Taking a Chance with the Burden of Proof: The But-For Test in Homicide Case Law

• Author: Bradley D. Price
• Position: J.D. Candidate, The University of Iowa College of Law, 2007; B.A., The University of Iowa, 2003
• Pages: 705-739

J.D. Candidate, The University of Iowa College of Law, 2007; B.A., The University of Iowa, 2003. I am grateful to Professor Eric A. Johnson for his thorough and provocative scholarship on this subject, making this Note possible. My thanks and respect also extend to my family for their guidance; to Matt Lynch for his assistance on an earlier draft; to my editors John Pantazis, Nathan Moenck, and Brandon Duncomb; and of course to Kate, for her constant patience and support. Errors and omissions are mine.

Page 705

I Introduction

American criminal law permits a finding of liability for homicide only upon proof beyond a reasonable doubt of (1) conduct by a defendant that is proscribed by statute, (2) the requisite mental state, (3) the concurrence of the defendant's conduct and mental state, and (4) a causal link between the defendant's conduct and some forbidden result.¹ This Note focuses on the fourth criterion for proving liability in homicide cases. Analytical problems arise when the causal link between conduct and harm is unclear, and sometimes even expert witnesses are unable to ascertain what caused the death of a homicide victim. Often, as in the cases relevant to this Note,² more than one act or omission results in a specified harm. Scholars refer to these cases as those of concurrent causation. Professor Eric Johnson³ has correctly identified these cases as ones in which the traditional but-for test seems inadequate to preserve the just notions of liability and punishment that our moral sensibilities demand.⁴

However, in order to retain a coherent doctrine of criminal law within the context of concurrent causation, the but-for test should be supplemented, not abandoned. The purposes of this Note are to (1) demonstrate that neither the doctrine of lost chance nor the rule of multiple sufficient causal sets should be used to determine whether causation has been proved in a homicide case, and (2) propose an alternative solution for the disposition of homicide cases that involve concurrent causation. Since concurrent-cause homicides more closely resemble attempts than completed crimes, prosecutors should consider altering their charging decisions accordingly. Alternatively, state legislatures should codify an appropriate offense that takes account of concurrent causation in order to preserve a coherent body of substantive homicide law.

This analysis begins in Part II, which identifies the common-law rules and policies of the causal element in homicide law, setting forth what hasPage 706 been the generally accepted doctrine up to this point.⁵ It addresses the general causation standard as it exists today in cases of both homicide and accomplice liability. Part III examines Professor Johnson's Article, "Criminal Liability for Loss of a Chance," recently published in the Iowa Law Review.⁶ Professor Johnson's Article is summarized in full as it relates to his "grand unified theory"⁷ of causation. Specifically, Part III explores two theories that Professor Johnson contends account for the outcomes of concurrent-cause homicide cases. First, Professor Johnson imports a tort doctrine known as the rule of multiple sufficient causal sets. "This rule states a defendant's conduct will qualify as a cause of the plaintiff's injury if it is a necessary element of a set of conditions that would have been sufficient to bring about the injury, even if another 'sufficient causal set' was also present."⁸ With this as his working definition of cause, Professor Johnson then argues that courts should use the doctrine of "lost chance" as "shorthand" for the tort rule of multiple sufficient causal sets.⁹ The doctrine of lost chance "requires the government . . . to prove that the defendant's conduct had the effect of substantially reducing the victim's chances of surviving . . . and that the victim might have survived but for the defendant's conduct."¹⁰ Professor Johnson uses a syllogism to argue that lost chance qualifies as a species of causation sufficient to convict a homicide defendant because conduct that qualifies as causation under lost-chance analysis also qualifies as causation under the rule of multiple sufficient causal sets.¹¹

Part IV argues that Professor Johnson's analysis is flawed in a number of respects.¹² First, Professor Johnson impermissibly imports tort principles into the realm of criminal law. Second, Professor Johnson's formulation of the rule of multiple sufficient causal sets suffers from a number of structural and conceptual defects. Most importantly, the rule fails the criminal-law standard for proof beyond a reasonable doubt, uses conceptually problematic language in demarcating its own limits, and would allow inappropriate manipulation of factual evidence presented to juries in order to fabricate liability. Finally, after denying the relevance of policy in the causal inquiry, Professor Johnson makes incorrect assumptions about the rationales for imposing criminal liability.

Page 707

Part V demonstrates the serious and troubling implications of both using lost chance to define causation in homicide cases¹³ and also of continuing to leave the issue of concurrent-causation homicide cases unresolved. It argues that leaving the issue unresolved will lead to the potential for the role of state legislatures to be usurped. Further, if Professor Johnson's definition of "cause" is accepted, not only will conduct that constitutes an attempted crime be treated as if it constitutes a completed crime, but the resulting legal framework will be one in which (1) an element of causation will be thrust into a body of law where it does not belong,¹⁴ and (2) the element of causation will broaden liability beyond current permissible statutory limits. This Note concludes that if criminal courts are to adhere to the requirement of proof beyond a reasonable doubt and prevent prosecutors from vitiating the distinction between principal conduct and accomplice conduct, the legislature must act by codifying criminal statutes that account for cases involving concurrent causation.

II Foundations of Causation in Criminal Law

Before assessing the viability of lost chance and multiple sufficient causal sets in the context of criminal law, this Note sets out the general principles, rules, policy rationales, and perspectives of causation in criminal law. What follows is an explanation of the scope of relevant causation in criminal law, the general requirements for criminal causation, and the policies underlying the application of these standards to criminal law.

A The Scope of the Causal Inquiry

Not all crimes raise the issue of causation. Professor LaFave¹⁵ notes that "some crimes are so defined that conduct accompanied by an intention to cause a harmful result may constitute the crime without regard to whether that result actually occurs."¹⁶ Examples of such crimes include perjury, forgery, and some other forms of reckless and negligent conduct.¹⁷ However, homicide is not such a crime. Rather, it is one of those crimes that is "defined in such a way that the occurrence of a certain specific result of conduct is required for its commission."¹⁸ Professor LaFave sets forth the standard applicable in the cases relevant to this Note with a series of short examples:

Page 708

A is not liable, of course, for murder though he shoots at B intending to kill, so long as his bullet hits no one; nor for manslaughter or battery though he drives his car recklessly, if his conduct fails to harm an endangered pedestrian or motorist. For these crimes of cause-and-result we must have at the very least the required fatal or injurious consequence.¹⁹

All of the cases in this Note deal with the substantive law of homicide and are therefore governed by the but-for causation standard, which is discussed in Part II.B.²⁰ Now that the causal context is clear, it is helpful to map out the general requirements for satisfying the but-for causation requirement in a homicide case.

B General Rule: The But-For Test

The general rule for crimes in which causation is an element of the offense is that the defendant's actions must be both (1) the actual or but-for cause of the result and (2) the legal or proximate cause of the result.²¹ Further, "when causing a particular result is an element of the crime charged, it must be proved by the prosecution beyond a reasonable doubt."²² Professor LaFave sets forth a more detailed description of those elements that make up the general rule for criminal causation, explaining:

[T]his ordinarily means (1) that the defendant's conduct must be the "but-for" cause . . . of the forbidden result . . . and in addition (2) that the forbidden result which actually occurs must be enough similar to, and occur in a manner enough similar to, the result or manner which the defendant intended (in the case of crimes of intention), or the result or manner which his reckless...