Ramos v. Bondi

Opinion

FILED

United States Court of Appeals Tenth Circuit

PUBLISH

October 15, 2025

UNITED STATES COURT OF APPEALS Christopher M. Wolpert

FOR THE TENTH CIRCUIT Clerk of Court

_________________________________

MODESTA RAMOS RAMOS; A.M.R.R.; J.A.R.R.,

Petitioners, v. Nos. 23-9567, 24-9524

PAMELA J. BONDI, United States Attorney General, *

Respondent.

_________________________________

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

_________________________________

Brandon R. Gould of Covington & Burling LLP, Washington, DC, for Petitioners.

Robert Dale Tennyson, Jr., of U.S. Department of Justice, Washington, DC, Attorney, (Brian Boynton, Principal Deputy Assistant Attorney General, Nancy Friedman, Senior Litigation Counsel, and Justin R. Markel, Senior Litigation Counsel, on the brief) of Office of Immigration Litigation, U.S. Department of Justice, Washington, DC, for Respondent.

_________________________________

Before PHILLIPS, ROSSMAN, and FEDERICO, Circuit Judges.

_________________________________

* On February 5, 2025, Pamela J. Bondi became Attorney General of the

United States. Consequently, she has been substituted for Merrick B. Garland as Respondent, per Fed. R. App. P. 43(c)(2).

2

ROSSMAN, Circuit Judge.

_________________________________

This case began in late 2014, when an immigration judge (IJ) granted asylum from Honduras to Modesta Ramos Ramos and her two minor sons.1 The government then commenced a series of appeals seeking to reverse that decision. In 2015, the government prevailed in its first appeal to the Board of Immigration Appeals (BIA), and the matter was remanded to the IJ for additional factfinding. More than three years passed. In 2019, the IJ again granted asylum to Ms. Ramos. The government appealed to the BIA once more, again contesting the IJ’s factfinding. Four more years passed. Eventually, in June 2023, a three-judge panel of the BIA vacated the IJ’s decision, holding the IJ made clearly erroneous factual findings. Ms. Ramos now petitions for review of the BIA’s June 2023 order.

Ms. Ramos contends the BIA misapplied the clear-error standard. When the standard is correctly applied, Ms. Ramos argues, the record compels the conclusion that the IJ’s factual findings are not clearly erroneous. The government does not attempt to defend the BIA’s decision in this appeal.

1 Ms. Ramos and her two children are the petitioners before this court.

The children are listed as riders on Ms. Ramos’s application for asylum and withholding of removal, meaning their eligibility for relief turns entirely on Ms. Ramos’s application. See 8 U.S.C. § 1158(b)(3)(A); 8 C.F.R. § 1208.21(a) (2024). For ease of discussion, and because these children do not have independent asylum claims, we use “Ms. Ramos” in our discussion to refer to all three petitioners.

3

Instead, it insists we should “allow the Board an opportunity to reconsider” its ruling. Resp. Br. at 20. Ms. Ramos has the availing position. Under the unique circumstances of this case, remand for further agency proceedings would be futile. Ms. Ramos’s asylum case has been pending for over a decade, and it is time to resolve it. Exercising jurisdiction under 8 U.S.C. § 1252(a)(1), we grant the petition for review, 2 vacate the BIA’s June 2023 order, and remand to the BIA with instructions to reinstate the IJ’s grant of asylum to Ms. Ramos.

I3

A

Ms. Ramos is a native and citizen of Honduras. There, on the night of New Year’s Eve 2012, she attended a party at her brother’s house with Arturo Robles, her partner, and their two children. Ms. Ramos has two other children from a previous relationship: one resides in Honduras, and the other moved to the United States in 2015.

2 As we will explain, Ms. Ramos filed two petitions for review in this

court, one in July 2023 and one in April 2024. We need not decide which petition enables our review as there is no question one of them is sufficient.

3 We take the facts from the BIA order on review and, because that order’s factual recitation is sparse, unchallenged portions of the two prior IJ orders and other uncontroverted parts of the record. No party disputes any of the historical facts recited here.

4

Later that night, after the family returned home and fell asleep, Jonis Erco (Jonis)—an acquaintance who played soccer with Mr. Robles—woke the adults. Jonis had “connections to”—and may have been “a member of”— Inestroza, an affiliate gang of MS-13. RI.106, 102. Jonis persuaded Mr. Robles to go outside with him. The next morning, Ms. Ramos’s sister told her Mr. Robles had been found dead and partially decapitated a few blocks from their home. Ms. Ramos believed Jonis was the killer. It turns out Jonis “and other gang members ‘hated Arturo’ for” reasons that, while not entirely clear, may have been “because of something related to a soccer match.” RI.105.

Mr. Robles’s brother reported the killing to police, but “the officers ‘did not do anything’ to find [Mr. Robles’s] killer.” RI.102. Ms. Ramos then noticed Jonis “was ‘always’ watching her whenever she was in public.” RI.102. Jonis began threatening harm to her and her two younger children. He suggested the children would not grow up and instead “would ‘become like’” their recently murdered father. RI.102. Jonis appeared unafraid of police; when Ms. Ramos threatened to report him to law enforcement in response to his threats, “he told her to ‘go ahead’ because the police ‘didn’t frighten [him].’” RI.106 (alteration in original).

In February 2013—more than a month after Mr. Robles’s murder—

Ms. Ramos approached police to ask why they had not pursued Jonis. She also intended to tell the officers about Jonis’s threats aimed at her and the children,

5

but they “‘did not pay attention’ to her and simply said that they ‘didn’t have time’ to help her.” RI.102. Ms. Ramos attributed their inaction to their own ties with gangs.

Afraid for their safety in light of these threats, Ms. Ramos and her children fled Honduras and entered the United States unlawfully in June 2014. She did not think she had “‘anywhere to go’ in Honduras where she could be safe” because she knew Inestroza and related criminal organizations have a nationwide reach. RI.106. Ms. Ramos further worried Jonis’s brazen actions, and his attitude of impunity toward police, suggested he could even track her down himself.

Since arriving in the United States, neither Ms. Ramos nor her relatives have been contacted by Jonis. Ms. Ramos thinks her two older sons—including the one who lives in Honduras—are safe from Jonis because they are unrelated to Mr. Robles. In Ms. Ramos’s view, Jonis’s “main reason” for targeting her two younger children “was because they ‘are Arturo’s family.’” RI.104.

B

1

In September 2014, shortly after Ms. Ramos and her young sons arrived in the United States, the government placed them in removal proceedings. Ms. Ramos conceded removability and applied for asylum, withholding of removal, and relief under the Convention Against Torture (CAT). See 8 U.S.C.

6

§ 1158(a)(1); id. § 1231(b)(3)(A); Convention Against Torture and Other Cruel, Inhuman or Degrading Treatment of Punishment, Dec. 10, 1984, S. Treaty Doc. No. 100-20, 1465 U.N.T.S. 113. In November 2014, an IJ held a hearing, where Ms. Ramos was represented by pro bono counsel. Ms. Ramos testified, as did her brother who knew Mr. Robles well. Ms. Ramos submitted country- condition evidence describing, among other things, the influence of gangs in Honduras—including a report from Dr. Elliott Young, a history professor at Lewis & Clark College who specializes in conditions in Honduras.

On December 2, 2014, the IJ issued a written order granting asylum. 4 The IJ found Ms. Ramos and her brother were credible witnesses. The IJ also determined Ms. Ramos’s asserted basis for persecution—her membership in the particular social group (PSG) “of the nuclear family of Arturo Jovony Romero Robles,” RI.619—was a cognizable protected group for asylum purposes. See 8 U.S.C. § 1158(b)(1)(A) (for asylum, requiring the applicant to be “a refugee”); id. § 1101(a)(42)(A) (defining “refugee” to require past or future “persecution on account of race, religion, nationality, membership in a particular social group, or political opinion” (emphasis added)).

Turning to the so-called nexus requirement, the IJ concluded Ms. Ramos had a “well-founded fear of persecution” in the future, 8 U.S.C.

4 Because she granted asylum, the IJ declined to reach the alternative claims for relief via withholding of removal or CAT protection.

7

§ 1101(a)(42)(A), and her membership in that family-based PSG was “at least one central reason for the feared persecution.” RI.621; see 8 U.S.C. § 1158(b)(1)(B)(i) (for asylum, requiring that a protected ground “was or will be at least one central reason for persecuting the applicant”). According to the IJ, based on country-condition evidence and Dr. Young’s report, “the government of Honduras is unable to control the MS-13 gang members” and their affiliates, like Jonis. 5 RI.621. Those findings sufficed, the IJ held, to grant asylum.

The government appealed to the BIA. In a three-page, single-judge order dated November 4, 2015, the BIA reversed. The BIA acknowledged—and declined to disturb—the IJ’s favorable credibility determinations. But the BIA determined remand was required so the IJ could better explain her findings. Specifically, the BIA remanded and instructed the IJ to “further address the issues of nexus, the ability of the respondents to relocate within Honduras, and whether the government is unable or unwilling to control the criminals threatening the respondents.” RI.401.

2

In April 2019—more than three years after the 2015 BIA decision directing the remand—the IJ held another evidentiary hearing. Ms. Ramos

5 Because the IJ granted asylum on grounds of the family-based PSG,

she declined to analyze the alternative asserted basis of Ms. Ramos’s “imputed anti-gang opinion.” RI.621–22. That alternative ground was never before the BIA and is not before this court.

8

submitted more evidence on Honduran country conditions, including an update to Dr. Young’s report. Ms. Ramos and Dr. Young also testified.

Again, the IJ granted asylum. The IJ issued a written order, elaborating on the findings the BIA had determined were inadequately explained in 2014. As to nexus under 8 U.S.C. §§ 1101(a)(42)(A) and 1158(b)(1)(B)(i), citing several specific examples, the IJ found “Ms. Ramos provided credible testimony to establish that her relationship to Arturo motivated Jonis to persecute her and her children—specifically Arturo’s biological children.” RI.104. As to government protection under 8 U.S.C. § 1101(a)(42)(A), discussing Dr. Young’s expert testimony and reports the parties had submitted, the IJ found “Ms. Ramos’ personal experiences in Honduras, combined with witness testimony and documentary evidence concerning Honduran country conditions, indicate that the Honduran government is unable or unwilling to control gang members from perpetrating violence against Ms. Ramos and her children.” RI.105. As to internal relocation, again referencing Ms. Ramos’s credible testimony and evidence on country conditions showing Inestroza had a nationwide reach in Honduras, the IJ found “Ms. Ramos has met her burden to show that she could not relocate within Honduras to avoid persecution.” RI.106.

The government once more appealed to the BIA. In its notice of appeal, the government suggested the IJ’s three newly bolstered factual findings were clearly erroneous. The government’s brief to the BIA raised those same issues

9

and added two more: the family-based PSG is not cognizable, and future persecution is not sufficiently likely.

Ms. Ramos’s case remained pending before the BIA for over four years.

On June 16, 2023, a three-judge BIA panel again reversed. The panel found the IJ’s three factual findings—that Jonis’s persecution carried a nexus to the family-based PSG, that Ms. Ramos could not reasonably relocate within Honduras, and that the Honduran government is unable and unwilling to protect her—were clearly erroneous. Because that conclusion was dispositive of the asylum claim, the BIA did not reach any other issues. The BIA therefore vacated the asylum grant and determined that disposition necessarily foreclosed relief through withholding of removal. Instead of dismissing the case, however, the BIA remanded so the IJ could address Ms. Ramos’s CAT claim, which had not before been at issue because the IJ had twice granted asylum.

3

Proceedings then commenced in this court. The procedural history is somewhat technical, but we recite it here because it is relevant to our analysis and disposition.

10

On July 10, 2023, Ms. Ramos submitted a petition for review (the First Petition). 6 The First Petition was filed less than thirty days after the latest BIA order of June 16, 2023. Recall, there was not then a final removal order because the CAT claim remained unadjudicated before the IJ. While the First Petition was pending before this court, Ms. Ramos and the government unsuccessfully attempted to mediate this appeal. In March 2024, at the government’s suggestion, Ms. Ramos moved to withdraw her CAT claim in order to obtain a final order of removal from the IJ. The IJ granted that motion, thereby resolving all claims, and entered a final order of removal on March 18, 2024. Ms. Ramos did not appeal that order to the BIA, however.

Instead, on April 16, 2024, less than thirty days after the March 18 removal order, Ms. Ramos filed another petition for review in this court (the Second Petition). We issued a show-cause order asking whether we can hear this case under the First or Second Petitions. The parties responded, and those threshold questions were referred to this panel. The normal appellate briefing schedule resumed.

In July 2024, Ms. Ramos timely filed her opening brief. She argued “the BIA failed to review the IJ’s factfinding only for clear error,” instead engaging

6 Later that month, this court issued a show-cause order asking the parties to explain whether and how this court has appellate jurisdiction over that petition. The parties never responded to this order because, as we will explain, subsequent events rendered it moot.

11

in an impermissible and more searching review in violation of 8 C.F.R. § 1003.1(d)(3)(i). Op. Br. at 12 (heading capitalization omitted). She also argued the record before the BIA—including records related to both IJ hearings and decisions—compels the finding that “the IJ’s factfinding was not clearly erroneous.” Op. Br. at 44. Ms. Ramos asked this court “to vacate the BIA’s 2023 decision, determine that the IJ’s factfinding was not ‘clearly erroneous,’ and instruct the BIA to restore the IJ’s 2019 order granting asylum.” Op. Br. at 47.

On September 20, 2024, the day its response brief was due, the government instead filed a motion to remand. Our local rules allow parties to file “a motion to remand for additional . . . administrative proceedings,” but only “within 14 days after the appeal or other proceeding is docketed in this court, unless good cause is shown for later filing.” 10th Cir. R. 27.3(A)(1)(c), (A)(3)(b). In its motion, without addressing timeliness or good cause, the government argued this court should “allow the Board an opportunity to reconsider its clear error determinations.” Motion to Remand at 1. In support, the government identified errors made by the BIA in its review of the IJ’s factual findings, pointing out evidence the IJ considered that the BIA purportedly ignored and findings the IJ reached that the BIA seemingly misinterpreted. Nevertheless, the government maintained “[t]his remand

12

request is not a confession of error on the agency’s part.” Motion to Remand at 5.

Ms. Ramos opposed the motion to remand. She argued the motion—filed well after Rule 27.3’s fourteen-day limit—was untimely, the government lacked good cause under that rule to file the motion so late, and the motion’s reasons for remanding were unpersuasive. This court referred the motion to this panel and ordered merits briefing to resume.

The government then filed its response brief as directed. There, the government only reprised the arguments in its motion to remand. “As argued in the Respondent’s motion to remand,” the government explained, “Respondent requests remand to allow the Board an opportunity to reconsider its clear error determinations . . . .” Resp. Br. at 20; see also Resp. Br. at 20, 22, 23 (suggesting the BIA “may wish to reconsider” its three clear-error holdings). 7 Ms. Ramos’s reply brief observed the government does not “defend[] any aspect of the BIA’s decision on appeal.” Reply Br. at 1. She again argued the IJ’s findings were not clearly erroneous and urged this court to order the BIA

7 Under our local rules, although the government’s motion to remand was

filed late and did not argue good cause existed, “[f]ailure to file a timely motion under this rule does not preclude a party from raising the issue in a merits brief.” 10th Cir. R. 27.3(A)(3)(d). We thus may consider these arguments as they are presented in the response brief.

13

to reinstate the IJ’s grant of asylum, instead of remanding the case for further agency proceedings.

II

Before turning to the merits of the petition, we must address a threshold issue, born of the somewhat unusual fact that Ms. Ramos filed two petitions for review with this court. Here, Ms. Ramos filed one petition within thirty days of the BIA’s June 16, 2023 order, and another petition within thirty days of the IJ’s March 18, 2024 removal order. Ms. Ramos insists both petitions enable our review but is “indifferent” as to which one is ultimately adjudicated. Petrs’. Jurisdictional Memo. at 3, 11–22. The government agrees we can hear this case on appeal, but only under the Second Petition, arguing we should deny the First Petition because it is not a final order of removal or dismiss it as moot.

We have jurisdiction to review only “a final order of removal.” 8 U.S.C.

§ 1252(a)(1). Immigration petitioners have thirty days from “the final order of removal” to file a petition for review in a circuit court. 8 U.S.C. § 1252(b)(1). There is no question there was a “final order of removal” in this case. 8 U.S.C. § 1252(a)(1). Under the circumstances, we ultimately need not decide which of

14

the June 2023 BIA order or the March 2024 IJ order is final. 8 That is because Ms. Ramos filed a timely petition for review after each possible final order. 9 If the June 2023 BIA order is final, then the timely First Petition plainly enables our review of that order. If, instead, the March 2024 IJ order is final, then the timely Second Petition allows us to review the June 2023 BIA order. 10 And importantly, the government does not contest the Second Petition allows us to reach the entirety of the agency’s reasoning underlying Ms. Ramos’s

8 No party suggests, and we see no argument that, neither is final. Given

our holding, we need not consider whether or how the Supreme Court’s recent decision in Riley v. Bondi bears on the finality analysis. See generally 145 S. Ct. 2190, 2197–201 (2025).

9 In addition to timely filing, before seeking this court’s review, a petitioner also must “ha[ve] exhausted all administrative remedies available to the [petitioner] as of right.” 8 U.S.C. § 1252(d)(1). We find this exhaustion requirement does not bar our review even though Ms. Ramos did not appeal the IJ’s 2024 removal order to the BIA. Ms. Ramos argues, and the government agrees, there was then nothing left for her to appeal. Because exhaustion is nonjurisdictional and thus waivable, Santos-Zacaria v. Garland, 598 U.S. 411, 423 (2023), the government’s concession decides the issue. Independently, the parties’ reasoning is persuasive; nothing remained, at that point, for Ms. Ramos to appeal to the BIA—i.e., to exhaust. The BIA had already considered all claims of error in the IJ’s asylum grant, and it denied that form of relief and withholding of removal. And Ms. Ramos had, by that point, withdrawn her only remaining claim for relief, under CAT.

10 Independently, the government explicitly agreed not to level a timeliness objection as to the Second Petition. Under Riley—which held “§ 1252’s 30-day filing rule is not jurisdictional”—the government’s “conce[ssion]” means we “ha[ve] no obligation to consider” timeliness under this petition. 145 S. Ct. at 2203, 2201.

15

challenge to removal on appeal. This means that under either petition, we can consider the BIA’s June 2023 order. 11 We therefore conclude one of Ms. Ramos’s petitions enables our review of the June 2023 BIA order. We thus need not decide which of the two orders is final, as Ms. Ramos undoubtedly filed a timely petition for review. See People for the Ethical Treatment of Prop. Owners v. U.S. Fish & Wildlife Serv., 852 F.3d 990, 1008 (10th Cir. 2017) (“[I]f it is not necessary to decide more, it is necessary not to decide more.” (alteration in original) (quoting PDK Labs. Inc. v. DEA, 362 F.3d 786, 799 (D.C. Cir. 2004) (Roberts, J., concurring))). We now turn to the parties’ arguments concerning the BIA’s June 2023 order.

11 There is a good reason for the absence of dispute on this point. “[A]

noncitizen’s various challenges arising from [a] removal proceeding must be ‘consolidated in a petition for review and considered by the courts of appeals.’” Nasrallah v. Barr, 590 U.S. 573, 580 (2020) (emphasis added) (quoting INS v. St. Cyr, 533 U.S. 289, 313 n.37 (2001)); see also 8 U.S.C. § 1252(b)(9) (providing for consolidation “of all questions of law and fact . . . arising from any action taken or proceeding brought to remove an alien from the United States” in the same judicial action). Indeed, a final order of removal “encompass[es]” more than just the order finding the noncitizen removable, but “all matters on which the validity of the final order is contingent.” Nasrallah, 590 U.S. at 582 (quoting INS v. Chadha, 462 U.S. 919, 938 (1983) (internal quotation marks omitted)). Here, there is no question the June 2023 BIA order affects the “validity” of the final order of removal. The June 2023 BIA order is thus part of what must be consolidated for our review, meaning we can, even must, review it—even if the March 2024 IJ order is the final order of removal that enables our review.

16

III

Ms. Ramos first contends the BIA misapplied the clear-error standard in 8 C.F.R. § 1003.1(d)(3)(i) when reviewing the IJ’s factual findings. For its part, the government does not suggest otherwise. Even so, because the government does not formally concede the point, we analyze the merits of this issue.

A

“We review the BIA’s legal determinations de novo, and its findings of fact under a substantial-evidence standard.” Kabba v. Mukasey, 530 F.3d 1239, 1244 (10th Cir. 2008) (quoting Niang v. Gonzales, 422 F.3d 1187, 1196 (10th Cir. 2005)). At issue in this appeal is the BIA’s standard for reviewing an IJ’s factual findings in 8 C.F.R. § 1003.1(d)(3)(i). That regulation states, “The Board will not engage in de novo review of findings of fact determined by an immigration judge. Facts determined by the immigration judge, including findings as to the credibility of testimony, shall be reviewed only to determine whether the findings of the immigration judge are clearly erroneous.” Id. The BIA’s regulations legally bind the agency, as it is “expected to render a decision in accord with its own collective belief.” United States ex rel. Accardi v. Shaughnessy, 347 U.S. 260, 266 (1954), superseded by statute on other grounds as stated in DHS v. Thuraissigiam, 591 U.S. 103 (2020). Whether the BIA has properly applied the clear error standard in § 1003.1(d)(3)(i) is a legal determination reviewed de novo. Kabba, 530 F.3d at 1244; see id. at 1245

17

(holding “we must consider de novo whether the BIA, in making its own factual findings, actually reviewed the IJ’s decision only for clear error,” as required by § 1003.1(d)(3)(i)).

1

Nobody disputes the BIA correctly recited the clear-error standard in its June 2023 order. The BIA accurately observed “[a] finding is ‘clearly erroneous’ when although there is evidence to support it, the reviewing court on the entire evidence is left with the definite and firm conviction that a mistake has been committed.” RI.5 (quoting United States v. U.S. Gypsum Co., 333 U.S. 364, 395 (1948)). The BIA also appropriately acknowledged “[c]lear error is a high standard and we do not take the reversal of an [IJ]’s findings . . . lightly.” RI.6.

But to adjudicate Ms. Ramos’s appellate challenge, we must “determine whether the BIA applied the correct legal standard, not simply whether it stated the correct legal standard.” Kabba, 530 F.3d at 1245. To that end, it bears describing more fully what the clear-error standard of review requires in application and why the BIA must employ it when reviewing an IJ’s factual findings.

“‘Clear error’ is a term of art derived from Rule 52(a) of the Federal Rules of Civil Procedure, and applies when reviewing questions of fact.” Ornelas v. United States, 517 U.S. 690, 694 n.3 (1996); see Fed. R. Civ. P. 52(a)(6) (“Findings of fact . . . must not be set aside unless clearly erroneous . . . .”). In

18

Anderson v. City of Bessemer City, North Carolina, the Supreme Court confirmed this clear-error standard is a high bar. The Court held “[a] finding is ‘clearly erroneous’ when although there is evidence to support it, the reviewing court on the entire evidence is left with the definite and firm conviction that a mistake has been committed.” 470 U.S. 564, 573 (alteration in original) (quoting U.S. Gypsum Co., 333 U.S. at 395); accord RI.5 (the BIA reciting this standard). “This standard plainly does not entitle a reviewing court to reverse the finding of the trier of fact simply because it is convinced that it would have decided the case differently.” Anderson, 470 U.S. at 573. And “[i]f the [factfinder]’s account of the evidence is plausible in light of the record viewed in its entirety, the [appellate body] may not reverse it even though convinced that had it been sitting as the trier of fact, it would have weighed the evidence differently.” Id. at 573–74. In other words, if a “view[] of the evidence” is “permissible,” that view “cannot be clearly erroneous.” Id. at 574 (emphasis added); see also Holdeman v. Devine, 572 F.3d 1190, 1192 (10th Cir. 2009) (similarly observing the existence of another “permissible” “view of the evidence” cannot mean a factual finding is clearly erroneous).

In requiring the BIA to review the IJ’s factual findings for clear error, see 8 C.F.R. § 1003.1(d)(3)(i), the Code of Federal Regulations invokes the same standard found in Rule 52(a)(6). In Estrada-Martinez v. Lynch, the Seventh Circuit observed the BIA’s “own commentary on [§ 1003.1(d)(3)(i)] cited

19

Anderson v. Bessemer City, the case providing the definitive interpretation of Rule 52(a)(6)’s clear-error standard.” 809 F.3d 886, 895 (7th Cir. 2015) (citing Procedural Reforms to Improve Case Management, 67 Fed. Reg. 54878, 54889– 90 (Aug. 26, 2002); Anderson, 470 U.S. at 573–75). In Kabba, we likewise acknowledged the BIA is subject to Anderson’s holding that a choice among “permissible views of the evidence . . . cannot be clearly erroneous.” 530 F.3d at 1245–46 (quoting Anderson, 470 U.S. at 574). “Many circuits” have recognized the same, as they also “use the Anderson standard to evaluate whether the Board adhered to its required clear-error standard of review.” Estrada-Martinez, 809 F.3d at 895 n.5 (citing cases).

It is helpful to consider why the agency adopted the clear-error standard over de novo review for an IJ’s factual findings. In the final rule establishing this standard, the Executive Office of Immigration Review recognized “[d]uplication of the trial judge’s efforts [by an appellate body] would very likely contribute only negligibly to the accuracy of fact determination at a huge cost in diversion of judicial resources.” 67 Fed. Reg. at 54889 (alterations in original) (quoting Anderson, 470 U.S. at 574–75). Given “[t]he parties to a case on appeal have already been forced to concentrate their energies and resources on persuading the trial judge that their account of the facts is the correct one,” “requiring them to persuade three more judges at the appellate level is requiring too much.” Id. (quoting Anderson, 470 U.S. at 575). As the rule put

20

it, “[t]he ‘clearly erroneous’ standard of review recognizes that an evidentiary hearing on the merits should be the ‘main event,’” not “a ‘tryout on the road.’” Id. (quoting Wainwright v. Sykes, 433 U.S. 72, 90 (1977)).

Thus, successfully demonstrating that a factual finding is clearly erroneous is “no easy hurdle to clear.” United States v. Ludwig, 641 F.3d 1243, 1247 (10th Cir. 2011). 12 It “requir[es] the [challenger] to show that the findings are more than possibly or even probably wrong but pellucidly so.” Ludwig, 641 F.3d at 1247. Of course, the standard in § 1003.1(d)(3)(i) is not impossible to meet, but as the Second Circuit succinctly explained, “the phrase ‘clear error’ is to be taken literally: the error must be clear.” Lin v. Lynch, 813 F.3d 122, 127 (2d Cir. 2016); see id. at 125–29 (explaining the BIA’s demanding clear- error standard).

It falls to reviewing courts, then, to determine “de novo whether the BIA, in making its own factual findings, actually reviewed the IJ’s decision only for clear error.” Kabba, 530 F.3d at 1245. Our sister circuits, in undertaking that inquiry, have outlined three circumstances in which the BIA, while “purporting to apply clear error,” may be “actually review[ing] de novo”:

[A] Board decision that does not address the [IJ]’s key factual findings, gives more weight to certain facts in the record than others, or fails to explain how the [IJ]’s alleged errors showed a

12Notably, like in the immigration context, “the considerations underlying Rule 52(a) . . . all apply with full force in the criminal context.” Maine v. Taylor, 477 U.S. 131, 145 (1986).

21

lack of logic, plausibility, or support in the record suggests the use of a standard of review less deferential than clear error.

F.J.A.P. v. Garland, 94 F.4th 620, 638 (7th Cir. 2024) (first alteration in original) (quoting Brito v. Garland, 40 F.4th 548, 553 (7th Cir. 2022)); see also Soto-Soto v. Garland, 1 F.4th 655, 659–60 (9th Cir. 2021) (similar). We find these guideposts instructive.

Applying these principles, and particularly absent any contrary argument from the government, we conclude the BIA erred by failing to correctly apply the clear-error standard of review in § 1003.1(d)(3)(i). As we will show, the BIA mistakenly undertook something closer to de novo review as to each of the three factual findings at issue: nexus, internal relocation, and the Honduran government’s ability and willingness to protect Ms. Ramos. We consider these in turn.

a

We begin with the IJ’s finding of nexus. Asylum is available only to those who qualify as “a refugee within the meaning of” the Immigration and Nationality Act (INA). 8 U.S.C. § 1158(b)(1)(A). And the INA defines “refugee” to require, as relevant here, “persecution or a well-founded fear of persecution on account of race, religion, nationality, membership in a [PSG], or political opinion.” Id. § 1101(a)(42)(A). The statute clarifies the “on account of,” or nexus, requirement is met when the protected ground “was or will be at least

22

one central reason for persecuting the applicant.” Id. § 1158(b)(1)(B)(i). As relevant to this appeal, the only protected group Ms. Ramos asserted is her membership in Mr. Robles’s nuclear family.

The IJ found “Ms. Ramos’ nuclear relationship to Arturo was one central reason for her persecution in Honduras,” RI.104, satisfying the nexus requirement, because:

• Ms. Ramos cited that relationship as “the ‘main reason’” she thought Jonis was persecuting her, RI.104;

• Jonis’s threats referenced Mr. Robles specifically;

• The threats began only after Mr. Robles’s murder, suggesting “a direct connection between” the murder and Ms. Ramos’s persecution, RI.104;

and • “Jonis’ threats extended to everyone in her proposed social group—

including Arturo’s biological children—and did not include anyone outside of Arturo’s nuclear family,” RI.104.

The IJ then rejected the government’s proffered evidence purportedly showing a lack of a connection between the family-based PSG and the persecution. For example, the IJ reasoned, even if other motives—like “a desire to prevent [Ms. Ramos] from reporting [Jonis’s] actions to the police”—may have played a role, that does not have any bearing on whether “Ms. Ramos’ relationship to Arturo [w]as at least one of the central reasons for the threats.”

23

RI.104–05 (emphasis added). And, while the details surrounding Jonis’s feud with Mr. Robles were not entirely clear, the IJ noted the relevant relationship is between Jonis and Ms. Ramos—and those two lacked any relationship outside of their shared connection to Mr. Robles.

The BIA concluded “the [IJ’s] findings of fact regarding the motivation of” Jonis, RI.5, was clearly erroneous because:

• The IJ “necessarily clearly erred in relying on the ‘timing’ of past harm”

because Ms. Ramos “did not suffer any direct past harm in Honduras,”

RI.6; and • Nothing in Jonis’s “extremely vague” threats indicated the asserted PSG motivated those threats, RI.6.

The BIA rejected the IJ’s reasoning that the PSG more likely motivated the threats because they extended to everyone within the PSG and no one outside it. According to the BIA, since Ms. Ramos did not affirmatively testify that no one outside the PSG had been harmed, the IJ’s inference was “unsupported” by record evidence. RI.6.

Reviewing de novo, see Kabba, 530 F.3d at 1245, we readily conclude the BIA’s analysis failed to comport with the clear-error standard of review. We first agree with both parties that the BIA misconstrued the IJ’s findings in multiple ways. As the government points out, the BIA’s finding regarding timing misunderstands that the IJ’s determination “appears to be based not

24

on the timing between any past harm to herself and the threats she received, but on the time between Arturo’s death and the threats Petitioner received.” Resp. Br. at 21 (citing RI.104). 13 And the BIA similarly misunderstood the IJ’s finding about who Jonis threatened. The BIA believed the IJ to have found Jonis never threatened anyone outside the PSG. That is wrong. As the government correctly explains, the IJ is best understood as finding that the threats Ms. Ramos knew of were aimed at everyone inside, and no one outside, that PSG. See Resp. Br. at 21–22; see RI.102–03 (the IJ concluding the record shows “Ms. Ramos believes that, unlike her children with Arturo, her older sons are not vulnerable to Jonis, because they are not Arturo’s biological children”).

As the government does not dispute, the IJ recited circumstantial evidence tending to suggest the asserted PSG motivated Jonis’s persecution and rejected arguments to the contrary. See INS v. Elias-Zacarias, 502 U.S. 478, 483 (1992) (confirming “direct or circumstantial” evidence of a persecutor’s motive can suffice to establish nexus (emphasis added)). The BIA was required to limit its inquiry to the narrow question of whether the IJ’s nexus findings were “permissible,” Kabba, 530 F.3d at 1245 (quoting Anderson, 470 U.S. at 574), or whether any error was “clear,” Lin, 813 F.3d at 127, or “pellucid[],”

13 Tellingly, the government’s response brief provides some of the most persuasive reasons for concluding the BIA legally erred in its June 2023 order.

25

Ludwig, 641 F.3d at 1247. Instead of engaging in that deferential review, the BIA considered the evidence in the first instance, finding more persuasive that Jonis’s threats were, in its view, “extremely vague.” RI.6. That is tantamount to saying “the [BIA] would have weighed the evidence differently or decided the facts differently had it been the factfinder.” Kabba, 530 F.3d at 1245 (alteration in original) (quoting 67 Fed. Reg. at 54889); accord F.J.A.P., 94 F.4th at 638 (finding the BIA may be “actually review[ing] de novo” and not for clear error when it “gives more weight to certain facts in the record than others” (quoting Brito, 40 F.4th at 553)). At best, the BIA might have thought the IJ’s findings were “possibly or even probably wrong”—but disagreement with a factual finding does not make it clearly erroneous. Ludwig, 641 F.3d at 1247. We therefore conclude the BIA “failed to give deference to the IJ’s findings and improperly engaged in its own fact finding,” exceeding what the clear-error standard prescribes. Kabba, 530 F.3d at 1248.

b

We next turn to the IJ’s finding that Ms. Ramos could not relocate within Honduras. Recall, Ms. Ramos’s asylum claim is based not on past harm but on an asserted “well-founded fear of persecution” in the future. 8 U.S.C. § 1101(a)(42)(A). “Fear of persecution is not well-founded,” we have held, “if the applicant can avoid persecution by relocating to another part of the country and it would be reasonable to expect her to do so.” Ritonga v. Holder, 633 F.3d

26

971, 976–77 (10th Cir. 2011) (citing 8 C.F.R. § 1208.13(b)(2)(ii), (b)(3)). Thus, when asylum is sought based on feared future persecution, applicants like Ms. Ramos must typically show they cannot reasonably relocate within the country to which they would be removed.

The IJ found Ms. Ramos could not reasonably relocate within Honduras.

According to the IJ, the record showed Jonis had “connections to organized crime—including Inestroza and MS-13,” which make Ms. Ramos “vulnerable to persecution everywhere in Honduras.” RI.106. “Given MS-13’s countrywide network of criminal operatives and sophisticated communication systems,” the IJ found, “Ms. Ramos’ risk of persecution is not confined to one location in Honduras.” RI.106. The IJ based its finding about Honduran gangs’ reach on country-condition evidence submitted during the asylum proceeding, including Dr. Young’s report and testimony. The IJ continued, “[T]he fact that Jonis does not fear the police and is well-known within Honduran criminal organizations, indicates that he could personally track down Ms. Ramos on his own.” RI.106. “It follows, therefore, that there is nowhere in Honduras where Ms. Ramos would be safe from Jonis.” RI.106. The IJ found Ms. Ramos’s relationship with Mr. Robles and status as a single mother made her more vulnerable. According to the IJ, that evidence explains why Ms. Ramos, unlike her relatives, could not live safely in Honduras.

27

The BIA concluded the IJ’s internal-relocation finding was clearly erroneous. In the BIA’s view, Ms. Ramos “did not establish that Jonis was in fact a member of an organized criminal group; rather, she testified that she surmised he was an associate of MS-13, because he was always on the streets, did not have a job, and had a brother in MS-13.” RI.7. The BIA understood the record on this point to include nothing “beyond [Ms. Ramos’s] own speculation, that Jonis was a gang member with country-wide connections.” RI.7.

The BIA misapplied the clear-error standard in rejecting the IJ’s finding that Ms. Ramos could not reasonably relocate within Honduras. The BIA inaccurately recounted what the IJ found and then effectively “engaged in its own fact finding.” Kabba, 530 F.3d at 1248. The government provides the best explanation for why the BIA erred here: the IJ “did not base her finding on Jonis being a member of an organized criminal group,” but instead, the IJ found “that Jonis was well-known within Honduran criminal organizations.” Resp. Br. at 22 (emphasis added) (citing RI.106). We agree. The IJ’s finding was based on a permissible view of the record. See RI.106 (the IJ basing her conclusions on “Jonis’ connections to organized crime—including Inestroza and MS-13,” not purported membership in those groups (emphasis added)). The BIA also appeared to overlook other parts of the record referenced by the IJ, including evidence that Jonis “could personally track down Ms. Ramos on his own,” RI.106, and that Ms. Ramos could not safely live with a Honduran

28

relative because her relationship with Mr. Robles and single-mother status made her more vulnerable.

The IJ’s decision squarely relied on record evidence that Ms. Ramos would be vulnerable to Jonis or his gang connections. The BIA’s reasons for finding that evidence insufficiently persuasive are tantamount to, again, reweighing evidence, ignoring parts of the record, finding facts in the first instance, or discrediting permissible inferences, which the BIA may not do on clear-error review. Kabba, 530 F.3d at 1245; accord F.J.A.P., 94 F.4th at 638 (finding the BIA often legally errs when it “does not address the [IJ]’s key factual findings[ or] gives more weight to certain facts in the record than others” (quoting Brito, 40 F.4th at 553)).

c

Finally, we consider the IJ’s finding that the Honduran government was not able or willing to protect Ms. Ramos. We have acknowledged that persecution sufficient to merit asylum may be “inflicted by the government itself, or by a non-governmental group that ‘the government is unwilling or unable to control.’” Wiransane v. Ashcroft, 366 F.3d 889, 893 (10th Cir. 2004) (quoting Batalova v. Ashcroft, 355 F.3d 1246, 1253 (10th Cir. 2004)); see 8 U.S.C. § 1101(a)(42)(A) (establishing that asylum applicants must show they are “unable or unwilling to avail [themselves] of the protection of” their home country). No party contends the Honduran government would be directly

29

involved in the persecution Ms. Ramos fears, so we—like the IJ and BIA—focus on the “unwilling or unable” inquiry.

On this issue, the IJ found “the Honduran government is unable or unwilling to control gang members from perpetrating violence against Ms. Ramos and her children.” RI.105. That is because, based on credible testimony in the record from Ms. Ramos, “when [she] went to tell the police about Jonis’ ongoing threats to harm her and her children, the officers ‘didn’t really listen’ to her; without listening to her complaint, the officers simply said that ‘there was nothing [they] could do.’” RI.105 (second alteration in original). Ms. Ramos testified the reason “the police ‘did not pay attention’ and refused to help her” was “because the officers are ‘tied up to the same group’ as her persecutors,” meaning they were friendly with gangs like Inestroza. RI.105. As the IJ explained, country-condition evidence showed “the Honduran government collaborates with organized crime and fails to protect individuals from gang violence.” RI.105. The IJ found Honduras has high rates of “crime and violence,” including “one of the highest murder rates in the world”—and the government lacks resources and willpower to address those issues. RI.105 (citing sources). The evidence about these conditions, the IJ observed, is consistent with other record evidence. Ms. Ramos testified Jonis did not fear the police, and police officers flatly refused to help her. Likewise, Dr. Young’s report indicated the Honduran government inadequately protects victims of

30

violent crime. The IJ thus concluded, on the basis of this evidence, “[R]egardless of whether Jonis or another MS-13 member target[s] Ms. Ramos, either individually or collectively, the Honduran government does not have the resources or inclination to protect Ms. Ramos and her children.” RI.106.

According to the BIA, the IJ based this unable-or-unwilling finding on what it called Ms. Ramos’s “frustrating interactions with police, and, in a more general fashion, country conditions in Honduras.” RI.8. The BIA conceded “there is evidence in the record of corruption and inefficiency in Honduran law enforcement at all levels,” and in Ms. Ramos’s particular situation, “the police did not pay attention to her, and told her they had no time” to deal with her complaints about Jonis’s threats. RI.8. In the BIA’s view, however, those complaints were “not the sort of threats on which the police could take concrete action.” RI.8. And the record reflects a contrary “country-condition” fact: “the Honduran government actively combats criminal organizations.” RI.8.

Again, what we see here is the BIA’s misapplication of the clear-error standard of review. Crediting some record facts over others is a quintessential example of reweighing evidence on appeal—a function the BIA’s own regulations squarely prohibit. See Kabba, 530 F.3d at 1247 (faulting the BIA for “selectively examin[ing] some evidence while ignoring other evidence presented to it”); accord F.J.A.P., 94 F.4th at 638 (similarly identifying “giv[ing] more weight to certain facts in the record than others” as an indicator

31

of de novo review (quoting Brito, 40 F.4th at 553)). And concluding Ms. Ramos’s complaints are not the sort “on which the police could take concrete action,” RI.8, is simply saying the BIA would have “decided the facts differently had it been the factfinder,” Kabba, 530 F.3d at 1245 (quoting 67 Fed. Reg. at 54889). At most, the BIA has demonstrated that another “view of the evidence” might be “permissible,” but that is not enough to establish clear error. Holdeman, 572 F.3d at 1192.

* * *

We therefore conclude the BIA misapplied the clear-error standard in § 1003.1(d)(3)(i) when evaluating the IJ’s factual findings on nexus, internal relocation, and the government’s ability and willingness to protect.

IV

We next consider Ms. Ramos’s further contention that, when the clear-

error standard is properly applied, the record compels the conclusion that the IJ’s findings were permissible and thus not clearly erroneous. For its part, the government does not dispute the historical facts or suggest further factual development is needed. Still, the government contends we should be predisposed to remand to the BIA for further proceedings. This “court’s role in immigration cases is typically one of review, not of first view,” the government argues, “and ‘agencies should be the primary decision makers over matters

32

which Congress has vested in their authority.’” Resp. Br. at 18 (quoting Mickeviciute v. INS, 327 F.3d 1159, 1164 (10th Cir. 2003)).

We cannot endorse—under the circumstances of this case—the government’s speculation that the BIA “may wish to reconsider” some of its determinations on remand. Resp. Br. at 20, 22, 23. Rather, we agree with Ms. Ramos “there is ‘no need to remand’” “because ‘the BIA [has] given [] a reasoned explanation of its decision’” that we can review in the normal course, Reply Br. at 13 (alterations in original) (quoting Mickeviciute, 327 F.3d at 1163), and “because the correct outcome is beyond doubt,” Reply Br. at 4.

A

It is well settled, when we review BIA decisions, “we can avoid a remand if it would be futile”—that is, if “governing law would ‘“require[ ]” [the agency] to reach a “necessary result.”’” Zapata-Chacon v. Garland, 51 F.4th 1191, 1196 (10th Cir. 2022) (alterations in original) (quoting Gutierrez-Zavala v. Garland, 32 F.4th 806, 810 (9th Cir. 2022)). For instance, “[a] remand is not required where ‘nothing remains for the agency to investigate or explain,’” Granados Arias v. Garland, 69 F.4th 454, 465 (7th Cir. 2023) (quoting Ghebremedhin v. Ashcroft, 392 F.3d 241, 243 (7th Cir. 2004) (per curiam)), or when “the record evidence compels the result that we have reached,” Ghebremedhin, 392 F.3d at 243; see also Hussain v. Gonzales, 477 F.3d 153, 158 (4th Cir. 2007) (finding remand unnecessary when “the result . . . is a foregone conclusion such that

33

remand would amount to nothing more than a mere formality”). The Third Circuit has recognized, “in rare circumstances ‘where application of the correct legal principles to the record could lead only to the same conclusion, there is no need to require agency reconsideration’” before granting a petitioner’s requested relief. Yusupov v. Att’y Gen., 650 F.3d 968, 993 (3d Cir. 2011) (quoting Kang v. Att’y Gen., 611 F.3d 157, 168 (3d Cir. 2010)). That is, “[w]hen the outcome is clear as a matter of law . . . remand [to the BIA] is not necessary” before ruling for the petitioner. Id. (ellipsis in original) (quoting Mahmood v. Gonzales, 427 F.3d 248, 253 (3d Cir. 2005)). Indeed, at oral argument, the government conceded—as these persuasive authorities confirm—remand is not required. See Oral Arg. at 22:50–23:20; 24:44–25:03. When asked, “What would prevent us legally from” reaching Ms. Ramos’s contention that the IJ did not clearly err, the government responded, “I don’t think there’s anything, but under the ordinary-remand rule, [this court] should” remand, as a matter of discretion. Oral Arg. at 23:10–19.

In suggesting remand is the appropriate course, the government relies primarily on INS v. Ventura, 537 U.S. 12 (2002) (per curiam), and Gonzalez v. Thomas, 547 U.S. 183 (2006) (per curiam). Ventura and Thomas instruct that “the ordinary ‘remand’ rule” favors giving agencies the chance to exercise their judgment and correct errors themselves in most cases. Ventura, 537 U.S. at 18; Thomas, 547 U.S. at 187. We recognize the Supreme Court’s exhortations in

34

Ventura and Thomas that, “[g]enerally speaking, a court of appeals should remand a case to an agency for decision of a matter that statutes place primarily in agency hands.” Ventura, 537 U.S. at 16 (emphasis added); see also Thomas, 547 U.S. at 185–87 (similar). But, as applied in the unique circumstances here, Ventura and Thomas are mostly instructive by contrast. See Estrada-Martinez, 809 F.3d at 897 (confirming “[r]emand is not always necessary,” the ordinary-remand rule embodied in Ventura and Thomas notwithstanding (emphasis added)).

In each case, the court of appeals had decided an issue “without giving the BIA the opportunity to address the matter in the first instance in light of its own expertise.” Ventura, 537 U.S. at 17; see Thomas, 547 U.S. at 184–85; see also Ventura, 537 U.S. at 17 (indicating the agency should “make an initial determination”); Thomas, 547 U.S. at 186 (same). Here, of course, the BIA reviewed the IJ’s factual findings for clear error “in the first instance.” The conditions that prompted reversal in Ventura and Thomas—reaching an issue the agency never touched on—are thus absent here. See Almaghzar v. Gonzales, 457 F.3d 915, 923 n.11 (9th Cir. 2006) (“Neither Ventura nor Thomas require us to remand an issue to the agency when the agency has already considered the issue.”); Zhao v. Gonzales, 404 F.3d 295, 311 (5th Cir. 2005)

35

(finding Ventura did not require a remand for the BIA to decide an issue it “ha[d] already rejected”). 14 Similarly unavailing is the government’s reliance on Mickeviciute. See Resp. Br. at 16–19. There, we remanded to give the BIA another chance to flesh out a sparse, six-sentence explanation on a critical issue. Mickeviciute, 327 F.3d at 1162. Here, the BIA’s explanation as to clear error spans several single- spaced pages—more than enough to say the agency has adequately explained its reasoning. The BIA does not need a chance to redo the same inquiry it already fully undertook. The government has offered no persuasive argument to the contrary. 15

14 We acknowledge the generally compelling policy rationale behind Ventura and Thomas: “The agency can bring its expertise to bear upon the matter; it can evaluate the evidence; it can make an initial determination; and, in doing so, it can, through informed discussion and analysis, help a court later determine whether its decision exceeds the leeway that the law provides.” Ventura, 537 U.S. at 17; see also Thomas, 547 U.S. at 186–87 (same). Often, it is true agencies will offer greater expertise. Yet we are just as expert as the BIA when it comes to applying clear-error review, meaning agency expertise provides no additional reason to consider remanding on this record. See Calle v. U.S. Att’y Gen., 504 F.3d 1324, 1330 (11th Cir. 2007) (similarly concluding remand is unnecessary when no further factfinding is needed, and agency expertise would be unhelpful).

15 At times, the government appears to suggest that, because the BIA

never applied the clear-error standard correctly, it never applied it at all. See Resp. Br. at 16, 17 (suggesting this court’s clear-error ruling would be “in the first instance”); Oral Arg. at 17:36–54 (government counsel arguing “the Board should [have] the first crack at applying the clear-error standard” without any “mistaken apprehensions of the immigration judge’s findings of fact”); Oral

36

We recognize that taking the further step of reaching whether the IJ’s factual findings were clearly erroneous diverges from our approach in at least two other cases where we found the BIA exceeded its narrow role in applying clear-error review. In Kabba, after finding “the BIA did not apply th[e] deferential [clear-error] standard,” we vacated the affected parts of the BIA’s decision and remanded to the agency to redo the analysis. 530 F.3d at 1249 (citing Thomas, 547 U.S. at 186; Ramirez-Peyro v. Gonzales, 477 F.3d 637, 641 (8th Cir. 2007)); see also Ramirez-Peyro, 477 F.3d at 641–42 (remanding to the BIA in a similar posture based on the ordinary-remand principles from Ventura and Thomas). Similarly, in Villegas-Castro v. Garland, we found the BIA did not apply clear-error review to the IJ’s credibility determinations. 19 F.4th 1241, 1247–49 (10th Cir. 2021). But we held we should consider whether the IJ’s factual finding at issue was clearly erroneous “only after the Board applies the clear-error standard in the first instance.” Id. at 1249 n.3.

These cases do not persuade us to take a similar approach here, for several reasons. First, we find it significant, as Ms. Ramos correctly points out,

Arg. at 20:05–13 (government counsel arguing “the Board hasn’t actually applied the clear-error standard on an appropriate understanding of the facts in the record”); Oral Arg. at 21:19–34 (government counsel arguing “once the Board properly apprehends what the immigration judge’s findings are, . . . then the Board has something to do: it has to apply the clear-error standard there,” and this court “can’t get ahead of their application of that standard”). That argument is a nonstarter. The BIA undisputedly made clear-error determinations in this case. See RI.5–8.

37

the government does not “defend[] any aspect of the BIA’s decision on appeal.” Reply Br. at 1. Second, we conclude “application of the correct legal principles to the record could lead only to the same conclusion” on clear error review, meaning “there is no need to require agency reconsideration.” Yusupov, 650 F.3d at 993 (quoting Kang, 611 F.3d at 168). There is no indication the analyses in Kabba or Villegas-Castro were similarly clear cut. Third, there is no suggestion either case had then lasted more than a decade with piecemeal litigation before the BIA, as here.

Simply put, the ordinary-remand rule may apply in ordinary cases, but this is not an ordinary case. We therefore have no problem reaching whether the IJ’s findings are clearly erroneous. As we will show, “on remand, governing law would” compel holding the IJ’s findings were permissible and not clearly erroneous. Zapata-Chacon, 51 F.4th at 1196. Thus, “remand . . . would be futile.” Id.

B

According to Ms. Ramos, “the BIA’s decision that the IJ’s factfinding was ‘clearly erroneous’ is indisputably incorrect.” Reply Br. at 3. We agree. Recall, the BIA must review an IJ’s factual findings “only to determine whether [they] are clearly erroneous.” 8 C.F.R. § 1003.1(d)(3)(i). Applying the law on clear- error review here, the record permits only one conclusion: that each of the IJ’s

38

three findings at issue was permissible, and thus not clearly erroneous. 16 See Kabba, 530 F.3d at 1245 (confirming a finding cannot be clearly erroneous so long as it constitutes a “permissible view[] of the evidence” (quoting Anderson, 470 U.S. at 574)).

1

The evidence before the IJ rendered permissible her finding “that the persecution Ms. Ramos suffered and continues to fear in Honduras is on account of her membership in Arturo’s nuclear family.” RI.105. This finding is

16 As discussed, whether the BIA exceeded clear-error review is a legal

question reviewed de novo. Kabba, 530 F.3d at 1245. Normally, “[w]e review the BIA’s legal determinations de novo, and its findings of fact under a substantial-evidence standard.” Id. at 1244 (quoting Niang v. Gonzales, 422 F.3d 1187, 1196 (10th Cir. 2005)). We have never decided whether the BIA’s determination that the IJ’s factual findings are clearly erroneous is legal or factual in nature. On one hand, Ms. Ramos seems to suggest that determination is factual. See Reply Br. at 5 (describing the substantial- evidence standard, applicable to “questions of fact,” in analyzing this issue). On the other hand, perhaps whether an ultimate finding from a set of historical facts is clearly erroneous is a legal question. Cf., e.g., Xue v. Lynch, 846 F.3d 1099, 1104 (10th Cir. 2017) (observing “the BIA has specifically determined that the ultimate resolution whether a given set of facts amount to persecution is a question of law reviewed de novo,” but this court treats even that “ultimate resolution” as a factual determination).

But we need not consider this further. Even under the substantial-

evidence standard, we would still reject the BIA’s determination and find the IJ’s factual findings were permissible. The record supports only one conclusion, meaning “the record demonstrates that any reasonable adjudicator would be compelled to conclude” the IJ’s findings were, in fact, not clearly erroneous. Niang, 422 F.3d at 1196 (quoting Yuk v. Ashcroft, 355 F.3d 1222, 1233 (10th Cir. 2004)). That suffices to reject the BIA’s determination, and to accept the IJ’s findings, under either possible standard.

39

plainly a “permissible view[] of the evidence”—all that is required, Kabba, 530 F.3d at 1245 (quoting Anderson, 470 U.S. at 574)—in light of Ms. Ramos’s credible testimony about why Jonis targeted her, Jonis’s references to Mr. Robles in his threats, the fact that the threats began only after Mr. Robles’s murder, and the fact that Ms. Ramos had heard the threats directed at everyone within the asserted PSG and no one outside it. The BIA concluded the IJ’s “findings concerning the motivation of Jonis are based on speculation and on assumptions that are not supported by evidence in the record,” RI.6, but the record shows the IJ relied on plenty of circumstantial evidence, see Elias-Zacarias, 502 U.S. at 483 (clarifying circumstantial evidence can show nexus). In light of that evidence, a “reviewing court” cannot be “left with the definite and firm conviction that” the IJ’s nexus finding was “a mistake.” Kabba, 530 F.3d at 1245 (quoting U.S. Gypsum Co., 333 U.S. at 395).

2

We further conclude the IJ’s finding that Ms. Ramos could not reasonably relocate internally because “there is nowhere in Honduras where [she] would be safe from Jonis” or the gangs to which he is connected, RI.106, was permissible and not clearly erroneous. The IJ based her conclusion that the relevant gangs enjoy a nationwide reach on ample country-condition evidence, including from Dr. Young, whose qualifications the government does

40

not dispute. The IJ also observed Jonis’s threats, gang connections, and attitude of impunity toward police mean he might track her down and harm her himself. And she found Ms. Ramos’s status as a single mother and relationship with Mr. Robles, with whom gang members had a feud, make her especially vulnerable. Again, this record compels the conclusion that the IJ’s finding—that Ms. Ramos could not reasonably relocate within Honduras—is “a permissible view[] of the evidence” and thus not clearly erroneous. Kabba, 530 F.3d at 1245 (quoting Anderson, 470 U.S. at 574).

3

We also have no trouble concluding the IJ’s factual finding on the Honduran government’s inability and unwillingness to protect was permissible and not clearly erroneous. As the government points out, Ms. Ramos credibly testified “that she was never given an opportunity to even tell the police about Jonis’ threats because the police ‘didn’t give [her] time to speak anything.’” Resp. Br. at 23 (alteration in original) (quoting RI.144). The leap from Ms. Ramos not being heard by police to her not being sufficiently protected by police is not large. And the IJ also relied on other facts the BIA downplayed or ignored, like Honduran law enforcement’s friendly ties with gangs, evidence of

41

police indifference or ineffectiveness from Honduras’s high crime rates, and Jonis’s own lack of fearing police. 17

* * *

We thus hold, when the correct standard of review is applied, the record compels only one conclusion—that the IJ’s findings on nexus, internal relocation, and the government’s ability and willingness to protect were permissible and not clearly erroneous. That means remanding for any further substantive proceedings “would be futile.” Zapata-Chacon, 51 F.4th at 1196.

V

Given our conclusion, nothing remains but to reinstate the IJ’s 2019 order granting asylum. The BIA fully considered a completely developed record, comprising two IJ hearings, and explained its reasoning at some length. The government does not suggest the record needs to be further developed. The government has made no argument for any other form of relief—only advancing the argument, rejected above, that we should remand to give the

17 Moreover, though the IJ did not rely on this fact in this section, we

note Ms. Ramos credibly testified that the police “did not do anything” to track down Jonis after he murdered someone. RI.102. We struggle to imagine better evidence of police being unable or unwilling to help protect against Jonis. See Kabba, 530 F.3d at 1245 (confirming the “reviewing court” is to look to “the entire evidence” when reviewing for clear error (quoting U.S. Gypsum Co., 333 U.S. at 395)).

42

BIA “an opportunity to reconsider its clear error determinations.” 18 Resp. Br. at 20. Of course, reconsideration is not an appellate remedy. The BIA’s errors may have provided a valid basis for a motion to the agency to reconsider its decision. See 8 C.F.R. § 1003.2(a)–(b). But the time to make that motion has long passed. See id. § 1003.2(b)(2) (limiting these motions to “30 days after the mailing of the Board decision”).

18 After the second IJ order granting asylum, the government made five

arguments to the BIA: each of the three new factual findings is clearly erroneous, the asserted PSG is not cognizable, and future persecution is not sufficiently likely. The BIA agreed with the government on the three clear- error arguments, which sufficed to reject asylum, so the agency did not reach those remaining two arguments. But the government does not suggest to this court it is still pursuing those two arguments, even though it was fully on notice that Ms. Ramos sought to “finally resolve Petitioners’ asylum proceedings” in this appeal. Op. Br. at 44; see also Op. Br. at 12, 47 (requesting that we “end Petitioners’ decade-long asylum proceedings”). It is even unclear whether these two other claims of error were adequately presented to the BIA, as the arguments were not made in the government’s notice of appeal, beyond a broad statement that it “reserves the right to raise additional issues or claims or error in its brief” to the BIA. RI.96; see 8 C.F.R. § 1003.3(b) (“The party taking the appeal must identify the reasons for the appeal in the Notice of Appeal . . . in order to avoid summary dismissal . . . .”). And the cognizability argument appears to have been available to, but not raised by, the government in its first BIA appeal. See RI.609 (the government conceding in its first notice of appeal “a family group can constitute a PSG,” and not raising that argument); RI.563–81 (the government’s first BIA brief, not making this argument).

43

Recall, the IJ first granted asylum on December 2, 2014. Ms. Ramos now asks this court to “end Petitioners’ decade-long asylum proceedings.” 19 Op. Br. at 12, 47. We will do so. We agree with the Eighth Circuit that “[n]o immigrant should have to live over ten years with the uncertainty as to whether she can stay in this country or not.” Mayo v. Ashcroft, 317 F.3d 867, 875 (8th Cir. 2003); see also Castañeda-Castillo v. Holder, 676 F.3d 1, 3 (1st Cir. 2012) (per curiam) (recognizing a case’s “unusually prolonged and convoluted history” can justify even “extraordinary step[s]” to end a petitioner’s “long ordeal” (quoting Castañeda-Castillo v. Holder, 638 F.3d 354, 363 (1st Cir. 2011))). The government seems to suggest the unusually lengthy asylum proceedings in this matter inured to Ms. Ramos’s benefit, as “[a]sylum is a purely discretionary form of relief from removal.” 20 Resp. Br. at 18 (citing 8 U.S.C. § 1158(b)(1)(A)).

19 We must note the government’s role in prolonging some of the proceedings in this court. Recall, the government’s motion to remand was filed untimely without explanation. See 10th Cir. R. 27.3(A)(3)(b) (explaining this motion had to be filed “within 14 days after the appeal or other proceeding is docketed in this court, unless good cause is shown for later filing”). Because of this late-filed motion, appellate resolution of Ms. Ramos’s petition for review was delayed by many months.

20 We recognize that whether a petitioner is entitled to asylum relief

is a discretionary decision vested in the Attorney General, and by regulation, delegated to IJs. 8 U.S.C. § 1158(b)(1)(A) (granting discretion to the Attorney General); 8 C.F.R. §§ 1003.10, 208.14(a) (delegating the Attorney General’s discretionary authority to IJs). For this reason, we generally remand a finding of asylum eligibility to the agency so the Attorney General can exercise her discretion to grant asylum in the first

44

In the government’s view, Ms. Ramos and her children “ha[ve] been afforded the privilege of being able to reside in the United States during this time even though [Ms. Ramos] is removable as charged.” Resp. Br. at 19. There is no windfall here for Ms. Ramos. We reject any contrary notion.

Given our conclusion in this case, “[r]emand” for any further agency-level determinations would be “a gross waste of time for an agency that can ill afford such impositions.” Rangel-Fuentes v. Garland, 99 F.4th 1191, 1206 (10th Cir. 2024), reh’g granted and opinion vacated on other grounds, No. 23-9511, 2024 WL 3405079 (10th Cir. July 10, 2024).

VI

We GRANT the petition for review, VACATE the BIA’s June 2023 order, and REMAND with instructions to reinstate the IJ’s April 2019 grant of asylum. We DENY the government’s motion to remand as moot.

instance. But here, where the IJ has unequivocally exercised her discretion favorably by twice granting Ms. Ramos asylum relief, and absent contrary argument by the government on this point, we harbor no hesitation in ordering the reinstatement of the IJ’s grant of asylum.