Lynch v. McDonough

21 F.4th 776 (Fed. Cir. 2021), decided December 17, 2021. Court of Appeals for the Federal Circuit, No. 2020-2067, Circuit Judge Prost; Part II.B decided en banc. Circuit Judge Reyna concurring in part and dissenting in part from Part II.B, joined by Judges Newman and O'Malley. Precedential. Affirmed.

Certiorari denied, 143 S. Ct. 369 (2022) — so this is the last word on the question.

This opinion replaced an earlier one. The first decision in this same case, at 999 F.3d 1391 (Fed. Cir. 2021), was withdrawn and superseded. A source citing 999 F.3d 1391 is citing an opinion that no longer exists.

The rule it is about

38 U.S.C. §5107(b): "When there is an approximate balance of positive and negative evidence regarding any issue material to the determination of a matter, the Secretary shall give the benefit of the doubt to the claimant." The regulation, 38 C.F.R. §3.102, says the same in terms of reasonable doubt.

What it holds

Approximate balance means "nearly equal." It does not mean a dead tie.

Mr. Lynch argued that the earlier case law had read the word "approximate" out of the statute by demanding equipoise. The court disagreed that the case law said that — and then made sure it could not be read that way:

"Under §5107(b) ... a claimant is to receive the benefit of the doubt when there is an 'approximate balance' of positive and negative evidence, which Ortiz interpreted as 'nearly equal' evidence. This interpretation necessarily includes scenarios where the evidence is not in equipoise but nevertheless is in approximate balance. Put differently, if the positive and negative evidence is in approximate balance (which includes but is not limited to equipoise), the claimant receives the benefit of the doubt."

The court named a Veterans Court decision that had said otherwise — Chotta v. Peake, 22 Vet. App. 80 (2008), which stated that if the evidence "is not in equipoise ... the benefit of the doubt rule would not apply." That recitation, the court said, "is incorrect," and Chotta is abrogated on the point.

And it abandoned the "preponderance" framing. The older formulation said the rule does not apply when "the preponderance of the evidence is found to be against the claimant." That was correct in substance — it meant the factfinder had been persuaded — but the court recognized that "preponderance of the evidence" carries equipoise baggage from other areas of law. So, "to eliminate the potential for confusion going forward, we depart from Ortiz's 'preponderance of the evidence' language."

The replacement: the rule applies when the evidence is in approximate balance, and does not apply "when the evidence persuasively favors one side or the other."

Two things this case does not do

It does not reopen old denials. Footnote 6 forecloses it: "Today's change in our construction of §5107(b) does not provide grounds for claims of clear and unmistakable error for prior Board decisions," because CUE "does not include the otherwise correct application of a statute or regulation where, subsequent to the Board decision challenged, there has been a change in the interpretation." That is the first thing anyone wants to do with this case, and it is closed.

It does not change "at least as likely as not." Those are different standards doing different jobs — one is about whether a nexus is established, the other about what happens when the evidence on an issue is close. At least one later decision has had to say so to a veteran who conflated them.

The dissent, which is worth reading

Judge Reyna agreed that the preponderance formulation needed fixing but would have overruled Ortiz outright, and objected that the new "persuasive evidence" standard is the old one in different words. His practical worry is the sharper point:

"Where the evidence is close, but the VA ultimately determines that the evidence 'persuasively' forecloses a veteran's claim, the VA can make its determination without explaining that the case was in fact a close call."

He would require VA to say so when it finds a case close but not in approximate balance, so the question is actually reviewable.

That worry became the next case. Three years later the Supreme Court held in Bufkin v. Collins that the Veterans Court reviews the approximate-balance determination only for clear error.

And Mr. Lynch himself

He lost. The Board made extensive findings that it was persuaded he was not entitled to more than a 30% rating for PTSD, and the Veterans Court found the evidence "quite clearly against" him rather than in approximate balance. The rule was clarified; the result did not change.

Finding the opinion

The Federal Circuit's official PDF: 20-2067.OPINION.12-17-2021_1881557.pdf. Published at 21 F.4th 776.

Reviewed October 6, 2026