Retributive and utilitarian goals for criminal justice decisions are in conflict. In part, this is because the retributive aim rejects prediction, while all utilitarian aims require it. In the context of this debate, we review research concerning the prediction of violence, and conclude that because such predictions are of low accuracy, they are only modestly useful for public policy formulation or for individual decision-making. Because we believe prediction, and utilitarian goals, to be important, this paper focuses on two issues that have potential for increasing the accuracy with which predictions may be made. One is the measurement of the seriousness of crime and ways to improve it. Second, we introduce the concept of societal stakes and suggest that this must be assessed as well. Finally, we propose a model that may be useful for lessening the conflict between retributive and utilitarian perspectives.
Prediction is fundamental to science, and hence to the application of scientific methods to problems of crime and criminal justice. Prediction occupies a central position not only in criminological research, but in criminal justice decisions as well. Behavioral predictions are ubiquitous in the decision-making processes that characterize the criminal justice system-from the victim's decision to evoke the process, to the offender's final discharge from custody or supervision.
In addition to being central to the decisions made in the criminal justice system, prediction is central to all plausible social policies governing such decisions when those policies have a utilitarian or consequentialist purpose such as crime reduction (for example, through rehabilitative, incapacitative, or deterrent strategies) (Gottfredson & Gottfredson, 1988). Prediction, then, is among the basic concepts underlying both institutional (policy) and individual (case) decisions made about offenders against the law.
The major nonconsequentialist criminal justice goal for which this is not the case is that of deserved punishment (von Hirsch, 1976, 1985). Thus, two popular views on justifiable purposes for dealing with offenders-desert and utility-conflict fundamentally. One absolutely rejects prediction; the other relies upon it.
In the context of this tension, we review selected research aimed at the prediction of violence, and conclude that because such predictions are of very low accuracy, they are only modestly useful for public policy formulation or for individual decision-making. Because we believe prediction to be important, however, this paper focuses on two issues that have potential for increasing the accuracy with which such predictions may be made.
The first is the measurement of the seriousness of crime and ways to improve these measurements. The second requires that we frame decision problems somewhat differently than we have in the past, taking instruction from the informed and prudent gambler: we must consider and measure the "stakes" as well as the risks. Accordingly, after a discussion of offense seriousness, we will suggest and illustrate the development of measures of the related-but independent-concern of societal stakes. Finally, we propose a conceptual model that may be useful for lessening the conflict between punishment and utilitarian perspectives, and for the development of practical decision tools.
DETERMINANTS OF CRIMINAL JUSTICE DECISIONS
Decisions made about offenders, including those involving violent behaviors that most shock the public conscience, rely heavily on judgments about three issues (Gottfredson & Gottfredson, 1988). First, although measured differently in various studies, offense seriousness has been shown to be the dominant factor in all criminal justice decisions. This is true for the decisions of victims (e.g., whether to report a crime), of police (e.g., to arrest), of bail magistrates (e.g., to release on recognizance or to set a bail …