In recent years there has been substantial concern over the issue of whether the judiciary is a representative institution. Most research on the matter suggests that judges are to some degree sensitive to public opinion, but confusion exists over the process through which the public affects the courts. This research is designed to reduce the level of confusion.

Taking advantage of the quasi-experimental research design afforded by the “circuit” system of court organization used by the Iowa trial courts, this research investigates the “sharing” model of representation. In order to insure against spurious results, controls are introduced for three types of influences on sentencing decisions: defendant and case attributes, judge attributes and role orientations, and local system practices. Analysis of a path model consisting of measures of sentencing behavior, seriousness of crime in the local jurisdiction, and perceptions of crime seriousness reveals that one-sixth of the variance in sentences can be explained. However, substantial variation across judges in responsiveness to local norms is also discovered.

Further investigation of these data was undertaken to determine the factors that account for variation in the strength of environmental linkages. The analysis suggests that judges with greater contact with their constituencies, who have experienced electoral defeat, and who assume a “delegate” role orientation, are far more influenced by environmental factors.

The article concludes with some observations on how the process of recruitment and early career patterns affect representation.