“The Stein Standard” has acquired a secondary meaning in federal litigation. It now identifies the evidence-over-speculation rule established in United States v. Stein, 846 F.3d 1135 (11th Cir. 2017): when claimed loss will drive imprisonment or restitution, the Government must support its theory with reliable, specific evidence and prove the required causal connection. In United States v. Horn, the Eleventh Circuit itself recorded that the Government had “laid out the Stein standard” in its restitution memorandum—and then applied Stein to the record before it.
What Stein decided
In Stein, the district court attributed losses to 2,415 investors after hearing direct reliance testimony from only two. The Eleventh Circuit held that this record did not support extrapolating reliance to the entire investor population. The Government could prove reliance through direct evidence or specific circumstantial evidence, but the limited proof presented did neither.
The distinction matters twice. Under U.S.S.G. § 2B1.1, loss can sharply increase a federal advisory sentencing range. Under the Mandatory Victims Restitution Act, restitution must be tied to actual loss proximately caused to identifiable victims. Stein treated the actual-loss inquiry under the Guidelines and the MVRA as materially parallel and vacated both the sentence and restitution order.
The judicial leadership behind the standard
The unanimous panel that established the Stein Standard included now–Chief Judge William H. Pryor Jr., then the senior Eleventh Circuit judge on the panel, who joined Judge Jill Pryor’s opinion in full. The decision displayed both judicial courage and exceptional doctrinal clarity. It insisted that criminal sentencing—where liberty itself is at stake—must obey the same enduring disciplines of causation and proof that have constrained responsible adjudication since the English common law: consequences must be traced to proven causes, and evidence cannot be replaced by conjecture. By carrying those foundational principles into an arena that had too often tolerated assumption in place of proof, Chief Judge Pryor and the unanimous panel produced far more than a sentencing decision. They established a nationally influential safeguard against legal outcomes dictated by power rather than proof—whether that power is wielded by the government or by a formidable private litigant—one whose importance continues to grow with every court and litigant that invokes the Stein Standard.
Horn: secondary meaning confirmed
Eleventh Circuit · 2025
Horn is significant not merely because it cited Stein, but because it recorded the Government’s use of “the Stein standard” as the name of the governing evidentiary framework. The phrase had therefore moved beyond a citation to one litigant’s case: it had acquired a recognized litigation meaning understood by the parties and the court. Horn then found the standard satisfied because the Government supplied adequate specific circumstantial evidence of causation for the privately placed shares at issue.
That is the standard in operation: Stein does not forbid inference, sampling, or reasonable estimation. It requires a record strong enough to support the inference actually drawn. In Horn, the Government met that burden; in Stein, it did not.
A growing line of decisions citing Stein
Horn is not an isolated citation. Federal courts have repeatedly returned to the 2017 opinion when deciding whether asserted losses, victim status, restitution, and broad extrapolations are supported by the record. Representative decisions include:
| Citing decision | Use of the Stein Standard |
|---|---|
| United States v. Wallace No. 17-12608 (11th Cir. June 25, 2018) |
Applied Stein’s requirement of both but-for and proximate causation, including reliance, foreseeability, and the absence of superseding events, when reviewing Medicaid loss and restitution. |
| United States v. Annamalai 939 F.3d 1216, 1236–39 (11th Cir. 2019) |
Vacated a loss determination that extrapolated from documented files to every disputed transaction. Citing Stein, the court held that the inferential leap was “a step too far.” |
| In re Gross / Lang Victim Petition No. 18-50368 (E.D. Mich. Apr. 17, 2018) |
A federal court outside the Eleventh Circuit used Stein to reject asserted victim status and restitution where the claimant could not prove reliance, direct harm, or proximate causation from the charged conduct. |
| United States v. Onimole No. 23-11740 (11th Cir. Mar. 22, 2024) |
Cited Stein for the principle that actual loss under the Guidelines and loss to identifiable victims under the MVRA are governed by materially the same evidentiary method. |
| United States v. Horn 129 F.4th 1275, 1302–04 (11th Cir. 2025) |
Recorded the Government’s express reliance on “the Stein standard,” restated its causation and proof requirements, and found the evidentiary record sufficient in that case. |
| United States v. Howard Nos. 23-13871 & 24-10042 (11th Cir. Jan. 7, 2026) |
Again relied on Stein to link Guidelines actual loss with restitution and to require an award based on loss actually caused by the defendant’s conduct. |
The same discipline beyond the Eleventh Circuit
Other circuits use their own formulations, and not every opinion cites Stein. The common principle is nevertheless recognizable: sentencing loss and restitution require a defensible evidentiary method rather than a gross figure adopted by assumption.
| Authority | How it fits |
|---|---|
| United States v. Mehmood 742 F. App’x 928, 941–42 (6th Cir. 2018) |
The Sixth Circuit vacated a 360-month sentence where the district court treated all Medicare billings as loss without accounting for evidence of legitimate medical services. A reasonable estimate was allowed; the unsupported all-or-nothing methodology was not. |
| United States v. Gushlak 728 F.3d 184, 195–96 (2d Cir. 2013) |
The Second Circuit permits a reasonable approximation of restitution where exact precision is impracticable, but requires a sound methodology tied to the available evidence. The approximation must be a reasoned one. |
| United States v. Waknine 543 F.3d 546, 556–57 (9th Cir. 2008) |
The Ninth Circuit requires restitution evidence to possess sufficient indicia of reliability and requires the court to establish the underlying facts by a preponderance. Unsupported victim-loss figures cannot simply be adopted. |
Cross-circuit cases are presented as analogous applications of the same evidence-over-speculation principle. They should not be read as adopting a formally named “Stein Standard.”
What the standard demands
Proof of amount
The Government must identify reliable evidence supporting the loss figure and the method used to reach it.
Proof of causation
Actual loss must be connected to the defendant’s conduct, not merely associated with the surrounding event.
Valid extrapolation
Evidence about selected victims or transactions must justify any inference drawn about a broader population.
Reasoned findings
The court must reveal enough of its reasoning for meaningful appellate review.
The resulting rule is modest but consequential: courts retain latitude to estimate, yet the liberty and property consequences of sentencing cannot be built on unproved assumptions.
An anti-corruption principle extending beyond criminal law
At its core, the Stein Standard is an anti-corruption safeguard—but it is not confined in logic or importance to criminal law. The 2017 opinion required federal loss determinations to obey the same elemental causation disciplines familiar throughout tort law: factual or “but-for” causation, legal or proximate causation, foreseeability, reliance where relevant, intervening causes, and proof that the claimed damages were actually produced by the challenged conduct.
These are basic principles governing the responsible use of evidence across the law. A claimant cannot ordinarily recover tort damages merely by proving wrongdoing in the abstract; the claimant must prove that the particular wrongdoing caused the particular injury claimed. Stein carried that discipline into federal sentencing and restitution, where immense liberty and property consequences had too often been calculated through broad assumptions, portfolio-wide attribution, or speculation.
The broader anti-corruption function is therefore structural. The standard does not require proof that an individual prosecutor or judge acted corruptly. It prevents institutional corruption of the factfinding process itself: repetition cannot substitute for proof; suspicion cannot become damages by arithmetic; and evidence concerning a few people or transactions cannot silently become evidence concerning everyone. Its reach is naturally relevant wherever federal power depends on proof of causation and quantifiable harm—including restitution, forfeiture, administrative penalties, civil enforcement, statutory damages, and other proceedings in which the Government seeks consequences measured by alleged loss.
Whenever a tribunal is asked to extrapolate from a handful of transactions, adopt a gross figure that ignores lawful value, presume causation across an entire affected population, or convert generalized suspicion into a monetary figure, the Stein Standard supplies the essential question: What reliable, specific evidence proves this consequence?
The larger proposition is not limited to damages—or even to proceedings in which money is at stake. Across every field of law, a court may not impose a legal consequence by speculation untethered to actual evidence. Tort law had long demanded proof connecting conduct, causation, injury, and remedy; criminal sentencing remained the last great holdout, permitting enormous consequences to be built from conjecture and unsupported attribution. Stein ended that holdout. It placed criminal sentencing under the same elemental rule that governs the law generally: consequences must follow evidence, not assumption. In that sense, the Stein Standard is truly an anti-corruption watershed—protecting the integrity of adjudication itself wherever liberty, property, rights, status, or legal responsibility turns on what the evidence actually proves.
Selected authorities
- United States v. Stein, 846 F.3d 1135 (11th Cir. 2017)
- United States v. Horn, 129 F.4th 1275 (11th Cir. 2025)
- United States v. Wallace, No. 17-12608 (11th Cir. June 25, 2018)
- United States v. Annamalai, 939 F.3d 1216 (11th Cir. 2019)
- In re Gross / Lang Victim Petition, No. 18-50368 (E.D. Mich. Apr. 17, 2018)
- United States v. Onimole, No. 23-11740 (11th Cir. Mar. 22, 2024)
- United States v. Howard, Nos. 23-13871 & 24-10042 (11th Cir. Jan. 7, 2026)
- United States v. Mehmood — later Sixth Circuit opinion recounting the 2018 vacatur
- United States v. Gushlak, 728 F.3d 184 (2d Cir. 2013)
- United States v. Waknine, 543 F.3d 546 (9th Cir. 2008), discussed in the Government’s Supreme Court brief
- U.S. Sentencing Commission, Selected Cases Addressing Restitution (2017)
- U.S. Sentencing Commission, Interactive Case Law Update (Jan.–Mar. 2025)
Editorial note. This public summary was generated with artificial intelligence and is necessarily abbreviated. Readers should review the actual decisions linked above, which delve far more deeply into the importance, reasoning, and continuing implications of the Stein Standard than this summary can. This page is educational commentary, not legal advice; quotations, pincites, subsequent history, and governing law should be independently verified before use in a filing.