Historically, judges have had considerable discretion over the determination of criminal sentences under so-called indeterminate sentencing. While such discretion theoretically allows judges to tailor sentences to the circumstances of individual crimes and criminals, thereby achieving a sort of ex post fairness, it also permits variation in sentences that may not be warranted by the observable facts of the case, reflecting instead the judge's own preferences. Moreover, judicial discretion makes it difficult for potential offenders to predict the consequences of their actions, thus impeding the deterrence function of criminal punishment. These and other dissatisfactions with indeterminate sentencing led to the establishment of federal sentencing guidelines in 1987, which greatly curtailed the discretion of judges (Champion, 1989; Freed, 1992).
The sentencing guidelines dictate punishment ranges for crimes based on the seriousness of the offense and the defendant's criminal history. Judges are then bound, with few exceptions, to select a sentence from that range. (1) And while judges retain some discretion under this regime, the guidelines clearly transferred much of the control over sentencing back to the legislature, which sets the ranges. The key question in this context concerns the optimal interaction between the stringency of legislative guidelines and the degree of judicial discretion within this sequential process, given that legislatures and judges act at different points in time and may hold differing views regarding the social function of punishment.
According to the standard economic model of crime, the primary goal of punishment is deterrence. (2) Sentences are therefore set prospectively, with the aim of inducing only those offenders who value a crime more than its social harm to commit it. In this theory, optimal sentences are independent of an offender's characteristics, reflecting instead the harm caused by the illegal act. Extensions to the basic theory, however, have recognized that other values, like fairness, may also be important in setting penalties. (3) Here, the characteristics of the offender as well as the crime become relevant in determining the appropriate punishment.
Broadly speaking, legislatures are in a better position to further the goal of deterrence by establishing fixed penalties before the fact, which then serve as signals to potential offenders regarding the harmfulness of certain acts. In contrast, judges are better suited to pursue fairness because they are able to take into account the circumstances of specific cases. Viewed in this light, the debate over sentencing guidelines can be seen as balancing these competing social values. The objective of this article was to provide some insights into this interplay between rules and discretion in criminal sentencing. (4)
The article is organized as follows. Section II sets up and analyzes the model. Section III then discusses the results in the context of sentencing reform and reviews the empirical evidence on the impact of reform. Finally, Section IV concludes.
II. THE MODEL
The structure of the model is as follows. First, potential offenders decide whether or not to commit criminal acts by comparing the gain to the expected punishment. Once an act is committed, the offender expects to be caught with probability p and is then subject to a sanction s. For simplicity, we focus only on fines, but our results generalize to other forms of punishment. (5) We also treat the probability of apprehension as fixed. Although in standard economic models of crime, both p and s are choice variables, in reality, the punishment and probability of apprehension are chosen at different points in time by separate government agencies. And since the focus here is on the punishment choice by legislatures and judges (neither of whom directly controls p), it does not seem inappropriate to treat p as a parameter. …