Awil Mohamed v. Merrick B. Garland

United States Court of Appeals For the Eighth Circuit ___________________________

No. 20-1829 ___________________________

Awil Abdi Mohamed

Petitioner

v.

Merrick B. Garland, Attorney General of the United States

Respondents ____________

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

Submitted: April 15, 2021 Filed: August 13, 2021 ____________

Before KELLY, GRASZ, and KOBES, Circuit Judges. ____________

KOBES, Circuit Judge.

Awil Mohamed, a Somalian national, committed several felonies and was placed in removal proceedings. The Immigration Judge concluded that he would more likely than not be tortured in Somalia, but the Board of Immigration Appeals disagreed and ordered him removed. Mohamed says the BIA applied the wrong legal standard and improperly found new facts. We deny his petition for review.

I.

Awil Mohamed, a Somalian national who entered the United States in 2001, committed several felonies and was placed in removal proceedings in 2016. Mohamed introduced evidence that al-Shabaab, an Islamic militant group that controls large parts of Somalia, attacks people with ties to foreign governments and those who do not conform to its beliefs. Mohamed said he would more likely than not be tortured if he were returned to Somalia because his brother-in-law worked for the United States Navy. He also reported that his brothers had been kidnapped and beaten in Somalia by al-Shabaab members for using cell phones, and that his brother- in-law’s grandfather was abducted because he had connections to the United States.

The Immigration Judge found that it was more likely than not that Mohamed would be tortured if he were removed to Somalia. The IJ also concluded that the Somalian government would acquiesce in the torture because the government has been infiltrated by al-Shabaab members and because Mohamed is a member of a minority clan in Somalia, is westernized, and uses drugs and alcohol.

The BIA reversed the IJ and ordered Mohamed’s removal, finding that the IJ’s factual conclusions were clearly erroneous. The BIA concluded that the IJ’s reasoning was based on a hypothetical “chain of occurrences” instead of a plausible view of the facts and evidence in the record. Add. 10 (citation omitted). The BIA rejected the IJ’s conclusions about the likelihood of torture and about government acquiescence.

Mohamed appeals, raising two issues: (1) whether the BIA applied the wrong legal standard; and (2) whether the BIA failed to address relevant evidence and impermissibly engaged in its own factfinding.

-2-

II.

We review the BIA’s conclusions of law de novo. Doe v. Holder, 651 F.3d 824, 829 (8th Cir. 2011). “The BIA may review [the IJ’s] factual findings, including predictive findings of future events, only for clear error.” Uzodinma v. Barr, 951 F.3d 960, 965 (8th Cir. 2020). “[T]he Board must adequately explain why it rejected the IJ’s finding and identify reasons grounded in the record that are sufficient to satisfy a reasonable mind that there was clear error.” Omar v. Barr, 962 F.3d 1061, 1064 (8th Cir. 2020). The BIA may not perform independent factfinding, cannot reweigh the evidence, and can only “tak[e] notice of commonly known facts.” Id.

A.

Mohamed first argues that the BIA applied the wrong legal standard. We review that question and whether the BIA complied with the proper standard’s requirements de novo. Id.

Mohamed faults the BIA for relying on Matter of J-F-F-, 23 I. & N. Dec. 912 (A.G. 2006). He argues that Matter of J-F-F- is inapposite because it analyzed a chain of events leading to torture rather than the combined effect of several independent sources of risk. Mohamed says that the BIA should have instead applied the standard from Matter of G-A-, which considered all risk factors in the aggregate. 23 I. & N. Dec. 366, 368 (B.I.A. 2002). He argues that because the BIA applied the wrong standard, it failed to consider how his minority clan membership, his alcohol and drug use, his westernization in the United States, and his relative’s work for the United States Navy all come together to make it more likely than not that he will be tortured.

The BIA applied the proper legal standard. While the BIA acknowledged that Mohamed was at serious risk of violence and possibly torture if he were captured by al-Shabaab, it explained that the IJ clearly erred because Mohamed’s evidence could not establish a “greater-than 50% chance” that he would fall into al-Shabaab’s hands -3-

in the first place—something that must happen before al-Shabaab could torture him. Add. 11. The BIA noted that it reversed because the IJ strung together a “‘hypothesized chain of occurrences’ about what might happen to [Mohamed] in a worst-case scenario, without evidence that each link in the chain of inference is more likely than not to occur.” Add. 10 (citation omitted). That is, the BIA said the IJ failed to properly analyze the threshold issue: whether Mohamed would be apprehended by al-Shabaab to begin with.

Because the BIA analyzed a necessary chain of events, the BIA correctly applied Matter of J-F-F- and properly considered whether Mohamed’s evidence could show that each link in the chain is more likely than not to occur. See Matter of J-F-F-, 23 I. & N. Dec. at 917–18; see also id. at 918 n.4 (“It is the likelihood of all necessary events coming together that must more likely than not lead to torture, and a chain of events cannot be more likely than its least likely link.”).

B.

Mohamed next argues that the BIA impermissibly found facts and improperly reweighed the evidence in the record. He identifies three alleged new findings of fact: (1) the BIA’s statement that “only a small fraction of Somalia’s population is killed, injured, or tortured by al-Shabaab militants each year,” Add. 10; (2) the BIA’s note that “most victims” of al-Shabaab are “concentrated in conflict areas and regions under al-Shabaab control,” Add. 10; and (3) the BIA’s conclusion that “[t]here is no evidence in the record that the Somali government would force [Mohamed] to relocate” from Mogadishu to areas under al-Shabaab control or that Mohamed would “do so of his own volition,” Add. 10–11. He also says the BIA ignored evidence that supported the IJ’s conclusions.

The three BIA statements Mohamed points to are not independent factfinding. Each instead explains how the IJ’s conclusions outpaced the evidence. Plus, all three are supported by the record. First, Mohamed’s own evidence established that only a small fraction of Somalia’s total population of approximately twelve million is -4-

injured, killed, or tortured by al-Shabaab. Second, Mohamed’s evidence likewise established that most of al-Shabaab’s victims are located in areas the group controls—and not in the urban center of Mogadishu, Mohamed’s place of birth. Third, the BIA did not engage in factfinding when it pointed out the lack of evidence that Mohamed would be forced to move to areas under al-Shabaab control and apprehended. The BIA is tasked with evaluating the record to decide whether the IJ went beyond the evidence. The BIA did that here.

The BIA did not improperly find facts, ignore contrary information in the record, or reweigh the evidence. It instead explained how “the evidence relied on by the [IJ] does not support her inference that [Mohamed’s] personal characteristics give him a greater-than 50% chance of falling” into the hands of al-Shabaab. Add. 11. The BIA showed how the IJ took “an implausible view of the evidence,” Add. 11, and it put forward enough analysis grounded in the record to “satisfy a reasonable mind that [the IJ committed] clear error,” Omar, 962 F.3d at 1064.

C.

Mohamed’s alternative argument is that the BIA improperly analyzed the acquiescence element required for relief under the Convention Against Torture. Applicants for CAT relief must prove that: (1) “it is more likely than not that [the applicant] would be tortured if removed to the proposed country of removal”; and (2) “the relevant act(s) must be inflicted by or at the instigation of or with the consent or acquiescence of a public official or other person acting in an official capacity.” Mouawad v. Gonzales, 485 F.3d 405, 413 (8th Cir. 2007) (citation omitted). Mohamed says the BIA impermissibly “reweighed the evidence” about the second element “and substitute[d] its own finding for the IJ’s.” Mohamed Br. 33.

The IJ determined that Mohamed showed that the Somalian government would acquiesce in his torture for either of two reasons: first, the government would acquiesce because al-Shabaab is reported to have infiltrated the Somalian government; and second, the government would acquiesce because Mohamed is a -5-

minority clan member who is westernized, uses drugs and alcohol, and has ties to the United States military.

When reviewing the IJ’s decision, the BIA explained that the record evidence supporting the first reason was too slim. The evidence consisted primarily of a memorandum that made “a conclusory assertion of such infiltration, but none of the evidentiary exhibits cited in the memorandum [were] addressed to the question of infiltration.” Add. 11 (citations omitted). Plus, the BIA also noted that Mohamed’s own expert testified that he was not aware of “any incident in which a Somali government official acquiesced in torture by a member of al-Shabaab.” Id. Mohamed’s limited evidence makes it possible that al-Shabaab has infiltrated the Somalian government to some extent. But the BIA properly pointed out that the record could not support the IJ’s conclusion that al-Shabaab has infiltrated the Somalian government so completely that Somalian “public officials will acquiesce in his torture.” Garcia v. Holder, 746 F.3d 869, 873 (8th Cir. 2014).

Although the BIA did not address the IJ’s second reason separately from the first, the BIA held that it was also too thinly supported. See Add. 11 (“Under the circumstances, the [IJ] also clearly erred when she found that [Mohamed’s] torture at the hands of al-Shabaab would more likely than not occur with the acquiescence of a Somali public official.”). We agree. “A government does not acquiesce in the torture of its citizens merely because it is aware of torture but powerless to stop it.” Mouawad, 485 F.3d at 413 (citation omitted). Mohamed’s own evidence established this. His expert testified that the Somalian government would be unable, but not unwilling, to prevent torture. Plus, his expert said that he was unaware of any time where Somalian government officials acquiesced to torture at the hands of al-Shabaab.

Mohamed says minority clans like his do not enjoy the same protection from al-Shabaab as major clans and that major clans play a large part in the Somalian government. But that evidence addressed clan protection, not the Somalian

-6-

government’s acquiescence or consent to the torture of minority clan members. Plus, the on-point evidence from Mohamed’s expert goes the other way.

Mohamed does not direct us to any other evidence supporting the IJ’s conclusion that his westernization, drug use, or links to the United States Navy would lead the Somalian government to acquiesce in his torture. This evidence establishes that he would face a substantial chance of torture if he were to fall into al-Shabaab’s hands, but it does not show that the Somalian government would acquiesce in his torture because of those characteristics. As the BIA properly understood, the record reveals nothing more than how the Somalian government is “aware of torture but powerless to stop it.” Id. (citation omitted). Because of the lack of evidence showing that Mohamed’s characteristics would spur government acquiescence, the BIA did enough to “satisfy a reasonable mind that [the IJ committed] clear error.” Omar, 962 F.3d at 1064.

The BIA cabined itself to reviewing the record and concluded that the evidence could not support the IJ’s conclusions about Mohamed’s likelihood of torture and the Somalian government’s acquiescence. It did not reweigh the evidence or find its own facts. Because the BIA committed no legal or factual error, we deny Mohamed’s petition for review.

III.

The petition for review is denied. KELLY, Circuit Judge, dissenting.

Awil Abdi Mohamed applied for deferral of removal to Somalia under the Convention Against Torture (CAT). See 8 C.F.R. § 208.16. The IJ granted his application after concluding, based on witness testimony and other evidence Mohamed presented during the final removal hearing, that Mohamed will more likely than not be tortured in Somalia. The Department of Homeland Security (DHS)

-7-

appealed, and the BIA vacated the IJ’s decision in part and ordered Mohamed removed to Somalia. I would grant Mohamed’s petition for review because the BIA applied the wrong legal standard to his claim and engaged in impermissible factfinding.

In her final decision in this case, the IJ warned “that torture will happen and that it is not a matter of if, but when it would happen to [Mohamed].” She identified three independent risk factors for torture that, whether considered “individually or cumulatively,” create a greater than 50% chance that Mohamed will be tortured. These three risk factors, which find support in the record, are (1) al-Shabaab’s knowledge that Mohamed’s brother-in-law recently supported the United States Navy’s anti-terrorism operations in Somalia by serving as an interpreter; 1 (2) Mohamed’s visibly “Western” behavior, such as his addiction to drugs and alcohol; and (3) Mohamed’s membership in the minority Tunni clan, whose members are subjected to killings, torture, kidnaping and other crimes by militias and majority clans “with impunity.”

The BIA did not dispute that Mohamed will be targeted by al-Shabaab because of these risk factors. Indeed, it observed that “[t]he evidence upon which the [IJ] relied establishes that [he] will be at serious risk of violence, and perhaps torture or death, if he is unfortunate enough to fall into the hands of al-Shabaab.” Nevertheless, the BIA ruled that Mohamed is not entitled to deferral of removal because “the

1 Mohamed’s brother-in law, a United States citizen, testified that several Somali nationals living in the United States have referred to him as a “snitch” because of his work for the Armed Forces in Mogadishu. Two of these individuals were later ordered removed to Somalia and blamed the brother-in-law for that result. They threatened to seek retribution by disclosing the names and locations of his family members to al-Shaabab once removed to Somalia.

Indeed, the brother-in-law’s grandfather has already been abducted because of his grandson’s ties with the United States. Ante at 2. And Mohamed’s brothers, who live in Somalia, were beaten by al-Shabaab after one of them was observed using a cell phone, which the organization associates with Western values.

-8-

evidence relied on by the [IJ] does not support her inference that the respondent’s personal characteristics give him a greater-than 50% chance of falling under the custody or control of al-Shabaab militants in Somalia, such that it is more likely than not that they will torture him.” Essentially, the BIA agreed that Mohamed will be tortured if he is captured by al-Shabaab but found that he failed to establish he will more likely than not be captured by al-Shabaab in the first place. Yet the IJ never made any specific finding as to the likelihood of Mohamed falling into the hands of al-Shabaab. Instead, she determined only that the record, considered in its entirety, supports a greater than 50% chance of torture. This focus on the aggregated risk of torture is consistent with governing law. See Abdi Omar v. Barr, 962 F.3d 1061, 1065 (8th Cir. 2020) (explaining that a CAT petitioner must establish the probability of torture “considered in terms of the aggregate risk of torture from all sources”); 8 C.F.R. § 208.16(c)(4) (“If the [IJ] determines that the [noncitizen] is more likely than not to be tortured in the country of removal, the [noncitizen] is entitled to protection under the [CAT].”

Neither the BIA nor any court of appeals has previously required a petitioner seeking relief under the CAT to fulfill a separate prerequisite, see ante at 3, of establishing the sequence of events that will cause them to fall within the control of the person or entity that will torture them. See, e.g., In re G-A-, 23 I. & N. Dec. 366, 368 (B.I.A. 2002) (en banc) (concluding that the petitioner’s multiple risk factors for torture, in combination with evidence of widespread use of torture in Iran, established he would more likely than not be tortured there); Zewdie v. Ashcroft, 381 F.3d 804, 810 (8th Cir. 2004) (vacating portion of BIA order denying relief under the CAT because petitioner’s testimony about past torture and threats of future harm, in addition to evidence of human rights abuses in Ethiopian prisons, “provide[d] substantial grounds for believing that it is more likely than not [petitioner] will be tortured if forced to return to Ethiopia”); Xochihua-Jaimes v. Barr, 962 F.3d 1175, 1188 (9th Cir. 2020) (finding record compelled the conclusion that members of the Los Zetas gang in Mexico would more likely than not target petitioner for torture based on her testimony about past abuse and ongoing threats from gang members and their attempted kidnaping of her siblings). Whether a CAT -9-

petitioner has demonstrated a likelihood of capture by their torturer is not a separate prerequisite or threshold question but simply one consideration in evaluating whether they have met their burden of establishing a likelihood of torture. See Kang v. Att’y Gen. of U.S., 611 F.3d 157, 165 (3d Cir. 2010) (“All evidence relevant to the possibility of future torture should be considered when assessing if a petitioner is more likely than not to be tortured if removed.” (citing 8 C.F.R. § 208.16(c)(3))).

Here, the IJ determined that Mohamed has a greater than 50% chance of being tortured by al-Shabab. That he will more likely than not be captured by the group is a necessary component of that conclusion. The IJ’s assessment of the aggregate risk of torture was based on Mohamed’s readily discernible personal characteristics— known family ties to the United States, Western behavior and habits, and clan membership—that will lead al-Shabaab to target him for torture. I believe it was error for the BIA to require him to separately establish the series of events that will lead to his capture, in addition to demonstrating the likelihood of torture. Moreover, if the BIA believed the IJ erred by not specifically addressing the probability that Mohamed will be captured by al-Shabaab (in my view, she did not) the BIA “should have remanded the case to the IJ for further factfinding on this point.” Waldron v. Holder, 688 F.3d 354, 361 (8th Cir. 2012).

Instead of remanding to the IJ, however, the BIA engaged in impermissible factfinding See 8 C.F.R. § 1003.1(d)(3)(i) (“The [BIA] will not engage in de novo review of findings of fact determined by an [IJ]” and shall review the IJ’s finding of fact “only to determine whether [they] are clearly erroneous.”). It went beyond the scope of the IJ’s factfinding by stating that (1) “only a small fraction of Somalia’s population is killed, injured or tortured by al-Shabaab militants each year;” (2) “most victims” of al-Shabaab are “concentrated in conflict areas and regions under al- Shabaab control”; and (3) “[t]here is no evidence in the record that the Somali government would force [Mohamed] to relocate” there or that he would “do so of his own volition.” None of these findings appear in the IJ’s decision, and whether they are supported by the administrative record, see ante at 4, is not relevant because the BIA is not permitted to comb through the record to make new factual -10-

determinations. See Flores v. Holder, 699 F.3d 998, 1003–04 (8th Cir. 2012) (explaining that when the IJ makes “no factual findings” on a particular issue “any potential findings by the BIA [on the same issue] would be the result of an independent, improper factual analysis by the BIA”). Further, the appropriate inquiry for a CAT claim is whether “it is more likely than not that a particular individual will be tortured.” Ngure v. Ashcroft, 367 F.3d 975, 992–93 (8th Cir. 2004). That only a small portion of Somalia’s population may be tortured or harmed by al-Shabaab each year is immaterial. Comparing Mohamed’s risk of torture to the risk faced by members of the general public who do not share his risk factors is a meaningless exercise in the context of protection under the CAT.

Finally, I disagree with the BIA’s characterization of the IJ’s opinion as “stringing together a hypothesized chain of occurrences about what might happen to [Mohamed] in a worst-case scenario.” (internal quotations and citation omitted). Although the Attorney General held in Matter of J-F-F-, 23 I. & N. Dec. 912 (B.I.A. 2006) that an IJ cannot aggregate a series of dependent events—each of which has a less than 50% chance of occurring—to conclude that the petitioner has established a likelihood of torture, id. at 921, the facts of Mohamed’s case are distinct from those of Matter of J-F-F-. In that case, the IJ concluded the petitioner would more likely than not be tortured in the Dominican Republic because he required medication to behave lawfully; such medication was unavailable in the Dominican Republic; he would engage in “rowdy” behavior as a result of being unmedicated; this behavior would lead to his incarceration; and the police would torture him during his incarceration. Id. at 917. The Attorney General reversed, explaining that “[t]he evidence d[id] not establish that any step in this hypothetical chain of events [was] more likely than not to happen, let alone that the entire chain w[ould] come together to result in the probability of torture.” Id. at 917–18.

In Mohamed’s case, rather than imagining a domino effect of possible future events that could lead to his torture, the IJ identified three independent risk factors that, whether standing alone or considered cumulatively, make it more likely than not that he will face torture in Somalia. As such, the BIA should have assessed -11-

whether these multiple, independent risk factors “when considered in the aggregate” establish a likelihood of torture. In re G-A-, 23 I. & N. Dec. at 368 (B.I.A. 2002); see Shakkuri v. Barr, 780 F. App’x 286, 291–92 (6th Cir. 2019) (distinguishing between “independent” and “interdependent” probabilities of torture and explaining that “the probability that an applicant would be tortured if deported to the country of removal is the sum of the weighted probability of torture from each entity and for each reason”) (cleaned up); Kamara v. Att’y Gen. of U.S., 420 F.3d 202, 213–14 (3d Cir. 2005) (explaining that a CAT petitioner need only show that “the cumulative probability of torture [posed by different risk factors] exceeds 50%”).

I would grant the petition for review, vacate the BIA’s decision, and remand to the BIA to consider whether Mohamed has established a likelihood of torture based on the aggregate risk posed by the three risk factors the IJ identified. ______________________________

-12-