Foreword: the criminal law and the luck of the draw.

• Author: Kadish, Sanford H.
• Position: Supreme Court Review

I propose to consider what to make of a doctrine of the criminal law that seems to me not rationally supportable notwithstanding its near universal acceptance in Western law, the support of many jurists and philosophers, and its resonance with the intuitions of lawyers and lay people alike. This is the doctrine--the harm doctrine, I'll call it--that reduces punishment for intentional wrongdoers (and often precludes punishment for negligent and reckless wrongdoers) if by chance the harm they intended or risked does not occur. I will also consider a corollary of the harm doctrine which offers a full defense if it so happens that, unbeknownst to the defendants, the harm they intended could not possibly have been done.

Whether the harm doctrine can be justified is, as George Fletcher has said, a "deep, unresolved issue in the theory of criminal liability."(1) Indeed, a German scholar, Bjorn Burkhardt, recently concluded his comparative review of the law on this subject with the sobering words that "little progress has been made toward a solution of this issue in the last two hundred years."(2) He continued: "The arguments of the past still dominate contemporary discussion.... [H]ardly anything of substance has been added."(3) And he concluded: "In the final analysis, it is questionable whether a compelling and rational argument on this issue is possible."(4) That may well be so. The debate over the issue remains unresolved notwithstanding the earnest attention of generations of scholars. But though the ground is well trod,(5) the subject continues to have a fascination for those of us who worry about the criminal law (perhaps just because it has defied successful resolution), and I am not immune to its attraction.

I should explain at the start what I mean by saying that the harm doctrine is not rationally supportable. I mean that it is a doctrine that does not serve the crime preventive purposes of the criminal law, and is not redeemed by any defensible normative principle. Suppose, for example, the law provided that any crime committed during the night of a half-moon may not be punished with more than one-half the punishment appropriate on all other occasions. This distinction is patently irrelevant to any crime preventive purpose of the criminal law. Yet it might still be rationally supportable if it could be justified by moral principle. But no such principle can relate guilt or desert to the phases of the moon. Here, then, we would have an extreme instance of a rationally indefensible doctrine. One qualification: sometimes the law must defer to people's irrationalities to maintain the acceptance needed to govern. This might possibly be the case even with my half-moon doctrine of punishment. The doctrine, however, would still be rationally indefensible, even though its adoption by the law would not be.

Of course our criminal law has for centuries included many irrational doctrines--whole Augean stables full. Some of them were that way from the start. Others got that way when changed conditions made them anomalous, like the murder rule requiring the victim to have died within a year and a day of the injury.(6) But these differ from the harm doctrine in that they are widely recognized as insupportable, and their long persistence in the law is simply evidence that the law is slow to change. The harm doctrine is special (although, as we will see, not singular) in that large segments of the legal and lay community regard it as sound.

I will begin by setting out the law that most clearly exhibits the harm doctrine at work. This is the law governing the punishment of failed efforts to do some prohibited harm (the law of attempts) and of actions that create the risk of the harm without producing it (the law of culpable risk creation). These rules are well known and I will only sketch them briefly.

First, the law of attempts. Consider the case of a man who stabbed his son in anger, pleaded guilty and was convicted of a crime equivalent for our purposes to attempted murder.(7) After serving several months of a two year sentence he was paroled. However, three months later his son, who had been hospitalized since the attack, took a turn for the worse and died, whereupon the prosecutor, quite within the law, charged the father with murder, a crime punishable with life imprisonment or death.

What did the father do in jail or on parole that merited the greater punishment? Not a thing. If a good constitution or a good surgeon had saved the son, the father could not have been further punished. The occurrence of the resulting death alone raises the crime and the punishment. In most jurisdictions this same principle operates for all crimes, not just homicidal crimes.(8) In California, for example, an attempt to commit a crime is punishable with half the punishment for the completed crime.(9) Thus, the reward for failing, no matter how hard you try to succeed or how close you come, is a lesser punishment.

Now consider crimes of culpable risk creation-crimes in which a person is punished, not for attempting a harm, but for culpably risking it. The punishment of these crimes is also made to depend on chance. Take the case of Mr. Malone.(10) He and his friend decided to play a game of Russian Roulette in which each took turns spinning the chamber of a revolver, with one round in it, and firing at the other.(11) When Malone's turn came to pull the trigger the gun fired and killed his friend.(12) Malone was convicted of second degree murder, based on the egregious risk to life he needlessly created.(13)

That sounds fair enough. But suppose instead, that the bullet only inflicted a flesh wound, or that the bullet was not in the firing chamber when Malone pulled the trigger. Could Malone then have been convicted of any crime? Perhaps he could have been convicted of some ad hoc statutory offense concerning firearms, but such an offense would carry nothing like the penalty for murder. And if there had been no special statute of this kind, he could not be convicted of any crime at all, since traditionally just recklessly endangering another was itself not criminal--except in specific contexts, like driving a car. Some jurisdictions have in recent years made it criminal to recklessly endanger another person in all situations, but even these statutes treat the offense as a minor one.(14)

Finally, I need to mention one more doctrine that exhibits the law's preoccupation with a resulting harm. It is the doctrine of impossibility, which takes the harm doctrine one step further. The harm doctrine calls for a lesser punishment when no harm was done; the impossibility doctrine calls for no punishment at all when the harm could not possibly have been done. Though now a minority view, it still has its defenders, both on and off the bench.(15)

The gist of this doctrine is that a serious effort to commit a crime, even one which includes what the actor thought was the last thing he needed to do to commit it, is not punishable if the crime could not have succeeded. For example, assume a person uses lies and deceit to dupe another into giving up his property, but all to no avail because the owner sees the lies for what they are. Can the would-be deceiver be convicted of attempting to obtain the property of another by false pretenses? Not under the impossibility doctrine--since the owner saw through the lies, the defendant could not possibly have succeeded.(16) In another example a hunter, meaning to shoot a deer before the hunting season, shoots at a straw dummy the game warden erected, with reflectors for eyes to give it verisimilitude. Is this an attempt to take a deer out of season? No, said the court: "If the [s]tate's evidence showed an attempt to take the dummy, it fell far short of proving an attempt to take a deer."(17) A final example: Professor Moriarty shoots at what he takes to be Sherlock Holmes, but which in reality is the shadow of Holmes' paper cutout profile that Holmes has set revolving on a phonograph in front of a lamp. Too clever by half, Mr. Holmes. Moriarty escapes again, this time thanks to the impossibility doctrine, since his action--shooting at a shadow--couldn't possibly kill Holmes.

Today, these impossibility cases would go the other way in most jurisdictions--though the attachment of courts to the doctrine is sometimes remarkable.(18) But with the exception of the impossibility doctrine, all of the doctrines I have described are essentially still the law in most places.(19)

Having illustrated the workings of the harm doctrine, the two major tasks I have set for myself lie ahead. First I must make good, if I can, my claim that the doctrine, in all its applications, is not rationally defensible. Second, I need to consider what to make of the durability of this doctrine.

1. THE RATIONAL INDEFENSIBILITY OF THE HARM DOCTRINE

   To make my case that the harm doctrine cannot be rationally defended, I must establish two things: (a) that the doctrine cannot be justified in terms of the crime preventive purposes of criminal punishment; and (b) that neither can it be justified in terms of any convincing principle of justice.

   1. THE ARGUMENT FROM THE PURPOSES OF PUNISHMENT

      Does the harm doctrine further the law's interest in crime prevention? There are two main ways in which criminal punishment is thought to reduce crime. One is by preventing further criminal acts by the offender. The other is by discouraging criminal acts by others. How far the law succeeds in particular times and places in attaining this goal is an empirical question I need not pursue here, for the question is whether, on the law's premise that punishment does work in this way, the distinction in punishment required by the harm doctrine is defensible.(20) Let's first consider the goal of preventing further crime by the offender being punished.

      Convicting offenders serves to identify those who have shown themselves to threaten further breaches of the law...