Abstract

Excerpted From: Hillel J. Bavli, Character Evidence as a Conduit for Implicit Bias, 56 U.C. Davis Law Review 1019 (February, 2023) (211 Footnotes) (Full Document)

It is fitting that the word character has its origin in the Greek word kharassein, meaning to engrave or stamp: the image of one's character is as much a creation of the observer as a creation of the observed. In spite of this, courts routinely invite jurors to make judgments based on their perceptions regarding an individual's character. Courts thereby sanction judgments that are based on the subjective prior beliefs and prejudices of jurors rather than on the evidence in a case. Consequently, courts invite jurors to rely on their priors--including their implicit biases-- regarding an individual's race, sex, appearance, accent, education, economic status, and other background characteristics when determining a verdict.

The admissibility of evidence regarding a criminal defendant's prior bad acts has been described as “the single most important issue in contemporary criminal evidence law.” Indeed, in a large proportion of cases-- including cases like the recent Chauvin trial, Rittenhouse trial, and Cosby trial--the court must decide whether to admit evidence regarding the prior acts of the defendant or a victim, and “[t]he stakes for the prosecution and the defense are enormous.” This decision is often critical to--if not altogether determinative of--a case's outcome. 404 of the Federal Rules of Evidence (“FRE”) prohibits evidence of an individual's character traits or past acts to prove that the individual acted in accordance with a certain character or propensity on the occasion in question. It reflects the principle that individuals should be judged legally based on what they did rather than who they are. It is not grounded in an absence of probative value. Rather, the ban on character evidence is based on the unfair prejudice that it causes. Under Rule 403, courts “exclude relevant evidence if its probative value is substantially outweighed by a danger of ... unfair prejudice” and other risks, such as causing confusion and wasting time. But Rule 404 replaces Rule 403's ordinary balancing analysis for character evidence: it “reflects the judgment of Congress that as a matter of law the probative value of propensity evidence is substantially outweighed by the risk it poses of unfair prejudice, juror confusion, and waste of time.” notwithstanding the rule against character evidence, courts routinely admit other-acts character evidence through an expanding set of exceptions and ad hoc departures. Some exceptions are legislative. For example, under Rule 413 of the FRE, in a criminal sexual-assault case, “the court may admit evidence that the defendant committed any other sexual assault” to prove that the defendant has a propensity to commit such acts and is therefore likely to have committed the crime in question. Other departures from the rule against character evidence have developed by common law, often through misinterpretations of Rule 404(b)(2).

Rule 404(b)(2) clarifies that other-acts evidence (as distinct from other-acts character evidence) may be admissible if it is offered for a non-character purpose--that is, for a purpose that does not require propensity reasoning, or inferences based on an individual's character or propensity--such as to prove motive, intent, or knowledge. This is important because other-acts evidence may be highly probative for non-character purposes without giving rise to the concerns associated with other-acts character evidence. However, courts also regularly misapply Rule 404(b)(2) by permitting other-acts evidence that, although relevant to a Rule 404(b)(2) purpose such as intent or motive, indeed requires propensity reasoning. For example, courts regularly admit evidence that a defendant previously engaged in drug trafficking to prove that the defendant has a propensity to commit such acts and is therefore likely to have had the requisite knowledge and intent to distribute drugs in the incident in question. Indeed, some of these misapplications are so deeply ingrained in the common law that courts do not even acknowledge the rule when departing from it. summary, the rule against character evidence is not a firm rule. It is riddled with exceptions, and the exclusion of even basic forms of character evidence is at best unpredictable. Moreover, these exceptions are only expanding, and courts and legislatures have grown increasingly permissive of character evidence. in creating exceptions to the rule against character evidence, lawmakers frequently rely on a misleading account of the harms associated with this evidence. Specifically, in the usual narrative, repeated regularly by courts and scholars, the rule against character evidence is based on two forms of unfair prejudice: (1) a jury is likely to give excessive weight to other-acts character evidence, and (2) it may punish a defendant based on the defendant's prior bad acts rather than determine a verdict based on the act in question in the case. Courts and legislatures rely on this narrative when they create exceptions for circumstances in which character evidence seems highly probative relative to the unfair prejudice described by this narrative, or in which it seems possible to mitigate such unfair prejudice through counteracting measures such as providing juries with limiting instructions. this narrative arguably misses a highly troublesome feature of character evidence and far understates its pernicious effects. Moreover, these effects are intrinsic to character-propensity reasoning and are unlikely to be mitigated through counteracting measures. Relatedly, this narrative--firmly entrenched in caselaw and scholarship--has created a substantial blind spot for courts and scholars: while certain harmful effects of character evidence and Rule 404 erosion have been examined extensively, the effect on inequality in the U.S. legal system has received relatively little attention. in this Article, I develop a new model of character evidence that refocuses the question of character evidence on the distortions associated with the prior beliefs and prejudices inherent in a juror's perception of character evidence. I draw on disciplines outside of law--including Bayesian statistics and cognitive psychology--to explain how jurors use character evidence to arrive at a verdict. I apply this framework to show that when a court admits character evidence through exceptions, it invites jurors to rely on their prior beliefs and prejudices when determining a verdict, and that, consequently, judgments based on character evidence are inherently biased against certain groups of people based on their race, sex, appearance, accent, education, economic status, and other background characteristics.

To introduce what I believe to be a more accurate account of character evidence, I employ a set of basic principles from an area of statistics known as Bayesian inference. This type of reasoning sounds complex (and it can be), but its central principles are intuitive and provide a natural framework for understanding character evidence. At its center, Bayesian inference provides a formula--known as Bayes rule--for combining prior beliefs (a “prior”) with new evidence to arrive at a new belief or conclusion (a “posterior”) that accounts for the new evidence. I employ this rule to explain how jurors use character evidence to arrive at a verdict.

Specifically, I consolidate principles of Bayesian inference into a single intuitive rule--which I refer to as the “shrinkage principle”--that tells us how much weight to afford a prior belief relative to new evidence when such evidence becomes available. In short, the principle states that when combining prior beliefs with new evidence, each should be weighted in proportion to its relative precision (the inverse of uncertainty) with respect to the matter in question. In particular, a prior belief will be most valuable and will be relied on most heavily when it provides precise information regarding a matter for which there is poor or imprecise new evidence. Conversely, a prior belief will be less valuable and will be relied on less heavily when either it provides imprecise information regarding the matter that it informs, or there is precise new evidence regarding that matter. apply this Bayesian framework to describe a process whereby a juror incorporates character evidence into a verdict in two stages, each of which can be understood as an application of the shrinkage principle to combine prior beliefs with new evidence. First, the juror combines her implicit biases regarding the defendant's background characteristics with court-sanctioned character evidence to formulate an assessment of the defendant's character. Second, the juror combines her character assessment with other (non-character) evidence in the case to arrive at a verdict. argue that, based on this model--which I call the “reflective model” of character evidence--character evidence in a sense activates a juror's implicit biases regarding a defendant's race, sex, appearance, and other background characteristics. Specifically, I apply the shrinkage principle to argue that a juror's implicit biases--her priors regarding the defendant--remain relatively subdued in their effect when the juror is asked to make non-character judgments, such as whether a defendant committed the act in question; on the other hand, when a court admits character evidence, and a juror is asked (implicitly or explicitly) to engage in character reasoning, this activates the full force and effect of the juror's priors. Indeed, I argue that character evidence inherently requires a juror to rely on her prior beliefs and prejudices. Meanwhile, from the start of the trial, the juror has been developing an extensive set of prior beliefs and prejudices based on the defendant's background characteristics and regarding precisely what the juror is now asked to assess: the defendant's character.

Importantly, although the model proposed herein is grounded in a theory of information optimization based on Bayesian principles, it finds substantial support in the psychology of juror decision-making. Specifically, I draw on scholarship surrounding the concepts of “stereotype activation,” “confirmation bias,” and other concepts in cognitive psychology to demonstrate that the behavioral assumptions on which the reflective model relies are well grounded in empirical research regarding how people--and jurors in particular--process information. I argue that the model proposed herein has critical implications for policy surrounding the admissibility of character evidence--policy that is currently premised on a narrative that does not adequately account for the pernicious nature and effects of character evidence. In particular, I examine implications for exceptions to the rule against character evidence and for the courts' increasingly permissive approach to character evidence. I also highlight the importance of developing sound empirical research regarding the interaction between character evidence and implicit bias, and I describe a basic approach to empirically test the proposed model.

Ultimately, I argue that, based on my analysis herein, a stricter rule against character evidence is necessary. In particular, courts should read Rule 404(b) as a rule of exclusion, interpreting Rule 404(b)(2) as providing for permissible non-propensity uses of other-acts evidence rather than providing for exceptions to the ban on character evidence. Moreover, although there may be extremely important concerns underlying some formal exceptions to the rule against character evidence--such as the exception for other-acts evidence in sexual-assault cases--these exceptions should be evaluated in light of their inherent relationship with jurors' prior beliefs and prejudices regarding a defendant's background characteristics, and the critical aims of fairness and equality should be protected along with those underlying such exceptions. proceed as follows: In Part I, I discuss the rule against character evidence, its underlying policy concerns according to the traditional account, and its many exceptions. In Part II, I develop the reflective model of character evidence. In particular, I apply basic concepts in Bayesian inference to describe how jurors use character evidence to arrive at a verdict, and to explain the shrinkage principle in simple and intuitive terms. I then apply this principle to demonstrate how character evidence may activate a juror's prior beliefs and prejudices and invite jurors to rely on their implicit biases regarding a defendant in determining a verdict. In Part III, I discuss implications of the reflective model for evidence law--including implications for exceptions to the rule against character evidence and for the trend toward a more permissive approach to character evidence. Finally, I discuss implications for empirically examining the interaction between character evidence and implicit bias with respect to case outcomes. In Part IV, I conclude.

[. . .]

Notwithstanding the rule against character evidence, courts regularly admit character evidence through an expanding set of exceptions. These exceptions arise for various reasons, including evidentiary gaps, circumstances in which character evidence seems particularly probative, and non-evidentiary policies. In creating exceptions, legislatures and courts implicitly or explicitly rely on a common narrative regarding the risks associated with the admission of character evidence. In particular, they create exceptions for types of character evidence that seem particularly probative relative to the unfair prejudice described by this narrative, or for circumstances in which courts seem to be able to mitigate such unfair prejudice through limiting instructions and other counteracting measures.

However, this narrative fails to account for a highly pernicious effect of character evidence, and this effect is intrinsic to character-propensity reasoning and is unlikely to be mitigated through counteracting measures. According to the reflective model, regardless of the concerns that have led to an exception--some of which are of critical importance--character evidence inherently involves inferences that are substantially based on a juror's prior beliefs and prejudices regarding the background characteristics of a defendant. Consequently, verdicts that are based on character evidence are inherently biased against certain groups of people, and the admissibility of character evidence in many contexts may therefore be a significant driver of inequality and unfairness in the U.S. legal system. If this model is correct, it provides a strong reason to limit legislative and judicial exceptions to the rule against character evidence and to curb the current trend toward a more permissive rule.

Hillel J. Bavli. Assistant Professor of Law, SMU Dedman School of Law. For helpful comments and exchange