Kisor v. Wilkie and the Limits of Auer Deference

Kisor v. Wilkie is the 2019 Supreme Court decision that preserved Auer deference but rebuilt it around a demanding multi-step test that courts must now work through before accepting a federal agency’s interpretation of its own regulation. The five-justice majority refused to overrule the doctrine, citing stare decisis, but tightened it enough that many interpretations which would once have received near-automatic acceptance now fail at the courthouse door.¹

The Dispute That Reached the Court

James Kisor, a Vietnam War veteran, first applied for PTSD disability benefits from the Department of Veterans Affairs in 1982. The VA denied the claim after its evaluating psychiatrist concluded he did not have PTSD, even while acknowledging his involvement in a military action called Operation Harvest Moon.¹

In 2006 he moved to reopen the claim with a new psychiatric report. The VA agreed this time that he had PTSD but granted benefits only from the 2006 filing date. Under 38 CFR § 3.156(c)(1), the agency could have made benefits retroactive if it found there were “relevant official service department records” it had not considered the first time around. The VA read “relevant” to exclude the later-discovered service records and denied retroactive benefits. The Federal Circuit found the regulation ambiguous and deferred to the Board of Veterans’ Appeals under Auer. Kisor asked the Supreme Court to overrule Auer deference entirely.¹

What the Court Decided

On June 26, 2019, the Court declined to overrule Auer but changed how it must be applied. Justice Elena Kagan wrote for the majority, joined by Chief Justice Roberts and Justices Ginsburg, Breyer, and Sotomayor. The majority explained that Auer deference rests on “a presumption that Congress would generally want the agency to play the primary role in resolving regulatory ambiguities,” since agencies drafted the rules and understand the technical subject matter.²

Justices Gorsuch, Thomas, Kavanaugh, and Alito agreed with the outcome but would have overruled Auer outright. Justice Gorsuch wrote a lengthy concurrence arguing the doctrine was inconsistent with the Administrative Procedure Act and the separation of powers.

The majority grounded its refusal to overrule Auer in stare decisis. It pointed to a line of precedents stretching back roughly 75 years to Bowles v. Seminole Rock & Sand Co., the disruption that overruling would cause to countless settled regulatory interpretations, and Congress’s decision to leave the doctrine alongside the APA for nearly a century.³ The Court vacated the Federal Circuit’s ruling and sent the case back, finding the lower court had “jumped the gun” in declaring the regulation ambiguous and “assumed too fast” that Auer deference applied.²

The Kisor Framework

The core of the decision is a sequence of requirements a court must satisfy before deferring to any agency’s reading of its own regulation. Each step is a gate. Fail one, and deference is denied.

Is the Regulation Genuinely Ambiguous

A court must first decide whether the regulation is genuinely ambiguous, and this is not a casual inquiry. The court must exhaust all “traditional tools of construction” before concluding the text can reasonably be read more than one way.⁴ Those tools include the ordinary meaning of words, grammatical structure, context within the broader regulatory scheme, the purpose of the regulation, and the regulatory history. If applying them resolves the meaning, the inquiry ends and the court applies its own reading. No deference enters the picture.

This is where most deference claims now fail. Before Kisor, courts often declared a regulation ambiguous after a surface-level reading and moved straight to deference. The majority made clear that genuine ambiguity is a high bar. A regulation is not ambiguous simply because the parties disagree about what it means.

Does the Interpretation Come From an Authoritative Source

If the regulation survives that step, the court must confirm the interpretation comes from an authoritative source. The reading must “emanate from those actors, using those vehicles, understood to make authoritative policy” within the agency. A formal guidance document or official staff memorandum published in the Federal Register would qualify. An offhand remark by a mid-level employee or a legal argument invented for litigation would not.⁴

The requirement blocks agencies from retrofitting an interpretation to win a particular case. The reading has to reflect the institution’s position, not the view of whichever lawyer happens to be briefing the matter.

Is the Reading Reasonable

The interpretation must also be a reasonable construction of the regulation’s text, structure, and history. A reading that strains the language beyond what it can fairly bear gets no deference, even coming from the agency head personally.² The interpretation has to fall within the zone of meanings the words can support.

Does It Reflect the Agency’s Fair and Considered Judgment

Finally, the interpretation must reflect the agency’s “fair and considered judgment.” The Court identified several situations that disqualify a reading from meeting that standard:

Convenient litigating positions adopted to win a particular lawsuit.
Post hoc rationalizations offered after the fact to defend a past agency action.
Readings that create unfair surprise by imposing retroactive liability for conduct the agency had never previously addressed.
Interpretations that conflict with the agency’s own earlier reading of the same regulation.

The point is to keep agencies from using interpretation as a backdoor around formal rulemaking. If an agency wants to change how a regulation works in practice, it should generally do so through notice-and-comment rulemaking, where the public can participate.³

When Auer Deference Is Off the Table

Beyond the framework, Kisor and earlier precedent identify categorical situations where Auer deference does not apply at all. A regulation that simply restates the language of the underlying statute — a “parroting” regulation — receives no deference, because the agency is really interpreting the statute rather than its own rule. The Court recognized this in Gonzales v. Oregon (2006). Deference is also inappropriate when an agency’s interpretation creates an unfair surprise that disrupts regulated parties’ settled expectations, a principle courts have applied where an agency reversed longstanding practice without warning.

The Skidmore Fallback

When an interpretation fails Kisor’s requirements, it does not become worthless. Courts can still apply Skidmore deference, a weaker standard from the 1944 case Skidmore v. Swift & Co. Under Skidmore, an agency’s reading receives weight proportional to its persuasiveness, factoring in the thoroughness of the reasoning, how consistent the interpretation has been over time, and whether it reflects genuine expertise.

The practical difference is real. Auer deference puts a thumb on the scale for the agency: a court accepts a reasonable reading. Skidmore lets the agency’s view inform the court’s independent judgment. The agency must persuade rather than merely satisfy a reasonableness threshold. With Kisor’s narrowing, more agency interpretations now receive only Skidmore-level consideration.

What Happened to James Kisor on Remand

The Federal Circuit applied the new framework and concluded that the word “relevant” in 38 CFR § 3.156(c)(1) was not genuinely ambiguous. The court held it had only one reasonable meaning in context: a record must speak to a matter actually in dispute and affect the outcome.⁵

The basis for the VA’s original 1983 denial was the lack of a PTSD diagnosis, not any dispute over whether Kisor had served in combat. The later-discovered service records documented his involvement in Operation Harvest Moon, which had never been in question. Because those records did not address the actual reason for the denial, they were not “relevant” under the regulation, and Kisor was not entitled to retroactive benefits. The Federal Circuit affirmed the denial of an effective date earlier than June 5, 2006.⁵

The outcome illustrates how the framework operates. The court never reached the deference question because the regulation was not ambiguous to begin with. The first step did exactly what the Supreme Court intended: it forced the court to do its own interpretive work rather than defaulting to the agency.

Kisor After Loper Bright

In June 2024, the Supreme Court overruled Chevron deference in Loper Bright Enterprises v. Raimondo, holding that courts may not defer to an agency’s interpretation of a statute simply because the statute is ambiguous.⁶ That immediately raised the question of whether Auer could survive under similar reasoning.

The two doctrines address different things. Chevron governed how courts review an agency’s interpretation of a statute Congress wrote. Auer governs how courts review an agency’s interpretation of a regulation the agency itself wrote. Some legal scholars argue this subsidiary interpretive power survives Loper Bright’s logic, since the agency drafted the regulation and retains authority to explain what it intended.⁷

Loper Bright did not explicitly overrule Auer or address the Kisor framework. It cited Kisor only in passing when discussing interpretive questions that “fall more naturally into a judge’s bailiwick.”⁶ Because a Supreme Court majority upheld Auer on stare decisis grounds only five years earlier, lower courts should not read Loper Bright as having silently overruled it. Commentators have noted, however, that applying Loper Bright’s own stare decisis analysis to what remains of Auer after Kisor suggests the doctrine stands on unstable ground.⁸

What Kisor Means in Practice

For agencies, the decision creates pressure to draft regulations precisely. Vague language that once provided interpretive flexibility is now a liability, because a court will do its own close reading before considering the agency’s view. Agencies that want their interpretations to hold up need to issue them through formal channels, keep them consistent over time, and avoid springing new readings on regulated parties without warning.

For businesses and individuals dealing with federal agencies, Kisor opens more room to challenge agency interpretations. Before the decision, the practical advice was that fighting an agency’s reading of its own rule was usually a losing proposition. Now a challenger can argue that the regulation is not genuinely ambiguous, that the interpretation came from the wrong source within the agency, that it strains the text, or that it contradicts what the agency said last year. Each of those arguments gets a real hearing rather than being brushed aside by a blanket presumption of deference.

Lower courts have applied the framework with varying rigor. Some have used the genuine-ambiguity step to resolve cases entirely on their own interpretive analysis, never reaching the deference question. Others have been criticized for treating the framework as a speed bump rather than the substantial gatekeeping mechanism the Supreme Court described. How consistently courts apply each step will shape whether Auer deference remains a meaningful doctrine or fades toward irrelevance, a question that carries more weight than ever after the demise of Chevron.

1
Justia. Kisor v. Wilkie
2
Oyez. Kisor v. Wilkie
3
Supreme Court of the United States. Kisor v. Wilkie
4
Cornell Law School / Legal Information Institute. Kisor v. Wilkie | Supreme Court Bulletin
5
Justia Law. Kisor v. Wilkie, No. 16-1929 (Fed. Cir. 2020)
6
Supreme Court of the United States. Loper Bright Enterprises v. Raimondo
7
Harvard Law Review. What Loper Bright Might Portend for Auer Deference
8
Yale Journal on Regulation. Auer after Loper Bright, by Chad Squitieri