Yahye Herrow v. Attorney General United States of America

PRECEDENTIAL

UNITED STATES COURT OF APPEALS

FOR THE THIRD CIRCUIT

____________

No. 22-1854

____________

YAHYE HERROW,

Petitioner

v.

ATTORNEY GENERAL

UNITED STATES OF AMERICA

On Petition for Review of an Order of the

Board of Immigration Appeals

(Agency No. A075-713-985)

Immigration Judge: David W. Crosland

Argued on March 22, 2023

Before: RESTREPO, PHIPPS and ROTH, Circuit Judges

(Opinion filed: February 13, 2024)

Christopher M. Casazza

Caitlin J. Costello (ARGUED)

Palladino, Isbell & Casazza

1528 Walnut Street

Suite 1701

Philadelphia, PA 19102

Counsel for Petitioner

Michael Engler

Tracie N. Jones (ARGUED)

United States Department of Justice

Office of Immigration Litigation

P.O. Box 878 Ben Franklin Station

Washington, DC 20044

Counsel for Respondent

O P I N I ON

ROTH, Circuit Judge:

Yahye Herrow petitions for review of the Board of

Immigration Appeal’s (BIA) denial of his claims for

withholding of removal and relief under the Convention

Against Torture (CAT). He contends that the BIA erred in

failing to consider evidence favorable to his CAT claim and in

2

finding that “Repatriated Minority Somalis” does not

constitute a cognizable particular social group. We agree with

the BIA that “Repatriated Minority Somalis” does not

constitute a cognizable particular social group. Therefore, we

will deny his claim for withholding of removal. However, we

conclude that the BIA, in deciding his CAT claim, failed to

consider evidence favorable to Herrow. For that reason, we

will remand his petition as it applies to that claim.

I. BACKGROUND

A. Factual Background

Yahye Herrow is a member of the Bandabow Tribe,1

which is a sub-clan of the Reer Hamar Benadiri. The Reer

Hamar are ethnic Somali Bantus, a minority in Somalia. He

and his family, who lived in Mogadishu, have been targeted

because of their tribal affiliation. The Darod and Hawiye tribes

burned the homes of his people, who “did not have any

rights.”2 Al Shabaab3 and other terror groups have also

subjected his family to violence. Herrow’s brother, Usman,

died in a 2005 hotel explosion that killed 700 people. A

bombing in Mogadishu also killed Herrow’s sister, Fatima.

In 1992, Herrow and his family fled Mogadishu. Days

1

The record includes varied spellings of Bandabow, including

Bandawow and Bandhowow. Each refers to the same minority

clan to which Herrow belongs.

2

Certified Administrative Record (CAR) 112.

3

Al-Shabaab, Al-Shabbab, Al Shabbab, Al Shabaab, and the

Shabab are used interchangeably throughout the record to refer

to the same group in Somalia.

3

later, they arrived in Nairobi, Kenya, where Herrow stayed

until May 2000, when his aunt and clans people arranged for a

smuggler to bring him to the United States.

Herrow arrived in the United States through Mexico

with a Kenyan passport. He immediately applied for asylum.4

In July 2000, Herrow was granted asylum in Los Angeles,

California. He met his wife Munira Mohamed Adan, a U.S.

citizen, in 2001. They are currently married, and she resides in

Minneapolis, Minnesota. One of his brothers still lives in

Somalia, but he does not work and “hides around and lives in

different places.”5 Herrow is not aware of his location.

B. Procedural History

i. Conviction and Notice to Appear

In 2018, a jury in the United States District Court for the

District of Minnesota convicted Herrow of Conspiracy to

Commit Mail Fraud and Wire Fraud, in violation of 18 U.S.C.

§§ 1349 and 1341, for his part in a health-care fraud scheme.6

4

In his 2000 application for asylum, Herrow alleged false facts.

He claimed that the smuggler who brought him to the United

States “told him that he would ‘get farther’ if he adopted the

narrative his smuggler wrote out for him [and the smuggler]

insisted that this was the only way for [Herrow] to obtain

asylum.” CAR 78.

5

CAR 229.

6

As part of a scheme to bill insurance companies for services,

a chiropractic medical office, for which Herrow was already

working as a driver, paid him between $100 to $200 for

bringing in individuals involved in car accidents. Herrow

4

The court sentenced Herrow to twelve months and one day

incarceration.

In 2019, the U.S. Department of Homeland Security

issued Herrow a Notice to Appear, charging him with removal

under § 237(a)(2)(A)(iii) of the Immigration and Nationality

Act (INA) due to his conviction.

ii. Herrow’s I-28, I-485, I-605, I-589, and Accompanying

Evidence

To obtain relief from removal, Herrow submitted a

Form I-485 (Application to Register Permanent Residence or

Adjust Status), a Form I-602 (Application by Refugee for

Waiver of Grounds of Excludability), and a Form I-589

(Application for Asylum, Withholding of Removal, and Relief

under CAT).

In his brief in support of his submissions, Herrow

asserted a fear of persecution and torture in Somalia due to his

membership in five separate particular social groups, including

“Repatriated Minority Somalis.”7 He also claimed that the

Somali government and Al Shabaab will subject him to torture

because of his “westernization”; his extended time abroad; the

suspicion that he is a Western spy; his minority status as a

Bantu; his lack of clan ties; the perception that he could pay

ransom; and his refusal to adhere to Al Shabaab’s rigid and

extreme views of Islam.8 In his submissions, he also pointed

referred approximately 19 patients to the clinic, making around

$2,000 in total as a “runner.” CAR 267.

7

CAR 275.

8

CAR 278.

5

to the government’s acquiescence in Al Shabaab’s conduct.

With his brief, Herrow submitted a wide range of

evidence: The U.S. Department of State’s 2019 Human Rights

Report on Somalia; Human Rights Watch’s 2020 Somalia

report; Amnesty International’s 2019 Somalia report; and the

Canadian Government Report on Somalia. Herrow also

attached a Georgetown Immigration Law Journal article

surveying the experience of twenty Somali Bantus (the Somali

minority to which Herrow and other Reer Hamar belong) who

were deported and returned to Somalia between 2016 and 2018

(the Georgetown Article). In addition, Herrow included two

declarations from a 2017 case concerning deportees in

Somalia, as well as numerous articles highlighting clan

relationships and status, Al Shabaab and its treatment of

returnees, and acquiescence and instability in the government

and police forces.

iii. May 2020 Merits Hearing

In May 2020, the Immigration Judge (IJ) held a merits

hearing. Herrow withdrew his I-485 and I-602 applications

and proceeded on only his I-589 (Application for Asylum,

Withholding of Removal, and Relief under CAT).

Herrow testified that if he were to go back to Somalia,

Al Shabaab would target him due to his membership in a

particular social group, his life in the United States, and his

status as a repatriated minority Somali. In Herrow’s words, Al

Shabaab has “infiltrated every part of the government. They’re

at the airport. They are, you know, everywhere. And [if] they

6

find me, they will kill me.”9 He believes he would be targeted

because he is perceived as a U.S. spy and his culture has

changed. He testified that “it’s not if, it’s when they’ll cut me

into pieces.”10 He explained that people who have previously

returned have been killed or jailed. He testified that Al

Shabaab has infiltrated the police and government, so there is

nowhere to go. Al Shabaab is in the government and works

with the government. They have an agenda and they

“assassinate ministers and anybody they want . . . , put[ting]

bombs on the street and kill[ing] people that way.”11

iv. Decision of the Immigration Judge

The Immigration Judge (IJ) terminated Herrow’s

asylum status and denied his applications for relief, ordering

him removed to Somalia. The IJ found Herrow credible.

However, because of his criminal conviction and because his

original asylum petition recited false facts, the IJ terminated

his asylee status. Due to his conviction, the IJ denied his new

application for asylum.

The IJ also denied Herrow’s application for withholding

of removal under INA § 241(b)(3). First, the IJ found that, of

Herrow’s proposed groups, only the “Reer Hamar or

Bandabow clan/tribe” constitutes a particular social group.12

Second, the IJ found that the record did not support a clear

probability of future persecution of Herrow based on his

9

CAR 213–14.

10

CAR 233.

11

CAR 232.

12

CAR 129.

7

membership in this group.13 Finally, the IJ rejected Herrow’s

contention that the government of Somalia is unwilling or

unable to protect him from the persecution he fears upon return

to Somalia.

Turning to CAT, the IJ rejected Herrow’s application

for protection. The IJ concluded that Herrow is not likely to

face torture beyond the risk common to all returnees.

v. Appeal to the Board of Immigration Appeals

Herrow appealed the IJ’s decision. With regard to

withholding of removal, Herrow made three arguments: first,

that the IJ erred in finding that “Minority Somalis,” “Minority

Somalis who lack clan and/or family ties,” and “Repatriated

Minority Somalis” are not cognizable social groups;14 second,

that the IJ erred in finding that the evidence does not show that

he would be subject to a pattern or practice of persecution if he

returned to Somalia; third, that the IJ erred in finding that the

“record does not reflect that the government of Somalia is

unwilling or unable to protect [Herrow] from the persecution

he fears.”15

With regard to CAT, Herrow made several arguments:

first, that the IJ erred in failing to apply the legal standard we

set out in Myrie v. Attorney General;16 second, that the IJ

applied the wrong standard in determining whether Herrow

would be tortured if he returned to Somalia; finally, that the IJ

13

CAR 131.

14

CAR 72.

15

CAR 73.

16

855 F.3d 509 (3d Cir. 2017).

8

did not consider all of the evidence in the aggregate in making

its determination.

The BIA dismissed Herrow’s appeal, ordering his

removal to Somalia. As is relevant here, the BIA agreed with

the IJ’s determination that “Repatriated Minority Somalis” is

not a cognizable particular social group because the terms

“minority” and “repatriated minority” are “too amorphous,

overbroad, and diffuse.”17 Further, the BIA, in applying Myrie

and citing to the IJ’s opinion, found that Herrow failed to allege

a sufficient likelihood of harm or to assert whether the harm

amounted to torture and whether the government of Somalia

would acquiesce in that torture.

Herrow now petitions for our review of three issues: (1)

whether the BIA erred in finding that “Repatriated Minority

Somalis” is not a cognizable social group; (2) whether the BIA

erred in finding that Herrow was not likely to face torture in

Somalia; and (3) whether the IJ erred in finding that the Somali

government would not acquiesce in torture.18

II. JURISDICTION AND STANDARD OF

REVIEW

The BIA had jurisdiction under 8 C.F.R. § 1003.1(b)(3).

We have jurisdiction under 8 U.S.C. § 1252(a).19

17

CAR 12–13.

18

Petitioner Br. 1–2.

19

The government submitted a letter pursuant to Federal

Rules of Appellate Procedure Rule 28(j) acknowledging our

jurisdiction under § 1252(a).

9

Where the BIA affirms the decision of the IJ and “set[s]

forth somewhat its own rationale and analysis,” we review both

the BIA’s and the IJ’s decisions.20 However, “[i]f the Board

relies only on some of the grounds given for denying relief, we

review only th[ose] grounds.”21 We review the BIA’s

affirmance of an IJ’s factual findings under a substantial

evidence standard.22 In essence, “administrative findings of

fact are conclusive unless any reasonable adjudicator would be

compelled to conclude to the contrary.”23 We review the IJ’s

and BIA’s legal determinations, including mixed questions of

law and fact, de novo.24

III. DISCUSSION

A. Convention Against Torture

Herrow claims that the BIA and IJ erred in denying his

CAT claim and in finding that (1) he is unlikely to face torture

and (2) the Somali government would not acquiesce in such

torture. Because the BIA and IJ ignored evidence favorable to

Herrow, we will grant his petition in part and remand for a

more comprehensive review of the evidence.

20

Leia v. Ashcroft, 393 F.3d 427, 433 n.5 (3d Cir. 2005);

Myrie, 855 F.3d at 515.

21

Myrie, 855 F.3d at 515 (citing Chukwu v. Att’y Gen., 484

F.3d 185, 193 (3d Cir. 2007)).

22

Tarrawally v. Ashcroft, 338 F.3d 180, 184, 186 (3d Cir.

2003); Myrie, 855 F.3d at 516–17.

23

INA § 242(b)(4)(B), codified at 8 U.S.C. § 1252 (b)(4)(b).

24

Arreaga Bravo v. Att’y Gen., 27 F. 4th 182, 185 (3d Cir.

2022); Myrie, 855 F.3d at 516–17; Matter of Cabrera, 24 I. &

N. Dec. 459, 460 (BIA 2008).

10

The CAT provides that “[n]o State Party shall expel,

return . . . or extradite a person to another State where there are

substantial grounds for believing that he would be in danger of

being subjected to torture.”25 CAT became binding on the

United States in 1994, at which time it became “the policy of

the United States not to expel . . . or otherwise effect the

involuntary return of any person to a country in which there are

substantial grounds for believing the person would be in danger

of being subjected to torture . . . .”26 The burden of proving the

likelihood of torture is on the applicant.27 If the applicant

satisfies that burden, “withholding of removal or deferring of

removal [under CAT] is mandatory.”28

An act constitutes torture if it is:

(1) an act causing severe physical or mental pain

or suffering; (2) intentionally inflicted; (3) for an

illicit or proscribed purpose; (4) by or at the

instigation of or with the consent or acquiescence

of a public official who has custody or physical

control of the victim; and (5) not arising from

25

Art. 3(1), S. Treaty Doc. No. 100–20, 1465 U.N.T.S. 85.

26

Silva–Rengifo v. Att’y Gen., 473 F.3d 58, 64 (3d Cir. 2007)

(quoting Foreign Affairs Reform and Restructuring Act of

1998, Pub. L. No. 105–277, div. G., tit. XXII, § 2242, 112

Stat. 2681–822 (codified at 8 U.S.C. § 1231)); U.N. Doc. 571

Leg/SER.E/13.IV.9 (1995).

27

8 C.F.R. § 208.16(c)(2).

28

Silva–Rengifo, 473 F.3d at 64 (citing 8 C.F.R. §§ 1208.16–

18).

11

lawful sanctions.29

Herrow contends that the IJ and BIA erred because they

improperly ignored evidence that would lead a rational

factfinder to conclude that he is likely to face torture and that

the Somali government would acquiesce in that torture. We

agree. As is clear under our existing law, the IJ and the BIA

“may not ignore evidence favorable to the [noncitizen].”30

i. Likelihood of Torture

To establish a likelihood of future torture, the record

must demonstrate an aggregate risk of torture to the noncitizen

that exceeds fifty percent.31 In making this determination, the

IJ must address what is likely to happen to the petitioner if

removed,32 and whether “what is likely to happen amount[s] to

the legal definition of torture.”33 In answering these questions

29

Myrie, 855 F.3d at 515 (quoting Auguste v. Ridge, 395 F.3d

123, 151 (3d Cir. 2005)). Herrow need not, however, “establish

that torture is inflicted ‘on account of’ any protected status.”

Silva-Rengifo, 473 F.3d at 64. In essence, Herrow “must

establish a likelihood of being subjected to torturous acts

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.” Id. (quoting 8 C.F.R. § 1208.18(a)(1)).

30

Huang v. Att’y Gen., 620 F.3d 372, 388 (3d Cir. 2010) (citing

Espinosa-Cortez v. Att’y Gen., 607 F.3d 101, 107 (3d Cir.

2010)); Quinteros v. Att’y Gen., 945 F.3d 772, 786 (3d Cir.

2019).

31

Kamara v. Att’y Gen., 420 F.3d 202, 213–14 (3d Cir. 2005).

32

Id.

33

Myrie, 855 F.3d at 515.

12

here, the BIA and IJ found that Herrow did not demonstrate a

likelihood of torture. We conclude, however, that this

determination could not have been made if all the evidence

presented by Herrow had been properly considered.

The government contends that we should deny

Herrow’s petition under Hernandez Garmendia v. Attorney

General, 28 F.4th 476, 484 (3d Cir. 2022), which requires that

a petitioner put forth evidence that he is likely to be singled out

and tortured.34 We correctly denied Hernandez Garmendia’s

petition under this standard because there was no evidence in

the record to suggest that Hernandez Garmendia would be

singled out and tortured. Here, however, Herrow provided

ample support for his claim that he will be singled out and

tortured if he returns to Somalia due to his status as a returnee

from the United States and as a member of a minority clan.

For example, the IJ concluded that even though “Al-

Shabaab operatives have condemned returnees from Western

countries as ‘infidels’ and threatened to retaliate, and even to

kill, those suspected of spying,” Herrow “was unable,

especially in light of his twenty-eight-year absence from the

country, to provide any reason for why he would be singled out

for torture.”35 It is exactly that twenty-eight-year absence,

however, that supports his being singled out. The IJ ignored

extensive evidence in the record, including reports and

declarations of experts, that returnees from the United States

are easily identifiable as “American” because of clothing,

learned gestures, weight, lighter skin color, and accent. The IJ

did not engage with this evidence. In fact, the IJ found Herrow

34

Dkt. 33 at 13:21–25.

35

CAR 102.

13

credible but ignored his assertions that Al Shabaab would

target him due to his shaved head, size, cultural differences,

and tribe. The evidence demonstrates that returnees like

Herrow are “considered inherently suspect” and “targeted for

violence and death on this basis.”36

The IJ also failed to engage with other evidence in the

administrative record demonstrating a likelihood of torture to

Herrow. For example, the IJ makes no reference to the

Georgetown Article, in which the author surveyed the twenty

Somali Bantus deported and returned to Somalia between 2016

and 2018. Of these, fifty-five percent reported being tortured

upon their return. In fact, of those returned in 2018, 66.7

percent were tortured. Based on Herrow’s identifiable traits

and the statistics garnered in the Georgetown Article, the

evidence does not support the IJ and the BIA’s conclusion that

Herrow “could not articulate why [Al Shabaab] would target

him specifically, aside from happenstance.”37

The IJ asserts that Herrow can mitigate his risk of

torture by relocating to a major urban center, like Mogadishu.

This finding is contrary to the evidence in the record.38 As an

initial matter, the record suggests relocation is difficult, and

36

CAR 312.

37

CAR 135.

38

Herrow rightfully notes that “diminished risk is not the

standard under which we determine the likelihood of torture.”

Dkt. 33 at 25:6–19. Even if Herrow could diminish his risk of

torture by moving to Mogadishu—a claim that finds minimal

support in the record—the BIA and IJ provide no support from

the record to suggest that the risk would diminish to the extent

that Herrow’s risk of torture is less than 50 percent.

14

relocation itself can open individuals up to violence and

torture. The U.S. Department of State’s Somalia report states

that “[c]heckpoints operated by government forces, allied

groups, armed militias, clan factors, and al-Shabaab inhibit[]

movement and expose[] citizens to looting, extortion,

harassment, and violence.”39 The airport provides no better

means of relocating. The Georgetown Article notes that “35

percent [of the surveyed individuals] could not even leave the

Mogadishu International Airport without being detained.”40

Ninety percent of those surveyed experienced some form of

abuse, torture, interrogation, or extortion at the airport.

The IJ suggests that Herrow would be safe in

Mogadishu or another city due to the rise in status of his clan,

the Reer Hamar. However, the IJ cherry picks facts from the

record. While one report notes that some Benadiri and Reer

Hamar have acquired positions in the government, most

positions are in local and regional governments. A different

report posits that this representation is minimal and “token.”41

In fact, there is evidence that only a few thousand Benadiri and

Reer Hamar remain in Somalia at this point, with even fewer

in Mogadishu—to the point where the clan has “almost ceased

to exist.”42 They are forced to pay another clan or private

militia for protection or gain protection through the marriage

39

CAR 398.

40

CAR 512

41

CAR 532.

42

CAR 576 (“The minorities’ areas are ghost towns and it is

unknown how many minority group members remain in

Mogadishu.”); CAR 576 (“90% of the Rer [sic] Hamar

population in Mogadishu have left the city as a consequence of

civil war and lack of security for this group.”).

15

of their daughters to stronger clans.43 This level of protection

is not afforded to all clan members; those who lack

marriageable daughters or the funds to afford protection are out

of luck. After a twenty-eight-year absence, Herrow, like all

those “who have been in the United States for years or decades

lack[s] any meaningful clan or social networks that are crucial

to maintain what safety can be had in Somalia.”44

Additional evidence also conflicts with the IJ’s finding

that Herrow can mitigate his risk in Mogadishu due to greater

government and police presence. Al Shabaab “maintains a

presence in Mogadishu and is able to attack . . . buildings . . .

and other targets almost at will. . . . It carried out at least 33 car

bombings in Mogadishu from January to November 2017.”45

Herrow points to a number of other attacks in Mogadishu,

including one by a political aide inside the Mogadishu mayor’s

office that led to the mayor’s death.46 The evidence repeatedly

notes that Al Shabaab has significant technical and intelligence

capabilities and has infiltrated government and securities

institutions throughout Somalia, including Mogadishu.

Finally, in addition to issues within the airport itself, the

43

CAR 192 (“for the [Reer Hamar] ‘protection’ has meant

forced payment of bribes and being subjected to extortion by

majority clans’ armed and lawless militias, as a means of

raising illegal revenue—and is more accurately akin to

‘protection rackets’ . . . it is axiomatic, therefore, that a person

must be from a dominant clan to be able to access this kind of

protection.”).

44

CAR 359.

45

CAR 352.

46

In fact, two of Herrow’s siblings were killed in explosions in

Mogadishu.

16

Georgetown Article found that twenty percent of those

surveyed were kidnapped and tortured outside the airport

before being able to enter Mogadishu. Further, all of those

surveyed stated “that they were orally or physically abused,

detained for more than four hours, forced to pay a bribe,

interrogated, kidnapped, tortured, or forced to pay ransom”

while in Mogadishu.47 In the City of Kismayu, 66.7 percent of

those surveyed were treated similarly. The IJ engages with

none of this evidence, which taken as a whole cannot support

the conclusion that Herrow “may be able to mitigate some of

his risk by locating to a major urban center.”48

While the IJ noted that it considered all the evidence,

none of the above is specifically mentioned. The IJ need not

address every piece of evidence in the administrative record,

but it cannot ignore evidence favorable to Herrow.49 Because

the IJ does ignore such evidence, we grant Herrow’s petition

for review and remand this case to the BIA for reconsideration

of all evidence before it.

ii. Acquiescence to Torture

Herrow also challenges the IJ’s conclusion that the

evidence does not support a finding that the Somali

government would acquiesce to torture.50

47

CAR 499.

48

See CAR 102.

49

Huang, 620 F.3d at 388.

50

Because the BIA did not reach the issue of acquiescence, we

review the IJ’s findings. See CAR 5 (“Because we uphold this

finding [that there is no likelihood of torture], we need not

address whether the government would acquiesce in

17

“Acquiescence of a public official requires that the

public official, prior to the activity constituting torture, have

awareness of such activity and thereafter breach his or her legal

responsibility to intervene to prevent such activity.”51 In

assessing acquiescence, the IJ must ask “how public officials

will likely act in response to the harm the petitioner fears” and

“whether the likely response from public officials qualifies as

acquiescence.”52 An individual can establish acquiescence by

showing that a public official was actually aware of torture or

willfully blind to the torture.53 “[T]he question of whether

likely government conduct equates to acquiescence is a mixed

question of law and fact . . ..”54

Again the IJ’s conclusion that the government’s conduct

would not amount to acquiescence was based on cherry picked

record evidence. The IJ ignored ample evidence favorable to

Herrow that is contrary to its findings. First, the IJ states that

the government would not acquiesce because Reer Hamar has

gained political status and position in the government. The IJ

explains that because of this gained power, “it is unlikely that

[Herrow’s] torture.”).

51

8 C.F.R. § 1208.18(a)(7).

52

Myrie, 855 F.3d at 516. The first question is reviewed for

clear error, while the second question is reviewed de novo. Id.

at 516–17; Arreaga Bravo, 27 F. 4th at 185.

53

Myrie, 855 F.3d at 516; Silva-Rengifo, 473 F.3d at 65 (“The

CAT does not require an alien to prove that the government in

question approves of torture, or that it consents to it. Rather . .

. an alien can satisfy the burden established for CAT relief by

producing sufficient evidence that the government in question

is willfully blind to such activities.”).

54

Myrie, 855 F.3d at 516.

18

the government would condone the targeting of one of its

members.”55 As discussed above, that overstates Reer

Hamar’s status and position in the government. It remains a

small community with limited power. And most security

forces are closely identified with a majority clan that likely has

no incentive to protect Herrow.56 A 2017 U.S. Department of

State Human Rights Report on Somalia aptly describes the

implications of this situation, explaining that “[m]inority

groups, often lacking armed militias, continued to be

disproportionately subjected to killings, torture, rape,

kidnapping for ransom, and looting of land and property with

impunity by faction militias and majority clan members, often

with the acquiescence of federal and local authorities.”57

Second, our case law is contrary to the IJ’s finding that

because the Somali government is engaged in an active internal

conflict against Al Shabaab, it has not acquiesced in their

torture. We have held that such facts do not preclude a

showing that a government was willfully blind because “[a]n

applicant can establish governmental acquiescence even if the

government opposes the [] organization that is engaged in

torturous acts.”58

55

CAR 103.

56

See CAR 564 (“To the extent that the population turns to the

police for assistance, they turn . . . to police officers from the

same clan. Other sources also emphasize that the loyalty of the

police and other members of the government forces normally

rests with their own clan.”).

57

CAR 486, 495, 552 (emphasis added) (quoted throughout

record).

58

Myrie, 855 F.3d at 518 (quoting Pieschacon-Villegas v.

Att’y Gen., 671 F.3d 303, 312 (3d Cir. 2011)) (second

19

Third, the IJ ignores evidence suggesting the Somali

government’s own participation and acquiescence in the

torture of minority returnees. For example, the IJ never

engages with the fact that “[m]ost instances of detention,

kidnapping, and torture [of the Bantu deportees surveyed in the

Georgetown Article] were carried out by one or more

uniformed Somali government security personnel.”59 In fact,

ninety percent of those surveyed “experienced oral or physical

abuse, detention for more than one hour, forcible payment of a

alteration in original); see also id. at 516 (“Circumstantial

evidence may establish acquiescence to targeted acts of

violence even when the government has an official policy or is

engaged in a campaign of opposition against the entity the

applicant fears.”).

59

CAR 518; see also CAR 486 (“Survey results from the

Somali Bantu deportees reveal that most were kidnaped and

tortured for ransom by uniformed Somali police or armed

groups that the Somali Government was unwilling or unable to

control. Some were kidnapped and tortured for ransom upon

arrival at the Mogadishu International Airport (MIA) by

Somali government security personnel . . . .”); CAR 518 (“Of

the seven instances at MIA, uniformed Somali government

security personnel were responsible for 57.1 percent of

abductions. Of the 4 instances in Mogadishu, uniformed

Somali government security personnel were responsible for 75

percent of the abductions and torture. Of the two instances in

Kismayu, uniformed Somali government security personnel

were responsible for one of the abductions and torture.”

(citations omitted)); CAR 500 (“Of the eight Somali Bantu

deportees who were abused in the City of Kismayu, 50 percent

identified their abusers as the Jubbaland State police and 50

percent as Al Shabaab.”).

20

bribe, interrogation, torture, or forcible payment of a ransom

by uniformed Somali security personnel stationed at the

airport.”60 To the extent that individuals involved are plain-

clothed and impersonating customs officials, the article authors

suggest that they are “doing so with the consent of actual public

officials,” “evident from the fact that these plain-clothed

individuals are able to detain Somali Bantu deportees in the

airport, a government-controlled facility.”61 Additionally, a

Canadian Government report suggests that, while authorities

have used military courts where security forces abused

civilians, “they generally did not investigate abuse by police,

army, or militia members” and “a culture of impunity was

widespread.”62

Fourth, the IJ ignores evidence that members of Al

Shabaab constitute the government, instead treating them as

separate entities. The IJ’s conclusion that members of Al

Shabaab are rogue operatives attempting to sabotage the

government and that their actions are not that of the

government itself is unsupported by case law. In 2020, the

U.S. Attorney General (AG) rejected the contention that

misuse of government authority or extrajudicial acts do not

qualify as government action.63 Instead, the AG explained that

“[i]t is misuse of authority, ‘made possible only because the

wrongdoer is clothed with the authority’ . . . of law, that may

violate the CAT regulations.”64 Similarly, we agree with the

60

CAR 499.

61

CAR 518.

62

CAR 564.

63

Matter of O-F-A-S-, 28 I. & N. Dec. 35, 41 (2020).

64

Id. (citing West v. Atkins, 487 U.S. 42, 49 (1988)); see also

id. (“By immunizing extrajudicial action by low-level officials

21

Fifth Circuit that:

[P]roving action in an officer’s official capacity

“does not require that the public official be

executing official state policy or that the public

official be the nation’s president or some other

official at the upper echelons of power. Rather .

. . the use of official authority by low-level

officials, such a[s] police officers, can work to

place actions under the color of law even where

they are without state sanction.”65

The IJ’s conclusion is unsupported by the evidence. In

many regions of Somalia, Al Shabaab fully governs.

from the CAT's scope, a freestanding ‘rogue official’ rule

would appear to disqualify much of what the ‘under color of

law’ rule might otherwise qualify as ‘torture.’”). To the extent

that rogue officials do act for purely private reasons, the

Second Circuit has suggested that the fact that torture is

“routine” and tied to other criminal justice system objectives

can demonstrate knowledge by officials or willful blindness.

CAR 516 (citing Khouzam v. Ashcroft, 361 F.3d 161, 171 (2d

Cir. 2004)). However, the court noted that “when it is a public

official who inflicts severe pain or suffering, it is only in

exceptional cases that we can expect to be able to conclude that

the acts do not constitute torture by reason of the official acting

for purely private reasons.” CAR 516–17 (quoting Khouzam,

361 F.3d at 171).

65

Marmorato v. Holder, 376 F. App’x. 380, 385 (5th Cir.

2010) (alterations in original) (quoting Ramirez-Peyro v.

Holder, 574 F.3d 893, 901 (8th Cir. 2009)); CAR 516 (quoting

Marmorato, 376 F. App’x at 385).

22

Meanwhile, in areas run by the Somali government, Al

Shabaab has infiltrated all levels of the government and

security forces. Herrow himself asserts that “they’re the most

powerful in the government.”66 As at least one declaration in

the record notes, “an individual can be a policeman by day and

an Al Shabaab operative by night.”67 The record demonstrates

Al Shabaab’s use of government authority in abducting,

torturing, and attacking civilians, including returnees to

Somalia. It is through its members’ roles in the government

that Al Shabaab can target and keep track of people. It is with

the authority of law, working as Somali security personnel

stationed at the airport, that Al Shabaab is able to kidnap,

abduct, and torture returnees. They can locate individuals and

take action because Al Shabaab “exists right at the heart of

federal government and is ‘present during cabinet

meetings.’”68

Based on ample evidence in the record that the IJ failed

to consider and that contradicts the IJ’s findings on

acquiescence, we will grant Herrow’s petition with respect to

his claim that the BIA ignored favorable evidence, remanding

to the BIA for consideration of all the evidence under the

appropriate legal standards.

B. Withholding of Removal

Herrow claims that the BIA and IJ erred in denying his

withholding of removal claim based on his membership in the

66

CAR 231.

67

CAR 353.

68

CAR 311 (emphasis omitted).

23

“Repatriated Minority Somalis” group.69 Because we agree

with the BIA and IJ that “Repatriated Minority Somalis” is not

a cognizable particular social group, we will deny Herrow’s

petition in part.

A person is not subject to removal and is granted

withholding of removal if his “life or freedom would be

threatened in that country because of [his] race, religion,

nationality, membership in a particular social group, or

political opinion.”70 To establish membership in a particular

social group, a person must show 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.”71

A characteristic is immutable if it cannot be changed or

should not “be required to [be] change[d] as a matter of

conscience to avoid persecution.”72 “A group is particularized

if it is discrete, has definable boundaries—as opposed to being

overbroad, diffuse, or subjective . . . .”73 A group’s size,

however, does not preclude it from being cognizable under the

69

The determination of whether a group constitutes a

cognizable social group is reviewed de novo. Matter of W-Y-

C- & H-O-B-, 27 I. & N. Dec. 189, 191 (BIA 2018).

70

8 U.S.C. § 1231(b)(3).

71

Radiowala v. Att’y Gen., 930 F.3d 577, 583 (3d Cir. 2019)

(citing S.E.R.L. v. Att’y Gen., 894 F.3d 535, 540 (3d Cir.

2018)); Matter of M-E-V-G-, 26 I. & N. Dec. 227, 237 (BIA

2014).

72

Radiowala, 930 F.3d at 583.

73

Id.

24

INA.74 That said, there must be a clear “benchmark for

determining who falls within” the group.75 To be socially

distinct, there must be “evidence that the society in question

recognizes a proposed group as distinct.”76

Here, the BIA held that “Repatriated Minority Somalis”

is not a cognizable social group because the term “repatriated

minority” is “too amorphous, overbroad, and diffuse to satisfy

the particularity requirement.”77 In doing so, it agreed with the

IJ that “minority” can encompass any “relatively small group

of people in Somalia” and is neither “limited to one’s clan

membership” nor “narrowed by any other criterion.” 78 We

agree with the IJ’s assessment.

Herrow argues that his proposed group is no more

amorphous than other groups courts have recognized, such as

“Somali females” or “Iranian women.”79 He is mistaken. All

members of those groups share a commonality: they are

women. Meanwhile, as the IJ correctly noted, “minority” can

encompass persons belonging to “tribal, ethnic, religious,

74

N.L.A. v. Holder, 744 F.3d 425, 438 (7th Cir. 2014) (“[I]t

would be antithetical to asylum law to deny refuge to a group

of persecuted individuals who have valid claims merely

because too many have valid claims.”); see also Singh v. I.N.S.,

94 F.3d 1353, 1359 (9th Cir. 1996) (“[W]e reject the notion

that an applicant is ineligible for asylum merely because all

members of a persecuted group might be eligible for asylum.”).

75

Radiowala, 930 F.3d at 583.

76

Id. (quoting S.E.R.L., 894 F.3d at 551).

77

CAR 3.

78

CAR 12.

79

Petitioner Br. 23–24.

25

cultural, political, and other minorities.”80 While being part of

a minority may be immutable, society may not recognize a

group encompassing all members of all minorities in Somalia

as socially distinct.

The term “repatriated” is similarly overbroad, as it can

encompass individuals absent for days or decades. While

evidence in the record demonstrates that a returnee who has

been absent from Somalia for years may be distinguishable,

there is no evidence that a returnee absent for a few weeks or

months might be. Thus, we agree that, as framed, this group is

“insufficiently socially distinct” as well.81 Therefore, we reject

Herrow’s argument that “Repatriated Minority Somalis” are a

cognizable social group and deny Herrow’s petition for review

of his withholding of removal claim.

IV. CONCLUSION

We will grant in part Herrow’s petition for review and

remand this case to the BIA for consideration of all evidence,

including that evidence which is favorable to Herrow, on his

CAT claim. For the reasons stated above, we will affirm the

dismissal of the remaining claims.

80

CAR 96.

81

CAR 97.

26

Herrow v. Attorney General, No. 22-1854

PHIPPS, Circuit Judge, dissenting in part.

The Majority sets aside the final order of removal and

remands this case to the agency on the grounds that the agency

failed to consider evidence favorable to Herrow’s CAT claim.

While I agree with the other holding in the case (that

repatriated minority Somalis do not constitute a cognizable

particular social group), I respectfully disagree with the

conclusion that the CAT claim merits reconsideration by the

agency.

From a casual reading of the Majority Opinion – one

unaccompanied by the administrative record – one could leave

with the impression that the agency ignored an almanac of facts

regarding the torture of repatriated Somalis. That takeaway

would be inaccurate because the facts that the agency did not

specifically address come from a single source: an article

published in the Georgetown Immigration Law Journal. See

Daniel J. Van Lehman & Estelle M. McKee, Removals to

Somalia in Light of the Convention Against Torture: Recent

Evidence from Somali Bantu Deportees, 33 Geo. Immigr. L.J.

357 (2019) (AR484–524). And there was readily apparent

good cause for not crediting those facts.

The article itself lacks many of the traditional indicia of

reliability. Nothing in the administrative record indicates that

the journal is peer-reviewed; rather, it appears to be a student-

run journal. Moreover, the article itself is an advocacy piece.

Its two authors are paid to advocate for refugees – one leads a

law school immigration clinic and the other testifies as an

expert at removal hearings. See id. at 357 n.* (AR484) (stating

that one author “teaches the Asylum and Convention Against

Torture Appellate Clinic” at Cornell Law School, and that the

other “has testified as an expert witness in criminal and

immigration court for Somali minorities, including the Somali

Bantu”). In addition, the article expressly announces its

purpose of demonstrating that “current United States law does

not properly consider the reality of torture and persecution

suffered by the Somali Bantu.” Id. at 360 (AR487). And lest

its character as an advocacy piece be overlooked, the article

closes with an appendix containing an open letter addressed to

the United Nations High Commissioner for Refugees urging

that Somali Bantus be “not returned to Somalia against their

will.” Id. at 396 (AR523).

The article’s methodology further undercuts its reliability.

The article makes broad conclusions about the conditions in a

country of over 20 million people based on interviews of

18 Somalis. That sample size is not only statistically

insignificant, but also the methodology is subjective (since it

bakes in the biases of the participants) and does not lend itself

to independent verification (since it neither includes a list of

interview questions nor offers to make available the

responses). Moreover, the article suffers from selection bias

because a criterion for inviting the participants was the ease of

contacting them, and the article does nothing to rule out the

possibility that the participants may have been former clients

of the clinic administered by one of the authors or persons in

whose favor the other author testified.

Finally, even on its own terms, the article is not reliable.

The demographic data it provides about the participants

indicates that they have differences in birth country, native

language, clan and subclan membership, and religious practice.

Yet the article makes no effort to determine the role that those

other variables may have on the likelihood of torture in

Somalia. Without any attempt at a multivariate analysis, the

article does nothing to rule out another variable or combination

of variables as the cause for the experiences reported by the

participants.

In light of the patent concerns about the article’s reliability,

it is inappropriate, in my view, to set aside the removal order

based on the agency’s failure to explain its decision not to

credit the findings in the article. While an agency has an

2

obligation to consider the whole record, including evidence

that detracts from its position, see Universal Camera Corp. v.

N.L.R.B., 340 U.S. 474, 488–89 (1951), an agency need not

explain its decision not to credit evidence that has minimal

persuasive value, see Huang v. Att’y Gen., 620 F.3d 372, 388

(3d Cir. 2010). Here, the obvious reliability concerns with the

article make it unnecessary for the agency to specifically

explain why it was not persuaded by the article’s factual

findings. And even if the agency did have to explain why it

gave no meaningful weight to the article’s fact-finding, the

omission of that explanation would be nothing more than

harmless error since, even if considered, the article would not

compel a reasonable adjudicator to reach a contrary

conclusion. See 8 U.S.C. § 1252(b)(4)(B); Garland v. Ming

Dai, 593 U.S. 357, 373 (2021) (“[T]he court of appeals must

accept the agency’s findings of fact as ‘conclusive unless any

reasonable adjudicator would be compelled to conclude to the

contrary.’” (quoting 8 U.S.C. § 1252(b)(4)(B))).

For these reasons, I respectfully dissent in part, and I would

deny the petition.

3