<<
>>

Combining deterrence and retribution

Some philosophers recently have suggested a sort of halfway posi­tion between Bentham and Kant. Respect for the distinction between guilt and innocence means that we must not inflict penal­ties on the innocent.

But even if someone is guilty, we may punish him or her only if the penalty is inflicted by a system that succeeds in deterring crime. It might be argued that if people were not deterred by punishments, then no good would be achieved by pun­ishing them, so we ought not to do it. The middle way is to say that punishment may be carried out for its deterrent effects, but only when it is applied to the guilty.

This middle way between Bentham and Kant is initially attrac­tive. But Bentham's way of justifying punishment by reference to overall utility also suggests another reason why it might be justified, even if deterrence were ineffective. Once a person has committed a crime, we might decide to lock them up, not because this would deter anyone else, but because it would stop them from doing it again. Once more, the disutility to offenders would be justified by the utility to potential victims of their potential future offenses.

Even this rationale is open, however, to the same sort of Kantian objection. If we lock criminals up because they are a danger to the public and call this “punishment,” why should we not lock up peo­ple who are a danger to the public before they have committed any crime? As psychological theory gets better it may become possible to predict who will commit crimes. We could try to treat such peo­ple, but if the treatment failed, what objection could Bentham have to locking them up? (Once more, there's a film that is based on this idea: Steven Spielberg's Minority Report.)

It is plain that Kant would have an objection to this sort of policy, too.

For a person who is going to commit a crime has not yet done anything to deserve punishment. It is one thing to punish someone for planning a crime or for conspiring to commit crimes with others, another to penalize a person who is going to commit a crime but has not yet formed an intention to do so. To inflict a penalty on such a person would, once more, be to treat him or her as a means to the public good, ignoring the question of whether the person was guilty of any offense.

We could modify the middle way, then, to say that we may pun­ish offenders in any way that contributes to the public good—by protecting the innocent or in some other way—and not just in ways that produce deterrence. We would thus keep the core of retribu- tivism—only the guilty may be punished—while taking into account the deterrence theorist's basic idea that we should do this only if some good comes from it.

But this concession will not satisfy the retributivist. For the ret­ributivist insists that punishment is retribution for an offense: not only can we punish only offenders, but we can punish them only for their offenses. And that means that, in some sense, the penalty inflicted must reflect the nature of the crime.

There are at least two ways in which it can be thought that the pun­ishment must “fit” the crime. I shall consider one less obvious way later. But the obvious way in which punishments may fit crimes is that there may be some proportion between punishment and offense.

Thus, suppose that Virginia has parked her car illegally and that a police car chasing an assassin has hit her car and thus allowed the assassin to escape. Suppose the assassin has killed a much-loved public figure. Then many of us might gain a great deal of relief if Virginia is severely punished, even though what she did—parking illegally—is not a serious offense.

Even if it would produce a great deal of utility for many people if an offender was severely punished (because, say, it satisfied a desire for revenge), it would be wrong, the retributivist says, to punish her more than she deserves.

Most people, I think, would accept these retributivist claims about punishment. However much good it may do for the rest of us, the degree of suffering we may impose on an offender must be lim­ited by the seriousness of the wrong he or she has done. More than this, a harm inflicted on a person who is innocent, or that is out of all proportion to the offense, should not be called a punishment at all. There are certain moral constraints internal to the concept of punishment—constraints captured in the idea of desert—just as the natural law theorists claimed there were certain moral constraints internal to the concept of law. Even if it is a good thing that punish­ment deters—even if there are reasons for increasing the efficiency of its deterrent effects, such as by publicizing trials and sentences— these are goals that we can use the criminal justice system to pursue only if we first respect the rule that the penalties must be deserved.

The difference between retributivists and deterrence theorists is another example of a dispute between the two types of moral prin­ciple that Nozick called end result and historical principles. Retributivists, unlike Bentham and deterrence theorists, require that we look beyond the end result of our system of punishment, beyond the allocation of utility that it produces; they require us to respect the historical principle that punishment should be given only to those who have done something to deserve it.

7.9   

<< | >>
Source: Appiah Kwame Anthony. Thinking It Through: An Introduction to Contemporary Philosophy. Oxford University Press,2003. — 425 p.. 2003

More on the topic Combining deterrence and retribution:

  1. Justifying punishment: Deterrence
  2. Diplomacy and deterrence
  3. Kant's universalizability principle
  4. TYPES OF INJUSTICE