Welcome back to Ninth Circuit Review-Reviewed, your monthly recap of administrative law before arguably “the second most important court in the land.”

On 28^th March, in a thoughtful concurring opinion to Aguirre Barbosa v. Barr, Judge Marsha Berzon “join[ed] the chorus of voices calling for renewed consideration as to whether the phrase ‘crime involving moral turpitude’ is unconstitutionally vague.” Only a month before, Berzon’s colleague Judge William Fletcher made the same point with a concurrence to Islas-Veloz v. Whitaker.

Here’s the backstory to these similarly-minded concurring opinions. Under immigration laws, an alien can be deported if he or she has been convicted of a “crime involving moral turpitude” (CIMT) within five years of admission, or if she/he is convicted of two CIMTs at any time. Although the statute authorizes the Attorney General to cancel the removal of any alien, this discretion does not extend to anyone convicted of CIMTs in the manner described above. Of course, removal proceedings implicate the highest of stakes—the Supreme Court recognizes that deportation can deprive a person “of all that makes life worth living.” Accordingly, there is a lot riding on the meaning of a morally turpitudinous crime.

Looking to the statutory term, per se, lends little insight; Ninth Circuit panels repeatedly have observed that CIMT “is perhaps the quintessential example of an ambiguous phrase.” For its part, Congress failed to provide any interpretive guidance, as “there are no statutorily established elements for a crime involving moral turpitude.” Instead, the meaning of CIMT “is left to the [Board of Immigration Appeals] and courts to develop through case-by-case adjudication.”

While such a common law approach might appear appealing—at least on paper—the reality has proved anything but. Despite many years of trying, prior Ninth Circuit panels have conceded “the consistent failure of either the [Board of Immigration Appeals] or our own court to establish any coherent criteria for determining which crimes fall within that classification and which crimes do not.”

For Judges Berzon, Fletcher, and like-minded others (including me), it’s a fool’s errand to try to lend coherence to CIMTs in immigration law, because the statutory term is unconstitutionally vague. The Supreme Court, however, held to the contrary in 1951. Out of due respect for stare decisis, Judges Berzon and Fletcher filed their arguments in concurrences, rather than dissenting opinions. But see Lindsay M. Kornegay & Evan Tsen Lee, Why Deporting Immigrants for “Crimes Involving Moral Turpitude” Is Now Unconstitutional, 13 Duke J. Const. L. & Pub. Pol’y 47, 48–49 (2017) (arguing that recent Supreme Court cases have revitalized the doctrine for constitutional vagueness).

Short of offending precedent, I suggest a more modest reform to address the growing concern over CIMTs among judges on the U.S. Court of Appeals for the Ninth Circuit. Specifically, the court should overturn its 2009 en banc decision in Marmolejo-Campos v. Holder, which established that the Board of Immigration Appeals receives binding Chevron deference for its interpretations of CIMTs that are rendered through precedential administrative adjudications. By so doing, the court could lesson considerably the uncertainty surrounding CIMTs in immigration law.

As an initial matter, it bears noting that Chevron is a strange fit for discerning CIMTs. Deference is rooted in agency expertise, but it’s absurd to suggest that the Board of Immigration Appeals is an expert in morality.

In addition, the rule of lenity—the principle that ambiguities in criminal statutes should be resolved in favor of the defendant—militates against Chevron’s applicability for interpreting CIMTs. Although immigration policy is governed by civil statutes, the Supreme Court has recognized that deportation is a “penalty” of a “grave nature,” and, for this reason, analyzed the phrase “CIMT” as though it were part of a criminal statute.

Chevron, moreover, generally is unavailable when the interpretive question broaches constitutional anxieties. From this standpoint, the Chevron doctrine seems particularly inapt in the CIMT context, as deference serves to compound the underlying constitutional concern. On the one hand, the phrase’s vagueness raises due process unease by potentially leaving too much discretion in the hands of agencies. On the other hand, Chevron compels judges to give binding respect to an agency’s policymaking discretion. In this manner, the doctrines complement each other, with the combined effect of enhancing agency power to penalize aliens without forewarning. The upshot is that even if the phrase “CIMT” is not unconstitutionally vague under current Supreme Court precedent, Chevron’s synergistic effect raises a discrete constitutional question—one that courts properly should avoid by declining to employ the Chevron framework.

Apart from these doctrinal considerations, there are policy reasons why the Ninth Circuit should overturn Marmolejo-Campos v. Holder and reject Chevron deference for the Board’s CIMT interpretations. In practice, Chevron deference introduces even greater instability into the already uncertain scope of a CIMT. After all, a galvanizing principle of the Chevron holding is that statutory interpretations are not “carved in stone,” but instead are reappraised by agencies “on a continuing basis.” To this end, the Supreme Court in Brand X identified “a change in administrations” as a viable justification for according Chevron to an agency’s historically inconsistent statutory interpretation. It follows that Chevron encourages politically-driven shifts in statutory interpretation.

Relative to agencies whose power is enhanced by deference, Article III courts provide much greater interpretative stability. In stark contrast to an agency’s discretion to flip-flop under the Chevron doctrine, stare decisis is the “preferred course” for courts, “because it promotes the evenhanded, predictable, and consistent development of legal principles, fosters reliance on judicial decisions, and contributes to the actual and perceived integrity of the judicial process.” Whereas an agency can change its statutory interpretations every four years in response to an election, the Supreme Court upsets its precedent only when confronted with a “special justification.”

In sum, the Chevron doctrine exacerbates the constitutional problems associated with interpreting CIMTs. By giving binding deference to the Board’s identification of crimes involving moral turpitude, the Ninth Circuit linked CIMTs to the vagaries of the electorate, and thus introduced even greater instability into the process of identifying convictions of this type. Primarily for this reason—but also because the Chevron doctrine comports poorly with agency interpretations of CIMTs as a general matter—the Ninth Circuit should look for an appropriate vehicle to convene an en banc panel and overturn Marmolejo-Campos v. Holder.

Notwithstanding the growing sentiment that the statutory phrase “CIMT” is unconstitutionally vague, I can find no instances of a Ninth Circuit judge objecting to the use of Chevron in these circumstances. Indeed, Judge Berzon agreed with the majority in Marmolejo-Campos v. Holder in holding that Chevron “is the correct framework” for reviewing the Board’s interpretation of CIMTs. Respectfully, I recommend that she reconsider this conclusion, and lead the charge against grants of binding judicial respect for the Board’s interpretation of morality.

Dueling Approaches for Managing the Ninth’s Workload

As I’ve noted before, the Ninth Circuit is by far the largest circuit in terms of geographic size and population served. Accordingly, it has by far the largest caseload. The most recent data from 2018 show that over 21 percent of all federal appeals were filed in the Ninth Circuit. The circuit generally handles more than 11,000 cases a year, triple the average for the other regional circuits, and the court’s massive backlog accounts for nearly one-third of all pending federal appeals

Considering these sobering statistics, one possible solution is to expand the court. That’s what the Judicial Conference of the U.S. recommended in its biennial report to Congress, which was delivered in mid-March. To be precise, the Conference advised Congress to add 5 judges to the Ninth Circuit, increasing the circuit’s total to 34 authorized judgeships.

There are, however, possible drawbacks to simply adding judges. In a 2000 empirical study, then-Judge Richard Posner concluded that “adding judgeships tends to reduce the quality of a court’s output” and increase the probability of summary reversal. Posner identified two potential reasons behind his finding:

The first is that federal judges … are “kept in line” if at all mainly by informal norms of judicial propriety and restraint, and informal norms are likely to be more effective the smaller the group in which they are operative, because violations are more easily detectable and repeat transactions (offering opportunities for retaliation against a violator) more likely.

Second, the large size of the Ninth Circuit has led that court, with Congress’s authorization, to adopt a bobtailed en banc procedure. Cases accepted for rehearing en banc are assigned to an 11-judge panel consisting of the circuit’s chief judge plus a randomly selected 10 of the remaining judges in active service. Because of the random assignment of a fraction of judges to the en banc panel, a three-judge panel that decides to defy circuit precedent or otherwise go out on a thin limb has a reasonable prospect of getting away with it. Even if rehearing en banc is sought and granted, the luck of the draw may result in an en banc panel’s being dominated by the original panel’s members and their allies, although the probability that all three of the judges on the original panel will be on the en banc panel is only about .05.

If Posner is right, then the Ninth Circuit’s jurisdiction is too big for its number of judges, but, at the same time, the court already has too many judges. As a result, the Judicial Conference’s recommendation to add 5 judges would risk increasing efficiency at the expense of quality. Under Posner’s reasoning, a better recommendation would be to both add judges and simultaneously break up the Ninth Circuit, so that the progeny courts born from the Ninth Circuit would correspond to the average size and workload of sister circuits.

In an excellent new paper, my colleagues Ilya Shapiro and Nathan Harvey make an apolitical case for breaking up the Ninth Circuit so as to better regulate its workload. While I recommend the entire paper, I was particularly struck by their observation that there is nothing new to managing judicial caseloads by splitting circuit courts. The most recent such division occurred in 1980, when Congress split the Fifth Circuit in response to many of the same issues currently facing the Ninth Circuit.

William Yeatman is a research fellow at the Cato Institute’s Levy Center for Constitutional Studies.