PERCEPTIONS OF CRIME SEVERITY: AN ASSESSMENT OF SOCIAL CONSENSUS (CLUSTER)

During the 1970s penal philosophy moved toward what has become known as the Neo-Classical Revival. The \u27just deserts\u27 model of punishment has assumed substantial importance, despite the fact that questions remain respecting its philosophic grounding and practical implications. One such issue involves the basis for the generation of proportionality-based punishments. Although a number of potential candidates have emerged in the literature, social consensus regarding the severity of crime seems to be well suited to providing such a basis. Given the potential of this approach, the task then becomes the measurement of social consensus. If citizens agree on the severity of crime then it is perhaps reasonable to argue for the use of such perceptions as the basis for schemes of punishment. The research reported in this work is thus concerned with citizen attitudes toward a wide variety of crimes. It examines the hypothesis that citizens tend to agree in their perceptions of the severity of crime. Data obtained from the National Survey of Crime Seriousness was analyzed utilizing three varieties of cluster analysis. Cluster analysis is a technique capable of discerning \u27natural\u27 clusters in data. If respondents agree on the severity of crime, their responses should tend to aggregate in clusters. The three techniques used included complete linkage, average linkage, and k-means analyses. The results of all three analyses support the view that citizens agree in their perceptions of the severity of crime. A number of limitations of the cluster techniques utilized prohibit unqualified rejection of the claim that citizens disagree about the seriousness of crime. Irrespective of the impact of these potential problems, several more fundamental issues suggest caution in applying these findings to the creation of penal policy