Part Two: Implicit Bias in Criminal Cases

A. Prosecuting Attorneys and Prosecutorial Bias

Implicit bias can operate at many stages when we consider the choices prosecutors face during the various stages of a criminal case. Prosecutors have discretion when deciding whether to charge a person for a crime at all. Studies show prosecutors are more likely to charge Black suspects than white suspects in similar circumstances. For instance, on the issue of justified homicide and self-defense, the cell phone/weapon IAT results suggest that “prosecutors might be more likely to believe that the white victim was reaching for his cell phone, and thus, that the suspect acted unreasonably in shooting the deceased;” but when the victim is Black they are more likely to find that the suspect acted reasonably in discharging the weapon. The bias operates doubly here, because the white victim is more likely to be perceived as reaching for a cell phone whereas the Black victim is more likely to be perceived as reaching for a weapon, and the Black suspect is more likely to be perceived “as reacting unreasonably in discharging his weapon” against the white victim, whereas the white suspect is more likely to be perceived as “being in reasonable fear” of the Black victim, and therefore acting appropriately in discharging his weapon.

The prosecutor has discretion as to what level of crime to charge to each individual. The Black or Latino male “drug dealer” stereotype can impact whether a prosecutor files a simple possession charge or “with intent to distribute” charge. Similarly, the decision to charge juveniles in adult court can be impacted by the race of the individual.

Determining the level (or even availability) of bail is impacted by presumed ties to the community, which often are based on race, ethnicity, and socioeconomic status. Justice Hyman notes that “these discretionary decisions are closely tied to the prosecutor's evaluation of the suspect's behavior and whether a suspect seems likely to be a future danger to society.” In short, “the perception of the defendant ... in turn, alters the perception of the seriousness of the crime.”

But when is the prosecution evaluating the suspect's behavior? If before arrest, such as in the decision whether to seek a warrant, the prosecutor is generally seeing things unfold through the lens of the police or investigator, who is interpreting and reporting the suspect's actions, influenced by his or her own biases. When the prosecutor interprets the officer's report, that interpretation is influenced by the prosecutor's own biases. Thus, two levels of bias could creep in before the prosecutor even makes a charging decision.

If the prosecutor's evaluation of the defendant's behavior is post-arrest, common and expected reactions to being in jail--where one must not show fear or remorse, act “hard” and most of all be silent and self-protective--may leave an unfavorable impression with the prosecutor. The same person likely behaves very differently out in society on bail than when in prison or jail.

On the issue of plea-bargaining, the power of “in-group favoritism” impacts how empathetic a person will be when seeing or hearing about another person experiencing pain. For instance, studies have shown people empathize more with a lighter-skinned person being subjected to pain than with a darker-skinned person being subjected to pain, particularly when the people being measured have lighter skin. Empathy can lead to lower sentences; and the evidence shows that white defendants also receive more favorable plea bargains than Black defendants.

In jury selection, peremptory challenges allow implicit racial bias to seep in when Black jurors are disproportionately stricken. Implicit bias seems to play a large role during the voir dire process with preemptory strikes often reflecting an attorney's own unconscious stereotypes that govern which jurors to strike without reason. For instance, the prosecutors may strike African American jurors who live in the inner city--applying a stereotype about Black Lives Matter supporters when any particular person in that group could be “tough on crime” based on its devastating impact on the community. The remaining jurors are likely to be those who the attorney believes fit a favorable stereotype and they too could actually hold counter-stereotypic views.

Thus, prosecutors will attempt to keep those perceived to be pro-law enforcement jurors (just as the defense may strike otherwise pro-defense jurors or keep pro-prosecution jurors), all because they assess the jurors based on what could be erroneous or inapplicable stereotypes. Even though the defense can offer a nonracial reason for the strike, that reason may be covering for implicit bias as an after-the-fact, dissonance-reducing justification.

This potential for prejudice can continue into the conduct of the trial. The prosecutor hears and evaluates the testimony of the witnesses, and makes judgments about the credibility of those witnesses, which influence the substance of the prosecution's closing arguments. Black defendants are more often dehumanized by the words and phrasing used in the prosecution's closing arguments, such as referring to Black defendants as “animals.” Professor Frank cautions against references that implicate racial stereotypes. Courts “seldom find that such references deny nonwhite defendants a fair trial,” and therefore:

[U]ncharged acts admitted against non-White [sic] defendants likely assist jurors in filling any evidentiary gaps in the prosecution's case with implicit race associations. This phenomenon is consistent with studies demonstrating the jurors are more likely to convict racial minorities than non-White [sic] defendants when the evidence is ambiguous, versus weak or strong.

Prosecutors are using language that is persuasive to the jurors, and defense attorneys must employ counter-stereotype techniques in their efforts to vigorously represent the clients because judges are not likely to step in. Nevertheless prosecutors, as representatives of the state, must first pursue justice rather than any individual's interest, and this asymmetry suggests a more compelling case for curbing implicit bias in prosecutors over defense lawyers.

B. Criminal Defense Attorney Bias

Criminal defense attorneys also demonstrate implicit bias. Defense attorneys may be affected when primed with words or conduct associated with gangs or violence in their everyday practice. They may also be more likely to believe that ambiguous evidence relates to guilt or that the defendant is actually guilty in conformance with the stereotype. Defense attorneys may also attribute a higher level of violence to conduct allegedly committed by Black defendants over white defendants. Defense lawyers also rely on stereotypes when they pick juries, making assumptions based on their race, and Lyon cautions:

[I]f we are using this process of elimination based on stereotypes, jurors will know it. And then we cannot get angry if the jurors return the favor by making the assumption that our young male minority client is guilty, a gang member, or otherwise dangerous and not deserving of respect.

For public defenders, implicit bias may particularly affect decision making, given the triage circumstances of a need for quick action with few resources. Triage begins immediately, when public defenders prioritize cases based on whether or not the state can prove its case beyond a reasonable doubt or whether the public defender believes the client is “factually innocent.”

In a triage situation, public defenders allocate their resources first to the cases they can or should win. If public defenders fall under the spell of generalizations or stereotypes, they may miscalculate the merits of the case and the odds of winning it and thus decide to allocate fewer resources to what they perceived to be losing cases. For instance, public defenders subscribing to a stereotype that those who do not look their questioner in the eye are lying, may discount a witness's potential testimony or choose not to put that witness on the stand. Without that witness, who while truthful, appears to the defense attorney to be untruthful, the case against the defendant may be stronger, resulting in a disservice to the client.

The prophecy becomes self-fulfilling, reinforcing the perception for similar clients or circumstances in the future. These implicit biases “can cause attorneys to treat stereotyped individuals in stereotype-consistent ways” and the attorneys' “unconscious negative expectations may produce perceptions and attributions consistent with them.” The client may sense this negative reaction and respond in a way that causes the attorney to confirm her opinion and “can create a vicious cycle of mutual distrust and dislike, adversely affecting the attorney's triage decisions. Spending time with a client can, of course, change these initial impressions. But, an unpleasant initial interaction may reduce the defender's desire to do so.”

One might think public defenders would be “better” on race issues, but as one former public defender learned, “there is no person without prejudices, myself included.” For instance, when Professor Andrea Lyon had a Black defense client who became hostile and raised his voice, she succumbed to the “angry Black man” stereotype and decided to walk away from the client. Only later did she realize that the client might have been suffering from developmental disabilities, acting “tough” to cover-up the fact he could not follow her conversation and did not understand the multiple options she was presenting to him. She attributes her decision in the moment in part to the concept of “race loyalty,” which occurs when people are accustomed to “assuming the best in people who match our race because of our desire to see our race (and ourselves) positively.” The converse is also true, assuming the worst in people whose race is different from one's own race, which can impact how one represents criminal defendants.

Implicit bias may be even more of a concern in death penalty defenses, though little has been recorded about the attitudes of death penalty attorneys specifically. Taking the data sets of presentations made at training sessions for capital defense attorneys, researchers divided them into three groups: habeas corpus, trial lawyers, and law students. They found statistically significant differences in the results of the white-with-good-and-Black-with-bad IAT over the Black-with-good-and-white-with-bad for all three groups.

Interestingly, subjects did better when associating their own race with “good.” Having more Black capital defense attorneys would likely lead to more comfortable associations of Black-with-good. Based on this study, diversity in this attorney pool can have an impact on the volume and level of bias made manifest. Given that forty-two percent of individuals on death row are Black, some states have taken the initiative to enact laws that give defendants a claim of racial discrimination in their post-conviction appeals. For example, North Carolina enacted the Racial Justice Act in 2009, allowing defendants to make a claim of racial discrimination in their post-conviction appeals. Professor Shelley Song suggests additional reforms in the criminal context. While her article specifically addresses death penalty cases, her suggestions are broadly applicable to criminal cases generally and several also apply to civil cases.

C. Effects on Jurors

We know skin color may impact juror decision making. With otherwise identical scenarios, the darker the skin of the alleged perpetrator, the more likely jurors in mock situations are to find the alleged perpetrator guilty. The “violent Black man” stereotype discussed above can also operate as character evidence as Professor Mikah Thompson notes, referring to the nature of implicit bias as “unspoken evidence” often used at trial against African Americans. Various concepts of implicit bias come together to paint a picture of “stereotypical Blackness” that is often used against criminal defendants because “rather than offering inadmissible evidence of a Black defendant's character for violence, the government can instead offer evidence of the defendant's stereotypical Blackness, thereby playing upon the jurors' implicit biases to establish the guilt of the defendant.”

The George Zimmerman trial (for the killing of Black teenager Trayvon Martin in Florida) provides an illustration of the idea of “stereotypical Blackness.” The racialized criticism of prosecution witness Rachel Jeantel (as “dumb,” uneducated and not credible) provides insight into the way jurors and Zimmerman viewed Jeantel and also how they possibly viewed the victim. If people watching the televised trial thought Trayvon had a similar character to Jeantel because of their friendship and common color, then her “stereotypical Blackness” could have been passed on to him throughout the trial given that he was deceased and therefore not present. If no evidence was offered to “humanize” Trayvon Martin, all that remained was stereotypical Blackness, which could have acted as his character evidence. As such, the jurors likely would perceive him as threatening or violent and would perceive Zimmerman as being more reasonable in fearing for his life and using deadly force in self-defense. Hence, the acquittal.

In addition to being influenced by implicit bias, Professor Thompson suggests that white decision-makers may be affected by the “transparency theory.” The transparency theory is “the tendency of whites not to think about whiteness.” In short, because whiteness is the norm, being a white person is not something that many white people tend to think about. They do not actually reflect on their whiteness unless they are in a situation that calls them to compare themselves to someone who is not a white person. This transparency phenomenon is dangerous for African Americans because it causes white people to disregard the existence and salience of white-specific norms for those who are not white people.

Professor Thompson examines Federal Rule of Evidence 404(a), which limits when character evidence (other than for credibility purposes) may be used against a defendant in a criminal trial. Although the rule is meant to protect defendants from having unduly prejudicial evidence admitted, she suggests implicit bias combined with the transparency theory creates a different kind of character evidence not regulated by Rule 404(a): evidence of “stereotypical Blackness,” which is just as prejudicial. Therefore, when African Americans do not assimilate to these norms or seems unwilling to assimilate, they will face discrimination from jurors, even if those jurors have good intentions. Further, a Black defendant's perceived failure to assimilate to white-specific norms can cause that defendant to fall into the category of “stereotypical Blackness,” which implicitly includes evidence “that he or she has a propensity for engaging in certain behavior.” That behavior, based on the IAT studies discussed above, is aggressive, violent and often criminal.

D. How Judges' Implicit Bias May Affect Rulings

A few studies identify implicit biases in judges. Judges are “just as susceptible as are jurors to three cognitive illusions that hinder accurate decision making: anchoring, hindsight bias and egocentric bias.” An IAT study showed:

[T]he white judges mostly showed a white preference while black [sic] judges showed no clear overall preference. When subliminally primed with black-associated words in a hypothetical vignette, judges who expressed a white preference on the IAT were more likely to impose harsher punishments on defendants in the story than those primed with neutral words. When race of the defendant was made explicit in a hypothetical vignette, black [sic] judges were appreciably more willing to convict the white defendant rather than the defendant identified as African-American.

Justice Hyman's article recalls a moment when an attorney saw an African American woman exiting the judge's chambers and exclaimed “you must be the law clerk.” It turned out the woman was not the law clerk but in fact the judge. The justice then questioned whether this incident would affect the way the judge viewed that attorney moving forward in the case and how the judge's rulings would be affected. If the judge's actions were somehow influenced moving forward, this incident would affect not only the attorney but ultimately the client being represented.

Even this informed judicial perspective reveals a deeper level of hidden bias. Why would the judge let a personal slight impact her rulings? Perhaps because the author's common sense tells him that some female judges would be offended by the assumption that they are not judges. Professors Kang and Banaji and others caution against the use of so-called “common sense” by judicial actors because theories and notions that were once considered common sense are less accurate now based on the literature and empirical research. For instance, many may have considered it “common sense” that men on average are better drivers than women, but automobile insurance companies have studied the data and give women (especially when compared to young men) lower rates because their risk of accidents is lower.

Why would his common sense tell him the judge might be offended? Is it because of the stereotype that African American women are “unforgiving,” “hard,” “cold,” and perhaps are more likely to hold grudges over real or perceived slights than white males or other females? Quite probably. And so, she is second-guessed when a young white male judge would get the benefit of the doubt that he would not hold any such grudge if he instead had been mistaken for the law clerk or an off-duty bailiff. Implicit bias has an impact.

Another recent study of over 200 sitting judges analyzed the strength of positive stereotypes of white people and Christians, along with both negative and positive stereotypes of Asian people and Jewish people. Self-reporting using IAT tests, the study found most judges “displayed strong to moderate implicit bias against Asians relative to whites and against Jews relative to Christians.” The participants also completed a sentencing task--reading a file and making a sentencing decision where the race and religion of the defendant varied. Other findings include: (1) State judges gave longer sentences to white defendants than to Asian defendants; (2) Anti-Jewish, pro-Christian biases accurately predicted lesser sentences for Christian defendants, and (3) male judges showed stronger anti-Jewish biases than female judges.

The behavioral realism approach posits that judges need to keep up with empirical research if they are going to implicitly or explicitly base their decisions on theories about how people behave. Do judges even want to address the issue of implicit bias? Professor Eric Girvan argues, “the law-science gap exists and persists in significant part because judges believe they lack the ability to effectively remedy non-purposeful discrimination of the kind described by work on implicit bias and are unwilling to take steps necessary to develop ways to do so.” He criticizes judges for not doing a better job of “conforming their assumptions about human behavior to available social science.” He notes a Westlaw search he did, revealing “more examples of cases in which a majority of the judges indicating that they are aware of the concept refused to alter anti-discrimination doctrine accordingly in cases in which a majority of judges acknowledged evidence of implicit bias as justification for liability.”

In other words, despite reasoning that implicit bias played a role in allegedly discriminatory behavior, the judges declined to extend the interpretation of antidiscrimination doctrine to permit or support a finding of liability. In one recent case, the U.S. Court of Appeal for the Fourth Circuit broke from this pattern, holding that “[i]nvidious discrimination steeped in racial stereotyping is no less corrosive of the achievement of equality than invidious discrimination rooted in other mental states.” The court's comment explained that while the circuit had “admonished district courts, albeit in unpublished, non-precedential decisions,” it is now “past the time when that admonishment should be given precedential force.” Recognizing the “real risk” that legitimate claims based on subtle “stereotyping or implicit bias” may be dismissed by judges substituting their own rationales, the court reversed the district court's dismissal. Professor Girvan concludes, “if the antidiscrimination [sic] law-science gap is caused primarily by a lack of judicial knowledge, ... the solution to it should be to inform judges of research supporting the psychological science of implicit bias and evidence of its applicability to the cases they are deciding.” Part four addresses these suggestions.

E. Implicit Bias in the Courthouse

In the courthouse, further research is needed on the impact of implicit bias on court personnel. Court workers may treat people differently based on what they look like and based on implicit biases. As Professor Debra Bassett notes:

Court clerks who accept court filings may unconsciously respond differently to individuals of different races leading them to provide more help to some individuals than to others. As a mundane example, suppose a litigant presents paperwork for filing, and the paperwork lacks a required two-hole punch across the top. The court clerk may reject the paperwork when offered by some individuals but in other cases may accept the paperwork and simply punch it themselves.

Differences in treatment can lead to real consequences for litigants--even impacting the outcome of their cases.

Another important facet of civil litigation is how few cases go to trial. We know that the attorney's assessments of the case and its settlement value play a large role in the recovery options for civil plaintiffs. Implicit biases of settlement judges, mediators, and other dispute resolution actors may exacerbate the prejudice to the fair administration of justice.