“Death Results” in Federal Kidnapping Requires Only But‑For Causation, Not Proximate Cause

Introduction

In United States v. Parks (2d Cir. June 22, 2026), Edward Michael Parks appealed convictions for two counts of kidnapping resulting in death under 18 U.S.C. § 1201(a)(1) and one count of witness tampering by killing under 18 U.S.C. §§ 1512(a)(1)(C) & 1512(a)(3)(A), for which he received three consecutive life sentences.

The appeal centered on a causation question with broad sentencing consequences: whether § 1201(a)(1)’s enhanced punishment “if the death of any person results” requires the government to prove only actual (“but-for”) causation, or also legal (“proximate”) causation—i.e., that death was a reasonably foreseeable result of the kidnapping. Parks also challenged the sufficiency of the evidence on all counts and argued that supplemental jury instructions improperly expanded the government’s theory.

Summary of the Opinion

The Second Circuit affirmed. It held that the district court correctly instructed the jury that the “death results” element in § 1201(a)(1) requires but-for causation, not proximate causation. The court further held that, even if proximate cause were required, any omission was harmless because the evidence made foreseeability overwhelming. The panel also rejected Parks’s remaining sufficiency challenges and upheld the witness-tampering-by-killing conviction under the “reasonable likelihood” standard for federal communication.

Analysis

Precedents Cited

  • United States v. Felder, 993 F.3d 57 (2d Cir. 2021)
    The decisive interpretive anchor. Felder construed similarly worded “if death results” language in the federal carjacking statute (18 U.S.C. § 2119(3)) and concluded that “results from” generally demands actual (but-for) causation, not proximate causation. Parks extended that approach to § 1201(a)(1), treating kidnapping’s “if the death of any person results” as the same kind of death-results enhancement.
  • Burrage v. United States, 571 U.S. 204 (2014)
    The Supreme Court’s central causation decision on “death results” language. Burrage held that such phrasing imposes a but-for requirement and that, when it increases sentencing exposure, it is an element for the jury. Parks relied on Burrage to reinforce the textual baseline: “results” points to actual causation. The panel noted Burrage did not decide whether proximate cause is also required, leaving room for circuit interpretation—filled here by Felder.
  • United States v. Guillette, 547 F.2d 743 (2d Cir. 1976)
    Parks’s principal counter-authority. Guillette construed the “if death results” enhancement in 18 U.S.C. § 241 and approved liability where defendants’ conduct proximately caused death, even if not the immediate cause. The Parks court treated Guillette as describing what may be sufficient in some contexts rather than announcing that proximate cause is always required whenever Congress uses “death results.”
  • Elonis v. United States, 575 U.S. 723 (2015) and Carter v. United States, 530 U.S. 255 (2000)
    Used to support a mens rea/“innocent conduct” rationale: courts read into statutes only the mens rea needed to separate wrongful from innocent conduct. Because kidnapping requires “willful intent” to confine, the panel saw no need to add a foreseeability gloss to avoid punishing innocent behavior.
  • Fowler v. United States, 563 U.S. 668 (2011)
    Governs the witness-tampering-by-killing count: the government must show a “reasonable likelihood” that the victim would have communicated with a federal officer—more than remote or hypothetical, but not more likely than not. The panel applied Fowler to uphold Count Three.
  • Standards-of-review and post-verdict motion precedents shaping the court’s posture: Musacchio v. United States (Rule 29 sufficiency), United States v. Ho (circumstantial inferences), United States v. Landesman, United States v. Alston, United States v. Aguiar (Rule 33 new trial), United States v. Zheng, United States v. Guldi, United States v. Raniere (jury instructions), United States v. Gomez and United States v. Silver (harmless-error framework), and United States v. Denkberg (contextual reading of instructions).

Legal Reasoning

1) Text and grammar: “results” and the passive voice

The court emphasized that § 1201(a)(1) does not mention “proximate cause,” “foreseeability,” or “direct result.” Following Felder, it treated Congress’s use of “results” (rather than “causes”) and the passive construction (“if the death … results”) as signaling focus on whether death occurred as a consequence of the conduct, not on layered notions of foreseeability.

2) Comparative statutory drafting: Congress knows how to say “proximate”

The opinion strengthened its textual inference by pointing to other federal statutes in which Congress expressly required “direct or proximate” causation (e.g., 18 U.S.C. §§ 247(d)(2) & (3), 844(d), 844(f), 844(i), and § 38). The implication is a negative one: when Congress omits proximate-cause language in a death-results enhancement, courts should be reluctant to insert it.

3) Offense nature and redundancy concerns

Borrowing Felder’s logic, the court reasoned that kidnapping is generally a violent crime and inherently risky (whether by physical or psychological coercion). Adding a foreseeability requirement for death could be “redundant and even confusing” for jurors, particularly where the predicate offense already requires willful, nonconsensual confinement.

4) Reconciling Guillette with modern causation doctrine

The panel rejected the claim that Guillette constitutionalized or universalized proximate cause for all “death results” enhancements. It treated Guillette as consistent with holding a defendant liable when death is within the consequences of the criminal conduct (including through intermediaries), but not as mandating that the government prove foreseeability in every statute with “death results” wording.

The court also noted that Guillette predates the modern Supreme Court’s sharper textual approach to “death results” language (particularly Burrage’s but-for holding), supporting the Second Circuit’s reliance on more recent interpretive baselines.

5) Harmlessness alternative (belt-and-suspenders)

Even assuming arguendo that proximate cause were required, the court held any omission harmless: the evidence that Parks held the victims at gunpoint and prevented them from leaving made death plainly foreseeable, so a rational jury would have found proximate cause beyond a reasonable doubt.

6) Supplemental instruction and “actions” language

Parks argued that the court’s supplemental instruction (referring to defendant’s “actions”) allowed conviction based on conduct other than kidnapping (e.g., gun trafficking). Applying a contextual approach (and plain-error review because the point was not preserved), the panel read the instruction against the full charge, which repeatedly tied the “resulted in death” element to the kidnapping elements immediately preceding it.

The court also rejected the due process claim that the supplemental instruction improperly shifted the government’s theory. The district judge gave both sides a chance to present additional argument after the jury’s question, and Parks had notice from the charging conference that the court viewed the law as not requiring Parks to be the shooter.

7) Witness-tampering-by-killing under Fowler

On Count Three, the court held the evidence satisfied Fowler’s “reasonable likelihood” standard because: (i) the FBI was involved from the day of the murders; (ii) the murders were connected to illicit firearm sales (a federal investigative hook); (iii) all surviving eyewitnesses ultimately testified; and (iv) Parks’s statement that Williams “talked too much” supported the inference that Parks killed him to prevent disclosure to law enforcement, including federal officers.

Impact

  • Clarified burden in § 1201(a)(1) death-results prosecutions: The government in the Second Circuit need not prove foreseeability/proximate cause to trigger the life/death enhancement—only that the kidnapping was a but-for cause of death.
  • Alignment across “death results” enhancements: By extending United States v. Felder’s logic from carjacking to kidnapping, the court promotes a more uniform, text-driven approach to similarly phrased federal enhancements, making outcomes less dependent on common-law causation glosses.
  • Litigation shift to factual causation and intervening actors: Defendants may focus more on contesting but-for causation (e.g., independent causes, intervening acts that actually break causation) rather than arguing foreseeability. The opinion leaves open that intervening events could still defeat but-for causation in an appropriate case.
  • Jury-instruction practice: Trial courts may increasingly give “but-for” instructions for “death results” clauses and, when juries ask questions, clarify that shooter identity is not necessarily dispositive—so long as causation and other elements are met.
  • Fowler remains a comparatively low (but meaningful) threshold: The witness-tampering-by-killing analysis underscores that the government need not prove the victim would more-likely-than-not speak to federal agents; it is enough that federal communication is reasonably likely given the investigative context.

Complex Concepts Simplified

But-for causation (actual cause)
A death “results” from conduct if the death would not have happened when it did without that conduct. If the kidnapping set the fatal sequence in motion such that, without the kidnapping, the deaths would not have occurred, but-for causation is satisfied.
Proximate causation (legal cause / foreseeability)
A limiting principle that asks whether the result was a reasonably foreseeable outcome of the conduct, or whether the causal chain is too attenuated to justify criminal liability for the result.
“Death results” enhancement as an element
Because “death results” increases the sentencing range (often drastically), it must be found by the jury beyond a reasonable doubt, not merely by a judge at sentencing.
Rule 29 vs. Rule 33
Rule 29 asks whether any rational juror could find guilt beyond a reasonable doubt (highly deferential to the verdict). Rule 33 allows a new trial only in extraordinary circumstances to prevent manifest injustice.
Harmless error vs. plain error
Harmless error applies when an issue is preserved; the government must show the error did not affect the outcome. Plain error applies when an issue was not preserved; the defendant must show a clear, outcome-affecting error that undermines the proceedings’ fairness or integrity.
Fowler “reasonable likelihood”
For § 1512(a)(1)(C), the government need not prove the victim was about to speak to a federal agent, only that it was reasonably likely that the information would reach federal law enforcement—more than a speculative possibility.

Conclusion

United States v. Parks establishes (and, within the Second Circuit, solidifies) a key rule for federal kidnapping cases: the “if the death of any person results” enhancement in 18 U.S.C. § 1201(a)(1) requires proof of but-for causation, not proximate causation. The decision extends the court’s earlier reasoning in United States v. Felder, narrows the practical reach of foreseeability arguments under “death results” wording, and reinforces a text-first approach informed by Congress’s express use of “proximate” language elsewhere. It also confirms the resilience of witness-tampering-by-killing prosecutions under Fowler where federal involvement is realistically in play, even if not certain.