Optimal Jury Design

• NCJ Number: 94053
• Journal: Jurimetrics Journal Volume: 24 Issue: 3 Dated: (Spring 1984) Pages: 218-235
• Author(s): A B Urken; S Traflet
• Date Published: 1984
• Length: 18 pages
• Publication Format: Article
• Language: English
• Country: United States of America

Annotation

The results of the presented computer simulation experiments suggest that changing a jury's voting system should be considered as a means of achieving due process.

Abstract

Analyses of how juries can be controlled to maximize jury competence have concluded that changes in group size or decision rule would have an insignificant marginal effect on the probability that a jury would make a correct choice. This study presents an alternative perspective on this conclusion based on a different model of the process of making a correct collective choice under different voting systems. The model used in the Condorcet jury theorem can be interpreted as a special case for a one person-one vote system. After showing how juror preferences and voting systems can be combined to produce different collective outcomes, this study used computer simulation experiments to compare one person-one vote and approval systems using two standards of evaluation: control of the collective outcome and control of the probability of making a correct choice. Under 'approval voting', the juror would be allowed to cast a single vote for each alternative he/she would approve as a verdict. In this system, the voter must decide on the minimum utility rating an alternative must achieve to receive an approval vote. Taking a negligence case as an illustration, the study found that as the number of possible correct choices increases, the probability of making a correct choice under approval voting exceeds the performance of a one person-one vote system as long as the decision rule is plurality. As one approaches unanimity, approval voting becomes inferior to one person-one vote voting; however, if the number of correct alternatives cannot be predicted, the average probability of making one, two, or three correct choices from two or three alternatives may be higher under an approval voting system than under one person-one vote rules. Twenty-eight footnotes are provided.