The "desert" model for sentencing: its influence, prospects, and alternatives

Social Research, Summer, 2007 by Andrew von Hirsch

INTRODUCTION

THE DECLINE OF THE REHABILITATIVE ETHOS IN SENTENCING THEORY in the post-1960s is a story that has been told often (see, for example, Allen, 1981) and need not be rehearsed here. * Penal treatment programs, once tested for their effectiveness, showed scant success--or at most, succeeded only in limited categories of cases. Doubts grew also about the fairness of making the severity of a person's sentence depend upon his responsiveness to treatment. As penal rehabilitation diminished in influence, the key question for penologists and reformers became: what conception of sentencing should follow it? My article, as well as Michael Tonry's and John Donahue's in the present volume, is devoted to this question, of the choice of sentencing rationale.

I will focus, particularly, on a sentencing theory that became influential in the late 1970s and 1980s in this country (see Tonry in this volume) and western Europe (Bottoms, 1995; Jareborg, 1995; Hornle, 1999) and continues to have considerable influence today (von Hirsch and Ashworth, 2005, chap. 1): namely, the proportionalist or "desert" model for sentencing.

PROPORTIONATE SENTENCING: THE "DESERT" MODEL

The idea of proportionate, deserved sentences has had a curious history of long being ignored, and then becoming very influential. Judges, as a matter of practice, have traditionally based their sentences in considerable part on the seriousness of crimes. Philosophers for centuries debated retribution as the basis for punishment. In modern sentencing theory, however, retribution was scarcely taken seriously: criminologists either ignored the idea or dismissed it as obsolete, reactionary, or obscure. (1)

The interest in desert theory originates from the mid-1970s with publication of several works on reform of sentencing policy, including my own Doing Justice (von Hirsch, 1976), the report of the Committee for the Study of Incarceration. Once broached, the idea quickly became influential in penological thinking in the United States and Europe, (2) and a number of jurisdictions explicitly drew on it in their sentencing reform efforts. (3)

This literature addressed two of the common objections to traditional versions of retributivism. One objection was that the notion of deserved punishment was essentially incomprehensible, that it rested on "metaphysical" notions such as the requital of evil for evil. Desert theorists provided a simpler explanation, however. Punishment, they pointed out, is a blaming institution. The difference between a criminal and a civil sanction lies, generally, in the fact that the former involves censure of the actor for his criminal conduct (von Hirsch, 1993, chap. 2; von Hirsch and Ashworth, 2005, chap. 2). Fairness thus requires that penalties be allocated consistently with their blaming implications. The severity of the punishment (and thereby its degree of implied censure) should comport with the blameworthiness (that is, the seriousness) of the defendant's criminal conduct. Disproportionate or disparate punishments are unjust, not because they fail to requite suffering with suffering, but because they impose a degree of penal censure on offenders that is not warranted by the comparative reprehensibleness of their criminal conduct.

The other objection was to the harshness of retributive punishment, to its apparent exaction of an eye for an eye. Desert theorists' response has been that desert does not require visitation of suffering that is equal that done by the offender to his victim. What is required, instead, is punishments that are proportionate to the seriousness of the criminal conduct. Proportionate punishments may be imposed without increasing (indeed, while substantially decreasing) prevailing severity levels, (4) so long as penalties are ranked in the order of a crime's seriousness, and equally reprehensible criminal acts are penalised with roughly comparable severity.

It should be noted that the two conceptions are linked. Censure, on this perspective, should not be regarded as mere denunciation of crime and criminals. It should be understood, rather, as a normative communication that views the actor as an agent capable of moral deliberation and as having continued status as a member of the community. High overall severity levels, such as those involved in talionic punishments, would thus be unacceptable--because they would operate purely in terroram, and would fail to recognise the actor's continued status as a moral agent and member of the community (see von Hirsch, 1993, chaps. 2 and 5; von Hirsch and Ashworth, 2005, chaps. 2 and 9).

The central principle of sentencing, according to this view, is the principle of proportionality: that sentences are to be proportionate in their severity to the seriousness of the defendant's criminal conduct (von Hirsch, 1976, chap. 8; von Hirsch and Ashworth, 2005, chap. 9). The main criterion for deciding the quantum of punishment is thus retrospective: the seriousness of the violation the defendant has committed. Future-oriented considerations--for example, the defendant's likelihood of returning to crime or the expected deterrent effect of the penalty--should not substantially determine the comparative severity of punishments. (5) On this approach, imprisonment--because of its severity--should be visited only on those convicted of the more serious felonies. For less serious crimes, penalties less severe than imprisonment are to be used. The degree of intrusiveness of these noncustodial sanctions is to be determined not by predictive (or rehabilitative) considerations but, again, by primarily the degree of gravity of the criminal conduct. Warnings, limited deprivations of leisure time, and monetary penalties are among the sanctions to be used. The desert literature has thus supported penalties well below the prevailing severity levels in the United States and most other jurisdictions. (6)