# At the Supreme Court, Chevron Deference Has Morphed into the Application of the Tools of Construction Since the Supreme Court is moving away from relying on Chevron and finding ambiguity, the Court should either overrule Chevron or at least formally adopt Justice Kavanaugh’s test in a majority opinion, thus providing clear guidance to the circuit courts. January 9, 2023 • Commentary By [Isaiah McKinney](https://www.cato.org/people/isaiah-mckinney) This article appeared in *[Yale Journal on Regulation](https://www.yalejreg.com/)* on January 9, 2023. Donald L.R. Goodson’s [excellent response](https://www.yalejreg.com/nc/chevron-continue-to-apply/) to my recent [piece](https://www.yalejreg.com/nc/chevron-ended/) on the circuits’ application of *Chevron* argues that it is up to the Supreme Court to overrule *Chevron*, not lower courts. It is true, of course, that only the Supreme Court can overrule its precedents, and only the Supreme Court can overrule *Chevron*. But there are actually two questions of precedent at issue, not one. One question is whether *Chevron* itself should be overruled. The other is whether the lower courts, when applying *Chevron*, have properly followed Supreme Court precedent on when to find ambiguity. On the issue of ambiguity, in the past seven terms the Supreme Court has found statutory language to be ambiguous in only [10%](https://www.yalejreg.com/nc/chevron-ended/) of *Chevron* cases. The lower courts, however, have found ambiguity in [50%](https://www.yalejreg.com/nc/chevron-ended/) of cases. This disconnect strongly suggests that the Supreme Court and lower courts are, in practice, not applying the same legal tests to determine ambiguity. Either the Supreme Court and circuit courts apply the [“traditional tools of statutory interpretation”](https://www.supremecourt.gov/opinions/21pdf/20-1114_09m1.pdf) differently, or the lower courts are not emphasizing these canons as much as the Supreme Court. Either way, this significant discrepancy points to a methodological difference. This doctrinal difference has been [well documented](https://scholarcommons.sc.edu/cgi/viewcontent.cgi?article=2477&context=law_facpub) for years, [earning the titles](https://repository.law.umich.edu/cgi/viewcontent.cgi?params=/context/mlr/article/1667/&path_info=) of “*Chevron* Supreme” and “*Chevron* Regular” by [Professors Christopher Walker and Kent Barnett](https://repository.law.umich.edu/cgi/viewcontent.cgi?params=/context/mlr/article/1667/&path_info=).[\[1\]](#_ednref1) The Supreme Court rigorously applies the canons of statutory construction, finding ambiguity only when it has truly exhausted those canons. That rigorous approach is itself precedential and binding on the lower courts. But the Supreme Court’s lack of specific guidelines on how to determine whether a statute is ambiguous and how far lower courts must push the canons have unfortunately left the lower courts with little to work with, leading to this discrepancy in the outcomes of *Chevron* cases. This discrepancy is itself a problem. If the Supreme Court does not overrule *Chevron*, it should at least provide clarity on when to find ambiguity. As [then-Judge Kavanaugh urged](https://harvardlawreview.org/wp-content/uploads/2016/06/2118-2163-Online.pdf), rather than quickly resorting to finding ambiguity, courts should “find the best reading of the statute by interpreting the words of the statute, taking account of the context of the whole statute, and applying any appropriate semantic canons.” Since the Supreme Court is moving away from relying on Chevron and finding ambiguity, the Court should either overrule Chevron or at least formally adopt Justice Kavanaugh’s test in a majority opinion, thus providing clear guidance to the circuit courts. I want to highlight three cases that show the disparity between the Supreme Court’s current approach to *Chevron* and the lower courts’ approach. First, in [*Pereira v. Sessions*](https://www.supremecourt.gov/opinions/17pdf/17-459_1o13.pdf) (2018), the [First Circuit concluded](http://media.ca1.uscourts.gov/pdf.opinions/16-1033P-01A.pdf), along with several other circuits, that the immigration statute at issue was ambiguous, and that the Board of Immigration’s (“BIA’s”) interpretation was permissible. The [Supreme Court](https://www.supremecourt.gov/opinions/17pdf/17-459_1o13.pdf), however, declined to defer, instead holding that the statutory language was unambiguous and the BIA’s interpretation was incorrect. In a further critique of deference, Justice Kennedy concurred, urging the Supreme Court to reexamine the “reflexive deference” that many circuits were giving agencies. There are also cases where the Supreme Court could easily have found ambiguity and deferred to make its work easier and quicker, but instead remained silent on *Chevron*. In [*BNSF Railway Co. v. Loos*](https://www.supremecourt.gov/opinions/18pdf/17-1042_7mio.pdf) (2019), the [Eighth Circuit](https://ecf.ca8.uscourts.gov/opndir/17/08/153355P.pdf) had earlier rejected the IRS’s statutory interpretation, determining the statute’s interpretation was unambiguous under *Chevron*. The Supreme Court, however, reversed the Eighth Circuit and decided that the [“IRS’s long held construction”](https://www.supremecourt.gov/opinions/18pdf/17-1042_7mio.pdf) was the correct interpretation of the unambiguous statute. The Supreme Court’s opinion did not even cite *Chevron*. Since the Supreme Court and the Eighth Circuit’s excellent jurists directly disagreed on the statute’s meaning, more deferential courts may have considered the statute ambiguous. Under *Chevron*, the Supreme Court needed only to find ambiguity to reach the same result it ultimately came to and overrule the Eighth Circuit. But rather than deferring when it had the opportunity to do so, the Court performed the necessary statutory analysis and ignored *Chevron*. Similarly in [*Becerra v. Empire Health Foundation*](https://www.supremecourt.gov/opinions/21pdf/20-1312_j42l.pdf) (2022), the [Ninth Circuit vacated](https://cdn.ca9.uscourts.gov/datastore/opinions/2020/05/05/18-35845.pdf) the Department of Health and Human Service’s (“HHS’s”) rule, determining that the statute unambiguously precluded HHS’s interpretation. The [Supreme Court reversed](https://www.supremecourt.gov/opinions/21pdf/20-1312_j42l.pdf) the Ninth Circuit without citing *Chevron*, instead deciding that the statute “disclose\[s\] a surprisingly clear meaning—the one chosen by HHS.” Just as in *Loos*, the Supreme Court could easily have reached the same result by finding the statute ambiguous and deferring. But rather, the Court chose to work through the “\[t\]ext, context, and structure” of the statute to determine its meaning. The Court’s silence in *Loos* and *Empire Health* is not indicative in and of itself of a *Chevron* revolution. The Supreme Court is silent on seemingly salient issues for many reasons. However, the Court’s decision to not defer in both, and its reversal of the circuit court’s deference in *Pereira*, are further signs that the Supreme Court has largely abandoned finding ambiguity—unlike [50% of recent circuit court cases](https://www.yalejreg.com/nc/chevron-ended/)—and instead thoroughly applies the canons of construction. Since it has been [clear](https://harvardlawreview.org/wp-content/uploads/2016/06/2118-2163-Online.pdf) for [many years](https://www.administrativelawreview.org/wp-content/uploads/2014/04/Chevrons-Demise-A-Survey-of-Chevron-from-Infancy-to-Senescence.pdf) that the Supreme Court is not keen on resorting to *Chevron* and finding ambiguity, why do the lower courts continue to proceed past Step One? Goodson says the answer is because they must under *stare decisis*. But I would argue that under *stare decisis*, the circuits must also follow the Supreme Court’s lead and more rigorously interpret statutes rather than deferring reflexively. The circuit court panels in *Loos* and *Empire Health* did just that—they just happened to be wrong in their statutory interpretation according to the Supreme Court. More circuit courts should follow these panels’ example rather than finding ambiguity and proceeding to Step Two. Finally, Goodson notes that one reason it may appear the Supreme Court is deferring less is because the Court hears significantly fewer cases than it used to. This prompted me to compare the number of *Chevron* cases at the Supreme Court with the total number of Supreme Court cases heard per year. I found that even though the Supreme Court hears fewer than half the cases it did when *Chevron* was decided, the average percentage of *Chevron* cases has also gone down. In the twenty-first century, majority opinions discussing *Chevron* have dropped from an annual average of 4.1% of all Supreme Court cases in the 2000s, to 3.0% in the 2010s, down to 1.6% in the start of the 2020s.[\[2\]](#_ednref2) The drop in average percentage of *Chevron* cases further indicates the Court’s increased reluctance to find ambiguity and defer, since in many decisions where the Court finds no ambiguity the Court does not even mention *Chevron*. Therefore, although *Chevron* is still good law, the circuit courts have not followed the Supreme Court’s lead in thoroughly applying the tools of statutory construction before resorting to ambiguity. Though the Supreme Court applies the tools of statutory interpretation exactingly and rarely finds statutes ambiguous, the lower courts have not caught on. Since the Supreme Court is moving away from relying on *Chevron* and finding ambiguity, the Court should either overrule *Chevron* or at least formally adopt [Justice Kavanaugh’s test](https://harvardlawreview.org/wp-content/uploads/2016/06/2118-2163-Online.pdf) in a majority opinion, thus providing clear guidance to the circuit courts. ##### About the Author ##### [Isaiah McKinney](/people/isaiah-mckinney) Legal Associate Show Endnotes Hide Endnotes ##### Notes [\[1\]](#_edn1) This discrepancy between the high Court and the lower courts has significant real-world impact, as litigants can expect very different [results](https://www.yalejreg.com/nc/chevron-ended/) depending on whether they can make it to the Supreme Court or not. [\[2\]](#_edn2) This data was obtained by searching in Lexis for all Supreme Court cases referencing “467 U.S. 837” once and “*Chevron*” at least four times. I discarded all cases that did not reference *Chevron* in the majority. This was the same search criteria I used for my earlier *Chevron*-in-the-circuits study, except this time I did not examine what the Court said about *Chevron*. I then divided the case numbers per year by the number of total Supreme Court cases for the [year](https://harvardlawreview.org/supreme-court-statistics/).