Restrepo Castano v. Bondi: Partial but “Fruitful” State Protection and the “Unwilling or Unable” Standard in Gang-Based Asylum Claims

I. Introduction

In Restrepo Castano v. Bondi, No. 24‑2117 (1st Cir. Nov. 26, 2025), the First Circuit denied a Colombian family’s petition for review of the Board of Immigration Appeals’ (BIA) decision rejecting their applications for asylum and withholding of removal. The case turns on a central and recurrent question in modern asylum law: when is a foreign government deemed “unable or unwilling” to protect an applicant from persecution by non-state actors, such as criminal gangs?

The petitioners—Jhon Eduardo Restrepo Castano, his wife, and their child—fled Colombia after receiving repeated death threats from the notorious Gulf Clan, a powerful criminal organization. They claimed that the Colombian authorities were effectively incapable of protecting them, despite some police measures. Both the Immigration Judge (IJ) and the BIA concluded that the Colombian state response, though imperfect, was sufficiently “fruitful” to show that the government was both willing and able to protect them. The First Circuit, applying the highly deferential substantial evidence standard, upheld that conclusion.

This commentary explores the opinion’s factual background, legal framework, and reasoning, and situates it within the First Circuit’s broader “unwilling or unable” jurisprudence. It also explains the implications of the Court’s emphasis on “fruitful” state action, the burden to report ongoing threats, and the limits of asylum protection when governments confront powerful criminal elements.

II. Factual and Procedural Background

A. The Petitioners and the Gulf Clan Threats

Castano, born in Medellín, Colombia, opened a bakery in Don Matías in 2019. Beginning in 2020, he received phone calls from members of the Gulf Clan (Clan del Golfo), a violent criminal enterprise involved in drug trafficking and money laundering. The callers demanded monetary and political support and threatened to kill his wife and child if he did not comply.

In September 2021, the threats escalated: two armed Gulf Clan members, wearing motorcycle helmets, entered the bakery and delivered a direct, in-person death threat—comply with their demands or be killed, or leave the area. Castano filed a complaint with the police.

The Colombian police took several steps:

They assigned a police guard who stayed at the bakery during working hours until Castano went home.
They opened an investigation, though Castano never received a formal update on its results.
They “blocked” the phone lines of Castano and his family to prevent further calls from the known threatening numbers.

After the guard was posted, there were no further in-person threats. However, the Gulf Clan continued making telephonic death threats from new numbers. Castano testified that he felt unsafe and decided to flee with his family to the United States in March 2022. Notably, he acknowledged that neither he nor his family was ever physically harmed by the gang.

B. Removal Proceedings and IJ Decision

After entering the United States without inspection, the family was served with Notices to Appear, initiating removal proceedings. Castano applied for:

Asylum
Withholding of removal
Protection under the Convention Against Torture (CAT)

His wife and child were derivative beneficiaries of the asylum application. The IJ found his testimony credible and generally consistent with his affidavit and supporting evidence, including a letter from the police chief. However, the IJ denied all requested relief.

Key IJ findings:

The Gulf Clan is a private entity, not a governmental or quasi-governmental actor (Castano’s “quasi-government” theory was explicitly rejected).
Because the persecutor was private, Castano had to show that the Colombian government was unwilling or unable to protect him.
The Colombian government was clearly willing, as evidenced by assigning an on-site police guard and opening an investigation.
The government was also able to protect him:
- The in-person threats stopped entirely after police intervention.
Because he failed to show a sufficient “government nexus” to persecutory harm, Castano did not meet the requirements for asylum or for withholding of removal.
The IJ also denied CAT protection, a ruling Castano did not challenge on appeal.

C. BIA Decision

The BIA:

Adopted and affirmed the IJ’s decision, dismissing the appeal.
Found no clear error in the IJ’s determination that Colombia was both willing and able to protect Castano and his family.
Emphasized that the police’s stationing of a guard and other actions showed a concerted response, even if not completely successful.
Agreed that failure to establish eligibility for asylum meant that the more demanding standard for withholding of removal could not be met.

Castano then filed a timely petition for review in the First Circuit.

III. Summary of the Opinion

The First Circuit, in an opinion by Judge Rikelman, denied the petition for review. The Court’s principal holdings are:

The “unwilling or unable” inquiry is factual and reviewed for substantial evidence.
The Court reaffirmed that whether a government is unwilling or unable to control private persecutors is a question of fact, subject to highly deferential review.
Substantial evidence supports the finding that Colombia was able to protect Castano.
The Court emphasized:
- The police promptly assigned an on-site guard and opened an investigation.
- The in-person threats ceased entirely after that intervention.
- The telephone threats were not acted upon and there was no evidence of attempts to carry them out.
- The record was unclear whether Castano reported the post-guard phone threats; he bears the burden to show government inaction or futility of reporting, and he had not done so.
Incomplete success in combating crime does not establish governmental inability.
The Court reiterated that limited success against criminal organizations is not, by itself, sufficient to show that a government is unable to protect an asylum applicant.
The BIA applied the correct standard of review and did not “rubber-stamp” the IJ’s decision.
The BIA’s clear-error review of the IJ’s unwilling-or-unable finding was appropriate, and its brief decision adequately reflected its reasoning and consideration of the continued telephonic threats.
Because there was no government nexus, the Court did not reach whether the threats rose to the level of “persecution.”
Invoking INS v. Bagamasbad and SEC v. Chenery Corp., the Court confined its review to the dispositive ground relied upon by the agency.

As a result, the petitioners remain ineligible for asylum and withholding of removal.

IV. Legal Framework

A. Asylum and the Definition of “Refugee”

To obtain asylum under 8 U.S.C. § 1158(b)(1)(A), an applicant must qualify as a “refugee” under 8 U.S.C. § 1101(a)(42)(A). A refugee is a person who is unable or unwilling to return to their country because of:

“persecution or a well-founded fear of persecution on account of race, religion, nationality, membership in a particular social group, or political opinion.”

Key elements include:

Persecution: Serious harm (or threat of such harm), beyond mere harassment.
Nexus: The persecution must be “on account of” a protected ground.
Government involvement: The harm must be inflicted by the government or by private actors whom the government is unwilling or unable to control.
Standard of proof: A “well-founded fear” is a relatively generous standard; even a moderate chance of persecution can suffice.

B. Withholding of Removal

Withholding of removal, 8 U.S.C. § 1231(b)(3)(A), provides that an individual may not be removed to a country where their life or freedom would be threatened on account of a protected ground. The elements parallel asylum, but:

The standard of proof is higher: the applicant must show a “clear probability” of persecution (more likely than not), a stricter standard than a “well-founded fear.”
The applicant’s subjective fear is not relevant; the focus is on the objective likelihood of harm.

As the Court noted, an applicant must prove, among other things:

A threshold level of serious harm (past or future), and
A government nexus—harm involving government action or inaction, including a showing that the government is unwilling or unable to control private persecutors.

Because the government-nexus requirement overlaps with asylum, failure to establish it for asylum typically defeats withholding claims as well.

C. Government Nexus and Private Actors

When the persecutor is a private actor (like a gang), the applicant must show that:

“the government is unwilling or unable to control that private conduct.” (Khan v. Holder, 727 F.3d 1, 7 (1st Cir. 2013))

This “unwilling or unable” test appears both in statutory text (via the refugee definition) and case law. Its practical applications include:

If the government cooperates with the persecutors, or refuses to intervene out of bias or corruption, it is “unwilling.”
If the government tries to provide protection but lacks capacity—e.g., no functioning police, rampant impunity—it may be “unable.”
If the government takes meaningful steps that halt or significantly mitigate harm, it is generally considered both willing and able.

Thus, evidence of police reports, official responses, investigations, and the effectiveness of those measures is critical.

D. Standard of Review

The Court reaffirms two important review principles:

Factual findings (e.g., whether the government is “unwilling or unable” to protect) are reviewed under the substantial evidence standard. The Court must uphold such findings unless:
“any reasonable adjudicator would be compelled to conclude to the contrary.” (Barnica-Lopez v. Garland, 59 F.4th 520, 527 (1st Cir. 2023))
The BIA reviews an IJ’s factual findings for clear error, not de novo. Legal conclusions are reviewed de novo.

The “unwilling or unable” determination falls squarely on the factual side of this divide, a point that plays a central role in the Court’s analysis and in rejecting the petitioner’s procedural challenges.

V. Detailed Analysis

A. The “Unwilling or Unable” Question as Purely Factual

The Court expressly states:

“Whether a government is unwilling or unable to protect an asylum applicant from harm caused by a private actor is a question of fact.” (citing Medina-Suguilanda v. Garland, 121 F.4th 316, 322–23 (1st Cir. 2024))

This has two immediate consequences:

The IJ’s determination is entitled to strong deference.
The BIA’s review is limited to clear-error review of the IJ’s factual findings on this point; the BIA may not substitute its own view of the facts simply because it would weigh them differently.

The Court thus rejects the suggestion that the BIA was required to independently reassess (de novo) whether the undisputed facts established governmental inability. Instead, as long as the IJ’s finding is supported by “reasonable, substantial, and probative evidence,” the agency’s conclusion stands.

B. Government Responses and the Concept of “Fruitful” Protection

A central theme in the Court’s reasoning is that imperfect state protection can still be legally sufficient. The opinion draws on a line of First Circuit cases that focus on whether government responses are “fruitful,” not whether they completely eradicate the threat.

1. Key precedents: Khan, Burbiene, Ortiz-Araniba, Singh

Khan v. Holder, 727 F.3d 1 (1st Cir. 2013).
The Pakistani police:
- Investigated reported threats and violence.
- Made arrests.
- Called in the army to secure the petitioner’s area.
Despite threats not being fully “eradicated” and some investigative shortcomings, the Court held that these actions were sufficiently “fruitful” to show that Pakistan was able to protect the petitioner.
Burbiene v. Holder, 568 F.3d 251 (1st Cir. 2009).
The Court stressed that a government’s general difficulty in combating crime—“struggles to combat a criminal element”—does not equate to an inability to protect this applicant. Limited success in curbing crime is not enough to show inability.
Ortiz-Araniba v. Keisler, 505 F.3d 39 (1st Cir. 2007).
The Court memorably observed:
“no government c[an] provide the sort of absolute protection [the petitioner] seeks.”
This principle is explicitly invoked against Castano’s apparent desire for 24-hour, personal police protection.
Singh v. Garland, 87 F.4th 52 (1st Cir. 2023).
In Singh, government action was deemed “fruitful” because persecutors fled after the army was deployed, even though there were no arrests. This further confirms that arrests or convictions are not a prerequisite for finding that the government is able to protect.

2. Application to Castano’s case

In light of these precedents, the Court characterizes Colombia’s response as “fruitful”:

The police responded immediately to the in-person threat by sending a guard to the bakery.
In-person threats stopped entirely after the guard’s arrival.
The police took an additional step of blocking the known threatening phone numbers.
There was no evidence that the Gulf Clan tried to carry out the telephonic threats by physical attacks after the guard was posted.

Although the state did not fully eliminate telephonic threats, the Court sees this as akin to the situations in Khan, Burbiene, and Singh: governments confronting serious criminal organizations but making meaningful protective efforts that substantially reduce or prevent harm.

The Court thus concludes that the agency’s finding—that Colombia was able to protect Castano—was supported by substantial evidence. In other words, a reasonable adjudicator could certainly view this response as adequate, even if another adjudicator might be more sympathetic to the petitioner.

C. Reporting Requirement and the (Unused) Futility Exception

The opinion also underscores that asylum applicants bear the burden of showing that the government either did not act or could not act in response to their reports. Citing Morales-Morales v. Sessions, 857 F.3d 130, 135 (1st Cir. 2017), the Court notes:

“the burden of showing the requisite government . . . inaction falls on the asylum applicant.”

In Castano’s case, the record was:

Clear that he reported the initial in-person and telephonic threats.
Unclear or inconsistent about whether he reported the continued phone threats after the guard was posted.

During cross-examination, Castano admitted he could not recall whether he reported those continued calls. The Court points out that:

If he did not report them, it is hard to argue that the government was unable to protect him from something it did not know about.
If he did report them, but the record is ambiguous or silent about the state’s response, that ambiguity does not compel a contrary conclusion under substantial evidence review.

There is a recognized exception where victims need not report if doing so would be futile or dangerous. But:

Castano did not advance a “futility” argument.
Even if he had, the Court suggests the record would not compel such a finding because the initial reporting did yield a fast and meaningful police response (the guard and phone blocking).

This aspect of the opinion implicitly instructs future applicants and their counsel: if you contend the government is unable or unwilling to protect you, you must either:

Show that you reported continued or escalating threats and that the state’s response was effectively illusory, or
Provide a strong, well-documented futility/danger argument explaining why further reporting would have been pointless or too risky.

D. Precedents Cited and Their Role

1. Core First Circuit “unwilling or unable” cases

In addition to Khan, Burbiene, and Ortiz-Araniba, the Court situates its analysis in several recent cases:

Medina-Suguilanda v. Garland, 121 F.4th 316 (1st Cir. 2024) – cited for the proposition that the “unwilling or unable” determination is a question of fact and for the importance of whether local authorities responded promptly to each incident.
Gómez-Medina v. Barr, 975 F.3d 27 (1st Cir. 2020) – quoted for the principle that “the most telling datum” is whether authorities “responded immediately to each incident.” This frames the Colombian police’s swift posting of a guard as particularly significant.
Singh v. Garland, 87 F.4th 52 (1st Cir. 2023) – confirms that government action can be deemed “fruitful” even without arrests, as long as persecutors desist or flee in response.
Barnica-Lopez v. Garland, 59 F.4th 520 (1st Cir. 2023) and Khalil v. Garland, 97 F.4th 54 (1st Cir. 2024) – used for the substantial evidence standard, underscoring the Court’s reluctance to disturb agency fact findings except in extreme cases.

2. Rosales Justo v. Sessions and its Limits

Castano relied on Rosales Justo v. Sessions, 895 F.3d 154 (1st Cir. 2018), where the Court found that Mexico was unable to protect a family targeted by a criminal organization that murdered the petitioner’s son and continued to pursue the family despite internal relocation. There, the Court concluded the IJ’s finding of governmental inability was not clearly erroneous and that the BIA had erred by overturning it.

The panel distinguishes Rosales Justo on two main grounds:

Procedural posture and standard of review – In Rosales Justo, the Court was reviewing whether the BIA had erroneously reversed the IJ’s factual findings. Here, by contrast, the first-level fact finder (the IJ) found that the government was able to protect Castano, and the BIA affirmed. The First Circuit’s task is thus to see if the record compels the opposite conclusion—which is much harder to show.
Factual severity and investigation quality – In Rosales Justo, the evidence indicated:
- A son had actually been murdered shortly after receiving threats.
- The family was pursued even after relocating.
- The record affirmatively suggested that the investigation would not make the family safer.
In Castano’s case:
- There was no physical harm to any family member.
- Threats did not escalate into attacks after the guard arrived.
- The record was silent, not negative, about the investigation’s quality.

Thus, the Court treats Rosales Justo as consistent with, not contrary to, its current decision: where state responses are demonstrably ineffective, inability may be found; where responses are meaningfully protective even if incomplete, substantial evidence can support a finding of ability.

3. Out-of-circuit cases: J.R. v. Barr and others

Castano cited the Ninth Circuit’s decision in J.R. v. Barr, 975 F.3d 778 (9th Cir. 2020), for the proposition that an applicant need not wait for death threats to be carried out to qualify for asylum. The First Circuit finds J.R. inapposite:

Issue difference: J.R. primarily involved the government’s unwillingness to protect the petitioner after previously affording special protection that was then withdrawn, not its ability per se.
Fact difference: The petitioner in J.R. had:
- Two fingers cut off.
- Been shot seven times (losing his lung).
- Lost his son to murder on his front porch.
- Seen relatives murdered after he relocated.

By contrast, Castano and his family were never physically harmed, and threats diminished after police intervention. The Court notes that other circuits have declined to extend J.R. to more typical gang-threat cases (Aguilar v. Garland, 9th Cir. 2024 (unpublished); Hidalgo-Nunez v. Garland, 10th Cir. 2022 (unpublished)). The First Circuit likewise declines to treat J.R. as controlling or even analogous in an “ability” case with much less dire facts.

4. Administrative law principles: Bagamasbad and Chenery

The Court also relies on two foundational administrative law decisions:

INS v. Bagamasbad, 429 U.S. 24 (1976) – Agencies are not required to decide issues that are unnecessary to the outcome. Here, once the IJ and BIA found no government nexus, they had no obligation to decide whether the threats constituted “persecution.”
SEC v. Chenery Corp., 332 U.S. 194 (1947) – A reviewing court must judge the propriety of agency action solely by the grounds invoked by the agency. Consequently, the First Circuit confines itself to evaluating the government-nexus rationale rather than reaching arguments the agency never addressed.

These principles reinforce the narrowness of the Court’s holding: it is about the government-nexus element, not the overall severity of harm.

E. Procedural Challenges to the BIA’s Decision

Castano made two procedural arguments:

The BIA should have reviewed the “unwilling or unable” issue de novo rather than for clear error, and instead impermissibly “rubber-stamped” the IJ’s factual finding.
The BIA “turned a blind eye” to the fact that the telephonic death threats continued despite some police protection, thus committing a legal error by failing to consider important evidence.

The Court rejects both challenges.

1. Standard of review at the BIA

Reiterating Medina-Suguilanda, the Court explains that the “unwilling or unable” determination is factual and therefore subject to clear-error review by the BIA. There was no error in the BIA’s approach. To the extent Castano argued for de novo review of this mixed question, that argument is inconsistent with controlling First Circuit precedent.

2. Adequacy of the BIA’s reasoning

The Court notes that both the IJ and the BIA explicitly acknowledged:

The existence of continuing telephonic threats, and
The absence of further in-person threats after the guard was stationed.

Quoting Lopez Perez v. Holder, 587 F.3d 456, 460–61 (1st Cir. 2009), the Court emphasizes that:

“[A]n agency’s decision must illuminate the path of its reasoning, but it need not do so at great length or in exquisite detail,” especially where the BIA adopts the IJ’s reasoning.

The BIA’s brief but focused treatment—emphasizing that where a government makes a concerted effort to combat a particular problem, limited success does not equal inability—sufficed to reveal its “decisional calculus.” Thus, there was no procedural or legal error.

F. Scope and Limits of the Holding

It is important to be precise about what Restrepo Castano does—and does not—decide:

The Court does not decide whether the Gulf Clan’s threats constituted “persecution” under asylum law. It assumes arguendo that they are serious but holds that the government-nexus element is dispositive.
The Court does not revisit the IJ’s rejection of the theory that the Gulf Clan is a quasi-governmental entity, nor the denial of CAT protection. Those issues were not pressed on appeal.
The holding is limited to the facts of this case and the substantial evidence standard. The Court does not create a categorical rule that protection is always adequate when police post guards or block phone numbers. Rather, it holds that, given this record, a reasonable adjudicator need not have concluded that Colombia was unable to protect.
The opinion reinforces and systematizes existing First Circuit law: meaningful, prompt, and at least partially successful protective efforts can establish governmental “ability,” even when threats or crime persist at some level.

VI. Impact and Implications

A. Consequences for Gang-Based Asylum Claims

For asylum seekers fleeing gangs, especially from countries like Colombia, El Salvador, Honduras, and Mexico, Restrepo Castano underscores several points:

Evidence that the police responded quickly and took concrete steps (e.g., assigning guards, conducting operations, blocking numbers) will weigh heavily in favor of a finding that the government is able to protect.
The mere continuation of threats—especially if not acted upon physically—may not suffice to show governmental inability, absent evidence that the authorities were informed and failed in some meaningful way.
Applicants and counsel must pay close attention to the timeline of reporting and responses: whether threats were reported, what steps the police took, and what happened thereafter.

B. Burden of Proof and Documentation Strategy

Practically, the decision sends a clear message to practitioners:

Document every police report or attempt to report: dates, locations, officers’ names, and any written complaints or case numbers.
Document the shortcomings: lack of follow-up, dismissive responses, corruption, or refusal to help.
If further reporting is too dangerous or pointless, develop a robust futility argument with country conditions reports, expert affidavits, or prior instances of police complicity or inaction.
Explain the causal chain—how, despite any facially protective steps (like assigning a guard), the persecutors could still act with impunity.

In the absence of such evidence, courts are likely to regard at least partially effective state responses as “fruitful” and thus as evidence of governmental ability, as in this case.

C. BIA Decision-Writing and Standards of Review

For the BIA, the opinion is a reminder that:

It must apply clear-error review to “unwilling or unable” findings.
Even concise decisions must show that the BIA has:
- Identified the key factual findings of the IJ.
- Recognized the relevant legal framework.
- Explained, however briefly, why the IJ’s view is not clearly erroneous.

Restrepo Castano suggests that relatively short BIA decisions can be sufficient when they explicitly endorse and rely on the IJ’s reasoning and cite controlling precedent.

D. Doctrinal Consolidation of “Fruitful” State Action

Perhaps the most enduring doctrinal contribution of Restrepo Castano is the consolidation of the idea that:

Partial, “fruitful” state protection—showing real efforts and discernible results—precludes a finding that the government is “unable” to protect, even when criminal actors continue to pose some risk.

This aligns with earlier First Circuit cases but applies the principle crisply to a situation involving a prominent transnational gang and modern threats (phone calls) rather than exclusively in-person violence. It also clarifies that the lack of an arrest or conviction does not, by itself, undermine a finding of state ability.

VII. Complex Concepts Simplified

A. “Persecution” vs. Threats or Harassment

In everyday terms, “persecution” means something more than unpleasantness or generic danger. It generally involves:

Serious harm—physical, psychological, or severe economic harm,
Or credible, targeted threats of such harm,
Directed at you because of your race, religion, nationality, political opinion, or membership in a particular social group.

Here, the Court did not decide whether the threats alone would qualify as “persecution” because the case was resolved on the separate government-nexus ground.

B. “Government Nexus” and Private Actors

Asylum does not protect people from all bad things that happen in their home countries. When persecutors are private actors (like gangs or criminals), you must show:

You asked the government for help (or can explain convincingly why you could not), and
The government either:
- Refused to help (unwilling), or
- Tried but could not help in any meaningful way (unable).

If police respond quickly, provide protection that makes a difference, and at least partly succeed in preventing harm, courts will often conclude the government is both willing and able to protect you.

C. Substantial Evidence Standard

“Substantial evidence” review is very deferential. The Court does not ask:

“Do we personally believe the government was unable to protect?”

Instead, it asks:

“Could a reasonable fact finder look at this record and reach the same conclusion as the IJ and BIA?”

The petitioner must show that the evidence is so one-sided that any reasonable adjudicator would be compelled to find the opposite. That is a very high bar.

D. “Well-Founded Fear” vs. “Clear Probability”

There are two main standards of proof in protection cases:

Asylum – “Well-founded fear”
Even a less-than-50% chance of being persecuted may be enough if the fear is objectively reasonable.
Withholding of removal – “Clear probability”
You must show it is more likely than not that you will be persecuted if returned.

If you cannot meet the lower “well-founded fear” standard for asylum (because, for example, you cannot show a government nexus), you automatically cannot meet the higher standard for withholding.

E. Derivative Beneficiaries

In asylum, spouses and children can often be added as “derivative beneficiaries.” This means:

The principal applicant’s eligibility is what matters.
If the principal wins asylum, the derivative family members generally do as well.
If the principal loses (as here), the derivatives’ applications fail too, unless they have their own independent claims.

VIII. Conclusion

Restrepo Castano v. Bondi reinforces a consistent First Circuit approach to the “unwilling or unable” standard: where the government responds quickly and meaningfully to threats from private actors—by assigning protection, launching investigations, and reducing or halting in-person threats—courts will generally view that response as evidence of governmental “ability,” even if some threats persist and even if no arrests are made.

The decision emphasizes:

The fact-intensive nature of the “unwilling or unable” inquiry and the strong deference owed to IJ and BIA fact-finding.
The importance of documented reporting of threats and the narrowness of the “futility” exception.
The idea that partial, “fruitful” protection suffices to defeat claims of governmental inability, aligning this case with earlier First Circuit precedents like Khan, Burbiene, and Ortiz-Araniba.
The limited role of courts under Bagamasbad and Chenery: they review agency decisions on the grounds actually invoked, here the lack of a government nexus, and do not reach unnecessary issues such as whether certain threats alone would count as persecution.

For future asylum seekers, particularly in gang-related claims, the case underscores that success will often depend less on the gravity of threats alone and more on demonstrating that the home state’s protection was either non-existent, corrupt, or manifestly ineffective, despite being properly informed—a demanding evidentiary burden that Restrepo Castano shows was not met on this record.