The courts can be harsh playgrounds, I know, but defense attorneys must be inventive and intrepid to overcome skepticism and achieve hard-earned victories, especially in the technical arena of the “categorical approach.”  I spend much of my career in criminal-immigrant defense and often deploy the “categorical approach,” challenging many settled notions about what crimes should have immigration consequences, including the last decade-plus pursuing judicial answers to life’s most vexing questions about drugs, like “is ‘cocaine’ really cocaine?” and “what if a drug statute lacks an element of criminal intent?”

This path has led to (in my view) some cheap shots. Several years ago, a case I co-represented was flagged by the Board of Immigration Appeals (BIA) for supplemental and amicus briefing on the question of cocaine definitions. Florida’s “cocaine” statute is differently phrased than the federal definition and the BIA wanted more ink spilled before it determined whether that semantic definition was both scientifically real and legally meaningful. With the help of a PhD chemist and some amicus curiae partners, we waded into the molecular world of “stereoisomers” and “nongeometric diastereomers,” to show that the way that Florida defined cocaine was actually and meaningfully broader than the way that the federal government defined cocaine.  We further showed that Florida prosecutors never care which type of cocaine molecule was found in a drug bust, as they always (as in the case of our client) charge the defendant “generically,” for having had “cocaine.”

In response to this earnest effort, we eventually received a negative decision from the BIA, albeit unpublished, complete with a concurring opinion using the term “absurd,” referencing the television program Breaking Bad (“I am pretty sure that Walter White never asked [this] question”), and concluding that “[a]t some point, common sense needs to prevail, and we are way, way beyond that point.” More frustrating, the concurrence offered “I am no chemist, so the question … is way, way beyond my level of expertise,” as if a statute’s facial complexity should be dispositive of immigration consequences, even when an expert explains that language to a specialized court (the BIA), and where the language at issue was used by a legislature, its members ostensibly not all being chemists themselves.

Appeal of that case to the U.S. Court of Appeals yielded a similar outcome in 2022, this time published, with similar prose.  The court provided what it found to be a helpful metaphor, that our theory was akin to arguing that a criminal code protecting animals might unfairly be applied to acts involving … (wait for it) … a dragon.  To win, apparently, we would have to prove that “dragons” (i.e., non-generic cocaine molecules) exist and that prosecutors lodge cases for criminal acts involving them. Ouch.

Meanwhile in the intervening five years, numerous other courts had “gone there,” acknowledging that it matters if a state defines cocaine, methamphetamine, ethylone, and other drugs in unusual ways. Within a few weeks after our “cocaine” loss, even the Eleventh Circuit applied this logic and agreed that Florida “cannabis” is not the same as federal “cannabis” (as the court could picture the over-inclusivity of the Florida definition). Shortly thereafter, I got a share of a moral (and actual) victory as half of AILA’s amicus team that helped nail down that New York “cocaine” is not “cocaine,” before the Second Circuit, when the Department of Justice moved for remand.

It is no small wonder that the Supreme Court repeatedly has used “burglary” offenses to illustrate the categorical approach, as it is much easier to picture variations of that crime, compared to chemical definitions. In the classic, “elements-based” categorical approach, the courts compare the action (did the “burglary” require an unlawful entry (or remaining) in a structure, that is part of “generic” burglary”?) and/or the locational element (did the state offense include boat trailers or jet skis as potential forums of “burglary,” contrary to the sole generic locations of “structures” and “dwellings”?) to determine if the offense necessarily matches a standard definition. This methodology helpfully explains terms like “theft,” “violence,” “minor,” and “firearm.”  If, on its face, a state criminal statute either misses an element of the generic offense or a prosecution under that statute includes extra options that are broader than the generic offense, the crime does not trigger an immigration consequence. However, if the statutory language does not clearly establish unusual overbreadth, defenders can still win, by identifying state convictions that applied the statute to unusual, non-generic facts, i.e., find the proverbial unicorn.

We do have to always remember that the strict “elements-based” approach is not applicable in every situation. As highlighted by the “fraud” aggravated felony, Courts have held that where the immigration statute establishes a consequence for a crime “involving _______,” the crime is subject to a conduct-based inquiry, in which the consequence can be triggered by reviewing generally what the defendant did, not what the elements of the crime literally established.  In 2000 and 2023, respectively, the Supreme Court reiterated this “conduct-based” approach to statutes involving “involving ______” and apparently expanded it to statutes “relating to ______.”

With so many criminal codes being interpreted by the various circuits, many AILA members devote their practices solely to criminal defense and the categorical approach. Law review articles and entire books are written on the subject. It takes vigilance to stay abreast of the current playing field and monitor “issues to watch” that are percolating before the BIA and Courts of Appeals.   AILA has now dedicated a conference to the subject, with an initial panel of experts teaching the categorical approach in depth, in a model familiar to the immigration bar. However, the remainder of the day will be truly novel: experts will summarize the dominant lines of cases from each Court of Appeal, so that attendees can see how the precedent in each Circuit compares to the cases in others, and gain a sense of trends, and potential copy-cat theories to utilize or buttress against.

As in the case of the “cocaine” example, a theory can spread across the country and become normalized. As I have learned, loss can feel harsh, but the proper perspective is that each case is part of a larger movement, and earnest defense means playing the long game by raising and preserving all available issues. A “theory” is only novel until someone wins with it, at which point it becomes a weapon “for slaying dragons,” in the hands of all defenders.


AILA members interested in learning more about the categorical approach, your opportunity awaits at the Fall Conference: