Marta Tista-Ruiz de Ajualip v. Merrick B. Garland

114 F.4th 487

• Court: Court of Appeals for the Sixth Circuit
• Decided: August 9, 2024
• Docket: 23-3274
• Status: Published

Opinion

RECOMMENDED FOR PUBLICATION Pursuant to Sixth Circuit I.O.P. 32.1(b) File Name: 24a0170p.06

UNITED STATES COURT OF APPEALS FOR THE SIXTH CIRCUIT

┐ MARTA LIDIA TISTA-RUIZ DE AJUALIP; GLENDY │ JOMARA AJUALIP-TISTA; KENDEL AMISAEL AJUALIP- │ TISTA; OSWIN ALEXIS AUGUSTO AJUALIP-TISTA; │ JAMILTON OSVIEL TISTA-AJUALIP, │ No. 23-3274 Petitioners, > │ │ v. │ │ MERRICK B. GARLAND, Attorney General, │ Respondent. │ ┘

On Petition for Review from the Board of Immigration Appeals. Nos. A 088 925 800; A 208 893 829; A 208 893 830; A 208 893 831; A 208 893 832.

Decided and Filed: August 9, 2024

Before: COLE, GILMAN, and LARSEN, Circuit Judges. _________________

COUNSEL

ON BRIEF: Stephen C. Knight, BAYDOUN & KNIGHT, PLLC, Brentwood, Tennessee, for Petitioners. Roberta O. Roberts, UNITED STATES DEPARTMENT OF JUSTICE, Washington, D.C., for Respondent.

COLE, J., delivered the opinion of the court in which GILMAN, J., joined. LARSEN, J. (pp. 28–43), delivered a separate dissenting opinion. _________________

OPINION _________________

COLE, Circuit Judge. Applicant Marta Lidia Tista-Ruiz De Ajualip is a native citizen of Guatemala who applied for asylum and withholding of removal for herself and on behalf of her No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 2

three children and grandson,1 all of whom entered the United States on March 5, 2016.2 Marta’s family left Guatemala because they were afraid that Marta’s son-in-law, Marvin, would act on his threats to kill them.

After initiating removal proceedings and holding an administrative hearing in 2016, an immigration judge (IJ) found that Marta and her family were subjected to persecution as defined in the Immigration and Nationality Act (INA), but the IJ denied Marta’s asylum and withholding claims for other reasons. In making her asylum determination, the IJ relied on previously applicable immigration precedent to declare that victims of domestic violence are not protected by the asylum laws of the United States. And when analyzing Marta’s withholding claim, the IJ applied the wrong legal standard and premised her denial on a separate, inapplicable immigration statute.

In March 2023, the Board of Immigration Appeals (Board) issued an opinion affirming both points. The Board issued a separate opinion on the asylum claim that acknowledged a significant change in precedent since the IJ’s decision, but the Board affirmed instead of remanding for further review. The Board summarily affirmed the denial of Marta’s withholding claim. Because the Board’s denial of asylum and withholding of removal is inconsistent with this court’s precedent and other immigration authority, we grant Marta’s petition for review and remand for further proceedings.

I.

A.

Marta and her family lived in a remote village in the mountains of Guatemala named Volcancillo. In her testimony, all deemed credible by the IJ, Marta described her village as a

1 The IJ’s and the Board’s final decisions omitted the name of minor child, Jamilton Osviel Tista-Ajualip (A208-893-832); (AR 247). These omissions seem to be clerical errors, and we reinsert his name into the case caption as a result. 2 Marta is the lead applicant in this case, and her three children and grandson act as derivative applicants for purposes of Marta’s asylum application only, and not for the Board’s denial of withholding of removal. (AR 003, BIA Order.) Unless otherwise noted, references to Marta in the singular also encompass the arguments made by her children in their asylum applications. No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 3

“town in the mountains” where there were only “five houses [] and . . . the other closest houses [were] about an hour walking distance.” (AR 275.) The village did not have any local police or any mechanism to “protect women [from] violence.” (Id.)

Marta’s eldest daughter, Glendy, entered into a domestic relationship with Marvin, an adult male, when she was fifteen years old. Marvin moved in with the family soon thereafter, but started abusing Glendy when he learned that she was pregnant. Glendy initially refused Marvin’s requests to have an abortion, so Marvin resorted to physical abuse in order to force her to abort the baby. As just one example of the reoccurring abuse, while Marta would tend to livestock outside of the house, Marvin would lock Glendy in a bedroom and beat her repeatedly.

The severity of Marvin’s abuse quickly escalated and spread to the rest of Marta’s immediate family members after the baby was born. Marvin was dependent on drugs and alcohol, was not working, and would leave for days at a time, but he always returned to abuse Glendy further. The family also heard from others that Marvin “would go and steal” and that he was “associated with gangs[.]” (Id. at 266−67.)

Marta eventually confronted Marvin and told him to leave because of the frequent abuse and his inability to support Glendy’s baby. This angered Marvin further, prompting him to take the family’s phones, sharpen a machete in front them, and explain that he planned on killing them with it. Marvin eventually left, but he returned approximately three months later to threaten the family again. Specifically, Marvin stated, “he had gotten money together and that he came back so that [Glendy] would give him his son,” noting that he could find someone to kill the entire family “for 400 quetzales,” the Guatemalan currency. (Id.)

After this incident, Marta took her family on a day-long hike to the nearest city to file a police report because there were no local authorities in her village. Glendy gave a statement outlining Marvin’s abuse and his most recent threats, and she “ask[ed] for protection for [her] family.” (AR 371.) Instead of arresting Marvin or offering protection, the police required that Glendy undergo a psychological exam before implementing any “security measures.” (AR 115.) Taking the exam would “require [Marta’s family] to risk the one-day journey back to their home [] through the mountainous region of Guatemala; then to make the one-day hike to return to the No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 4

city to undergo the ‘psychological exam,’ and then make the one-day trek back home”—a three- day endeavor. (Appellant Br. 17.)

Marta, aware of the dangers these journeys presented, as well as the reality that it is common practice for the Guatemalan Police Force to ignore or fail to protect victims of domestic violence, did not believe that her family would survive Marvin’s threats if they stayed in Guatemala long enough to follow through with this process. In fear for their lives, they fled to the United States three days later. Marta also testified that she fears going back to Guatemala, that there is nowhere “safe for [her] to go if [she] went back there,” and that she does not “have a place or any family there.” (AR 268−69.)

B.

The Department of Homeland Security initiated removal proceedings against Marta and her family in 2016, the same year they entered the U.S. The IJ held a hearing on June 27, 2019, where she heard testimony from both Marta and Glendy. (Id. at 208−46.) At the beginning of the hearing, Marta’s counsel stated that Marta’s proposed particular social groups (PSGs) were “victims of domestic violence” and “family members of victims of domestic violence.” The IJ responded by questioning the cognizability of both PSGs, asking “[h]ow is that going to pass given the Attorney General’s decision in Matter of A-B-?”, which was applicable immigration authority during Marta’s hearing in 2019. Matter of A-B-, 27 I&N Dec. 316, 320 (2018) (hereinafter “A-B-I”) (since overruled). In relevant part, A-B-I stated that “[g]enerally, claims by aliens pertaining to domestic violence or gang violence perpetrated by non-governmental actors will not qualify for asylum.” Id.

The IJ’s reliance on A-B-I, now overruled, colored her actions and decisions throughout the rest of the proceedings. During the hearing, Marta’s counsel tried to ask questions about Marta’s personal characteristics, as well as questions regarding Marta and her family members’ employment status, marital status, and education level. But the IJ precluded Marta from developing this testimony by interrupting Marta’s counsel on four separate occasions: instructing counsel to “skip the biometric information,” asking “[w]hat’s the relevance of that to [Marta’s No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 5

asylum claim?”, and ordering counsel to “start with [why Marta was] afraid to return to her home country.” (AR 262−264, 269.)

The IJ also conducted her own, very brief, examination of Marta during the hearing. (AR 273−74.) The IJ first asked Marta whether “[she] or any family member would be seen as distinct or in any way stand out in Guatemalan society,” but Marta—apparently misunderstanding the IJ’s legal question—instead answered by explaining that her family would be susceptible to extortion if they returned to Guatemala. (Id.) Immediately after Marta’s answer, the IJ again asked Marta whether she “would be seen as [] distinct” or whether “[she] or any family member [is] a part of any group within Guatemala.” (Id.) Marta responded “[n]o” to both questions. (Id.)

After testimony concluded, Marta’s counsel again emphasized that Marta’s family lived in a small town where “[t]here’s not police, and [t]here’s not any kind of protection for women that are being abused.” (AR 281.) Marta’s counsel also submitted a State Department report titled “Guatemala 2015 Crime and Safety Report,” which explained that the Guatemalan police force “lacks personnel and training . . . [and] is significantly under-trained and under-funded.” (Id. at 391.) The report further explained that “[m]ore often than not a police investigation fails to result in arrest, much less a conviction,” and, specifically, “[s]upport for victims of sexual assault is lacking outside of major cities” and “there are not enough trained personnel who can help victims either in the capital or outlying areas.” (Id. at 391, 404.)

At the end of the hearing, the IJ issued an oral decision that summarily denied Marta’s claims. The IJ stated that “although the Court believes what you told it today, the asylum laws of the United States do not protect against general domestic violence claims from another country, and therefore, the Court has been unable to grant your applications today.” (Id. at 282.) The IJ found that the harm Marta and her family suffered amounted to persecution. But the IJ concluded that Marta and her family’s proposed PSGs—“victims of domestic violence” and “family members of victims of domestic violence”—were circular and therefore not sufficiently defined under A-B-I. The IJ also concluded that Marta could not show that the Guatemalan government was unwilling or unable to protect them from private persecution because Marta and Glendy chose not to follow through with the psychological exam requested by the Guatemalan No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 6

authorities. (Id. at 210−13.) These two conclusions, the IJ stated, disqualified Marta and her family from attaining asylum under the INA.

The IJ relied almost exclusively on A-B-I for both of her conclusions. See 27 I&N Dec. at 320 (stating that “claims by aliens pertaining to domestic violence . . . perpetrated by nongovernmental actors will not qualify for asylum”); (see also AR 210−12, 282) (IJ stating that “[t]he Attorney General noted in [A-B-I] that the fact that the local police have not acted on a particular report of an individual crime does not necessarily mean the government is unwilling or unable to control crime.”).

The IJ also concluded that Marta did not qualify for withholding of removal under Section 241(b)(3) of the INA. In making this determination, the IJ stated that because Marta could not meet the burden for asylum, she also could not “meet the higher burden required for withholding of removal under [the INA],” and the evidence in the record “without more [was] insufficient to carry the high burden that [Marta] would be specifically targeted for torture if she returned to Guatemala.” (AR 213.)

However, after the IJ denied relief to Marta—but before the Board could review the IJ’s denial and issue its own decision—Attorney General Garland issued an immigration decision in 2021 that overruled A-B-I. See Matter of A-B-, 28 I&N Dec. 307 (A.G. 2021) (hereinafter “A-B- III”). A-B-III vacated A-B-I in its entirety, stating that A-B-I’s “broad language could be read to create a strong presumption against asylum claims based on private conduct” and “threatens to create confusion and discourage careful case-by-case adjudication of asylum claims.” Id. at 308−09. Moving forward, Garland instructed immigration courts to follow “pre-A-B-I precedent, including Matter of A-R-C-G-.” Id. at 309.

A-R-C-G-, unlike A-B-I, recognizes that victims of domestic violence can qualify for asylum relief. See Matter of A-R-C-G-, 26 I&N Dec. 388, 394. A-R-C-G- also clarifies that asylum claims premised on domestic violence contain unique issues that require a careful evaluation of the record:

[C]ases arising in the context of domestic violence generally involve unique and discrete issues not present in other particular social group determinations, which extends to the matter of social distinction. [And] even within the domestic No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 7

violence context, the issue of social distinction will depend on the facts and evidence in each individual case, including documented country conditions; law enforcement statistics . . .; the [applicant’s] past experiences; and other reliable and credible sources of information.

Id. Due to the unique issues in domestic violence cases, A-R-C-G- emphasizes that the PSG determination, specifically “[w]hen evaluating the issue of social distinction,” requires immigration courts to look to the evidence to determine “whether a society . . . makes meaningful distinctions based on [] common immutable characteristics.” Id.

The Board affirmed the IJ’s decision in March 2023. The Board acknowledged that A-B-III overruled A-B-I, but nonetheless agreed with the IJ because (1) even if A-B-I was overruled, “victims of domestic violence” does not qualify as a PSG under Sixth Circuit precedent because the group is “circular[ly]” defined by the harm,” and (2) the IJ’s conclusion that Marta did not meet her burden of showing that “the Guatemalan government would be unwilling or unable to protect her . . . is not clearly erroneous, as [her] personal experience with the police does not suggest” otherwise. (AR 003−05.) The Board then summarily adopted and affirmed the IJ’s denial of withholding of removal.

Marta appealed the Board’s decision, requesting that we reverse the denial of her applications for asylum and withholding of removal. She argues that the Board’s conclusions are inconsistent with A-B-III and our precedent, and that there was not substantial evidence in the record supporting either conclusion.

II.

This court has jurisdiction under 8 U.S.C. § 1252 to review the Board’s final determination ordering removal. Umaña-Ramos v. Holder, 724 F.3d 667, 670 (6th Cir. 2013). We review the Board’s order if it issued its own decision, but we review the IJ’s decision to the extent that the Board adopted it. Id. Questions of law are reviewed de novo and factual findings under the substantial-evidence standard. Id.; see also Loper Bright Enter. v. Raimondo, 144 S.Ct. 2244, 2258–59, 2273 (2024) (explaining that deferential review of an agency decision is “cabined to factbound determinations” because “[c]ourts must exercise their independent No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 8

judgment in deciding whether an agency has acted within its statutory authority”) Under the substantial-evidence standard, “we uphold a [Board] determination as long as it is supported by reasonable, substantial, and probative evidence on the record considered as a whole.” Juan Antonio v. Barr, 959 F.3d 778, 788 (6th Cir. 2020) (quotation marks omitted). And “we will not reverse a factual determination . . . unless we find ‘that the evidence not only supports a contrary conclusion, but compels it.’” Ceraj v. Mukasey, 511 F.3d 583, 588 (6th Cir. 2007) (quoting Marku v. Ashcroft, 380 F.3d 982, 986 (6th Cir. 2004) (emphasis in original). Lastly, because the IJ found the parties credible, “we accept their factual statements as true.” Mandebvu v. Holder, 755 F.3d 417, 424 (6th Cir. 2014).

An asylum applicant has the burden of proving that she is a “refugee” under the INA, meaning someone “unable or unwilling to return to her home country because of past persecution or a well-founded fear of future persecution on account of race, religion, nationality, membership in a [PSG], or political opinion.” Juan Antonio, 959 F.3d at 789 (quotation marks omitted). An asylum applicant who alleges that the persecution stems from a private actor’s violence towards her must therefore “independently prove” that (1) the “violence [arises or] would arise on account of her membership in a [PSG]” and (2) “the government is unable or unwilling to control [the] private violence.” Ortiz v. Garland, 6 F.4th 685, 686, 691 (6th Cir. 2021) (the “state-action element”).3

In Marta’s case, the Board concluded that her first proposed PSG, “victims of domestic violence,” was impermissibly circular without reviewing any of the evidence in the record. The Board’s analysis, and subsequent conclusion, were therefore deficient. The Board issued its own decision on the PSG issue, so we review its decision as the “final agency determination,” but we still review the IJ’s decision to the extent that the Board adopted it. Khalili v. Holder, 557 F.3d 429, 435 (6th Cir. 2009). And we review questions of law de novo. Id.

3 There is no dispute in this case that Marvin’s violent acts and later threats towards Glendy and Marta’s family constitute persecution. No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 9

A cognizable PSG is “a group that shares a common, immutable and fundamental characteristic that either cannot be changed or should not be required to [change] because it is fundamental to the members’ individual identities or consciences.” Juan Antonio, 959 F.3d at 789 (cleaned up). Moreover, the PSG must be both particular and socially visible. Id. (citation omitted); see also A-R-C-G-, 26 I&N Dec. at 393 (“[T]here is some degree of overlap between the particularity and social distinction requirements because both take societal context into account.”) (citation omitted).

Further, a PSG cannot be “circular,” which means that the “group must share a narrowing characteristic other than [its members’] risk of being persecuted” to be cognizable. Juan Antonio, 959 F.3d at 789 (internal quotation omitted). And we have recognized that “the [Board] and IJ have an independent role to assess the plausibility of the application on the record as a whole.” See Zometa-Orellana v. Garland, 19 F.4th 970, 979 (6th Cir. 2021); Juan Antonio, 959 F.3d at 788, 791−92; see also Turcio-Flores, 67 F.4th at 356 (explaining that a “group [can be] vulnerable to persecution in part due to its social distinction”). This factfinding obligation is in line with current immigration authority and our sister circuits, which have recognized that the Board cannot “conclude that [a PSG] is defined by the harm or potential harm inflicted merely by the language used rather than determining what underlying characteristics account for the fear and vulnerability.” Grace v. Barr, 965 F.3d 883, 903−04 (D.C. Cir. 2020) (explaining that immigration authority and “[g]uidance make clear that asylum officers must analyze each case on its own merits in the context of the society where the claim arises”) (emphasis added); see also Cece v. Holder, 733 F.3d 662, 672 (7th Cir. 2013) (en banc); A-R-C-G, 26 I&N (“A determination of this issue will be dependent upon the facts and evidence in a case.”).

In its decision, the Board acknowledged Marta’s argument that A-B-I was overruled, but it still agreed with the IJ’s conclusion that “victims of domestic violence” was a circular PSG because it is “defined solely by being victimized by domestic violence.” (AR 004.) The Board is correct in its understanding that we are normally precluded from recognizing as cognizable a circularly proposed PSG. But the Board incorrectly applied this principle to Marta’s asylum application by (1) not remanding for further review after the IJ relied on outdated precedent, A- B-I, to ignore pertinent evidence in the record and impede Marta’s counsel from further No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 10

developing relevant testimony, and (2) failing to consider the record as a whole when making its PSG determination.

1.

We must first return to the record, and the timeline of relevant immigration authority, to adequately address the many mistakes made in this case. At the outset of Marta’s hearing, the IJ made clear that she did not believe Marta’s proposed PSGs were cognizable under A-B-I, which previously held that, generally, “claims by aliens pertaining to domestic violence . . . perpetrated by nongovernmental actors will not qualify for asylum.” See 27 I&N Dec. at 320. The IJ’s significant reliance on A-B-I then colored how she conducted the remainder of Marta’s hearing.

Marta’s counsel, for example, tried to ask general questions about Marta’s personal characteristics and circumstances, and specific questions regarding Marta and her family members’ employment status, marital status, and education level. But the IJ prevented the development of this testimony by questioning its relevance and interrupting Marta’s counsel on four separate occasions. Then, in issuing her oral decision at the hearing’s conclusion, the IJ stated broadly that “the asylum laws of the United States do not protect against general domestic violence claims from another country.” (See also AR 282) (IJ stating that “[t]he proposed [PSGs] are delineations of the social group discussed in [A-B-I], and the Court finds that such a proposed group is defined solely by the persecution of its members, rendering it a circular social group”).

After the IJ’s decision, however, Garland issued A-B-III, which overruled A-B-I and left the door open for “asylum claims based on private conduct,” such as claims made by victims of domestic violence. See A-B-III, 28 I&N Dec. at 307. A-B-III explicitly commanded that “immigration judges and the Board should no longer follow A-B-I . . . when adjudicating pending or future cases,” and they should instead “follow pre-A-B-I precedent, including [A-R-C-G-].” Id. at 309. Collectively, A-B-III, A-R-C-G-, and our own precedent make clear that the Board should have remanded Marta’s case back to the IJ for further factual review.

a. The Board Should Have Remanded for Further Factual Development. No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 11

As a preliminary matter, the IJ’s reliance on A-B-I to broadly proclaim that victims of domestic violence are generally barred from asylum protection is the exact type of “confusion” that A-B-III sought to avoid in “pending or future cases.” See id. at 309 (emphasis added) (collecting cases to criticize immigration decisions that “stat[e] without qualification that [A-B-I] declined to extend asylum and withholding of removal relief to victims of private criminal activity”) (citation omitted). To clarify, in issuing A-B-III, Attorney General Garland reiterated that “A-B-I threaten[ed] to create confusion and discourage careful case-by-case adjudication of asylum claims.” Id. No such careful adjudication was afforded to Marta, whose counsel was prevented from developing testimony that was relevant to her asylum claim.

We need only look at A-R-C-G- to see that the IJ, if she were not permitted to rely on A-B-I to categorically bar Marta’s claim, would have been required to consider the testimony Marta’s counsel tried to develop. First, A-R-C-G- mandates that “any claim regarding the existence of a particular social group must be evaluated in the context of the evidence presented regarding the particular circumstances in the country in question.” 26 I&N Dec. at 392 (emphasis added). Second, in discussing whether there is a common immutable characteristic, A-R-C-G- states that “[a] determination of this issue will be dependent upon the particular facts and evidence in a case,” and explains that “adjudicators must consider a respondent’s own experiences, as well as more objective evidence, such as background country information.” Id. at 393 (emphasis added) (stating, for example, that “marital status can be an immutable characteristic where the individual is unable to leave the relationship”); see also Juan Antonio, 959 F.3d 791 (same). Finally, when discussing social distinction, A-R-C-G- states that “within the domestic violence context, the issue of social distinction will depend on the facts and evidence in each individual case,” including, amongst other things, “the respondent’s past experiences.” A-R-C-G-, 26 I&N Dec. at 394−95 (“We point out that cases of domestic violence generally involve unique and discrete issues not present in other particular social group determinations.”).

Yet notably absent from the IJ’s decision are the different bodies of evidence that A-B-III and A-R-C-G- state are highly relevant to the PSG determination. Compare (AR 210) (IJ stating Marta’s PSG is not “particular because it is defined by the member’s vulnerability to private No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 12

activity”), with A-B-III, 28 I&N Dec. at 309 (explaining that A-B-I incorrectly “create[d] a strong presumption against asylum claims based on private conduct”), and A-R-C-G-, 26 I&N Dec. at 392 (stating the social-distinction determination requires us to review evidence such as whether (1) “the society in question [offers] protection to victims of domestic violence,” (2) “the country has criminal laws designed to protect domestic abuse victims,” (3) “those laws are effectively enforced,” and (4) “other sociopolitical factors”). Here, the IJ’s reluctance to cite or discuss this type of evidence stems from her reliance on A-B-I to prevent Marta from developing relevant testimony and to also disregard the documentary evidence that Marta submitted.

To be fair, because A-B-I was valid immigration precedent at the time her decision was issued, we cannot fault the IJ for her significant reliance on that case. The Board, however, was wrong to affirm the IJ’s decision despite acknowledging the significant change in precedent that came with A-B-III. Here, “[i]n essence, the [IJ], relying on Matter of A-B-, disposed of the element on the grounds that [PSGs] presumptively cannot be defined by domestic violence by non-government actors.” See Zometa-Orellana v. Garland, 19 F.4th 970, 978 (6th Cir. 2021). Next, the Board—by ignoring the underdeveloped record and an IJ decision in direct conflict with A-B-III and A-R-C-G-—essentially rubber stamped the IJ’s denial of asylum by citing to Sixth Circuit precedent on circularity. (AR 004). But “[i]n cases where a decision on which the IJ or BIA relied to make a determination was vacated or abrogated, we have determined that this change in the law ‘counsels remand.’” Id. at 979 (quoting Juan Antonio v. Barr, 959 F.3d 778, 790 n.3 (6th Cir. 2020)). The Board cannot simply ignore the change in law here by claiming circularity in a perfunctory manner. Moreover, notwithstanding its procedural mistake, the Board’s application of the circularity rule was also incorrect.

b. The Board Misapplied the Circularity Rule.

We have stated that “the [Board] and IJ have an independent role to assess the plausibility of the application on the record as a whole.” See Zometa-Orellana, 19 F.4th at 979 (emphasis added); see also Juan Antonio, 959 F.3d at 784, 788−89, 791 (holding that “the Board’s conclusion [was] not supported by substantial evidence on the record considered as a whole” and that the PSG determination “will be dependent upon the particular facts and evidence in a case”) (quoting A-R-C-G-, 26 I&N Dec. at 392−93). This factfinding duty is in line with No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 13

settled immigration authority, A-B-III, which requires “careful [and] case-by-case adjudication of asylum claims.” See 28 I&N Dec. at 307. And it is further clarified by our precedent, which states that when assessing whether the PSG is “supported by the record in [a] case, we emphasize that the IJ and BIA have certain obligations under international law to extend refuge to those who qualify for such relief and bear the responsibility for assuring that refugee protection is provided where such protection is warranted by the circumstances of an asylum applicant’s claim.” See Zometa-Orellana, 19 F.4th at 979 (cleaned up) (emphasis added).

These obligations remain at the forefront when making the circularity determination; this is because “whether a group exists independently of the harm alleged is not always so apparent,” and thus “whether a given group is circular depends on the facts of a particular case.” See Grace, 965 F.3d 883 at 903−04 (collecting cases from other circuits to discuss the “circularity rule” and stating these “examples demonstrate it is not fair to conclude that the group is defined by the harm or potential harm inflicted merely by the language used rather than determining what underlying characteristics account for the fear and vulnerability”) (cleaned up); see also Cece, 733 F.3d at 672 (same). Indeed, our own precedent, the caselaw in other circuits, and current immigration authority support this principle. See, e.g., Espinoza-Ochoa v. Garland, 89 F.4th 222, 223 2023 (1st Cir. 2023) (stating “[o]ur sister circuits also have recognized . . . that it is not fair to evaluate a PSG based merely on a cursory assessment of the words used”); Turcios-Flores v. Garland, 67 F.4th 347, 356 (6th Cir. 2023) (rejecting circularity as applied and stating that “th[e] group is not circularly defined by the fact that it suffers persecution. Rather, th[e] group is vulnerable in part due to its social distinction”) (cleaned up); A-R-C-G-, 26 I&N Dec. at 394 (stating “within the domestic violence context, the issue of social distinction will depend on the facts and evidence in each individual case”). Simply put, if the Board wishes to deny an asylum claim on the basis of circularity, then it must review the record and conduct the proper analysis. See Espinoza-Ochoa, 89 F. 4th at 223 (stating a “review of a PSG for legal validity must be based on a substantive analysis, not a superficial ‘quick look’”).

Let us return to Marta’s case with this authority in mind. Here, the Board incorrectly concluded that Marta’s proposed PSG was circular without discussing, or even citing to, the many pieces of evidence in the record suggesting otherwise. (AR 004). And it is apparent that No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 14

the Board’s circularity determination is even more flawed upon learning that, as explained above, the record is incomplete because the IJ prevented Marta from developing potentially relevant testimony. (AR 262−264, 269.)

Reviewing the record as a whole, the particular facts of Marta’s case show that her proposed PSG was not circular because her application reflects additional immutable and narrowing characteristics separate and apart from her suffering of domestic violence. Credible testimony and documentary evidence in the record show, for example, that Marta’s family (a) lived in a remote village in the mountains, where (b) there were no local authorities or means to protect women from domestic violence, and (c) they had no apparent ability to escape their abuser, who was allegedly associated with gangs, while remaining in Guatemala. (AR 265, 267−68, 272, 274− 75, 281); see also Grace, 965 F.3d at 905 (reiterating that “Guatemalan women unable to leave their relationships . . . is not [a] categorically barred [PSG] . . . and that its validity would turn on the specific factual circumstances of an applicant’s claim”) (quotation marks omitted).

The Board was required to at least consider some of this evidence in making its circularity determination. See Turcios-Flores, 67 F.4th at 356 (rejecting the IJ’s “one sentence” conclusion regarding social distinction to state that “the record compels a different conclusion”); see also A-R-C-G-, 26 I&N Dec. at 394 (summarizing the evidence that “adjudicators must consider,” including evidence such as “a respondent’s own experiences,” “background country information,” and “other sociopolitical factors”). In fact, the Board itself recognized the substantial overlap between “victims of domestic violence” and “married women who are unable to leave a domestic relationship” when it stated in A-R-C-G- that:

When evaluating the issue of social distinction, we look to the evidence to determine whether a society, such as Guatemalan society in this case, makes meaningful distinctions based on the common immutable characteristics of being a married woman in a domestic relationship that she cannot leave. Such evidence would include whether the society in question recognizes the need to offer protection to victims of domestic violence, including whether the country has criminal laws designed to protect domestic abuse victims, whether those laws are effectively enforced, and other sociopolitical factors. No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 15

26 I&N Dec. at 394 (emphasis added). These types of claims—from both a commonsense perspective, and in the Board’s view—are inextricably linked.

The dissent contends that we “faul[t] the BIA for not scouring the record on its own” in performing the PSG analysis. Dis. Op. at 1. First, all of the evidence referenced above is reflected in Marta’s testimony, which spans thirteen pages and is also summarized in one paragraph by Marta’s counsel at the hearing’s conclusion. (AR 262− 275, 282) (reflecting Marta’s testimony on direct, cross, and redirect, and counsel’s summary thereto). The same transcript includes the IJ’s oral decision to deny relief—which the Board undoubtedly had to review to make its own determination—and is just one page after counsel’s summary. (AR 282.) Second, it is the Board that chose, in a precedential decision, to link the social-distinction analysis for “married women in a domestic relationship that [they] cannot leave” to a society’s treatment of “victims of domestic violence” and “domestic abuse victims.” A-R-C-G-, 26 I&N Dec. at 394. And as previously mentioned, the Board also “point[ed] out that cases arising in the context of domestic violence generally involve unique and discrete issues not present in other particular social group determinations.” Id.

A-R-C-G- was good law when the Board adjudicated Marta’s appeal, and Marta expressly argued that she qualified for asylum pursuant to A-R-C-G- in her brief to the Board. The Board should have addressed whether the record as a whole supported Marta’s assertion that she was a member of the PSG at issue in this case. Yet, when discussing “the proposed group of ‘victims of domestic violence,’” the Board did not mention or discuss any evidence in the record whatsoever in hastily concluding that Marta’s PSG “is defined solely by being victimized by domestic violence and therefore is not cognizable.” (AR 004). This complete disregard for the record, especially given the change in immigration authority, is wholly inadequate under A-R-C-G- and our own precedent.

Finally, although we do not make the cognizability determination here, it is worth noting that our court has previously found proposed PSGs with records very similar to Marta’s cognizable. See, e.g., Juan Antonio, 959 F.3d at 792 (“The Board has found that married women in Guatemala who are unable to leave their relationship can constitute a cognizable particular social group.”) (quotation marks omitted); Turcios-Flores, 67 F.4th at 356 (holding that “single No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 16

mothers living without male protections . . . is a socially defined group” because they “share narrowing characteristics . . . [and] [i]t just so happens that these characteristics also make them vulnerable to [persecution]”) (emphasis in original); cf. Zometa-Orellana, 19 F.4th at 978, 979 (remanding to determine whether “El Salvadorian women of childbearing age in domestic partnerships” is a cognizable PSG on remand).4

2.

The Board also erred in concluding that Marta “waived the issue” regarding her second proposed PSG, consisting of “family members of victims of domestic violence.” (AR 3−4). In contrast to its analysis of Marta’s first proposed PSG, “victims of domestic violence,” the Board recognized that the circularity analysis as applied to “family members of victims of domestic violence” was more complicated. (Id.) And the Board expressly stated that “[the second proposed PSG] may share a narrowing characteristic other than their risk of being persecuted.” (Id. at 4) Rather than addressing the issue, however, the Board concluded that Marta waived her argument because she allegedly “[did] not challenge the [IJ’s] determination” that her proposed PSGs were not “socially distinct within Guatemalan society, particularly in light of [Marta’s] testimony that she did not believe that she or her family . . . belong to any particular group within Guatemala.” (Id. at 4−5). The Board’s conclusion was incorrect for three separate reasons.

First, we are hesitant to construe Marta’s brief responses to the IJ’s legal questions as a waiver of this argument. For context, in her brief examination of Marta, the IJ asked in quick succession whether “[Marta] or any family member would be seen as distinct or in any way stand out in Guatemalan society,” or whether “[Marta] or any family member [is] a part of any group within Guatemala.” (AR 273−74.) Marta ultimately responded “[n]o” to both questions. (Id.)

4 To the extent that the government or dissent argues that Marta’s proposed PSGs are unexhausted arguments that were not raised before the Board, this argument is unpersuasive because it (1) misreads Marta’s argument and (2) hinges on an incorrect reading of our precedent on “circularity.” Contrary to what the government contends, it is not our understanding that Marta now argues that she and her family “are members of a PSG consisting of indigenous Mayan women.” Rather, Marta relies upon our precedent, Juan Antonio, “to emphasize certain immutable and distinct characteristics” that are reflected in the record—such as her gender, residence in a remote village, and inability to seek protection from authorities—that she believes are comparable to other cases granting asylum post A-B-I. (Reply Br. at 1−2) (citing Juan Antonio, 959 F.3d at 792). Nonetheless, even if these arguments were not exhausted (they were), we are still permitted to review arguments, like circularity, that the Board raises independently and addresses on the merits. See Khalili v. Holder, 557 F.3d 429, 433 (6th Cir. 2009). No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 17

The IJ then relied on Marta’s responses to conclude that her PSG was not socially distinct, and this was the only piece of evidence that the IJ relied on to make the determination. (AR 210.) The Board then exclusively relied on Marta’s two, one-word responses (“no”) to agree with the IJ’s social-distinction finding, and to also conclude that Marta “waive[d] this issue.” (See AR 004−05) (Board agreeing that Marta “has not demonstrated that her proposed groups are socially distinct within Guatemalan society, particularly in light of her testimony”).

The Board also disregarded that when Marta answered the IJ’s first question, her answer demonstrated that she did not understand the question. (See AR 211, 273−74). More troubling, the IJ insisted on asking Marta—rather than her counsel—this legal question twice, even after Marta apparently did not comprehend it. See Zometa-Orellana, 19 F.4th at 979 (quoting Cantarero-Lagos v. Barr, 924 F.3d 145, 154 (5th Cir. 2019) (Dennis, J., concurring)) (“‘Someone who faces persecution on account of a protected ground is no less deserving of asylum’s protections because of her inability to exactly delineate a convoluted legal concept.’”); see also Quintero v. Garland, 988 F.3d 612, 632 (4th Cir. 2021) (stating “it [is] unreasonable and fundamentally unfair to expect pro se asylum seekers—many of whom suffer from the effects of trauma and lack literacy, English proficiency, formal education, and relevant knowledge—to even understand what a [PSG] is, let alone fully appreciate which facts may be relevant to their claims and articulate a legally cognizable group”). Our precedent states that “[a]n immigration judge has a responsibility to function as a neutral, impartial arbiter and must refrain from taking on the role of advocate for either party.” Elias v. Gonzales, 490 F.3d 444, 451 (6th Cir. 2007). At the very least, that responsibility extends to considering the record as a whole, and not relying solely on the applicant’s one-word responses to an IJ’s legal questions. See Zometa-Orellana, 19 F.4th at 979

Second, the Board’s statement regarding waiver is not an accurate reflection of the IJ’s limited social-distinction analysis. Although the IJ analyzed Marta’s first proposed PSG, the IJ did not discuss, analyze, or even allude to whether “family members of victims of domestic violence” are a socially distinct group. At most, the IJ appears to have relied on her initial conclusion that “victims of domestic violence” are not socially distinct. But Marta certainly challenged that conclusion on appeal to the Board. (See AR 032) (Marta, in relying on the No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 18

change in precedent, argues to the Board that “[i]n [A-R-C-G-], the [Board] further recognized female victims of domestic violence as a cognizable and socially distinct group of people.”) The Board might have disagreed with Marta’s argument, but its conclusion that Marta waived her second proposed PSG is undercut by both the IJ’s failure to address “family members of victims of domestic violence” in her oral decision and Marta’s articulated challenge to the only social- distinction determination that the IJ did make.

Third, A-B-III instructs here that the appropriate recourse would have been for the Board to remand to the IJ for further factfinding regarding social distinction. See 26 I&N Dec. at 309 (ordering that “[IJs] and the Board should no longer follow [A-B-I] when adjudicating pending or future cases” and should instead “follow pre-A-B-I precedent, including [A-R-C-G-]”) (emphasis added)). This is especially true for Marta’s proposed PSG of “family members of victims of domestic violence” because the Board’s decision acknowledged that this PSG “may share a narrowing characteristic other than the risk of being persecuted.” (AR 004).

The dissent highlights that the IJ’s decision references the phrase Marta’s “social group or groups,” and that the decision also cites to Marta’s responses to legal questions about her “family”; the dissent then argues that these two points “can only be read as the IJ concluding that Marta had not produced sufficient evidence of social distinction for either group.” Dis. Op. at 8−9, n.3. To be clear, we do not agree that the IJ’s fleeting references to the words “groups” and “family” are enough to satisfy the IJ’s burden of “ensur[ing] that the specific group being analyzed is included in his or her decision.” See Matter of W-Y-C- & H-O-B-, 27 I&N. Dec. 189, 191 (BIA 2018); see also Espinoza-Ochoa, 89 F. 4th at 223 (“A review of a PSG for legal validity must be based on a substantive analysis, not a superficial quick look.”) (internal quotation marks omitted).

Even accepting the dissent’s argument as true, however, it would simply mean that remand is an even more appropriate course of action. Meaning, if the IJ actually did conclude that “family members of domestic violence” was not a socially distinct PSG given the two reasons the dissent highlights, then the Board would still be required to remand because A-R-C-G- requires a more careful analysis premised on specific evidence in the record. See A-R-C-G-¸ 26 I&N Dec. at 394 (stating “within the domestic violence context, the issue of social No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 19

distinction will depend on the facts and evidence in each individual case, including documented country conditions . . .; the respondent’s past experiences; and other reliable and credible sources of information”). Again, none of this evidence was considered, or even referenced, in the IJ’s decision. Maj. Op. at 15−16. Given this deficient analysis, and the change in precedent thereafter, the Board was not permitted to dismiss this PSG by employing its (evidently incorrect) theory of waiver. Instead, the Board should have remanded to the IJ for factfinding consistent with A-B-III and A-R-C-G-’s mandates. See A-B-III, 26 I&N Dec. at 309; see also Zometa-Orellana, 19 F.4th at 979.

Accordingly, because Marta did not waive this argument, the Board must also consider whether “family members of victims of domestic violence” is a cognizable PSG on remand. See Arangure v. Whitaker, 911 F.3d 333, 347–48 (6th Cir. 2018) (holding that remand is appropriate for an issue not addressed by the Board “so that the Board can consider it in the first instance”).

3.

With regard to the PSG issue, the dissent claims that we assert a new PSG on behalf of Marta that she did not propose before the IJ or the Board. Dis. Op. at 1−2, 8−9. The dissent then references this “new” PSG to argue that our decision conflicts with immigration precedent on exhaustion. Id. at 10−12. Both arguments misread our conclusions. We address them in turn.

The dissent first claims that the majority invents a new PSG for Marta, “consisting of victims of domestic violence who: ‘(a) lived in a remote village in the mountains, where (b) there were no local authorities or means to protect women from domestic violence, and (c) . . . had no apparent ability to escape their abuser, who was allegedly associated with gangs, while remaining in Guatemala.’” Dis. Op. at 2, 11−12 (quoting Maj. Op. at 16). Respectfully, the dissent faults us for a conclusion that we did not reach.

Instead, and as explained in detail above, we vacate the Board’s decision because it failed to conduct a proper circularity analysis. The Board’s boilerplate invocation of circularity was not premised on any evidence in the record, and it was in direct conflict with our own precedent, current immigration authority, and the caselaw in other circuits addressing the circularity rule. See, e.g., Zometa-Orellana, 19 F.4th at 979; A-B-III, 28 I&N Dec. at 307; A-R-C-G-, 26 I&N at No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 20

392−93; Grace v. Barr, 965 F.3d at 904. To the extent we reference evidence in the record that the dissent claims is a “new” PSG, our references are instead examples of “narrowing characteristics separate and apart from [Marta’s] suffering of domestic violence” that the Board should have considered under the aforementioned authorities. See Maj. Op. at 16.

The dissent next argues, by relying on its own claim of a newly asserted PSG, that our decision to vacate the Board’s asylum order conflicts with immigration authority on exhaustion, specifically, “Matter of W-Y-C-.” See Dis. Op. at 6−8 (citing 27 I&N Dec. at 191). As an initial matter, because Marta did not raise—nor have we addressed—a newly asserted PSG, any argument regarding exhaustion is inapposite here. We nonetheless address the dissent’s comments regarding exhaustion and the record to provide further clarity.

Regarding exhaustion, the dissent relies on W-Y-C- and an unpublished decision to argue that Marta did not “clearly indicate the exact delineation of [the] [PSG(s)] to which she claimed to belong.” Cf. W-Y-C-, 27 I&N Dec. at 191; see also Dis. Op. at 9, 15 (citing Gonzalez- Valencia v. Barr, 764 F. App’x 510, 512 (6th Cir. 2019)). The dissent ignores, however, that W- Y-C- stresses that its decision “is underscored by the inherently factual nature of the social group analysis,” and that “[a] determination whether a social group is cognizable is a fact-based inquiry made on a case-by-case basis.” Id. (emphasis added) (citation omitted). Given this factfinding obligation, the language in W-Y-C- itself even emphasizes that “the [IJ] should seek clarification . . . [i]f an applicant is not clear as to the exact delineation of the proposed [PSG].” See W-Y-C-, 27 I&N Dec. at 191.

This emphasis on an IJ’s duty to adjudicate claims through careful, case-by-case factfinding is directly in line with the other immigration decisions discussed above, which were issued (or reinstated) after W-Y-C-. See A-B-III, 28 I&N Dec. at 309 (emphasizing the necessity for “careful case-by-case adjudication of asylum claims”); see also A-R-C-G, 27 I&N Dec. at 392−96) (discussing the different types of evidence that IJs should consider throughout the PSG determination). And even in Gonzales-Valencia, the unpublished decision that the dissent relies on, the panel made certain to note that “Matter of W-Y-C- correctly held that an IJ must make factual findings to decide an applicant’s claim for relief based on membership in a [PSG].” 764 F. App’x 510, 512 (6th Cir. 2019) (emphasis added) (unpublished). No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 21

Thus, considering W-Y-C- within the larger context of the immigration authority addressed above, we do not, as the dissent argues, announce any new rule. To the contrary, by vacating the two immigration decisions before us that wholly disregard the record, we instead ensure that the Board and the IJ can review the record in line with current immigration authority to make the requisite factual findings. See W-Y-C-, 27 I&N Dec. at 191 (stating that “the social group analysis” is “inherently factual [in] nature”). Further, this factfinding obligation, under our own precedent, is nothing new. See Zometa-Orellana, 19 F.4th at 979 (“[T]he Board and IJ have an independent role to assess the plausibility on the record as a whole.”); see also Juan Antonio, 959 F.3d at 791−92 (holding that “the record considered as a whole, which includes both [the applicant’s] testimony and the broader societal context,” compels a different conclusion); see also Turcios-Flores, 67 F.4th at 356−357 (rejecting the Board’s social- distinction and circularity determinations, and explaining that “[t]he record shows that [applicant] was vulnerable to persecution because she is a part of a socially defined group.”)

***

Despite the dissent’s assertions to the contrary, we do not “propos[e] a new social group for Marta,” nor do we “impos[e] a new procedural rule on the agency.” See Dis. Op. at 17. For sake of clarity, we reiterate here that we do not make the cognizability determination in this case. Rather, to summarize, our decision to remand Marta’s application directly responds to: (1) the IJ’s near-exclusive reliance on A-B-I to broadly proclaim that victims of domestic violence do not qualify for asylum protection, and to ignore her factfinding obligations thereto; (2) the subsequent change in immigration authority that directs the agency to afford careful, case-by- case adjudication to asylum claims relating to domestic violence; (3) the Board’s disregard of immigration authority and Sixth Circuit precedent requiring the remand of pending applications in light of the glaring change in immigration authority here; and (4) the Board’s misapplication of the circularity rule in Marta’s case. See Zometa-Orellana, 19 F.4th at 979 (“In cases where a decision on which the IJ or BIA relied to make a determination was vacated or abrogated, we have determined that this change in the law ‘counsels remand.’”) (quoting Juan Antonio, 959 F.3d at 790 n.3); see also A-R-C-G-, 26 I&N Dec. at 394−95; Turcios-Flores, 67 F.4th at 356. No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 22

To the extent the dissent suggests that our decision mandates a “keep looking approach,” that characterization is incorrect. See Dis. Op. at 20. The dissent largely relies on unpublished decisions that—unlike Marta’s case—do not “aris[e] in the context of domestic violence,” which “involve unique and discrete issues not present in other social group determinations.” See A-R-C-G-, 26 I&N at 394. These nonprecedential decisions are inapposite. Dis. Op. at 20. And the two published decisions that the dissent relies on regarding circularity, Rreshpja and Kante, both predate A-R-C-G- and A-B-III. Those panels were not bound to consider the unique issues present in the domestic-violence context that are explicitly acknowledged by agency authority now. See Dis. Op. at 20 (citing Rreshpja, 420 F. 3d at 555, and Kante, 634 F.3d at 326−27). Finally, and perhaps more important here, the decisions that the dissent relies on do not present the same issue regarding the significant change in precedent—the abrogation of A-B-I—while Marta’s case was pending. As explained above, A-B-III explicitly commanded that “immigration judges and the Board should no longer follow A-B-I . . . when adjudicating pending or future cases,” and they should instead “follow pre-A-B-I precedent, including [A-R-C-G-].” 28 I&N at 309. No such reevaluation was afforded to Marta’s case here, and, as such, we remand for further review on this issue.

D.

The Board’s conclusion that Marta did not show that the Guatemalan government was unwilling or unable to protect her family does not comport with our precedent, nor is there substantial evidence in the record to support it.

When making the state-action determination, the Board must review (1) the government’s response to the applicant’s persecution if reported, and (2) broader documentary evidence surrounding the country’s conditions. Zometa-Orellana, 19 F.4th at 979. Factors informing this determination include the extent to which the alleged persecution was investigated and punished, the degree of protection the government offered to the applicant, and any concessions made by the government. K.H. v. Barr, 920 F.3d 470, 476 (6th Cir. 2019). Importantly, “a government’s specific response to [an applicant’s] persecution cannot be the only relevant evidence an immigration judge considers.” K.H., 920 F.3d at 476 (emphasis added). No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 23

Addressing this element in its own separate decision, the Board noted that the IJ acknowledged the evidence of police corruption in the record, and the Board adopted the IJ’s reasoning to conclude that the state-action element was not met because Glendy did not follow through with the psychological exam. (AR 005.) The Board cited a non-precedential opinion to reiterate that the IJ’s finding “[was] not clearly erroneous [because] [Marta’s] personal experience with the police d[id] not suggest an inability or unwillingness to protect her.” Id. (citing Reyes Almendarez v. Barr, 817 F. App’x 35, 41–42 (6th Cir. 2020)).

As a preliminary matter, the IJ again relied heavily on A-B-I (now abrogated) to make the state-action determination. When addressing the asylum claim, the IJ briefly mentioned the documentary evidence in the record when she stated that “[t]he mere fact that a country may have problems effectively policing certain crimes or that certain populations are more likely to be victims of crime cannot itself establish an asylum claim.” (AR 212) (citing A-B-I). By relying on A-B-I, the IJ effectively disregarded the documentary evidence in the record that addressed the country’s conditions, and—despite A-B-I’s abrogation—the Board summarily adopted the IJ’s one-dimensional reasoning. But this one-dimensional approach also conflicts with our precedent, which, as explained in detail above, requires a more holistic analysis of the record. See K.H., 920 F.3d at 476 (rejecting a “one-dimensional approach” and stating that “we must evaluate past persecution based on the overall context of the applicant’s situation”) (cleaned up). It is worth noting, however, that even if the Board’s approach followed our instruction (it did not), there is not substantial evidence in the record to support its conclusion.

First, when looking at the Government’s actual response to Marta’s alleged persecution, we must consider, among other factors, “whether the police investigated, prosecuted, [or] punished the persecuto[r],” and “the degree of protection offered to an asylum applicant following the persecution.” Id. at 476−77. Here, there was no investigation, prosecution, punishment, or protection offered. To the contrary, after learning of Marvin’s prior abuse and his threat to kill Marta’s family, the record reflects that the police did not inquire as to Marvin’s location, probe further into Marvin’s abuse or the threat, or attempt to make an arrest. (AR 272, 279.) Instead, the police questioned the psychological fitness of Glendy—who had experienced the most severe domestic abuse and now faced an immediate death threat from that same No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 24

abuser—and required that Glendy return for a psychological exam before they would implement any “security measures.” (AR 115).

Without immediate protection, Marta and her family would be unable to stop Marvin if he chose to carry out his threat or continue his abuse. Further, returning for the exam would require hiking through dangerous territory over several days with no promise of a future arrest. Rather than continue this futile and risky process, Marta fled to the United States with her family to escape Marvin’s death threats. See Juan Antonio, 959 F.3d at 794 (“But it cannot be that an applicant must wait until she is dead to show her government’s inability to control her perpetrator.”)

Notably, we have previously remanded and ordered the Board to reassess the sufficiency of the state-action element where an applicant took much less initiative than Marta by “not [even] report[ing] the incidents regarding her abuse to the police department.” See Zometa- Orellana, 19 F.4th at 979−980 (noting that the applicant “explained that she feared retaliation and believed that the police would not act in any meaningful way to protect her”). And even in the unpublished circuit opinion that the Board cited, Reyes, the court concluded that the Board’s decision was supported by substantial evidence only where the police actually tried to protect the applicant-victim and “responded, investigated, offered protection at least once, and were able to make arrests.” 817 F. App’x at 42. The Guatemalan police force undertook no such efforts on behalf of Marta and her family.

Second, the documentary evidence addressing the country’s conditions in the “Guatemala 2015 Crime and Safety Report,” coupled with Marta’s individual circumstances, further suggests that the government was unwilling or unable to protect Marta’s family. (See AR 377.) Again, that report explained that “[m]ore often than not a police investigation fails to result in an arrest, much less a conviction,” and “[s]upport for [assault victims] is lacking outside of major cities” in areas such as the remote village that Marta and her family lived in. (AR 391, 404.) Thus, based on the evidence in the record, Marta could not “reasonably expect the assistance of the government in controlling [Marvin’s] actions.” See Juan Antonio, 959 F.3d at 795 (quotation marks omitted). No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 25

As such, because the Board erred in making both the PSG and state-action determinations in Marta’s case, we remand to the Board so that it may review the record as a whole consistent with this opinion’s reasoning and current immigration authority.

III.

The Board’s denial of Marta’s application for withholding of removal was also in error. Here, the IJ analyzed Marta’s applications for asylum and withholding of removal under the INA in concert, and the Board summarily adopted the IJ’s decision as to Marta’s withholding-of- removal claim. (AR 005, 212−13.) We review the IJ’s decision because the Board adopted it for this claim. See Khalili, 557 F.3d at 435.

The IJ stated that “[h]aving failed to establish eligibility for asylum under Section 208 of the [INA], it necessarily follows that [Marta] cannot meet the higher burden required for withholding of removal under Section 241(b)(3) of the [INA].” (AR 212.) The IJ then explained that its denial of withholding of removal was predicated on the fact that evidence in the record was “insufficient to carry the high burden to show that [Marta] would be specifically targeted for torture if she returned to Guatemala.” (Id.) (emphasis added). The IJ’s conclusion is incorrect as a matter of law for two reasons: (1) under the INA, the IJ was required to conduct an independent analysis as to the probability of persecution under the separate withholding-of- removal standard—not the asylum standard—and (2) the IJ erred when determining that Marta did not meet the INA’s withholding- of-removal requirement because she could not show that she would be “targeted for torture” if she returned to Guatemala. (Id.)

First, we have “explained that the nexus showing required for withholding of removal differs from the nexus showing required for asylum.” Sebastian-Sebastian v. Garland, 87 F.4th 838, 851 (6th Cir. 2023); see also Turcio-Flores, 67 F.4th at 357 (“A court need look no further than the text of the withholding statute in comparison with the text of the asylum statute: ‘a reason’ [in the withholding statute] is different from—and weaker than—‘a central reason [in the asylum statute].’”) (cleaned up) (emphasis added). It is true that noncitizens must generally meet a “higher burden” to establish eligibility for withholding of removal under the INA. No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 26

Guzman-Vazquez v. Barr, 959 F.3d 253, 274 (6th Cir. 2020). But we have clarified that this “higher burden” is limited to an applicant’s ability to prove the likelihood of persecution. Id. (“[A] [noncitizen] must meet a higher burden in establishing a right to withholding of removal than in demonstrating asylum eligibility, meaning that, with respect to [the] likelihood of persecution, a [noncitizen] who fails to qualify for asylum necessarily does not qualify for withholding of removal.”) (cleaned up) (emphasis added).

Relevant here, the IJ already concluded in her asylum analysis that the harm that Marta’s family faced rose to the level of persecution. (AR 210.) And to the extent that the IJ denied that past persecution existed, that denial was premised only on Marta’s alleged failure “to show a nexus between the alleged harm and a cognizable [PSG].” (Id.) The IJ therefore failed to conduct the persecution analysis with respect to the withholding claim, and we are unable to appropriately review this claim as a result. Specifically, our precedent states that “[w]hen the [Board] analyzes asylum and withholding of removal in concert and ‘does not examine the evidence and make specific findings regarding the probability of persecution with respect to the withholding of removal claims, this court does not have sufficient grounds to review the decision under the substantial evidence standard.’” Sebastian-Sebastian, 87 F.4th at 851 (quoting Juan Antonio, 959 F.3d at 797).

Second, the IJ’s denial of withholding of removal in Marta’s case is based on an inapplicable immigration standard. The IJ referenced a “higher burden” for withholding-of- removal claims before explaining that Marta would not be “specifically targeted for torture if she returned to Guatemala.” (AR 213.) As the Board noted, however, Marta did not seek relief under the Convention Against Torture (CAT), (AR 3 n.2), and withholding of removal under the INA has nothing to do with torture. See Kamar v. Sessions, 857 F.3d 811, 817 (6th Cir. 2017) (“Withholding of removal under § 241(b)(3)(A) of the [INA] is mandatory if an alien shows a ‘clear probability’ that, if she was removed, her ‘life or freedom would be threatened’ on a protected ground such as her ‘race, religion, nationality, membership in a particular social group, or political opinion.’”) (quoting 8 U.S.C. § 1231(b)(3)(A)). Thus, the IJ’s reliance on the wrong immigration statute, CAT, also results in legal error. No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 27

In the same breath, the IJ applied the wrong nexus standard and relied on an inapplicable statute to deny Marta’s withholding claim. The Board summarily affirmed. This flurry of mistakes cannot stand. Accordingly, the Board should also consider Marta’s withholding-of- removal claim on remand, and in making this determination under the INA—not CAT—the Board should consider “all of the evidence [in the] record.” See Juan Antonio, 959 F.3d at 797 (“When this court remands to the agency to reconsider the asylum claim, the Board should consider on remand whether petitioners are entitled to withholding of removal based on all of the evidence on the record.”) (internal quotation marks omitted).

IV.

Marta and her family were assaulted and repeatedly threatened by Marvin, a serial abuser with gang affiliations. They went to the Guatemalan government for help, but they were not provided relief. Marta’s family fled to the United States, seeking relief for asylum and withholding of removal under the INA. For their asylum claim, the IJ relied on previously applicable immigration authority to proclaim that victims of domestic violence do not qualify for relief. When denying the withholding claim, the IJ applied the wrong standard for the nexus determination and relied on an inapplicable statute to make her final conclusion. The Board incorrectly affirmed the denial of relief as to both claims.

For the reasons stated, we grant Marta’s petition for review, vacate the Board’s denial of her applications for asylum and withholding of removal, and remand to the Board for reconsideration consistent with this opinion. No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 28

_________________

DISSENT _________________

LARSEN, Circuit Judge, dissenting. The majority grants the petition for review on the ground that the lead petitioner, Marta Ajualip, presented the Immigration Judge (IJ) with a non- circular “particular social group.” Yet neither Marta, nor her counsel, ever advanced the social group the court suggests. The majority nonetheless faults the BIA for not scouring the record on its own to find the characteristics the majority deems sufficient to support a cognizable social group. In the process, the court casts aside established BIA precedent and ignores the limits on our statutory authority. Respectfully, I dissent.

An applicant for asylum or withholding of removal must show that she faces past persecution or a “well-founded fear of persecution on account of race, religion, nationality, membership in a particular social group, or political opinion.” 8 U.S.C. §§ 1101(a)(42)(A), 1158(b)(1)(B)(i), 1231(b)(3)(A) & (C). “A ‘particular social group’ must meet three criteria: (1) immutability (members must share an immutable characteristic), (2) particularity (the group has discrete and definable boundaries), and (3) social distinction (society actually perceives the purported group as a distinct class of persons).” Cruz-Guzman v. Barr, 920 F.3d 1033, 1036 (6th Cir. 2019).

In addition, “a social group may not be circularly defined by the fact that it suffers persecution. In other words, the individuals in the group must share a narrowing characteristic other than their risk of being persecuted.” Kante v. Holder, 634 F.3d 321, 327 (6th Cir. 2011). “This narrowing requirement enforces the recognition that the social group category was not meant to be a ‘catch all’ applicable to all persons fearing persecution.” Solis-Gonzalez v. Holder, 523 F. App’x 320, 321 (6th Cir. 2013) (citation and internal quotation marks omitted). The non- circularity requirement was discussed in the Board’s now-overruled decision in Matter of A-B- (A-B-I), 27 I. & N. Dec 316, 334–35 (B.I.A. 2018). But, as our decision in Kante shows, the non-circularity principle pre-dated A-B-I; and as the majority acknowledges, it remains a part of No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 29

our law, and the Board’s, even now. Maj. Op. at 9; see Reyes Galeana v. Garland, 94 F.4th 555, 559 (6th Cir. 2024).

The majority opinion determines that the lead petitioner in this case, Marta Ajualip, has set forth a non-circular “particular social group” consisting of victims of domestic violence who: “(a) lived in a remote village in the mountains, where (b) there were no local authorities or means to protect women from domestic violence, and (c) . . . had no apparent ability to escape their abuser, who was allegedly associated with gangs, while remaining in Guatemala.” Maj. Op. at 14. But Marta has never proposed such a group. Not before the IJ; not before the Board; and not even before this court. So whether or not this group is cognizable, we have no license to consider it.1

Marta, who has been represented by counsel throughout these proceedings, presented the IJ with two particular social groups: “victims of domestic violence” and “family members of domestic violence.” Specifically, when asked, Marta’s attorney informed the IJ that there was no “other social group or groups under which [the petitioners] will” be proceeding beyond “victims of domestic violence and family members of domestic violence.” See AR 261. The IJ expressed appropriate concern about counsel’s specification of these particular social groups and questioned counsel about their legal viability:

JUDGE TO MS. HARRIS How is that going to pass given the Attorney General’s decision in Matter of A-B-[I]? How is going to be a cognizable social group under the law where the Attorney General held that to be cognizable, a particular social group must exist independently of the harm asserted? And you’ve just defined groups specifically by the harm.

MS. HARRIS TO JUDGE Okay. Your Honor, they are in fear of being persecuted if they go back to their country.

1 The majority purports not to make the ultimate cognizability determination here, though it suggests that the group it has fashioned for Marta is “very similar” to particular social groups our court has previously recognized. See Maj. Op. at 15. I am not so sure. But I do not engage the question because the groups Marta actually proposed—“victims of domestic violence” and “family members of domestic violence”—cannot pass muster under our cases. No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 30

JUDGE TO MS. HARRIS I understand that because they’re victims of domestic violence. *** My question to you, counsel, is how is that a legally cognizable social group?

MS. HARRIS TO JUDGE Your Honor, we plan on the testimony here, what they are afraid of.

JUDGE TO MS. HARRIS My question to you, we’re going to hear testimony, is how is that a legally cognizable social group? You as counsel have to put forth a legally cognizable social group.

MS. HARRIS TO JUDGE So they are family members of somebody who has been persecuted.

Id. at 261–62. The IJ then gave up and asked counsel to call her first witness. When lead petitioner, Marta, took the stand, the IJ tried again. In an evident attempt to elicit testimony supporting a non-circular particular social group, the IJ asked Marta:

Q. Do you believe you or any family member would be seen as distinct or in any way stand out in Guatemalan society? A. Yes. Q. What way? A. Yes, because when someone arrives to Guatemala after being in the United States, they think that you have money, and then they start to extort you. Q. Is there any other reason you would be seen as being distinct within Guatemalan society? A. No. Q. Do you consider yourself or any family member to be a part of any group within Guatemala? A. No.

Id. at 273–74. No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 31

In light of this testimony, and counsel’s insistence on her two particular social groups, the case proceeded.2 The IJ rejected the proffered particular social groups on the grounds that they were circular, not particular, and lacked social distinction.

On appeal to the BIA, Marta reiterated that she was claiming “membership in a particular social group, defined by the fact that they are victims of domestic violence.” Id. at 31; see also id. at 32 (describing the “social group defined” as “victims of domestic violence”). She did not mention “family members of domestic violence” at all, but at times she suggested another social group—“female victims of domestic violence.” Id. at 32–33 (emphasis added). This group, she apparently hoped, would be acceptable because she believed the Board’s opinion in Matter of A- R-C-G-, 261 I. & N. Dec. 388 (B.I.A. 2014), had “recognized female victims of domestic violence as a cognizable and socially distinct group of people.” AR 32. She offered no argument to develop the point, however.

The BIA did not explicitly entertain the newly proposed social group. But it did acknowledge that the Attorney General had overruled A-B-I while Marta’s appeal was pending. See Matter of A-B- (A-B-III), 28 I. & N. Dec. 307 (B.I.A. 2021) (overruling A-B-I and reinstating A-R-C-G-). And it explained why the group Marta had proposed to the IJ (“victims of domestic violence”) still differed from the one found cognizable in A-R-C-G-. In A-R-C-G-, the BIA explained, “the group was defined by an inability to leave a relationship, which we noted was not limited solely to harm but ‘may be informed by societal expectation about gender and subordination, as well as legal constraints regarding divorce and separation.’” AR at 4. (quoting A-R-C-G-, 26 I. & N. Dec. at 393). But Marta’s “proposed group is defined solely by being victimized by domestic violence and therefore is not cognizable.” Id. As for the proposed social

2 The majority attempts to paint the IJ’s treatment of Marta and the IJ’s handling of the case in an unfavorable light. For example, the majority says, “The IJ’s reliance on A-B-I, now overruled, colored her actions and decisions throughout the rest of the proceedings.” Maj. Op. at 4. There is no evidence to support this and no indication that the IJ would have reached a different result after the BIA overruled A-B-I. After all, circular social groups were not permitted before A-B-I, nor are they permitted now. The majority also claims that that the IJ conducted only a “very brief examination of Marta.” Id. at 5. But that is hardly surprising. The brevity of the questioning was directly related to Marta and her counsel’s having delineated a circular social group that required little analysis under the relevant caselaw. The majority also accuses the IJ of “summarily den[ying] Marta’s claims.” Id. That accusation is perplexing, as the IJ’s reasoned oral decision spans six transcript pages. No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 32

group of “family members of domestic violence,” the BIA concluded that Marta had abandoned her claim for relief.

The question for us is whether the BIA erred by rejecting the two particular social groups Marta proposed to the IJ. It did not. The BIA was right that “victims of domestic violence” is not cognizable. It has long been the rule in this circuit that “the group cannot be defined exclusively by the fact that its members have been subject to harm.” Kante, 634 F.3d at 327 (citation and internal quotation marks omitted). That was true before A-B-I, see id., and it remains the rule after A-B-III, see Reyes Galeana, 94 F.4th at 559. So, as the BIA correctly concluded, the overruling of A-B-I did nothing to transform “victims of domestic violence” into a non-circular social group. See AR 4 (Despite A-B-III, “it is nonetheless well-established . . . that a particular social group cannot be circularly defined by the fact that it suffers persecution . . . .”). Marta and the majority offer no case to suggest otherwise.

As for “family members of victims of domestic violence,” the BIA thought “that this group may share a narrowing characteristic other than their risk of being persecuted.” Id. But, even if so, the Board noted that Marta’s appeal had failed to contest the IJ’s “determination that she ha[d] not demonstrated that her proposed groups are socially distinct within Guatemalan society, particularly in light of her testimony that she did not believe that she or her family would be seen as distinct or belonged to any particular group within Guatemala.” Id. at 4–5 (emphasis added). The Board held that Marta’s failure to challenge the IJ’s social-distinction ruling on appeal, “which is dispositive of her asylum and withholding of removal claims,” waived the question before the BIA. Id. at 5 (“[W]here an issue addressed in an Immigration Judge’s decision is not raised on appeal it may be deemed waived.” (citing Matter of R-A-M-, 25 I. & N. Dec. 657, 658 n.2 (B.I.A. 2012) and Matter of O-R-E-, 28 I. & N. Dec. 330, 336 n.5 (B.I.A. 2021))).

This ruling was not in error. A “particular social group” must be socially distinct. Cruz- Guzman, 920 F.3d at 1036. The IJ held that Marta had provided insufficient evidence of social distinction. Marta’s brief to the BIA did not challenge that conclusion. Indeed, it did not No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 33

contest, or even mention, the IJ’s ruling with respect to her “family members of domestic violence” claim at all. And we have held that “[t]he BIA [is] not required to consider issues not raised to it on appeal.” Lopez-Hernandez v Garland, 2023 WL 4626785, at *5 (6th Cir. July 19, 2023). That should end our inquiry.3

But what of the group Marta alluded to in her brief before the BIA—female victims of domestic violence? The BIA did not explicitly address it, but that was not error either. Longstanding BIA precedent permitted the BIA to limit its review to the two groups Marta had presented to the IJ. See Matter of W-Y-C- & H-O-B-, 27 I. & N. Dec. 189, 191 (B.I.A. 2018); Matter of A-T-, 25 I. & N. Dec. 4, 10 (B.I.A. 2009); In re J-Y-C-, 24 I. & N. Dec. 260, 261 n.1 (B.I.A. 2007). It was Marta’s duty “to clearly indicate the exact delineation of any particular social group(s) to which she claims to belong” to the IJ. Gonzalez-Valencia v. Barr, 764 F. App’x 510, 512 (6th Cir. 2019) (emphasis added) (quoting W-Y-C-, 27 I. & N. Dec. at 191). And, as an appellate body, the Board’s general policy is not to consider new social groups not first presented to the IJ. W-Y-C-, 27 I. & N. Dec. at 191–92. In unpublished authority, we have upheld W-Y-C’s “exact delineation” and “forfeiture” standards. Gonzalez-Valencia, 764

3 The majority opinion claims that the BIA was wrong to find waiver. As the majority sees it, Marta preserved her claim before the BIA because the IJ “fail[ed] to address ‘family members of victims of domestic violence” at all, and Marta did challenge the IJ’s “social distinction determination” with respect to her other group, “victims of domestic violence.” Maj. Op at 18. But if the IJ had dismissed her asylum claim without ruling on one of her social groups, Marta needed to raise that to the BIA. She did not. She did not even mention the group “family members of domestic violence” to the BIA. And preserving a challenge to the IJ’s “social distinction” determination regarding a different group could not have sufficed. Be that as it may, the majority’s premises are wrong. First, the IJ did rule on “family members of domestic violence.” The IJ noted the proposed social groups— “victims of domestic violence and family members of victims of domestic violence.” AR 210. In the very next paragraph, the IJ found “that such a social group or groups are not socially distinct within Guatemalan society because the respondent has provided insufficient evidence or testimony to indicate that Guatemalan society at large views victims of domestic violence as a particular group within society.” Id. at 210–11 (emphasis added). And the IJ pointed to Marta’s testimony that she did not “believe[] she or anyone in her family were a member of any particular group seen as distinct within Guatemalan society.” Id. at 211 (emphasis added). Despite the majority’s contention to the contrary, this can only be read as the IJ concluding that Marta had not produced sufficient evidence of social distinction for either group. If Marta believed that dispositive conclusion was in error, she needed to appeal it to the BIA. She did not. Second, Marta did not meaningfully contest social distinction at all. Marta’s only mention of social distinction in her brief to the BIA consists of two sentences generally describing social distinction and this line: “In [A-R-C-G-], the [Board] further recognized female victims of domestic violence as a cognizable and socially distinct group of people.” Id. at 32 (emphasis added). That doesn’t reference “family members of domestic violence,” and in any event, it is not enough. We have explained that preservation requires “that [an] issue must be reasonably developed in the petitioner’s brief to the BIA.” Khalili v. Holder, 557 F.3d 429, 433 (6th Cir. 2009). No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 34

F. App’x at 512–13. Other circuits have upheld them as well. See Cantarero-Lagos v. Barr, 924 F.3d 145, 153 (5th Cir. 2019); Honcharov v. Barr, 924 F.3d 1293, 1296–97 (9th Cir. 2019).

Even if the BIA had considered the new group, neither Marta nor the majority suggest that merely adding “female” to “victims of domestic violence” could have changed the outcome under our caselaw. In Rreshpja v. Gonzales, we rejected “young . . . attractive Albanian women who are forced into prostitution” as a proposed social group. 420 F.3d 551, 555 (6th Cir. 2005). We explained that the attribute of being “forced into prostitution” is defined by the harm; and that, even if the group were stripped of its circular attributes, the remaining group of “young, attractive Albanian women,” was too “subjective” and “generalized” to qualify. Id. at 555–56. So too here. The attribute of being “victims of domestic violence” is defined by the harm. And even stripping away that circular attribute would leave us only with “females,” a group far too “‘sweeping’” to constitute a particular social group. Paplekaj v. Holder, 411 F. App’x 844, 846 (6th Cir. 2011) (quoting Rreshpja, 420 F.3d at 555). There can be no doubt then that the group “female victims of domestic violence” is not a cognizable social group under our caselaw.

The BIA appropriately rejected each of the groups Marta offered to the agency. So I would deny the petition for review.

The majority opinion nonetheless vacates the Board’s denial of Marta’s applications for asylum and withholding of removal. Despite the fact that Marta has repeatedly insisted that her two social groups are “victims of domestic violence” and “family members of domestic violence,” the majority faults the Board for accepting Marta’s persistent representations— through counsel—that these are the two groups she wished to present.4 In so doing, the majority

4 The reader might wonder whether I exaggerate Marta’s insistence on her particular social groups. I do not. Her commitment to these groups before the agency is outlined above. In her opening brief in this court, Marta did, at times, suggest another group: “female victims of domestic violence who find themselves unable to leave said relationships.” Petitioner Br. at 10. And at times, her case citations gave the impression that perhaps she was suggesting she belonged to a group of indigenous Mayan women. See id. at 13, 15–16, 18–19, 21. The government responded by arguing a lack of exhaustion. See Respondent Br. at 23–24. Marta’s reply brief then took pains to clarify that she was not pressing any group other than the two she offered to the IJ. See Reply Br. at 1 (“Petitioners have requested, and continue to request, classification solely as victims of domestic violence and family members of victims of domestic violence throughout the duration of these proceedings.”); id. at 2 (“[F]rom the onset of these proceedings, the Petitioners have always asserted their membership in the particular social group consisting of No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 35

opinion casts aside the Board’s rule that it is the applicant’s burden to “clearly indicate the exact delineation of any particular social group(s) to which she claims to belong” to the IJ. W-Y-C-, 27 I. & N. Dec. at 191 (emphasis added) (citation and internal quotation marks omitted). In its place, the majority opinion imposes a new obligation on the Board: it must plumb “the record as a whole” to determine whether an “application reflects additional immutable and narrowing characteristics separate and apart from [the] suffering of domestic violence.” Maj. Op. at 14. This, the majority says, is a “proper circularity analysis.”

Had the BIA complied with this new rule, the majority says that the agency could have pieced together a “particular social group” consisting of victims of domestic violence who “(a) lived in a remote village in the mountains, where (b) there were no local authorities or means to protect women from domestic violence, and (c) [who] had no apparent ability to escape their abuser, who was allegedly associated with gangs, while remaining in Guatemala.” Id. Because the Board “should have considered” these characteristics in its analysis, the majority vacates the Board’s decision. Id. at 19–20.

Many problems plague the majority’s approach. One that should particularly concern us is that we have no statutory authority to decide the questions the majority raises.

Section 242(d)(1) of the Immigration and Nationality Act (INA) bars courts from reviewing final orders of removal unless the petitioner has “exhausted all administrative remedies available to [her] as of right.” 8 U.S.C. § 1252(d)(1). This is a “mandatory” claims- processing rule that we may not excuse, if it is timely invoked. Santos-Zacaria v. Garland, 598 U.S. 411, 421 (2023); United States v. Alam, 960 F.3d 831, 833 (6th Cir. 2020). Here, the government timely invoked the rule in its answering brief, so we must apply it.

‘victims of domestic violence’ and ‘family members of domestic violence.’”); id. at 3 (Petitioners “have classified themselves both as victims of domestic violence and family members of victims of domestic violence.”); id. at 4 (“Petitioners have consistently held that their particular social group is in fact victims of domestic violence and family members of victims of domestic violence.”); id. (“Petitioners in this matter have not sought to redefine their particular social group at any point throughout these proceedings, as to do so would alienate certain members of this particular family from eligibility, as they would not meet the narrowing characteristics that would be necessary to fall within the limited class of individuals as described.”); id. at 6 (“Petitioners classified as both victims of domestic violence and family members of victims of domestic violence.”); id. (“Petitioners would again point to the fact that they have not asserted any particular social group other than those articulated at the onset of these proceedings.”). No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 36

To satisfy the INA’s exhaustion requirements, Marta had to “identify every legal or factual issue” before the BIA. Singh v. Rosen, 984 F.3d 1142, 1155 (6th Cir. 2021). That included the “particular social groups” she wanted the agency to consider. See Gil-Cerqueda v. Rosen, 841 F. App’x 815, 819 n. 3 (6th Cir. 2021); Orellana v. Sessions, 722 F. App’x 443, 449– 50 (6th Cir. 2018); De Mejia v. Sessions, 691 F. App’x 245, 248 (6th Cir. 2017); Reyes-Cardona v. Holder, 565 F. App’x 366, 368 (6th Cir. 2014) (per curiam); see also Del Carmen Amaya-De Sicaran v. Barr, 979 F.3d 210, 215 (4th Cir. 2020) (collecting authority from a “broad array of our sister circuits” acknowledging the statutory prohibition against considering “newly raised groups” on appeal). The INA’s exhaustion requirement simply prevents us from considering new social groups in a petition for review.

As noted, Marta, through counsel, told the IJ in no uncertain terms that her two “particular social groups” were “victims of domestic violence” and “family members of victims of domestic violence.” Before the BIA, she failed even to mention “family members of domestic violence” but, at times, suggested that the group might instead be “female victims of domestic violence.” See AR 33. Even if we thought that the IJ’s mention of the “family members” group was sufficient to preserve it, but see Khalili, 557 F.3d at 433, we would at best have statutory authority to consider those three groups. We have no license to consider the particular social group the majority invents for Marta here: victims of domestic violence who “(a) lived in a remote village in the mountains, where (b) there were no local authorities or means to protect women from domestic violence, and (c) [who] had no apparent ability to escape their abuser, who was allegedly associated with gangs, while remaining in Guatemala.” Maj. Op. at 14.

Nor do we have statutory authority to impose on the Board the entirely new requirement that it rummage through “the record as a whole” in search of “additional immutable and narrowing characteristics” that might help define a viable “particular social group.” Id. The Board has squarely placed on the applicant’s shoulders the duty to articulate to the IJ “the exact delineation of any particular social group(s) to which she claims to belong.” W-Y-C-, 27 I. & N. Dec. at 191. Marta did not ask the BIA to reconsider this rule; nor did she argue that the Board or the IJ had a duty to search the record for “additional immutable and narrowing No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 37

characteristics.” Indeed, she didn’t even ask this court to impose such a requirement. So we lack statutory authority to decide this issue as well.

One final exhaustion note. The majority hangs much of its analysis on its conclusion that the IJ “prevented” Marta from offering testimony relevant to her proposed social group, when the IJ told Marta to “skip the biometric information.” See, e.g., Maj. Op. at 4, 10, 14. Marta made no such argument to the BIA, so that claim is not exhausted either. Singh, 984 F.3d at 1155. In any event, as I read the transcript, the IJ was merely asking Marta to skip testimony about her current situation, the names of her children and grandchild, her education, and her present marital and employment status. All of this was well established in the record and had no bearing on Marta’s past in Guatemala. I can find no place in the record where the IJ prevented Marta’s counsel from asking the witnesses questions about their time in Guatemala. And Marta’s counsel offered no objection to the court’s instruction. The majority’s conclusion that the IJ prevented Marta from making her case is not grounded in the record, and we may not consider such a claim anyway.5

Even if the INA’s exhaustion requirement did not limit our review, the majority’s procedural innovation should fail. The new rule is inconsistent with W-Y-C-. Yet the majority has not adequately explained why the agency’s rule must be set aside. Courts may “hold unlawful and set aside agency action . . . found to be arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” 5 U.S.C. § 706(2)(A). But the majority has provided no reason to think that the agency’s rule is legally wrong; the closest it comes is a brief allusion to “certain obligations under international law.” Maj. Op. at 13 (emphasis omitted) (quoting Zometa-Orellana v. Garland, 19 F.4th 970, 979 (6th Cir. 2021)). As for the agency’s reasoning, the majority has not shown why that was deficient either. The BIA supplied its rationale in a

5 The majority also seems to reinstate the derivative claim made on behalf of Marta’s grandson. See Maj. Op. at 2 and n.1. But the IJ ruled that the grandson could not make a derivative claim. And after that ruling, Marta’s counsel asked the IJ “to sever [the grandson] from the proceeding and then move him to the adjustment of status pending his application.” AR 255. The IJ granted that motion. Id. at 266. Marta didn’t appeal those determinations to the BIA, or this court, so it appears Marta’s grandson is no longer part of these proceedings. No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 38

published decision.6 The agency explained that “[i]t is an applicant’s burden to establish her claim for relief or protection on the record before the Immigration Judge.” W-Y-C-, 27 I. & N. Dec. at 191; see also 8 U.S.C. § 1158(b)(1)(B)(i) (asylum); id. § 1231(b)(3)(C) (withholding of removal). It follows, the Board said, that the applicant “must ‘clearly indicate’ on the record before the Immigration Judge ‘what enumerated ground(s) she is relying upon,’” including the “exact delineation of any particular social group(s) to which she claims to belong.” W-Y-C-, 27 I. & N. Dec. at 191 (citation omitted). “[A]rticulating the contours” of the social group, the Board explained, allows the IJ to make “relevant factual findings” related to the group (like immutability, particularity, social distinctiveness, and membership), “which [the BIA] cannot do in the first instance on appeal.” Id. And because the Board is an appellate body, the Board generally will not consider particular social groups not presented to the IJ. Id. at 190–92.

The Fifth Circuit upheld the W-Y-C- rule on petition for review. Cantarero-Lagos, 924 F.3d at 151–52. Unpublished authority in this circuit also upheld it, Gonzalez-Valencia, 764 F. App’x at 512–13, and other circuits have readily applied it. See Del Carmen Amaya-De Sicaran, 979 F.3d at 214; Honcharov, 924 F.3d at 1295–97; Mayorga-Rosa v. Sessions, 888 F.3d 379, 382–83 (8th Cir. 2018); Vazquez v. Garland, 2021 WL 5822688, at *4 (10th Cir. Dec. 8, 2021); Lopez-Monroy v. U.S. Dep’t of Homeland Sec., 751 F. App’x 303, 307 (3d Cir. 2018). To my knowledge, only the Fourth Circuit has rejected the W-Y-C- rule by imposing a duty on the agency to “help the respondent articulate a cognizable social group.” Quintero, 998 F.3d at 630. Whatever the merits of that decision, the Fourth Circuit made clear that its rule applies only to pro se petitioners. See id. at 636 (“[W]e hold that W-Y-C-’s exact-delineation requirement is inapplicable in pro se cases.”).

In this case, Marta has always been represented by counsel. The IJ took great pains to impress on counsel that her proposed social groups were entirely circular. The IJ even questioned Marta herself, evidently hoping that the witness might produce some different, non-

6 In other litigation, the government has explained that the Board’s W-Y-C- rule is an “interpretation of its standard-of-review regulation, 8 C.F.R. § 1003.1(d)(3).” See Quintero v. Garland, 998 F.3d 612, 636 (4th Cir. 2021). Here, the government had no opportunity to explain or defend its rule at all. It had no notice that its W-Y-C- rule was in question, since Marta did not raise the issue either before the Board or in her briefing to this court. Indeed, in her reply brief, Marta forcefully retreated from any suggestion that she was relying on any groups other than the two she had presented to the IJ. See supra n.4. No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 39

circular, social group. Marta identified none. And counsel insisted that there was no “other social group or groups under which [the petitioners] will” be proceeding beyond “victims of domestic violence and family members of domestic violence.” See AR 261. How can we fault the IJ for ultimately allowing counsel to present her case as she wished? Being “neither diviner nor advocate,” it is not the IJ’s role to “graft on or shave off adjectives until the proposed groups are narrow enough to be particular, yet visible enough to be socially distinct.” Viuda de Mejia v. Sessions, 691 F. App’x 245, 248 (6th Cir. 2017). And if Marta’s attorney was at fault for not offering other proposed social groups, there’s a remedy for that—ineffective assistance of counsel. See Hanna v. Holder, 740 F.3d 379, 387 (6th Cir. 2014).

V.

The majority claims there is nothing to see here. In its view, it has not proposed a new social group for Marta. Nor has it imposed a new procedural rule on the agency. So concerns about exhaustion and the casting aside of agency precedent are misplaced. The majority’s claims do not hold up.

As for the social group, the majority says that it has not defined a “new” social group beyond the one proposed by Marta’s counsel. Instead, it has merely identified additional “narrowing characteristics separate and apart from” the domestic violence “that the Board should have considered.” Maj. Op. at 20. I struggle to see the difference. The very definition of a “particular social group” is a group whose members share “common, immutable characteristic[s].” Umana-Ramos v. Holder, 724 F.3d 667, 671 (6th Cir. 2013) (citation omitted). Adding or subtracting characteristics obviously redefines the group. Our cases have repeatedly recognized this. We have concluded, for example, that the group “widows in El Salvador,” presented on appeal, differed from various groups presented to the BIA: “women in El Salvador,” “widow[s] who had declined the sexual advances of a gang member,” and “unmarried or widowed women who either lacked a ‘male protector’ or worked.” Viuda de Mejia, 691 F. App’x at 247–48. Similarly, we concluded that the group “Salvadoran girls,” presented to this court, differed from the group “Salvadoran children” presented to the BIA. Orellana v. Sessions, 722 F. App’x 443, 449, 450 (6th Cir. 2018). And, in each case we held that the exhaustion rule prevented us from reviewing the new groups on appeal. Id.; Viuda de Mejia, No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 40

691 F. App’x at 248. The same rule forbids the majority’s attempt to add its own “narrowing characteristics” to save the circular social groups Marta’s counsel presented to the BIA.

As for the procedural rule, the majority claims that its “proper circularity analysis” is just business as usual, required by our precedent. That is not right. But even if it were, we could not consider a claim that the IJ had fallen short on this score unless Marta had presented it to the BIA. See 8 U.S.C. § 1252(d)(1); Singh, 984 F.3d at 1155. She did not. She never argued to the BIA that the IJ had neglected the “narrowing” characteristics that the majority now says it “should have considered.” Nor did Marta ever suggest to the BIA that the IJ had impeded her from developing other facts relevant to her particular social group. So we are prohibited from considering those matters.

What’s more, the majority opinion’s “proper circularity analysis” is far from business as usual. If the majority were right, asylum seekers could present any circular group (“victims of beatings” or “Guatemalans who are harmed”), leaving the IJ, then the BIA, and ultimately this court with the obligation to review the record in search of immutable, narrowing, and socially distinctive characteristics that could form a cognizable group. Essentially, the majority opinion’s “proper circularity analysis” erases the duty of proposing a cognizable social group at all.

That approach cannot be reconciled with the Board’s rule in W-Y-C-, which makes it the “applicant’s burden to specifically delineate her proposed social group.” 27 I. & N. Dec. at 191. Nor do our cases support such a drastic change in the law. The majority points to Zometa- Orellana v. Garland, 19 F.4th 970 (6th Cir. 2021), and Juan Antonio v. Barr, 959 F.3d 778 (6th Cir. 2020). But neither is on point. In both cases, the petitioners themselves explicitly delineated a proposed social group beyond “victims of domestic violence.” In Zometa, the petitioner argued before the IJ and BIA that the proposed social group was “El Salvadorian women of childbearing age in domestic partnerships.” 19 F.4th at 975, 978. By the time the case got to this court, the caselaw that had guided both the BIA and the IJ in rejecting that group had been vacated. Id. at 979. So the panel remanded the case for the agency to reconsider this proposed social group under new law. Id. Although the panel noted, in dictum, that the agency has “certain obligations under international law to extend refuge to those who qualify for such relief” and “an independent role to assess the plausibility of the application on the record as a No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 41

whole,” nothing in that opinion holds that the agency or this court may or must redefine a petitioner’s proposed social groups, particularly those crafted by counsel. See id. Whatever Zometa-Orellana’s dictum means, the majority opinion loads it with more weight than it can possibly bear.

Juan Antonio is equally unhelpful. Unlike the petitioners here, the petitioner in Juan Antonio argued before the IJ and BIA that she was a member “of the particular social group of ‘married indigenous women in Guatemala who are unable to leave their relationship.’” 959 F.3d at 786; see also id. at 788. The IJ and BIA found that social group cognizable but determined that the evidence didn’t support a finding that she was a member of the group. Id. at 789. This court vacated the agency’s fact finding on petitioner’s membership in the group because it was not supported by substantial evidence. Id. at 792. So Juan Antonio likewise offers no support for the majority’s new rule that the agency (or the reviewing court) must comb through the record to construct a social group for the petitioner.

Perhaps most telling, the majority’s “keep looking” approach cannot be reconciled with our cases addressing circularity. In Rreshpja, for example, we rejected the group the applicant proposed, “young . . . attractive Albanian women who are forced into prostitution.” 420 F.3d at 555. We did not require the IJ, the BIA, or this court itself to keep searching the record for other characteristics that might save the claim. The same was true in Kante, where we rejected as circular “Kante’s submitted classification”—“women subjected to rape as a method of government control.” 634 F.3d at 326–27. So too for Soto-Ambrocio v. Sessions, where we rejected the proposed social group because “Soto Ambrocio defined her proposed group as young women from Guatemala subject to abuse from families.” 724 F. App’x 456, 458 (6th Cir. 2018). In each of these cases, and many others, we analyzed the group proposed by the petitioner, and finding it to be circular, our inquiry stopped. See also Solis-Gonzalez v. Holder, 523 F. App’x 320, 321–22 (6th Cir. 2013) (“By defining her proposed social group in terms of those targeted by Guatemalan gangs, Solis-Gonzalez has defined the group in a circular fashion.”); Kalaj v. Holder, 319 F. Appx 374, 377 (6th Cir. 2009) (“[A] petitioner must set out some common, immutable social-group characteristic beyond the risks of persecution.”); Castro- Paz v. Holder, 375 F. Appx 586, 589–91 (6th Cir. 2010) (rejecting the proposed social group of No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 42

“individuals who are targeted by gang members because they possess sensitive information obtained through their employment” because the proposed social group was “defined by its risk of being persecuted”); Torres v. Sessions, 728 F. App’x 584, 586–87 (6th Cir. 2018) (rejecting as circular a group the petitioner defined as “Mexican Nationals who were kidnapped and beaten by Knights Templar Cartel who will be targeted for violence by Knights Templar because they fled Mexico because they were unable to meet extortion demands”).

Neither A-B-III nor A-R-C-G- changes this. Yes, as the majority notes, A-B-III requires “careful [and] case-by-case adjudication of asylum claims.” See 28 I. & N. Dec. at 309. And yes, A-R-C-G- says that “any claim regarding the existence of a particular social group in a country must be evaluated in the context of the evidence presented regarding the particular circumstances in the country in question.” 26 I. & N. Dec. at 392. But both cases asked whether the group the petitioner proposed was cognizable. Id.; A-B-I, 27 I. & N. Dec. at 321. In A-R-G- C-, that group was not “victims of domestic violence” but “married women in Guatemala who are unable to leave their relationship.” A-R-C-G-, I. & N. Dec. at 389. The BIA explained that “marital status can be an immutable characteristic where the individual is unable to leave the relationship,” but that any “determination of [that] issue will be dependent upon the particular facts and evidence in a case,” such as whether divorce would be “contrary to religious or other deeply held moral beliefs or if dissolution is possible when viewed in light of religious, cultural, or legal constraints.” Id. at 392–93. Marta did not present that social group, or anything like it, despite warnings from the IJ that her proposed group was entirely circular. Nothing in A-R-C-G- or A-B-III indicates that the agency had a duty to create a different group for her or to find additional facts beyond her proposed social group of “victims of domestic violence.”

The majority opinion also asserts that its “proper circularity analysis” is consistent with W-Y-C-, highlighting W-Y-C-’s recognition that “whether a social group is cognizable is a fact- based inquiry made on a case-by-case basis.” W-Y-C-, 27 I. & N. Dec. at 191. But the majority errs by concluding that the fact-intensive nature of the inquiry somehow relieves the petitioner from setting forth the “exact delineation” of her proposed social group. W-Y-C- explains that fact-intensiveness is instead the reason for the “exact delineation” requirement: “The importance of articulating the contours of any proposed social group before the Immigration No. 23-3274 Tista-Ruiz de Ajualip, et al. v. Garland Page 43

Judge is underscored by the inherently factual nature of the social group analysis.” Id. (emphasis added). Only with a precise description of petitioner’s groups firmly in hand can the IJ find the critical facts: “whether the group is immutable and is recognized as particular and socially distinct in the relevant society.” Id.

That is why W-Y-C- emphasizes that the IJ “should seek clarification” “[i]f an applicant is not clear as to the exact delineation of the proposed social group.” Id. And that is exactly what the IJ did here. See supra at 38–39. In response, counsel made clear that Marta wished to proceed exclusively on “victims of domestic violence” and “family members of victims of domestic violence.” AR 261–62. The IJ had nothing left to do but consider those groups.

The majority opinion overturns established agency precedent and creates a new rule requiring the agency, and this court, to unearth arguments that were never presented to it. We don’t have statutory authority to reach that result, and even if we did, the majority has identified nothing in our caselaw or a proper understanding of administrative law supports that position. The agency did not err by considering the only particular social groups offered by counsel or by concluding that they were not cognizable. We should deny the petition for review. I respectfully dissent.