By Maureen Sweeney
On January 14, 2015, the Supreme Court will hear oral argument in the case of Moones Mellouli v. Eric Holder, in which Mr. Mellouli challenges his removal from the United States for having been convicted of possession of drug paraphernalia (in his case, a sock).
[This essay is part of an online symposium on crImmigration.com addressing Mellouli v. Holder. For other contributions, visit the symposium introduction here.]
The Court has quite recently affirmed, and then reaffirmed, a straightforward application of a categorical approach to assessing the immigration consequences of convictions, in which the elements of a state offense must match exactly or be narrower than the removal category created by the federal generic offense. Moncrieffe v. Holder, 133 S.Ct. 1678 (2013) (discussed on crImmigration.com here), and Descamps v. United States, 133 S.Ct. 2276 (2013) (discussed on crImmigration.com here). As I’ll describe below, the government in Mr. Mellouli’s case does a bit of interpretive gymnastics to try and avoid this analysis, but I think the Court will likely re-reaffirm the categorical approach in this context for all the reasons it has given before.
More interesting to me, as a scholar of the categorical analysis (and a litigator with cases pending at the BIA), is whether the Court will take this opportunity to address the BIA’s recent application of its so-called “realistic probability” test, which can serve to frustrate the categorical analysis.
Mr. Mellouli argues before the Court that Descamps and Moncrieffe require a categorical analysis for determining whether his offense was a removable controlled substance offense under the INA, and that the plain language of the INA dictates that the offense relate to a substance that is listed in federal controlled substance schedules pursuant to 21 U.S.C. § 802, part of the federal Controlled Substances Act (CSA). Because the Kansas schedule of controlled substances under which he was convicted is broader than the federal schedule and the record of conviction submitted by the government in his case did not specify the substance he was accused of holding in his sock, Mr. Mellouli argues that the government has failed to prove that he was necessarily convicted of an offense relating to a federally controlled substance. Under the categorical analysis, his conviction cannot then support removal. He also argues that the categorical analysis does not require him to prove a “reasonable probability” that Kansas will prosecute substances beyond the federal schedule, because the express terms of the state statute establish its reach by incorporating the broader schedule.
The government, on the other hand, reads INA § 237(a)(2)(B)(i) to say that if a state’s drug law “relates” as a whole to substances controlled under § 802, any of the law’s provisions can trigger removability. This reading would avoid applying the categorical approach to drug convictions, and there would be no requirement that the specific offense of which the individual is accused be tied to any specific substance regulated under § 802. Because Kansas enacted a drug law that mirrors and regulates many of the same substances as the federal CSA, creating “an interlocking trellis of federal and state law,” the Kansas statute is a “law … relating to … a controlled substance (as defined by section 802…),” and therefore all of its provisions, including the paraphernalia possession offense, trigger removability. The government also relies on the reasoning of the BIA’s decision in Matter of Martinez-Espinoza which held that convictions for possessing drug paraphernalia are not required to be related to any particular federally controlled substance because they “relate to the drug trade as a whole.” 25 I&N Dec. 118, 121 (BIA 2009).
I don’t expect the Court to abandon its clear, bright-line categorical analysis in favor of either of the catch-all theories offered by the government. The government’s arguments have no support in a plain reading of the statutory language or in the Court’s recent case law. To the contrary, they run directly counter to the rigorous categorical approach that this same Court endorsed less than two years ago in Descamps and Moncrieffe (with 7-member majorities each time). Rather than ensuring that any conviction that could lead to deportation necessarily fit within a clearly defined federal category, the government’s analysis would allow a finding of removability under any provision a state chose to enact as long as the overall network of laws “related” sufficiently to federal drug enforcement efforts. I think the Court will more likely re-reaffirm its categorical approach for all the reasons it has given before – efficiency in the administrative process, constitutional protections, greater consistency, and the will of Congress to authorize deportation for convictions within specified, categorical limits.
The Court should do more, though – it should go beyond simple reaffirmation and take the opportunity to nip in the bud the BIA’s new “realistic probability” test, which has served to confuse and impair application of the categorical analysis.
In applying the categorical approach in the wake of Moncrieffe and Descamps, the Board has developed a requirement that respondents demonstrate a “realistic probability” that a state in fact prosecutes an overbroad statute in a way that falls outside the federal offense category. This “realistic probability” requirement, announced by the Board in September 2014 in the Matter of Ferreira, 26 I& N Dec. 415 (BIA 2014), comes from language in Moncrieffe and the earlier case of Gonzales v. Duenas–Alvarez, 549 U.S. 183, 193 (2007) which cautions that the categorical approach is not an invitation to exercise “legal imagination” to find outlandish but unrealistic ways in which a state statute might be prosecuted outside the federal category. Moncrieffe, 133 S.Ct. at 1685.
The Board has applied its new “realistic probability” test broadly and seemingly beyond the modest scope envisioned in the Court’s language. The Board is now requiring respondents generally to prove actual prosecutions – as a factual matter – even where statutory language indicates that the offense is broader than the federal generic category. For example, in an unpublished decision issued in November, the BIA remanded a case for factual finding on the realistic probability of prosecution because:
[a]ssuming, arguendo, that the respondent’s argument [on overbreadth] is correct, where a State statute on its face covers conduct that is not punishable under the generic definition of the crime, there must be a realistic probability that the State would prosecute conduct under the statute that falls outside the generic definition of the removable offense to defeat a charge of removability under the categorical approach.
In Re: Edwin Alexander Jandres-Aguiluz a.k.a. Edwin Handres, 2014 WL 6883040 (BIA Nov. 13, 2014) (emphasis added). Even where a statute is broader on its face than the federal removal category, the BIA is requiring respondents to present factual proof of actual prosecutions before it will accept the implications of the categorical approach.
Obtaining such proof of actual prosecutions can be enormously difficult as a practical matter – and virtually impossible for unrepresented detainees – because the offenses often involve low-level prosecutions that don’t go to trial and in which there is no factual record created or kept. The many challenges are laid out well in the brief of Mellouli amici curiae National Immigrant Justice Center and American Immigration Lawyers Association. 2014 WL 4804043 (U.S.) (Appellate Brief).
More fundamentally, though, the BIA’s test transforms what is meant to be a streamlined legal determination (a comparison of the elements of the offense of conviction with the generic removal category) into a factual inquiry. This puts immigration courts in the position of having to hold mini-trials and weigh evidence to determine removability – exactly what the categorical analysis is designed to avoid. It also makes inconsistent outcomes more likely, as success in the mini-trial will be affected by the ability of respondents to obtain evidence.
And finally, insofar as it introduces a truly factual component to the categorical analysis, the test also promises to increase the Board’s confusion on the relationship between the categorical analysis and the burden of proof in removal proceedings. The danger here is apparent even in the language of Ferreira. The Board talks about the showing of reasonable probability being required “to defeat a charge of removability.” It is not a respondent’s burden to defeat removability; it is the government’s burden to prove it. And while the Court has repeatedly presented the categorical analysis as a legal determination, the BIA continues to hold that a criminal record that does not – as a matter of law – establish removability can still disqualify a respondent in the relief phase of proceedings. Because it introduces questions of factual evidence and proof to the categorical analysis, the BIA’s new test will undoubtedly further cloud these already muddy waters.
While it may be simple and therefore tempting for the Court to simply restate the need to apply the categorical analysis in assessing the immigration consequences of a conviction, it should avoid unnecessary litigation (not to mention unnecessary detention and deportations) by taking this opportunity to address the BIA’s overzealous use of the “realistic probability” test head-on.
Maureen Sweeney directs the Immigration Clinic at the University of Maryland Carey School of Law. She teaches and writes about the immigration consequences of crimes and about racial profiling and other problems inherent in involving state and local officers in immigration enforcement. She has worked closely with the Maryland Office of the Public Defender in developing their response to Padilla v. Kentucky, 559 U.S. 356 (2010), and she is a founding member of the Maryland Immigrant Rights Coalition. She has created a short animated video explaining the categorical analysis of immigration consequences, which you can see here. She is working on a sequel on the modified categorical approach.