In January, the U.S. Supreme Court decided Burrage v. United States. The Court there interpreted the eligibility of a heroin-distributing defendant for a sentencing enhancement under the penalty-enhancement provision of the Controlled Substances Act, for selling drugs from the use of which “death . . . resulted.” The Court held that a defendant qualifies for the enhancement only if the prosecution proves to the jury that the particular drugs the defendant sold were the “but for” cause of the user’s death (unless, perhaps, the prosecution can prove instead that those drugs would have been sufficient to cause the defendant’s death, even if he in fact simultaneously took other drugs that were also sufficient to cause his death). I argued in an earlier column, parts 1 and 2 of which can be found here and here, for the result that the Court has now reached, so I am pleased with that outcome. In this column, I will explore further whether and why establishing causation should play such a crucial role in determining guilt and punishment in our criminal justice system.
Causation in Burrage: But-For Cause Versus Contributing Cause
In Burrage, the defendant, Marcus Andrew Burrage, distributed heroin to Joshua Banka, and the latter subsequently died after injecting that heroin. In addition to using heroin, however, Banka had simultaneously taken several other potentially dangerous drugs, including oxycodone. Experts testified at Burrage’s trial that the heroin contributed to the odds that he would die, but they would not say either (1) that absent the heroin, Banka would have survived, or (2) that absent the other drugs, Banka would have died from the heroin alone.
In other words, the prosecution failed to satisfy the ordinary meaning of the phrase “resulted in,” whereby the item that “resulted” in an outcome was necessary to that outcome or, perhaps, was sufficient to cause that outcome, even if it was not necessary (because of the other drugs taken). The United States maintained that a lesser standard of “contributing cause” would satisfy the causation language and that Burrage was therefore properly convicted of count 2 of the indictment, which alleged the unlawful distribution of heroin from the use of which death resulted.
The Supreme Court unanimously rejected the government’s argument and concluded that there was insufficient evidence, given the language of the statute, to support a conviction for distribution that resulted in a death. This ruling seems faithful to the way in which we use words like “caused” and “resulted in,” both in legislation and as a general matter. And because a potentially ambiguous criminal law must be interpreted to favor the more lenient reading, one might describe the case as a “no-brainer.”
Does Causation Really Matter?
A different question, however, that in a sense precedes the statutory interpretation issue is why causation plays such an important role in determining guilt and punishment. Why does the outcome of a person’s behavior play a central role in determining how harshly we judge the person’s behavior?
The most dramatic illustration of the significance of causing harm in the criminal law may be the critical role that causation plays in determining eligibility for the death penalty. The only crime against the person that qualifies a defendant for the death penalty is murder, an offense that contains a causation element: the defendant’s conduct, intended to bring about the death of his victim, must in fact bring about the victim’s death.
Say that two defendants, [A] and [B], each load a pistol, point it at an intended victim’s head ([C] and [D], respectively) and fire. Both [A] and [B] hit their mark, and both victims, [C] and [D] are subsequently rushed to the hospital in ambulances. A brilliant surgeon operates on [C] and is able to save [C]’s life, while [D] dies of his wounds, because his surgeon is competent but less able than [C]’s surgeon.
In judging the behavior of [A] and [B], it is difficult to identify a moral distinction between the two; their conduct is identical and the surrounding circumstances at the time of their behavior in the two cases were identical as well. There is no basis for concluding that [A] is a better person than [B], that [A] is less dangerous than [B], or that [A] did a better job of controlling his murderous impulses than [B] did. The survival of [C] and not [D] was a fortuity that turned on the respective competence of the two victims’ surgeons, not on anything for which [A] or [B] could plausibly be credited or blamed.
Yet consider the fact that [A] now cannot be found guilty of homicide and therefore cannot qualify for the harshest penalty that our constitutional system countenances: the death penalty. [B], on the other hand, may be death-eligible. Similarly, if [A] and [B] both aim and fire at their respective victims, but [C] suddenly and unpredictably drops dead of a stroke before [A]’s bullet hits him, there too, [A] cannot be guilty of murder (because although [A] acted with the intention of killing [C] and [C], in fact, died, [A]’s action did not cause [C]’s death) and thus [A] may not be executed.
One might observe that, in general, when an action brings about its intended result, it is more likely that the actor truly intended the result than when the action did not bring about its intended result. Therefore, as a rule, those whose actions actually cause a harmful outcome are probably more culpable than those whose actions fail to cause that outcome.
Yet this fact does not truly address the problem of distinguishing successful from failed attempts in either of the above two cases, because the crime of “attempted murder,” or attempted anything else, necessarily requires the prosecutor to prove that the defendant really did intend the hypothesized outcome. To the extent that the failure of the attempt gives rise to reasonable doubt about the real intention of the defendant, then, the jury must acquit the defendant of the attempt crime. Conversely, achieving a particular result does not itself guarantee the actor’s intention, and the prosecutor must therefore prove intent in such cases as well, notwithstanding the existence of causation. Juries must therefore make an actual determination about a defendant’s intentions, notwithstanding the ordinary factual coincidence between intentions and their intended consequences.
Asking whether causation truly matters is another way of asking whether harm truly matters. That is, if your actions bring about a harm, then do those actions thereby become more blameworthy in some sense than if they bring about no harm? Can moral luck fairly play a direct role in determining punishment?
Our everyday interactions and behavior would suggest that the answer is yes. When our children, for example, are careless at dinner in a manner that risks breaking a glass, we are far more likely to be upset with them if the risk comes to pass (and the glass shatters) than we are if the glass remains intact. When the glass breaks, we are upset about both the children’s conduct (and the negligent or reckless state of mind that drives that conduct) and the outcome that directly followed from that conduct and state of mind. And when a drunk driver actually kills a pedestrian, we experience far greater moral outrage at the drunk driver than we do if the drunk driver merely swerves around on the road, but hits no one. Our blaming practices, then, do seem routinely to take causation of harm into account.
Heroin Resulting in a Death
When a drug-dealer, whom we will call Dave Dealer, sells heroin to a customer, Carl Customer, the very act of the sale risks bringing about a death. Heroin use is a risky activity, and people who purchase heroin from a dealer are—a good deal of the time—in less than complete control of their impulses. When you sell a person heroin, you accordingly generate a risk of the customer’s dying through the use of that heroin. For that reason, when a customer such as Carl can be proven to have died because he ingested the heroin that Dave sold to him, it may seem fair to blame Dave for that death (though the user probably bears some responsibility too).
Yet the moral luck question still plagues us here. Imagine a different drug-dealer, Diana Dealer, who sells exactly the same quantity and quality of heroin to Colleen Customer (no relation to Carl), that Dave sold to Carl. Colleen, however, who is as addicted (or as unaddicted) as Carl, injects a non-lethal quantity of heroin, while Dave’s customer Carl inadvertently injects a lethal dose. Why must Dave be sentenced to a mandatory prison term of twenty years under federal law, when Diana might be sentenced to substantially fewer years’ imprisonment for her conduct? The two people carried out precisely the same act, with exactly the same degree of recklessness with respect to the safety of their respective customers. The difference in outcome thus turned entirely on what the customers later went on to do, and that distinction had no connection whatsoever to anything distinct that either Dave or Diana did.
To highlight the potential unfairness of punishing Dave more severely than Diana, consider the facts of Burrage. The dealer there sold heroin to his customer, Banka, and the customer used that heroin and then died. Experts could not say definitively whether the heroin that Banka used was what caused his death, or whether he would have died anyway, because of the other drugs that he took. Judging things from Burrage’s perspective, however, the fact of whether Banka’s death did or did not result from the heroin he purchased lay completely outside of Burrage’s control. If it was the oxycodone or the clonazepam, and not the heroin, that killed Banka, in other words, this is not something for which we can credit Burrage.
And yet I argued in my own column, referenced above, that Burrage should not be punished as someone whose heroin sale “resulted” in a death, because the prosecution was unable to prove that Banka would not have died “but for” the heroin. And I took this position, despite the fact that it is difficult to imagine how Burrage might have done things differently if he had known that the district court would apply a less demanding standard to establishing causation than “but for” cause. Burrage was surely on notice that his act of selling drugs could cause a death and that he would be punished more severely than otherwise if it did in fact cause a death.
In explaining the law’s attaching greater blame to wrongdoers who cause harm than to wrongdoers who fail to cause harm (despite identical states of mind and conduct), it is useful to turn once again to the Supreme Court’s death penalty jurisprudence, where we saw that causation is an essential element of eligibility for the harshest permissible punishment of death.
The Supreme Court held in Woodson v. North Carolina that a mandatory death penalty violates the Eighth Amendment ban on cruel and unusual punishments. This holding, coupled with rulings in a series of other cases including Eddings v. Oklahoma, mean that no matter how heinous a murder is (and how plainly death-eligible a defendant is), the sentencing body must—as a constitutional matter—have the option of extending mercy to the defendant and sparing his life. This constitutionally-mandatory option necessarily results in there being some defendants who have committed precisely the same heinous and horrible acts that other defendants have committed, who have been caught doing so, and who have had exactly the same aggravating (and mitigating) factors on the basis of which a sentencing decision could be made, and yet some of those people will be sentenced to death, while others will not, as a consequence of constitutional doctrine.
By the Court’s own lights, then, people who are deserving of a penalty in virtue of what they did (and in virtue of the outcome) are nonetheless constitutionally entitled to be spared that penalty if the sentencing body (a jury or a judge) chooses to exercise mercy. In some sense, this means that, for the Court’s purposes, mercy as a value may be more important than absolute fairness.
If mercy is this important, then it may help explain the role of harm-causation in crime and punishment. When a person intentionally (or otherwise culpably) acts in a manner that could bring about a desired (or known, or risked) outcome, then that person has essentially done everything in his power to merit the severe punishment associated with those whose attempts/risks are successful. Nonetheless, we feel merciful toward the drunk driver who hurts no one, or the drug dealer whose customer survives his heroin use. This mercy may translate into a less severe penalty (and, sometimes, a less severe criminal offense) associated with the failed attempt/risk.
It may be that when a person’s actions bring about the precise harm the risk of which renders the actions wrongful, we feel the full wrath associated with the culpable conduct. If, on the other hand, the actor is lucky enough to fail to bring about the desired/known/risked result, then that failure extinguishes some of our righteous indignation and reduces our will to visit harsh punishments upon the actor. As we do with the negligent child who fails to break the glass, despite her negligent behavior, we have an easier time extending some version of forgiveness to those whose bad conduct avoids inflicting injury and harm in the world. It is perhaps the criminal law version of “No harm, no foul.”