What Happens in the Jury Room? A Process Nobody Is Allowed to Examine

Deliberations are private and in several systems the secrecy is protected by law, with jurors prohibited from disclosing what was said and researchers prohibited from asking. The reasons offered are that jurors must feel able to speak freely without later scrutiny, that verdicts would be undermined if every decision could be picked apart afterwards, and that jurors would be exposed to pressure and harassment. The cost is that the central decision-making process in a criminal trial is the one part of it nobody may examine, so the evidence about how it works comes from simulated juries, from post-verdict interviews where they are permitted, from jurisdictions with different rules, and from the small number of cases where something went wrong conspicuously enough to be litigated.

Simulation studies produce consistent findings that are worth knowing:

• •The majority position at the first vote predicts the final verdict in the great majority of cases
• •Deliberation more often confirms initial leanings than reverses them
• •A small number of jurors do most of the talking, and participation correlates with status outside the room
• •Groups instructed to reach unanimity deliberate longer and consider more evidence than those permitted a majority
• •Juries understand judicial instructions considerably less well than courts assume
• •Groups tend to move towards a more extreme version of their initial leaning

The criticisms are better known than the findings in favour, which are substantial. Twelve people pooling recollection reconstruct the evidence considerably more completely than any individual does, and errors by one juror are frequently corrected by another, which is the single strongest argument for a group. A jury brings knowledge of ordinary life that judges as a group do not have, which matters for assessing whether an account is plausible. It distributes a decision that would otherwise rest on one person. It is difficult to corrupt, since jurors are numerous, anonymous and temporary. And comparisons between verdicts and the opinions of the trial judge find agreement in a large majority of cases, which suggests juries are not reaching eccentric conclusions.

Selection determines a great deal and the arrangements differ sharply between countries. Most systems draw randomly from a list such as the electoral register, which is only as representative as that list, and people who move frequently or are not registered are systematically absent. Excusals and disqualifications remove further groups, with some occupations formerly excluded entirely and most systems now excusing far fewer people than they once did. Failure to attend is a substantial problem wherever summonses are ignored without consequence. The United States permits extensive questioning of potential jurors and challenges by both sides, which produces a selection process that can last longer than the trial and has generated a professional industry. England and Wales permit almost no questioning, on the view that random selection is the point.

The documented problems concentrate in a few areas. Comprehension of complex evidence, particularly scientific and financial evidence, is limited, and the same is true of any lay tribunal. Understanding of legal directions is measurably poor, with studies finding that a substantial proportion of jurors misunderstand key instructions, which has prompted work on writing them in plain language and providing them in writing. Prejudice affects verdicts in ways that are measurable and difficult to remove. Improper research by jurors using the internet has become a serious problem and has caused trials to collapse. And the length and cost of jury trial limits its use, so most criminal cases in most systems are resolved without one, which is the largest practical fact about the institution.

Deliberations are private and researching them is prohibited in several systems, so the evidence comes from simulations and from other jurisdictions. The majority at the first vote usually prevails, a few jurors do most of the talking, and comprehension of legal directions is measurably poor. Pooling recollection reconstructs evidence better than any individual, which is the strongest argument for the arrangement.