United States v. Escalante

JENNIFER WALKER ELROD, Circuit Judge

Appeal from the United States District Court for the Northern District of Texas Before SMITH, WIENER, and ELROD, Circuit Judges. JENNIFER WALKER ELROD, Circuit Judge:

Johnny Escalante failed to register as a sex offender when he travelled to Texas. The district court concluded that his prior Utah conviction for unlawful sexual activity with a minor classified him as a tier II sex offender, and he was sentenced based on the corresponding Guidelines range. Because the district court deviated from the categorical approach to classify him as a tier II sex offender, we VACATE and REMAND for resentencing.

I.

The Sexual Offense Registration and Notification Act of 2006 (SORNA) requires sex offenders to update their registration after a change in residence. See 34 U.S.C. § 20913(c). Failing to do so is a federal crime when the offender travels in interstate commerce. See 18 U.S.C. § 2250. Section 2A3.5 of the Guidelines provides three base offense levels when a sex offender is found guilty of failing to register. Those levels correspond with the sex offender tiers in 34 U.S.C. § 20911. Relevant to this case, a tier II sex offender is someone "whose offense . . . is comparable to or more severe than the following offenses, when committed against a minor[:] . . . abusive sexual contact (as described in section 2244 of title 18)[.]" 34 U.S.C. § 20911(3).

Pub. L. No. 109-248 (codified at 34 U.S.C. § 20911 et seq.).

E.g., If the defendant was a tier III sex offender, the base offense level for failing to register is 16. If tier II, then 14; if tier I, then 12. U.S.S.G § 2A3.5 (U.S. Sentencing Comm'n 2017).

In 2010, Escalante was convicted in Utah for unlawful sexual activity with a minor. At the time of the offense, Escalante was 35 years old and the victim was 14. After being released from prison, Escalante travelled to Texas and failed to update his registration. He was subsequently identified by law enforcement during an unrelated traffic stop and charged for failing to register as a sex offender. He pleaded guilty. In the factual resume that he signed as part of his plea agreement, he admitted that: (1) he was required to register as a sex offender due to the 2010 Utah conviction; (2) he travelled to Texas; and (3) he knowingly failed to update his registration.

See Utah Code Ann. § 76-5-401 (2010) (criminalizing sexual activity with a person aged 14-15 under circumstances not amounting to, inter alia, rape or aggravated sexual assault).

The Pre-Sentence Report (PSR) concluded that Utah's crime of unlawful sexual activity with a minor was comparable to abusive sexual contact as described in 18 U.S.C. § 2244, and therefore recommended that Escalante be categorized as a tier II offender with a Guidelines imprisonment range of 27-33 months. However, the PSR also urged the court to consider an upward departure based on Escalante's history of domestic violence, parole violations, and high risk of recidivism.

18 U.S.C. § 2244 cross-references several other statutes. Of those, only 18 U.S.C. § 2243—sexual abuse of a minor or ward—arguably corresponds with Utah's crime of unlawful sexual activity with a minor. 18 U.S.C. § 2243(a) criminalizes sexual acts with someone who is (1) 12-15 years old and (2) at least four years younger than the offender.

Escalante objected to the PSR, arguing, as relevant here, that the Utah statute "sweeps more broadly than the federal statute" and therefore, under the categorical approach, cannot serve as a predicate for classification as a tier II offender. Specifically, Escalante pointed to the facts that: (1) 18 U.S.C. § 2243(c)(1) permits an affirmative defense if the defendant reasonably believed the victim to be over 16, whereas Utah's relevant statute did not; and (2) 18 U.S.C. § 2243(a)(2) requires the government to prove a four-year age differential, whereas Utah's relevant statute did not.

Instead, Utah considered the four-year age differential a mitigating factor which, if established by the defendant, will reduce the crime to a misdemeanor and not require registration. Compare Utah Code Ann. § 76-5-401(3)(b) (2010) with 18 U.S.C. § 2243(a)(2).

At sentencing, the district court overruled Escalante's objections, adopted the PSR as its factual findings, and upwardly varied from the Guidelines to sentence Escalante to 48 months' imprisonment. Escalante repeated his objections at sentencing and timely appealed. We have jurisdiction to review Escalante's sentence under 18 U.S.C. § 3742.

II.

"For properly preserved claims, this court reviews the district court's interpretation and application of the Sentencing Guidelines de novo." United States v. Young, 872 F.3d 742, 745 (5th Cir. 2017) (citation omitted).

III.

We employ the categorical approach when classifying the SORNA tier of a defendant's state law sex offense. See Young, 872 F.3d at 746 (joining four other circuits in employing the categorical approach in such cases). "Under the categorical approach, the analysis is grounded in the elements of the statute of conviction rather than a defendant's specific conduct." United States v. Rodriguez, 711 F.3d 541, 549 (5th Cir. 2013) (en banc), abrogated on other grounds by Esquivel-Quintana v. Sessions, 137 S. Ct. 1562 (2017). If the statute of conviction "sweeps more broadly" than the referenced federal offense, the state offense cannot serve as a proper predicate. See Descamps v. United States, 570 U.S. 254, 261 (2013).

On appeal, Escalante repeats the objections he made to the PSR, arguing that the district court erred by not considering that the Utah conviction offense and the corresponding federal offense had different possible affirmative defenses. He also argues that the district court erred by considering the specific circumstances of his offender-victim age differential when categorizing his sex offender tier level. In response, the government contends that the district court did not err, but, even if it did, that any such error would be harmless. We address each argument in turn.

1.

First, we address Escalante's affirmative defenses argument. Escalante observes that 18 U.S.C. § 2243(c)(1) permits an affirmative defense if a defendant can establish, by a preponderance of the evidence, that he reasonably believed that the other person was at least 16 years old. In contrast, the Utah statute that he was convicted under provided no such affirmative defense. See Utah Code Ann. § 76-5-401 (2010).

Escalante argues that whether that affirmative defense was available "reflects an enormous difference in culpability" because it distinguishes between intentional and unintentional conduct. He argues that because a mental state defense exists for one offense and not for the other, the two offenses reach "significantly different" classes of offenders. Specifically, he argues that § 2243 offenders are more culpable and blameworthy as a class than § 76-5-401 offenders because they all knew, or should have known, that the victim was under 16. Therefore, he asserts, it is immaterial whether the defendant's mental state is an element that needs to be proven by the government or an affirmative defense that needs to be proven by the defendant. Either way, Escalante argues, the class of people who are § 76-5-401 offenders could include people with less culpability than the class of people who are § 2243 offenders, and, therefore, § 76-5-401 cannot serve as a predicate for classifying him as a tier II offender under the categorical approach.

Escalante cites United States v. Roebuck, 2015 WL 13667427 (D.N.M. Jan. 26, 2015) (unpublished), as supporting case law. In Roebuck, the district court addressed this question when determining the tier level of a Texas statutory rape offense that, like Utah's, did not include an affirmative defense for reasonably believing the victim to be 16. Id. at *5-6. Without offering much analysis, the Roebuck court concluded that because § 2243 permits an affirmative defense that the Texas law did not, the "Texas statute sweeps more broadly than the federal statute" and could not serve as a predicate offense for classification as a tier II sex offender. Id. at *6.

We reject Escalante's argument. The Supreme Court has repeatedly articulated that the categorical approach looks exclusively to the elements of the offenses to be compared. See, e.g., Mathis v. United States, 136 S. Ct. 2243, 2248 (2016) (directing that courts "focus solely on whether the elements of the crime of conviction sufficiently match the elements of [the predicate crime]" (emphasis added)); Descamps, 570 U.S. at 261 (holding that courts "may look only to the statutory definitions—i.e., the elements—of a defendant's prior offenses" (emphasis added, citation and quotation marks omitted)); Taylor v. United States, 495 U.S. 575, 602 (1990) (holding that courts "look only to the fact of conviction and the statutory definition of the prior offense" (emphasis added)).

The Supreme Court has defined "elements" in this context to be "the 'constituent parts' of a crime's legal definition—the things the 'prosecution must prove to sustain a conviction.'" Mathis, 136 S. Ct. at 2248 (quoting Black's Law Dictionary 634 (10th ed. 2014)). Despite Escalante's assertion to the contrary, it is black letter law that an affirmative defense (or the absence thereof) is not the same thing as an element of the crime. See generally Martin v. Ohio, 480 U.S. 228, 234-35 (1987) (discussing affirmative defenses and the elements of a crime as different things, and holding it constitutional for the defendant to bear the burden of proving the former but not the latter). Moreover, we agree with the government that the unpublished district court opinion in Roebuck fails to offer any appreciable analysis on this point and so we give its holding to the contrary little weight.

This conclusion aligns with that of at least two of our sister circuits. See United States v. Velasquez-Bosque, 601 F.3d 955, 963 (9th Cir. 2010) ("The availability of an affirmative defense is not relevant to the categorical analysis."); Donawa v. U.S. Attorney Gen., 735 F.3d 1275, 1282 (11th Cir. 2013) (rejecting a comparison under the categorical approach when the mens rea component was an element of the federal statute, but its absence was an affirmative defense in the state statute because "[a]n affirmative defense generally does not create a separate element of the offense that the government is required to prove in order to obtain a conviction"). Conversely, we have not identified any circuits expressly holding that different affirmative defenses should be considered under the categorical approach.

As a secondary point, we note the practical difficulty that would accompany expanding the categorical approach to include consideration of all the different permutations of potential affirmative defenses. One of the reasons that the Supreme Court gave for implementing the categorical approach in the first place was that the "practical difficulties" of the alternative were deemed to be "daunting." See Taylor, 495 U.S. at 601-02. However, requiring an examination of all the affirmative defenses that could have possibly been raised under state law for someone charged with a state crime (even if, like here, the defendant personally could not assert the defenses, and thus they were not raised in the proceedings below), and then trying to align all those possible defenses with the defenses permitted under federal law would make applying the categorical approach an even more daunting task than it already is. Escalante argues that the expansion he seeks would only apply to this defense for this specific offense, but, as a matter of logic, it is hard to see why it would stop there.

But see Lewis v. Chicago, 560 U.S. 205, 217 (2010) ("[I]t is not our task to assess the consequences of each approach and adopt the one that produces the least mischief. Our charge is to give effect to the law Congress enacted.").

For those reasons, we reject Escalante's argument to consider different permissible affirmative defenses, and we consider only the elements of the offenses to be compared when applying the categorical approach.

2.

Second, we address Escalante's argument that the district court erred by considering the specific circumstances of his offender-victim age differential when determining his sex offender tier level.

Escalante argues that Utah's § 76-5-401 offense sweeps more broadly than "abusive sexual contact (as described in section 2244 of title 18)"—which cross-references to 18 U.S.C. § 2243(a)—because § 2243(a) requires, as an element, that the government prove at least a four-year age differential. See 18 U.S.C. § 2243(a)(2) (requiring the government to prove the victim "is at least four years younger than the person so engaging"). Utah's § 76-5-401, as it existed in 2010, did not require the government to prove any age differential. The Utah statute still took into account whether the offender was within a four-year age range, but made that a mitigating factor which, if proven by the defendant, reduced the crime to a misdemeanor. See Utah Code. Ann. § 76-5-401(3)(b). Looking solely at their elements, the Utah statute criminalized consensual sexual activity between an 18-year-old and a 15-year-old, but the federal statute does not. Escalante therefore asserts that applying the categorical approach means that the Utah offense is overly broad and cannot be considered comparable to § 2243(a) (as cross-referenced by § 2244) for the purpose of classifying him a tier II sex offender.

Though it does not impact the analysis, Utah's statute has since been amended to only apply to persons 18 and older who engage in sexual activity with persons aged 14-15. Thus, Utah's law does in fact now require the government to prove an age differential of at least two years (which is still smaller than the four-year age differential required under 18 U.S.C. § 2243).

The government responds by noting that this court has expressly left open the question of whether the categorical approach should be used when considering the specific circumstances of the victim's age for SORNA tier classifications. See Young, 872 F.3d at 747-48 ("[T]his opinion should not be read as holding that the categorical approach applies . . . when it comes to the specific circumstance of the victims' ages. . . . We save discussion of any argument on [that] point for a day when it is properly raised."). The government points out that at least three other circuits have adopted a circumstance-specific approach when considering the victim's age for the purpose of SORNA's §§ 20911(2)-(4) tier classifications. See United States v. Berry, 814 F.3d 192, 197 (4th Cir. 2016) ("[T]he language of Section 16911(3)(A), like the language of Section 16911(4)(A), instructs courts to apply the categorical approach when comparing prior convictions with the generic offenses listed except when it comes to the specific circumstance of the victims' ages."); United States v. White, 782 F.3d 1118, 1135 (10th Cir. 2015) ("[W]e conclude Congress intended courts to apply a categorical approach to sex offender tier classifications designated by reference to a specific federal criminal statute, but to employ a circumstance-specific comparison for the limited purpose of determining the victim's age."); United States v. Byun, 539 F.3d 982, 991 (9th Cir. 2008) ("[A]s to whether an individual is a 'tier II' offender, the language of the statute points strongly toward a non-categorical approach with regard to the age of the victim."). Conversely, the parties do not brief, and we are not aware of, any circuits to expressly hold the opposite.

We now address the question we left open in Young—does SORNA require courts to perform a circumstance-specific inquiry to determine whether the victim was a minor when applying the categorical approach to classify sex offender tier levels? In Nijhawan v. Holder, the Supreme Court recognized that federal statutes may impute the categorical approach by referring to generic or cross-referenced crimes, but then require circumstance-specific inquiries to determine whether specific conditional or modifying requirements are also met. 557 U.S. 29, 37-40 (2009) (noting, inter alia, that the phrases "if committed for commercial advantage," "for the purpose of assisting . . . ," and "in which the loss . . ." required a circumstance-specific inquiry into those underlying facts even though the offenses themselves were cross-referenced statutes that should be compared under the categorical approach). This hybrid approach—for lack of a better term—turns on how the circumstance-specific conditions modify the generic or cross-referenced offenses.

Applying that hybrid approach to this case, SORNA's sex offender tier classification imposes circumstance-specific conditions on the cross-referenced offenses. Title 34 U.S.C. § 20911(3) modifies the cross-referenced offenses with "when committed against a minor," and 34 U.S.C. § 20911(4)(A)(ii) modifies the cross-referenced offenses with "against a minor who has not attained the age of 13 years." Therefore, in alignment with every other circuit to consider the question, we hold that when classifying sex offender tier levels under 34 U.S.C. §§ 20911(2)-(4), the text of SORNA requires a circumstance-specific inquiry into the victim's age to determine whether the victim was, in fact, a minor at the time of the offense.

Or, in the case of a tier III categorization under § 20911(4)(A)(ii), whether the victim was, in fact, younger than 13.

However, that does not end the inquiry in this case. Here, the government urges us to look at the facts not only to determine whether the victim was a minor, but also to determine whether the offender-victim age differential existed. Accordingly, the government needs to establish that the relevant statutory text permits departure from the categorical approach not only when considering the victim's age (i.e., to ascertain whether he or she is a minor), but also when a cross-referenced federal crime includes, as an element, an age differential that the state crime of conviction does not. This is an apparent matter of first impression, as the parties have not briefed, nor have we identified, any other circuits to specifically address this question.

Coming back to the basics, a state sex offense is a tier II sex offense for SORNA sentencing purposes when, inter alia, it "is comparable to or more severe than [abusive sexual contact as described in § 2244] when committed against a minor." 34 U.S.C. § 20911(3) (emphasis added). Abusive sexual contact under 18 U.S.C. § 2244 includes, inter alia, sexual abuse of a minor under 18 U.S.C. § 2243(a). Section 2243(a) requires that the victim be 12-15 years old and that the offender be four years older than the victim. The Utah statute under which Escalante was convicted required that the victim be 14-15 years old, but would not require the government to prove an age differential.

The conditional language of "when committed against a minor" permits the government to conduct a circumstance-specific inquiry into whether the victim was a minor (which she was in this case), but it does not suggest that the court can abandon the categorical approach and conduct a circumstance-specific inquiry when looking at an offender-victim age differential that is required as an element of the cross-referenced federal offense. Sections 20911(2)-(4) of SORNA do not contain any conditional language referring to the offender's age or the offender-victim age differential—nor does 18 U.S.C. § 2244's description of abusive sexual contact, which cross-references to sexual abuse of a minor under 18 U.S.C. § 2243. Looking solely at the elements then, the Utah offense criminalizes consensual sexual contact between an 18-year-old and a 15-year-old, whereas the federal statute does not. Thus, under the categorical approach, the Utah offense "sweeps more broadly" than the comparable federal offense and cannot serve as a proper predicate for a SORNA tier II sex offender designation.

The government draws our attention to United States v. Coleman, 681 F. App'x 413 (5th Cir. 2017) (unpublished), for the proposition that under SORNA a comparable state offense can be "slightly broader" than the corresponding federal offense. In Coleman, the appellant argued that a Minnesota statute was not a proper SORNA predicate because that statute required "sexual or aggressive intent" while the relevant federal statute required "intent to abuse . . . or gratify the sexual desire[.]" Id. at 417; compare 18 U.S.C. § 2246(3), with Minn. Stat. § 609.341. We held that the abusive intent of the federal statute was analogous to the aggressive intent of the Minnesota statute. Coleman, 681 F. App'x at 417. However, we then also noted that "even if" aggressive intent was slightly broader than abusive intent, other "[c]ourts have stated that, given SORNA's broad purpose, a comparable statute can be 'slightly broader' than the federal crime." Id. at 416-18 (citing United States v. Forster, 549 F. App'x 757, 769 (10th Cir. 2013) (unpublished), and United States v. Morales, 801 F.3d 1, 7-8 (1st Cir. 2015)).
The Tenth Circuit's Forster opinion, for its part, suggested that because SORNA uses the word "comparable," a corresponding state offense could possibly be "slightly broader"; however, it then went on to hold that there was no distinction between the SORNA tier and state offense in question, both of which referenced minors younger than 13. 549 F. App'x at 769. And the First Circuit's Morales opinion, while suggesting that "'comparable to' may, as the government argues, provide us some flexibility in examining the offenses," declined to give any flexibility in that case because "the question of age is so essential to the framework that the congressional cutoff must be strictly construed." 801 F.3d at 78.
We are not bound by our unpublished opinions. Ballard v. Burton, 444 F.3d 391, 401 n.7 (5th Cir. 2006); 5th Cir. R. 47.5.4. Moreover, we are skeptical that courts applying the categorical approach have leeway to hold that a broader offense can still be a predicate when it is deemed only "slightly broader." The Supreme Court has had many opportunities to articulate how lower courts should conduct the categorical approach, and the government does not point us to any decision wherein the Supreme Court suggested that a state offense which criminalized broader conduct than the corresponding federal crime could constitute a valid predicate if it was only "slightly broader." Cf., e.g., EsquivelQuintana, 137 S. Ct. at 1568 (articulating the categorical approach as an examination of whether the elements of the state offense "would fall within the federal definition of the crime[,]" not whether they come close to falling within that federal definition). Nonetheless, even assuming that the categorical approach could be "slightly broader" under SORNA, a statutory rape offense that criminalizes conduct without an age differential requirement is more than "slightly broader" than one which does require an age differential.

The government places a lot of weight on United States v. Gonzalez-Medina, wherein we held that we must consider the specific circumstances of the victim-offender age differential when determining whether an offense is a "sex offense" at all under 34 U.S.C. § 20911(5)(C). 757 F.3d 425, 431 (5th Cir. 2014). Section 20911(5)(C) is an exception to SORNA, stating that "[a]n offense involving consensual sexual conduct is not a sex offense for the purposes of this subchapter . . . if the victim was at least 13 years old and the offender was not more than 4 years older than the victim." Gonzalez-Medina involved a situation where a 24-year-old offender violated a Wisconsin statute that criminalized sexual activity with persons 16-17 years old. 757 F.3d at 427. He subsequently argued that he was not required to register as a "sex offender" at all because, under the categorical approach, the Wisconsin statute did not include the age differential exception for when an offense is deemed a "sex offense" under SORNA. Id. at 428-29. We disagreed with that argument, concluding that "the language, structure, and broad purpose of SORNA all indicate that Congress intended a non-categorical approach to the age-differential determination in [34 U.S.C. § 20911(5)(C)]." Id. at 432.

Gonzalez-Medina refers to 42 U.S.C. § 16911 when discussing SORNA. That provision has since been re-codified at 34 U.S.C. § 20911 (effective Sep. 1, 2017).

Nonetheless, there are several reasons why the government's emphasis on Gonzalez-Medina is unavailing in this case.

First, and most importantly, the age differential for the § 20911(5)(C) exception is actually in SORNA's text as an exception to when state offenses can even constitute a "sex offense." Conversely, the age differential relevant for classifying tier II sex offenders under § 20911(3) is not actually found in SORNA, but rather it is built in as an element of one of the many corresponding offenses that SORNA cross-references.

Second, the text of the § 20911(5)(C) exception states that it is applicable to offenses "involving . . . conduct . . . [if] the offender was not more than 4 years older than the victim." That "involving conduct" verbiage is consistent with circumstance-specific inquiries. See Gonzalez-Medina, 757 F.3d at 430. However, the text of § 20911(3), which is at issue in this case, does not use that verbiage when classifying a tier II sex offender, and instead lists corresponding "offenses" against which to compare the conviction offense, modified only by the conditional language of "when committed against a minor."

And third, Gonzalez-Medina held that SORNA was intended to cast a wide net, and that a categorical exception for any conviction involving consensual conduct that did not require a four-year age differential as an element—to include child pornography—"would frustrate SORNA's broad purpose" of requiring sex offenders to register. Id. at 431. However, those purposive concerns are not as strong for sex offender tier classifications. Applying the categorical approach to the age differentials built into the cross-referenced corresponding offenses would not mean that Escalante does not have to register under SORNA at all; instead, it means that his failure to register imposes a sentencing baseline of 12 points rather than 14.

Moreover, applying the categorical approach to age differentials built into the cross-referenced statutes does not leave §§ 20911(2)-(4), or any subsection thereof, without meaningful application. Cf. Nijhawan, 557 U.S. at 39 (not applying the categorical approach where doing so would leave a subparagraph "with little, if any, meaningful application"). Instead, § 20911(3)'s "when committed against a minor" language still causes tier II classification for an array of offenses based only on the fact that they were committed against minors, as well as for statutory rape offenses wherein the government had to prove a four-year age differential.

The offenses for which an offender is classified as tier II based solely on the fact that the victim was a minor include sex trafficking, coercion and enticement, transportation with intent to engage in criminal sexual activity, and abusive sexual contact (which, in addition to the offense of sexual abuse of a minor or ward at issue here, also includes sexual abuse and aggravated sexual abuse.). See 34 U.S.C. § 20911(3) and, as cross-referenced, 18 U.S.C. §§ 2244 and 2243.

For statutory rape offenses where the government did not have to prove a four-year age differential, offenders still have to register under SORNA, but as tier I offenders rather than tier II.

As a final point, our decision in Rodriguez warrants discussion. In Rodriguez, our court, sitting en banc, had the task of defining the generic term "sexual abuse of a minor" for the crime-of-violence enhancement under § 2L1.2(b)(1)(A)(ii) of the Guidelines. We rejected the argument that Texas's statutory rape law was too broad to serve as a predicate offense because it only required a three-year, not a four-year, age differential. 711 F.3d at 544, 562 n.28 ("We reject this argument because the definitions of 'sexual abuse of a minor' in legal and other well-accepted dictionaries do not include such an age-differential requirement."). Esquivel-Quintana abrogated Rodriguez's holding that the generic age of a "minor" for consensual sex offenses could be higher than 16; however, Esquivel-Quintana did not abrogate Rodriguez's holding that the generic crime of "sexual abuse of a minor" does not require an age differential. That holding remains the law of this circuit.

See also Shroff v. Sessions, 890 F.3d 542, 545 (5th Cir. 2018) (holding that because Esquivel-Quintana "focused on the age requirement and did not make an express holding on the requirement of sexual contact" that it did not abrogate this court's precedent regarding the contact requirement); United States v. Montes-Barrientos, 742 F. App'x 18, 19 (5th Cir. 2018) (unpublished) (noting that Shroff's holding on that point was not dicta).

However, that holding is distinguishable from the facts of this case in an important way: Rodriguez dealt with the generic term "sexual abuse of a minor"; this case deals with the phrase "abusive sexual contact (as described in section 2244 of title 18)." See 34 U.S.C § 20911(3)(A)(iv). In other words, SORNA does not classify someone as a tier II offender based on comparing their offenses to a generic offense; It classifies them based on comparing their offenses to a specified other offense. In this case, the specified other offense—sexual abuse of a minor under 18 U.S.C § 2243(a)—does in fact require a four-year age differential.

Based on that reasoning, we hold—in alignment with every other circuit to consider the issue—that SORNA requires a circumstance-specific inquiry into the victim's age when classifying sex offender tier levels to determine whether the victim was a minor, or, in the case of a tier III categorization under § 20911(4)(A)(ii), whether the victim was younger than 13.

We also hold, as a matter of apparent first impression, that the text of SORNA does not permit a court, when applying the categorical approach to determine sex offender tier levels, to conduct a circumstance-specific inquiry into an offender-victim age differential that is built into one of the corresponding cross-referenced offenses as an element of the crime. In this case, looking solely at the elements, the Utah offense under which Escalante was convicted sweeps more broadly than the cross-referenced federal offense corresponding to tier level II. Accordingly, the district court erred by categorizing Escalante as a tier II sex offender.

3.

Last, we address the harmlessness argument. Escalante's sentence was based on a Guidelines calculation that categorized him as a tier II offender. The government argues that even if the district court erred by classifying Escalante as a tier II sex offender, any such error would be harmless because the district court upwardly varied from the Guidelines at sentencing. That variance, the government asserts, was based on his significant criminal history and likelihood of recidivism, and therefore had nothing to do with the sex offender tier at which he was classified.

Nonetheless, we agree with Escalante that the government has not cleared the "high hurdle" necessary for holding the sentencing error to be harmless. See United States v. Halverson, 897 F.3d 645, 652 (5th Cir. 2018) (quoting United States v. Ibarra-Luna, 628 F.3d 712, 714 (5th Cir. 2010)). Under our precedent, the harmless error doctrine only applies to Guideline calculation errors when the government "convincingly demonstrates both (1) that the district court would have imposed the same sentence had it not made the error, and (2) that it would have done so for the same reasons it gave at the prior sentencing." Id. The "crux" of the inquiry is "whether the district court would have imposed the same sentence, not whether the district court could have imposed the same sentence." Id. (quoting United States v. Delgado-Martinez, 564 F.3d 750, 753 (5th Cir. 2009)). And "[t]he record must show 'clarity of intent' expressed by the district court[.]" Id. (quoting United States v. Shepherd, 848 F.3d 425, 427 (5th Cir. 2017)).

Here, the government has not argued, nor does a review of the sentencing transcript clearly indicate, that the district court would have imposed the same 48-month sentence regardless of the Guidelines range suggested by the sex offender tier. Instead, all the government argues is that the district court's rationale would have supported an upward variance from whatever baseline that the Guidelines suggested. That is not enough. See, e.g., Ibarra-Luna, 628 F.3d at 719 ("[W]e are convinced that the explanation the district court gave for imposing an above-Guidelines sentence would have led it to do so even if it had considered the correct Guidelines range. . . . [But w]e cannot state with the requisite certainty, however, that the district court would have imposed precisely the same sentence."); United States v. Martinez-Flores, 720 F.3d 293, 300 (5th Cir. 2013) (holding the error harmful where "the district court did not clearly state (and we cannot glean from the record) that it would impose the same sentence if [the Guidelines calculation had been corrected]"). Escalante's erroneous classification as a tier II sex offender was not harmless.

IV.

In summary, the district court erred by deviating from the categorical approach to consider the circumstance-specific facts of the offender-victim age differential when classifying Escalante as a tier II sex offender under SORNA, and by sentencing him based on a Guidelines range that was derived from that tier II categorization. Because the record does not convey the district court's clear intent to impose the same sentence absent that Guidelines error, we must VACATE and REMAND for resentencing.

* * * *

This outcome is required by faithful adherence to precedent that has struggled to grapple with the expansion and byzantine-like application of the categorical approach. However, it is not lost on us that adherence to the categorical approach leads to a result in this case that is almost certainly contrary to any plain reading of the statute. Persons who commit sex offenses against minors would, under any plain reading of SORNA, be expected to register as tier II sex offenders. In this case, the victim was 14. Escalante was 35. He abused a minor under any jurisdiction's definition of minor. Nonetheless, application of the categorical approach means that he is not a tier II sex offender because Utah's statute for unlawful sexual activity with a minor did not require the same age differential that the referenced federal statute did—even though Escalante was not even close to being within that age differential at the time of the offense.

In the nearly three decades since its inception, the categorical approach has developed a reputation for crushing common sense in any area of the law in which its tentacles find an inroad. What began as an effort by the Supreme Court to simplify the judiciary's job when determining whether a state crime constituted generic "burglary" has now metastasized into something that requires rigorous abstract reasoning to arrive at the conclusion that a 35-year-old who sexually abused a 14-year-old cannot be categorized as a tier II sex offender—notwithstanding the fact that his crime was actually "committed against a minor"—because it is theoretically possible that someone else could be convicted under the statute without being four years older than the victim.

See Sessions v. Dimaya, 138 S. Ct. 1204, 1252 (2018) (Thomas, J., dissenting) ("The categorical approach originated with Justice Blackmun's opinion for the Court in Taylor v. United States, 495 U.S. 575 (1990).").

See, e.g., United States v. Lewis, 720 F. App'x 111, 120 (3d Cir. 2018) (Roth, J., concurring in the judgment) (describing the categorical approach as "willful blindness—which may allow violent offenders to evade accountability"); United States v. Davis, 875 F.3d 592, 595 (11th Cir. 2017) (observing that the categorical approach carries judges "down the rabbit hole . . . to a realm where we must close our eyes as judges . . . . It is a pretend place in which a crime that the defendant committed violently is transformed into a non-violent one . . . . Curiouser and curiouser it has all become[.]"); United States v. Chapman, 866 F.3d 129, 136-38 (3d Cir. 2017) (Jordan, J., concurring) (expressing dismay at the "kudzu quality of the categorical approach, which seems to be always enlarging its territory[,]" and which "often asks judges to feign amnesia," and to "ignore facts already known and instead proceed with eyes shut"); United States v. Faust, 853 F.3d 39, 61 (1st Cir. 2017) (Lynch, J., concurring) (observing that the categorical approach "can lead courts to reach counterintuitive results, and ones which are not what Congress intended"); United States v. Doctor, 842 F.3d 306, 313-15 (4th Cir. 2016) (Wilkinson, J., concurring) (stating that the categorical approach has caused judges to "swap[] factual inquiries for an endless gauntlet of abstract legal questions[,]" and recommending that the categorical approach should "loosen[] its present rigid grip upon criminal sentencing").

Members of this court have been critical of the counterintuitive—and all too often absurd—conclusions that can result when the categorical approach wraps its tentacles around a sentencing decision. We have said that "[e]xcept as otherwise directed by the Supreme Court, sentencing should not turn on reality-defying distinctions." United States v. Reyes-Contreras, 910 F.3d 169, 186 (5th Cir. 2018) (en banc) (citation omitted). Unfortunately, application of the categorical approach here leads us inexorably to the conclusion that this is such a case.

See, e.g., United States v. Herrold, 883 F.3d 517, 550 (5th Cir. 2018) (en banc) (Haynes, J., dissenting) (observing that "arcane technicalities" should not allow criminals to evade sentencing enhancements (quoting Taylor, 495 U.S. at 589)), cert. granted, judgment vacated, No. 17-1445, 2019 WL 2493911 (U.S. Jun. 17, 2019)); United States v. Castillo-Rivera, 853 F.3d 218, 244 (5th Cir. 2017) (en banc) (Higginson, J., concurring in part and dissenting in part) ("Our ongoing struggle to apply the categorical approach while respecting the congressional purpose to enhance punishment for similar recidivists may justify Supreme Court intervention yet again."); Rodriguez, 711 F.3d at 545 n.2 (observing that the current form of the categorical approach begets "confusion and gymnastics," and suggesting that the Supreme Court should provide clarity on the issue).

Somewhere along the way, when developing the categorical approach, the federal judiciary has gotten lost. See also Mathis, 136 S. Ct. at 2266-68 (Alito, J., dissenting) (analogizing the evolution of the categorical approach to someone who reportedly had the wrong address in her GPS for what was supposed to be a less-than-an-hour trip, but who then drove across Europe in the other direction for multiple days before entertaining the thought that she was not going where she first set out to go). Perhaps one day the Supreme Court will consider revisiting the categorical approach and setting the federal judiciary down a doctrinal path that is easier to navigate and more likely to arrive at the jurisprudential destinations that a plain reading of our criminal statutes would suggest.

See also Quarles v. United States, 139 S. Ct. 1872, 1881 (2019) (Thomas, J., concurring) (suggesting that the Supreme Court reconsider this approach and noting that "the categorical approach employed today is difficult to apply and can yield dramatically different sentences depending on where a [crime] occurred"). --------