Historical Evolution of Jury Decision-Making Research
Modern empirical research on jury decision-making generally traces to the Chicago Jury Project of the 1950s, one of the earliest systematic efforts to study juries in operation using social scientific methods (Devine et al., 2001, p. 622). Funded in part by the Ford Foundation, the project marked a decisive shift away from purely doctrinal, anecdotal, or impressionistic accounts of jury behavior toward empirical investigation grounded in observation, interviews, and quantitative analysis. Although controversial at the time, the Chicago Jury Project established a methodological foundation that continues to shape contemporary jury research by demonstrating the feasibility and value of studying juror decision-making as a social and psychological process rather than solely as a legal abstraction.
Beginning in the mid-1950s, Harry Kalven and Hans Zeisel, working under the auspices of the Chicago Jury Project, collected data on jury decision-making using surveys, interviews, and mock-jury methodologies. As the project expanded, researchers sought access to actual jury deliberations, including the use of audio recordings (Devine et al., 2001, p. 623). These developments generated substantial public and governmental backlash and ultimately led federal authorities and many states to prohibit access to research on jury deliberations (p. 623). Nevertheless, the Chicago Jury Project itself produced a substantial body of empirical work and stimulated a sustained stream of scholarly publication. In 1966, The American Jury reported the project’s findings by comparing actual jury verdicts with those favored by presiding trial judges. Drawing on data from approximately 3,500 civil and criminal trials, the study represented an unprecedented empirical examination of jury behavior.
Among the most influential findings from The American Jury was the identification of what Kalven and Zeisel termed the liberation hypothesis (Devine, 2009, pp. 136-148). When trial evidence clearly favored one side, judge and jury verdicts converged at high rates, suggesting that both decision-makers responded similarly to unambiguous evidence. However, when evidence was closely balanced or subject to reasonable dispute, juries diverged from judicial assessments far more frequently, and this divergence consistently favored the defendant (Kalven & Zeisel, 1966). Kalven and Zeisel proposed that evidentiary ambiguity liberates jurors to rely on values, sentiments, and extralegal considerations that would otherwise remain subordinate to the factual record. Under conditions of uncertainty, jurors are freed to express community standards, sympathies, or moral intuitions that judges, bound more tightly by legal doctrine, would suppress or override.
The liberation hypothesis reframes jury leniency as a value-laden response to evidentiary uncertainty. In cases where the facts clearly established guilt, juries convicted at rates comparable to jurists. In close cases, however, juries favored acquittal, producing a disagreement rate of approximately 19 percent: juries acquitted defendants whom judges would have convicted (Kalven & Zeisel, 1966). Subsequent empirical research has confirmed this pattern. Devine and colleagues (2008) analyzed data from 179 criminal jury trials and found that extralegal variables were significantly related to jury verdicts primarily when evidence strength was moderate rather than clearly strong or weak, suggesting that when evidence permits multiple reasonable interpretations, jurors apply a functional presumption of innocence that exceeds what legal doctrine formally requires, resolving doubt in favor of the accused even when judges might find sufficient basis for conviction. The liberation hypothesis identifies evidentiary strength as a mediating variable that influences whether jury verdicts reflect narrow legal judgment or broader community sentiment.
The durability of these findings was confirmed nearly four decades later. Eisenberg and colleagues (2005), working with the National Center for State Courts, conducted a systematic replication of Kalven and Zeisel’s judge-jury agreement methodology using contemporary criminal case data from multiple jurisdictions. Their results demonstrated essentially the same rate of judge-jury agreement, approximately 78 percent (Eisenberg et al., 2005). When judges and juries disagreed, the pattern of disagreement replicated the original findings: judges exhibited lower conviction thresholds than juries and tended to convict more frequently in cases of moderate evidentiary strength. The study found little evidence that evidentiary complexity or legal complexity explained rates of disagreement; rather, the data supported Kalven and Zeisel’s original explanation that the differential reflects judges’ systematically lower bar for finding guilt beyond a reasonable doubt (Eisenberg et al., 2005).
While the Chicago Jury Project yielded important insight into jury decision-making, its methodological constraints also revealed the importance of examining the stages of the jury process that remain accessible to empirical inquiry. Limitations in observing deliberations have directed scholarly attention to earlier phases of juror involvement, where participation can be measured, observed, and systematically analyzed.