Perhaps the greatest attraction of the utilitarian view of punishment presented so forcefully in Smart's paper is its empirical character. This empirical, scientific approach to social and moral problems typical of utilitarianism is made possible by its monistic nature: by its claim to have found in consequences of actions the principle of a uniform interpretation of all the various moral concepts, rules, ideals, a universal moral currency in which all the pros and cons of any moral issue that might arise can be cashed. This moral monism is seen by utilitarians as one of the main advantages of their theory; but it is also a source of some of its most persistent troubles. For whereas some moral notions may lend themselves to utilitarian interpretation, some others seem resistant to it.

1) Just so soon as I was asked to send in a final, revised version of my “Retrospect and Prospect, Retribution and Deterrence” for eventual publication, it became obvious to me that all the alterations or additions which I wanted to make were alterations or additions in response to the “Comments” by Igor Primoratz. But, if I were to make the changes which I now believe to be necessary, in that original paper, then those Comments would be left in the air — apparently provoked by nothing at all. So instead I present my afterthoughts here, appropriately, as what they are, Responses to his Comments.

First, and urgently, we need to introduce a distinction between justifying (in general) and justicizing (in particular).

There is no single “problem of punishment”. The activity of punishing is problematic at many levels. There are problems relating to the authority to punish, the severity of punishment, the method of punishment, and of course the moral basis for punishment. Given this, we should be surprised were we to find that all these problems could be comprehended by a single, simple rationale. Nevertheless, there is one, fundamental, underlying moral challenge posed by all punishment, and it is especially (though not exclusively) to this that the classical justifications of punishment have been primarily directed. “Retributivist” and “utilitarian” justifications, however we understand them, are, above all, responses to what I see as this central problem of punishment.

A consequentialist holds that systems of criminal punishment must be justified, if they can be justified at all, by their consequential benefits. It is a contingent fact, if it is a fact at all, that these benefits are most efficiently attained by a system of punishment, and by punishments which ordinary moralists would regard as just; and a thorough-going consequentialist must be ready in principle to justify punishments which many would condemn as unjust. A retributivist, on the other hand, insists that the justice of a system or instance of punishment is essential to its justification; and that the demands of justice cannot be reduced to those of consequential utility.

Retributivist accounts of punishment come in a variety of forms: we must distinguish those who insist only that guilt is a necessary, not a sufficient, condition of justified punishment from those who insist that punishment must be fully justified by reference to a past offence; and amongst the latter we find various accounts of how it is that an offence justifies or requires punishment. But essential to any retributivist account is the claim that an adequate justification of punishment must cite not, or not only, its consequential benefits, but its relationship to a past offence: it is a non-consequentialist requirement of justice that punishment must be of an offender, for an offence; that only the guilty may be punished, and that the severity of their punishments should be limited, if not completely determined, by the seriousness of the offences for which they are punished.

In this comment I wish to defend Antony Duff's final conclusion; namely, that punishment is to be seen as a deterrent system constrained by requirements of justice. I shall do this, however, mainly by using John Kleinig's arguments. My aim is to show that although Kleinig himself rejects deterrence as part of punishment's justification, some of his arguments can best be defended if deterrence is taken into account. In the final section of the comment I shall return to Duff's arguments in this context, but only in order to raise some doubts as to their underlying assumption.

Before doing so, let me compare, briefly, Duff's and Kleinig's attitudes which seem to be relevant to the purpose of this comment.

Both Kleinig and Duff believe that, ideally, punishment is to be justified by retributive considerations. According to Kleinig, even in such an imperfect society as ours, punishment should still remain retributive.

The moral argument contends that crimes deserve retribution, that to impose the deserved punishment is to do justice, and that justice is a moral requirement, an end in itself, independent of any useful effects. The moral desirability of justice can be explained, but not demonstrated in non-deontological fashion. In penal matters, justice and deserved retribution are synonymous. It is for the sake of justice that we try to punish the guilty and not the innocent. Innocents do not deserve punishment. Because of justice, we would not impose punishment on them, even if it were useful to do so.

Argument over the death penalty — especially in the United States during the past generation — has been concentrated in large part on trying to answer various disputed questions of fact. Among them two have been salient: is the death penalty a better deterrent to crime (especially murder) than the alternative of imprisonment? Is the death penalty administered in a discriminatory way — in particular, are black or other nonwhite offenders (or offenders whose victims are white) more likely to be tried, convicted, sentenced to death, and executed than whites (or than offenders whose victims are nonwhite)? Other questions of fact have also been explored, including these two: what is the risk that an innocent person could actually be executed for a crime he did not commit? What is the risk that a person convicted of a capital felony but not executed will commit another capital felony?

It seems that nothing new can be said about the issue of capital punishment. And of course even this apologetic proposition itself is hardly original as is attested by the opening remark of Justice Mosk's paper. Yet, this is not an uncommon plight for a commentator in a philosophical conference on capital punishment, and particularly so when his task is to join into a long, comprehensive, and sophisticated debate between two of the major contestants in this battle. Actually, I feel almost like a child invading the privacy of its parents engaged in one of their quarrels. I use this metaphor because, for me, the names Bedau and van den Haag connote the “pro” and “con” attitudes to capital punishment since my philosophical childhood. I still have my notes, taken twenty years ago while reading as a student the exciting exchange in Ethics between these two sharp, persistent, uncompromisingly critical philosophers. They have introduced me into the subject which has interested me since, without of course imagining at the time that I will ever have the chance to get personally involved in their debate.

In a way, I suppose, everything has been said about the death penalty that can be said. Yet we continue to discuss the penalty, the legal processes involved, the actual means of execution, the crimes for which death is to be imposed, and at the cornerstone, the moral justification for society deliberately taking human lives.

At the risk of appearing immodest, I claim to be peculiarly equipped to enter into this discussion because I have been on all sides of the issue — not, I hasten to explain, because of unconcern or ambivalence. First, as an idealistic young man, I debated for abolition of the death penalty. Then, as executive secretary to the Governor of California, I had the duty of actually interviewing, in prison, men — and one woman — under sentence of death and making a report to the Governor. He, in turn, had the responsibility of commuting some sentences to life and allowing others to go to their death.

I, too, want to express my gratitude to Mishkenot Sha'ananim and to Ms. Karin Moses for their organization of this Conference and for their accommodating hospitality, and to Prof. Igor Primoratz of the Hebrew University for his thoughtful invitation to me to participate and for his almost inexhaustible patience in waiting for my paper.

I should mention also the pleasure of being with so many distinguished scholars from Israel, Europe and America, some of them old comrades (or antagonists) in the work for social justice, all of them, I fear, the intellectual betters of someone who, like myself, is (as it were) a professional activist rather than a systematic thinker. I am honoured to be in their company.

One of the most disturbing aspects of examining the extensive capital punishment debate, with its clear indications of discriminatory practices, ambiguous judicial directives, undeniable miscarriages of justice, controversial statistical data, and inept, inconsistent and/or unjust implementation, is the constantly gnawing thought that if this is the situation vis-à-vis what is considered the most extreme penalty with its special super due process, then what is happening in the cases of lesser penalties. These latter cases of petty thieves sentenced to years of incarceration for relatively minor delinquencies, of accused inadequately defended without appeals being lodged within the judicial system or public interest shown, of compulsory life imprisonment without parole, no doubt reflect all of the faults and errors of capital punishment.

This paper has two parts. The first (sections I-III) compares retributive theories measuring criminal desert largely by the harm the criminal did or risked, with a theory measuring it by the unfair advantage the criminal necessarily took, what I shall call “the fairness theory”. This part of the paper summarizes arguments I have made elsewhere. The second part (sections IV-V) defends the fairness theory against five objections recently made against it by two important theorists, Andrew von Hirsch and Hyman Gross. The objections are variations on the charge that the fairness theory of criminal desert is implausible, incoherent, or otherwise fundamentally flawed. Each objection collapses almost as soon as one or another standard distinction is made. I conclude the paper wondering why theorists have not noticed how weak these objections are.

Recent philosophical writing about punishment has been devoted mainly to the “why punish?” question. While a variety of theories — utilitarian, reciprocity based, and expressive — have been put forward to answer this question, there has been wide agreement among philosophers that the institution of legal punishment should be retained. Proponents of the various theories disagree chiefly why institutionalized punishment is necessary.

The philosophical writings have paid comparatively little attention to the criteria for distributing punishments — particularly to the criteria for deciding how much to penalize convicted offenders.

Von Hirsch has presented me with a dilemma. As he says, “Proportionality in the Philosophy of Punishment” is a “substantially revised” version of the paper originally prepared for the Jerusalem conference. Its predecessor, “Allocating Penalties”, was in part a response to my “Criminal Desert, Harm, and Fairness”, the paper I prepared for the conference. “Proportionality” is instead (if only in part) a response to what I said at the conference — the “Postscript” cited in its footnotes.

Here is the dilemma. On the one hand: I could leave “Postscript” more or less as I gave it. That would preserve the historical record, but at some cost. I would have missed a chance to advance discussion of important issues. I would also have caused an anomaly in the order of papers printed here. On the other hand: I could rewrite “Postscript”, making it a response to “Proportionality” rather than to “Allocating Penalties”.

Any system of justice purporting to be civilized must pursue two basic principles: (1) that people similarly circumstanced are to be treated equally under the law; (2) that people administering the law are not permitted to act arbitrarily or to prescribe individually the rules governing their actions and decisions. Both of these principles were persistently violated through well over half the twentieth century in the sentencing of people for crimes in the United States. The steady violations were produced by a combination of sentimental good intentions, puritanical severity and irrational misconceptions concerning the effects of punishment and the capacity of those commissioned to administer punishment. As the century wanes, the long course of error has been identified. This is a time of intensive sentencing reform. We are vexed now in a familiar way by doubts about the efficacy and the side effects of the reforms.

In thinking about the sentence-setting process, it is important to recognize the different perspectives associated with the two different levels at which sentencing questions must be addressed. The first is the macro policy level. This is the level of issues faced by a judicial council, a legislature, or a sentencing commission. Such a body must establish a range of permissible sentences from which the individual judge chooses the sentence for any particular case. The policy body must develop principles for establishing those ranges. Most of the discussion in this paper appeals to this policy level.

The second level is the micro level of the individual judge who must choose, from the range of sentences that the policy permits, the particular sentence that is most appropriate for the particular case. As part of its policy, the policy body may also provide guidance to the individual judge on how to choose a particular sentence from within the available range.