Cognition and Incentives in Plea-Decisions: Categorical Differences in Outcomes as the Tipping Point for Innocent Defendants

Though factual guilt or innocence, variations in sentence at plea and trial, probability of conviction at trial, and individual risk preferences all influence one’s decision to plead guilty, incentives affect guilty and innocent defendants differently. 3,375 British adult users of the Prolific platform participated in a three-to-four-minute vignette study in which they had been accused of a theft and were required to choose between going to trial and accepting a guilty plea. Participants were randomly assigned guilt conditions, probabilities of conviction, and one of four sentence discounts. A subset of participants also answered questions about the considerations that were most important to them when pleading. Discounts resulting in categorically- or qualitatively-different sentence types (e.g., probation instead of custody) prompted both guilty and innocent participants to plead guilty. In particular, the categorical difference between plea and trial made factual innocence a less-important consideration. In contrast, discounts resulting in shorter sentences for the same sentence type (e.g., halving sentence length) prompted only guilty participants to plead guilty even when the incentive was relatively modest. These findings may help plea researchers better understand the relationship between guilt and probability of conviction at trial.

“. . . [W]hile purely quantitative incentives to plead are likely to appeal only to guilty defendants, more categorical, qualitative incentives to plead are likely to appeal to both guilty and innocent defendants. Participants in the innocent condition in this study showed significant resistance to pleading guilty when the more quantitative incentive was offered, even at high levels of probability of conviction. However, when a categorical incentive was offered . . . , they showed a similar trend to guilty defendants[.]” p. 352