To Deny or Confess Guilt?: A Model of Confessions Describing Suspects’ Decision-Making

Yueran Yang, University of Nevada, Reno
Max Guyll, Iowa State University
Stephanie Madon, Iowa State University

This article presents a new model of confessions referred to as the interrogation decision-making model. This model provides a theoretical umbrella with which to understand and analyze suspects’ decisions to deny or confess guilt in the context of a custodial interrogation. The model draws upon expected utility theory to propose a mathematical account of the psychological mechanisms that not only underlie suspects’ decisions to deny or confess guilt at any specific point during an interrogation, but also how confession decisions can change over time. Findings from the extant literature pertaining to confessions are considered to demonstrate how the model offers a comprehensive and integrative framework for organizing a range of effects within a limited set of model parameters.

confessions, decision-making, expected utility, police interrogation

“In the criminal justice system, a confession is among the most persuasive forms of incriminating evidence. A confession is so persuasive, in fact, that in many cases convictions have been determined on the basis of confession evidence alone… Because of its incriminating power, a confession typically leads to legal sanctions. Because these sanctions can be severe, common sense would seem to suggest that even guilty suspects would deny guilt during a custodial interrogation. On the contrary, however, between 42% and 55% of all interrogated suspects confess, some of whom are innocent” (p. 80).

“Recognizing the advantage of developing broader theory, scholars have advanced a variety of theoretical perspectives that explain why suspects choose to confess when interrogated, including models derived from psychoanalytic, cognitive-behavioral, and decision-making perspectives… These theoretical models have advanced the field by providing explanatory accounts of suspects’ decisions to confess guilt. However, they are limited in the sense that they do not provide a comprehensive analysis of the multifaceted processes that collectively operate within a custodial interrogation… In this article, therefore, we introduce a model of confessions that we refer to as the interrogation decision-making model. Our goal in developing this model was to provide a comprehensive account of suspects’ decision-making during custodial interrogations-an account that is capable of integrating the diversity of ideas and empirical findings that have emerged within the confession literature over the past several decades” (p. 80-81).

“In making a decision, an individual chooses among the options in a decision space, which refers to the collection of all of the choices that one perceives as being available to them at a given point in time. For example, during an interrogation, suspects may perceive themselves as having any number of choices, such as denying guilt, confessing guilt, invoking their Miranda rights, refusing to speak, speaking about unrelated matters, and so forth” (p. 81).

“Deciding between a denial and a confession requires a suspect to consider and weigh the outcomes associated with both choices. To model this complex decision process we draw upon expected utility theory to account for the psychological mechanisms that underlie suspects’ decisions to deny or confess guilt during an interrogation, and specifically make use of the concept of expected utility from behavioral economics. Utility represents the experience of satisfaction, happiness, or “goodness.” Correspondingly, expected utility is the overall amount of utility one expects to experience as the result of a particular outcome. An outcome’s utility has two components: valence and magnitude. A desirably valenced outcome …has a positive utility… whereas an undesirably
valenced outcome … has a negative utility… The magnitude of an outcome’s utility reflects how strongly the outcome is viewed as desirable or undesirable” (p. 81).

“It is important to note that a single decision to deny or to confess is associated with multiple outcomes. Thus, in evaluating the utility of a particular choice, suspects evaluate the overall utility associated with all the outcomes they expect to flow from that decision. Ultimately, suspects choose to either deny or confess guilt based on which choice they expect to yield greater utility” (p. 81).

“[S]uspects evaluate the amount of satisfaction, happiness, or goodness they expect to experience as a result of each particular outcome… In addition to evaluating the utility of each outcome, expected utility theory holds that individuals will also assess the probability that an outcome will actually occur…” (p. 82). “It is well established that human beings temporally discount future outcomes.
That is, people more heavily weight proximal outcomes than distal outcomes because a given benefit is valued more highly the more quickly it is to be received, and a given cost is judged more dear the more quickly it is to be incurred. Empirical research suggests that temporal discounting could influence suspects’ confession decisions wherein they give disproportionate weight to the proximal outcomes experienced during an interrogation, without sufficient consideration of the distal outcomes that may be levied by the judicial system if they are convicted. Drawing on this body of work, the model accounts for suspects’ discounting of distal outcomes…” (p. 82).

“It is important to emphasize that suspects’ perceptions of possible outcomes are subjective rather than objective. Moreover, because the information that is available to suspects can be both insufficient and inaccurate, their subjective judgments may not be reliable. In other words, it is suspects’ subjective beliefs about the probability and utility of likely outcomes, which may or may not be accurate, that influence their decisions in combination with their idiosyncratic discounting of future outcomes. When individuals’ perceptions are inaccurate, decision-making can be flawed, ultimately leading to errors and ill-advised poor choices” (p. 82). This article proposes a mathematical equation incorporating these factors and considerations to predict whether a suspect will decide to confess or deny.

“Suspects’ decisions are not made in a vacuum, but rather in the course of dynamic circumstances that entail social influences, learning that occurs within the interrogation, and suspects’ personal characteristics and competencies. Accordingly, the interrogation decision-making model places the suspect’s expected utility evaluations that drive the immediate choice of making a denial or a confession within a larger context. Each decision-making cycle begins with a set of prevailing conditions that includes a broad array of factors that have the potential to influence suspects’ decision-making. Pertinent factors pertain to the crime under investigation, the characteristics of the interrogator and the interrogation setting, and the suspect’s own characteristics, the latter of which include traits, experiences, competencies, knowledge, beliefs, expectations, and scripts” (p. 84).

“Many factors may influence the subjective judgments that suspects make in assessing the expected utilities of a denial and a confession. These factors may be broadly classified according to whether they are primarily associated with characteristics of the crime under investigation, characteristics of the suspect, or characteristics of the interrogation. The interrogation decision-making model provides explanations for the mechanisms that operate in suspects’ interrogation decisions, and can account for the effects of these various factors on suspects’ choices to deny or confess guilt” (p. 84).

“The proposal that crime-relevant facts should affect suspects’ decision-making is straightforward. Naturally, confessions to more serious crimes are discouraged because conviction entails more serious punishment, which is represented by a more negative utility of the distal outcomes associated with confession. As an example, suspects accused of murder will be less likely to confess because they anticipate lengthy incarceration or execution, whereas those accused of larceny will be more likely to confess because they anticipate only probation. An additional crime factor that can influence decision-making is the state of the evidence. Guilty suspects who believe that objective evidence exists indicating their guilt will view conviction as highly probable for both a denial and a confession. By contrast, guilty suspects who believe that no other evidence exists that will indicate their guilt should judge the likelihood of conviction to be much lower if they choose to deny than to confess” (p. 85).

“Suspects’ decision-making processes can also be influenced by a variety of personal characteristics, such as personality traits, intelligence level, preexisting beliefs and expectations, knowledge of the legal system, factual innocence or guilt, and so forth. In particular, the rate of confessions is higher among juveniles than adults, higher among those with cognitive disabilities than those without, higher among first-time offenders than recidivists, and higher among the guilty than the innocent. The interrogation decision-making model is useful for examining how such individual differences can affect suspects’ subjective judgments pertaining to the model parameters and thereby influence decision-making” (see article for details; p. 86).

“A third category of factors that can influence suspects’ subjective judgments relates to characteristics of the interrogation itself, including the physical and social environment, interpersonal influence tactics applied, and the interrogator. Interrogation techniques are particularly relevant because they can alter suspects’ decision-making by manipulating their perceptions of the probabilities and utilities of the outcomes associated with denials and confessions. By applying the model, it is possible to analyze how specific interrogation tactics can impact the parameters that affect expected utility assessments, thereby understanding how they influence suspects’ decisions to deny or to confess guilt. The influence of interrogation tactics further implies the importance of the interrogator. Indeed, it is the interrogator who chooses and implements the tactics, and whose characteristics may also affect the suspect’s subjective evaluations” (p. 87).

“The discussion to this point has primarily focused on how suspects choose to deny or confess guilt at a specific point in time during an interrogation. However, suspects may not make just a single decision during an interrogation. Instead, one can imagine a suspect making a series of decisions-at first denials, and, at last, perhaps, a confession. The interrogation decision-making model accounts for the dynamic nature of suspects’ interrogation decisions by incorporating an iterative process that can yield a chain of successive decisions. The choices and consequences of one decision-making process can shape the conditions that prevail at the start of the next cycle. These factors can then influence the decision-making parameters, including the outcomes under consideration, their corresponding probabilities and utilities, as well as the discount factor … The explicit inclusion of time in the interrogation decision-making model is critical for understanding how suspects may change their decisions within an interrogation, such as by the accumulation of interrogation effects over time” (p. 88).

“Also supporting the usefulness of expected utility in determining confession decisions is that the model explains a wide range of empirical findings regarding how the crime under investigation, suspect characteristics, situational features, interrogation techniques and interrogator behaviors can influence confession decisions via their effects on the parameters of the expected utility function. Furthermore, the model also incorporates the dynamic character of interrogations by the inclusion of recursive effects that account for temporal variability in suspects’ confession decisions stemming from a variety of causes, such as fatigue, cumulative effects of interrogation techniques, or learning from choice-consequence pairings experienced during the interrogation” (p. 90).

As always, please join the discussion below if you have thoughts or comments to add!