More on Those Great New Categorical Approach Cases

February 2, 2016
By Hanging Out with Carl Gunn


  • United States v. Garcia-Jimenez, 807 F.3d 1079 (9th Cir. 2015), provides a nice summary of the sources to look to in establishing the federal generic definition of a crime for purposes of the categorical approach.
  • Garcia-Jimenez and United States v. Dixon, 805 F.3d 1193 (9th Cir. 2015), also nicely illustrate the sort of investigation and research you need to engage in to apply the categorical approach, including both examination of the universe of state statutes and researching case law on how they’ve been interpreted and applied.
  • Dixon also reaffirms and provides a nice practical explanation of Rendon’s holding that an alternative is an “element” for purposes of divisibility under Descamps only if there has to be jury unanimity about the element.



I ended last week’s post about the opinions in United States v. Dixon, 805 F.3d 1193 (9th Cir. 2015), and United States v. Garcia-Jimenez, 807 F.3d 1079 (9th Cir. 2015), by noting that they also clarify, illustrate, and reaffirm some general principles about the categorical approach and what we have to do in investigating and developing potential categorical approach challenges.  They do this in a number of ways.

To begin, Garcia-Jimenez summarizes “a number of sources” courts should survey “to establish the federal generic definition of a crime” – “including state statutes, the Model Penal Code, federal law, and criminal law treatises.”  Id., 807 F.3d at 1084.  The opinion goes on to state that “[m]ost often, ‘[t]he generic definition of an offense “roughly corresponds to the definitions of the offense in a majority of the States’ criminal codes.”’” Id. (quoting United States v. Garcia-Santana, 774 F.3d 528, 534 (9th Cir. 2014), and Taylor v. United States, 495 U.S. 575, 589 (1990)).  It further cautions that “the Model Penal Code, while a helpful tool in the categorical analysis, does not dictate the federal generic definition of a crime.”  Id., 807 F.3d at 1086.  Garcia-Jimenez then bases its ultimate holding on “[t]he weight of authority – approximately two-thirds of the states, the common law, federal law, and at least one treatise, as compared to the Model Penal Code and one-third of the states.”  Id. at 1087.

Garcia-Jimenez – and Dixon – also illustrate the sort of investigation in which we need to engage to investigate and make categorical approach challenges.  In order to calculate the two-thirds/one-third division in the states on which it relied, Garcia-Jimenez had to examine every one of the 50 state aggravated assault statutes – no doubt with the help of the briefing of the defense attorney in that case, our esteemed panel colleague Davina Chen.  See Garcia-Jimenez, 807 F.3d at 1085-86 nn. 5, 6, 7.  And in instances where the state statutes weren’t clear on their faces, Garcia-Jimenez needed to look at case law interpreting the statutes.  Id. at 1085-86 n.5.  Indeed, while it may not have done so, the court arguably needed to conduct this case law research for every state because language that seems clear on its face doesn’t always mean what it seems to mean.  (For a more in depth discussion of the need to be careful in this respect, see one of my early posts long ago – “More on the Categorical Approach for Applying Federal Prior Conviction Enhancement Statutes,” in the April 2012 link at the right.)

This need to look to the case law is another reminder provided by Garcia-Jimenez and DixonGarcia-Jimenez illustrates the use of case law to reveal that language similar – indeed, identical – to language in the generic definition isn’t dispositive – by noting language in the New Jersey attempt statute that’s identical to the generic federal definition but pointing to case law that reads the language more broadly.  See id., 807 F.3d at 1088.  Dixon illustrates the use of case law to provide examples of a state statute’s broad application – by pointing to a California Supreme Court case applying the California robbery statute to conduct outside the definition of “violent felony.”  See id., 805 F.3d at 1197-98.  Dixon also reminds us that finding just one such example is enough.  As the court explained after discussing this one example:

We acknowledge that the facts underlying the vast majority of convictions under CPC § 211 [the California robbery statute] likely meet the [ACCA definition of “violent felony”].  However, Taylor’s categorical approach is an “elements-based inquiry,” not an “evidence-based one.”  Descamps [v. United States], 133 S. Ct. [2276,] 2287 [(2013)].  Because Anderson [the California Supreme Court case] shows that one can realistically violate CPC § 211 in a manner that is not covered by the ACCA’s definition of “violent felony,” a violation of CPC § 211 is not categorically a “violent felony” under the ACCA.

Dixon, 805 F.3d at 1197-98.

Dixon also reaffirms and further elucidates the holding in Rendon v. Holder, 764 F.3d 1077 (9th Cir. 2014), that a statute’s listing of alternatives in the disjunctive does not automatically make the statute “divisible” and so subject to the “modified categorical approach” under the Supreme Court’s Descamps opinion.  Dixon reminds us the court “must determine whether a disjunctively worded phrase supplies ‘alternative elements,’ which are essential to a jury’s finding of guilt, or ‘alternative means,’ which are not.”  Id., 805 F.3d at 1198 (quoting Rendon, 764 F.3d at 1085-86).  It then applies that requirement to two disjunctive phrases in the California robbery statute – that the property be taken from the “person or immediate presence” and that the property be taken by “force or fear” – and explains in a very practical way why these phrases set forth alternative means rather than alternative elements:

The disjunctively worded phrases in the statute and jury instructions are alternative means, not alternative elements.  To return a guilty verdict on a CPC § 211 charge, a jury must find that the elements are satisfied, but jurors need not agree on the disjunctively worded alternatives.  For example, jurors must agree that element (3) [the “person or immediate presence” element] is met – but the jury can return a guilty verdict even if some jurors believe the defendant took property from the victim’s person and other jurors believe the defendant took the property from the victim’s immediate presence.  Similarly, for element (5) [the “force or fear” element], a jury can return a guilty verdict even if some jurors believe the defendant used force and other jurors believe the defendant used fear.

Dixon, 805 F.3d at 1198.

Keep in mind a caveat about Dixon – and the other Rendon progeny – however.  The approach of Rendon and its progeny is now before the Supreme Court in Mathis v. United States, ___ S. Ct. ___, 2016 WL 207258 (Jan. 19, 2016).  The lower court in Mathis disagreed with Rendon and enhanced a split in the circuits that the Supreme Court has now decided to resolve – hopefully in our favor.

In any event, subject to whatever the Supreme Court decides in Mathis, remember the principles articulated and illustrated in Garcia-Jimenez and Dixon when you consider a prior conviction under the categorical approach.  It’s surprising what you can come up with.