Framing Biases in Plea Bargaining Decisions: Insights From the Psychology of Medical Decision Making

The vast majority of criminal cases in the United States are resolved by plea bargains, yet the philosophical underpinnings of “bargaining in the shadow of trial” are a strong version of rationality that appears to be at odds with research conducted by cognitive psychologists on framing. Reyna et al. (2025) provide empirical evidence that directly demonstrates that the way identical plea bargain options are framed has a significant influence on decision making. Of course, even the best of studies are not in and of themselves the final word on a topic of this importance. With a dearth of research on framing effects in plea bargaining, it seems useful to consider relevant research in related domains. The psychology of medical decision making, especially judgments and decisions made by patients in consultation with healthcare providers, is ripe for instructive comparisons. Like the decisions of criminal defendants, medical decisions are among the most important ordinary people can make. Such decisions are often made between unfavorable options (e.g., chemotherapy or invasive surgery), in unfamiliar and complex knowledge domains, and with the assistance of professionals who may not be fully trusted. Yet, they are also distinct from the well-worn lessons of market economics.

Given these structural parallels, it is worth asking whether the way options and decision-relevant information are framed have any measurable influence on judgments and decisions in medicine. The answer is decisively yes. For example, Armstrong et al. (2002) presented participants with graphs of curves representing how long people live after being treated for a disease over several years. These representations are useful, for example, in comparing different treatment options. Two mathematically identical ways of presenting the same data are survival curves, showing how many people are still alive after different time intervals, and mortality curves, showing how many people have died after those same intervals. Despite their mathematical equivalence, Armstrong et al. (2002) found that participants who were given survival curves (or both survival and mortality curves together) preferred preventive surgery significantly more often, and were significantly more accurate in answering knowledge questions, than participants who only received mortality curves. Donovan and Jalleh (2000) told participants about a hypothetical immunization that protected infants against respiratory problems. Side effects were framed either negatively (10% chance of side effects) or positively (90% chance of no side effects). They found that positive framing yielded better performance for participants without a young infant at home. Farrell et al. (2001) presented information to participants about the safety of donated blood in either a gain frame (lives saved), a loss frame (lives lost), or a combined loss frame presented in a positive context. They found that participants in the gain-frame condition were significantly more confident of blood safety for transfusions than in the loss-frame or combined conditions. These findings are far from unique. The Making Numbers Meaningful systematic literature review covers dozens of published studies demonstrating framing effects in judgment and decision making in medical contexts (see Ancker et al. 2025; Zikmund-Fisher and Ancker 2025 and other related articles in the same issue of Medical Decision Making Policy & Practice for a detailed review of the literature). My own lived experience as a survivor of Stage III renal cell carcinoma is quite consistent with these findings. An 80% chance of survival feels different than a 20% chance of dying—even for a reasonably numerate JDM researcher.

There are several take-aways from the literature on framing effects in medical judgment and decision making relevant to plea bargaining. First, they are pervasive, encompassing a number of dimensions including probability, risk, and the quality of outcomes. Thus, with respect to decisions about plea bargaining, in addition to the effects Reyna et al. (2025) found for time served, it is reasonable to predict comparable framing effects for other decision-relevant judgments including the likelihood of conviction, time pressure before a plea deal “goes off the table,” and the harshness of detention facilities. Second, framing effects often interact with other factors including participant characteristics. Reyna et al. (2025) discovered that criminal history interacts with framing due to the presumed higher propensity for risk taking of those who acknowledge previous criminal behavior. It would not be surprising if other individual differences such as tolerance for ambiguity and need for cognition also mediated or moderated framing effects in plea bargaining. Third, simply presenting two opposite frames together (e.g. lives saved and lives lost) has been surprisingly ineffective in medical contexts. It is therefore unlikely that a simple fix such as this will prove sufficient in future research and practice related to plea bargain decisions. Finally, theoretically based cognitive interventions have been found to reduce framing effects in medical decision making. These interventions have theoretical as well as practical implications for plea bargaining.

Some of the demonstrably effective interventions in medical decision making are based on the fuzzy-trace theory notion that people make judgments and decisions based on a mental representation of the bottom-line meaning—or gist—of decision relevant information (Reyna 2008; Reyna et al. 2022). For example, people make decisions differently when medical risk information is presented as absolute values compared to relative risk, and the use of stacked bar graphs to convey risk information improves judgments of medical risk (Brust-Renck et al. 2013). In my own research on perhaps the first web-based Intelligent Tutoring Systems to help women understand and make decisions about genetic cancer risk, icon arrays (sometimes called bathroom figures) along with other gist-based visual representations, gist-based tutorial dialogues, and gist explanations significantly improved risk assessment, knowledge, and comprehension (Wolfe et al. 2015; Wolfe et al. 2016). Another risk-related judgment is when two options are roughly the same or approximately equal in riskiness. This is important, for example, when two medical treatments are approximately equal in efficacy, but one has more serious side effects than the other. Here too, the way risks are framed influences judgment. However, a simple verbal intervention, “Think about the bottom-line gist. Is this a meaningful difference …” (Wolfe et al. 2018, 158), significantly increased appropriate approximately equal judgments, as did semantic context and source attribution. Given the evidence from the realm of medical decision making, comparable techniques involving asking defendants for gist explanations of plea-related decisions, simple gist-invoking wording, and even images to counteract quantitative framing effects are all promising approaches to helping defendants understand the gist of information related to plea bargains.

The US public seems deeply divided over whether too many people are suffering imprisonment in an age of mass incarceration (see Fleetwood 2020 for a deeply humanistic treatment) or whether the justice systems coddles criminals (“hugs for thugs”) who should be kept off the streets for as long as possible. Yet surely we can agree that it would be incredibly expensive if every criminal case that was currently resolved via plea bargaining went to trial. The American tradition of judicial fairness demands that people understand what they are agreeing to before waiving their right to a trial by a jury of their peers. Adherents to a belief in the strong form of human rationality sometimes argue that without this assumption, the very basis for the right to autonomous legal decision making disappears. However, decades of research in cognitive psychology, including multidisciplinary research on medical decision making, suggest that people who understand the bottom-line meaning of their circumstances can, and regularly do, make good decisions aligned with their own goals and values, about important and complex issues. This is the case with complex decisions such as whether to undergo invasive surgery, and it should be the case with decisions about plea bargaining too. Although there are lively debates in the field of judgment and decision making, I am persuaded that people are most likely to make judgments and decisions in their own best interest when they understand the bottom-line meaning (i.e., have a sound gist mental representation) of the decision-relevant information at hand. If I had my way, officers of the court, including public defenders and prosecutors, would come to understand that they have an obligation to ensure that defendants actually understand the gist of plea agreements, however framed. These mental representations would include a gist understanding of the prevalence of plea bargaining in American jurisprudence and the likelihood and nature of positive and negative outcomes in both chosen and relinquished decision options. The more judges, attorneys, and prosecutors understand the way people actually make life-altering decisions, the better positioned they will be to ensure that every time a defendant pleads guilty as part of a plea bargain, justice is served.