Federal Defenders of New York Second Circuit Blog

Sentencing court cannot accept a police report at face value but must assess its specific indicia of reliability; and cannot consider third-party misconduct unrelated to the defendant as “context.”

United States v. Dralle, No. 24-2424 (2d Cir. May 12, 2026) (Bianco, joined by Robinson; Sullivan dissenting).

On plain-error review, the Second Circuit vacated Chase Dralle’s 30-month sentence for receipt of a trafficked firearm and remanded for resentencing. Judge Bianco’s opinion identifies two procedural errors: The sentencing court erred in (1) accepting at face value the allegations of a police report (about an offense Dralle allegedly committed on bail) without considering its reliability; and (2) relying on a co-defendant’s prior shootings (unrelated to and unknown by Dralle) as “context” in determining Dralle’s sentence. Judge Sullivan dissents on both.

 

Relevant Facts

Dralle purchased a pistol from co-defendant Stefan Bagley in August 2023. About a month before, Bagley was involved in two shootings that Dralle did not participate in (or even know about). Dralle was arrested federally in 2024. While he was on bail, someone called the local police in Trumbull, Connecticut, to complain that Dralle had assaulted him at a gas station. Officers spoke to the complainant, reviewed video footage, and completed a police report. But Dralle was never arrested or charged.

 

The police report was not introduced into evidence in federal court and was not in the PSR. The only document before the court was the Probation Office’s petition to revoke Dralle’s bail, which summarized the Trumbull PD’s report, which in turn summarized the complainant’s account of Dralle attacking him. Probation’s petition further recounted the report as claiming that “[o]fficers reviewed CCTV footage and concluded that the complainant’s account appeared to be accurate.” Op. 9. A federal magistrate judge revoked Dralle’s bail.

 

At sentencing, the district court took both Bagley’s misconduct and the Trumbull incident into account in sentencing Dralle. On Bagley, the court said that while Dralle was “not being sentenced” for Bagley’s shootings, it “ha[d] to consider the nature and circumstances of your offense and does so with the greater context of Mr. Bagley’s activities.” Op. 29-30.  On Trumbull, the district court, considering solely Probation’s double-hearsay petition, excoriated Dralle for committing a new crime while on bail. The court did not receive into evidence or even review the police report or any video. Nor did it make a reliability finding as to the report (or petition).

 

The Panel’s Ruling

  1. The district court erred in accepting the police report at face value

The centerpiece of Judge Bianco’s opinion is its framework for handling police-report allegations at sentencing. While a sentencing court may consider uncharged conduct, including hearsay, at sentencing, the underlying facts “must have some minimal indicium of reliability beyond a mere allegation” and must be found by a preponderance of the evidence. United States v. Juwa, 508 F.3d 694, 700-01 (2d Cir. 2007). The question is what those requirements demand, and whether they were met, when a sentencing court relied only on a Probation petition summarizing a police report, which in turn summarized a complainant’s account and claimed that video evidence confirmed that account, without consider the reliability of the underlying allegations.

 

The court ruled that this was error: “[T]o the extent the district court relied upon information in the police report (or the petition summarizing the police report) to establish that the uncharged conduct occurred, it erred by failing to make a determination as to the reliability of the information contained in the police report, as well as whether such evidence met the preponderance of the evidence standard.” Op.18.

 

“Police reports are not presumed to be categorically reliable.” Op. 19 (citing United States v. Jordan, 742 F.3d 276, 280 (7th Cir. 2014), and United States v. Harrison, 809 F.3d 420, 424 (8th Cir. 2015)). Thus, “there must be a finding that the specific document at issue contains sufficient indicia of reliability . . . .” Id. (emphasis in original). Adopting the framework from United States v. Padilla, 793 F. App’x 749 (10th Cir. 2019), Judge Bianco emphasized that “such a finding is logically appropriate only where the document has been entered into evidence so that the district court may, on the record, assess factors bearing on the document’s veracity such as the level of detail, internal consistency, and overall quality of that document.” Op. 20-21.

 

Even better, the Court’s remand instructions provide a blueprint of what a sentencing court must do. Before relying on the Trumbull incident at sentencing, the district court must “identify any documents or evidence upon which it intends to rely,” “ensure that any such documents or evidence have been provided to the defendant,” give Dralle “an opportunity to object to the factual accuracy of those allegations,” and determine “whether the evidence offered by the government is sufficiently reliable to prove those allegations under the preponderance-of-the-evidence standard.” Op. 28-29.” The court added that the magistrate’s probable cause finding upon revoking bail does not suffice and is no substitute: “[T]he magistrate judge’s probable-cause finding at the bail-revocation hearing five days before sentencing could not substitute for a preponderance finding, because ‘probable cause is a lower standard than preponderance of the evidence.’ Op. 25-26 n.3 (quoting Juwa)).

 

  1. The district court erred in relying on unrelated third-party misconduct

The Government conceded at oral argument that the district court could not have attributed Bagley’s shootings to Dralle. Op. 31. Under United States v. Johnson, 378 F.3d 230, 238 (2d Cir. 2004), and U.S.S.G. § 1B1.3(a)(1)(B), holding a defendant accountable for the acts of others at sentencing requires findings that the acts were within the scope of their jointly undertaken criminal activity and reasonably foreseeable to the defendant. Neither showing could be made here.

The Government’s pivot was that Bagley’s shootings were not attributed to Dralle but were merely “context” for the offense — they “helped illustrate why it was so dangerous for someone like Dralle to illegally purchase a gun.” Op. 32.

 

The panel rejected that move, holding that the “mere use of the word ‘context’ by a sentencing judge does not insulate a district court’s consideration of a co-defendant’s violent conduct from review where the district court does not connect its consideration of such conduct to one or more Section 3553(a) factor.” Op. 32. The court must “articulate how, even though it would be improper to hold the defendant accountable for the co-defendant’s conduct by increasing the defendant’s sentence without making the findings required under Johnson, the co-defendant’s activities still could be properly considered for some other reason under the Section 3553(a) factors.” Op. 32-33.

 

Judge Bianco distinguished the government’s two principal cases. United States v. Small, 2024 WL 1173046 (2d Cir. Mar. 19, 2024) (summary order), turned on a causal link — the third-party violence flowed from the defendant’s drug operation. United States v. Davis, 82 F.4th 190 (2d Cir. 2023), tied an above-Guidelines variance to general deterrence and specific local-crime data, and rested on a defense concession. Op. 35-36. Neither is present here.

 

Practitioner’s Notes

First, the panel’s adoption of the Padilla framework is significant. The practice it rejects — a district court absorbing the contents of a Probation petition summarizing a police report, never identifying the source document, never receiving it into evidence, and never making a reliability finding — is depressingly common, and now plainly improper. Counsel should treat Dralle as imposing a checklist at sentencing whenever the government seeks to import police-report allegations. At the least, the court must identify the source document on the record and must make a particularized reliability finding tied to the document’s level of detail, internal consistency, and overall quality.

 

One caveat. The majority opinion stops short of holding that the district court’s consideration of the Trumbull incident was plain error requiring vacatur and remand because the court’s consideration of the co-defendant’s conduct was a sufficient basis for resentencing: “[W]e need not resolve whether the district court’s consideration of the uncharged conduct … amounted to a clear or obvious error, or whether any such error affected Dralle’s substantial rights and the fairness of his sentencing, because … a full resentencing is separately required because of the district court’s plain error with respect to its consideration of the co-defendant’s shootings.” Op. 28. Judge Sullivan argues that the majority’s discussion of the Trumbull issue is therefore dicta. Dissenting Op. 7. You can push back on Judge Sullivan’s argument, should you face it, by observing that the majority specifically says that the district court “erred” (Op. 21), and that the only issue the majority reserved was whether that error rose to the level of plain error for appellate purposes. When dealing with a district court in the first instance, you can and should argue that Dralle’s framework is binding guidance that, if disregarded, would result in legal error.

 

Second, Dralle establishes that “context” is not a magic word for importing third-party misconduct. A sentencing court must either make the Johnson relevant-conduct findings or articulate, on the record, how the third-party conduct bears on a specific § 3553(a) factor as applied to this defendant.

 

Leave a Comment