Uribe Andrade v. Garland

FOR PUBLICATION UNITED STATES COURT OF APPEALS FOR THE NINTH CIRCUIT MIGUEL ANGEL URIBE No. 21-1244 ANDRADE, Agency No. A200-739-003 Petitioner, v. OPINION MERRICK B. GARLAND, Attorney General, Respondent. On Petition for Review of an Order of the Board of Immigration Appeals Argued and Submitted October 20, 2023 Phoenix, Arizona Filed March 1, 2024 Before: Sandra S. Ikuta, Bridget S. Bade, and Daniel A. Bress, Circuit Judges. Opinion by Judge Bress 2 URIBE V. GARLAND SUMMARY* Immigration The panel denied Miguel Angel Uribe Andrade’s petition for review of the Board of Immigration Appeals’ dismissal of an appeal of the denial of asylum and related relief, concluding that the Board did not commit legal error in finding that Uribe’s proposed social group lacked particularity, and that substantial evidence supported the denial of CAT relief. Although Uribe was removable based on a criminal offense covered by 8 U.S.C. § 1252(a)(2)(C), the panel concluded that this provision did not deprive it of jurisdiction to review Uribe’s asylum and withholding claims, because Uribe did not challenge the factual findings underlying the agency’s rejection of his proposed social group, and pursuant to 8 U.S.C. § 1252(a)(2)(D), the court retained jurisdiction to review Uribe’s legal arguments concerning that determination. The panel concluded that the Board did not err in determining that Uribe’s proposed social group, “Mexicans with mental health disorders characterized by psychotic features who exhibit erratic behavior,” was not cognizable because it lacked particularity. The panel noted that Uribe did not challenge the agency’s factual finding that the term “erratic behavior” was not defined in the record and was not used by Uribe’s treatment providers in describing his conditions or symptoms. Additionally, nothing inherent in * This summary constitutes no part of the opinion of the court. It has been prepared by court staff for the convenience of the reader. URIBE V. GARLAND 3 the phrase “erratic behavior” required the Board to treat it as a self-defining term, much less to supply a lay definition. The Board could therefore conclude that “erratic behavior” did not provide firm enough indication of who might be in the proposed group given that the phrase may cover a range of conduct that varies in frequency, duration, and character. The panel held that substantial evidence supported the Board’s denial of CAT protection. As to Uribe’s first theory—that he will be tortured in a Mexican mental health facility—the panel concluded that the record did not compel the conclusion that Uribe will be unable to obtain his medication or other treatment in Mexico. Thus, the Board reasonably concluded that Uribe was not more likely than not to be committed to a mental health institution. Substantial evidence also supported the Board’s determination that, even if Uribe were committed to a mental health institution, he would not likely be tortured because the poor conditions in Mexico’s mental health facilities are not created with the requisite specific intent to inflict suffering. As to Uribe’s fear of being tortured on account of his former gang membership and tattoos, the panel concluded that evidence of widespread cartel violence in Mexico did not show that Uribe’s past or perceived gang affiliation would make him particularly vulnerable to such violence. Nor did the record compel the conclusion that Mexican officials would acquiesce to Uribe’s torture. 4 URIBE V. GARLAND COUNSEL Kristin Macleod-Ball (argued) and Trina A. Realmuto, National Immigration Litigation Alliance, Brookline, Massachusetts, for Petitioner. Jaclyn G. Hagner (argued) and Allison Frayer, Trial Attorneys; Sarah A. Byrd, Senior Litigation Counsel; Jennifer Levings, Assistant Director; Brian M. Boynton, Principal Deputy Assistant Attorney General; United States Department of Justice, Civil Division, Office of Immigration Litigation, Washington, D.C.; for Respondent. OPINION BRESS, Circuit Judge: This is a petition for review of a Board of Immigration Appeals (BIA) decision dismissing an appeal of an Immigration Judge (IJ) order denying the petitioner’s applications for asylum, withholding of removal, and protection under the Convention Against Torture (CAT). We hold that the BIA did not commit legal error in finding that the petitioner’s proposed social group—“Mexicans with mental health disorders characterized by psychotic features who exhibit erratic behavior”—lacked particularity, and so could not be a basis for granting asylum or withholding of removal. We also hold that substantial evidence supports the denial of CAT relief. We therefore deny the petition for review. URIBE V. GARLAND 5 I Petitioner Miguel Angel Uribe Andrade (Uribe) is a native and citizen of Mexico who entered the United States with family members in 1999, at the age of nine. In 2005, the government placed Uribe and his mother in deferred action status, meaning it chose to give their cases lower priority for removal. As an adolescent growing up in California, Uribe joined the “Southsider” gang and began using drugs. He had behavioral problems and spent time in juvenile detention for robbery and skipping class. After he was released, he remained affiliated with the gang until he moved to Oregon in 2009. Shortly after moving to Oregon, Uribe, then age nineteen, was convicted of felony assault and other offenses following a fight with his girlfriend. Uribe served nine months in prison before he was transferred to immigration detention and placed in removal proceedings. In April 2012, he was granted lawful permanent resident status and was released. From 2013 to 2020, Uribe was convicted of various offenses in Oregon, including methamphetamine possession. In August 2020, the Department of Homeland Security (DHS) served Uribe with a Notice to Appear, charging him with removability under 8 U.S.C. § 1227(a)(2)(B)(i) as an alien convicted of an offense related to a controlled substance. In support of this charge, DHS cited Uribe’s Oregon methamphetamine convictions. At a hearing held pursuant to Franco-Gonzalez v. Holder, 2014 WL 5475097 (C.D. Cal. Oct. 29, 2014), the IJ determined that Uribe was not competent to represent himself and appointed counsel for him in the immigration proceedings. Through counsel, Uribe conceded 6 URIBE V. GARLAND removability and, as relevant here, applied for asylum, withholding of removal, and CAT protection. Uribe’s asylum and withholding of removal claims were premised on his fear of persecution in Mexico as a member of a proposed particular social group, “Mexicans with mental health disorders characterized by psychotic features who exhibit erratic behavior.”1 In support of these claims, Uribe submitted a report by a mental health evaluator, Dr. Kathryn Lanthorn-Cardenas, who diagnosed Uribe with “Major Depression, Moderate, with Psychotic Features,” in addition to stimulant and opiate use disorders in sustained remission in a controlled environment. Dr. Lanthorn- Cardenas described Uribe’s prognosis as “cautiously optimistic” with “appropriate treatment” but “poor” without “appropriate mental health and substance use treatment.” Before the IJ, Uribe testified about his mental health history. He described first experiencing depression, anxiety, and psychotic hallucinations as a teenager. He testified that his drug use at the time “mentally broke [him] down,” causing him to “hear voices, and see things that weren’t there.” Although Uribe initially attributed his symptoms in part to his drug use, his symptoms persisted after he became sober in 2019. Uribe received limited treatment for these symptoms and had never been institutionalized in a psychiatric hospital. Uribe received mental health treatment, including medication and counseling, while incarcerated. He continued to receive treatment while in immigration detention. 1 Before the agency, Uribe also raised an alternative proposed social group, “Mexicans with disabilities.” He abandoned this alternative on appeal. URIBE V. GARLAND 7 Relying on this medical history, Uribe sought asylum and withholding of removal on the theory that he would be at risk of persecution because he would be unable to obtain in Mexico the “intensive mental health and substance abuse treatment” that he believes he requires. Without this treatment, Uribe feared he would “fall back into using drugs” and that his behavior would be “impaired, making him noticeable to others.” These circumstances, he argued, made it likely that he would be involuntarily institutionalized in Mexico, where he would face “appalling, abusive conditions.” Uribe’s CAT application was premised on his fear of torture by health care providers in Mexico’s state-run institutions, and by cartel members or Mexican officials based on his previous gang affiliation. As to the first theory, Uribe testified that he did not know where he would seek addiction and mental health treatment in Mexico and that abstaining from treatment would cause him to return to using drugs. He testified that he had an uncle and a grandmother still living in Mexico, but that he could not remain sober if he lived with them because his uncle “uses drugs and is part of the cartel.” Uribe’s counsel pointed to Uribe’s historically “poor prognosis without treatment” and lack of familial support in Mexico (most of his family members are U.S. citizens or lawful permanent residents) as evidence that he would likely be arrested and involuntarily hospitalized if returned to Mexico. As to the second basis for CAT protection, Uribe testified that he would be targeted in Mexico because his tattoos, including one on his face, identified him as a member of the Southsider gang. The IJ denied relief. Uribe’s asylum and withholding claims failed because his proposed social group— “Mexicans with mental health disorders characterized by 8 URIBE V. GARLAND psychotic features who exhibit erratic behavior”—was not legally cognizable. The first problem, the IJ explained, was that Uribe had not established that Mexican society perceived his proposed group as “socially distinct.” The IJ also found the group “insufficiently particular” because the terms “psychotic features,” “exhibit,” and “erratic behavior” were “broad and ambiguous, particularly where those terms are not defined by the evidence in the record.” The IJ also denied Uribe’s application for CAT protection because Uribe had not shown it was “more likely than not” that he would be tortured if removed to Mexico. The IJ recognized that Uribe’s “mental health condition may decline if he is returned to Mexico” and that some Mexican mental health institutions had “extremely poor” conditions. But the IJ rejected Uribe’s argument about the possibility of torture in such a facility because Uribe had not established that he would be institutionalized or that the Mexican government created the poor conditions to torture individuals with mental illness. Instead, evidence suggested that existing deficiencies stemmed from budget constraints rather than a specific intent to torture. The IJ also rejected Uribe’s argument that he would be tortured by cartel members or Mexican officials based on his gang-related tattoos because Uribe “ha[d] not received any particularized threat from any criminal organization in Mexico, and the evidence [did] not establish that the Mexican government would acquiesce to [Uribe’s] mistreatment . . . .” The BIA dismissed Uribe’s appeal. It affirmed the IJ’s denial of asylum and withholding of removal because, even assuming that Uribe’s proposed social group was socially distinct in Mexico, it was not defined with sufficient particularity. Noting that “erratic behavior” was “not defined in the record, clinically or otherwise” and “was not URIBE V. GARLAND 9 used by [Uribe’s] treatment providers in describing [Uribe’s] conditions or symptoms,” the BIA reasoned that there was “no discernible basis for readily identifying members of [the] proposed group.” The BIA also affirmed the denial of CAT protection. Because the record did not indicate that Uribe would lack access to his prescribed medication in Mexico, Uribe failed to show that he would more likely than not be unmedicated and involuntarily hospitalized. The BIA also found that the evidence plausibly established that the poor conditions in Mexico’s mental health facilities “were the result of limited financial resources, rather than a specific intent to torture.” As to Uribe’s argument that he would be tortured by cartel members or government officials due to his past gang affiliation, the BIA agreed with the IJ that Uribe had not presented evidence of any particularized threat as to him. Uribe timely petitioned for review. We have jurisdiction under 8 U.S.C. § 1252. II We begin with Uribe’s claims for asylum and withholding of removal. A Uribe concedes that he is removable for his past drug crimes under 8 U.S.C. § 1227(a)(2)(B)(i). Although 8 U.S.C. § 1252(a)(2)(C) deprives us of jurisdiction “to review any final order of removal against an alien who is removable by reason of having committed a [covered] criminal offense,” and Uribe’s methamphetamine convictions so qualify, see id. §§ 1227(a)(2)(B)(i), 1252(a)(2)(C), we retain jurisdiction to review “constitutional claims or questions of law.” Id. § 1252(a)(2)(D). Uribe disclaims any factual 10 URIBE V. GARLAND challenges to the BIA’s decision, and we agree that his arguments present questions of law. See Diaz-Reynoso v. Barr, 968 F.3d 1070, 1076 (9th Cir. 2020) (explaining that whether a “particular social group is cognizable is a question of law”). We therefore consider whether the BIA legally erred in denying Uribe asylum and withholding of removal, reviewing questions of law de novo. Hernandez-Mancilla v. Holder, 633 F.3d 1182, 1184 (9th Cir. 2011). To be eligible for asylum, an alien must demonstrate that he is a “refugee,” 8 U.S.C. § 1158(b)(1)(B)(i), defined as “any person who is outside any country of such person’s nationality . . . and who is unable or unwilling to return to . . . that country because of persecution or a well-founded fear of persecution on account of race, religion, nationality, membership in a particular social group, or political opinion . . . .” Id. § 1101(a)(42)(A). The asylum applicant must show that he faces a “likelihood of ‘persecution or a well-founded fear of persecution’” on account of one of these five enumerated bases. Sharma v. Garland, 9 F.4th 1052, 1059 (9th Cir. 2021) (quoting 8 U.S.C. § 1101(a)(42)(A)); see also 8 U.S.C. § 1158(b)(1)(B)(i). To establish eligibility for withholding of removal, an alien must show “that it is more likely than not” that he will be persecuted on these grounds. Barajas-Romero v. Lynch, 846 F.3d 351, 357, 360 (9th Cir. 2017); see also 8 U.S.C. § 1231(b)(3)(A). Uribe claims that he is eligible for asylum because he has a “well-founded fear of persecution” on account of his “membership in a particular social group.” 8 U.S.C. § 1101(a)(42)(A). His argument in support of withholding of removal is the same. The statute does not define the phrase “membership in a particular social group,” but we have adhered to the BIA’s determination that an applicant URIBE V. GARLAND 11 “must establish that the group is: ‘(1) composed of members who share a common immutable characteristic, (2) defined with particularity, and (3) socially distinct within the society in question.’” Diaz-Reynoso, 968 F.3d at 1077 (quoting Matter of M-E-V-G-, 26 I. & N. Dec. 227, 237 (B.I.A. 2014)); see also Villegas Sanchez v. Garland, 990 F.3d 1173, 1180 (9th Cir. 2021); Akosung v. Barr, 970 F.3d 1095, 1103 (9th Cir. 2020). Uribe bears the burden of proving these elements. See Nguyen v. Barr, 983 F.3d 1099, 1104 (9th Cir. 2020). Here we consider whether Uribe’s proposed social group satisfies the second requirement, particularity. Particularity requires that a proposed social group “be ‘discrete’ and have ‘definable boundaries.’” Acevedo Granados v. Garland, 992 F.3d 755, 762 (9th Cir. 2021) (quoting Matter of M-E- V-G-, 26 I. & N. Dec. at 239). And the group may not be “amorphous, overbroad, diffuse, or subjective.” Matter of M-E-V-G-, 26 I. & N. Dec. at 239. The ultimate question when assessing particularity is whether the proposed social group is defined by “characteristics that ‘provide a clear benchmark for determining who falls within the group.’” Nguyen, 983 F.3d at 1103 (quoting Matter of W-G-R-, 26 I. & N. Dec. 208, 214 (B.I.A. 2014)). B In his petition for review, Uribe claims membership in the proposed particular social group, “Mexicans with mental health disorders characterized by psychotic features who exhibit erratic behavior.” The BIA found that this group was insufficiently particular because it is “amorphous and subjective,” specifically because of the reference to “erratic behavior.” Although Uribe argued that inclusion of the phrase “erratic behavior” narrows the proposed group, the 12 URIBE V. GARLAND BIA explained that “the term ‘erratic behavior’ is not defined in the record, clinically, or otherwise, and . . . was not used by [Uribe’s] treatment providers in describing [Uribe’s] conditions or symptoms.” Citing the lack of any record- based definition, the BIA affirmed the IJ’s conclusion that the proposed social group lacked particularity because “there is no discernible basis for readily identifying members of this proposed group.” Uribe, as we have noted, does not challenge the factual findings underlying the IJ and BIA’s rejection of his proposed particular social group, namely, that the term “erratic behavior” was not defined in the record and was not used by Uribe’s treatment providers in describing his conditions or symptoms. As a general matter, the BIA may conclude that a proposed social group is insufficiently particular when the group definition involves terminology that is undefined or unexplained in the record, particularly terminology that purports to be medical in nature, as here. Cf. Pirir-Boc v. Holder, 750 F.3d 1077, 1084 (9th Cir. 2014) (describing the particularity inquiry as “a case-by-case determination”); Diaz-Reynoso, 968 F.3d at 1088 (similar). In this case, nothing inherent in the phrase “erratic behavior” required the BIA to treat it as a self-defining term, much less to supply a lay definition. See Nguyen, 983 F.3d at 1103 (holding that the BIA did not err in concluding that “known drug users” lacked particularity when the petitioner “d[id] not provide any evidence” to demarcate the group’s boundaries within the relevant society). The BIA could therefore conclude that “erratic behavior” did not provide firm enough indication of who might be in the proposed group given that the phrase may cover a range of conduct that varies in frequency, duration, and character. See id. (explaining that drug “user” was insufficiently particular URIBE V. GARLAND 13 because it “could vary broadly based on the amount and frequency of an individual’s drug use” and “could encompass first-time users, occasional users, habitual users, or rehabilitated individuals”). Indeed, Uribe acknowledges in his opening brief that his proposed group “included a modifier that does not necessarily have an established medical definition.” In our experience, petitioners sometimes include additional attributes in their proposed social groups as a basis for establishing the social distinction of their group. But when petitioners do so, the BIA may fairly conclude that the group lacks particularity in the absence of evidence that gives sufficient meaning to the terms describing the additional attributes. Uribe nonetheless argues that the BIA committed legal error because, in his view, the BIA’s rejection of his proposed particular social group conflicts with Acevedo Granados v. Garland, 992 F.3d 755 (9th Cir. 2021), a precedent that the BIA here recognized and cited in its decision. In Acevedo Granados, the agency rejected the proposed group of “El Salvadoran men with intellectual disabilities who exhibit erratic behavior” because the “imprecise contours of the terms ‘intellectual disability’ and ‘erratic behavior’ render[ed] the proposed group ‘amorphous, overbroad, diffuse, and subjective.’” Id. at 761–62 (brackets and citation omitted). We granted the petition for review and remanded for further proceedings. Id. at 758. Although “erratic behavior” was included in the proposed social group in Acevedo Granados, we did not address the import of that term. Rather, the problem we identified was that “[t]he BIA and IJ treated the term ‘intellectual disability’ as if it were applied by a layperson.” Id. Psychologists had in fact “diagnosed [the petitioner] with 14 URIBE V. GARLAND an Intellectual Disability, as defined in the Diagnostic and Statistical Manual of Medical Disorders (‘DSM-5’).” Id. at 759–60. But instead of hewing to the diagnosis in the record and the clinical definitions associated with it, the BIA “assumed that a determination of mental illness was a subjective one, to be carried out by a judge.” Id. at 762. The IJ even went so far as to conclude that the petitioner did not suffer from mental illness based on the IJ’s lay perception of the petitioner’s seemingly normal behavior in court. Id. This was erroneous, we held, because “[t]he record in th[at] case contained evaluations conducted by recognized psychologists, retained by the government, who reported their findings in professional terms.” Id. at 762–63. We acknowledged that “the term ‘intellectual disability’ can be used, especially by laypersons, in a way that lacks precision.” Id. at 762. But because the petitioner “was diagnosed with ‘intellectual disability’ as that term is used within the psychological profession,” id., the agency could not “disregard . . . the evidence in the record” in favor of its own subjective determinations that the term lacked particularity. Id. at 763; see also id. at 762 (“[T]he particularity standard does not expect that immigration judges make independent diagnoses based on their observations in the courtroom.”). Far from demonstrating legal error in the agency decision we consider here, Acevedo Granados if anything underscores that the IJ and BIA acted appropriately in finding that Uribe had not met the particularity requirement. The agency here did not ignore clinical evidence and diagnoses in favor of a lay judgment. Instead, the problem was a lack of supporting evidence in the first place: the term “erratic behavior” was “not defined in the record, clinically or otherwise,” and “was not used by [Uribe’s] treatment URIBE V. GARLAND 15 providers in describing [Uribe’s] conditions or symptoms.” This case is therefore unlike Acevedo Granados, where there was both a clinical definition and a documented diagnosis of “intellectual disability,” which the agency disregarded in favor of a lay judgment, and where we did not consider the meaning of the phrase “erratic behavior.” 992 F.3d at 762– 63. Acevedo Granados confirms that absent appropriate record-based evidence, terms associated with mental health can otherwise “lack[] precision.” Id. at 762. That is the problem the BIA permissibly identified here. Uribe points out that the proposed group in Acevedo Granados was defined similarly to his own, as “El Salvadoran men with intellectual disabilities who exhibit erratic behavior.” Id. at 760 (emphasis added). From this he argues that Acevedo Granados already held that a proposed group that incorporates the phrase “erratic behavior” is cognizable as a matter of law. That is not correct. As we noted above, Acevedo Granados did not address the import of the phrase “erratic behavior.” It thus cannot be read to sign off on the particularity of all proposed social groups containing that phrase, much less without regard to the record before the agency. See, e.g., United States v. Corrales-Vazquez, 931 F.3d 944, 954 (9th Cir. 2019) (“Cases are not precedential for propositions not considered, or for questions which merely lurk in the record.” (citations, alterations, and quotation marks omitted)). And, more broadly, contrary to Uribe’s position, Acevedo Granados did not hold that the proposed group there was, in fact, legally cognizable. We instead held that “the agency misunderstood [the petitioner’s] proposed social group,” and that record evidence that the agency ignored “may satisfy” the requirements for a particular social group. 992 F.3d at 758 16 URIBE V. GARLAND (emphasis added). In remanding “for further fact-finding on an open record,” id. at 765, we did not hold that proposed social groups involving “erratic behavior” (or any specific terms) necessarily meet all the legal requirements for a particular social group. Nor could such a holding be reconciled with Acevedo Granados’s emphasis on the central relevance of clinically based record evidence for defining medical and psychological terms. See id. at 762– 63. In arguing that the BIA committed legal error, Uribe next invokes the Fourth Circuit’s decision in Temu v. Holder, 740 F.3d 887 (4th Cir. 2014). But Temu is distinguishable. In Temu, the court in a divided decision held that the BIA erred in finding insufficiently particular the proposed group of Tanzanian “individuals with bipolar disorder who exhibit erratic behavior.” Id. at 891. Although the court in Temu “doubt[ed] that ‘individuals who exhibit erratic behavior’ would qualify as a particular social group,” “[u]nlike ‘erratic behavior,’ the term bipolar disorder has well-defined, identifiable characteristics.” Id. at 895 (citing clinical sources). In the Fourth Circuit’s view, the proposed social group was cognizable when considered “as a whole.” Id. at 894; see also id. at 895 (“Thus, erratic behavior has unclear boundaries that the other component of Mr. Temu’s group supplies. In turn, bipolar disorder covers a broad spectrum of behavior that is sharply limited by the requirement of erratic behavior.”). The Fourth Circuit’s analysis in Temu is inapplicable here for several reasons. Uribe’s proposed group does not use the term “bipolar disorder,” but the broader phrase “mental health disorders characterized by psychotic features.” And in Temu, there was record evidence, including expert testimony, that “visibly erratic behavior” is URIBE V. GARLAND 17 seen as a manifestation of demonic possession in Tanzania; as the Fourth Circuit noted, “Tanzanians even have a label for the group . . . .” Id. at 890. There is no equivalent record evidence here. Finally, to the extent that Temu relied on a lay understanding of “erratic behavior” in connection with mental illness, that would appear contrary to our observation in Acevedo Granados that mental health terms “can be used, especially by laypersons, in a way that lacks precision.” 992 F.3d at 762; see also Temu, 740 F.3d at 901 (Agee, J., dissenting) (explaining that “‘erratic behavior’ is inherently subjective and amorphous”). Although Temu indicated that under the specific circumstances in Tanzania the phrase “erratic behavior” limited “bipolar disorder” and made that term more particular, id. at 895–96, there is no basis here for concluding that the phrase “erratic behavior” limits the broader phrase in our case, “mental health disorders characterized by psychotic features.” Furthermore, as we explained above, nothing about the term “erratic behavior” required the BIA to accord it any particular definition absent clarifying record evidence. Nor did the BIA commit legal error in considering Uribe’s proposed social group “piecemeal,” as Uribe contends. The BIA did not hold that the reference to persons “who exhibit erratic behavior,” standing alone, had to meet the requirements of a particular social group. And the BIA did not conceive of “erratic behavior” as an isolated concept disconnected from mental illness. Instead, the BIA evaluated Uribe’s proposed group as a whole when it discussed “erratic behavior,” as that phrase was integral to the group as Uribe conceived it. See Diaz-Reynoso, 968 F.3d at 1084 (“We agree that courts cannot rewrite proposed social groups . . . .”). Uribe does not argue his proposed group is cognizable without reference to “erratic behavior.” 18 URIBE V. GARLAND Although Uribe appears to have intended for “erratic behavior” to provide a limiting principle, that does not mean the BIA was required to conclude that the group definition in toto “provide[d] a clear benchmark for determining who falls within the group . . . .” Nguyen, 983 F.3d at 1103 (quotation marks and citation omitted). A component of a proposed social group can narrow the group while at the same time failing to provide clear boundaries for ascertaining who is a member of the group. That was the case here based on the BIA’s unchallenged finding that “the term ‘erratic behavior’ is not defined in the record, clinically, or otherwise, and . . . was not used by [Uribe’s] treatment providers in describing [Uribe’s] conditions or symptoms.” For these reasons, we discern no legal error in the BIA’s denial of asylum and withholding of removal. III We next consider Uribe’s CAT claim. “The Convention Against Torture provides mandatory relief for any immigrant who can demonstrate that ‘it is more likely than not that he or she would be tortured if removed to the proposed country of removal.’” Gutierrez-Alm v. Garland, 62 F.4th 1186, 1200–01 (9th Cir. 2023) (quoting Hamoui v. Ashcroft, 389 F.3d 821, 826 (9th Cir. 2004)); see also 8 C.F.R. § 1208.16(c)(2). “To constitute torture, an act must inflict ‘severe pain or suffering,’ and it must be undertaken ‘at the instigation of, or with the consent or acquiescence of, a public official.’” Hernandez v. Garland, 52 F.4th 757, 769 (9th Cir. 2022) (quoting 8 C.F.R. § 1208.18(a)(1)). We have jurisdiction to review the agency’s factual findings underlying the denial of CAT relief, as well as its legal conclusions. See Nasrallah v. Barr, 140 S. Ct. 1683, 1693 (2020). We review the agency’s factual findings for URIBE V. GARLAND 19 substantial evidence. Gutierrez-Alm, 62 F.4th at 1201 (citing Nuru v. Gonzales, 404 F.3d 1207, 1215 (9th Cir. 2005)). This standard requires Uribe to show that the record compels the conclusion that the agency’s decision was incorrect. See Sharma, 9 F.4th at 1060. We review legal questions de novo. Zheng v. Ashcroft, 332 F.3d 1186, 1193 (9th Cir. 2003). Here, Uribe has not experienced past torture in Mexico. And the BIA determined that Uribe failed to establish a likelihood of torture in Mexico, whether in a mental health facility or on account of his gang affiliation and tattoos. We hold that substantial evidence supports this decision, which is also free of legal error. Uribe’s first theory—that he will be tortured in a Mexican mental health facility—relies on the premise that he will not be able to obtain necessary mental health treatment in Mexico, and that without this treatment he will be involuntarily committed in a facility where he will then be tortured. Because the allegations of torture rest on a hypothetical chain of events, “CAT relief cannot be granted unless each link in the chain is ‘more likely than not to happen.’” Velasquez-Samayoa v. Garland, 49 F.4th 1149, 1154 (9th Cir. 2022) (quoting Matter of J-F-F-, 23 I. & N. Dec. 912, 917–18 (A.G. 2006)); see also Medina- Rodriguez v. Barr, 979 F.3d 738, 751 (9th Cir. 2020). Uribe’s claim fails at the first link because he has not shown it is more likely than not that he would be institutionalized based on his inability to obtain treatment in the country of removal. That is, the record does not compel the conclusion that Uribe will be unable to obtain his medication or other treatment in Mexico. Uribe’s argument that he will be unable to afford treatment in Mexico is also unavailing because it rests on a conclusory assertion that he cannot maintain employment without his family’s support. 20 URIBE V. GARLAND On this record, and considering that Uribe has never previously been institutionalized, the agency reasonably concluded that Uribe was not more likely than not to be committed to a mental health institution in Mexico. Uribe argues that the BIA committed legal error by applying an inappropriately high standard when assessing whether he showed a likelihood of institutional commitment. We disagree. In context, when stating that “one cannot know from the evidence whether [Uribe] will have access to medication [in Mexico],” the BIA was explaining Uribe’s failure to show error in the IJ’s factual findings and why Uribe’s theory rested on speculation. Indeed, the BIA articulated the “more likely than not” standard in the second part of the same sentence. Substantial evidence also supports the agency’s determination that, even if Uribe were committed to a mental health institution, he would not likely be tortured because the poor conditions in Mexico’s mental health facilities are not created with the requisite specific intent. For an act to constitute torture within the meaning of the CAT, it must be “specifically intended to inflict severe physical or mental pain or suffering.” Acevedo Granados, 992 F.3d at 765 (quoting 8 C.F.R. § 1208.18(a)(5)). Uribe produced evidence documenting poor conditions in Mexican mental health institutions. But, as the BIA explained, the record also supports a plausible inference that budgetary constraints, not an intent to torture those with psychiatric disabilities, are to blame. As we have explained in confronting similar claims about the risk of torture in Mexican mental health institutions, “evidence that conditions were squalid did not prove that any Mexican official had the specific intent to inflict suffering,” as URIBE V. GARLAND 21 required for CAT relief. Id.; see also Villegas v. Mukasey, 523 F.3d 984, 989 (9th Cir. 2008). Uribe’s claim that he will be tortured on account of his former gang membership and tattoos similarly fails. Although Uribe proffers evidence of widespread cartel violence in Mexico, he has not shown that his past or perceived gang affiliation will make him particularly vulnerable to such violence. See Dhital v. Mukasey, 532 F.3d 1044, 1051 (9th Cir. 2008) (per curiam) (“[T]he petitioner must demonstrate that he would be subject to a ‘particularized threat of torture . . . .’” (quoting Lanza v. Ashcroft, 389 F.3d 917, 936 (9th Cir. 2004))); see also Medina-Rodriguez, 979 F.3d at 750 (holding that the record did not compel the conclusion that the petitioner’s “tattoos make it more likely than not he will be tortured at the hands of a drug cartel with either the direct involvement or acquiescence of the Mexican government”). Nor does the record compel the conclusion that Mexican officials would acquiesce to Uribe’s torture. See Hernandez, 52 F.4th at 769 (quoting 8 C.F.R. § 1208.18(a)(1)). The agency therefore reasonably concluded that Uribe’s claim of torture was speculative. Finally, the BIA did not err by failing to consider the aggregate risk of torture. See Velasquez-Samayoa, 49 F.4th at 1154 (“[I]n assessing a CAT claim from an applicant who has posited multiple theories for why he might be tortured, the relevant inquiry is whether the total probability that the applicant will be tortured—considering all potential sources of and reasons for torture—exceeds 50 percent.”). Uribe acknowledges that the IJ explicitly considered the risks of torture “in aggregate.” Though the BIA did not specifically acknowledge that it was considering the aggregate risk, it agreed with the IJ’s findings and cited the pages of the IJ 22 URIBE V. GARLAND decision in which the IJ undertook the “aggregate” analysis. This is sufficient. See Hernandez, 52 F.4th at 772 (“Although the Board did not specifically address the combined probability of torture from different sources, its discussion of the CAT claim contained nothing to suggest that the Board had not adopted that aspect of the immigration judge’s reasoning.”). Considering the risks of torture in the aggregate, Uribe has not shown that the record compels a finding that he is likely to be tortured if returned to Mexico. PETITION DENIED.