By Nicholas Anderson and Linus Chan
The Supreme Court’s decision on Tuesday to grant cert in the Texas v. United States case has gotten quite a bit of deserved attention among the media, pundits, and immigration advocates and scholars. However, crimmigration nerds should be paying attention to a different Supreme Court cert. grant on the same day. Despite not featuring an immigrant or even immigration law directly, United States v. Mathis, 786 F.3d 1068 (8th Cir. 2015) (docket number 15-6092), will have a significant impact on anyone facing removal from the United States based on a criminal conviction. The Mathis decision will become the latest attempt by the Supreme Court to clarify the often-used but confusingly applied “categorical approach” – a methodology that is used to decide when previous criminal convictions lead to deportation in the immigration context or, in Mathis’s case, a sentence enhancement under the Armed Career Criminals Act (ACCA).
The defendant Mathis was convicted of a felon-in-possession of a firearm under federal law. Under the ACCA, a fifteen-year minimum sentence would be imposed if Mathis had been convicted of three or more “violent” felonies, which by definition under the residual clause of the ACCA included crimes that qualified as burglaries. Mathis rejected the government’s claim that his five Iowa burglary convictions were “violent” felonies because the Iowa statute used a term “occupied structure,” which is separately defined to include vehicles. Breaking into a vehicle does not qualify as a federal “burglary” as the generic definition of burglary only involves buildings. See Taylor v. United States, 495 U.S. 595 (1990).
Most readers of this blog should be familiar with the “categorical approach.” Since the Iowa burglary statute at issue in Mathis included conduct that is not a generic burglary (i.e., breaking into an occupied vehicle), the conviction should not be a burglary unless the statute was “divisible” – if it was, then it would allow a court to examine court records to determine whether Mathis’s convictions involved the subsection of the statute that listed buildings rather than vehicles. If the statute was merely “overbroad” then no matter what the criminal record contained, the inquiry would be over. But how does a court decide whether a criminal statute is divisible or not.
The Elements Approach to Divisibility in Descamps v. United States
In 2013 the Supreme Court in Descamps v. United States, 133 S.Ct. 2276 (2013), rejected a broad definition of divisibility. The Descamps Court described a divisible statute as “a statute [that] lists multiple, alternative elements and so creates ‘several …different crimes.’” Id. at 2285 (quoting Nijhawan v. Holder, 557 U.S. 29, 41 (2009)). The body of the Descamps decision repeatedly refers to “elements” and an “element-centric” approach, but fell short of clearly defining what it meant by the term “element.” This apparent ambiguity has given rise to a circuit split, with the Eighth Circuit in Mathis abandoning a strict “elements-centric” approach and instead allowing statutory “means” or definitions to also be divisible. In order to get to the bottom of the confusion, it is important to dig into the justifications for the elements based approach required by Descamps.
The Court in Descamps refers to “element” as having the same meaning found in other criminal cases, such as Schad v. Arizona, 501 U.S. 624 (1991). The Schad rule states that the “elements” of a crime are the facts that require jury unanimity in order to secure a conviction. This requirement has several important justifications. Most notably, it mandates adherence to the Court’s oft-repeated statement that the fact of “conviction” is all that matters. It also removes all doubt as to what an individual necessarily pleaded guilty to or what a jury necessarily found in its deliberations, alleviating potential Sixth Amendment concerns over the potential for immigration or ACCA consequences. For instance, assume a defendant was charged with possessing a dangerous weapon. Under the statute of conviction, the term “dangerous weapon” is defined as a knife, a firearm, or an explosive. Despite the fact that the indictment alleged the defendant was carrying a firearm, he insisted he only had a knife. Nevertheless, the jury members at trial could disagree whether the defendant had a gun or a knife, but still vote for conviction. (Or the defendant wouldn’t bother at plea to fight about a knife or gun as it results in conviction either way). However, if the element of the statute requires a firearm, any guilty plea would necessarily involve admitting to carrying a firearm. Accordingly, a strict elements-based analysis leads to greater fairness for criminal defendants, and clarifies much of the uncertainty for judges surrounding divisibility.
The Ninth Circuit, the Fourth Circuit and even the BIA have found statutes to be indivisible despite instances of disjunctive phrasing (i.e., separated by a comma or the disjunctive “or”), finding that the phrases merely outlined alternative means of satisfying an element of an offense. Those courts read the Supreme Court’s decision in Descamps and either noticed the ubiquitous references to the word “element” or took seriously the Court’s description of divisibility as requiring a statute to define separate crimes. By “means” the courts are generally referring to “means” in which an element can be met-or when the legislature is trying to define an element itself. Following the Ninth Circuit’s approach in Rendon v. Holder, 764 F.3d 1077 (9th Cir. 2014) (en banc rehearing denied) for example, the term “occupied structure” in Mathis would not be divisible even though a separate Iowa statute defined “occupied structure” by providing a disjunctive list of structures. Because a jury need not agree on the type of occupied structure, it would not be an “element” of the offense.
Footnote Two of Descamps: Furthering the Confusion Between Means and Elements
If the Supreme Court has been so insistent on an elements-based analysis, why do some courts, including the Eighth Circuit in Mathis, continue to find statutes divisible among the terms that seem to merely represent means of committing an offense? Much of the problem can be traced to footnote two of the Descamps opinion, which reads, “And if the dissent’s real point is that distinguishing between ‘alternative elements’ and ‘alternative means’ is difficult, we can see no real world reason to worry. Whatever a statute lists (whether elements or means), the documents we approved in Taylor and Shepherd… would reflect the crime’s elements.” Descamps, 133 S.Ct. at 2285 n.2. The Eighth Circuit hinged its opinion on this footnote, stating “Whether [the list of ‘occupied structures’ in the statute] amount to alternative elements or merely alternative means to fulfilling an element, the statute is divisible, and we must apply the modified categorical approach.” The Mathis decision’s only citation to Descamps was to its footnote two. Essentially, the Supreme Court’s use of the term “elements” without providing a definition, along with the tautological reasoning in footnote two gave the Eighth Circuit room to decide that “means” can render a statute divisible just as “elements” would.
Does footnote two really render statutes divisible as to alternative means, even though the remainder of the Descamps opinion keeps referring to elements? On one hand, the Court in Shepard v. United States, 544 U.S. 13 (2005), did note that it is often quite difficult to determine what a trial court is required to find to obtain a conviction, which arguably lends credit to Mathis’s use of charging documents in the record. It seems unlikely, however, that the Descamps court intended this result for the opinion’s whole purpose is to decide when to use the rule in Shepard.
Rejecting a strict elements approach would seem to renew many of the problems that the Descamps court attempted to fix in terms of trying to splice different phrases of a statute. A strict elements approach however, may not prove easier to apply as it would require courts to dig into jury instructions and case law to figure out what phrases are “elements” and which are merely definitional, or “alternate” means. Mathis has proposed a third approach that would not require courts to decide between elements or means, but rather to decide divisibility based on where the alternative phrases are located. See Petition for Certiorari here. A statutory phrase is only divisible if the alternate wording is contained in the actual statute of conviction and not by reference elsewhere. In his case, the Iowa statute defined “occupied structures” by reference to a separate statutory section then the burglary statute. Mathis argues that this cannot be divisible. The Petitioner’s argument focuses quite a bit on the ease of application of his rule and a means to sidestep the confusion between elements and means.
When it finally turns to deciding Mathis, the Court will have three choices before it: Decide that (1) statutes are divisible only when alternate elements (which require jury unanimity) are listed, or (2) statues may be divisible as long as alternate phrases or disjunctives appear in the statute or by reference, or (3) use a more “appearance” based approach and decide that divisibility can only occur when the alternate phrases are in the actual statute of conviction rather than referenced elsewhere. Regardless of which way the Supreme Court decides, it is unlikely to be the last word on the categorical approach.
Nicholas Anderson is a third year law student enrolled in the Detainee Rights Clinic at the University of Minnesota. Linus Chan is a visiting associate professor of clinical law at the University of Minnesota and the Center for New Americans where he teaches a clinic on representing people who are detained by Immigration and Customs Enforcement.