Litigating Mellouli: Adventures in Team Whack-a-Mole

By Kate Evans

In Mellouli v. Lynch, the Supreme Court reached what may now look like an inevitable result when, for the fourth time in a decade, it rejected an attempt by the government to deport a lawful permanent resident for a minor drug offense using the categorical approach. But the case did not start out that way. Rather, the straightforward requirement that “the Government must connect an element of the alien’s conviction” to a federally controlled substance, slip op. at 14, came only as the result of a far-reaching collaboration to winnow down the case.

 

 

Mr. Mellouli—an actuary and college math instructor–was deported based on a conviction for concealing an unnamed Kansas controlled substance in his sock. He was deported under INA § 237(a)(2)(B)(i), 8 U.S.C. § 1227(a)(2)(B)(i), which authorizes the removal of a noncitizen “convicted of a violation of . . . any law or regulation of a State . . . relating to a controlled substance (as defined in section 802 of Title 21), other than a single offense involving possession for one’s own use of 30 grams or less of marijuana.” Section 802 of Title 21, in turn, defines the term “controlled substance” and limits it to the substances included in five federal schedules. 21 U.S.C. § 802(6). At the time of Mr. Mellouli’s conviction, Kansas’ definition of “controlled substance” included a handful of substances not on the federal lists. Mr. Mellouli’s future in the United States therefore hinged on whether a conviction associated with some Kansas controlled substance (perhaps jimson weed, pseudoephedrine, or salvia) was enough to deport him under a law that references only federally controlled substances.

He had lost before the immigration judge who applied the BIA’s decision in Matter of Martinez-Espinoza, which says that paraphernalia crimes involve “the drug trade in general” and thus relate to “any and all controlled substances,” making irrelevant the particular substance involved. Mellouli, slip op. at 10. The BIA affirmed and the Eighth Circuit deferred to Martinez-Espinoza and its application here. Mr. Mellouli had been deported, was separated from his U.S. citizen fiancé, and had no alternative paths to relief. A petition for certiorari was his only hope. On June 30, 2014, the Court granted the request.

Though Mr. Mellouli’s case came in the wake of the Supreme Court’s decisions in Lopez v. Gonzales, Carachuri-Rosendo v. Holder, and Moncrieffe v. Holder, each of which “applied a strict ‘categorical approach’ to the analysis of the noncitizen’s underlying conviction”, the controlled substance ground of deportability at issue here presented three distinct challenges. It does not reference a specific criminal offense; it is not an aggravated felony; and it is grammatically complex. On top of these textual concerns, Kansas’ list of controlled substances and the federal list include most of the same drugs, while the court record (but not the conviction record) contained documents that indicated the sock contained Adderall, a substance controlled under both Kansas and federal law.

Consequently, for Mr. Mellouli and his legal team, of which I was a part, to win we would have to fend off: (1) Chevron deference; (2) a reading of the immigration law that would allow any state drug conviction connected to any substance—federal or not—to support removal so long as the state controlled substance list overlaps with the federal one; (3) Nijhawan’s examination of the specific circumstances surrounding an offense; and (4) Duenas-Alvarez’s test to require the realistic probability that the state applies its criminal statute to conduct that does not match the immigration statute. (See Brief for the United States in Opposition.) We would also need the Court to agree that Mr. Mellouli’s guilty plea to concealing an unspecified substance did not make him removable despite the indication that the substance actually involved may have been federally controlled. In short, the Court would have to apply a strict categorical approach again.

To achieve this would require an army of allies. Thankfully, we had one.

Before certiorari was granted, advocates, coordinated by the Immigrant Defense Project through the Supreme Court Immigration Law Working Group (read about the SCILWG and litigating immigration cases in the Supreme Court here) had put in place an amicus strategy with three main briefs: (1) a law scholar’s brief reinforcing the categorical approach and its application to this statute, (2) a criminal defense attorney brief on paraphernalia charges and plea bargaining, and (3) an immigrant rights brief on the harsh impact of the Eighth Circuit’s decision on noncitizens in removal proceedings. The attorneys writing the briefs further filled out the team. Alina Das, who authored a similar and critical amici brief in Moncrieffe, would represent the law scholars; attorneys at Wilmer Cutler Pickering Hale and Dorr LLP, who had litigated numerous Supreme Court immigration cases, would represent the criminal defense groups; and attorneys at McDermott Will & Emery LLP, who represented a noncitizen challenging the Ninth Circuit’s ruling on the parallel inadmissibility provision (see the cert petition in Madrigal-Barcenas held with Mellouli here) would represent the immigrant rights groups. These advocates supported the merits team, a collaboration of attorneys from the University of Minnesota’s Center for New Americans, the law firm Faegre, Baker, Daniels LLP, the Immigrant Law Center of Minnesota, and Michael Sharma-Crawford, who had represented Mr. Mellouli throughout his proceeding and appeals.

The principle challenge in the case arose from the specter of Chevron deference and the possibility that deference would displace the categorical approach. The categorical approach involves matching the elements of a state conviction to the elements of the offense described in the immigration law. But what about a removal ground that does not reference a discrete crime, like burglary or illicit trafficking, but rather any violation of a law of a state relating to a federally controlled substance? Does the categorical approach’s matching process still apply? To further complicate matters, because the controlled substance ground was not an aggravated felony and did not cross-reference a specific criminal statute, it did not appear in criminal provisions the agency does not administer. Thus, the argument that deference is inappropriate was less clear.

The removal ground does, however, contain the critical statutory hook of conviction. We argued that the word “convicted” operates no differently here than in statutes involving aggravated felonies. ILCM attorney and former University of Minnesota Law School student, Julia Decker, outlined the “historical background . . . demonstrat[ing] that Congress . . . [has] long required a direct link between an alien’s crime of conviction and a particular federally controlled drug.” See slip op. at 8, 13. And the amici brief submitted by nearly 100 law professors reviewed the lengthy history of the categorical approach, its purpose in the administration of immigration law, and the BIA’s decades-long application of that approach to require a federal drug under this removal ground. (See Alina Das’ discussion here and Jennifer Lee Koh’s discussion here.) Critically, these amici also explained away the Board decisions the government invoked as contrary cases.

Because of the uncertainty surrounding the overall analytic framework the Court would apply, the merits brief had to take a two-tiered approach. We advanced a statutory argument rooted in the categorical approach and the term “convicted” along with one using the “traditional tools of statutory construction” referenced in Chevron that would give the “controlled substance” definition referenced in the statute its proper limiting effect even if the Court declined to apply the categorical approach.

Attacking the reasonableness of Martinez-Espinoza under Chevron was greatly facilitated by the Board’s simultaneous, conflicting opinions on the scope of the removal ground—at once requiring a conviction to establish a federal controlled substance for possession and distribution offenses in Matter of Paulus and Matter of Ferreira but disavowing that same requirement for paraphernalia offenses in Martinez-Espinoza. The Eighth Circuit decided that Martinez-Espinoza was reasonable so long as there was “nearly a complete overlap” in the substances controlled by the state and federal laws–an interesting twist on Chevron deference and one that ultimately signaled the decision’s downfall. 719 F.3d at 1000. As the Court correctly noted, under the government’s view “[a] statute with any overlap would seem to be related to federally controlled drugs.” Slip op. at 14. The Eighth Circuit’s attempt to remedy the problem with conditional deference only underscored our argument that the Board’s “drug-trade” exception lacked any basis in the text and, as amici law professors explained, any basis in precedent too. Caitlinrose Fisher, a third-year law student at University of Minnesota, pushed further by distinguishing the breadth of a term from its ambiguity in a survey of circuit court and BIA case law examining the phrase “relating to”. Ultimately, the Court employed a robust analysis of the text and history that requires immigration law to be interpreted “as a symmetrical and coherent regulatory scheme” and rejected the Board’s “anomalous,” “incongruous” results that make “scant sense.” Slip op. at 11. (Read Nancy Morawetz’s analysis of the government’s position here and here.)

Next, we had to sweep away Duenas-Alvarez’s realistic probability test. Kansas criminal courts do not provide their records to national electronic databases and are not searchable by offense, much less substance, making it nearly impossible to identify state prosecutions for non-federally controlled substances. With the help of several public defenders in Kansas, though, we were able to collect examples. An extensive search of those state courts that do make their records available revealed additional examples in other states. We argued the test should not apply based on the statute’s explicit reach and we described the push by state prosecutors to add substances to the Kansas controlled substance schedule along with the extensive publication of the new offenses across the criminal justice system. Amici National Immigrant Justice Center (NIJC) and American Immigration Lawyers Association (AILA), further documented the impossible burden created where few prosecutions are appealed, state court records are often nonexistent or inaccessible, and noncitizens are frequently detained and without counsel to assist in this search. Though, one of the government’s principle arguments in its opposition to cert. relied on the realistic probability test (BIO at 9), by the time the government filed its merits brief, it had relegated the test to a footnote.

While the Court avoids discussion of Duenas-Alvarez, it rejected its application here. Instead, “state-court convictions” must contain an element connected to a federally controlled substance. Slip op. 12-13, 14. And statutes such as that California law that covered one additional substance and the Kansas paraphernalia law that covers nine, are not necessarily linked to a federally controlled substance, slip op. at 9, 15, without regard to “the realistic probability” of prosecution for those substances

Finally, the case required proportion and context. It was, after all, about deporting a mathematician for possessing a sock with unspecified pills. At the same time, Kansas prosecutors abandoned more serious charges that mentioned Adderall when Mr. Mellouli agreed to plead guilty to misdemeanor paraphernalia possession without naming the substance. The plea agreement was the result of the Court’s requirement that defense attorneys counsel their clients on the immigration consequences of criminal convictions. But we were concerned about the Court’s response to the effect of that decision: a plea constructed by the defense attorney and prosecutor to try to shield a noncitizen from immigration consequences. Here, we needed to shift the focus to the systemic role of paraphernalia charges in the plea bargaining process.

Amici National Association of Criminal Defense Lawyers, Immigrant Defense Project, and National Immigration Project of the National Lawyers Guild, explained that paraphernalia possession crimes are minor, broad, and serve as a catch-all offense that allow for the quick resolution of charges without the costs of chemical testing and trial. They also described the minimal procedural protections for noncitizens charged with these offenses who are often pressured into pleas and have few mechanisms to help them assess the immigration consequences of a conviction. The groups warned of the impact on the criminal justice system if defense attorneys cannot predict immigration consequences and protect their clients through the plea-bargaining process. Amici NIJC and AILA explained that defendants charged with paraphernalia possession have no right to counsel in most states and needed a rule that provided clear limits to the reach of the removal ground. These amici also showed that Mr. Mellouli’s offense is so minor it is not criminalized in many states. In the merits brief, student Caitlinrose added analysis demonstrating that most jurisdictions penalized paraphernalia possession akin to marijuana possession – the minor offense Congress explicitly exempted from the removal ground. Third-year Anna Finstrom, along with a team from Faegre, produced a massive appendix to assure the Court that the sky would not fall if Mr. Mellouli wins. They inventoried state crimes for drug manufacturing, distribution, and possession to show that the vast majority would likely require identification of the substance and be analyzed using the modified categorical approach. In other words, even if a ruling for Mr. Mellouli led to fewer deportations for paraphernalia possession, it would impact only the most minor of drug crimes and would benefit the criminal justice system in far more significant ways.

Due to the work of countless advocates, the categorical approach prevailed and the qualifications in Nijhawan and Duenas-Alvarez were further contained. Mr. Mellouli’s conviction, in which “[no] controlled substance (as defined in [§ 802]) figures as an element,” is thus outside the category of removable offenses defined by the federal law. Slip op. at 6, 13. In addition, the Court took the significant step of accepting the use of “safe harbor” pleas, including Mr. Mellouli’s, to guard against immigration sanctions. Slip op. 7 n.5. (See Sejal Zota’s discussion here) The result is a return to the requirement that a state conviction necessarily relates to a federally controlled substance, and with it enhanced protection and predictability for thousands of noncitizens.

Kate Evans is a Teaching Fellow at the University of Minnesota Law School’s Center for New Americans.