Introduction

In Villalta Martinez v. Bondi, the United States Court of Appeals for the Second Circuit issued a published, precedential opinion clarifying how the Board of Immigration Appeals (BIA) must apply clear-error review when overturning an Immigration Judge’s (IJ’s) grant of protection under the Convention Against Torture (CAT). The court held that the BIA’s conclusory characterization of the record as “generalized” and “anecdotal” was insufficient to justify a finding of clear error, particularly where the IJ had credited expert testimony that was corroborated by multiple country-conditions sources and where the record included substantial evidence of government intent and systemic abuses under El Salvador’s ongoing “State of Exception.”

Petitioner Jose Saul Villalta Martinez, a Salvadoran national with tattoos and a U.S. criminal record, sought CAT protection based on (1) the risk of torture by MS-13 and (2) a greater risk of arrest, detention, and torture by Salvadoran authorities under the State of Exception, which has led to mass arrests, suspensions of due process, deadly prison conditions, and reports of torture. An IJ granted CAT relief. The BIA vacated, finding the IJ’s conclusion “clearly erroneous,” and later denied reconsideration. The Second Circuit granted the petition for review, remanded to the BIA for further consideration, and dismissed the consolidated petition (from the denial of reconsideration) as moot. The panel (Judges Newman, Nardini, and Merriam) issued a per curiam opinion, converting a prior summary order into a published opinion.

Summary of the Opinion

• The Second Circuit held that the BIA failed to provide sufficient justification for its conclusion that the IJ clearly erred in granting CAT relief.
• The court emphasized that under clear-error review the BIA must:
  • Give “cogent reasons” for rejecting the IJ’s factual findings.
  • Engage the record as a whole, including corroborated expert testimony and country-conditions evidence.
  • Consider the aggregate risk of torture and address evidence bearing on specific intent and government responsibility.
  • Avoid impermissible factfinding or de novo re-weighing of the evidence.
• Key record categories that the BIA failed to meaningfully address included:
  • Evidence that tattoos and a criminal record would make the petitioner a likely target of detention under the State of Exception.
  • Corroborated reports of systemic, state-sanctioned torture, in-custody deaths, starvation, and extreme overcrowding.
  • Official statements and policies indicating intent to inflict severe suffering on suspected gang members.
  • Regime-driven opacity that limits “complete information,” undermining any inference that missing data weakens the applicant’s case.
  • The limited credibility and access of the Salvadoran Human Rights Ombudsman, given evidence of denied prison access and politicization.
• Disposition: Petition for review GRANTED and REMANDED to the BIA; consolidated petition (from denial of reconsideration) DISMISSED as moot; motion to publish GRANTED.

Analysis

1. Precedents Cited and Their Influence

• Wu Lin v. Lynch, 813 F.3d 122 (2d Cir. 2016):
  • Anchors the requirement that the BIA provide “cogent reasons” when concluding an IJ committed clear error.
  • Warns against the BIA “starting anew,” conducting its own analysis, or engaging in impermissible factfinding instead of reviewing for clear error.
  • In Villalta Martinez, the Second Circuit applies Wu Lin to stress that the BIA may not dismiss corroborated expert and country-conditions evidence with conclusory labels and must explain why the IJ’s reasoning was clearly wrong.
• Anderson v. Bessemer City, 470 U.S. 564 (1985):
  • Classic articulation of clear-error review: a reviewing body cannot reverse just because it might have decided differently; error must be “clear.”
  • Supports the panel’s insistence that the BIA cannot lightly displace IJ findings that are supported by record evidence.
• Yan Chen v. Gonzales, 417 F.3d 268 (2d Cir. 2005):
  • Clarifies that when the BIA issues its own opinion (as it did here), appellate review focuses on the BIA’s decision.
• Chun Gao v. Gonzales, 424 F.3d 122 (2d Cir. 2005):
  • Defines CAT’s “more likely than not” standard as greater than a 50% chance of torture.
  • The Second Circuit relies on this standard when assessing the IJ’s probability assessment compared to the BIA’s characterization.
• Pierre v. Gonzales, 502 F.3d 109 (2d Cir. 2007):
  • Prison conditions can constitute torture if they cause severe pain or suffering and authorities intend the severity to punish, intimidate, coerce, or the like.
  • Vital to rebutting the BIA’s conclusion that the petitioner would face only “difficult conditions” rather than intentionally inflicted torture.
• Galina v. Wilkinson, 988 F.3d 137 (2d Cir. 2021):
  • “Lawful sanctions” under local law cannot defeat CAT where the sanctions “defeat the object and purpose of the Convention,” including where penalties are brutal and disproportionate.
  • Supports the panel’s rejection of reliance on formal legality (or official tolerance) in El Salvador to negate intent or torture.
• Aliyev v. Mukasey, 549 F.3d 111 (2d Cir. 2008):
  • Procedural: once the court grants review of the final removal order and remands, any petition from denial of reconsideration is typically moot.

2. The Court’s Legal Reasoning

The crux of the opinion is not a reweighing of the CAT evidence by the court. Rather, it is an administrative-law holding about the BIA’s duty when using clear-error review to overturn an IJ’s CAT grant:

• Clear-error review is narrow. The BIA may not replace the IJ’s plausible view of the record with its own. It must identify an actual, “clear” mistake.
• “Cogent reasons” are required. The BIA cannot simply label expert testimony “generalized” or say “prison conditions are difficult” without engaging record evidence the IJ credited and explaining why the IJ’s view is clearly wrong.
• Aggregate-risk analysis is required. Under 8 C.F.R. § 1208.16(c)(3), the agency must consider risks “in the aggregate,” including:
  • Past harm (e.g., prior MS-13 attack).
  • Systemic human rights violations in El Salvador under the State of Exception.
  • Other relevant country conditions (e.g., detention practices for tattooed deportees; suspension of due process; official policies).
• Specific intent can be inferred from patterns, policies, and official statements. The IJ identified evidence that:
  • Authorities intentionally overcrowded prisons, severely restricted light and movement, and curtailed food to punish and terrorize suspected gang members.
  • Senior officials publicly touted deprivation as a feature, not a bug (e.g., reducing meals; “they won’t eat a grain of rice”; “do not see sunlight”).
  • Reports documented systemic beatings, group electrocutions, simulated drowning, and hundreds of in-custody deaths with signs of torture.
  The BIA was required to address this intent evidence before pronouncing the IJ clearly wrong on “intent.”
• Expert testimony cannot be dismissed for lack of a personal interview where the expert’s analysis is corroborated by independent sources. The IJ acknowledged some limitations but credited the expert given:
  • The expert’s subject-matter qualifications (El Salvador country conditions and the State of Exception).
  • Independent corroboration by State Department and NGO reports, press accounts, and other record materials.
  The BIA’s failure to grapple with the corroboration rendered its clear-error conclusion inadequately justified.
• Regime opacity cannot be weaponized against CAT applicants. Where the record shows systematic suppression of information, attacks on media, limits on independent oversight, and denial of prison access to the Ombudsman, “lack of complete information” does not, by itself, undermine credible, corroborated evidence of systemic abuse.
• Local lawfulness is irrelevant to whether conduct constitutes “torture.” Consistent with Galina and Pierre, the court reiterates that formally “lawful” sanctions that inflict severe suffering for punitive purposes can still constitute torture for CAT purposes.

3. Record Categories the BIA Was Required to Address

• Tattoos and criminal history as triggers:
  • Evidence showed that persons with tattoos or prior contact with the criminal system are frequently rounded up as suspected gang members.
  • The expert reviewed the I-213 reflecting tattoos and a conviction, making the risk analysis individualized.
• Suspended due process and impunity:
  • Official suspension of procedural rights; evidence of extrajudicial killings; pressure on officers to meet arrest quotas; retaliatory arrests of critics.
• Systemic torture and in-custody deaths:
  • Reports of hundreds of in-custody deaths, many with signs of torture; routine beatings and “group electrocution” of detainees; killings of deportees.
• Overcrowding, starvation, and deprivation as intentional policy:
  • Prison overcrowding multiple times capacity; inadequate food and medical care; official statements celebrating deprivation and commanding darkness/no sunlight.
• Ombudsman’s limited credibility and access:
  • Ombudsman’s office was denied access; public statements minimized abuses; State Department documented concerns about independence.

4. Impact

This opinion reinforces and concretizes the Second Circuit’s Wu Lin line: the BIA cannot recast the record through conclusory labels or selective engagement when overturning IJ grants under CAT.

• On BIA practice:
  • Expect more rigorous written explanations when the BIA finds clear error in IJ factfinding, especially in CAT cases involving systemic abuses.
  • Discounting an expert for not interviewing a noncitizen will not suffice where country-conditions evidence independently corroborates the expert’s analysis and the IJ credited the expert’s methodology.
  • The BIA must engage with intent evidence—even when intent is inferred from policies, public statements, and patterns—before rejecting an IJ’s CAT findings.
• On CAT jurisprudence:
  • Affirms that prison conditions can constitute torture when intentionally imposed to punish, intimidate, or coerce (Pierre), and that “lawful sanctions” do not immunize torture (Galina).
  • Strengthens the requirement to assess aggregate risk (8 C.F.R. § 1208.16(c)(3)) rather than isolating each risk factor.
  • Recognizes the probative value of state censorship and opacity as part of the evidentiary matrix, discouraging adverse inferences from “incomplete” data where the regime itself obstructs transparency.
• On El Salvador “State of Exception” cases:
  • Signals that evidence of systemic, officially endorsed deprivation and violence, combined with individualized risk factors (e.g., tattoos, deportee status, criminal history), can be sufficient for an IJ to find CAT eligibility—and the BIA must confront that evidence head on.

Complex Concepts Simplified

• Clear error (as used by the BIA):
  • It’s a deferential standard. The BIA cannot reverse an IJ just because it would have weighed the facts differently.
  • “Clear” means the IJ made an obvious mistake: no supporting evidence, a misunderstanding of testimony, or the opposing evidence overwhelmingly compels a different finding.
• CAT “more likely than not”:
  • The applicant must show a greater than 50% chance of future torture in the country of removal.
• Specific intent to torture:
  • Authorities must intend the severe pain or suffering for a prohibited purpose (punishment, intimidation, coercion, etc.).
  • Intent can be inferred from policies, patterns, official statements, and systemic practices—not only from direct confessions of intent.
• Aggregate risk:
  • The agency must consider all sources of potential harm together (e.g., gang threats, state detention practices, prison conditions), not in isolation.
• “Lawful sanctions” and CAT:
  • Local law cannot convert torture into something permissible under CAT. Brutal, disproportionate penalties are still torture even if labeled “lawful.”
• Impermissible factfinding by the BIA:
  • On appeal from the IJ, the BIA cannot make new factual findings or re-weigh the evidence de novo; it reviews the IJ’s findings for clear error.
• Mootness of reconsideration petition:
  • When the appellate court grants review of the principal removal order and remands, any petition challenging the BIA’s denial of reconsideration becomes moot.

Practical Takeaways for Advocates

• Build the record for aggregate risk:
  • Document personal risk factors (e.g., tattoos, deportee status, prior convictions) that interact with systemic abuses to increase the likelihood of torture.
• Corroborate expert testimony:
  • Pair expert opinions with State Department reports, reputable NGOs, and credible media to forestall “generalized” or “anecdotal” dismissals.
• Evidence of intent:
  • Collect official statements, policy directives, and structured practices (e.g., ration cuts, denial of sunlight, command-level instructions) that signify punitive or intimidating purposes behind harsh conditions.
• Opacity as evidence:
  • Explain how censorship, denial of access, and statistical manipulation impair data, and why this should not be used to discount credible accounts of abuse.
• Address “lawful sanctions” head-on:
  • Situate any purportedly legal measures within CAT’s purpose and Pierre/Galina to show why formal legality does not negate torture.
• On appeal to the BIA:
  • Pin the BIA to clear-error boundaries; insist on cogent reasons and an on-the-record engagement with the IJ’s findings and the corroborated evidence.

What Remains Open on Remand

• The BIA must either accept the IJ’s factual findings or supply a supportable basis for rejecting them under true clear-error review.
• The BIA must evaluate:
  • Whether the petitioner’s tattoos and criminal history make detention under the State of Exception more likely than not.
  • Whether record evidence shows a greater than 50% chance that he will be tortured, considering systemic abuses and personal factors in the aggregate.
  • Whether the government’s intent to inflict severe suffering (e.g., through deliberately imposed starvation, deprivation, and violent practices) is established by the record.

Conclusion

Villalta Martinez v. Bondi underscores and sharpens a critical administrative-law constraint: when the BIA overturns an IJ’s fact-bound CAT grant on clear-error grounds, it must offer cogent, record-grounded reasons that engage the IJ’s rationale and the corroborated evidence. The opinion reaffirms that:

• Prison conditions can amount to torture when deliberately used to punish or intimidate.
• Local legality is irrelevant to whether conduct is “torture” for CAT purposes.
• Aggregate risk must be assessed holistically, including evidence of systemic abuses and personalized risk triggers.
• Government intent may be proved circumstantially through policies, practices, and public statements, especially under authoritarian conditions that limit transparency.

As a published opinion, the decision is an important precedent for CAT adjudications within the Second Circuit. It provides a roadmap for how the BIA must articulate and justify any departure from IJ findings in CAT cases—particularly those involving countries with systemic state-directed violence and institutional opacity. The court’s remand ensures that the petitioner’s claims receive a lawful, evidence-engaged assessment, while signaling to the agency that conclusory clear-error determinations will not suffice.

Case Details

• Case: Villalta Martinez v. Bondi
• Court: U.S. Court of Appeals for the Second Circuit
• Panel: Newman, Nardini, Merriam (per curiam)
• Nos.: 24-115(L), 24-1222(Con.)
• Submitted: June 30, 2025; Decided: July 23, 2025; Published Opinion: October 8, 2025
• Disposition: Petition for review GRANTED; REMANDED to the BIA; consolidated petition DISMISSED as moot; motion to publish GRANTED