Economics Research Seminar Series: Michelle Baddley

Trial by jury is a long-standing democratic institution with its historical origins in Anglo-Saxon England, based on the historic principle that defendants will be judged more fairly by a panel of their peers than by a judge. This principle relies on an assumption that a jury will reach its verdict via a rational and objective assessment of the facts and evidence presented at trial, with the judge confined to providing the jury with guidance on matters of law. Unfortunately, the jury system is susceptible to complex range of behavioural and social biases – not just in terms of the jury’s deliberations amongst themselves but also in the form and substance of the information presented to them. Identifying the nature and extent of juries' decision-making biases is complicated by the secrecy which necessarily surrounds juries and their deliberations, and most behavioural economic research is done via “mock juries” or jury surveys. This paper presents an alternative approach focussed on quantifying differences in probability of conviction depending on whether the decision to convict or not is taken by a judge (via a "bench trial") or a jury. This method for assessing potential jury bias is constrained in Australian jurisdictions because defendants have limited choice between a jury trial and a bench trial. However, US jurisdictions commonly offer defendants this choice and this paper presents an econometric analysis of US Federal Court Caseload statistics data 2001-2023. This analysis suggests that the probability of being convicted by a jury is significantly greater than the probability of being convicted by a judge, a worrying finding given the substantial economic and non-economic costs associated with wrongful convictions.