Guilt Status Influences Plea Outcomes Beyond The Shadow-Of-The-Trial In An Interactive Simulation Of Legal Procedures

Objective: More than 95% of criminal convictions in the United States arc secured by guilty pleas. Our current understanding of the "deals" that lead so many to plead guilty is often tied to the shadow-of-the-trial (SoT) model, which posits that plea outcomes rely solely on the penalty discrepancy they offer (represented as: [trial conviction probability X trial sentence] - plea sentence). This study compared the power of the SoT model to predict plea outcomes with two expanded models. Hypotheses: We hypothesized that the SoT model's power to predict whether defendants will accept a plea offer could be improved by expanding the model. The first expanded model added a main effect of guilt status. presuming that regardless of penalty discrepancy. the guilty will be more likely to accept plea oilers than the innocent. The second expanded model added an interactive effect of guilt status with penalty discrepancy: although greater discrepancies would increase both true and false guilty pleas, the magnitude of the effect would he greater among the innocent. Method: We recruited student (N= 596, 45.8% female. M= 19.9 years old) and non-student (N = 525, 45.1% female, M = 30.9 years old) adult samples to participate in a 2 (guilt status: innocent or guilty) X 3 (conviction probability: 20%. 50%, or 80%) X 3 (plea discount: 6. 12. or 18 months) mixed design. All participants experienced two crime scenarios in a counterbalanced order. We randomized the manipulated variables for each scenario such that each participant was administered two of eighteen potential conditions. Results: As hypothesized. the predictive power of the SoT model was significantly improved by expanding it to include guilt status-guilty participant-defendants were consistently more likely to accept the plea offer than innocent participant-defendants. However, an interactive effect of guilt status with penalty discrepancy did not significantly improve the power of the SoT model to predict plea outcomes. Conclusions: The power of the Sol' model could be significantly improved by incorporating guilt status as a predictor. Although there are many times at which guilt status is unknown, the acknowledgment of a separate effect of guilt status has important policy implications far the plea process.