In our prior decision, we held that the Board had not erred in construing the term “relevant” as it appears in § 3.156(c)(1). In reaching that holding, we concluded that the term “relevant” was ambiguous and had more than one reasonable meaning. Kisor I, 869 F.3d at 1367-68. We therefore deferred, under Auer v. Robbins, 519 U.S. 452, 461 (1997), to the Board‘s interpretation of the term, which we found to be reasonable. Kisor I, 869 F.3d at 1367-69.

The case is now before us again on remand from the Supreme Court. See Kisor v. Wilkie, 139 S. Ct. 2400 (2019) (”Kisor II“). In Kisor II, the Court held that, in Kisor I, we were too quick to extend Auer deference to the Board‘s interpretation of “relevant” as it appears in § 3.156(c)(1). The Court therefore vacated our decision and remanded the case to us with the instruction that we decide whether Auer deference “applies to the agency interpretation at issue.” 139 S. Ct. at 2408. The Supreme Court stated that “[f]irst and foremost, a court should not afford Auer deference unless the regulation is genuinely ambiguous.” Id. at 2415. The Court directed us on remand “to determine, based on indicia like text, structure, history, and purpose, whether the regulation really has more than one reasonable meaning.” Id. at 2424.

For the reasons stated below, we now conclude that, in the setting of § 3.156(c)(1), the term “relevant” is not “genuinely ambiguous.” Id. at 2415. Accordingly, Auer deference is not appropriate in this case. In our view, in the context of § 3.156(c)(1), the term “relevant” has only “one reasonable meaning,” the meaning the Board attributed to it. As the Board determined, under the regulation, in order to be “relevant,” a record must speak to a matter in issue, in other words, a matter in dispute. We therefore once again affirm the decision of the Veterans Court that affirmed the decision of the Board denying Mr. Kisor entitlement under § 3.156(c)(1) to an effective date earlier than June 5, 2006, for his PTSD.

BACKGROUND

I.

The pertinent facts are as follows: Mr. Kisor served on active duty in the Marine Corps from 1962 to 1966. Veterans Court Decision, 2016 WL 337517, at *1. In December of 1982, he filed an initial claim for disability compensation benefits for PTSD with the VA RO in Portland, Oregon. Id. Subsequently, in connection with the claim, the RO received a February 1983 letter from David E. Collier, a counselor at the Portland Vet Center. J.A. 17. In his letter, Mr. Collier stated: “[I]nvolvement in group and individual counseling identified . . . concerns that Mr. Kisor had towards depression, suicidal thoughts, and social withdraw[a]l. This symptomatic pattern has been associated with the diagnosis of Post-Traumatic Stress Disorder.” Id.

In March of 1983, the RO obtained a psychiatric examination for Mr. Kisor. In his report, the examiner noted that Mr. Kisor had served in Vietnam. The examiner also noted that Mr. Kisor recounted that he had participated in Operation Harvest Moon; that he was on a search operation when his company came under attack; that he reported several contacts with snipers and occasional mortar rounds fired into his base of operation; and that he “was involved in one major ambush which resulted in 13 deaths in a large company.” J.A. 19-20. The examiner did not diagnose Mr. Kisor as suffering from PTSD, however. Rather, it was the examiner‘s “distinct impression” that Mr. Kisor suffered from “a personality disorder as opposed to PTSD.” Id. at 21. The examiner diagnosed Mr. Kisor with intermittent explosive disorder and atypical personality disorder. Id. Such conditions cannot be a basis for service connection. See 38 C.F.R. § 4.127. Given the lack of a current diagnosis of PTSD, the RO denied Mr. Kisor‘s claim in May of 1983. J.A. 23. The RO decision became final after Mr. Kisor initiated, but then failed to perfect, an appeal. Veterans Court Decision, 2016 WL 337517, at *1.

II.

On June 5, 2006, Mr. Kisor submitted a request to reopen his previously denied claim for service connection for PTSD. J.A. 25. While his request was pending, he presented evidence to the RO. This evidence included a July 20, 2007 report of a psychiatric evaluation diagnosing PTSD, as well as a copy of the February 1983 letter from the Portland Vet Center. See J.A. 17, 100-11. The evidence also included service personnel records that had not been before the RO in 1983. These records included a copy of Mr. Kisor‘s Department of Defense Form 214 (subsequently corrected in 2007 to note, inter alia, a Combat Action Ribbon); and a Combat History, Expeditions, and Awards Record documenting his participation in Operation Harvest Moon. See J.A. 27-29. The RO also located an additional record it did not consider in 1983: a daily log from Mr. Kisor‘s unit, the 2nd Battalion, 7th Marines. J.A. 30-31. In June of 2007, the RO made a Formal Finding of Information Required to Document the Claimed Stressor. This was based on Mr. Kisor‘s statements; on his service medical records (which verified his service in Vietnam with the 2nd Battalion, 7th Marines); and on the daily log from his battalion, which detailed the combat events Mr. Kisor had previously described in connection with his claim. J.A. 30-31. In September of 2007, a VA examiner diagnosed Mr. Kisor with PTSD. J.A. 115.

In due course, the RO issued a rating decision reopening Mr. Kisor‘s previously denied claim. The decision granted Mr. Kisor service connection for PTSD and assigned a 50 percent disability rating, effective June 5, 2006. Veterans Court Decision, 2016 WL 337517, at *1. According to the decision, the rating was based upon evidence that included the July 2007 psychiatric evaluation report diagnosing PTSD, the September 2007 VA examination, and the Formal Finding of Information Required to Document the Claimed Stressor. J.A. 32-33. The RO explained that service connection was warranted because the VA examination showed that Mr. Kisor was diagnosed with PTSD due to experiences that occurred in Vietnam and because the record showed that he was “a combat veteran (Combat Action Ribbon recipient).” J.A. 33.

In November of 2007, Mr. Kisor filed a Notice of Disagreement. In it, he challenged both the 50 percent disability rating and the effective date assigned by the RO. Veterans Court Decision, 2016 WL 337517, at *1. Subsequently, in March of 2009, the RO issued a decision increasing Mr. Kisor‘s schedular rating to 70 percent. In addition, the RO granted Mr. Kisor an extraschedular entitlement to individual unemployability, effective June 5, 2006. J.A. 41-45. In January of 2010, the RO issued a Statement of the Case denying entitlement to an earlier effective date for the grant of service connection for PTSD. See J.A. 53-65.

III.

Mr. Kisor appealed to the Board. Although not raised by Mr. Kisor, the Board considered whether the records Mr. Kisor submitted in connection with his June 5, 2006 request to reopen and the additional record located by the RO warranted reconsideration of his claim under 38 C.F.R. § 3.156(c)(1). If it did, then Mr. Kisor would be eligible for an effective date for his disability benefits of December of 1982, “the date VA received the previously decided claim.” 38 C.F.R. § 3.156(c)(3).

After reviewing the evidence, the Board denied Mr. Kisor entitlement to an effective date earlier than June 5, 2006. J.A. 91. The Board found that the VA did receive service department records documenting Mr. Kisor‘s participation in Operation Harvest Moon after the May 1983 rating decision. J.A. 89-90. As noted above, the Board concluded, though, that the records were not “relevant” for purposes of § 3.156(c)(1). J.A. 90. The Board explained that the 1983 rating decision denied service connection because there was no diagnosis of PTSD, and because service connection can be granted only if there is a current disability. Id. (citing Brammer v. Derwinski, 3 Vet. App. 223 (1992)). The Board stated that “relevant evidence, whether service department records or otherwise, received after the rating decision would suggest or better yet establish that the Veteran has PTSD as a current disability.” Id. The Board noted that Mr. Kisor‘s “service personnel records and the daily [Battalion] log skip this antecedent to address the next service connection requirement of a traumatic event during service.” Id. Finally, the Board concluded with the observation that the records at issue were not “outcome determinative” and “not relevant to the decision in May 1983 because the basis of the denial was that a diagnosis of PTSD was not warranted, not a dispute as to whether or not the Veteran engaged in combat with the enemy during service.” J.A. 90-91.

DISCUSSION

I.

As noted, this case is before us on remand from the Supreme Court. On remand, we asked the parties to provide us with their views as to how we should proceed in view of the Court‘s decision in Kisor II. In response, both Mr. Kisor and the government take the position that the term “relevant,” as it appears in 38 C.F.R. § 3.156(c), is not “genuinely ambiguous” and that therefore Auer deference is not appropriate. See Appellant‘s Suppl. Br. 4 (“In this case, the term ‘relevant’ as used by the Secretary in 38 C.F.R. § 3.156(c) is not ‘genuinely ambiguous.‘“); Appellee‘s Suppl. Br. 4 (“A thorough examination of the text, purpose, structure, and history of 38 C.F.R. § 3.156(c) demonstrates that our reading of ‘relevant’ in that subsection is the only reasonable reading of the regulation.“).

Mr. Kisor‘s view is that the only reasonable reading of the regulation is that a service department record is “relevant” if it has “any tendency to make the existence of any fact that is of consequence to the determination of the action more probable or less probable than it would be without the evidence.” Appellant‘s Suppl. Br. 9-10 (quoting Counts v. Brown, 6 Vet. App. 473, 476 (1994)). In other words, Mr. Kisor reasons that a service record is “relevant” if it constitutes evidence probative of any fact necessary to substantiate a veteran‘s claim, even if the matter to which the record speaks is not in dispute. The government‘s view is that the only reasonable reading of the term “relevant” in § 3.156(c) is that, in order to be relevant, a record must “address a dispositive issue and therefore . . . affect the outcome of the proceeding.” Appellee‘s Suppl. Br. 14. The government reasons that, in order for a record to affect the outcome of the proceeding it “must speak to the basis for the VA‘s prior decision.” Id. at 16. That was not the case here because the basis for the VA‘s prior decision was the absence of a diagnosis of PTSD, not the absence of an in-service stressor (participation in combat). Thus, while the parties both take the position that “relevant,” as it appears in the regulation, is not genuinely ambiguous, they advocate different meanings for the term.

As explained below, we too conclude that the term “relevant” in § 3.156(c) is not genuinely ambiguous. At the same time, we agree with the government that, in the context of the regulation, the term has only one reasonable meaning. To be relevant, a record must address a dispositive issue and therefore affect the outcome of the case. As the Board determined, the record must speak to a matter in issue, in other words, a matter in dispute. In this case, in 1983 the VA denied Mr. Kisor‘s claim for service connection for PTSD because he had not been diagnosed with PTSD, not because of the absence of an in-service stressor. Indeed, in this case, the presence of an in-service stressor has never been disputed. As the Supreme Court pointed out, “[t]he report of the agency‘s evaluating psychiatrist noted [Mr.] Kisor‘s involvement in . . . battle” during Operation Harvest Moon. Kisor II, 139 S. Ct. at 2409. For this reason, we again affirm the decision of the Veterans Court that affirmed the decision of the Board denying Mr. Kisor an effective date earlier than June 5, 2006, for service connection for his PTSD.

II.

Establishing service connection for a PTSD claim requires (1) a medical diagnosis of PTSD; (2) “a link, established by medical evidence, between [the] current symptoms and an in-service stressor“; and (3) “credible supporting evidence that the claimed in-service stressor occurred.” AZ v. Shinseki, 731 F.3d 1303, 1310 (Fed. Cir. 2013) (quoting 38 C.F.R. § 3.304(f)).

A veteran can seek to revise a Board denial of a claim for disability benefits through different procedures. First, Board decisions are subject to review to determine whether a clear and unmistakable error exists under 38 U.S.C. §§ 7111, 5109A, and 38 C.F.R. § 20.1400. Second, before amendments promulgated in 2019, a claimant could reopen a claim by submitting “new and material evidence” under former 38 U.S.C. § 5108 and 38 C.F.R. § 3.156. See Garcia v. Wilkie, 908 F.3d 728, 732-33 (Fed. Cir. 2018) (citing Cook v. Principi, 318 F.3d 1334, 1337 (Fed. Cir. 2002) (en banc)). As noted, in 2006, Mr. Kisor sought to reopen his claim for PTSD. Benefits awarded pursuant to a reopened claim under the former statutory and regulatory framework were granted an effective date no earlier than the date of the request for reopening. 38 U.S.C. § 5110 (2012), 38 C.F.R. § 3.400(q)-(r) (2006); see also Sears v. Principi, 349 F.3d 1326, 1331 (Fed. Cir. 2003).¹

For claims based upon “new and material evidence” filed before 2019, such as Mr. Kisor‘s, 38 C.F.R. § 3.156(a) defined “new evidence” as “existing evidence not previously submitted to agency decisionmakers.” “Material” evidence was defined under the same subsection as “existing evidence that, by itself or when considered with previous evidence of record, relates to an unestablished fact necessary to substantiate the claim.” The regulation goes on to explain that “[n]ew and material evidence can be neither cumulative nor redundant of the evidence of record at the time of the last prior final denial of the claim sought to be reopened, and must raise a reasonable possibility of substantiating the claim.” 38 C.F.R. § 3.156(a).

Third, a veteran may seek to have the VA reconsider a previously-denied claim under 38 C.F.R. § 3.156(c)(1).² Section 3.156(c)(1) reads today as it did in 2006 and in 2014 when the Board considered Mr. Kisor‘s case. As noted above, the regulation states that the VA will reconsider a claim after a final decision if it receives “relevant official service department records that existed and had not been associated with the claims file when VA first decided the claim.” 38 C.F.R. § 3.156(c)(1).³ The regulation further states that “[a]n award made based all or in part on the records identified by [§ 3.156(c)(1)] is effective on the date entitlement arose or the date VA received the previously decided claim, whichever is later.” 38 C.F.R. § 3.156(c)(3). “In other words, § 3.156(c) serves to place a veteran in the position he would have been had the VA considered the relevant service department record before the disposition of his earlier claim.” Blubaugh v. McDonald, 773 F.3d 1310, 1313 (Fed. Cir. 2014).

Unlike the “new” and “material” terms defined in § 3.156(a), § 3.156(c) does not provide a definition for the term “relevant.” However, the context of § 3.156(c) makes clear that, in order to be “relevant” for purposes of reconsideration, additional records must speak to the basis for the VA‘s prior decision. Specifically, the effective date for an award under § 3.156(c) is retroactive to the “date entitlement arose or the date VA received the previously decided claim” only if the award is “based all or in part on” the newly identified records. § 3.156(c)(3). Duplicative records and records directed to an undisputed fact would not speak to the basis for the VA‘s prior decision; a claimant filing such records thus could not obtain an award “based all or in part on” the newly identified records. In this case, the records at issue did not speak to the basis for the VA‘s prior decision: the absence of a diagnosis of PTSD.

Moreover, in the context of veteran‘s benefits, we have explained that “relevant” evidence is evidence that “must tend to prove or disprove a material fact.” AZ, 731 F.3d at 1311; see also Black‘s Law Dictionary (10th ed. 2014) (defining “relevant” as “[l]ogically connected and tending to prove or disprove a matter in issue“). Similarly, the VA‘s duty to assist claimants under 38 U.S.C. § 5103A mandates that the VA make reasonable efforts to obtain “relevant” records, but this does not encompass the situation in which “no reasonable possibility exists that such assistance would aid in substantiating the claim.” 38 U.S.C. § 5103A(a)(2); see Golz v. Shinseki, 590 F.3d 1317, 1321 (Fed. Cir. 2010) (“Relevant records for the purpose of [38 U.S.C.] § 5103A are those records that relate to the injury for which the claimant is seeking benefits and have a reasonable possibility of helping to substantiate the veteran‘s claim.“). Conversely, evidence that “simply does not tend to prove a fact that is of consequence to the action[] . . . is not relevant.” AZ, 731 F.3d at 1311 (quoting 2 Jack B. Weinstein & Margaret A. Berger, Weinstein‘s Federal Evidence § 401.07 (2d ed. 2012)).⁴

Mr. Kisor‘s original claim was denied in 1983 because he had no diagnosis of PTSD, not because of any dispute as to whether he had suffered an in-service stressor. J.A. 23. The Rating Decision acknowledges consideration of the VA psychiatric examiner‘s evaluation. Id. In the evaluation, the examiner detailed Mr. Kisor‘s recounting of his participation in Operation Harvest Moon, noting “it . . . appear[ed] that [Mr. Kisor] was involved in one major ambush which resulted in 13 deaths.” Id. at 19-20. The examiner concluded, however, that it was his “distinct impression that this man suffers from a personality disorder as opposed to PTSD.” Id. at 21. It was on this lack of a PTSD diagnosis that the Board relied when it concluded that PTSD was “not shown by evidence of record.” Id. at 23.

As noted, the additional service records at issue here are Mr. Kisor‘s service personnel records, including his Form 214, corrected to add a Combat Action Ribbon; and his Combat History, Expeditions, and Awards Record noting his participation in Operation Harvest Moon. The additional service records also include the daily log of his battalion in Vietnam that confirmed Mr. Kisor‘s description of the ambush during Operation Harvest Moon. Although they provide further support to Mr. Kisor‘s prior statements that he participated in Operation Harvest Moon and indeed could provide “credible supporting evidence that the claimed in-service stressor occurred,” see AZ, 731 F.3d at 1310, these additional service records do not show that Mr. Kisor had a medical diagnosis of PTSD as of 1983, or that the agency otherwise erred in denying his claim in 1983.

The Board‘s decision that Mr. Kisor‘s records were not “relevant” is also consistent with our holding in Blubaugh, 773 F.3d at 1314. In Blubaugh, we held that § 3.156(c) did not apply when a newly discovered service record “did not remedy the defects” of a prior decision and contained facts that “were never in question.” Id. Indeed, we held that “[s]ection 3.156(c) only applies ‘when VA receives official service department records that were unavailable at the time that VA previously decided a claim for benefits and those records lead VA to award a benefit that was not granted in the previous decision.‘” Id. (quoting New and Material Evidence, 70 Fed. Reg. at 35,388 (proposed June 20, 2005)).

We therefore conclude that the Board did not err in holding that the records cited by Mr. Kisor were not “relevant” because they did not pertain to the basis of the 1983 denial, the lack of a diagnosis of PTSD. The records added nothing to the case because they bore on a matter relating to entitlement to service connection for PTSD that was not in dispute: the presence of an in-service stressor.

III.

As noted, Mr. Kisor argues that a service department record is “relevant” under 38 C.F.R. § 3.156(c) if it has any tendency to make the existence of any fact that is of consequence to the determination of the action more probable or less probable than it would be without the evidence. This view, however, is squarely contrary to what we have just explained is the correct reading of the regulation. We therefore reject it.

Mr. Kisor makes two additional arguments. First, he contends that his reading of the regulation is supported by the fact that § 3.156(c) is intended to be remedial in nature. According to Mr. Kisor, the regulation was promulgated “to address what occurs when VA fails to obtain all relevant service department records before adjudicating [a] claim in the first instance.” Appellant‘s Suppl. Br. 11. Since the regulation is remedial, Mr. Kisor argues, the term “relevant” should be construed broadly in a manner consistent with the interpretation above that he urges. Id. at 13-15.

We disagree. Although broad, the VA‘s duty to assist is not without limits. Under 38 U.S.C. § 5103A(a)(2), “[the] Secretary is not required to provide assistance to a claimant under this section if no reasonable possibility exists that such assistance would aid in substantiating the claim.” (Emphasis added). Thus, to the extent the VA‘s duty to assist encompasses evidence necessary, but not sufficient, to substantiate a veteran‘s claim, the duty does not extend to the situation where, like here, the evidence provides no reasonable possibility that the claim could be substantiated because the evidence does not establish a missing claim element.

Finally, Mr. Kisor argues that we should resort to the “pro-veteran canon” of construction, see, e.g., Brown v. Gardner, 513 U.S. 115, 117-18 (1994), and thereby arrive at the reading of the term “relevant” in § 3.156(c) that he urges. Appellants’ Suppl. Br. 16-18. The canon is not available in this case, however, because it only applies in the situation where the statute or regulation at issue is ambiguous. Paralyzed Veterans of Am. v. Sec‘y of Veterans Affairs, 345 F.3d 1334, 1340 (Fed. Cir. 2003) (noting that the pro-veteran canon “applies only to ambiguous statutes” and “cannot be invoked to override the clear meaning of a particular provision“) (internal quotation marks and citation omitted); Spicer v. Shinseki, 752 F.3d 1367, 1371 (Fed. Cir. 2014) (declining to apply the pro-veteran canon to a regulation when there was no ambiguity and no interpretive doubt).⁵

In this case, both Mr. Kisor and the government take the position that the term “relevant” in § 3.156(c) is not “genuinely ambiguous.” We agree with that position and hold today that the term has only “one reasonable meaning.” That is the meaning adopted by the Board when it denied Mr. Kisor an effective date earlier than June 5, 2006 for service connection for his PTSD.

We have considered Mr. Kisor‘s other arguments and have found them to be without merit.

CONCLUSION

For the foregoing reasons, we conclude the term “relevant” has only one reasonable meaning in the context of § 3.156(c)(1): the “relevant” service records must speak to a matter in issue, in other words, a matter in dispute. Accordingly, we affirm the decision of the Veterans Court that affirmed the decision of the Board denying Mr. Kisor an effective date earlier than June 5, 2006 for service connection for his PTSD.

AFFIRMED

Costs

No costs.

REYNA, Circuit Judge, dissenting.

This appeal is on remand from the U.S. Supreme Court.

Three years ago, this panel unanimously held that the plain text of 38 C.F.R. § 3.156(c) was ambiguous as to the scope of the word “relevant.” It was on that basis that we applied Auer deference to what we determined was a reasonable interpretation of the regulation by the Department of Veterans Affairs (“VA“). We did not at the time consider a countervailing tool used to resolve ambiguities in veterans’ benefits regulations, the pro-veteran canon. The Supreme Court vacated our decision because it relied prematurely on Auer, and on remand, the VA waived Auer altogether.¹ Now, the majority decides that the agency‘s construction—of the same text we held was ambiguous—has become unambiguously correct. Slip Op. 15-17. On that basis, the majority again avoids application of the pro-veteran canon.

I disagree with my colleagues’ new position that the “one reasonable meaning” of the word “relevant” in 38 C.F.R. § 3.156(c) is the VA‘s. Slip Op. 4, 9, 16. Nothing in the plain meaning or context of the provision requires “relevant” records to “speak to the basis for the VA‘s prior decision,” to address facts expressly “in dispute,” or to “affect the outcome.” See Slip Op. 8, 9, 12, 14. Rather, the established constructions of the terms “relevant records” and “material evidence” in related veterans’ benefit provisions support the conclusion that records are “relevant” so long as they help to establish unestablished facts that are necessary for substantiating the veteran‘s claim. The majority relies on flawed assumptions and inferences in adopting the VA‘s construction and points to nothing that renders the more pro-veteran construction unreasonable.

Fundamentally, when a veterans’ benefit provision is ambiguous on its face, the pro-veteran canon must be weighed alongside the other traditional tools in resolving interpretive doubt. Neither the Supreme Court‘s decision in this case, nor this court‘s precedent, supports the majority‘s assumption that the canon is to be treated like Auer deference as solely a tiebreaker of last resort. Rather, the Court clarified that the “traditional tools of construction” must precede deference to the agency. As such a tool, the pro-veteran canon requires that we discern the remedial purpose of a veterans’ benefit provision in the context of the veterans’ benefit scheme as a whole and ensure that our construction effectuates, rather than frustrates, that purpose. By brushing aside the canon, the majority adopts a construction of § 3.156(c) that substantially narrows the scope of its remedial function. Thus, Mr. Kisor, a veteran who was denied twenty-three years of compensation for his service-connected disability after a disgracefully inadequate VA review, is denied relief under a regulation specifically promulgated to benefit him and other veterans in his situation. The result will reverberate far beyond this case.

I dissent.

I

When James Kisor submitted his first claim for service-connected post-traumatic stress disorder (“PTSD“) in 1982, he had undergone over a year of counseling for his symptoms at the Portland Vet Center. Yet a VA examiner diagnosed him with personality disorders rather than PTSD, and based on that diagnosis, the VA denied his claim on a one-page form. J.A. 23.

There is no dispute that the agency made no effort, before or after receiving the examiner‘s report, to determine whether Mr. Kisor suffered a traumatic stressor during his service in Vietnam. This was in spite of the fact that (1) a legal element of any PTSD claim is a verified in-service stressor, and (2) the first clinical criterion for a medical diagnosis of PTSD (another legal element of a PTSD claim) is the experience of an objectively distressing traumatic

event.² There was no documentation whatsoever of combat experience in Mr. Kisor‘s file because the VA had never bothered to request his personnel records from the service department. The rating decision made no mention of his combat status. J.A. 23.

It bears emphasizing a few neglected details of the examination that led to this rating decision against Mr. Kisor. Although the examiner‘s report recounted Mr. Kisor‘s descriptions of his combat experience, it did so with palpable skepticism³ and noted that Mr. Kisor had reported “no battle problems or traumatic experiences” to his social worker. J.A. 18-20. At the time, Mr. Kisor‘s treating counselor had considered his symptoms to be consistent with PTSD. J.A. 21. The examiner noted he was “not impressed” with that diagnosis but provided no explanation of the basis for his own opinion. J.A. 21. This was because he had “lost” the “portion of the original dictation” setting forth a “specific review of symptoms related to the PTSD criteria” and could not “recall the specifics.” J.A. 21-22. All he could offer was his “impression.” Id. Despite all this, the rating board accepted the examiner‘s diagnosis and went no further with Mr. Kisor‘s claim.

For the next twenty-three years, Mr. Kisor received no disability compensation from the VA, although the symptoms of his condition continued to keep him from holding down a job. In 2006, Mr. Kisor went to check his VA claims file, and discovered that there were no records of his combat history. He wrote to the VA, attaching service records documenting his combat history and Combat Action Ribbon, and demanded that the agency look again at his claim. J.A. 28-29. The VA construed his first letter as a request to reopen his claim based on new and material evidence, and although nothing else about his claim had changed, the VA this time proceeded to investigate his alleged in-service stressor, requesting an entry from his battalion‘s daily log that documented the following attack:

battalion forward and rear elements taken under heavy fire by mortar, recoilless rifle, and automatic weapons. . . . VC [Viet Cong combatants] were well camouflaged and dug into concealed positions. All VC contacted were well armed and equipped . . . . VC KIA [killed in action] 105.

J.A. 30-31. Based on the information in the log—information that all along had been in the government‘s possession—the VA formally verified Mr. Kisor‘s stressor. Id.

Mr. Kisor then obtained and submitted an evaluation from a third-party psychiatrist, who concluded that Mr. Kisor met each of the diagnostic criteria for PTSD and had been suffering from the effects of his condition for the last 27 years. J.A. 109. In particular, the psychiatrist opined that the VA examiner in 1983 had likely “misunderstood the impact of the claimant‘s war trauma upon him,” as symptoms of PTSD were apparent from Mr. Kisor‘s medical records at that time. Id. A new VA psychiatric examination concurred with this diagnosis.⁴ J.A. 115-116. This time, the new examiner accepted the presence of “combat stressors” based on records of Mr. Kisor‘s combat action ribbon, J.A. 112, and proceeded to describe his combat accounts and symptoms fully and sympathetically. The examiner also received and reviewed the other records now in Mr. Kisor‘s claims file. Id.

Based on Mr. Kisor‘s new diagnosis of PTSD and his service records, the VA found that he had established the necessary elements of a service-connected PTSD claim and awarded compensation for the claim. J.A. 32-33. The agency, however, refused to treat its new review as a “reconsideration” under § 3.156(c), which would entitle him to an effective date retroactive to his 1982 claim. The Board of Veterans’ Appeals (“Board“) recognized that reconsideration is only triggered when the VA receives newly identified “relevant official service records.” The Board reasoned that the newly received combat records in Mr. Kisor‘s case—i.e., his combat expeditions form, his Combat Action Ribbon award, and his battalion‘s daily log—were not “relevant” because they did not address the “basis” of the VA‘s prior decision and did not “manifestly change” its outcome. J.A. 90-91.

II

In all cases, the VA has a statutory duty to assist the veteran by fully and sympathetically developing the veteran‘s claim to its optimum before deciding the claim on the merits. McGee v. Peake, 511 F.3d 1352, 1357 (2008). The VA bears this obligation so long as there is any “reasonable possibility” that such assistance would “aid in substantiating the claim.” 38 U.S.C. § 5103A; Golz v. Shinseki, 590 F.3d 1317, 1323 (Fed. Cir. 2010). This includes making reasonable efforts to obtain evidence necessary to substantiate the veteran‘s claim. 38 U.S.C. § 5103A. In particular, the VA must obtain “relevant records pertaining to the claimant‘s active [military] service that are held or maintained by a governmental entity.” 38 U.S.C. § 5103A(c)(1).

What happens when the VA fails to fulfill this duty? If, decades after a claim is denied, the veteran uncovers service records that prove a necessary element of his claim and should have been part of his file, will his claim be reconsidered, offering him a chance to prove entitlement dating back to his first claim? Or must he first bear the burden of showing that the missing records might have changed the VA‘s original decision? The answer turns on the construction of the word “relevant” in the VA‘s regulation, 38 C.F.R. § 3.156(c).

Section 3.156(c) provides for reconsideration of claims previously decided without the benefit of all relevant service records. Subsection (c)(1) requires the VA to reconsider a claim if it receives “relevant service department records” that had not been considered when it first decided the claim:

[A]t any time after VA issues a decision on a claim, if VA receives or associates with the claims file relevant official service department records that existed and had not been associated with the claims file when the VA first decided the claim, VA will reconsider the claim.

§ 3.156(c)(1). Reconsideration includes further VA assistance in developing any additional evidence needed to substantiate the claim. 38 U.S.C. § 5103A(a)(1); Vigil v. Peake, 22 Vet. App. 63, 67 (2008). If, after reconsideration of the claim, “an award [is] made based all or in part” on these records, then the award is effective as far back as the effective date of the previously decided claim, depending on when entitlement arose, as determined through a retroactive assessment of disability. § 3.156(c)(3), (c)(4).

The plain text of § 3.156(c)(1) does not specify whether the “relevant” records that trigger reconsideration must “cast[] doubt on the agency‘s prior rating decision” or only “relat[e] to the veteran‘s claim more broadly.” Kisor v. Shulkin, 869 F.3d 1360, 1367 (Fed. Cir. 2017) (“Kisor I“). However, the history and text of § 3.156(c) make clear that reconsideration serves the dual remedial purpose of (1) providing a fair claim review based on a fully developed record to veterans who had been denied such a review before and (2) compensating such veterans for any benefits to which they can now prove they should have been entitled. We have noted that ”§ 3.156(c) serves to place a veteran in the position he would have been had the VA considered the relevant service department record before the disposition of his earlier claim.” Blubaugh v. McDonald, 773 F.3d 1310, 1313 (Fed. Cir. 2014). That includes affording him both his procedural right to a complete review and his substantive right to full compensation.

In light of the ambiguity in § 3.156(c) and the regulation‘s remedial purpose, consistent with the Supreme Court‘s instructions on remand, I look to the provision‘s context and history for a construction of “relevant” that best effectuates the purpose of reconsideration. I turn first to our construction of “relevant records” in the context of the VA‘s duty to assist veterans. I then look to the historical scope of the “new and material evidence” standard for the reopening of claims, which served as the original standard for reconsideration under § 3.156(c). Both sources point to the conclusion that “relevant . . . records” need only address a necessary and unestablished element of the claim as a whole, not the specific disputed issue that served as the basis for the VA‘s prior decision.

A. “Relevant Records” and the Duty to Assist

As discussed, 38 U.S.C. § 5103A requires the VA to assist a claimant in obtaining “evidence necessary to substantiate the claimant‘s claim,” including obtaining “relevant records” of the claimant‘s military service, so long as there exists any “reasonable possibility that such assistance would aid in substantiating the claim.” 38 U.S.C. §§ 5103A(a)(1)-(a)(2), 5103A(c)(1)(A) (emphasis added). There is no dispute that “relevant” records for purposes of reconsideration should be construed consistently with the meaning of “relevant records” under § 5103A.

In interpreting § 5103A, this court has defined “relevant records” as “those records that relate to the injury for which the claimant is seeking benefits and have a reasonable possibility of helping to substantiate the veteran‘s claim.” Golz v. Shinseki, 590 F.3d 1317, 1321 (Fed. Cir. 2010). When determining the scope of “relevant records” for a given claim, we look to the elements necessary to substantiate it. See id. at 1322. In particular, when a veteran seeks compensation for service-connected PTSD, we have held that the “records relevant to his claim are those relating to a medical diagnosis of PTSD, evidence corroborating claimed in-service stressors, or medical evidence establishing a link between any in-service stressor and a PTSD diagnosis.” Id.

We have also made clear that the VA‘s obligation to obtain relevant records does not depend on whether the records would likely be “dispositive” of the claim. McGee, 511 F.3d at 1358 (“The statute [§ 5103A] simply does not excuse the VA‘s obligation to fully develop the facts of [the veteran‘s] claim based on speculation as to the dispositive nature of relevant records.“). We have held that relevant records need not “independently prove” the veteran‘s claim. Jones v. Wilkie, 918 F.3d 922, 926 (Fed. Cir. 2019).

The scope of the VA‘s duty to assist thus supports the conclusion that “relevant” records are those that help to establish a necessary element of a veteran‘s claim, regardless of whether the evidence would be dispositive of the outcome. By this standard, Mr. Kisor‘s combat records are relevant at least because they corroborate his in-service stressor, a necessary element of a PTSD claim that had not been established when the VA first decided his claim. 38 C.F.R. § 3.304(f); AZ v. Shinseki, 731 F.3d 1303, 1310 (Fed. Cir. 2013).

B. “New and Material Evidence”

Up until 2019, all of 38 C.F.R. § 3.156 fell under the heading “New and Material Evidence.” As originally enacted, the provision provided for (1) reopening of previously decided claims based on “new and material evidence” and (2) reconsider[ation] of previously decided claims based on new and material evidence that consisted of official records from the service department. See § 3.156, New and Material Evidence, 27 Fed. Reg. 11887 (Dec. 1, 1962) (emphasis added). The distinction between the two procedures was that reconsideration provided the veteran an opportunity to prove and receive retroactive entitlement to benefits, whereas reopening only entitled veterans to the effective date of the request to reopen.

Effective 2006, the VA amended the language in § 3.156(c) to delete the reference to “new and material evidence,” and replace it with the current phrase “relevant official service records.” In proposing the change, the VA stated that the change was intended to eliminate any confusion as to whether awards made upon reconsideration would be subject to the same effective date as awards made upon reopening. New and Material Evidence, 70 Fed. Reg. 35388, 35388-89 (Jun. 20, 2005). The VA was thus clear that the new “relevant . . . records” language was not intended to impose a higher threshold for triggering reconsideration than before. It follows that records are “relevant” under § 3.156(c)(1) if they would satisfy the definition of “material evidence” for purposes of reopening a claim.⁵

This is critical because the standard for “material evidence” has always been forward-looking toward the claim to be substantiated, not backward-looking toward the prior VA decision. Since 2001, the VA has defined “material” evidence as “evidence that, by itself or when considered with previous evidence of record, relates to an unestablished fact necessary to substantiate the claim.” § 3.156(a) (emphasis added); see also § 3.156, Duty to Assist, 66 Fed. Reg. 45620, 45630 (Aug. 29, 2001). Historically, when the VA promulgated its first binding definition of materiality in 1990, it stated that “it has always been VA‘s position that evidence may be new and material even though it does not warrant revision of a previous decision.” New and Material Evidence, 55 Fed. Reg. 52274 (Dec. 21, 1990) (emphasis added).

Accordingly, in Hodge v. West, we rejected the Veterans Court‘s requirement that a claimant seeking reopening establish “a reasonable possibility that the new evidence, when viewed in the context of all the evidence, both new and old, would change the outcome.” 155 F.3d 1356, 1363 (Fed. Cir. 1998) (citing Colvin v. Derwinski, 1 Vet. App. 171, 174 (1991)) (emphasis added). We concluded that an outcome determinacy requirement for reopening, even under an attenuated “reasonable possibility” threshold, was “inconsistent with the general character of the underlying statutory scheme for awarding veterans’ benefits.” Id. at 1362. We reasoned that the availability of review based on new evidence reflects “the importance of a complete record for evaluation of a veteran‘s claim” that considers “all potentially relevant evidence.” Id. at 1363. We recognized that “so much of the evidence regarding the veterans’ claims for service connection and compensation is circumstantial at best,” and in this context, new evidence may “contribute to a more complete picture of the circumstances surrounding the origin of a veteran‘s injury or disability,” and warrant another look at the claim, even if it does not demonstrably change the right outcome. Id. Moreover, both the reopening and reconsideration of a claim entitles the veteran to receive additional assistance from the VA, such as new medical examinations and requests for additional records. See Paralyzed Veterans of Am. v. Sec‘y of Veterans Affs., 345 F.3d 1334, 1339, 1343 (Fed. Cir. 2003); Vigil v. Peake, 22 Vet. App. 63, 67 (2008). The provisions thus contemplate that a claim that is not fully substantiated based on the new evidence alone may be substantiated after further factual development.

The VA adhered to these principles when it adopted the current definition of materiality in 2001. In particular, the VA withdrew as “too restrictive” a proposal that would have defined “material evidence” as “evidence that relates specifically to the reason why the claim was last denied.” Duty to Assist, 66 Fed. Reg. at 45629 (final rule) (emphasis added); cf. Duty to Assist, 66 Fed. Reg. 17834, 17838-89 (Apr. 4, 2001) (proposed rule). In its place, the VA promulgated the current definition of materiality that focuses on the “unestablished fact[s] necessary to substantiate the claim.” Duty to Assist, 66 Fed. Reg. at 45629 (emphasis added).⁶

If the VA now intends to condition reconsideration on records that relate to the basis of the prior decision or change its outcome, it must do so through notice and comment. The agency cannot urge us to read those requirements into the word “relevant” when they have repeatedly refused to incorporate them into the criteria for reopening and reconsideration in promulgating prior versions of the regulation. The history and context of § 3.156 thus make clear that records relating to unestablished facts necessary to substantiate the veteran‘s claim are sufficient to trigger reconsideration under subsection (c).

* * *

Viewed as a whole, the context, history, and purpose of reconsideration support a construction of “relevant” that entitles Mr. Kisor to relief: that service records are “relevant” when they help to establish an unestablished fact necessary to substantiate a veteran‘s claim. Moreover, this reading of § 3.156(c) accords with the pro-veteran canon because it most effectuates the provision‘s remedial purpose of (1) ensuring that veterans whose claims were denied without the benefit of full VA assistance receive the full review and assistance they were owed; and (2) compensating veterans for any past benefits to which they can prove they should have been entitled.

III

Nothing in the majority‘s reasoning undermines the soundness of this pro-veteran interpretation. The majority concludes that a combination of dictionary definitions, context, and case law “makes clear” that the VA‘s interpretation is correct, but its inferences and assumptions fail under scrutiny.

First, borrowing from definitions of “relevant” as pertaining to “a matter in issue,”⁷ the majority assumes that “in issue” means “in dispute,” and reasons that evidence can only be relevant if it pertains to facts that were “disputed” during the claim‘s prior adjudication. Slip Op. 9, 12. Not only is this inference unwarranted in common legal usage, see Fed. R. Evid. 401 advisory committee‘s note (“[t]he fact to which [relevant] evidence is directed need not be in dispute“),⁸ it is fundamentally out of place in the VA‘s “completely ex-parte system of adjudication.” Hodge, 155 F.3d at 1362-63. Because no adverse party is expected to contest a claimant‘s assertions, the question of whether a fact is “disputed” has no import for whether it must be supported by competent evidence and adjudicated by the VA; that question depends instead on whether the fact remains unestablished and necessary for substantiating the claim. Here, regardless of whether the presence of Mr. Kisor‘s in-service stressor was “disputed” by the VA, it was not established at the time of the VA‘s first decision because the only mention of his combat experience in the record—a second-hand account by a VA examiner—was not competent evidence of a stressor. See, e.g., Cohen v. Brown, 10 Vet. App. 128, 145-46 (1997) (noting, in remanding a case to the Board, that “[a]n opinion by a mental health professional based on a post[-]service examination of the veteran cannot be used to establish the occurrence of the stressor,” and that the VA is “not require[d] [to] accept[] . . . a veteran‘s assertion that he was engaged in combat with the enemy“).

Next, the majority infers from language in § 3.156(c)(3) that relevant records must “speak to the basis for the VA‘s prior decision.”⁹ Slip Op. 12. Subsection (c)(3) provides that an award granted after reconsideration can receive a retroactive effective date if it is “made based all or in part” on the records that triggered reconsideration. § 3.156(c)(3). The majority reasons that records that do not “speak to the basis for the VA‘s prior decision” cannot form all or part of the basis for the VA‘s current award of benefits after reconsideration. Slip Op. 12. But nothing in the text ties the basis of the subsequent award to the basis of the prior decision. Nor are the two logically linked. If the VA denies a claim based on lack of evidence for one element without reaching the others, a later decision granting the claim will still be “based” on evidence of all the elements. And here, the VA‘s 2007 award to Mr. Kisor was indisputably “based” at least “in part” on his combat records. The majority seems to admit as much, see Slip Op. 6, and the Board never found otherwise.

In addition, the majority suggests that language in § 5103A(a)(2) excused the VA from further assisting with or reconsidering Mr. Kisor‘s claim after the first VA examiner failed to diagnose him with PTSD. Slip Op. 12, 15 (citing § 5103A(a)(2)). Section 5103A(a)(2) provides that the VA is not obligated to assist with a claim if “no reasonable possibility exists that such assistance would aid in substantiating the claim.” The majority reasons that if “evidence does not establish a missing claim element,” then it “provides no reasonable possibility that the claim could be substantiated.” Slip Op. 15.

But that reading of § 5103A is irreconcilable with our precedent that the VA‘s duty to obtain records is not limited to “dispositive” evidence. McGee, 511 F.3d at 1358; Jones, 918 F.3d at 926. We have emphasized that the VA‘s duty to assist is excused only when ”no reasonable possibility exists that such assistance would aid in substantiating the claim.” Jones, 918 F.3d at 926 (emphasis in original) (citing 38 U.S.C. § 5103A(a)(2)).¹⁰ Even when the availability of a new record leaves a claim element unestablished, there often remains the possibility that the missing element will be established with further assistance. Indeed, the “no reasonable possibility” standard in § 5103A(a)(2) was enacted to replace the unduly burdensome “well-grounded claim” standard in § 5107(a) that had required a veteran to present plausible evidence of each element of his claim before triggering the VA‘s duty to assist. See Paralyzed Veterans of Am., 345 F.3d at 1343; Epps v. Gober, 126 F.3d 1464,

1468 (Fed. Cir. 1997). The majority‘s reading of the “reasonable possibility” standard would import the well-pleaded complaint rule into the very provision enacted to overrule it.

Moreover, there is no factual basis for concluding that Mr. Kisor‘s claim had “no reasonable possibility” of being substantiated. The VA treated his claim as capable of substantiation when it obtained unit records to substantiate his combat stressor. On appeal, the Board found only that the combat records did not “manifestly change [the] outcome” of the VA‘s decision, not that they had no reasonable possibility of helping to do so. J.A. 90. In fact, Mr. Kisor‘s claim was substantiated with the aid of his combat records, and not miraculously so. Once there was competent evidence of Mr. Kisor‘s stressor, all that was needed to substantiate his claim was a new psychiatric examination. Given the history of his first examination, and the circumstantial nature of a PTSD diagnosis, there was at least a reasonable possibility that a new examination in light of the newly collected evidence would yield a different diagnosis and substantiate his claim.¹¹

Finally, the majority relies on language from Blubaugh v. McDonald for the proposition that “relevant” service records must (1) “remedy the defects” of a prior decision, (2) pertain to facts that were “in question,” and (3) “lead VA to award a benefit that was not granted in the previous decision.” Slip Op. 14 (citing Blubaugh, 773 F.3d at 1314). But the majority reads these statements out of context. In Blubaugh, we were not construing the word “relevant” as the threshold for reconsideration. We were explaining that retroactive benefits are only available under § 3.156(c) if entitlement is in fact awarded upon reconsideration of the veteran‘s claim. In Blubaugh, the veteran‘s claim was denied when the VA reconsidered his claim in light of his newly identified service record—a document that was not probative of any fact necessary for substantiating his claim. Id. It was in the context of discussing this denial by the VA that we explained that the new record did not “remedy [the] defects” of the prior decision, pertain to facts that were “in question,” or “lead VA to award a benefit.” Id. at 1314. Thus, nothing in Blubaugh suggests that service records are not “relevant” when, as here, the VA awards a claim after considering the records and expressly relies on the records in making the award.

Ultimately, nothing in the majority‘s reasoning establishes that the VA‘s outcome determinacy requirement for relevance is compelled by the text of the regulation or otherwise unambiguously correct. Thus, the majority should have tested the strength of the VA‘s arguments against the weight of the pro-veteran canon. That the majority refused to do so here deprived Mr. Kisor of the solicitude and independent judgment he was owed in this appeal.

IV

Courts have “long applied the canon that provisions for benefits to members of the Armed Services are to be construed in the beneficiaries’ favor.” Henderson ex rel. Henderson v. Shinseki, 562 U.S. 428, 441 (2011) (citing King v. St. Vincent‘s Hosp., 502 U.S. 215, 220-21, n.9 (1991) (internal quotations omitted)). Thus, interpretive doubt in such provisions should be resolved for the benefit of veteran. Brown v. Gardner, 513 U.S. 115, 118 (1994). This canon is a corollary of the broader interpretive rule that remedial provisions are to be construed liberally to effectuate and not frustrate their remedial purpose. See Boone v. Lightner, 319 U.S. 561, 575 (1943); Beley v. Naphtaly, 169 U.S. 353, 361 (1898).

This panel unanimously held in Kisor I that the plain text of § 3.156(c) was ambiguous as to the scope of the word “relevant,” and that text has not changed since that decision. Kisor I, 869 F.3d at 1367. Yet the majority concludes that the canon is “not available in this case” because after considering arguments that favor the VA‘s position under the other tools of construction, the provision is not “genuinely ambiguous.” Slip Op. 15-16.

But while we have held that the pro-veteran canon applies only to ambiguous statutes and cannot override plain text, that rule does not render the canon a tool of last resort, subordinate to all others.¹² To the contrary, we have stated that the canon applies whenever the plain text does not expressly exclude the veteran‘s interpretation. Sursely v. Peake, 551 F.3d 1351, 1357 (Fed. Cir. 2009); Hudgens v. McDonald, 823 F.3d 630, 637 (Fed. Cir. 2016). Thus, we have accepted the canon‘s guidance over the VA‘s reliance on a dictionary definition. Hudgens, 823 F.3d at 637. We have weighed the canon against countervailing legislative history. Nat‘l Org. of Veterans’ Advocs., Inc. v. Sec‘y of Veterans Affs., 260 F.3d 1365, 1377-78 (Fed. Cir. 2001). We have favored the canon over arguments that the veteran‘s interpretation would lead to “irrational” results. Sursely, 551 F.3d at 1357-58.

While the canon may not be dispositive of a provision‘s meaning every time it is applied, we are obligated to weigh it alongside the other tools of construction when the text itself gives us doubt.

Here, the majority points to nothing in the text that precludes Mr. Kisor‘s interpretation of “relevant.” Indeed, this panel accepted in Kisor I that his position was reasonable. Kisor I, 869 F.3d at 1368. While the majority now rejects his view as “squarely contrary” to what it concludes is the “correct reading” of the regulation, it does not explain why his reading is now contrary to the text.

The majority wrongly assumes that the Supreme Court‘s “genuine ambiguity” criterion for Auer deference applies to the pro-veteran canon. See Slip Op. 15-16. In setting the preconditions for Auer deference, the Court requires courts to first exhaust the “traditional tools of construction” because “the core theory of Auer deference is that sometimes the law runs out, and [a] policy-laden choice is what is left over.” Kisor v. Wilkie, 139 S. Ct. 2400, 2415 (2019) (“Kisor II“). The pro-veteran canon is not based on this “deference” theory. The canon does not serve to provide a “policy-laden” position, adrift from traditional legal principles, that differs with each case. Id. Rather, the pro-veteran canon is squarely rooted in the purpose of veterans’ benefit provisions, which we are bound to consider and effectuate in every construction.

If, as the majority seems to suggest, we can set aside the pro-veteran canon unless and until all other considerations are tied, then the canon is dead because there is no such “equipoise” in legal arguments. Id. at 2429-30 (Gorsuch, J., concurring in the judgment). It is our role as the court to fully employ the canons available in our “traditional interpretive toolkit” to reach “the best and fairest reading of the law.” Id. at 2430, 2446. In this case, when the regulatory text provides no clear answer as to the scope of the word “relevant,” our consideration of other sources of its meaning should be guided by solicitude for the provision‘s pro-veteran remedial purpose.

Here, reconsideration under § 3.156(c) serves two remedial purposes: procedurally, it acknowledges to the veteran that the VA failed in its duty to assist him and provides him with the complete and sympathetic assistance and review that he was owed; substantively, it makes the veteran financially whole for the benefits that he can now prove he was entitled to. The VA‘s interpretation frustrates both of those purposes. It denies veterans the right to a fair review unless they make the often impossible showing that an unsought record would have changed the course of the VA‘s prior decision. And it bars veterans from recovering compensation that is rightfully theirs.

The unreasonableness of that construction is plain in this case. The VA undeniably failed Mr. Kisor in this case when it made no effort whatsoever to obtain records to substantiate his in-service stressor. Rather than acknowledge its failure and make amends for it, the VA placed the burden on Mr. Kisor to show that its mistake was dispositive of its decision against him. When the agency deemed its new requirement unsatisfied, it denied the veteran twenty-three years of benefits for PTSD that he can now prove he suffered as a result of his service.

Those payments were compensatory, not charitable. They rightfully belonged to Mr. Kisor and his family. When Mr. Kisor and millions of others joined the armed services in their youth, for modest pay, risking the rest of their lives, they did so with the government‘s promise that upon their return, it would make them as whole as possible, if only financially, for their wounds, and that, as veterans, they would be treated fairly and sympathetically in the process. That is the basic purpose of the VA‘s existence. Its governing statutes and regulations should always be construed liberally within the bounds of their text to effectuate that purpose. This recognition is at the core of the pro-veteran canon. The majority waves it aside.

On this remand, freed from deference to the agency, we owed Mr. Kisor our best independent judgment of the law‘s meaning. We fail in that obligation when we again accept the VA‘s arguments unmoored from both the text of the law and the basic principles underlying its purpose.

For these reasons, I dissent.

Notes

Under the Veterans Appeals Improvement and Modernization Act of 2017, Pub. L. 115-55 (“Modernization Act“), veterans may now file “supplemental claims” based on “new and relevant” evidence. 38 U.S.C. § 5108 (2019); 38 C.F.R. §§ 3.156(d), 3.2501 (2019). Section 3.2501 defines “relevant evidence” as “information that tends to prove or disprove a matter at issue in a claim [and] includes evidence that raises a theory of entitlement that was not previously addressed.” 38 C.F.R. § 3.2501. The comments accompanying the proposed rule explained that the definition of “relevant evidence” came from 38 U.S.C. § 101(35). VA Claims and Appeals Modernization, 83 Fed. Reg. 39,818, 39,822 (proposed Aug. 10, 2018). The comments accompanying the final rule explain that the “new and relevant” standard for supplemental claims is “a lesser standard and reduces the claimant‘s burden” as compared to the prior “new and material” standard. VA Claims and Appeals Modernization, 84 Fed. Reg. 138, 144 (Jan. 18, 2019) (codified at 38 C.F.R. pts. 3, 8, 14, 19, 20, and 21). The earliest effective date for an award of disability benefits pursuant to a supplemental claim is the date the supplemental claim was filed. 38 U.S.C. § 5110 (2019); 38 C.F.R. §§ 3.400, 3.2500(h)(2) (2019). Recording of Oral Argument at 16:10-16:22 (“The government is not contending that the agency‘s interpretation is entitled to deference.“); see also Transcript of Oral Argument at 64:6-20, , http://oralarguments.cafc.uscourts.gov/default.aspx?fl=2016-1929_1142020.mp3 (conceding that Auer deference only applies “if the determination reflects the considered judgment of the agency as a whole” and that “we [the government] don‘t think that any individual Board decision by the VA Board reflects the considered judgment of the agency as a whole“).

In this case, Mr. Kisor did not explicitly seek reconsideration, but the Board considered reconsideration under § 3.156(c) when it addressed his request for an earlier effective date for service connection for PTSD. J.A. 88. See J.A. 107 (citing the diagnostic criteria for PTSD); AM. PSYCHIATRIC ASS‘N, DIAGNOSTIC AND STATISTICAL MANUAL OF MENTAL DISORDERS, 309.81 (3d ed. 1980) (identifying the first diagnostic criterion for PTSD as “[e]xistence of a recognizable stressor that would evoke significant symptoms of distress in almost everyone“); see also, e.g., (“A VA medical examination . . . concluded that he ‘does not meet DSM-IV criteria for the diagnosis of PTSD, in terms of a specific, identified stressor that meets Criterion A, which is required for the diagnosis to be made.‘“).

Recently, in , we addressed a claim under § 3.156(c)(1). However, in that case, the government did not dispute that the newly associated records were “relevant” and that reconsideration was required. See, e.g., J.A. 19 (“The veteran seemed to be implying that the very exposure to potential combat and the implied danger did affect a change upon his adaptation.“) (emphasis added); id. (“When the veteran was asked to describe combat situations he seemed very defensive and wanted to make certain that I understood that he was always in situations of combat danger.“); id. (“[I]t would appear that he was involved in one major ambush which resulted in 13 deaths in a large company. The veteran does not remember how long this ambush lasted. He described the ambush in the context of the stupidity of his commanding officer‘s orders and judgment.“); J.A. 20 (“Whenever I would ask direct questions concerning the actual amount of combat activity, this subject would get lost as he would again launch into another detailed anecdotal monologue.“); J.A. 21 (“[H]is Vietnam combat situations were couched in the framework of his basic premise: that most people who have attempted to boss him around had been inferior to him either intellectually or morally.“).

This understanding of “relevant” is also consistent with the definition for that term in connection with “supplemental claims” under the Modernization Act noted above. 38 C.F.R. § 3.2501 (“Relevant evidence is information that tends to prove or disprove a matter at issue in a claim [and] includes evidence that raises a theory of entitlement that was not previously addressed.“). There are similar instances in which a Vietnam veteran, whose PTSD claim was initially denied based on absence of a PTSD diagnosis, is later diagnosed with PTSD in a new examination, and awarded benefits after the VA receives new evidence of an in-service stressor. See, e.g., ; ; . Notably, the Board found § 3.156(c) to be applicable in each of these cases without questioning the relevance of the newly identified stressor evidence.

See also (declining to apply pro-veteran canon in part because there was no ambiguity in the statute); (“[T]he Secretary correctly posits that the rule of interpretative doubt favoring veterans in , has no force if a statute properly interpreted leaves no ambiguity as to its meaning.“) (citation omitted); (declining to hold a statute unambiguous by resorting to [the pro-veteran canon of construction] that is “used to analyze ambiguous statutes“); (declining to apply pro-veteran canon after concluding statute was unambiguous by “applying other interpretive tools.“); (declining to apply pro-veteran canon when the statute was not ambiguous); (“[W]hen we find an ambiguity in a veterans’ benefit statute, interpretative doubt is to be resolved in the veteran‘s favor. At the same time, we have also recognized that a veteran cannot rely upon the generous spirit that suffuses the law generally to override the clear meaning of a particular provision.“) (internal quotation marks and citations omitted). This is consistent with the VA‘s definition for “new and relevant evidence” for purposes of re-adjudicating “supplemental claims” under the recently enacted § 3.156(d). The definition provides that “[t]he new and relevant evidence” standard is no higher than the “new and material evidence” standard under § 3.156(a).

The same amendment also added the requirement that new and material evidence must “raise a reasonable possibility of substantiating the claim.” Duty to Assist, 66 Fed. Reg. at 45629. The VA clarified that this language required only that “there be a reasonable possibility that VA assistance would help substantiate the claim,” in accordance with the threshold for the VA‘s duty to assist. Id. (emphasis added). As I further explain in Section III, this “reasonable possibility” standard does not require new evidence to be independently capable of changing the outcome of a claim. See infra, 17-18.

See Slip Op. 12 (citing Black‘s Law Dictionary (10th ed. 2014) (defining “relevant” as “[l]ogically connected and tending to prove or disprove a matter in issue“)).

Indeed, the Advisory Committee observed that evidence directed to an uncontroversial point is often relevant and admissible at trial to “aid in understanding” the case. Fed. R. Evid. 401 advisory committee‘s note. Relatedly, in , this court rejected the VA‘s argument that “relevant” questions of law must have been specifically raised and addressed in prior proceedings. In doing so, we construed “relevant” to mean “bear[ing] upon or properly apply[ing] to the issues before us” based on the term‘s dictionary definitions. Id. (emphasis added).

The VA‘s position on whether “relevant” records must pertain to the “basis for the VA‘s prior decision” has been a moving target. The Board relied on this requirement in denying Mr. Kisor an earlier effective date. J.A. 91. In its initial response to Mr. Kisor‘s appeal to this court, the VA disavowed that interpretation, calling it “distorted.” Resp. 18-19. This panel accepted that disavowal. . On remand, the VA changed course in its supplemental briefing, asserting unequivocally that “to be ‘relevant’ for purposes of reconsideration, the additional records must speak to the basis for the VA‘s prior decision.” Gov. Supp. 16. The majority now accepts that interpretation without skepticism. Slip Op. 8, 12, 14.

It is instructive that the VA‘s own regulations appear to construe the “no reasonable possibility” standard extremely narrowly, limiting its examples to claims that are incapable of substantiation as a matter of law or facially incredible as a matter of fact: e.g., a veteran with a dishonorable discharge applying for VA benefits; a compensation claim for prostate cancer from a female veteran or ovarian cancer from a male veteran; a compensation claim for a disability that is the result of willful misconduct; or a claim for service connection for alcoholism or drug addiction. 38 C.F.R. § 3.159(d); Duty to Assist, 66 Fed. Reg. at 17837.

As discussed, supra n.4, Mr. Kisor was not unique in having different VA examiners reach different diagnoses of his condition.

Indeed, plain text defeats all other tools of construction. See (holding that when “plain language . . . is unambiguous, [the court‘s] inquiry begins with the statutory text, and ends there as well” (internal citations omitted)); see also (holding that “legislative history cannot override the plain meaning of a statute.“); (stating that “principles of symmetry cannot override the plain text of the statute.“); (finding constitutional avoidance canon inapplicable where it would require rewriting the statute).

Read the detailed case summary