Caffrey v. Brown

STEINBERG, Judge,

concurring in part and dissenting in part:

I concur in the opinion of the Court except for part II.B and its holding that a “breach of *385 the duty to assist cannot form a basis for a claim of [clear and unmistakable error]” (CUE) under 38 C.F.R. § 3.105 (a) (1993). Ante at 383-84. I also do not agree that there necessarily was a breach of VA’s duty to assist in either 1975, 1978, or 1979 1 when the veteran sought to reopen his claim.

A. Duty to Assist

Before concluding that the duty to assist in this case was violated by the Veterans’ Administration (now Department of Veterans Affairs) (VA or Department) Regional Office (RO) in 1975, 1978, or 1979, it must first be determined whether such a duty to assist then existed. The current statutory duty to assist in 38 U.S.C. § 5107 (a) was not enacted until 1988 (then as § 3007(a)) when it was added by the Veterans’ Judicial Review Act, Pub.L. No. 100-687, § 103 (a), 102 Stat. 4105 , 4106 (1988) (“[Secretary] shall assist such a claimant [one who has submitted well-grounded claim] in developing the facts pertinent to the claim”). However, that enactment merely codified a regulatory obligation in 38 C.F.R. § 3.103 (a) (1987) that had existed since 1972 and that had provided, in terms indistinguishable from those later enacted in section 5107(a), that “[i]t is the obligation of [VA] to assist a claimant in developing the facts pertinent to his [or her] claim.”

In concluding that the VARO in 1979 did not err in not obtaining the March 1962 Institute report, the Board of Veterans’ Appeals (BVA or Board) stated: “It is the responsibility of the claimant to provide reports of all private medical treatment pertinent to his claim.” Thomas A. Caffrey, BVA 90-_(Oct. 9, 1990), at 4. That statement is clearly wrong as a matter of law as to today’s duty to assist under Littke v. Derwinski, 1 Vet.App. 90, 92 (1990) (quoting § 3.103(a)). However, since the Court in Littke was construing the same regulatory and statutory words as existed in regulation in 1979 I think it would be reasonable to conclude that there was a comparable duty to assist since at least the 1972 promulgation of § 3.103(a). At a minimum, a duty to assist arose in this ease in 1979 (if not in 1975 and 1978) when the veteran specifically advised the RO about his 1962 psychiatric hospitalization at the Institute and gave the Institute’s address (R. at 65), and that duty to assist was either for VA to seek the report or to tell the veteran to get it. Doing neither would clearly have been a violation in 1979 when the RO had a specific duty to respond to the veteran’s very specific 1979 communication to it. Cf. Godwin v. Derwinski, 1 Vet.App. 419, 425 (1991) (duty to assist includes obligation to respond “one way or the other” to request for assistance).

According to the 1989 Statement of the Case, in 1975, 1978, and 1979 the RO informed the veteran in general terms that he needed to submit new and material evidence. R. at 106-07. Perhaps that was enough — at least up to 1979, but the record does not show what specific responses the RO gave at any of those times. Hence, I would hold that the duty to assist as described above might have been violated here and, if it was, then I would hold further, for the reasons set forth in part B, below, that that error constituted CUE. Accordingly, I would remand for the Board to readjudicate the CUE claim in light of the full record, which is not available to the Court.

B. CUE

More fundamentally, I differ from the majority’s holding that a CUE claim may never be premised on a duty-to-assist violation. I would tend to agree that such a violation could rarely meet the CUE prerequisite that the error must be one “which had it not been made, would have manifestly changed the outcome at the time it was made.” Russell v. Principi, 3 Vet.App. 310, 313 (en banc) (1992). In this ease, however, that CUE criterion is clearly met since we know that as soon as the RO learned in 1989 about the 1962 Institute record, which showed that the veteran had been treated for schizophrenia in the month after his discharge from service, it granted him service connection for that condition. R. at 81-82. Further, I agree with the majority (ante at 382) that the veteran *386 timely raised this issue with sufficient specificity in his August 1989 Notice of Disagreement. R. at 98.

For the following reasons, however, I do not agree with the majority that even if a failure to assist is demonstrated as to the RO’s actions or inactions in 1975, 1978, or 1979 (or more than one of those years), that is a type of error which cannot constitute CUE under § 3.105(a). First, the Secretary has not contended that it could not be CUE — only that there was no duty-to-assist violation. Mot. at 7-8. Second, the Court’s rationale for its conclusion that a duty-to-assist violation is not cognizable as the basis for a CUE claim — that “an incomplete record is [not] also an incorrect record” and that “an incomplete record, factually correct in all other respects, is not clearly and unmistakably erroneous” (ante at 383-84) — seems to be circular. For example, the Court concludes that “an analysis of whether CUE has been committed may only proceed on the ... record at the time of the prior determina-tion_” Ante at 383-84. Yet, Russell, supra, this Court’s seminal opinion on CUE, very clearly established as a basis for CUE that “[e]ither the correct facts, as they were known at the time, were not before the adjudicator or the statutory or regulatory, provisions extant at the time were incorrectly applied.” Russell, 3 Vet.App. at 313 . Here both factors may exist. The then-regulatory duty to assist may have been violated in 1975, 1978, or 1979 (or more than one of those times), and there is no question that “the correct facts” (the schizophrenia diagnosis of the veteran in the month after service) were not before the RO when it denied service connection in each of those adjudications.

Hence, the Court’s tautological distinction between an incomplete and an incorrect record seems manifestly at variance with a fundamental precept of Russell . Surely, it can just as reasonably be said that a record which causes the adjudicator to deny a claim when the “correct facts” would manifestly have produced the opposite result was an “incorrect” record and not just an “incomplete” one. In any event, I find nothing in Russell to warrant the incorrecVincomplete-record distinction which the majority attempts to make.

Moreover, the majority has taken a quotation out of context from Russell. Ante at 383-84. The majority relied on the third sentence in the following quotation without any reference to the first two sentences:

The short answer is that the claim which is reversed or amended due to a “clear and unmistakable error” is not being reopened. It is being revised to conform to the “true” state of the facts or the law that existed at the time of the original adjudication. New or recently developed facts or changes in the law subsequent to the original adjudication may provide grounds for reopening a case or for a de novo review but they do not provide a basis for revising a finally decided case.

Russell, 3 Vet.App. at 313 . The reference to “new or recently developed facts” in the extracted sentence refers to evidence that did not exist at the time of the prior adverse determination as contrasted with “the ‘true’ state of the facts ... that existed at the time of the original adjudication”. Indeed, those latter words in Russell provide full, and fully contextual, justification for the analysis I have set forth above since, indisputably, the prior RO adjudications were not based on the “true state of the facts”.

I concede that there is some language in Russell that could form a basis for the majority’s conclusion here. That is the statement that a CUE determination “must be based on the record and the law that existed at the time of the prior [RO] or BVA decision.” Id. at 314 . That particular phrase in the Russell opinion, however, seems to have been focusing on the kind of example it went on to give of an error which could not be CUE — “a new medical diagnosis that ‘corrects’ an earlier diagnosis ruled on by previous adjudicators.... ” Russell, 3 Vet.App. at 314 . That is not the situation before us.

Moreover, the Russell language “record ... that existed at the ’ time” is not the equivalent of the majority’s paraphrase “record at the time of the prior determination”. Ante at 383-84. The Russell reference to “the record” must be read in the context of this Court’s decisions expanding what consti *387 tutes the “record” before the Department for purposes of determining the scope of review in this Court under 38 U.S.C. § 7252 (b), which limits that review to “the record of proceedings before the Secretary and the Board”. See Bell v. Derwinski, 2 Vet.App. 611, 612 (1992) (holding that documents which were not actually before the adjudicators but had been generated by VA employees or submitted to VA by claimant were, “in contemplation of law, before the Secretary and the Board and should be included in the record”). This concept of a “constructive” record greater than the record actually before the adjudicators provides a useful analogy for squaring the Russell statements in such a way as to conclude that an outcome-determinative document which existed at a time when fulfillment of an existing duty to assist would have discovered it was part of “the record” that should have been before the adjudicators if they were to adjudicate the claim on the “correct facts” or the “true state of the facts”. See also Murincsak v. Derwinski, 2 Vet.App. 368, 372-73 (1992) (“The Court cannot accept the Board being ‘unaware’ of certain evidence, especially [but not necessarily exclusively] when such evidence is in possession of the VA, and the Board is on notice as to its possible existence and relevance.”).

C. Response to Concurring Opinion’s Reliance on Damrel Opinion

In his concurring opinion, Judge Kramer asserts that his opinion for the Court in Damrel v. Brown, 6 Vet.App. 242, 245 (1994), “put to rest” the possibility that a duty-to-assist violation could constitute CUE because “Damrel did nothing more than synthesize the Russell test into three prongs”. Ante at 384. I do not quarrel with Damrel’s synthesis, although the third prong would constitute a binding holding only if the Court’s decision sustaining the Board’s determination that there was no CUE in the 1967 prior RO adjudication were dependent on that prong. However, as to each of the two evidentiary items relied upon by the appellant in Dam-rel, the Court there held explicitly that even if they had been included explicitly within the record before the RO in 1967, they would be “not controlling for VA determinations” (a Social Security unemployability determination) and “not controlling with respect to ratings for compensation or pension” (a VA determination of total disability for VA insurance purposes). Damrel, 6 Vet.App. at 246 . Hence, in Damrel there could not have been a valid CUE claim as to either piece of evidence since the Russell criterion of “manifestly changing] the outcome” could not have been met.

Moreover, the Damrel opinion, in synthesizing a third Russell prong, reiterated the exact language of Russell that a CUE determination “must be based on the record and the law that existed at the time of the prior” adjudication. Id. at 245; Russell, 3 Vet.App. at 314 . I have analyzed that very language in part B., above, and explained how I believe those words should be read in Russell in the context of the particular facts of this case where the excluded true facts are undebat-ably outcome determinative. Although Damrel may lend support to the majority’s view on CUE here, it hardly “put to rest” the issue before us in the instant case.

Finally, I want to stress again, as I did at the outset of part B., that the CUE prerequisite of manifestly changing the outcome could rarely be met where the CUE asserted is a duty-to-assist failure. Only where it is indisputable that compliance with the duty to assist would have changed the outcome of the prior merits adjudication would such a CUE claim even have to be considered. Making such an outcome-determinative judgment would not involve the speculation suggested, by the concurring opinion, but, rather, would eliminate the need even to consider such a CUE assertion (was there a violation of the duty to assist as it existed, if it did, at the time?) in the vast majority of instances in which a duty-to-assist CUE claim might be raised. As to the concurrence’s Chicken-Little cry that a Frankenstein’s monster of “newly created evidence” that “bears a date prior to that of the [old] adjudication” and “that purports to be outcome determinative” will stalk our CUE jurisprudence, it would be well to remember that the sky never did fall and to leave consideration of any such factors to the elected policy-making branches of government.

*388

D. Conclusion

For the above reasons, I would remand the appellant’s properly raised CUE claim for readjudication by the BVA on the basis of the above analysis.

. The majority holds that VA's duty to assist was breached but does not specify when that breach occurred. Ante at 383.