Chmukh v. Garland

FOR PUBLICATION UNITED STATES COURT OF APPEALS FOR THE NINTH CIRCUIT VITALIY CHMUKH, No. 21-1096 Agency No. Petitioner, A079-176-419 v. OPINION MERRICK B. GARLAND, Attorney General, Respondent. On Petition for Review of an Order of the Board of Immigration Appeals Argued and Submitted April 4, 2024 Pasadena, California Filed December 23, 2024 Before: Ryan D. Nelson, Lawrence VanDyke, and Gabriel P. Sanchez, Circuit Judges. Opinion by Judge R. Nelson; Concurrence by Judge VanDyke; Partial Concurrence and Partial Dissent by Judge Sanchez 2 CHMUKH V. GARLAND SUMMARY * Immigration Denying Vitaliy Chmukh’s petition for review a decision of the Board of Immigration Appeals, the panel concluded that Chmukh’s conviction for possession of a stolen vehicle, under Revised Code of Washington (RCW) § 9A.56.068, was an aggravated felony and a particularly serious crime. Because the panel concluded that the conviction was an aggravated felony, the panel explained that the criminal- alien jurisdictional bar applied, but the panel retained jurisdiction to review Chmukh’s claims of legal error. The panel also assumed without deciding that Chmukh’s claims were exhausted. The panel concluded that Chmukh’s conviction for possession of a stolen vehicle was an aggravated felony under 8 U.S.C. § 1101(a)(43)(G) that made him ineligible for asylum. The panel concluded that the statute is a categorical match to the generic offense because both require possession of stolen property, actual knowledge that the property was stolen, and intent to deprive the owner of property. The panel rejected Chmukh’s arguments that the statute was overbroad. The panel concluded that Chmukh’s conviction was a particularly serious crime barring withholding of removal under 8 U.S.C. § 1231(b)(3)(B). Chmukh contended that the BIA erred by omitting the elements of his conviction. The * This summary constitutes no part of the opinion of the court. It has been prepared by court staff for the convenience of the reader.  CHMUKH V. GARLAND 3 panel observed that the BIA must first analyze the elements of the crime to see if it is potentially within the ambit of a particularly serious crime, but concluded that failing to list the elements was not error in this case. The panel also rejected Chmukh’s contention that the agency erred by ignoring appropriate factors and proper evidence. Finally, Chmukh argued that the agency failed to explicitly explain why he is a danger to the community, but the panel concluded that omitting this analysis was not a misapplication of the governing standard. Concurring, Judge VanDyke wrote to address Abebe v. Gonzales, 432 F.3d 1037 (9th Cir. 2005) (en banc), which held that a petitioner need not raise an issue before the BIA to exhaust the issue if the BIA summarily affirms under Matter of Burbano, 20 I. & N. Dec. 872 (BIA 1994). Judge VanDyke wrote that Abebe encourages a perverse gamesmanship where petitioners could trim their strongest arguments out of their appeal in the hopes that the BIA will issue a summary affirmance, and then hammer their arguments before this court in hopes of securing a remand and stalling removal. Judge VanDyke wrote that this court should take an appropriate case en banc to fix Abebe’s wrongheaded approach to exhaustion. Concurring in part and dissenting in part, Judge Sanchez agreed that Chmukh’s conviction qualified as an aggravated felony, but wrote that the BIA violated its own precedent in its particularly serious crime determination and failed to explain why it departed from its own requirements. First, the BIA failed to undertake the threshold elements-only inquiry required by governing BIA precedent. Second, the BIA failed to explain how the relevant factors justified the presumption that Chmukh is a danger to the community. Judge Sanchez wrote that circuit and agency precedent has 4 CHMUKH V. GARLAND never held that the BIA may forego the threshold analysis of determining and explaining why the nature and circumstances of an offense indicate that a petitioner poses such danger. COUNSEL Marcus Curtis (argued), Blaine H. Evanson, Gibson Dunn & Crutcher LLP, Irvine, California, for Petitioner. Imran R. Zaidi (argued), Senior Litigation Counsel, Criminal Immigration Team; Lindsay B. Glauner, Senior Litigation Counsel; Office of Immigration Litigation; Brian M. Boynton, Principal Deputy Assistant Attorney General; Civil Division, United States Department of Justice, Washington, D.C.; for Respondent.  CHMUKH V. GARLAND 5 OPINION R. NELSON, Circuit Judge: Vitaliy Chmukh petitions for review of the Board of Immigration Appeals’ (BIA) decision denying his claims for asylum, withholding of removal, and Convention Against Torture (CAT) protection. The BIA correctly determined that Chmukh’s conviction for possession of a stolen vehicle was an aggravated felony subject to removal. Likewise, the BIA did not err in finding that this conviction was a particularly serious crime, rendering Chmukh ineligible for withholding relief. And Chmukh waived any argument related to his CAT claim. Thus, we deny his petition for review. I Along with his parents and four siblings, Vitaliy Chmukh, a native and citizen of Ukraine, came to the United States as a refugee in 2001. For many of the past 23 years, he lived in Washington with his family. Starting in 2017, Chmukh and a friend used a stolen vehicle to steal packages from others’ porches. Eventually, the police arrested Chmukh. He pleaded guilty to possession of a stolen vehicle in violation of Revised Code of Washington (RCW) § 9A.56.068. That same day, Chmukh also pleaded guilty to possession of a controlled substance—heroin—in violation of RCW § 69.50.4013, stemming from a separate incident. A Washington judge sentenced him to a concurrent sentence—43 months for possession of a stolen vehicle and 24 months for the controlled substance violation. He ultimately served 38 months. 6 CHMUKH V. GARLAND The Department of Homeland Security (DHS) charged Chmukh with removability based on these prior convictions. The DHS alleged that his conviction for possession of a stolen vehicle was an “aggravated felony” as defined in the Immigration and Nationality Act (INA). 8 U.S.C. § 1101(a)(43)(G). That section defines an “aggravated felony” as a “theft offense (including receipt of stolen property) or burglary offense” for which a term of imprisonment of at least one year was imposed. Id. The DHS also maintained that his conviction for possession of a controlled substance violated the Controlled Substances Act, 21 U.S.C. § 802. Cf. 8 U.S.C. § 1227(a)(2)(B)(i). An immigration judge (IJ) found Chmukh removable given his admissions and the government’s evidence. In response, Chmukh applied for asylum, withholding of removal, and protection under the CAT. He asserted his fear of persecution and torture by Ukrainian government authorities on account of his religious beliefs and political opinions. The IJ found Chmukh removable and denied his applications. The IJ found that his conviction for possession of a stolen vehicle was an aggravated felony, rendering him ineligible for asylum. The same conviction also made him ineligible for withholding of removal because it was a particularly serious crime. Finally, the IJ denied Chmukh’s application for CAT protection because the evidence failed to show it was more likely than not that he would be tortured if removed to Ukraine. The BIA affirmed the IJ’s decision in full. It agreed with the IJ’s finding and adopted the IJ’s reasoning that Chmukh’s conviction for possession of stolen property was an aggravated felony and a particularly serious crime that  CHMUKH V. GARLAND 7 rendered him ineligible asylum and for withholding of removal. The BIA also affirmed the denial of CAT protection. II We have jurisdiction under 8 U.S.C. § 1252. We review both the IJ and the BIA decisions when the BIA “adopts and affirms” the IJ’s decision and provides its own analysis (as it did here). Alcaraz-Enriquez v. Garland, 19 F.4th 1224, 1229 n.2 (9th Cir. 2021). We review de novo the agency’s determination that a crime constitutes an “aggravated felony.” United States v. Flores, 901 F.3d 1150, 1155 (9th Cir. 2018). Because we conclude that Chmukh was convicted of an aggravated felony, the criminal-alien jurisdictional bar of 8 U.S.C. § 1252(a)(2)(C) applies. See Park v. Garland, 72 F.4th 965, 973 (9th Cir. 2023). We retain jurisdiction, however, to review Chmukh’s arguments that the agency committed legal error by applying the wrong legal standard or misapplying its own precedent in determining that he was convicted of a particularly serious crime under the limited review provision. See 8 U.S.C. § 1252(a)(2)(D). “In reviewing whether the BIA applied the correct legal standard in its particularly-serious-crime analysis, we consider ‘whether the agency relied on the appropriate factors and proper evidence to reach [its] conclusion.’” Park, 72 F.4th at 974 (quoting Flores-Vega v. Barr, 932 F.3d 878, 884 (9th Cir. 2019) (alteration in original)). And we disturb the agency’s judgment “only if it ‘acted arbitrarily, irrationally, or contrary to law’ by failing to apply or misapplying the proper standard.” Id. (quoting Bare v. Barr, 975 F.3d 952, 961 (9th Cir. 2020)). 8 CHMUKH V. GARLAND III Chmukh asks us to review the agency’s denial of asylum and withholding of removal. 1 Before addressing the merits, the government asks us to decide whether Chmukh exhausted his claims before the agency. See Umana- Escobar v. Garland, 69 F.4th 544, 550 (9th Cir. 2023). We assume, without deciding, that Chmukh fulfilled the exhaustion requirements. On the merits, Chmukh’s conviction for possession of a stolen vehicle is an aggravated felony. Thus, he is ineligible for discretionary relief such as asylum. See Moncrieffe v. Holder, 569 U.S. 184, 187 (2013) (citing 8 U.S.C. §§ 1158(b)(2)(A)(ii), (B)(i); §§ 1229b(a)(3), (b)(1)(C)). Nor did the BIA err in finding that Chmukh was convicted of a particularly serious crime. So withholding of removal is also unavailable to him. See 8 U.S.C. § 1231(b)(3)(B)(ii); 8 C.F.R. § 1208.16(d)(2). A “Any alien who is convicted of an aggravated felony at any time after admission is deportable.” § 1227(a)(2)(A)(iii). The INA defines an “aggravated felony” to include “a theft offense (including receipt of stolen property) . . . for which the term of imprisonment [is] at least one year.” § 1101(a)(43)(G). Ultimately, it is the “conviction, not conduct” that serves “as the trigger for immigration consequences.” Mellouli v. Lynch, 575 U.S. 798, 806 (2015). Thus, we employ a “categorical approach” that examines “the statute of conviction, rather than [] the 1 Because his opening brief fails to address his CAT claim, that argument is waived. See Martinez–Serrano v. INS, 94 F.3d 1256, 1259 (9th Cir. 1996).  CHMUKH V. GARLAND 9 specific facts underlying the crime.” Esquivel-Quintana v. Sessions, 581 U.S. 385, 389 (2017) (cleaned up). Under this approach, we assess whether the state statute “fits within the ‘generic’ federal definition of a corresponding aggravated felony.” Moncrieffe, 569 U.S. at 190 (citation omitted). Assessment of fit requires us to “compare the elements of the crime of conviction with the elements of the ‘generic’ version of the listed offense—i.e., the offense as commonly understood.” Mathis v. United States, 579 U.S. 500, 503 (2016) (applying the same test in the context of the Armed Career Criminal Act) (emphasis in original); see also Flores, 901 F.3d at 1154, 1156 (determining that a California conviction for receipt of stolen property “fits within the generic definition of theft”) (citation omitted). If a state statute “sweeps more broadly” than the generic federal crime, then it is not a categorical match. Alfred v. Garland, 64 F.4th 1025, 1031 (9th Cir. 2023) (en banc) (citation omitted). In that situation, a conviction under that state law cannot be an aggravated felony. See id. 1 Chmukh was convicted for possession of a stolen vehicle under RCW § 9A.56.068. The IJ found that receipt of stolen property was an aggravated felony. The INA does not define receipt of stolen property, so we must use the generic federal definition or “the offense as commonly understood.” See Mathis, 579 U.S. at 503. We have previously held that a generic federal offense of receipt of stolen property includes these elements: (1) “receipt, possession, concealment, or retention of property,” (2) “knowledge or belief that the property has been stolen,” and (3) “intent to deprive the owner of his property.” Flores, 901 F.3d at 1160 (citing Matter of Deang, 27 I. & N. Dec. 57, 59–63 (B.I.A. 2017)). 10 CHMUKH V. GARLAND As for the Washington state statute, “[a] person is guilty of possession of a stolen vehicle if he or she possess [possesses] a stolen motor vehicle.” WASH. REV. CODE § 9A.56.068 (second brackets in original). Possessing a stolen vehicle means to “knowingly [] receive, retain, possess, conceal, or dispose of stolen property knowing that it has been stolen and to withhold or appropriate the same to the use of any person other than the true owner or person entitled thereto.” Id. § 9A.56.140(1). Because the terms “receive, retain, possess, conceal, or dispose” are “closely related,” the Washington Supreme Court has clarified that “possession of stolen property [under the statute] is a single means crime.” State v. Tyler, 422 P.3d 436, 439–40 (Wash. 2018) (en banc) (citation omitted). Put differently, the statute includes a “multifaceted description of the ways in which one may possess stolen property,” and this is “properly regarded as definitional, enhancing the understanding of the single means crime.” Id. at 439. Washington’s stolen vehicle statute also requires actual knowledge that the vehicle was stolen. See WASH. REV. CODE § 9A.56.140(1). The statute requires not only that the defendant “knowingly” possess the stolen property but also have a state of mind of “knowing that it has been stolen.” Id. A person “acts knowingly” when he is either “aware of a fact, facts, or circumstances or result described by a statute defining an offense,” or “has information which would lead a reasonable person in the same situation to believe that facts exist which facts are described by a statute defining an offense.” Id. § 9A.08.010(b)(i)-(ii); see also State v. Allen, 341 P.3d 268, 273 (Wash. 2015) (“To pass constitutional muster, the jury must find actual knowledge but may make such a finding with circumstantial evidence.”) (citation omitted).  CHMUKH V. GARLAND 11 We now must compare these elements against the elements of the generic federal offense, see Mathis, 579 U.S. at 503, and assess whether the state statute “sweeps more broadly” than the generic crime. Alfred, 64 F.4th at 1031 (citation omitted). If it does, then it cannot be an aggravated felony. Id. When comparing a state statute with its generic federal counterpart, we may examine the statutory texts to see if “greater breadth is evident from [the] text.” Id. at 1043 (quoting Lopez-Aguilar v. Barr, 948 F.3d 1143, 1147 (9th Cir. 2020)); see also Flores, 901 F.3d at 1160–61. In Flores, we held that California’s stolen property statute is a categorical match with the generic federal offense of receipt of stolen property. 901 F.3d at 1160–61. We concluded so because the statutory language included a matching possession element. Id. And it contained a matching mens rea since it required “actual knowledge of or belief that the property is stolen.” Id. at 1161. This mens rea language also satisfied the generic federal offense’s intent element because “the act of buying or receiving stolen property [while] knowing it was stolen inherently entails the intent to deprive the owner of rights and benefits of ownership.” Id. (cleaned up). Because this state statute matched the generic federal offense, the defendant’s state conviction was categorically an aggravated felony. Id. Applying our analysis from Flores leads us to conclude that the Washington statute criminalizing possession of a stolen vehicle is also a categorical match with the generic federal offense. Both the generic offense and the state statute require possession of stolen property. Compare WASH. REV. CODE § 9A.56.068 (“A person is guilty of possession of a stolen vehicle if he or she possess [possesses] a stolen motor vehicle.”) (brackets in original) with Flores, 901 F.3d at 1160 (generic federal offense requires “receipt, 12 CHMUKH V. GARLAND possession, concealment, or retention of property”). Both statutes also require actual knowledge that the property was stolen. Compare State v. Jones, 463 P.3d 738, 747 (Wash. Ct. App. 2020) (“Knowing” “demands a subjective standard of knowledge when the State must prove the mens rea of ‘knowledge’ in order to convict the accused of a crime”) with Flores, 901 F.3d at 1160. And since actual knowledge requires an intent to deprive the owner of his property, the state statute also matches the generic offense’s intent requirement. Flores, 901 F.3d at 1161 (citation omitted). Thus, an offense under RCW § 9A.56.968 categorically qualifies as an aggravated felony. Thus, Chmukh is subject to removal. See § 1227(a)(2)(A)(iii). 2 Chmukh also contends that the Washington law is overbroad. First, he argues that the statute’s knowledge element is “watered down” because a jury in Washington could presume “actual knowledge by a finding of constructive knowledge.” This argument is unpersuasive. True, Washington law allows a jury to find actual knowledge based on circumstantial evidence. See Allen, 341 P.3d at 273. But Chmukh fails to explain how that allowance is otherwise disallowed by the generic federal offense. Cf. United States v. Gallo, 543 F.2d 361, 367 (D.C. Cir. 1976) (“holding that the jury may conclude from circumstantial evidence that an accused possessed the requisite knowledge for the statutory offense”). Again, the point of the categorical approach is to compare the elements themselves. Here, what matters is that both the state statute and the generic federal offense require that the jury finds the defendant possesses the same fundamental mens rea element. See Flores, 901 F.3d at 1160 (citing Matter of Deang, 27 I. & N. at 59–63). Whether the generic offense  CHMUKH V. GARLAND 13 considers circumstantial evidence to prove actual knowledge is irrelevant. Chmukh raises a separate issue with the possession element of the Washington statute. The generic federal offense prohibits only the “receipt, possession, concealment, or retention” of property. The state law’s definition of “possession,” however, includes another prohibition on “disposal.” WASH. REV. CODE § 9A.56.140(1) (making it a crime “to receive, retain, possess, conceal, or dispose of stolen property”) (emphasis added). Given this statutory language, Chmukh argues that the Washington law makes “disposing of” a “distinct act that, by definition, does not depend on the other forms of ‘possession’ as antecedent acts.” In his view, the state statute is therefore overbroad because it criminalizes more conduct than the generic federal offense. This argument, however, contradicts state law. The Washington Supreme Court explains that “the five terms in RCW § 9A.56.140(1) are so closely related that they do not describe distinct acts apart from actually possessing the stolen vehicle but are merely facets of the same criminal conduct.” Tyler, 422 P.3d at 440 (cleaned up). Given this interpretation, no legally significant distinction exists between Washington’s statute and the generic federal definition of possession. See Chavez-Solis v. Lynch, 803 F.3d 1004, 1007–08 (9th Cir. 2015) (where “there is no legally significant distinction between [the federal and state] terms,” the difference does not matter for assessing whether a crime is an aggravated felony). The state statute is a categorical match with the generic federal offense. Chmukh’s counsel strained to present a hypothetical where someone could dispose of property without ever possessing it. But the Washington Supreme 14 CHMUKH V. GARLAND Court has soundly rejected this hypothetical example, noting that it is “hard to imagine a situation where a person . . . disposes of a stolen vehicle without also possessing it at some time.” Tyler, 422 P.3d at 440 (citation omitted). The categorical approach “is not an invitation to apply ‘legal imagination’ to the state offense.” Moncrieffe, 569 U.S. at 191. There is thus no “realistic” probability that Washington “would apply its statute to conduct that falls outside the generic definition” of possession. See id. In sum, Chmukh’s conviction for possession of a stolen vehicle under Washington law qualifies as an aggravated felony subject to removal. B Besides the fact that his crime was an aggravated felony, Chmukh does not qualify for relief from removal because his crime was also a particularly serious crime. Ordinarily, withholding of removal prevents the removal of a noncitizen to a country where that individual’s life or freedom would be threatened because of their race, religion, nationality, membership in a particular social group, or political opinion. § 1231(b)(3)(A). This relief, however, does not apply to noncitizens who have been convicted of a “particularly serious crime.” § 1231(b)(3)(B)(ii); 8 C.F.R. § 1208.16(d)(2); Delgado v. Holder, 648 F.3d 1095, 1101 (9th Cir. 2011) (en banc). That said, a “particularly serious crime” is not precisely defined under the statute. See § 1231(b)(3)(B)(ii). But for offenses like Chmukh’s, that result in more than one but less than five years imprisonment, the BIA (as the Attorney General’s delegate) retains discretion to decide whether the offense is particularly serious. § 1231(b)(3)(B); see also Delgado, 648 F.3d at 1098.  CHMUKH V. GARLAND 15 In evaluating whether a crime is particularly serious, the BIA considers the Frentescu factors: (1) the nature of the conviction, (2) the type of sentence imposed, and (3) whether the circumstances and underlying facts of the conviction “justify the presumption that the convicted immigrant is a danger to the community.” Delgado, 648 F.3d at 1107 (citing Matter of Frentescu, 18 I. & N. Dec. 244, 247 (B.I.A. 1982)). The BIA begins its analysis by determining whether the elements of the crime of conviction “potentially bring the crime into a category of particularly serious crimes.” Mendoza-Garcia v. Garland, 36 F.4th 989, 999 (9th Cir. 2022) (citing In re N-A-M-, 24 I. & N. Dec. 336, 342 (B.I.A. 2007)) (quotation marks omitted). “If so, the BIA then considers ‘all reliable information’ in analyzing the remaining two [Frentescu] factors.” Id. (quoting In re N-A-M-, 24 I. & N. Dec. at 342). Chmukh argues that the agency failed to properly follow this analysis. We disagree. 1 The BIA must first analyze the elements of the crime to see if it will “potentially bring the offense within the ambit of a particularly serious crime.” Bare, 975 F.3d at 961–62 (citing In re N-A-M-, 24 I. & N. Dec. at 342). Chmukh contends that the BIA erred by omitting the elements of his conviction. Failing to list the elements of the crime, however, was not error and does not justify disturbing the agency’s judgment. Park, 72 F.4th at 974–75. In Bare, where the defendant was convicted of an aggravated felony—possession of a firearm—it was “a straightforward, well-known federal crime with simple elements.” 975 F.3d at 962. Thus, under such a “common federal crime with simple and straightforward elements,” “we will not require an explicit consideration of the elements of the offense.” Id. at 963. 16 CHMUKH V. GARLAND That reasoning applies here. “As an aggravated felony,” Chmukh’s conviction for possession of a stolen vehicle “is one of the types of crimes ‘most likely to be’ particularly serious,” id. at 962 (quotation omitted), even when the aggregate sentence is less than five years. See id. at 964. In addition, his conviction contained simple elements— possession of stolen property and actual knowledge—that the agency recited. See id. at 962. And the agency “referenced facts that went directly to each element in [its] analysis.” Id. So the agency did not err by overlooking “an explicit consideration of the elements of the offense.” See id. at 962–63. 2 The agency also considers “all reliable information . . . including the conviction records and sentencing information, as well as other information outside the confines of a record of conviction” in making a particularly serious crime determination. Alcaraz-Enriquez, 19 F.4th at 1231 (citation omitted). Chmukh argues that the agency erred by ignoring “appropriate factors” and “proper evidence” in its particularly serious crime reasoning. Again, we find this argument unpersuasive. Chmukh argues that the agency impermissibly considered his possession of a controlled substance in evaluating whether his conviction for possession of a stolen vehicle was a particularly serious crime. That is an inaccurate characterization. The agency did mention that “[Chmukh] had been arrested for a separate offense of possession of heroin.” But nothing in the record suggests that the agency considered this offense as part of the particularly serious crime analysis. Indeed, the agency noted that it was a “separate” conviction.  CHMUKH V. GARLAND 17 Chmukh also criticizes the agency for finding that he “knew” the vehicle was stolen despite his contrary testimony. But Chmukh’s conviction required the prosecution to prove actual knowledge. See Allen, 341 P.3d at 273. And a defendant “may not assert a cross-examination right to prevent the government from establishing uncontested facts.” Hernandez v. Garland, 52 F.4th 757, 767 (9th Cir. 2022) (emphasis in original) (citation omitted). The government correctly notes that Chmukh’s own admissions establish an uncontested fact that he knew it was a stolen vehicle. Thus, there is nothing to persuade us that the agency’s judgment should be disturbed. 3 Finally, Chmukh argues that the agency failed to explicitly explain why Chmukh is a danger to the community. It is true that the agency did not specifically mention this standard. But omitting this analysis was not error. Only “citing the nature of the conviction, the circumstances and underlying facts of the conviction, and the type of sentence imposed” is sufficient for determining a particularly serious crime. Konou v. Holder, 750 F.3d 1120, 1127 (9th Cir. 2014) (citation omitted). The agency faithfully applied that standard. The agency described the nature and circumstances of Chmukh’s conviction, possession of a stolen vehicle to steal property. It also noted the type of sentence imposed by mentioning that Chmukh served 38 months of a 43-month sentence. That analysis satisfies the requirements outlined by Konou. Even though the agency omitted a specific discussion of how Chmukh posed a danger to the community, it was not a misapplication of the governing standard. The agency does 18 CHMUKH V. GARLAND not err when it considers the circumstances and nature of a petitioner’s conviction. That analysis is enough to satisfy consideration of whether the petitioner presents a “danger to the community.” When those considerations are addressed—as they were here—omitting a specific discussion of “danger” or “danger to the community” is not error when evaluating whether a petitioner committed a particularly serious crime. To reach this conclusion, we are guided by our decision in Gomez-Sanchez v. Sessions, 892 F.3d 985, 991 (9th Cir. 2018). There, we stated that “a crime is particularly serious if the nature of the conviction, the underlying facts and circumstances, and the sentence imposed justify the presumption that the convicted immigrant is a danger to the community.” Id. (cleaned up). And we noted that there is no “statutory requirement for a separate determination of dangerousness focusing on the likelihood of future serious misconduct on the part of the alien.” Id. (cleaned up). “Rather, once an individual is found to have been convicted for committing a particularly serious crime, he or she shall be considered to constitute a danger to the community.” Id. (cleaned up) (emphasis in original). The agency’s description and explanation of Chmukh’s offenses show why he committed a particularly serious crime and was therefore a danger to the community. Thus, the agency’s failure to explain in any more detail why Chmukh would be a “danger to the community” was not a misapplication of the governing standard. IV Chmukh was guilty of an aggravated felony and the agency did not err in concluding that he committed a  CHMUKH V. GARLAND 19 particularly serious crime. Thus, Chmukh was removable and ineligible for withholding relief. PETITION DENIED. VANDYKE, Circuit Judge, concurring: I fully agree with the majority opinion. I write separately only to address our circuit’s requirements for exhaustion of issues before the BIA when the BIA summarily affirms the IJ’s decision. It turns out there are hardly any. In Abebe v. Gonzales, this court erroneously concluded that when the BIA summarily affirms an IJ’s decision by citing Matter of Burbano, 20 I. & N. Dec. 872 (BIA 1994), the normal requirement that a petitioner must actually raise a specific issue before the BIA to present it in a petition to our court is essentially eliminated, because any issue generally related to the IJ’s decision is deemed exhausted. See Abebe v. Gonzales, 432 F.3d 1037, 1040–41 (9th Cir. 2005) (en banc). What this practically means is that a petitioner is strongly incentivized not to raise perceived flaws with the IJ’s analysis before the BIA, hope the busy BIA simply summarily affirms and adopts the IJ’s decision (which, why wouldn’t it do that, since the petitioner hasn’t pointed out anything wrong with the IJ’s decision?), and then raise any alleged flaws with the IJ’s decision for the first time with our court. Abebe in essence affords asylum petitioners a judicial bypass of BIA review. The only thing they need to do to qualify for the privilege of skipping over the agency? Keep their mouths shut. 20 CHMUKH V. GARLAND This perverse incentive to sandbag the agency is particularly powerful in cases like this. Everyone knows the usual game in asylum cases is not about making ultimately meritorious claims. It’s instead about buying the unlawfully present alien more time. So if the petitioner can “save” a procedural flaw in the IJ’s decision until this court reviews his case, and then get a remand from our court back to the agency, that’s a banking error in petitioner’s favor: go directly to the BIA, and collect a few more years on your stay-in-the-United-States-illegally-clock while you’re at it. And even if, as is typically true, the actual substance of his claims lack merit. This case demonstrates that absurdity perfectly. Chmukh’s arguments here relating to the particularly serious crime determination are that the IJ abused its discretion because it completely failed to address certain necessary issues. But the Petitioner failed to make this argument before the BIA. And the BIA, on appeal, summarily adopted the IJ’s analysis, so whether you look at the BIA decision itself or treat it as effectively repeating the IJ’s decision, the BIA too completely failed to address those same issues. So to summarize: the IJ failed to address this issue, the petitioner failed to present it to the BIA, and the BIA also failed to address it. At no point anywhere was the issue raised or discussed. But under Abebe, the issue is somehow exhausted. That is nuts. It completely defeats every purpose behind exhaustion. But thanks to Abebe, Chumkh can hope to buy himself more time with a remand on an issue that literally no one addressed until it was raised for the first time with this court. Even though his arguments have no merit, as the majority opinion well explains, and were not presented to the agency, we still must consider them. That is the silly law of this circuit under Abebe.  CHMUKH V. GARLAND 21 If Chumkh had raised the issue to the BIA that he now presents to us, then the BIA could have made an informed decision whether to give the IJ another opportunity to consider it. That is how our immigration review system is designed to work. But Chumkh never did, so the BIA obviously had no reason to address that issue or decide whether remand was appropriate. Under Abebe, if there were two judges on this panel that took the position of our dissenting colleague, this case would be going back to the agency for years more of proceedings, with Chumkh getting years more unmerited time in the United States, simply because he was smart enough to lay behind the log and not present an issue to the BIA—the body specifically designed to consider such arguments. I. In Abebe, our court in a 6-5 en banc decision held that a petitioner does not actually have to raise an issue before the BIA for that issue to be exhausted if the BIA summarily affirms the IJ’s decision. Abebe, 432 F.3d at 1040–41. The Abebe majority reasoned that, when the BIA summarily affirms, it is implicitly considering all the issues presented to the IJ and adopting the IJ’s position on those issues. See id. Any issue (broadly defined) that was presented to or addressed by the IJ is thus exhausted, regardless of whether the petitioner raised his specific objection to the IJ’s decision in his appeal to the BIA. But Abebe ignored that there are good reasons for requiring that issues actually be presented to the BIA. First, it encourages good decision making. Usually when the BIA summarily affirms the IJ’s decision it is precisely because no meritorious challenges to the IJ’s decision were presented to the BIA. That is only natural. We don’t expect panels of 22 CHMUKH V. GARLAND our court to sua sponte spot issues that were never presented to us. Indeed, we fault appellants who fail to raise flaws with the lower court’s decision in their opening briefs before us, even where the flaw may be obvious. See United States ex rel. Kelly v. Serco, Inc., 846 F.3d 325, 336 (9th Cir. 2017); Smith v. Marsh, 194 F.3d 1045, 1053 (9th Cir. 1999). The reason is that our adversarial appellate process fundamentally relies on the parties to spot and frame the issues for appeal. See United States v. Sineneng-Smith, 590 U.S. 371, 375–376 (2020); Stevens v. Davis, 25 F.4th 1141, 1169 (9th Cir. 2022). It is hypocritical and counterproductive to hold the agency’s multi-level adjudicative process to a higher standard than our own. Beyond that, Abebe’s rule also imposes obvious inefficiencies on the agency, the parties, and our court. Instead of raising and resolving issues in one trip through the various levels of agency and then court review, petitioners now bounce back and forth between the agency and our court addressing issues piecemeal. This deprives the agency of the opportunity to correct any errors internally, deprives our court of agency expertise and a developed record in these cases, and of course imposes on government attorneys tasked with enforcing our immigration laws the burden of litigating even more unnecessary and time-consuming appeals. There is only one beneficiary of these inefficiencies: the petitioner with non-meritorious claims who gets to spend extra years in the United States unlawfully before his claims are finally resolved. And lastly, as already mentioned, Abebe’s rule encourages a perverse gamesmanship where it is more beneficial for petitioners to trim their strongest critiques of the IJ’s decision out of their appeal to the BIA in the hopes that the BIA will issue a summary affirmance, and then  CHMUKH V. GARLAND 23 hammer their best arguments hard before our court in hopes of securing a remand. After Abebe, a good advocate still needs to think strategically before the agency. But instead of working hard at presenting their best case to the agency, they need to work hard at presenting the nothing-to-see-here case to the agency that will reward them with a summary affirmance of the IJ’s decision. They can then waste this court’s time, and extend the clock for their client, by raising their best issues for the first time before our court, hoping for a remand. Only judges who are really dumb, or are intentionally trying to undermine our immigration system, would deliberately encourage such an inefficient and perverse form of “exhaustion.” Judges on our court are not dumb. II. Abebe is so wrong and counterintuitive that panels of this court regularly decline to follow it, presumably often inadvertently. One notable example is Arsdi v. Holder, which parallels this case. 659 F.3d 925 (9th Cir. 2011). There, the IJ “concluded that … [Arsdi’s] armed robbery was ‘particularly serious,’” and the BIA “Burbano affirmed” and “adopted the IJ’s decision.” Id. at 928. Our court in Arsdi nonetheless found the issue unexhausted because “Arsdi did not ‘put the BIA on notice’ … that he took issue with the IJ’s analysis regarding whether his armed robbery was a ‘particularly serious’ crime.” Id. at 929 (citation omitted). Arsdi cited Abebe, id., but then basically ignored it. Arsdi is the most blatant example, but it’s not the only one. Numerous panels post-Abebe have simply disregarded Abebe and found issues unexhausted because they were not actually raised before the BIA. See, e.g., Samayoa-Martinez 24 CHMUKH V. GARLAND v. Holder, 558 F.3d 897, 899, 902 n.7 (9th Cir. 2009) (noting that the BIA “Burbano affirmed” but finding the panel lacked jurisdiction to consider an issue because it was not raised before the BIA); Jovel v. Holder, 501 F. App’x 708, 708 (9th Cir. 2012) (stating that the BIA “Burbano affirmed,” and citing Abebe, yet concluding that the issue was unexhausted because he did not raise it before the BIA); Gueorguov v. Holder, 415 F. App’x 3, 4 (9th Cir. 2011) (citing Abebe, and noting that the BIA “Burbano affirmed,” but finding that the issue of ineffective assistance of counsel was unexhausted because it was not raised before the BIA); Ouedraogo v. Holder, 391 F. App’x 656, 656 (9th Cir. 2010) (stating that the BIA “Burbano affirmed,” but still finding his claim unexhausted before the BIA); Baharudeen v. Mukasey, 279 F. App’x 580, 581 (9th Cir. 2008) (noting that the BIA adopted and affirmed the IJ’s decision, but still finding that it did not have jurisdiction to review petitioner’s challenge because it was not exhausted before the BIA); Molina v. Mukasey, 270 F. App’x 647, 648 (9th Cir. 2008) (stating that the BIA adopted and affirmed the IJ decision, but finding that it did not have jurisdiction because the issue was not raised before the BIA and was thus unexhausted). Had they applied Abebe, these panels should have found that, because the BIA adopted and affirmed the IJ’s decision, the issues before them were exhausted and therefore reviewable. The fact that panels of our court actually enforce the requirements of exhaustion, as they should, instead of applying our published, en banc Abebe decision, is a testament either to Abebe’s unpopularity or its implausibility. Probably both. Indeed, panels of our court have ignored Abebe’s rule even when they have ultimately found that issues are exhausted. Panels sometimes go through a comprehensive  CHMUKH V. GARLAND 25 exhaustion analysis in cases where the BIA issued a Burbano affirmance. But that should be completely unnecessary under Abebe. In Figueroa v. Mukasey, for example, the panel expressly noted that the BIA issued a Burbano affirmance, but nonetheless proceeded to go through an exhaustion analysis to see whether the petitioner properly raised the issue of extreme hardship before the BIA. 543 F.3d 487, 491–93 (2008). Another panel found issues exhausted in a manner similarly dismissive of Abebe in Ahmed v. Holder, 569 F.3d 1009, 1012 (9th Cir. 2009). There, the panel noted that the BIA had issued a Burbano affirmance, but found the issues that petitioner raised to be exhausted because the BIA had “specifically addressed” them. Id. Again, the panel could have simply relied on Abebe to find the issues exhausted. See also Yan Zhou v. Keisler, 255 F. App’x 133, 136–37 (9th Cir. 2007) (conducting an exhaustion analysis on the issue of credibility despite citing to Abebe and noting that the BIA “Burbano affirmed”). III. Ultimately, I reluctantly join the majority opinion in reaching the merits of this case because I think it correctly finds the issue in this case exhausted under our circuit’s Abebe rule, and I don’t think we can simply ignore our binding precedent—even though, as shown above, that apparently happens with some frequency with Abebe. But Abebe is obviously wrong. It only helps petitioners with non-meritorious claims whose main purpose in asserting them is to stall their removal. The agency, this court, and petitioners with meritorious claims would all benefit from a rule that encourages everything to be presented to the agency in one linear proceeding. We should take an appropriate 26 CHMUKH V. GARLAND case en banc to fix Abebe’s wrongheaded approach to exhaustion. SANCHEZ, Circuit Judge, concurring in part and dissenting in part: It is a bedrock principle of administrative law that an agency acts arbitrarily if it fails to follow its own precedent and does not provide a reasoned explanation for doing so. See Andrzejewski v. F.A.A., 563 F.3d 796, 799 (9th Cir. 2009) (citing Atchison, Topeka & Santa Fe Ry. Co. v. Wichita Bd. of Trade, 412 U.S. 800, 807–08 (1973)). That is precisely what the Board of Immigration Appeals (BIA) did when it determined that Vitaliy Chmukh’s conviction for possession of stolen property was a “particularly serious crime,” making him statutorily ineligible for withholding of removal. See 8 U.S.C. § 1231(b)(3)(B)(ii). BIA precedent requires the agency to assess whether the elements of a statute of conviction potentially bring the crime within the category of a particularly serious crime, and only if they do, to then consider all reliable information about the facts and circumstances of the offense before making a particularly serious crime determination. Here, it is undisputed that the BIA failed to engage in the required elements-only analysis of Chmukh’s state law conviction. Nor did the agency explain why the facts and circumstances of his conviction for possession of a stolen vehicle justify the presumption that he is a danger to the community. While I agree that Chmukh’s conviction qualifies as an aggravated felony, the BIA erred by failing to follow its own precedent or explain why it departed from its own requirements. I would therefore vacate and remand for the agency to  CHMUKH V. GARLAND 27 reevaluate its particularly serious crime determination under the governing agency framework. 1 I. As a preliminary matter, the Government argues that Chmukh failed to exhaust his challenge to the IJ’s particularly serious crime determination because he did not raise it in his appeal to the BIA. The Government is incorrect. Exhaustion is satisfied when the BIA chooses to forego procedural default and expressly address the merits of a claim, as well as when the BIA exercises its discretion to adopt the IJ’s own merits-based analysis of a claim. Chmukh’s challenge to the agency’s particularly serious crime determination was exhausted under both established standards. Exhaustion generally requires a petitioner to have first presented their legal claim in the proper administrative forum below. See Arsdi v. Holder, 659 F.3d 925, 928 (9th Cir. 2011); 8 U.S.C. § 1252(d)(1). When the BIA chooses to ignore a procedural default by the petitioner and instead considers an issue on its merits, “we cannot then decline to consider the issue based upon [the] procedural defect.” Abebe v. Gonzales, 432 F.3d 1037, 1041 (9th Cir. 2005) (en banc). “The BIA is presumably aware of its ability to decline to review an argument when a petitioner has not properly raised the argument on appeal to the BIA,” id., and by addressing the merits of a claim, the BIA has “had a full opportunity to resolve the controversy or correct its own errors before judicial intervention.” Arsdi, 659 F.3d at 930 1 I agree with my colleagues that Chmukh waived any challenge to the agency’s denial of his claim under the Convention Against Torture. 28 CHMUKH V. GARLAND (quoting Figueroa v. Mukasey, 543 F.3d 487, 493 (9th Cir. 2008)). Exhaustion is also satisfied when the BIA adopts the IJ’s decision as its own by citing Matter of Burbano, 20 I. & N. Dec. 872, 874 (B.I.A. 1994), and does not express disagreement with the IJ’s decision. See Abebe, 432 F.3d at 1040 (explaining that a Matter of Burbano citation “signif[ies] that [the BIA] ha[s] conducted an independent review of the record and ha[s] exercised its own discretion in determining that its conclusions were the same as those articulated by the IJ”). Under such circumstances, “if the IJ’s decision was facially premised on an incorrect factual finding or legal conclusion, and the BIA nonetheless adopted the decision, we will also deem the alien to have exhausted the claim.” Arsdi, 659 at 929–30. In short, “we may review any issue addressed on the merits by the BIA, regardless of whether the petitioner raised it before the agency.” Parada v. Sessions, 902 F.3d 901, 914 (9th Cir. 2018). Although Chmukh did not challenge the IJ’s particularly serious crime determination in his appeal to the agency, the BIA expressly adopted the IJ’s own analysis of that issue by citing Matter of Burbano and stating, “[w]e adopt and affirm the decision of the Immigration Judge finding the respondent removable as charged [and] finding him ineligible for asylum and withholding of removal . . . .” The agency proceeded to address the particularly serious crime determination on the merits by stating, “We also agree with the Immigration Judge’s conclusion that the respondent’s conviction of possession of stolen property was a conviction of a particularly serious crime.” The BIA then cited Anaya- Ortiz v. Holder, 594 F.3d 673, 678 (9th Cir. 2010) and Matter of Frentescu, 18 I. & N. Dec. 244, 247 (B.I.A. 1982),  CHMUKH V. GARLAND 29 with parentheticals quoting or summarizing these authorities. By adopting the IJ’s own analysis of the particularly serious crime determination and citing approvingly to circuit and agency precedent in support of that analysis, the BIA addressed the merits of this claim. We therefore have jurisdiction to consider whether the agency erred in its particularly serious crime determination. II. “The BIA acts arbitrarily when it disregards its own precedents and policies without giving a reasonable explanation for doing so.” Israel v. I.N.S., 785 F.2d 738, 740 (9th Cir. 1986); see also Atchison, Topeka & Santa Fe Ry. Co., 412 U.S. at 807–08. This bedrock principle of administrative law is particularly important when the BIA decides whether a given offense amounts to a particularly serious crime. That decision bears serious consequences. A refugee who committed a crime the agency deems “particularly serious” generally must be removed even if there is a greater than fifty percent chance that the refugee would face persecution upon removal. See 8 U.S.C. § 1231(b)(3)(B)(ii). The Immigration and Nationality Act (INA) “reserves such severe consequences for those criminal offenses that make an alien so ‘dangerous to the community of the United States’ that we are not willing to keep him here, notwithstanding the persecution he may face at home.” Delgado v. Holder, 648 F.3d 1095, 1109 (9th Cir. 2011) (en banc) (Reinhardt, J., concurring in part) (alterations adopted). 30 CHMUKH V. GARLAND An aggravated felony is a particularly serious crime per se if it resulted in a “term of imprisonment of at least 5 years.” 8 U.S.C. § 1231(b)(3)(B)(ii). For all other offenses, the Attorney General, or the BIA in its exercise of delegated authority, must decide whether an offense is particularly serious. Id. The BIA, in turn, has fashioned a legal framework for deciding whether a crime is particularly serious. 2 In Matter of Frentescu, 18 I. & N. Dec. 244, the agency developed a multi-factor test for determining on a case-by-case basis whether a crime is particularly serious. The Frentescu factors require consideration of “the nature of the conviction, the circumstances and underlying facts of the conviction, the type of sentence imposed, and, most importantly, whether the type and circumstances of the crime indicate that the alien will be a danger to the community.” Id. at 247; see also Gomez-Sanchez v. Sessions, 892 F.3d 985, 991 (9th Cir. 2018) (noting that the Frentescu steps apply in all cases except when the conviction qualifies per se as particularly serious under 8 U.S.C. § 1231(b)(3)(B)(ii)). The Frentescu analysis proceeds in two steps. At step one, the BIA must determine whether the elements of the crime of conviction “‘potentially bring the crime into a category of particularly serious crimes.’” Mendoza-Garcia v. Garland, 36 F.4th 989, 999 (9th Cir. 2022) (quoting Matter of N-A-M-, 24 I. & N. Dec. 336, 342 (B.I.A. 2007), 2 As the majority notes, we generally lack jurisdiction over final removal orders for noncitizens convicted of aggravated felonies, 8 U.S.C. § 1252(a)(2)(C), but retain jurisdiction to review legal or constitutional claims, id. § 1252(a)(2)(D); see also Park v. Garland, 72 F.4th 965, 973 (9th Cir. 2023). Whether the BIA misapplied its own standard in determining whether a petitioner is ineligible for withholding of removal is a legal question subject to our review. Park, 72 F.4th at 973.  CHMUKH V. GARLAND 31 overruled in part on other grounds by Blandino-Medina v. Holder, 712 F.3d 1338, 1347–48 (9th Cir. 2013)). The step one analysis must be conducted “without regard to the individual facts or circumstances in the case, but only by reviewing the elements of the crime.” Bare v. Barr, 975 F.3d 952, 963 (9th Cir. 2020) (citing Matter of N-A-M-, 24 I. & N. Dec. at 342). As the BIA explains, “[i]f the elements of the offense do not potentially bring the crime into a category of particularly serious crimes, the individual facts and circumstances of the offense are of no consequence, and the alien would not be barred from a grant of withholding of removal.” Matter of N-A-M-, 24 I. & N. Dec. at 342. 3 If the elements of the offense potentially bring the crime within a category of particularly serious crimes, the agency moves on to step two and considers “all reliable information” in analyzing the remaining Frentescu factors. Mendoza-Garcia, 36 F.4th at 999 (citing Matter of N-A-M-, 24 I. & N. Dec. at 342). As the BIA acknowledged below, the “most important[]” part of this analysis is “whether the type and circumstances of the crime indicate that the [noncitizen] will be a danger to the community.” (quoting Matter of Frentescu, 18 I. & N. Dec. at 247). We, too, have explained that a “crime is particularly serious if the nature of the conviction, the underlying facts and circumstances[,] and the sentence imposed justify the presumption that the convicted immigrant is a danger to the 3 The BIA has repeatedly reaffirmed the importance of this threshold determination, permitting immigration judges to consider the facts of a noncitizen’s offense only after first finding that the elements of the offense potentially bring it within the ambit of a particularly serious crime. See, e.g., Matter of T-C-A-, 28 I. & N. Dec. 472, 480 (B.I.A. 2022); Matter of D-L-S-, 28 I. & N. Dec. 568, 577 (B.I.A. 2022). 32 CHMUKH V. GARLAND community.” Alphonsus v. Holder, 705 F.3d 1031, 1041 (9th Cir. 2013) (quoting Delgado, 648 F.3d at 1107). “Thus, dangerousness remains the ‘essential key’ to determining whether the individual’s conviction was for a particularly serious crime.” Gomez-Sanchez, 892 F.3d at 991; see also Matter of B-Z-R-, 28 I. & N. Dec. 563, 563 (Att’y Gen. 2022) (same). The agency erred at both steps of the particularly serious crime determination. III. Vitaliy Chmukh and a friend used a stolen car to steal packages off porches. He was convicted of possession of a stolen vehicle under Washington law, and the agency was required to determine whether that conviction qualified as a “particularly serious crime,” rendering him ineligible for withholding of removal. At step one, however, the IJ and BIA did not analyze the elements of Chmukh’s state law conviction to determine if those elements potentially bring the crime into a category of particularly serious crimes. In the IJ’s oral decision, the IJ first explained that her analysis is “guided by” Matter of Frentescu and Matter of N-A-M-. She noted that Chmukh “was convicted for possession of stolen property under RCW 9A.56.068,” and recited the elements of his conviction: “The elements again, that the property was stolen, the defendant was in possession and finally, knowing it was stolen.” She then stated: “The Court finds the elements and that the offense is a felony deemed by the state and that the respondent was given a lengthy sentence.” The IJ went on to discuss the length of Chmukh’s sentence and the individual facts and circumstances of his offense. Her analysis omitted any discussion of the nature of Chmukh’s state law conviction. Mere recitation of the elements does  CHMUKH V. GARLAND 33 not explain why those elements potentially bring the offense within the ambit of a particularly serious crime. On appeal, the BIA adopted the IJ’s analysis, adding that it “agree[d] with the Immigration Judge’s conclusion that the respondent’s conviction of possession of stolen property was a conviction of a particularly serious crime.” Neither the IJ nor the BIA undertook the threshold elements-only inquiry required by governing BIA precedent. See Matter of N-A-M-, 24 I. & N. Dec. at 342. The agency’s later discussion of the circumstances of the crime and the length of sentence imposed did not save it from its error, for “[i]f the elements of the offense do not potentially bring the crime into a category of particularly serious crimes, the individual facts and circumstances of the offense are of no consequence.” Id. The BIA’s failure to apply the correct legal standard in assessing whether Chmukh’s offense was a “particularly serious crime” constitutes legal error, requiring remand. See Mendoza-Garcia, 36 F.4th at 999 (concluding the “BIA committed an error of law . . . in failing to apply the correct legal standards in assessing whether Petitioner’s offense was a ‘particularly serious crime,’” requiring remand). The BIA compounded its error by failing to explain how the Frentescu factors justify the presumption that Chmukh is a danger to the community. In discussing Chmukh’s conviction for possessing a stolen vehicle, the IJ recited the following facts: (1) Chmukh knew the vehicle was stolen; (2) he and a friend used it to steal packages off of porches (specifically, “mail and packages belonging to approximately 16 different victims”); and (3) he received a “lengthy” sentence of 43 months. But as Chmukh notes, the IJ never explained why these facts justify the presumption that he is a danger to the community. On appeal, the BIA 34 CHMUKH V. GARLAND quoted from Matter of Frentescu, including the importance of assessing “whether the type and circumstances of the crime indicate that the [non-citizen] will be a danger to the community,” see 18 I. & N. Dec. at 247, but never provided any reasoned explanation why these circumstances indicate Chmukh is a danger to the community—the touchstone of the particularly serious crime analysis. See Gomez-Sanchez, 892 F.3d at 991; Alphonsus, 705 F.3d at 1047–48. This was error. The agency’s analysis provides no basis to determine why possession of a stolen vehicle for mail theft—an offense that is obviously irresponsible but nevertheless involved no violence, weapons, or physical harm—is sufficiently dangerous to justify the particularly serious crime bar. Possessing a stolen vehicle to steal mail from a porch is not the kind of crime against persons (such as the use of force, violence, or threats), or drug trafficking, or possession of child pornography that the BIA has found to be particularly serious. See Alphonsus, 705 F.3d 1031 at 1047–48 (“A review of the Board’s precedential decisions . . . demonstrates that the ‘particularly serious crime’ designation has generally been reserved for more grave offenses than the conduct at issue here.”). The BIA has traditionally found crimes against property less likely to qualify as particularly serious. In Frentescu, for example, the agency found that burglary with intent to commit theft was not particularly serious in part because it “was an offense against property, rather than against a person.” 18 I. & N. Dec. at 247. That is not to suggest that crimes against property can never qualify as particularly serious crimes. They can. See, e.g., Arbid v. Holder, 700 F.3d 379, 385 (9th Cir. 2012) (holding that a “complex scheme” to defraud victims of nearly $2 million was particularly serious). But the agency’s  CHMUKH V. GARLAND 35 analysis must “explain why [a non-citizen] falls on the wrong side of the line—if indeed he does” and be “consistent with the [INA’s] statutory text, which indicates that the line must be drawn so that ‘particularly serious crimes’ are not a major proportion of crimes generally.” Alphonsus, 705 F.3d at 1048 (cleaned up). The agency’s analysis of Chmukh’s possession of a stolen vehicle conviction did not meet that standard. IV. The majority does not dispute that the agency failed to conduct an elements-only analysis at step one, but my colleagues conclude that the agency was excused from doing so under Bare, 975 F.3d at 962–63. In Bare, we upheld the agency’s determination that a petitioner’s conviction for felon in possession of a firearm was particularly serious even though the agency failed to assess whether the elements of the offense brought it within the ambit of a particularly serious crime. See id. We reasoned that because the petitioner was convicted of a common federal offense with simple elements, we would not “put form over substance” by remanding to the agency for an explicit consideration of the elements of the offense. Id. at 963. Bare’s analysis excusing the agency’s failure to apply the threshold elements-only test is out of step with several of our sister circuits. In cases where the BIA failed to examine whether the elements of a crime of conviction potentially bring the offense within the category of particularly serious crimes, the Second, Third, and Fourth Circuits have uniformly held that the agency erred by failing to apply its own precedent. See Ojo v. Garland, 25 F.4th 152, 165 (2d Cir. 2022) (“We conclude that the agency failed to apply the correct legal standard at step one of the analysis in its 36 CHMUKH V. GARLAND determination that Ojo’s conviction involved a particularly serious crime,” necessitating remand); Luziga v. Att’y Gen., 937 F.3d 244, 253–54 (3d Cir. 2019) (“[T]he IJ and BIA failed to correctly apply the analysis articulated in [Matter of] N-A-M-, skipping right over the preliminary consideration of elements,” and the “BIA’s failure to correctly apply its own precedent” required remand); Annor v. Garland, 95 F.4th 820, 827 (4th Cir. 2024) (vacating and remanding BIA decision where the agency failed to properly evaluate the elements of the offense at the first step of the Matter of N-A-M- inquiry). But see Lafortune v. Garland, 110 F.4th 426, 435–36 (1st Cir. 2024) (concluding that Matter of N-A-M- does not require an explicit discussion of the elements of the offense of conviction at step one). In any event, Bare did not purport to give the BIA blanket permission to overlook the elements-only analysis in its particularly serious crime determination. As Bare observed, the step one inquiry serves an important gatekeeping function, and under Matter of N-A-M-, if the elements of the offense do not potentially bring the crime into a particularly serious category, the individual facts and circumstances of the offense “are of no consequence.” 975 F.3d at 961–62. The elements of the offense and the circumstances in Bare are also markedly different. The petitioner there was convicted for being a felon in possession of a firearm, and his criminal past included incidents in which he boarded a school bus of middle-school students and threatened them with violence; he pointed a rifle at a teenager and threatened to kill him for allegedly stealing a television set and was part of a group that beat the teenager with a baseball bat; and he pointed a pistol at an unarmed neighbor and fired the gun into the air. Id. at 959. Because these facts referenced all the elements of the offense and  CHMUKH V. GARLAND 37 made clear why the agency found him sufficiently dangerous to justify the particularly serious crime bar, we determined that remand was unnecessary. Id. at 963. The same cannot be said here. It is not obvious why Chmukh’s use of a stolen vehicle to steal packages from porches poses a sufficient danger to the community. Together with the agency’s failure to examine the elements of the offense, these twin errors of law require remand. See Annor, 95 F.4th at 829 (concluding that the BIA erred at steps one and two of the agency’s framework when it failed to consider the elements of the offense and whether the petitioner’s conviction indicates that he poses a danger to the community). The majority acknowledges that the agency did not expressly consider dangerousness in its analysis at step two, but it concludes that omitting this analysis was not legal error. The majority notes that under Gomez-Sanchez, “there is no statutory requirement for a separate determination of dangerousness focusing on the likelihood of future serious misconduct on the part of the alien.” 892 F.3d at 991 (citation omitted). While that is true, Gomez-Sanchez also explained that “dangerousness remains the ‘essential key’ to determining whether the individual’s conviction was for a particularly serious crime.” Id. (emphasis added). Thus, while the BIA need not make a separate finding of future dangerousness, “[i]t must be determined that an applicant for relief constitutes a danger to the community of the United States to come within the purview of the particularly serious 38 CHMUKH V. GARLAND crime bar.” Delgado, 648 F.3d at 1107 (quoting Matter of Carballe, 19 I. & N. Dec. 357, 360 (B.I.A. 1986)). 4 In short, circuit and agency precedent has never held that the BIA may forego the threshold analysis of determining and explaining why the nature and circumstances of a given offense indicate that the petitioner poses a danger to the community. On the contrary, that is the essence of the particularly serious crime analysis. The BIA therefore violated its own precedent. Twice. It failed to assess whether the elements of Chmukh’s possession of a stolen vehicle conviction potentially bring it within the category of particularly serious crimes. Then it failed to explain why the nature and circumstances of Chmukh’s conviction justify the presumption that he is a danger to the community. In so doing, the BIA acted arbitrarily by failing to apply its own precedent. 4 Konou v. Holder, 750 F.3d 1120 (9th Cir. 2014), does not suggest otherwise. There, we reiterated that “whether the type and circumstances of the crime indicate that the alien will be a danger to the community” is the “most important[]” part of the analysis. Id. at 1127 (quoting Matter of Frentescu, 18 I. & N. Dec. at 247). And because “[c]rimes against persons are more likely to be categorized as ‘particularly serious crimes,’” we concluded that the “BIA did not abuse its discretion in determining that Konou’s assault-and-battery convictions were particularly serious crimes.” Id. (citation omitted).