Back to 2026 Programme

Bias and Accuracy in Jury Selection

Fontainebleau, France 13 July 2026 – 15 July 2026

Amna Salam (UCSD)

A1 AI, Markets, and Algorithmic Decisions
Chair: Sepehr Shahshahani
Amphi De Vitry
Political science / Institutions and organizations in the public sector

Abstract

Members of juries and other deliberative bodies can arrive at different decisions than they would reach individually. The role of how information is aggregated in jury-type settings has been well studied, but how this impacts who gets selected to serve on juries has received less attention. When forming a jury, attorneys consider how its composition will affect the outcome of the case. I propose a framework in which jurors, who vary in the accuracy of their information, as well as preferences against wrongful conviction or acquittal, decide over a case according to a unanimous rule. The defense and prosecution observe the differences among jurors and can make peremptory strikes. I find that under some conditions litigants may prefer to strike better informed jurors; and that whether more information is preferable depends on the bias of the jurors.

This paper has been marked as unpublished by the author.