Last June, in the wake of the Supreme Court’s decision in Mellouli v. Lynch, I wrote a brief post  responding to Chris Walker’s commentary on the Chevron deference quirk of that case. Although ostensibly concluding that the “controlled substance” ground of deportability was ambiguous and that a reasonable agency interpretation of the statute would be entitled to deference, a majority of the Court held that the Board’s interpretation made “scant sense” and imposed its own limiting construction on the statute. One of Walker’s posited explanations for this strange “application” of deference seemed most compelling to me: “The bizarreness of Chevron ’s invocation in [Mellouli] is more likely attributable to the substantive area of law—immigration—and the soft-stepping of basic administrative law principles that both the Supreme Court and the courts of appeals engage in when the threat of deportation looms.” As I concluded, “[w]here deportation is concerned [] the Supreme Court’s conception of room for deference to the agency’s interpretation of the removability provisions is more circumscribed than it otherwise would be in other administrative law contexts, or even in other areas of immigration law concerning relief from removal or the conferral of benefits.”

Michael Kagan now raises a similar point in the context of the Supreme Court’s recent decision in Torres v. Lynch, focusing more narrowly on the criminal grounds of removability. Although characterizing the strict ruling of the Court—that a state criminal statute need not replicate a federal statute’s jurisdictional element in order to be a categorical match—as a largely unremarkable clarification of how the categorical approach operates, Kagan views the decision’s silence on deference as important and indicative of a general trend away from applying Chevron principles where a criminal ground of removability is at issue. As he writes, “[n]ow with the decision in Luna Torres—where the deference question was a major point of argument—we have a mounting body of evidence that the Court does not in fact defer to the BIA on interpretation of criminal grounds of removal.”

Although I might be expected to express sympathy with Kagan’s position, given my similarly expressed views post Mellouli, it suffers from several shortcomings: 1) Torres is not a compelling specific illustration of a Chevron exception, as deference arguments played only a tangential role in the presentation of the case to the Supreme Court; 2) there is otherwise no evidence of an “exception” to Chevron in criminal removability cases, as the cases the Court has decided either turn on the plain language of the statute or involve federal criminal statutes, where deference never applies; and 3) criminal grounds of removal are not different in any meaningful way from other provisions where deference has been extended, and despite increased attempts to import the criminal rule of lenity into the immigration context, deference is still the more logical approach to interpreting ambiguous statutory language.

First, regardless of whether there is a general trend of the type Kagan suggests, Torres is a poor illustration of this point. Kagan writes that “the question of whether Chevron should apply was a major point of contention,” and that in failing to reference deference in its decision the Supreme Court “seem[ed] to silently reject the Attorney General’s argument” that deference applies. But these points are true in only a very superficial sense. The parties did dispute whether deference should apply if the statute was ambiguous, but both Torres and the government placed principal and overwhelming reliance on the plain meaning of the statute, context, and other normal tools of statutory construction. Put differently, both parties came at the case from a step-one angle, with the deference argument thrown in as a back-up argument should ambiguity be discerned. This is why the government spent 27 pages arguing that the Second Circuit’s decision “is supported by text, structure, and context,” and only four in making its affirmative case that if the statute is ambiguous the Board’s interpretation is reasonable. This is also fairly unremarkable given the exact statutory provision issue, as two of the three courts of appeals that had addressed the exact question presented in Torres found the statute plain on its face—only the Second Circuit had discerned ambiguity warranting deference to the Board’s decision, and even it had opined that it would have applied the same construction even in the absence of deference.

Deference was never a focal point of contention, nor did the Supreme Court have any reason to reach the issue given its conclusion that the prevailing interpretation of the statute was, as argued by the government, dictated by statutory context. Kagan notes the decision’s unsurprising observation that the statute at issue in this case, like virtually every other law passed by Congress, could have been more clear, but this is not a conclusion of ambiguity and instead hearkens back to a statement Justice Thomas made in Kawashima v. Holder, another case where deference played no meaningful role: the statute is clear enough on its face that no recourse to deference principles is necessary.

Second, the “trend” that Kagan identifies does not exist. The Court has only recently begun to see criminal removability cases where deference is potentially at issue, and in Mellouli the Court did address, and seemingly accepted on some level, the applicability of deference. In Moncrieffe v. Holder, Carachuri-Rosendo v. Holder, and Gonzales v. Duenas-Alvarez, no issues of deference were presented because what was at issue were criminal provisions—the Controlled Substances Act in Moncrieffe and Carachuri, and the generic definition of “theft” in Duenas-Alvarez. The agency receives no deference to its interpretation of criminal provisions or to its application of the categorical approach, so the lack of any reference to deference in these cases is simply reflective of this accepted fact. Likewise, in Nijhawan v. Holder and Kawashima v. Holder the government did not assert arguments from deference. Rather, it contended, and the Court accepted, that the plain language of the statute supported the reading adopted by the lower court. The fact is that the Supreme Court has not applied deference in criminal removability cases because none of these cases presented the issue. Its silence is thus not indicative of any trend away from deference.

Finally, as a jurisprudential matter, an exception to deference in this context does not make sense and does not fit with how the Court has applied measures of deference to agency interpretations of dual-use statutes—civil provisions a violation of which may entail subsequent criminal liability. It is not clear what Kagan is arguing in the final two paragraphs of his post. At first blush he seems to argue broadly for an “immigration exception” to Chevron, in part justifying this with the bizarre statement that Board members are not more expert than federal judges in the area of law to which their practice is dedicated. But the Supreme Court continues to apply deference to Board decisions, and has done so explicitly because of its recognition of the expertise the Board brings to bear on interpreting the INA. If Kagan’s point is instead narrower, that deference should not apply within the subset of criminal removability provisions, no compelling reason is given for this carve-out. If a justification for deference is, as the Supreme Court has consistently stated, the Board’s expertise in interpreting the civil provisions of the INA, that expertise is as present where a criminal ground of removability is at issue as it is where a visa-related provision is at issue, as it was in Cuellar de Osorio, a case where Justice Kagan applied a fairly robust conception of deference to uphold the agency’s interpretation. Or perhaps the explicit analogy Kagan draws to the criminal context is short-hand for the argument that deference should not apply at all to dual-use statutes, but rather that courts should review de novo such questions of statutory interpretation consistent with the approach in criminal cases. But no court of appeals has bought that argument, and the Supreme Court has always applied some level of deference to agency interpretations of the statute it is charged with administering, even where that provision may entail criminal liability.

To come full circle and return to Mellouli, there may be situations where the Supreme Court is inclined to apply a lesser measure of deference to Board decisions interpreting the criminal grounds of removability. That seems like what happened in Mellouli itself, but if that is the case it marks the beginning of an as-yet unrealized trend, not the continuation of one. And regardless of whether a trend may be beginning, Torres has nothing to do with it—Torres is a simple statutory interpretation case where the Court’s silence on deference is unsurprising. So, to paraphrase Kagan’s closing, I have little doubt that “the Department of Justice will [] continue to argue that deference should apply,” but it is far from clear that the Department’s “is losing the argument.” The Court has been able to avoid the deference question either because it was not implicated ( Moncrieffe , Carachuri-Rosendo, Duenas-Alvarez ) or because the statute was sufficiently clear to forestall a finding of ambiguity ( Kawashima , Nijhawan, Torres ). A reckoning may have to come sooner or later in a case where a finding of statutory ambiguity of a civil provision of the INA is unavoidable, but until the Court decides that case the Department’s arguments for deference are the more faithful and rational application of extant Supreme Court precedent. And it would certainly be false to argue that such arguments have been foreclosed on the “description” of the Court’s criminal removal cases that Kagan has provided.

Mr. Glen is a Senior Litigation Counsel in the U.S. Department of Justice’s Office of Immigration Litigation, and was on the brief for the government in Torres v. Lynch. The views and opinions expressed in this post are his own, however, and not those of the Justice Department.