Reaffirmation and Clarification of the Categorical and Modified Categorical Approaches

October 27, 2015
By Hanging Out with Carl Gunn


  • A recent Ninth Circuit case – Chavez-Solis v. Lynch – clarifies and reaffirms two important principles about the categorical approach to evaluating prior convictions used for enhancements.
  • First, a defendant does not always have to point to an actual case to show the required “reasonable probability” of an overbroad application of the statute; he may point just to the statutory text if the statutory text is clear.
  • Second, Chavez-Solis reaffirms the holding in Rendon v. Holder that a list of alternatives in a statute makes the statute divisible only if the jury has to unanimously agree on one of the alternatives.



The Ninth Circuit recently published an opinion that reaffirms and/or clarifies some of the general principles of the categorical and modified categorical approaches to evaluating whether a prior conviction qualifies as a predicate conviction for various enhancement statutes or guidelines, like the Armed Career Criminal Act or the career offender or illegal reentry guidelines.  The case is Chavez-Solis v. Lynch, ___ F.3d ___, 2015 WL 5806148 (9th Cir. Oct. 6, 2015).  It’s an immigration case, but recall that immigration cases use the same categorical and modified categorical approaches that criminal cases use, so the opinion applies equally to criminal cases.

The opinion is helpful in several respects.  First, it clarifies a requirement first suggested by the Supreme Court in Gonzales v. Duenas-Alvarez, 549 U.S. 183 (2007), that a defendant pointing to a possible overbroad application of the statute he was previously convicted under must show a “reasonable probability,” rather than just a “theoretical possibility,” that the statute would actually be applied to the conduct the defendant suggests it could apply to.  Duenas-Alvarez, 549 U.S. at 193.  Judges applying this requirement have generally tried to point to actual cases applying the statute to the conduct and fought about whether there really are such cases.  See, e.g., Nunez v. Holder, 594 F.3d 1124 (9th Cir. 2010); Nicanor-Romero v. Mukasey, 523 F.3d 992 (9th Cir. 2008), overruled on other grounds, Marmolejo-Campos v. Holder, 558 F.3d 903, 911 (9th Cir. 2009) (en banc).

I’ve always wondered whether and why a defendant should be required to come up with an actual case if the statutory text clearly includes the conduct in question.  And the Chavez-Solis opinion clarifies that an actual case isn’t required if the statutory text clearly includes the conduct in question.  The opinion explains:

One way a petitioner [or criminal defendant] can show the requisite ‘realistic probability’ of prosecution for conduct that falls outside the generic definition is to ‘point to his own case or other cases in which the state courts in fact did apply the statute in the special (nongeneric) manner for which he argues.’ [Duenas-Alvarez, 549 U.S. at 193.]

     But that is not the only way.  We have explained that if “a state statute explicitly defines a crime more broadly than the generic definition, no ‘legal imagination’ is required to hold that a realistic probability exists that the state will apply its statute to conduct that falls outside the generic definition of the crime.”  United States v. Grisel, 488 F.3d 844, 850 (9th Cir. 2007) (en banc) (citation omitted); see also United States v. Vidal, 504 F.3d 1072, 1082 (9th Cir. 2007) (en banc).  Accordingly, when a “state statute’s greater breadth is evident from its text,” a petitioner [or criminal defendant] need not point to an actual case applying the statute of conviction in a nongeneric manner.  Grisel, 488 F.3d at 850.  The petitioner [or criminal defendant] may simply “rely on the statutory language to establish the statute as overly inclusive.”  Vidal, 504 F.3d at 1082.

Chavez-Solis, 2015 WL 5806148, at *5.

     Chavez-Solis also reaffirms the interpretation of the Descamps divisibility requirement in Rendon v. Holder, 764 F.3d 1077 (9th Cir. 2014), which I’ve discussed in a couple of past posts (see “The Ninth Circuit Agrees With Us on What Divisibility Means Under Descamps” in the October 2014 link at the right and “Rendon Remains Standing” in the April 2015 link).  The Chavez-Solis court held the state statute before it in that case – a child pornography statute – was not divisible despite being written as a disjunctive list of alternatives because (1) “the jury is not required to find that the pornographic materials portray any particular type of sexual conduct [in the list]” and (2) “California juries are not required to unanimously agree on what sexual conduct appears in a particular image.”  Chavez-Solis, 2015 WL 5806148, at *8.

In reaching this last conclusion, the court also made an observation about model jury instructions which we should be able to use in other cases.  In addressing a government argument that the pattern jury instruction suggested divisibility because it had parentheses or brackets around the laundry list of acts that qualify as sexual conduct, the court opined that that revealed nothing more than “that at least one type of sexual conduct must be filled in” and that “this does not tell us that California juries must unanimously agree on which sexual conduct each image contains.”  Chavez-Solis, 2015 WL 5806148, at *8 (emphasis added).  So the use of bracketed alternatives that we often see in model jury instructions doesn’t mean anything.

All in all, Chavez-Solis is a nice addition to our categorical approach case law.  Store it away in your quiver of arrows.