Coram Nobis Harmless-Error Review: Kotteakos Governs (and Petitioner Bears the Burden)

1. Introduction

Stinn sits at the intersection of post-Ciminelli fraud litigation and the rarely granted remedy of coram nobis. Bradley Stinn, former CEO of Friedman's Inc., was convicted in 2009 of securities fraud, mail fraud, and conspiracy (18 U.S.C. §§ 1341, 1348, 1349). His jury was instructed it could convict under either (i) a “traditional” money-or-property fraud theory or (ii) the Second Circuit’s then-accepted “right-to-control” theory.

After Stinn finished his prison and supervised release terms, the Supreme Court repudiated the right-to-control theory in Ciminelli v. United States, 598 U.S. 306 (2023). Stinn therefore sought to vacate his conviction via coram nobis, arguing that the invalid theory might have driven the general verdict and that the error required reversal. The district court denied relief as harmless; the Second Circuit affirmed, using the case to resolve a key procedural question: what harmless-error standard applies to coram nobis?

2. Summary of the Opinion

The Second Circuit held:

• Traditional property theory supported the conviction. The record showed the government’s case centered on Stinn’s $352,000 bonus (2002) and $300,000 salary increase (2003) tied to manipulated earnings targets—classic property fraud.
• A Yates error existed but was harmless under the correct standard. Because the jury returned a general verdict, it was “impossible to tell” whether it relied on the now-invalid right-to-control theory, creating a Yates v. United States, 354 U.S. 298 (1957), problem. But the court concluded the instructional error did not have a substantial and injurious effect on the verdict.
• Governing standard: For coram nobis, harmless error is reviewed under Kotteakos (as adapted by habeas doctrine), not Chapman; and the petitioner bears the burden.

3. Analysis

A. Precedents Cited (and How They Drive the Result)

1) Right-to-control’s demise and the “traditional property” boundary

• Ciminelli v. United States, 598 U.S. 306 (2023): The catalyst for the petition. Ciminelli rejected the right-to-control theory as not involving “money or property.” The panel assumed (without deciding) a uniform “money or property” meaning across mail, wire, and securities fraud, tracking Ciminelli’s in pari materia logic.
• Carpenter v. United States, 484 U.S. 19 (1984): Used to confirm that “money or property” covers both tangible and intangible property, but still demands a “traditional property interest.” The court relied on this frame to reject Stinn’s attempt to characterize the case as falling outside traditional property fraud.
• Kousisis v. United States, 605 U.S. 114 (2025): Invoked to reinforce a broad “fraudulent-inducement” conception—schemes that trick victims into handing over money/property—while emphasizing materiality as the limiter rather than categorical carveouts. This supported the panel’s skepticism of a sweeping “salary maintenance” exemption.

2) Law-of-the-case and the re-litigation barrier

• Stinn v. United States ("Stinn II"), 515 F. App'x 4 (2d Cir. 2013): The panel treated Stinn’s “salary maintenance vs. bonus/raise” contentions as already rejected, invoking law-of-the-case.
• United States v. Frias, 521 F.3d 229 (2d Cir. 2008): Supplied the law-of-the-case articulation and exceptions (new evidence, intervening controlling law, clear error/manifest injustice), none of which the panel found applicable.

3) Yates error and harmlessness architecture

• Yates v. United States, 354 U.S. 298 (1957): Established the core problem—general verdict + one invalid theory—requiring reversal unless harmless. The panel accepted a Yates error existed because the verdict did not reveal the theory relied upon.
• Johnson v. United States, 144 F.4th 133 (2d Cir. 2025): Quoted for two propositions: (i) Yates errors are not structural and are subject to harmless-error review; (ii) the court clarified that certain language in Johnson about what the jury “did” convict on was dictum and inconsistent with the workable harmless-error inquiry.

4) The central procedural holding: harmless error on coram nobis

• Chapman v. California, 386 U.S. 18 (1967): The direct-review standard Stinn wanted (“harmless beyond a reasonable doubt,” government bears burden). The panel rejected it for coram nobis.
• Kotteakos v. United States, 328 U.S. 750 (1946): The adopted standard—whether the error had a “substantial and injurious effect or influence” on the verdict.
• Brecht v. Abrahamson, 507 U.S. 619 (1993): The bridge from direct review to collateral review; the panel extended Brecht’s collateral-review logic to coram nobis (finality, time passage, systemic costs).
• Brown v. Davenport, 596 U.S. 118 (2022): Used to support that Brecht/Kotteakos “inverted Chapman’s burden,” i.e., the petitioner bears the burden under Kotteakos.

5) Coram nobis as an extraordinary, finality-protecting remedy

• Foont v. United States, 93 F.3d 76 (2d Cir. 1996): Cited to characterize coram nobis as an extraordinary collateral attack.
• United States v. Keogh, 440 F.2d 737 (2d Cir. 1971) (Friendly, J.), and United States v. Keogh, 391 F.2d 138 (2d Cir. 1968): Provided the rationale that coram nobis petitioners—no longer imprisoned—face less severe consequences, and that retrial is often unlikely, making relief effectively irreversible.
• United States v. Mandanici, 205 F.3d 519 (2d Cir. 2000): Supplied the standard of appellate review for coram nobis denials (de novo for legal standard; abuse of discretion for ultimate denial) and supported the “expunge without reinstatement” finality concern.
• United States v. Frady, 456 U.S. 152 (1982): Quoted for the federal government’s interest in finality, undercutting the argument that only state comity concerns justify collateral-review strictness.

6) “Properly instructed jury would have convicted” framework

• Colotti v. United States, 71 F.4th 102 (2d Cir. 2023), and Stone v. United States, 37 F.4th 825 (2d Cir. 2022): Applied for the operative inquiry under collateral review—whether a properly instructed jury would have found guilt.
• Gomez v. United States, 87 F.4th 100 (2d Cir. 2023), and Peck v. United States, 106 F.3d 450 (2d Cir. 1997): Reinforced that the Second Circuit’s harmless-error approach asks what the jury would have done absent the error.

7) Trial-record inference discipline

• Donnelly v. DeChristoforo, 416 U.S. 637 (1974): Used to resist reading isolated prosecutorial references (“handsome salary”) as proof the case turned on an improper theory.

8) Additional contextual precedents

• Allen v. United States, 164 U.S. 492 (1896): Mentioned in recounting Stinn’s direct appeal issues (United States v. Stinn ("Stinn I"), 379 F. App'x 19 (2d Cir. 2010)).
• Skilling v. United States, 561 U.S. 358 (2010): Mentioned to describe Stinn’s earlier § 2255 theory rejected in Stinn II.
• United States v. Yates, 16 F.4th 256 (9th Cir. 2021): Considered (and distinguished) on salary-maintenance concerns; the panel also noted tension among circuits, referencing United States v. Guertin, 67 F.4th 445 (D.C. Cir. 2023), abrogated on other grounds by Kousisis.

B. Legal Reasoning

1) Traditional property fraud: bonus and raise, not “mere salary maintenance”

The panel’s merits analysis is deliberately record-focused. Even if one accepted the premise (urged via the Ninth Circuit’s United States v. Yates) that deceiving an employer to keep an existing salary can be problematic as the object of fraud, the Second Circuit concluded the prosecution’s theory was different: Stinn allegedly manipulated earnings to hit targets tied to a large bonus and raise—classic remuneration as property obtained by deception.

The court pointed to (i) repeated closing-argument references to the bonus and raise as fraud proceeds, (ii) trial testimony connecting compensation decisions to manipulated performance, and (iii) a forfeiture verdict expressly capturing the bonus and raise as traceable proceeds—making it implausible that the jury’s guilt finding rested primarily on generalized “salary maintenance” rhetoric.

2) Choosing Kotteakos over Chapman for coram nobis

The opinion’s core doctrinal move is to treat coram nobis as functionally aligned with collateral review (like habeas), not direct review:

• Institutional posture: Coram nobis is an extraordinary, belated attack where time has degraded evidence and memories (a Brecht-style concern).
• Finality and cost: Overturning a long-final conviction carries systemic and societal costs; finality interests apply to federal convictions too (Frady).
• Remedy severity asymmetry: Because coram nobis petitioners are no longer imprisoned (Keogh), the court saw no reason to apply a more demanding government-burden rule than on habeas.
• Practical irreversibility: A successful coram nobis petition is often tantamount to expungement without realistic retrial (Keogh; Mandanici), supporting a higher hurdle for petitioners.
• Anti-incentive-to-delay principle: Applying Chapman would perversely reward defendants who wait until their sentences end to re-litigate under a more government-hostile standard.

The court therefore adopted: (i) Kotteakos’s “substantial and injurious effect” test; (ii) petitioner-borne burden (citing Brown v. Davenport); and (iii) the “properly instructed jury would have convicted” methodology (Colotti, Stone).

3) Application: why the Yates error was harmless

Under the adopted framework, the panel held Stinn did not show the right-to-control instruction substantially influenced the verdict. The evidence that the scheme aimed to obtain traditional property—especially the bonus and raise tied to earnings manipulation—was strong, and the forfeiture verdict reinforced that the jury viewed those items as proceeds of the fraud. Thus, a properly instructed jury (traditional property theory only) would have convicted.

C. Impact

1) Post-Ciminelli collateral litigation: a narrowed path

Many defendants convicted in the Second Circuit before Ciminelli faced right-to-control instructions and general verdicts. Stinn substantially limits coram nobis as a vehicle for relief in such cases by:

• imposing the Kotteakos prejudice threshold (not Chapman);
• placing the burden on the petitioner; and
• directing courts to the counterfactual “properly instructed jury” inquiry that often favors affirmance where traditional property evidence is substantial.

2) Strategic consequences for defendants

The opinion discourages “wait-out-the-sentence” strategies by refusing to grant more favorable harmless-error review merely because the defendant is now outside custody. Defendants seeking to capitalize on Ciminelli should expect the key battleground to be evidentiary: showing the trial record plausibly depended on right-to-control rather than traditional property.

3) Doctrinal consolidation in the Second Circuit

Stinn harmonizes coram nobis with the Second Circuit’s broader collateral-review jurisprudence (e.g., Colotti, Stone), while neutralizing dicta that could have made Yates errors effectively non-harmless on collateral review.

4. Complex Concepts Simplified

Writ of error coram nobis

A rare post-conviction remedy for people who are no longer in custody. It attacks a criminal conviction after the sentence is complete, typically to address continuing collateral consequences (e.g., professional licensing, immigration, reputational stigma), but it is granted only in extraordinary circumstances.

Right-to-control theory

A former Second Circuit fraud theory allowing conviction when deception deprived a victim of economically valuable information affecting decision-making—even without aiming to obtain money or property. The Supreme Court rejected this in Ciminelli.

Yates error (invalid-theory general verdict)

When a jury is told it may convict on multiple theories and one is legally invalid, a general verdict creates uncertainty about the basis for conviction. The conviction stands only if the error is harmless under the applicable harmless-error standard.

Harmless error: Chapman vs. Kotteakos

Chapman (direct review, constitutional errors): the government must prove the error was harmless beyond a reasonable doubt.
Kotteakos (collateral-style review): the petitioner must show the error had a substantial and injurious effect on the verdict—often evaluated by asking what a properly instructed jury would have done.

Law-of-the-case doctrine

Issues decided in earlier appeals generally cannot be re-litigated in later stages of the same case, absent narrow exceptions (new evidence, intervening controlling law, or clear error/manifest injustice).

5. Conclusion

Stinn v. United States of America establishes a clear procedural rule in the Second Circuit: coram nobis petitions raising instructional errors—including post-Ciminelli right-to-control Yates errors—are reviewed for harmlessness under Kotteakos, with the burden on the petitioner to show substantial and injurious influence on the verdict, measured by the “properly instructed jury would have convicted” inquiry. On the facts, the court found overwhelming traditional-property evidence (bonus and raise tied to manipulated earnings) and affirmed the denial of coram nobis relief.