One Last Piece (at Least for Now) of Good News on the Modified Categorical Approach

October 23, 2012
By Hanging Out with Carl Gunn


  • A recent Ninth Circuit opinion narrowly reads Aguila-Montes de Oca
  • Only facts necessary to the factual basis for the plea on which a prior conviction is based can be considered under the modified categorical approach
  • It’s not what the defendant did but what he was convicted of that matters.


Before I leave the subject of the modified categorical approach that I’ve talked about in my last two posts, there’s one more case to mention that can be categorized as good news. It was actually brought to my attention by an e-mail circulated by DFPD Davina Chen in our district. It’s the recent Ninth Circuit opinion of Aguilar-Turcios v. Holder, 691 F.3d 1025 (9th Cir. 2012), which places some nice limits on – or at least strictly applies – the expanded reading of the modified categorical approach in the Aguila Montes de Oca case that I discussed back in April and also two posts ago (see “They May Reverse the Ninth Again! (But This Time We Want It.”) linked at the right).

What Aguilar-Turcios does that’s helpful (though it may not be needed if the Supreme Court rules in our favor in the Descampscase that I pointed out in the post two weeks ago), is emphasize language and reasoning in Aguila-Montes de Oca that places some significant limits on what facts and admissions can be considered. It started with emphasizing some of the limiting language in Aguila-Montes de Oca:

We counseled in Aguila-Montes that we “must exercise caution” when determining upon what facts a prior conviction necessarily rested. Id.[, 655 F.3d at 937.] “It is not enough that an indictment merely allege a certain fact or that the defendant admit to a fact.” Id. Instead, to conclude that “the factfinder necessarily found the elements of the generic crime,” the modified categorical approach requires that “the defendant could not have been convicted of the offense of conviction unless the trier of fact found the facts that satisfy the elements of the generic crime.” Id.

Aguilar-Turcios, 691 F.3d at 1032-33 (emphasis in original).

The court then went on to emphasize in particular the requirement that the fact found or admitted be necessary to finding a factual basis. It stated that “[u]nder Aguila-Montes, we first examine what facts were ‘necessary’ to support Aguilar-Turcios’ Article 92 conviction.” Aguilar-Turcios v. Holder, 691 F.3d at 1034. It stated that “even if we could look to Aguilar-Turcios’ Article 134 conviction [in the same plea proceeding], none of the factual admissions that related to the Article 134 charge were ‘necessary’ to support his Article 92 conviction.” 1037. (The prior convictions being considered in the case were two military court martial convictions, only one of which was alleged to justify deportation, which was what Aguilar-Turcios was about.)

The court then reiterated that “[t]he modified categorical approach limits our inquiry to the ‘facts the conviction necessarily rested on.’” Id. at 1038 (quoting Aguila-Montes de Oca, 655 F.3d at 940). It explained that in the case at bar that “would mean that unless the [judge] found facts satisfying the elements of [18 U.S.C.] § 2252(a)(2) or (a)(4) [which were the generic offenses at issue], Aguilar-Turcios could not have been convicted of violating UCMJ Article 92.” Aguilar-Turcios, 691 F.3d at 1038 (emphasis in original). It then held that wasn’t the case there because “the Article 92 charge says nothing about minors, and it was not necessary for the [judge] to find that minors were involved in order to convict Aguilar-Turcios.” Id. at 1038-39.

It’s a little hard to follow the case-specific analysis without getting into a lengthy discussion of the facts and statutes at issue – which would make this post far too long – but there are two things this case gives us. First, it gives us a basis for really emphasizing that the fact found or admitted at the plea has to benecessary to the factual basis, since that word “necessary” gets used in Aguilar-Turcios no less than four – and actually more than four – times. Second, if you want to take a judge or a probation officer through the facts of Aguilar-Turcios, it’s a nice illustration of how that requirement places significant limits on the modified categorical approach even if the approach isn’t limited to “divisible” statutes, as the court held in Aguila-Montes de Oca (but the Supreme Court may not hold in Descamps).

Also interesting in Aguilar-Turcios are a couple of comments in Judge Bybee’s dissent that suggest the underlying frustration of some judges about not applying an enhancement even though “everyone knows” what the defendant “really did.” I’ve often thought that the reason some judges try to twist the categorical and modified categorical approaches to apply more expansively is their view that the defendant’s getting off easier than he should because the possible way of committing the crime that takes it outside the generic federal definition is a relatively uncommon way of committing the crime. Judge Bybee complains that “[e]veryone knows what Lance Corporal Aguilar-Turcios was looking at on his computer,” id., 691 F.3d at 1042 (Bybee, J., dissenting); that the other possibilities the majority suggests are “contrary to all we know,” id.; that “[t]here is no need to speculate” and “even counsel at oral argument conceded,” id. at 1051; and that “[t]here is no doubt” about what the respondent possessed, id. at 1052.

The majority opinion then contains a nice response to these complaints, which is consistent with the idea (qualified a little by the relevant conduct rules in the sentencing guidelines, I’m afraid) that we sentence people based on what they’re convicted of doing, not what they’re alleged to have done. The majority explains:

[W]e disagree with the dissent’s vague “flexible” approach. First, “the concern motivating the entire categorical/modified categorical exercise” is not, in fact, that we “be certain of the conduct that an alien committedbefore the alien is removed,” Dissent[, 691 F.3d at 1053] (emphases added), but instead that we “focus on what the defendant was convicted of rather than the acts hecommitted.” Aguila-Montes, 655 F.3d at 928 (emphases added). This is not a minor distinction but the linchpin of the categorical and modified categorical approaches.

Aguilar-Turcios, 691 F.3d at 1033. The majority then adds, in a footnote: “Again, what occurred is not the focus of our inquiry under the categorical or modified categorical approach. Instead, we focus only on the offense of which the defendant wasconvicted.” Id. at 1033 n.11 (emphasis in original). Though not specifically noted in the majority opinion, this is also the explanation given by the Supreme Court’s opinion in Taylor v. United States, 495 U.S. 575 (1990), which is the genesis of the categorical approach. The Court construed the language of the statute at issue there – the Armed Career Criminal Act – as “support[ing] the inference that Congress intended the sentencing court to look only to the fact that the defendant had been convicted of crimes falling within certain categories, and not to the facts underlying the prior convictions” and “most likely refer[ring] to the elements of the statute of conviction, not to the facts of each defendant’s conduct.” Id. at 600-01.

In any event, Aguilar-Turcios is a third piece of good news on the modified categorical approach to keep in mind in your fights in this area. Next week we’re off – or perhaps back – to other topics.