FOR PUBLICATION

UNITED STATES COURT OF APPEALS

FOR THE NINTH CIRCUIT

LUDWIN ISRAEL LOPEZ-AGUILAR, No. 17-73153 Petitioner, Agency No. v. A074 394 680

WILLIAM P. BARR, Attorney General of the United States, OPINION Respondent.

On Petition for Review of an Order of the Board of Immigration Appeals

Argued and Submitted March 5, 2019 Portland, Oregon

Filed April 23, 2019

Before: Susan P. Graber and Marsha S. Berzon, Circuit Judges, and John R. Tunheim, * Chief District Judge.

Opinion by Judge Tunheim; Dissent by Judge Berzon

* The Honorable John R. Tunheim, Chief United States District Judge for the District of Minnesota, sitting by designation. 2 LOPEZ-AGUILAR V. BARR

SUMMARY **

Immigration Denying Ludwin Israel Lopez-Aguilar’s petition for review of a decision of the Board of Immigration Appeals, the panel held that third-degree robbery under Oregon Revised Statutes § 164.395 is a categorical theft offense and, therefore, an aggravated felony under 8 U.S.C. § 1101(a)(43)(G), and the panel concluded that the record supported the denial of deferral of removal under the Convention Against Torture.

The BIA concluded that Lopez-Aguilar was removable for an aggravated felony theft offense based on his conviction for third-degree robbery under Oregon Revised Statutes § 164.395.

The panel explained that, in the context of aggravated felonies, a generic theft offense is defined as (1) a taking of property or an exercise of control over property (2) without consent (3) with the criminal intent to deprive the owner of rights and benefits of ownership, even if such deprivation is less than total or permanent.

Lopez-Aguilar contended that section 164.395 is not categorically a generic theft offense because: (1) it incorporates theft by deception, which covers consensual takings, and (2) it incorporates unauthorized use of a vehicle,

** This summary constitutes no part of the opinion of the court. It has been prepared by court staff for the convenience of the reader.

LOPEZ-AGUILAR V. BARR 3

which does not require an intent to deprive the owner of the rights and benefits of ownership.

The panel concluded that section 164.395 theoretically could cover a consensual taking due to its incorporation of theft by deception, explaining that the statute does not require that force be used or threatened against the owner of the property. The panel observed that, for example, the statute could theoretically apply to a situation where a person obtained property from its owner, by deception, and then used force against a third party. However, the panel concluded that there is no realistic probability that Oregon would prosecute such conduct under the statute.

The panel also concluded that the incorporation of unauthorized use of a vehicle under Oregon Revised Statutes § 164.135(1)(b)-(c) into section 164.395 does not make the statute overbroad. Lopez-Aguilar contended that Oregon’s unauthorized use of a vehicle statute criminalizes the use of a vehicle even where the vehicle has been temporarily and consensually placed in the defendant’s care. However, emphasizing that the generic definition includes a taking where the deprivation is less than total or permanent, the panel concluded that none of the conduct covered by Oregon’s unauthorized use of a vehicle statute falls outside of the generic theft offense definition.

Finally, the panel concluded that substantial evidence supported the denial of CAT relief, explaining that the record did not compel a finding that Lopez-Aguilar will more likely than not face torture by his father or gang members.

Dissenting, Judge Berzon disagreed with the majority’s conclusion that Lopez-Aguilar was required to establish a realistic probability that the statute would be applied in a nongeneric manner. Judge Berzon wrote that, under the 4 LOPEZ-AGUILAR V. BARR

circuit’s case law, Lopez-Aguilar was not required to establish such a realistic probability because section 164.395’s text is on its face broader than a generic theft offense under the Immigration and Nationality Act. Judge Berzon also wrote that, even if Lopez-Aguilar were required to make that showing, Oregon case law demonstrates that section 164.395 realistically applies to conduct falling outside generic theft.

Therefore, Judge Berzon would hold that section 164.395 is not categorically a generic theft offense, deem the issue of divisibility waived by the government, and grant the petition for review.

COUNSEL

Jennifer K. Lesmez (argued), Law Offices of Jennifer Lesmez, Allyn, Washington, for Petitioner.

Imran Raza Zaidi (argued) and Matthew A. Spurlock, Trial Attorneys; John S. Hogan, Assistant Director; Joseph H. Hunt, Assistant Attorney General; Office of Immigration Litigation, Civil Division, United States Department of Justice, Washington, D.C.; for Respondent.

OPINION

TUNHEIM, Chief District Judge:

Petitioner Ludwin Israel Lopez-Aguilar, a native and citizen of Guatemala, petitions for review of a final order of the Board of Immigration Appeals (“BIA”) finding him removable pursuant to section 237(a)(2)(A)(iii) of the

LOPEZ-AGUILAR V. BARR 5

Immigration and Nationality Act (“INA”) because of his conviction under Oregon Revised Statutes section 164.395 and denying his application for protection under the Convention Against Torture (“CAT”). We have jurisdiction under 8 U.S.C. § 1252. We deny Lopez-Aguilar’s petition because we conclude that: (1) section 164.395 is a categorical theft offense and, therefore, an aggravated felony under section 101(a)(43)(G) of the INA; and (2) the record supports the BIA’s denial of CAT relief.

BACKGROUND

Lopez-Aguilar is a native and citizen of Guatemala. He entered the United States in 1989, when he was three years old, and became a legal permanent resident on March 12, 2001, when his application for suspension of deportation was granted.

In Guatemala, Lopez-Aguilar’s father abused his mother physically, sexually, and verbally. Lopez-Aguilar’s father also abused him, starting when he was less than a year old, and threatened to kill him. Since Lopez-Aguilar entered the United States, he has never returned to Guatemala, but his mother has returned twice, once for three months and once for three weeks. Lopez-Aguilar’s father did not contact her or bother her on those trips, and he has not tried to contact Lopez-Aguilar since 1995. Nevertheless, Lopez-Aguilar fears that, if he is returned to Guatemala, his father will follow through on the threat to kill him.

Lopez-Aguilar was formerly a member of the Norteño gang. He became affiliated with the Norteños at age 16 and was initiated at age 18. He has numerous visible tattoos that he believes make him identifiable as a Norteño. His role as a Norteño involved fighting with members of rival gangs, 6 LOPEZ-AGUILAR V. BARR

including Mara Salvatrucha and Barrio 18, both of which operate throughout Guatemala.

Lopez-Aguilar left the gang in 2009. He has not been targeted by any gangs in the United States, and no one has harmed or looked for him. However, he fears that he will be targeted by police or by rival gangs, if returned to Guatemala, because he will be recognizable as a Norteño and seen as suspicious and a foreigner.

I. Lopez-Aguilar’s Conviction

In 2014, Lopez-Aguilar was convicted of third-degree robbery in violation of Oregon Revised Statutes section 164.395 and sentenced to 13 months in prison. Section 164.395 provides:

(1) A person commits the crime of robbery in the third degree if in the course of committing or attempting to commit theft or unauthorized use of a vehicle as defined in ORS 164.135 the person uses or threatens the immediate use of physical force upon another person with the intent of:

(a) Preventing or overcoming resistance to the taking of the property or to retention thereof immediately after the taking; or

(b) Compelling the owner of such property or another person to deliver the property or to engage in other conduct which might aid in the commission of the theft or unauthorized use of a vehicle.

LOPEZ-AGUILAR V. BARR 7

(2) Robbery in the third degree is a Class C felony.

Section 164.395 incorporates Oregon’s theft definition, which includes “theft by deception.” Or. Rev. Stat. § 164.015(4). It also incorporates Oregon’s Unauthorized Use of a Vehicle statute, which criminalizes, in relevant part:

(b) Having custody of a vehicle, boat or aircraft pursuant to an agreement between the person or another and the owner thereof whereby the person or another is to perform for compensation a specific service for the owner involving the maintenance, repair or use of such vehicle, boat or aircraft, the person intentionally uses or operates it, without consent of the owner, for the person’s own purpose in a manner constituting a gross deviation from the agreed purpose; or

(c) Having custody of a vehicle, boat or aircraft pursuant to an agreement with the owner thereof whereby such vehicle, boat or aircraft is to be returned to the owner at a specified time, the person knowingly retains or withholds possession thereof without consent of the owner for so lengthy a period beyond the specified time as to render such retention or possession a gross deviation from the agreement.

Or. Rev. Stat. § 164.135(1)(b)–(c). 8 LOPEZ-AGUILAR V. BARR

II. Immigration Court Proceedings

Because of Lopez-Aguilar’s robbery conviction, an immigration judge (“IJ”) found him removable as an alien convicted of an aggravated felony as defined in two sections of the INA: (1) section 101(a)(43)(F), which defines crimes of violence, and (2) section 101(a)(43)(G), which defines theft offenses for which the term of imprisonment is at least one year. The IJ also denied Lopez-Aguilar’s petition for deferral of removal under the CAT. The IJ ordered Lopez- Aguilar removed to Guatemala.

III. BIA Appeal

The BIA dismissed Lopez-Aguilar’s appeal. It disagreed with the IJ’s conclusion that Lopez-Aguilar’s conviction under Oregon Revised Statutes section 164.395 was for a crime of violence under section 101(a)(43)(F) of the INA but agreed that the conviction was for a theft offense under section 101(a)(43)(G).

The BIA rejected Lopez-Aguilar’s argument that section 164.395 is overbroad because it covers consensual takings by incorporating theft by deception. The BIA concluded that the statute also requires taking of property by force, which negates the consensual nature of theft by deception. Accordingly, the BIA found that Lopez-Aguilar had not demonstrated a realistic probability that an individual could be convicted under section 164.395 for a consensual taking.

The BIA also rejected Lopez-Aguilar’s argument that section 164.395 is overbroad because it covers unauthorized use of a vehicle, which does not require an intent to deprive the owner of the rights and benefits of ownership. The BIA concluded that the remainder of the statute required an intent

LOPEZ-AGUILAR V. BARR 9

to prevent or overcome resistance to the taking, to retain the property immediately after the taking, or to compel another to deliver the property. Accordingly, the BIA found that Lopez-Aguilar had not demonstrated a realistic probability that an individual could be convicted under section 164.395 for unauthorized use of a vehicle without the requisite intent.

Finally, the BIA agreed with the IJ’s conclusion that Lopez-Aguilar had not established that he would more likely than not face a particularized risk of torture with the acquiescence of a public official in Guatemala.

This timely petition for review followed.

JURISDICTION AND STANDARD OF REVIEW

We have jurisdiction to review final orders of removal based on a petitioner’s commission of an aggravated felony to the extent that the petition “raises . . . questions of law.” Ngaeth v. Mukasey, 545 F.3d 796, 800 (9th Cir. 2008) (per curiam) (quoting Vizcarra-Ayala v. Mukasey, 514 F.3d 870, 872 (9th Cir. 2008)). Whether a particular offense is an “aggravated felony” under the INA is a question of law that we review de novo. Id.

We also have jurisdiction to review the BIA’s denial of CAT relief where, as here, “the IJ did not rely on [the petitioner’s] conviction . . . but instead denied relief on the merits.” Alphonsus v. Holder, 705 F.3d 1031, 1036–37 (9th Cir. 2013), abrogated on other grounds as recognized in Guerrero v. Whitaker, 908 F.3d 541 (9th Cir. 2018). We review denial of CAT relief for substantial evidence. Owino v. Holder, 771 F.3d 527, 531 (9th Cir. 2014) (per curiam). The substantial evidence standard is “extremely deferential,” and we must uphold the BIA’s findings “unless the evidence presented would compel a reasonable finder of fact to reach 10 LOPEZ-AGUILAR V. BARR

a contrary result.” Gebhart v. SEC, 595 F.3d 1034, 1043 (9th Cir. 2010) (quoting Monjaraz-Munoz v. INS, 327 F.3d 892, 895 (9th Cir.), amended by 339 F.3d 1012 (9th Cir. 2003) (order)).

DISCUSSION

I. Section 164.395 and Generic Theft Offenses

Under the INA, a conviction for a generic theft offense that results in a prison term of at least one year is an aggravated felony. See 8 U.S.C. § 1101(a)(43)(G). To determine whether a particular conviction is for a theft offense, we use the categorical and modified categorical approaches of Taylor v. United States, 495 U.S. 575 (1990), and Shepard v. United States, 544 U.S. 13 (2005). Under those approaches, we compare Lopez-Aguilar’s statute of conviction (Or. Rev. Stat. § 164.395) with the generic crime of theft to determine whether the latter encompasses the former. See Hernandez-Cruz v. Holder, 651 F.3d 1094, 1100 (9th Cir. 2011). In the context of an aggravated felony theft offense under section 101(a)(43)(G) of the INA, the generic crime of theft is defined as “[1] a taking of property or an exercise of control over property [2] without consent [3] with the criminal intent to deprive the owner of rights and benefits of ownership, even if such deprivation is less than total or permanent.” Id. at 1100–01 (quoting Carrillo-Jaime v. Holder, 572 F.3d 747, 750 (9th Cir. 2009)).

Lopez-Aguilar contends that section 164.395 is not categorically a generic theft offense because: (1) it incorporates theft by deception, which covers consensual takings, and (2) it incorporates unauthorized use of a vehicle, which does not require an intent to deprive the owner of the rights and benefits of ownership.

LOPEZ-AGUILAR V. BARR 11

A. Theft by Deception

We conclude that, although section 164.395 theoretically could cover a consensual taking due to its incorporation of theft by deception, there is no realistic probability that Oregon would prosecute such conduct under the statute. To find that a statute of conviction is broader than a generic removable offense definition, there must be “a realistic probability, not a theoretical possibility, that the State would apply its statute to conduct that falls outside the generic definition.” Gonzales v. Duenas-Alvarez, 549 U.S. 183, 193 (2007).

To be convicted under section 164.395, a defendant must:

use[] or threaten[] the immediate use of physical force upon another person with the intent of: (a) Preventing or overcoming resistance to the taking of the property or to retention thereof immediately after the taking; or (b) Compelling the owner of such property or another person to deliver the property or to engage in other conduct which might aid in the commission of the theft or unauthorized use of a vehicle.

Or. Rev. Stat. § 164.395(1) (emphasis added). Because the statute does not require that force be used or threatened against the owner of the property, the text of the statute could theoretically cover situations involving consensual takings. For example, under subsection (a), a defendant could be convicted if he entered a residential building, obtained property from a resident through deception, and used force against a security guard on his way out of the building in order to retain the property. Under subsection (b), a 12 LOPEZ-AGUILAR V. BARR

defendant could be convicted if she convinced an owner, by deception, to give her property but used force against a third party to compel that third party to deliver the consensually obtained property to her. In either scenario, the property would have been taken by consent of the owner, and the force used would not negate the owner’s consent because the force was used against a third party without the owner’s knowledge.

However, these two scenarios represent merely theoretical – not realistic – possibilities. Indeed, under subsection (a), the threat or force must be used “immediately after the taking.” Therefore, it is unlikely that a defendant would be convicted for using or threatening force against a third party unless the force occurred in the presence of the owner, which would negate consent. 1

1 The dissent disagrees, citing State v. Tolbert for the proposition that Oregon courts prosecute robbery in “situations involving the use of force during flight.” 433 P.3d 501, 505 (Or. Ct. App. 2018). But the Tolbert court also noted that “[a]bsent evidence of ‘close pursuit,’ . . . intervening time, distance, and events may well break the link between the theft and the subsequent use of force.” Id. at 506. In fact, the Tolbert court reversed the defendant’s conviction because “[a]bout 10 to 15 minutes had passed” between the theft and the use of force. Id. This case does not convince us that there is a realistic possibility that Oregon courts would prosecute conduct falling outside the generic definition of theft.

The dissent also cites two cases in which Oregon courts prosecuted defendants for using force against security guards as examples of takings followed by threats or use of force against third parties. But neither case involved a consensual taking. See Pereida-Alba v. Coursey, 342 P.3d 70, 71 (Or. 2015); State v. Tolson, 546 P.2d 1115, 1116 (Or. Ct. App. 1976). Furthermore, Oregon treats security guards as authorized agents of the owner. See Or. Rev. Stat. § 131.655; see also, e.g., AFSCME v. City of Lebanon, 388 P.3d 1028, 1038 (Or. 2017) (holding that

LOPEZ-AGUILAR V. BARR 13

Moreover, Lopez-Aguilar has not presented, and we are unable to find, any Oregon case in which a defendant was prosecuted for conduct falling outside the generic definition of theft. Accordingly, we find that section 164.395 is not overbroad, qualifies as a categorical theft offense and, therefore, constitutes an aggravated felony under section 101(a)(43)(G) of the INA.

B. Unauthorized Use of a Vehicle

We also conclude that the incorporation of unauthorized use of a vehicle in section 164.395 does not make it overbroad. Lopez-Aguilar argues that subsections (b) and (c) of Oregon’s Unauthorized Use of a Vehicle statute criminalize use of a vehicle without the intent to deprive the owner of the rights and benefits of ownership because they apply when the vehicle has been temporarily and consensually placed in the defendant’s care. See Or. Rev. Stat. § 164.135(1)(b)–(c). However, in the context of aggravated felonies, a generic theft offense is defined as “[1] a taking of property or an exercise of control over property [2] without consent [3] with the criminal intent to deprive the owner of rights and benefits of ownership, even if such deprivation is less than total or permanent.” Hernandez-Cruz, 651 F.3d at 1100–01 (emphasis added) (quoting Carrillo-Jaime, 572 F.3d at 750). None of the

employers can be liable for their employees’ conduct); Hoke v. May Dep’t Stores Co., 891 P.2d 686 (Or. Ct. App. 1995) (holding that a department store is liable for the wrongful acts of its security guard); Gibson v. Safeway Stores, Inc., 764 P.2d 548 (Or. 1988) (holding that the grocery store to which a security-guard service company assigned the guard was his joint employer). Thus, even a conviction for a consensual taking followed by force against a security guard would not fall outside the generic theft definition because security guards are not third parties under Oregon law. 14 LOPEZ-AGUILAR V. BARR

conduct covered by Oregon’s Unauthorized Use of a Vehicle statute falls outside this definition. Accordingly, we hold that section 164.395 qualifies as a categorical theft offense and, therefore, as an aggravated felony under section 101(a)(43)(G) of the INA.

II. CAT Relief

To establish eligibility for CAT relief, Lopez-Agular must establish that, if removed to Guatemala, he will more likely than not be tortured. See 8 C.F.R. 1208.16(c)(2). Torture is defined as the intentional infliction of severe pain or suffering, whether physical or mental, for purposes such as intimidation, punishment, coercion, or discrimination, if “inflicted by or at the instigation of or with the consent or acquiescence of a public official or other person acting in an official capacity.” 8 C.F.R. 1208.18(a)(1). In determining the likelihood that an applicant will be tortured, courts must consider “all evidence relevant to the possibility of future torture,” including evidence that the applicant was tortured in the past. 8 C.F.R. 1208.16(c)(3).

The BIA agreed with the IJ that Lopez-Aguilar did not show past torture and noted that Lopez-Aguilar did not challenge that finding on appeal. It found that Lopez- Aguilar had not shown that he would more likely than not face a particularized risk of torture in Guatemala at the hands of his father or rival gangs. The BIA noted that Lopez- Aguilar’s father had not tried to contact him since 1995, neither Lopez-Aguilar nor his mother knew where his father was or even if his father was alive, and his mother had returned to Guatemala twice without incident. The BIA also noted that Lopez-Aguilar had faced no serious problems with rival gangs since leaving his gang, Lopez-Aguilar testified that rival gangs in the United States had left him alone because he told them he was no longer an active gang

LOPEZ-AGUILAR V. BARR 15

member, and Lopez-Aguilar had not shown that gangs abide by different rules of conduct depending on their geographical location. Finally, the BIA reviewed country conditions evidence and expert testimony and found that Lopez-Aguilar did not establish that he would more likely than not be tortured by the government or that a public official would acquiesce to harm he might endure at the hands of private actors. The BIA’s conclusions are supported by substantial evidence.

Even considering evidence that Lopez-Aguilar was tortured by his father in the past, the record does not compel a finding that his father will more likely than not torture him upon his return. While Lopez-Aguilar suffered horrific abuse at the hands of his father, the abuse happened when Lopez-Aguilar was a small child. Lopez-Aguilar is now an adult, and his father has not tried to contact him in more than twenty years. Lopez-Aguilar’s mother, who likewise was subjected to horrific abuse at the hands of his father, was not contacted by his father on either of her return trips to Guatemala. Furthermore, Lopez-Aguilar points to no evidence in the record that any future mistreatment by his father would be inflicted with the consent or acquiescence of a public official or a person acting in an official capacity.

The record does not compel a finding that gang members will more likely than not torture Lopez-Aguilar or that police will more likely than not consent to such torture. Lopez- Aguilar notes that his expert witness, Dr. Kirkland, testified that he would be “caught up in violence” because of his tattoos and that police decline to intervene in gang conflicts. But while Dr. Kirkland concluded that Lopez-Aguilar would likely be “harassed” by gang members, which “could rise to the level of torture,” he did not go so far as to say that Lopez- Aguilar was likely to be tortured by gang members. 16 LOPEZ-AGUILAR V. BARR

The record also does not compel a finding that police will more likely than not torture Lopez-Aguilar. Dr. Kirkland testified that police were likely to monitor, stop, and interrogate him, but these actions may not rise to the level of torture.

Finally, the record does not compel a finding that police or the government would consent to Lopez-Aguilar’s torture. While there is evidence that the Guatemalan police are struggling to address violence, there is also evidence that the Guatemalan government is taking active steps to combat illegal activities. The BIA correctly noted that “a government does not ‘acquiesce’ to torture where the government actively, albeit not entirely successfully, combats the illegal activities.” Del Cid Marroquin v. Lynch, 823 F.3d 933, 937 (9th Cir. 2016) (per curiam).

Ultimately, “[w]e are not free to look anew at the testimony and then measure the soundness of the [BIA’s] decision by what we would have found.” Donchev v. Mukasey, 553 F.3d 1206, 1213 (9th Cir. 2009) (footnote omitted). Evidence does not “compel the opposite conclusion just because it would also support a different result.” Id. Because the BIA’s denial of CAT relief is supported by substantial evidence, we deny Lopez-Aguilar’s petition on this ground.

CONCLUSION

We deny Lopez-Aguilar’s petition for review. Lopez- Aguilar is removable for having committed an aggravated felony, and the BIA permissibly concluded that Lopez- Aguilar failed to show that he was likely to be tortured if returned to Guatemala.

Petition DENIED. BERZON, Circuit Judge, dissenting:

I respectfully dissent.

The majority recognizes that section 164.395 of the Oregon Revised Statutes is, by its terms, broader than a generic theft offense under the Immigration and Nationality Act (INA). Concluding that Lopez-Aguilar failed to establish a “realistic probability” the state statute would be applied in a nongeneric manner, the majority holds that Oregon third-degree robbery categorically constitutes generic theft under the INA. I do not agree that Lopez- Aguilar was required to make such a showing and, in any event, believe that Oregon case law does establish such a realistic probability. I would therefore hold that the state crime is not categorically a generic theft offense. 1

I

Under the INA, theft is defined as the “taking of property or an exercise of control over property without consent with the criminal intent to deprive the owner of rights and benefits of ownership, even if such deprivation is less than total or permanent.” Gonzales v. Duenas-Alvarez, 549 U.S. 183, 189 (2007) (emphasis added) (quoting Penuliar v. Gonzales, 435 F.3d 961, 969 (9th Cir. 2006)). The BIA has accordingly 1 Having concluded that third-degree robbery under section 164.395 is not categorically an aggravated theft offense, we would normally proceed to the next step of the categorical approach—that is, divisibility. Here, however, the government has failed to argue divisibility. I would therefore deem this issue waived and grant Lopez-Aguilar’s petition for review. See Aguirre Barbosa v. Barr, No. 15-72092, 2019 WL 1388298, at *4 (9th Cir. Mar. 28, 2019) (holding that the government waived divisibility as to whether section 164.395 constitutes a crime involving moral turpitude). Because I would conclude that Lopez-Aguilar is not removable, I would not address whether he is entitled to relief under the Convention Against Torture. 18 LOPEZ-AGUILAR V. BARR

recognized that a taking with consent does not constitute a generic theft offense. See Matter of Garcia-Madruga, 24 I. & N. Dec. 436, 440–41 (B.I.A. 2008).

Under section 164.395 of the Oregon Revised Statutes, third-degree robbery has three elements. First, the defendant must “commit[] or attempt[] to commit theft or unauthorized use of a vehicle.” Or. Rev. Stat. § 164.395(1). Second, the defendant must, during “the course of committing or attempting to commit” the theft or unauthorized use of a vehicle, “use[] or threaten[] the immediate use of physical force upon another person.” Id. Third and finally, as to the use or threatened use of physical force, the defendant must act with the specific intent either to “[p]revent[] or overcom[e] resistance to the taking of the property or to retention thereof immediately after the taking” or to “[c]ompel[] the owner of such property or another person to deliver the property or to engage in other conduct which might aid in the commission of the theft or unauthorized use of a vehicle.” Id. § 164.395(1)(a)–(b).

Here, the BIA correctly recognized that the plain text of section 164.395 does not require that the defendant engage in a nonconsensual taking. To the contrary, Oregon law expressly provides that the first element—theft or attempted theft—can be satisfied through a consensual taking. Under section 164.015 of the Oregon Revised Statutes, “theft” is defined as including “theft by deception,” a taking that requires consent. Id. § 164.015. As the BIA observed here, however, the statute’s second and third elements require the use of force. The BIA noted that it had previously concluded, in Matter of Ibarra, 26 I. & N. Dec. 809 (B.I.A. 2016), that “[t]here is no meaningful difference between a taking of property accomplished against the victim’s will and one where his ‘consent’ to parting with his property is coerced

LOPEZ-AGUILAR V. BARR 19

through force, fear, or threats.” Id. at 811. Given the force requirement in section 164.395, the BIA reasoned, it would not be possible to commit third-degree robbery while engaging in a consensual taking.

As the majority appears to recognize, the BIA misread the statute. Section 164.395 requires only that the “use[] or threaten[ed] . . . immediate use of physical force” be “upon another person.” Or. Rev. Stat. § 164.395(1). That person need not be the property owner. Thus, under the plain text of the statute, it is possible to apply the force needed for a third- degree robbery against a third person while engaging in a taking that is consensual with regard to the robbery victim. The state statute at issue in Matter of Ibarra, by contrast, explicitly required that “the felonious taking of personal property in the possession of another . . . against his will.” 26 I. & N. Dec. at 810 n.2 (emphasis added) (quoting Cal. Penal Code § 211); see also id. at 812 (“[T]he jury instructions for section 211 of the California Penal Code require as an element that the defendant take property from another ‘against that person’s will.’” (citation omitted)).

II

Considering the statutory text discussed above, the majority rightly concludes that, on its face, section 164.395 defines third-degree robbery more broadly than a generic theft offense under the INA. Under our case law, that should have been enough. “Where . . . 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) (citation omitted) (quoting Duenas-Alvarez, 549 U.S. at 193), abrogated on other grounds by United States v. Stitt, 20 LOPEZ-AGUILAR V. BARR

139 S. Ct. 399 (2018); accord Chavez-Solis v. Lynch, 803 F.3d 1004, 1009–10 (9th Cir. 2015). Under those circumstances, “[t]he state statute’s greater breadth is evident from its text.” Grisel, 488 F.3d at 850. 2

The majority instead posits that, because section 164.395 requires that force be used “immediately after the taking,” it would be “unlikely that a defendant would be convicted for using or threatening force against a third party unless the force occurred in the presence of the owner, which would negate consent.” Maj. Op. at 12. The majority is incorrect.

First, to the extent the majority suggests that application of the plain text of section 164.395 in this nongeneric manner is possible but unlikely, that argument is foreclosed by our case law. As long as the application of the statute’s text in the nongeneric manner is not a logical impossibility, the relative likelihood of application to nongeneric conduct is immaterial. See United States v. Valdivia-Flores, 876 F.3d 1201, 1208 (9th Cir. 2017). For that reason, where “the difference in breadth is apparent on the face of the statute,” we have rejected arguments that a state statute “does not ‘extend significantly beyond’ its federal analogue for purposes of categorical comparison.” Id. (emphasis added) (quoting Duenas-Alvarez, 549 U.S. at 193); see also Cerezo v. Mukasey, 512 F.3d 1163, 1167–68 (9th Cir. 2008). Any nongeneric application evident from the statutory text means that the state statute cannot be a categorical match. See Valdivia-Flores, 876 F.3d at 1208.

2 The BIA recognized this principle but reasoned that there was an “apparent conflict” between section 164.395’s force elements and the possibility of theft by deception. As already discussed, there is no such conflict.

LOPEZ-AGUILAR V. BARR 21

If the majority instead believes that section 164.395 can never be applied in this nongeneric manner, it misunderstands the statute. Under Oregon law, “[r]obbery can include situations involving the use of force during flight.” State v. Tolbert, 433 P.3d 501, 505 (Or. Ct. App. 2018). In shoplifting cases, for example, “[f]light readily includes situations in which the thief is confronted directly upon leaving the store, and it extends beyond the store during fresh pursuit.” Id. Thus, contrary to the majority’s suggestion, the force required for robbery need not be applied “in the presence of the owner.” Maj. Op. at 12. 3

Further, Oregon law provides that force used during flight can be applied against third parties. As the Oregon Supreme Court has emphasized, “the term ‘victim’ is not used in any of the robbery statutes” and “the statutes generally only describe the person against whom force must be used as ‘any person’ or ‘another person.’” State v. Hamilton, 233 P.3d 432, 435 (Or. 2010). Thus, “the legislature deliberately chose not to limit the reach of the robbery statutes to the use of force against the owner.” Id. at 436.

Cases involving such third-party force are commonly prosecuted. One regularly recurring scenario involves security guards. Pereida-Alba v. Coursey, 342 P.3d 70 (Or. 2015), for example, involved a defendant who entered a

3 By allowing force used during flight to satisfy section 164.395’s force requirement, Oregon law diverged from the common law definition of robbery. “[U]nder the traditional view it is not robbery to steal property without violence or intimidation (e.g., to obtain it by stealth or fraud or sudden snatching), although the thief later, in order to retain the stolen property or make good his escape, uses violence or intimidation upon the property owner.” 3 Wayne R. LaFave, Substantive Criminal Law § 20.3(e) (3d ed. 2017). 22 LOPEZ-AGUILAR V. BARR

grocery store, filled a backpack with food, and left the store without paying. Id. at 71. When confronted by a security guard, the defendant pulled a gun from the backpack and pointed it at the guard. Id. at 72. The defendant was eventually arrested and charged with first-degree robbery in violation of section 164.415, which incorporates the definition of third-degree robbery under section 164.395. Id. The indictment accordingly alleged that the defendant “did unlawfully and knowingly while in the course of committing theft . . . use and threaten the immediate use of physical force upon [the security guard].” Id. at 72 (alteration in original).

Likewise, in State v. Tolson, 546 P.2d 1115 (Or. Ct. App. 1976), a defendant shoplifted three pairs of pants from a department store. Id. at 1116. Upon observing the theft, security guards followed the defendant out of the store; the defendant began running before she was caught by one of the guards. Id. The defendant “fought him, using a shopping bag, fists and one of her shoes as weapons.” Id. The Oregon Court of Appeals held that this force was sufficient for a second-degree robbery conviction under section 164.405, which likewise incorporates the definition of third-degree robbery under section 164.395. Id. at 1117. As Tolson explained, “[i]t is not stretching the meaning of the word to hold that, where hot pursuit continues for 455 feet with overtaking and use of force at that point, the entire action was ‘immediate.’” Id.

The majority notes that “Oregon treats security guards as authorized agents of the owner.” Maj. Op. at 12 n.2. Whether that is true of the security guards in the cases discussed above is unclear. 4 But that ambiguity simply underscores the fact

4 Whether a security guard is considered an employee of a store is a fact-specific inquiry that depends on the degree of control exercised by

LOPEZ-AGUILAR V. BARR 23

that section 164.395 is phrased such that no analysis of whether the force victim is the property owner is required. See Hamilton, 233 P.3d at 435. Moreover, the principle underlying the security-guard cases is that force used against an intervening party is enough to trigger application of section 164.395. There is no reason that this principle would not also extend to a police officer, acquaintance, or other Good Samaritan attempting to stop a theft.

To be sure, where “[t]he government contends that such a literal application of the statute is not realistic,” we must “consider whether [state] courts have interpreted the scope of [the statute of conviction] more narrowly so as to make it applicable only to conduct” consistent with the federal analogue. Cerezo, 512 F.3d at 1167–68; accord Cortez- Guillen v. Holder, 623 F.3d 933, 935 (9th Cir. 2010). Oregon law does recognize that “intervening time, distance, and events may well break the link between the theft and the subsequent use of force.” Tolbert, 433 P.3d at 506. But that principle does not constrain section 164.395 “so as to make it applicable only to conduct” involving a nonconsensual taking. Cerezo, 512 F.3d at 1167–68. Even with the “immediately after” limitation, it is evident from the examples discussed above that the use or threatened use of force need not occur “in the presence of the owner” and “negate consent,” as the majority suggests. Maj. Op. at 12.

III

Because section 164.395’s text is on its face broader than the INA’s generic theft offense, I do not believe that Lopez- Aguilar was required to establish a realistic probability that

the store. See Gibson v. Safeway Stores, Inc., 764 P.2d 548, 548–49 (Or. 1988). 24 LOPEZ-AGUILAR V. BARR

Oregon would apply its statute in a nongeneric manner. But even if Lopez-Aguilar were required to make that showing, Oregon case law demonstrates that section 164.395 realistically applies to conduct falling outside generic theft.

In determining whether there is “a realistic probability, not a theoretical possibility, that the State would apply its statute to conduct that falls outside the generic definition of a crime,” Duenas-Alvarez, 549 U.S. at 193, our decisions have not required a case involving an actual prosecution of the state offense in a nongeneric manner. Indeed, we have observed that, for some crimes, it is entirely possible that “the absence of appellate decisions” should be expected, given “the relatively light sentence that would result,” thereby “encourag[ing] prosecutors to charge violations that carry more substantial sentences.” United States v. Brown, 879 F.3d 1043, 1050 (9th Cir. 2018). Under those circumstances, “[c]onvictions . . . generally result from negotiations and guilty pleas, which do not produce appeals.” Id.

Consistent with this observation, one scholar has noted that it would “make[] little sense” to require a state appellate decision involving an actual prosecution of nongeneric conduct before concluding that there is a realistic probability that the state statute would be so applied:

Nearly all criminal cases—approximately 90 percent—are resolved through plea bargain. And because plea agreements will typically include a waiver of a defendant’s appellate rights, most cases involve no appeal. That means the cases that end up appealed are a small fraction of total cases. And even where there is an appeal, most cases are disposed of in short decisions that

LOPEZ-AGUILAR V. BARR 25

come with almost no discussion of the facts of the case. As a result, only a miniscule percentage of all prosecutions under a particular statute will end up producing an appellate decision that includes a discussion of the facts of the case. Given that fact, there is no reason to think that the cases that end up in an appellate decision are in any way representative—let alone exhaustive—of the types of cases that the state prosecutes.

Doug Keller, Causing Mischief for Taylor’s Categorical Approach: Applying “Legal Imagination” to Duenas- Alvarez, 18 Geo. Mason L. Rev. 625, 659–60 (2011) (footnotes omitted).

Rather, we have concluded that the requirement that a “realistic probability” be shown is satisfied when application to nongeneric conduct would logically follow from the statute’s text and applicable precedents. Chavez-Solis v. Lynch, for example, considered whether a California conviction for the possession of child pornography was broader than its federal analogue. 803 F.3d at 1006. Although both the state and federal crimes required that the illicit material depict a minor engaging in “sexual conduct,” the state statute’s definition of “sexual conduct” swept more broadly. Id. at 1008. Chavez-Solis rejected the government’s argument that there was not a “realistic probability” that the state would apply the statute to a depiction of “sexual conduct” under the nongeneric aspects of the statute, noting that one state appellate decision had construed “sexual conduct” in the nongeneric manner. Id. at 1010. Although that state appellate decision did not involve the statute of conviction, Chavez-Solis reasoned that prosecution for nongeneric conduct under the statute of conviction would 26 LOPEZ-AGUILAR V. BARR

logically follow, because the state appellate decision involved the same statutory phrase, “sexual conduct.” Id. at 1011–12.

Similarly, Medina-Lara v. Holder, 771 F.3d 1106 (9th Cir. 2014), concluded that a California conviction for illegal firearm possession was not a categorical match with the generic offense for firearm possession under federal law, because the state statute criminalized possession of antique firearms, while the federal analogue did not. Id. at 1116. Although there were no cases involving prosecutions under the statute of conviction for antique firearms, Medina-Lara noted that there were cases involving prosecutions of antique-firearm possession under similar statutes. Id. Those cases, Medina-Lara reasoned, logically suggested that prosecution for antique-firearm possession under the statute of conviction was possible, thereby establishing a realistic probability under Duenas-Alvarez. Id.

Here, two strands of Oregon case law indicate that state courts would apply section 164.395 to a consensual taking. First, as already discussed, it is abundantly clear that the force required under section 164.395 can be applied against persons other than the property owner. That fact pattern that appears not infrequently in cases involving shoplifting. See, e.g., Pereida-Alba, 342 P.3d at 71–72; Tolson, 546 P.2d 1116–17.

Second, the Oregon Supreme Court has noted that a consensual taking—namely, theft by deception—can occur during a shoplifting offense. In State v. Fonte, 422 P.3d 202 (Or. 2018), a defendant on two occasions took a pair of jeans from the sales floor of a department store and returned them for cash. Id. at 203. On the second instance, the defendant was stopped by security guards before he was able to leave the store. Id. The defendant was ultimately charged with two

LOPEZ-AGUILAR V. BARR 27

counts of first-degree theft on a theory of “theft by receiving.” Id. at 203–04. The Oregon Supreme Court noted, however, that “[t]he facts also supported charges of theft by deception of the money, which, given the value, would have been a misdemeanor.” Id. at 204 n.1. Had the defendant in Fonte used force against the security guards, as the defendants in Pereida-Alba and Tolson had, there is at least a realistic probability he would have been prosecuted for robbery.

That Oregon law would allow for such a prosecution is consistent with modern developments in the crime of robbery. As already discussed, under the traditional common law, “it is not robbery to steal property without violence or intimidation,” even if “the thief later, in order to retain the stolen property or make good his escape, uses violence or intimidation upon the property owner.” 3 LaFave, supra, § 20.3(e); see also supra note 3. This limitation reflects that, in accordance with “the general principle of criminal law . . . that the defendant’s conduct and his state of mind must concur,” the traditional crime of robbery required “that the defendant’s larcenous conduct (his taking of the victim’s property) and his violence-or-intimidation conduct . . . concur.” 3 LaFave, supra, § 20.3(e). Because Oregon—as well as other states—has expanded the definition of robbery to include force during flight, “a different result is often possible today.” Id.; see also Tolbert, 433 P.3d at 505. One such result is that, as one treatise suggests, robbery may now include situations when the theft has been committed “by stealth or fraud.” 3 LaFave, supra, § 20.3(e).

IV

In sum, third-degree robbery under section 164.395 of the Oregon Revised Statutes is, on its face, not a categorical theft offense as defined by the INA, and that mismatch is 28 LOPEZ-AGUILAR V. BARR

confirmed by Oregon case law. Because the majority’s application of the realistic-probability test is not only unnecessary but also erroneous, I respectfully dissent.