The 1984 case Chevron v. Natural Resources Defense Council is both the most influential and most controversial case in administrative law. It has come under increasingly strong attacks alleging the doctrine is unconstitutional and in violation of the Administrative Procedure Act. The Supreme Court has applied Chevron less and less in the past few terms, but the lower courts continue to defer under Chevron regularly. To see just how often they do, I recently did a study for the Cato Institute examining how the lower courts are applying Chevron. Although not as deferential as they used to be, the circuits are far more willing to defer than the Supreme Court. While the Supreme Court has not deferred under Chevron in over six years, it remains silent on the constitutionality of the doctrine. But the circuits are not following the Supreme Court’s quiet lead, and they continue to defer. It is time for the Court to clarify the issue and overrule Chevron.
A. The Supreme Court Has Not Applied Chevron in Six Years
Although the Supreme Court has not overruled Chevron, it has not deferred under the doctrine at Step Two in six years. The Court also did not cite Chevron in a majority opinion in the 2021 term, and the Court cited it only three times in 2020. Since Michigan v. EPA and King v. Burwell, the Supreme Court has analyzed ten cases under Chevron. In only one of those, Cuozzo Speed Techs., LLC v. Lee, did the Court continue to Step Two. All the other cases were decided under Step One or under an exception, such as United States v. Mead. Out of the nine cases decided before reaching Step Two, the agency only won twice—in HollyFrontier Cheyenne Refining, LLC v. Renewable Fuels Ass’n and Johnson v. Guzman Chavez—both at Step One. Since the end of the 2015 term, the agency has lost 70% of Supreme Court cases that addressed Chevron. But even looking at only the eight cases that were decided at one of the Chevron steps (rather than an exception), the agency still lost over half the cases—five—all at Step One. Two of the agency wins were at Step One, and only one was at Step Two. The Supreme Court thus rarely finds statutes ambiguous at Step One, and it does not even implicitly defer to the government when interpreting statutes at Step One.
Rather than resorting to Chevron, the Supreme Court has been applying the rules of statutory interpretation even more closely. For example, in one of the two “Becerra” cases from last term, American Hospital Association v. Becerra, the Court concluded that “after employing the traditional tools of statutory interpretation, we do not agree with HHS’s interpretation of the statute.” The two Becerra cases exemplify the Supreme Court’s current statutory interpretation jurisprudence when it comes to agencies. Many thought that either case, especially American Hospitals, would be an opportunity for the Court to overrule Chevron.
But instead, the word “Chevron” does not appear in either decision. In American Hospitals, the Supreme Court applied the “traditional tools of statutory interpretation,” determining that “under the text and structure of the statute, this case is therefore straightforward.” Even though “Chevron” was mentioned 51 times during oral argument, the case was decided purely on statutory interpretation grounds.
Similarly, in Becerra v. Empire Health Foundation (“Empire Health”), the Supreme Court favored the agency’s interpretation because “text, context, and structure all support” HHS’s interpretation. Throughout the opinion, the Court emphasized the plain meaning of the text, noting that “the usual meaning . . . should govern” and that the text “disclose[s] a surprisingly clear meaning—the one chosen by HHS.” The Supreme Court also dismissed one of Empire Health’s arguments by explaining that Congress “would simply have said so,” if it “had wanted to accomplish that unexpected object.” “Chevron” was mentioned 17 times in oral argument, but not once in the opinion. Rather than deferring under Chevron, the Court analyzed the text and structure of the statute, and those are what decided the case.
Instead of focusing on whether there is ambiguity that would trigger Chevron, the Becerra cases are examples of the Supreme Court applying the canons “to find the best reading of the statute,” as then-Judge Kavanaugh urged in Fixing Statutory Interpretation. “[W]hen the text of the statute is clear, a court should not turn to other principles of statutory interpretation such as . . . Chevron deference.”
And though the Supreme Court has emphatically moved away from relying on Chevron in the last term, this development is not novel. The Supreme Court has been steadily applying Chevron less frequently for years. More and more exceptions to the doctrine have cropped up over time, especially in this century: United States v. Mead (interpretation must have force of law); Michigan v. EPA (interpretation must be permissible in light of statutory context as well as text);King v. Burwell (interpretation must not be in an area of “deep political and economic significance”) (internal quotations omitted).
Further, as Justice Gorsuch noted in his recent dissent from denial of cert in Buffington v. McDonough,multiple recent and current justices have critiqued the doctrine, including Chief Justice Roberts and Justices Scalia, Kennedy, Thomas, Gorsuch, and Kavanaugh.
In sum, the Supreme Court has gradually moved away from Chevron, to the point of no longer addressing it even when it is seemingly applicable, and instead applying the canons of construction. Chevron appears to be “effectively buried” at the Supreme Court, though yet without “a tombstone” acknowledging its death, as Justice Gorsuch urged.
B. The Lower Courts Still Apply Chevron Regularly
But if Chevron is buried at the Supreme Court, it is surely out of its grave and roaming like a zombie at the circuit courts. In Justice Gorsuch’s Buffington dissent, he asserted that “courts . . . rarely rely upon [Chevron],” and he called the Federal Circuit’s decision to defer “something of an outlier.” However, this is not the whole picture. The circuits still apply Chevron at a rapid pace, although not as consistently as they used to.
In a study looking at cases from 2003 through 2013, Professors Kent H. Barnett and Christopher J. Walker analyzed 1,327 circuit opinions that applied the Chevron doctrine. Out of all the cases in their study, Barnett and Walker found that agencies won 71.4% of the time. Chevron was applied 74.8% of the time, while courts used the Skidmore test 10.8% of the time. De novo review was granted 7.5% of the time. The remaining 6.9% of the time, the court did not specify what standard of review it applied. When Chevron was applied, agencies won at Step One 11.7% of the time and at Step Two 65.7%. Agencies lost at Step One 18.3% of the time and at Step Two 4.4%.Of the cases that made it to Step Two, 93.8% favored the agency.
Because the Barnett and Walker study ended nearly a year and a half before King v. Burwell, before Michigan v. EPA, and before the Supreme Court cases discussed above, I did my own empirical study of the circuits for the Cato Institute. My survey covered the last two calendar years, from January 1, 2020 through December 31, 2021. I found 142 cases analyzing Chevron. The results were less deferential overall than the Barnett and Walker study, which is unsurprising considering the direction of deference jurisprudence. But unfortunately, the circuits are still much more deferential than the Supreme Court.
I looked at cases either applying Chevron or deciding whether to apply it. I found the circuit courts applied the Chevron steps 84.5% of the time, and 7.0% of cases were decided via Skidmore deference or persuasion, 7.0% received de novo review, and 1.4% were decided on other exceptions. Of all cases studied, the agency won 57.0% of the time and 50% of all cases were decided at Step Two.
Of the cases applying Chevron, 59.2% held that the statute was ambiguous and thus proceeded to Step Two, while 40.8% held that the statute was unambiguous. Among cases that reached Step Two, the agency’s interpretation was held to be permissible 77.5% of the time. Among cases decided at Step One, by contrast, the agency’s interpretation prevailed only 32.7% of the time.
Looking at all cases decided under Chevron, 13.3% were agency wins at Step One, 45.8% were agency wins at Step Two, 27.5% were agency losses at Step One, and 13.3% were agency losses at Step Two. The 45.8% of cases with Step Two agency wins is less than the 65.7% in the Barnett and Walker study, but it is still significantly higher than the 12.5% at the Supreme Court in the past six years.
A comparison of the recent Supreme Court cases, the Barnett and Walker study, and Cato’s very recent study of the circuits shows that the circuits apply Chevron less than they used to, but not nearly as seldom as the Supreme Court. Among cases applying Chevron, only 12.5% of Supreme Court cases (one case) in the last seven terms were decided at Step Two, but the 2003–2013 circuit analysis showed that 70.0% of cases made it to Step Two, and our 2020–2021 study shows that 59.2% of cases were decided at Step Two.
These comparisons show that unfortunately, Justice Gorsuch may have understated the situation when he said “courts . . . rarely rely upon [Chevron].” Circuit courts still find ambiguity 59.2% of the time when Chevron is invoked. Too many courts continue to look for ambiguity rather than “find[ing] the best reading of the statute.” Chevron continues to boldly wreak havoc among the lower courts. Leading by example has proven to be not enough, and the Supreme Court should finally put Chevron to rest and overrule it, providing much needed clarity to the circuits. The Chevron ball needs to end for all, including the circuit courts.
Isaiah McKinney is a Legal Associate at the Cato Institute’s Robert A. Levy Center for Constitutional Studies. He graduated cum laude from Wake Forest University School of Law in 2022. This blog post was largely based on the Cato Institute’s amicus brief in support of a petition of certiorari in Loper Bright v. Raimondo. Isaiah was the main drafter of that brief.
 The first statistic is derived from a Lexis search for “467 U.S. 837” from July 1, 2021 to the present. The second is from a Lexis search for “467 U.S. 837” from July 1, 2020 to June 30, 2021, and excludes Little Sisters of the Poor v. Pennsylvania, since it was decided as part of the 2019 term.
 Empirical data used in this brief was collected from a Lexis search. Cases were limited to those that mentioned Chevron at least 4 times, discussed it in the majority, and analyzed whether to apply Chevron. The data excludes cases that applied Auer/Kisor deference rather than Chevron and cases that were decided without addressing whether Chevron applied. The data is also limited to reported cases.
 The Barnett and Walker study used slightly different search criteria than my study did. One significant difference between my study at Cato and the Barnett and Walker study is that Barnett and Walker treated each instance of statutory interpretation, including multiple interpretations in one case, as separate counts, while I only counted each opinion once, giving controlling weight to the interpretation that deferred.
 See note 2 for search criteria. Cases that involved multiple agency interpretations were counted as only one case.
 See Lexis search discussed in note 1.