Charged up and anchored down: A test of two pathways to judgmental and decisional anchoring biases in plea negotiations.

Abstract

Prosecutors sometimes threaten defendants with disproportionately severe potential trial sentences (PTS) to coerce their plea acceptance. Three experiments tested whether the PTS act as an anchor that biases evaluations of the plea deal at deep (vs. shallow) processing level by encouraging asymmetric consideration of information consistent—but not inconsistent—with the appropriateness of the PTS. I expected that a high (vs. moderate) PTS would lead to greater plea acceptance by increasing maximum acceptable sentencing preferences and influencing evidence appraisals. I tested two anchor debiasing strategies: selective consideration of anchor-inconsistent evidence (consider-the-opposite strategy [COS]) and holistic consideration of all evidence (via process accountability). Participants read detailed information about their criminal case and the PTS (high or moderate) before indicating their maximum sentencing preferences, evaluating case attributes, and rendering a plea decision. Mock defendants (Experiment 1; N = 479; drelevant = 0.83), defense attorneys (Experiment 2; N = 155; dpre-COS = 1.19), and mock attorney–client pairs (Experiment 3; Npairs = 265; dunaccountable = 0.60) were willing to accept (or recommend) longer sentences when presented with a higher (vs. moderate) PTS—even when the PTS was irrelevant to their case (Experiment 1; dirrelevant = 0.64). Mediation analyses across all studies demonstrated that higher maximum sentencing preferences were, in turn, associated with greater willingness to accept a plea deal, but not with biased evaluations of evidence strength. Selective consideration of evidence inconsistent with the high PTS attenuated anchoring effects on sentencing preferences (Experiment 2; dpost-COS = 0.75) and plea recommendations but discussing more evidence did not (Experiment 3; daccountable = 0.68). This work challenges the law’s assumption that defense attorneys can debias defendants’ cognitive errors and preliminarily proposes easy-to-implement debiasing solutions to overcharging.

The “Partial Innocence” Effect: False Guilty Pleas to Partially Unethical Behaviors

“…Although most past research has examined defendant plea decisions using exclusively hypothetical vignettes or field study self-reports (c.f. Cardenas, 2023;Dervan & Edkins, 2013;Hendersen & Levett, 2018;Perillo et al 2014;Pardieck et al 2020;Wilford & Wells, 2018;, the current work used a triangulation approach that combined a highstakes real-world decision-making laboratory paradigm that afforded greater experimental control and then replicated and extended findings in two immersive and interactive role-playing studies. Where results differed was in the more extreme responses of participants in the roleplaying compared to real-world experimental paradigms, who reported a lower likelihood of conviction, feeling less guilt, and ultimately pled guilty less often.…”

“…In the real world, increasing the severity of charges is typically associated with increased punishment which may also deter responsibility-taking. Conversely, other work finds that overcharging by threatening guilty mock defendants and their mock attorneys with excessively harsh (compared to more appropriate) sentences increases plea acceptance (Cardenas, 2023;c.f. Schneider & Zottoli, 2019).…”