chevron-down Created with Sketch Beta.
April 29, 2024 Feature

Implicit Bias and Jury Trials: A Report on an Experiment in Washington

Dave McGowan
Implicit bias teachings alone are not enough to decide specific challenges, and verdicts should not be overturned on possibilities or presumptions.

Implicit bias teachings alone are not enough to decide specific challenges, and verdicts should not be overturned on possibilities or presumptions.

Getty Images/Mint Images RF/Mint Images

Implicit bias concepts are advancing in the law. They may be seen in mandatory continuing legal education requirements, jury instructions, rules for selecting jurors, and, in two notable cases from Washington—State v. Berhe and Henderson v. Thompson—in rules governing post-verdict challenges. At the same time, psychologists are tempering claims once made for implicit measures of bias, such as the Implicit Association Test (IAT). They caution against using the test to select jurors, for example, and warn that IAT data should not be interpreted as identifying a person as racist or likely to engage in racist behavior. Current scholarship focuses on implicit bias concepts as educational and research tools that might heighten public awareness, as opposed to diagnostic tools to decide specific cases.

Washington’s experiment is to borrow from a jury selection rule a standard for determining whether a verdict is tainted by implicit bias. Under this standard, a prima facie case of taint can be shown if a reasonable person schooled in implicit bias could perceive race as a factor in a verdict, in which case the verdict is vacated unless the prevailing party proves such bias was not a factor. Proving such a negative is not feasible as a practical matter, however, because Washington has declared implicit bias to be pervasive, impervious to introspection, and undetectable by direct inquiry. Relative to prior law, this experiment uses implicit bias rhetoric to make it harder to exclude jurors through peremptory challenges and then uses the same rhetoric to make it easier to disregard verdicts.

The motivation for this experiment is laudable. The Washington Supreme Court wants to make jury rooms (in Berhe) and courtrooms (in Henderson) less hostile to Black persons. The decisions issued in each case contain useful insights. However, each case also provides a cautionary example of the limits of implicit bias rhetoric in assessing verdicts. Implicit bias concepts do not tell judges how to distinguish tainted verdicts from others.

Use of implicit bias rhetoric to justify limits on peremptory challenges is valuable because a permissive rule for seating jurors is desirable. Such a rule strengthens juries as the voice of a community. But this rhetoric should not be used to justify presumptions that discard jurors’ work, both because nothing in IAT data or other implicit bias research justifies such use and because a permissive rule for disregarding verdicts undermines that voice.

Washington is grappling with an important question: How should courts deal with jury trials where the record makes witness credibility fair game, stereotypes exist, words may have both literal and nonliteral meanings, and implicit bias premises have been endorsed as a matter of law? This article proposes that evidence of explicitly racist language or conduct should not be a necessary condition for challenging a verdict on the ground of racial taint, nor should a finding of intentional misconduct be required, but a claim of implicit bias should not trigger a presumption of bias and the prevailing party should not be given the impossible task of disproving the effect of a presumptively pervasive force. Instead, deference should be given to trial judges who can and should assess both what is said in court and what it means in the context of a specific case.

Because specific facts are more important than abstract concepts, it is best to begin with the stories of Henderson and Berhe.

Henderson v. Thompson

In June 2014, Alicia Thompson hit Janelle Henderson from behind while driving. Thompson was moving at 40 miles per hour; Henderson’s speed does not appear in the briefing or opinions. Henderson was not pushed into any cars in front of her. Both parties drove away. Henderson sued Thompson, who conceded liability. Thompson is white, as was her (female) counsel. Henderson is Black, as was her (female) counsel. The full racial composition of the jury is not in the record, but no Black person was a juror.

Henderson has Tourette’s syndrome. She argued that the accident made it worse. She sought $3.5 million in general damages. Thompson argued that this amount was too high; she denied that the accident worsened Henderson’s condition. Thompson suggested that, if the jury accepted Henderson’s damages estimate ($250 per day), an amount no greater than $60,000 was warranted.

During closing argument, Thompson’s counsel argued that Henderson’s demeanor in a pretrial physical examination (played for the jury) and in court showed her to be uncooperative and “combative.” Below are some examples of the argument cited on appeal:

Now, you’ll recall that during my cross-examination of Ms. Henderson a couple of days ago, she was confrontational with me, asking to know why I was putting her on trial. Her point was, I was hit; I was rear-ended; I have injuries. And she wants the inquiry to end there. . . . [W]hy are we going through this exercise? And it seems pretty evident that the reason we’re going through this exercise is because the ask is for three and a half million dollars.

Henderson presented testimony from some of her friends, three of whom were Black women. Thompson’s counsel argued:

I thought it was interesting also that all four of those witnesses used the exact same phrase when describing Ms. Henderson before the accident: life of the party. Almost—almost like someone had told them to say that. It was—it was like a tape on repeat. She was described as a model with a slender body to die for who gained significant weight after the accident. Obviously, Ms. Henderson was interested in fashion. They said she loved to shop and dress in colorful outfits, but could no longer shop for those outfits after the accident. But, again, information that’s directly controverted by even Ms. Henderson’s own medical—medical providers.

On appeal, Henderson quoted Thompson’s counsel as suggesting that jurors should “set aside” the “inherently biased testimony of Ms. Henderson’s friends and family.”

In what became the central issue on appeal, Thompson’s counsel argued that Henderson’s demeanor belied her credibility:

But when it’s my turn to cross-examine her, she’s not interested in the search for truth; she’s interested in being combative. Why are you putting me on trial? I don’t know what I told my doctors. I don’t know when I saw my doctors. I don’t know what they have in my reports. I didn’t read the medical records. . . . You know, it was—it was quite combative. There’s—there’s definitely no search for the truth there.

The jury was out for less than a day. It awarded Henderson $9,200. Henderson moved for a new trial or for additur. She argued:

Defendant’s closing argued that plaintiff was “combative” and that her attorney was “intimidating” which are racially biased code words frequently used to malign African-American women. The jury’s award of $9,300 [sic], 1/5th of the amount defendant acknowledged was fair, can only be the result of racial animus against plaintiff and her attorney, two of the five African American women who testified in the courtroom.

The trial judge was the Honorable Melinda J. Young. At the time a relatively new judge, she graduated from the University of Washington Law School and had been a prosecutor for 20 years prior to her appointment. She held a special interest in mental health issues; she helped to create a mental health court to emphasize treatment and housing for criminal defendants with persistent mental illness. At present, her website biography states that she is a member of the King County Bar Association Diversity Committee.

Judge Young denied Henderson’s motion. Because her order embeds what are in substance findings of fact, it is worth quoting in some detail. (Emphasis is added to passages material to the subsequent appeal.)

The Court recognizes that implicit bias exists. The Court recognizes the specific bias against African American women and the stereotypes of the “angry black woman,” or “welfare queen,” or “Jezebel.” The court further recognizes that using the terms combative in reference to the plaintiff and intimidated in reference to the defendant can raise such bias. What makes implicit bias insidious is the subtle nature of the animus and the difficulty in determining its presence. It can be difficult for a person with implicit bias to recognize it in him or herself, much less recognize when triggered by racial stereotypes. However, there is no case that finds that the possibility of implicit bias is grounds for a new trial or additur.

In this case, the use of the terms that the plaintiff now complains of was not objected to when defense counsel made her argument. The terms were tied to the evidence in the case, rather than being raised as a racist dog whistle with no basis in the testimony. Ms. Henderson was very uncomfortable being cross examined and submitting to the CR 35 examination. There are a multitude of ways to describe her demeanor and it was not unfair to describe her as combative given her unwillingness to answer questions. Ms. Thompson was also uncomfortable testifying, although she did not avoid plaintiff counsel’s questions. It was not unfair to describe her as intimidated, especially when the reference was to the process and not intimidated by plaintiff’s counsel. The court cannot require attorneys to refrain from using language that is tied to the evidence in the case, even if in some contexts the language has racial overtones. . . .

While the amount of the verdict was well below what the plaintiff had asked for, and below what defendant had suggested would be appropriate if the jury found plaintiff’s calculation of damages to be appropriate, that does not prove implicit bias. The defendant did not concede that Ms. Henderson’s Tourette’s worsened after the collision; indeed that fact was hotly disputed at trial. Nor did the defendant concede that the plaintiff’s method for calculating damages was the appropriate method. The court understands the plaintiff’s suspicions about how race may have influenced the verdict; race can influence many things and juries are not immune to bias. However, in the absence of specific evidence of impermissible racial motivations by the jury, or misconduct by defense counsel, the court declines to use the possibility of implicit racial bias to overturn the jury’s verdict or grant additur.

Judge Young issued her order denying Henderson’s motion on July 17, 2019. The following day, the Washington Supreme Court issued its opinion in State v. Berhe.

State v. Berhe

Tomas Berhe was convicted of first degree murder for shooting Everett Williams. The jury was diverse but had only one Black juror. (Berhe is Black.) Jurors deliberated for three and a half days; they were polled and confirmed their verdict. The next day, Juror 6 contacted defense counsel and the court. She said that she had acquiesced in the verdict against her wishes. Defense counsel then began contacting other jurors, whose contact information Juror 6 had provided. After these contacts came to light, the prosecution began contacting jurors as well. The trial court ordered counsel to stop contacting jurors. Instead, the court sent letters to each juror providing contact information for counsel on each side. Multiple affidavits were then obtained.

Berhe subsequently moved for a new trial based in part on juror misconduct. In support of the motion, Berhe relied upon the affidavit of Juror 6, who averred that she had been the last juror holding out for a not guilty verdict and that other jurors had accused her of being “partial,” which she understood as referring to her race. She stated that other jurors were dismissive of her comments and characterized as stupid or illogical some of her suggestions, such as that Berhe (who was arrested in a car containing the murder weapon) could have taken the gun from the real shooter. Her affidavit did not accuse other jurors of using explicit racial language, but Juror 6 stated that she “felt emotionally and mentally exhausted from the personal and implicit race-based derision from other jurors.”

The trial court denied the motion for a new trial. It noted that Juror 6 did not contend that any racist language had been used and stated: “I understand about implicit bias. But we can’t just assume. I think it’s equally likely that she felt pressured because she was the lone holdout as she did because of any other reason.” The trial court said that “[i]t is not inappropriate for jurors to press other jurors on their respective positions during deliberations” and that “[t]he remaining hold-out juror is frequently subject to pressure by fellow jurors and such pressure is not inappropriate.” The trial court held that Juror 6’s affidavit failed to make a prima facie showing of racist juror misconduct, implicit or explicit. The court of appeals affirmed.

On appeal from that ruling, the Washington Supreme Court reversed. The court held that the trial court abused its discretion by not holding an evidentiary hearing to determine whether a prima facie case of racist misconduct could be shown. The court recognized that such a hearing ran counter to the principle that secrecy of deliberations is foundational for a jury system, a principle reflected in the common law rule that jurors may not impeach a verdict and in provisions such as Federal Rule of Evidence 606. The court noted, however, that in Peña-Rodriguez v. Colorado, the United States Supreme Court recognized a constitutional exception for cases in which one juror accused another of overtly racist comments that raised a question whether the accused juror’s vote was tainted by racist bias. Washington had reached a similar conclusion earlier in State v. Jackson, a case also involving a juror affidavit stating that another juror repeatedly expressed overtly racist views directed at a category of persons that included the defendant, certain alibi witnesses, and the juror who submitted the affidavit. The Washington Court of Appeals held that where a juror affidavit established a “clear inference of racial bias,” a trial court should conduct an evidentiary hearing into allegations of bias before ruling on a new trial motion.

Acknowledging that there was no claim that explicit racial language had been used in deliberations, the Berhe court said:

A juror’s racial bias is uniquely difficult to identify because many people who harbor explicit biases will not admit to doing so, and everyone harbors implicit biases that are difficult to recognize in oneself. . . .

[D]etermining whether a person has been influenced by implicit racial bias is inherently challenging. Courts cannot simply ask the person about it and assess the credibility of his or her response because “people are rarely aware of the actual reasons for their discrimination and will genuinely believe the race-neutral reason they create to mask it.” Therefore, a person may honestly believe and credibly testify that his or her actions were not influenced by racial bias, even where implicit racial bias did in fact play a significant role.

Nevertheless, we should not “throw up our hands in despair at what appears to be an intractable problem. Instead, we should recognize the challenge presented by unconscious stereotyping . . . and rise to meet it.”

The court posited that implicit bias is unconscious, such that biased jurors may truthfully testify that they acted without bias. It follows that truthful, credible testimony alone may not be decisive because it is insufficiently probative of unconscious truth. Similarly, the court held that “courts cannot base their decisions on whether there are equally plausible, race-neutral explanations” for behavior as “[t]here will almost always be equally plausible, race-neutral explanations because that is precisely how implicit racial bias operates.”

The court instead announced a standard derived from Jackson and Washington’s General Rule 37, which is directed to the pretrial exercise of peremptory challenges. That rule altered the burden of proof established in Batson v. Kentucky, which disallows a peremptory challenge only upon a finding of intentional discrimination. Batson effectively creates a three-part test. A party wishing to object to a peremptory challenge must make a prima facieshowing that the challenge was racially motivated. Absent such a showing, the objection fails. Upon a prima facie showing of racial motivation, the party asserting the peremptory must articulate a permissible basis for the challenge. If it does so, the party objecting to the challenge can prevail only if it can demonstrate that the proffered explanation was pretextual.

Under Washington’s General Rule 37, if an objection to a peremptory challenge is made, then the party who asserted the peremptory must state its reasons on the record. The court then assesses them. “If the court determines that an objective observer could view race or ethnicity as a factor in the use of the peremptory challenge, then the peremptory challenge shall be denied.” An objective observer is defined as a person aware that implicit bias is among factors that have caused the unfair exclusion of jurors. Combined with the hearing requirement created in Jackson, this rule is the foundation for both Berhe and Henderson. As the court held in Berhe:

The ultimate question for the court is whether an objective observer (one who is aware that implicit, institutional, and unconscious biases, in addition to purposeful discrimination, have influenced jury verdicts in Washington State) could view race as a factor in the verdict. If there is a prima facie showing that the answer is yes, then the court must hold an evidentiary hearing.

Because the court had altered the standard for establishing a prima facie case, it remanded the case to the trial court to determine whether such a case existed and, if so, to hold a hearing. With respect to how the hearing was to be conducted, the court said:

Where the evidence is unclear or equivocal, as it will often be in cases of alleged implicit racial bias, the court must conduct further inquiries before deciding whether a prima facie showing has been made, for example, by asking the juror making the allegations to provide more information or to clarify ambiguous statements.

Berhe did not establish criteria for how the trial court should assess responses to these questions.

Impact of Berhe Ruling on Henderson

The Washington Supreme Court’s ruling in Berhe was issued on July 18, 2019. On July 26, Henderson filed a motion seeking a hearing under the rule of Berhe. The motion repeated some of the arguments made in the prior motion. Unlike the allegation in Berhe, which came from Juror 6 and was based on normally private deliberations, each ground in Henderson’s motion occurred in open court. The motion argued in part, for example, that “racial bias may have been a factor in the verdict” because “defense counsel based on no evidence falsely claimed that the only reason Ms. Henderson was there was because she wanted 3.5 million dollars.”

On August 7, 2019, Judge Young denied this motion. She distinguished Berhe on the facts and wrote that “[o]ther than the verdict being significantly less than plaintiff’s request, and the allegation that defense used racially coded language, there is no specific evidence that implicit bias was the cause of the verdict.” The court noted that, unlike Berhe, no juror had come forward to complain of bias. Judge Young reiterated that the court already had found that the defense argument was “tied to the evidence, rather than being used as a racist dog whistle.”

Judge Young also dropped a footnote to address an allegation Henderson had raised in her initial motion. When the jury returned its verdict, Henderson was asked to wait outside the courtroom in case jurors wished to speak to the lawyers. Based on her own affidavit and affidavits from her counsel and another lawyer in the courtroom, Henderson argued that jurors made this request and that it was evidence of racist taint. Judge Young stated that, following her standard practice, she, not the jurors, had requested Henderson’s exclusion. Judge Young apologized to Henderson and stated that she would no longer follow the practice.

At some point in this process, Henderson apparently hired a private investigator, who contacted a juror, presumably seeking a statement that could be used to support a claim of bias. No such statement appears in the record. Henderson’s action confirms that diligent counsel will understand Berhe to invite such investigation, a point emphasized in one of Berhe’s concurring opinions. Berhe and Henderson increase the chance that jurors in Washington will be subject to post-trial investigations whenever counsel thinks a claim of implicit bias may be made.

Continuing Proceedings in Berhe

Meanwhile, in December 2019, the trial court in Berhe convened the first of what became a series of remarkable hearings that appear to have been unprecedented in American law. After hearing live testimony from Juror 6, the court found a prima faciecase of bias based on Juror 6’s account of being treated differently from other jurors—her ideas being disrespected and, in particular, a juror’s accusation that Juror 6 was being “partial.” As required by the Washington Supreme Court’s ruling, the trial court then scheduled evidentiary hearings.

On May 25, 2020, Minneapolis police murdered George Floyd while he was in custody. The Washington Supreme Court issued an open letter, dated June 4, 2020, which in part urged the Washington legal community to “develop a greater awareness of our own conscious and unconscious biases in order to make just decisions in individual cases.”

The evidentiary hearings in Berhe were held between July 20 and August 5, 2020. Jurors from Berhe’s trial were called, sworn, and examined and cross-examined on their deliberations. The transcripts of those proceedings are instructive; to do them justice required an article of its own. In brief, they showed that Juror 6 felt sincerely that other jurors had disregarded her comments and in essence shut her down; other jurors felt sincerely that Juror 6 was unable or unwilling to give reasons tied to the evidence for the positions she took. It was plausible that other jurors could have attributed Juror 6’s reluctance to vote for conviction to her race; it was plausible that Juror 6’s reluctance was informed by her race.

On December 4, 2020, the trial court granted Berhe’s motion for a new trial. The court found that jurors worked diligently and in good faith to reach a verdict and that Juror 6 was not subject to duress. It noted that vacating the verdict would be hard on the family of the victim, Williams, but it also cited the Washington Supreme Court’s open letter and stated that “[i]mplicit racial bias can occur unbeknownst to the perpetrator of the bias, and it can be perpetuated by good people of good intentions because it is the product of our culture.” Technically, the trial court rested its decision on the burden of proof specified by the Washington Supreme Court—given the presumed pervasiveness of implicit bias, the State could not meet its burden of proving that race had not been a factor in the verdict. A juror could be forgiven for reading these comments as asserting that their truthful, credible testimony did not matter because the law presumes that they did not know their own minds.

Henderson’s Appeal

Henderson appealed to the Washington Supreme Court, which took her case on direct appeal. Her opening brief was filed on August 12, 2020. Henderson argued that Judge Young erred because she “acknowledged defendant Thompson used language with racial overtones in closing” but denied the motion for a new trial. Henderson argued that the language from closing “reflects implicit bias which is insidious and can trigger a jury without its awareness.” Henderson’s brief accused both Judge Young and Thompson’s counsel of racially biased conduct. She sought a new trial before a new judge “due to attorney misconduct and discrimination.”

As to Judge Young, Henderson argued that “several in-trial rulings were so unreasonable and prejudicial to Henderson as could only be bias.” The brief repeated the charge that the jury asked that Henderson be removed from the courtroom before the jury left. Henderson argued that this request proved that the jury had been biased into believing Henderson to be violent or dangerous. Judge Young had stated twice that no such request had been made; Henderson cited her own affidavit and those of her counsel and of another lawyer who had been present, which averred that they heard Judge Young attribute to jurors the statement that they were willing to speak with counsel but only if Henderson was removed.

Henderson’s brief accused Thompson’s counsel of unethical conduct: “Arguing falsehoods, opinions and vouching for one’s own client is improper and misconduct. RPC 3.4, RPC 8.4.” Henderson argued that counsel “attacked Henderson’s motivations, her demeanor, her ‘problematic attitude’, the way she answered questions, all of which [counsel] argued was steeped in fraud, collusion, or immoral behavior. Thompson juxtaposed this negative view of Henderson to her own ideal white female client who she attested was ‘authentic and genuine,’ ‘honest’ and truthful.”

Henderson’s brief alluded to the concept of “Karens” to frame the conduct of Thompson and her counsel at trial. The brief advanced two examples. Thompson evidently trembled while on the stand; Henderson’s brief asserted that she did not tremble when the jury was not present, “reminiscent of the actions of ‘Central Park Karen.’” In closing, Thompson’s counsel had argued that her client testified credibly even though her client was intimidated by the process, “and rightly so.” Henderson argued that the phrase “and rightly so” “sought to form a bond with the jury on grounds steeped in the stereotypes that black people are inherently dangerous and white women are in need of protection.” The footnote to this statement reads: “For example: ‘Central Park Karen’ aka Amy Cooper; Emmitt Till accuser Carolyn Bryant; Robin DiAngelo, White Fragility (2018).”

Thompson’s responsive brief largely defended the trial court’s rulings. She emphasized that the comments Henderson objected to were tied to the record, as the trial court had found. She argued that her comments on witness credibility fairly pointed out improbable similarities in their language and bias stemming from friendship with Henderson rather than from race. Thompson argued again that she had not conceded that Henderson’s claimed injuries had been proved or that the $250-per-day damages figure was justified.

Henderson’s reply framed the main legal issue effectively. The trial court had said two things: (1) defense counsel’s statements were tied to the record, and (2) they were not veiled racist appeals. It is possible, though not necessary, to read the trial court as holding that (2) followed from (1)—that the comments were not veiled racist appeals because they were tied to the record. Picking up on this possibility, Henderson argued that it cannot be right to say that if a lawyer can tie a racist concept to the record, then the lawyer can make racist arguments: “If the trial court’s ruling was law, it means that any civil attorney could make any racially charged statement as long as she tied the statement to the evidence. This is a veritable roadmap to Jim Crow.”

The trial court’s ruling was not as sweeping as this description implies (the case would have been very different had explicitly racist language been at issue), but this argument gets to the heart of an important legal issue raised in Henderson: How should courts deal with jury trials where the record makes witness credibility fair game, stereotypes exist, words may have both literal and nonliteral meanings, and implicit bias premises have been endorsed as a matter of law?

Washington Supreme Court Opinion in Henderson

The Washington Supreme Court reversed Judge Young’s order denying Henderson’s motion for a new trial and remanded the case for an evidentiary hearing before a different judge. The court held that “an objective observer could conclude that racial bias was a factor in the jury’s verdict. At that hearing, the court must presume racism was a factor in the verdict and Thompson bears the burden of proving it was not.” Regarding the prima facie case of racial taint, the court reviewed the record de novo and drew conclusions contrary to those made by Judge Young. For example, the Washington Supreme Court stated—and effectively held—that Thompson’s counsel employed racist stereotypes, the opposite of Judge Young’s conclusion:

The direct contrast between defense counsel’s depiction of Henderson as “confrontational” and “combative” and her depiction of Thompson as “rightly” “intimidated” and “emotional” distorted the roles of plaintiff and defendant, casting Thompson—the person responsible for injuring Henderson—in the role of the victim to whom the jury owed more sympathy than the actual injured party. This invited the jury to make decisions on improper bases like prejudice or biases about race, aggression, and victimhood. . . .

Thompson’s counsel alluded to racist stereotypes about Black women as untrustworthy and motivated by the desire to acquire an unearned financial windfall. Defense counsel argued that Henderson’s injuries were minimal and intimated that the sole reason she had proceeded to trial was that she saw the collision as an opportunity for financial gain. . . .

Additionally, defense counsel relied on racist stereotypes about Black people and us-versus-them descriptions to undermine the credibility of Henderson and her witnesses. For example, defense counsel suggested that Henderson had probably asked her friends and family to lie for her, as evidenced by their shared use of a popular idiom—“life of the party”—to describe her. This argument was akin to the prosecutorial misconduct we condemned in Monday, where the prosecutor asserted that Black witnesses were unreliable because there was a “code” that “‘[B]lack folk don’t testify against [B]lack folk.’”

The court accepted additional criticisms of Thompson’s argument as well.

The Henderson court also said that upon acknowledging that some of defense counsel’s arguments might have racist overtones “in some contexts,” the trial court “failed to engage in any analysis of what effect the racially coded language could have had on the jury.” As in Berhe, the court offered no guidance for how to discern such effects other than by presumption. More strikingly, the court chided Judge Young for failing to consider “the perspective of the objective observer under Berhe” and instead “view[ing] the facts from her own perspective.”

On this point, the court effectively held either that trial judges may not bring their own observations to bear in assessing accusations that bias tainted a trial or that Judge Young was not an impartial observer. Either conclusion conflicts with the rule that part of the purpose of trial is for jurors, and judges on some issues, to form judgments about witnesses. Impressions derived from trial do not count as “bias” for purposes of disqualification in normal cases, and there appears to be no legal or logical basis for removing a trial judge from a case because the judge’s order based on the record did not square with de novoappellate fact-finding. One would, however, expect to see an increase in requests for reassignment, and in disqualification motions, following Henderson.

Following this ruling, Henderson petitioned the United States Supreme Court for a writ of certiorari. The Court denied the petition. Justice Alito published an opinion, joined by Justice Thomas, noting that review was premature in view of the remand but stating that the remand hearing “appears to have no precedent in American law.” Justice Alito stated that the Washington Supreme Court’s opinion “raises serious and troubling issues of due process and equal protection. In some cases, it will have the practical effect of inhibiting an attorney from engaging in standard and long-accepted trial practices . . . .” Justice Alito observed that “if the Washington courts understand the decision below to be as sweeping as it appears, review may eventually be required.”

Practical Effects of Henderson and Berhe

As a practical matter, Henderson states a rule of conduct for counsel rather than a standard for assessing taint in a verdict. Henderson holds that it is misconduct for counsel at trial to use words consistent with a negative stereotype that a court deems applicable to persons in at least racial categories, even if the words do not explicitly refer to race and even if the words tie off to record evidence. This holding applies to parties in such categories, and its logic would extend to witnesses as well. Under Henderson, such words are deemed to establish a prima facie case that a verdict is tainted by racial bias, and a verdict can be preserved only if the prevailing party can prove a negative—that racial bias, implicit or explicit, was not “a factor” in the verdict.

Because Berhe proclaims implicit bias to be pervasive and undetectable by direct inquiry, one would expect the prima facie case to be decisive and vacatur to follow. The remand hearing in Berhe produced testimony that amounted to a strong endorsement of the jury system, but the trial court vacated the verdict, citing the burden of proof. Literal application of the standard in Berhe will render the result of such a hearing a foregone conclusion. In effect, therefore, Henderson’s application of the Berhe rule means that, even absent an objection at trial, the use of such words will lead to a new trial at the option of the party against whom they are directed.

It is useful to place this rule in context by comparing it to the rules governing new trial motions based on argument misconduct unrelated to race and the rules governing expressly racist statements. In criminal cases not involving explicit racial comments (vouching, for example), new trial motions require a defendant to show that the objectionable comments were improper and prejudicial, which requires a showing that they were substantially likely to affect the verdict. Comments are not viewed in isolation but are placed in the context of the record as a whole and the jury instructions. A defendant who does not object during trial waives objection “unless the comment is so flagrant and ill-intentioned that it causes an enduring and resulting prejudice that could not have been neutralized by a curative instruction to the jury.”

In contrast, a comment that is flagrantly or intentionally an appeal to bias—such as the astonishing argument in Monday that “Black folk don’t testify against Black folk”—is treated as per se prejudicial and a new trial is required. Harmless error analysis does not apply. This rule is sound. Expressly racist comments threaten to impugn public confidence in courts in a way that vouching, for example, does not, and a rule against using expressly racist language is administrable—it is easy to state and relatively easy to follow. The penalty for doing so may be strict without deterring permissible arguments because the line is relatively clear.

Though formally the court’s findings were directed only to establishing a prima facie case of bias, Henderson effectively treats words that are not overtly race-based, and that are tied to the record, but that may, in the appellate court’s view, convey racial overtones, as if they were explicitly racist comments. This reading of Henderson is supported by the majority opinion’s assertion that Thompson’s argument that Henderson’s witnesses might have coordinated their testimony because they used identical language (“life of the party”) was “akin” to the expressly racist assertion in Monday. (Two justices disagreed with this equation in a concurring opinion.)

In favor of this approach, one might argue that both (1) expressly racist words and (2) words that do not literally refer to race but carry racial meaning in context can taint a verdict with racism. That is right. That general principle is not enough to decide specific cases, however. A trial lawyer can only avoid using terms consistent with a stereotype if the content and parameters of the stereotype are known and words likely to trigger it are known as well. Unlike the case with explicitly racial appeals, Henderson creates significant line-drawing problems for counsel who must follow these rules in the dynamic context of trial. Assuming the court did not mean to hold that a demand by a Black woman plaintiff could never be called excessive, or that a Black woman witness’s demeanor was off-limits for argument, what would it allow defense counsel to say? The court offered no parameters for permissible argument.

Read in view of Berhe and the court’s open letter following the murder of George Floyd, Henderson is best understood as part of the Washington Supreme Court’s project to make courthouses more welcome places for persons for whom they have historically been places of fear, humiliation, and distress. The goal is laudable. Among other reasons, if a Black person comes to court expecting hostility and humiliation, that expectation may, in and of itself, affect their testimony. People who expect to get punched may go into a crouch, or they may fight questions rather than just answering them. They may seem combative, not because they are making excessive claims or abusing the system, but because they expect mistreatment. Henderson shows commendable sensitivity to this point. The goal may get complicated in practice—it is not clear how it should operate if, for example, a Black prosecution witness with an immunity deal testifies against a Black defendant in a criminal trial—but the ideal that trials should be about evidence and not stereotypes is worth pursuing.

To use jury selection rules to pursue that ideal in the context of post-verdict challenges, however, risks intruding into ground legitimately reserved to jurors. The Henderson rule is needed only when argument is not overtly racist and (because argument not grounded in the record can be penalized on that basis) only when argument is grounded in the record. Following Henderson, counsel in a case with a Black opponent would be well-advised to truncate such argument to avoid the risk that an appellate court would find some comment a veiled racist appeal, while counsel in a case with a white opponent could engage in sharper attacks. Justice Alito referred to this aspect of the case as “special, crippling rules.” One might question the strength of his rhetoric, but the basic claim that Henderson implies two standards for assessing argument is a fair reading of the opinion.

Having said that, as a practical matter the idea that counsel dealing with parties or witnesses in racial classes protected by law might need to be extra careful with their language does not seem all that radical. In many jurisdictions, it would just be common sense. Prudent trial counsel should always be concerned that an attack on a witness may backfire. Even legitimate points scored on cross-examination can be drowned out if the manner of questioning or argument alienates jurors. And a failed attack on a witness does not just fail to gain ground; it is likely to be costly in terms of at least trial time and probably in a loss of counsel’s ethos.

For this reason, though Justice Alito’s observation may point to a vulnerability in the Henderson opinion should certiorari ever be granted, it is not clear that the opinion’s incremental deterrent effect will be great in practice. And the law does need a way to give recognition to the fact that facially neutral language can, in some circumstances, pack a racist punch. The Henderson court was not wrong about that. The important questions are who decides whether that has happened, and how.

As a practical matter, the Henderson court engaged in de novo review of what were in effect informal findings of fact made by Judge Young based in part on her observation of Henderson’s demeanor. The court’s de novo fact-finding is troubling. What words mean depends on the context in which they are used—including the full record, jury instructions, and opportunity of Henderson’s counsel to object to the statements and perhaps to flip them to her advantage. Fact-based analysis of meaning in context is the only way implicit bias ideas can be used to inform rather than to avoid analysis of whether a verdict was tainted by improper conduct.

Henderson declared that the statements recited above raised a prima facie case of racist taint because (to take a representative example) the statements “evoke the harmful stereotype of the ‘angry Black woman.’” This statement embeds three factual assertions: (1) there is a demeaning stereotype depicting Black women as angry; (2) the content of that stereotype is that Black women may have reactions that are proportionate to some harm but that white people may perceive as disproportionate; and (3) it is reasonable to assume that the stereotype could be evoked in the minds of jurors by the term “combative.” The majority’s assertion that “life of the party” is a “common idiom,” such that it was racist to suggest that the use of the phrase by three witnesses implied coordination of testimony, is a factual assertion as well.

What is the foundation for these assertions? The court did not say explicitly, but they reflect a form of either judicial notice or judicial assumption. And because the court’s review was de novo and focused on small portions of argument, it effectively asserted that context does not matter—“combative” is a trigger no matter what, and “life of the party” is a common, anodyne phrase, no matter what. How the terms landed in the context of the full record—what they meant in the room—is not part of the court’s opinion because it was not in a position to know.

The most notable example of the court’s assertive approach regarding facts was its comment that Thompson’s counsel tried to persuade jurors that Henderson’s argument “distorted the roles of plaintiff and defendant, casting Thompson—the person responsible for injuring Henderson—in the role of the victim to whom the jury owed more sympathy than the actual injured party.” This assertion is problematic. Both Judge Young and Thompson’s briefing confirm that she denied that the accident worsened Henderson’s preexisting condition. That was the question to be tried. Yet the passage above is fairly read as the Washington Supreme Court presuming that Henderson’s demand was not excessive relative to any injury she suffered, a question reserved for the jury. Perhaps the court meant only to acknowledge that Henderson in fact won at trial—the jury’s $9,200 award reflected a finding that Thompson had in fact injured Henderson. Fair enough. But a plaintiff may be both the victim of an injury and the proponent of an unreasonable damages demand—one disproportionate to any harm the defendant caused. The defense is entitled to try to persuade jurors that a defendant who is made the target of such a demand is at least as much a victim as the proponent of the demand. Which argument is stronger is the jury’s call.

The “angry Black woman” stereotype exists and is used to trivialize the experiences of persons to whom it is applied, or worse. The Washington Supreme Court was not wrong about that. It is not clear how the stereotype pertains to damage demands by plaintiffs in civil litigation, however. Is a $9,200 award evidence of racism, conscious or unconscious, because it is so slight relative to a $3.5 million demand? Using conventional legal analysis, one would feel unable to reach such a conclusion without first knowing how the demand was set and whether it was justified by the facts, which, again, was the question to be tried.

Perhaps the court meant to hold that the demeanor of Black women witnesses may not be impugned because jurors are too apt to leap, consciously or not, to biased inferences. This rule would fit with the goal of making courthouses less hostile to Black witnesses, but the existence of the stereotype does not address or establish such a connection because one cannot judge whether a reaction is proportionate without assessing the harm to which it is tied and because the stereotype the court identified does not presume a trial in which parties are allowed to create their own frames for a party’s behavior to explain why their reactions or demands are proportionate. Like IAT testing, stereotypes as such do not contemplate trial procedures designed to provide context for assessing behavior and to expose any effort at stereotypic race-baiting.

Jurors were entitled to like and trust Henderson or to dislike and distrust her. They were entitled to accept that she was harmed in an amount compensable only in millions, or that she was harmed little, if at all. They were not entitled to reach any conclusion because of her race. Whether the $9,200 award was based on permissible or impermissible factors is a factual question rather than a legal one. And in a case not involving expressly racist language it depends vitally on context—the sense and feel of the courtroom and how testimony and argument landed in the room.

On a fair reading of her rulings, Judge Young saw and heard the testimony, found that the record supported multiple ways to characterize Henderson’s testimony, found that Henderson but not Thompson refused to answer questions, and determined that, in this case, “combative” was not a veiled racist appeal. On each point, she was in a better position than the Washington Supreme Court to make those findings. If Judge Young had seen things differently, there would be no basis to disturb an order granting additur or a new trial motion, but the principle works in both directions: when explicit language is not used, what matters is the meaning of words in the courtroom, which is a question of fact. Her findings should have been reviewed for substantial evidence. The court’s de novo approach, to borrow its own phrase, got it exactly backward.

Limits of Implicit Bias Rhetoric

Implicit bias ideas gained legal interest roughly at the same time the IAT began to receive popular attention. IAT data were cited in an early academic paper calling for abandonment of the Batson requirement that racist intent be shown to reject a peremptory challenge, and today the IAT remains the most common legal referent for implicit bias concepts. A link to the IAT website at Harvard is, for example, the first reference in the ABA’s implicit bias toolkit.

Optimism regarding whether IAT data are useful in practical decisions has waned in recent years. Even IAT optimists do not recommend that it be used in making employment decisions or in selecting jurors—a disclaimer to that effect appears on Harvard’s IAT website. A recent paper whose lead author is Professor Anthony Greenwald—a preeminent implicit bias researcher whose lab at the University of Washington developed the IAT—moderates some earlier claims for IAT data, stating that they measure “associative knowledge about groups, not hostility toward them.” Referring to early descriptions of the IAT, the authors state “it was a mistake to equate implicit bias with racism or prejudice.” Writing in 2019, NYU Professor John T. Jost, an IAT optimist, summarized developments in the field to that point: “As a ‘bona fide’ pipeline used to quantify levels of ‘unconscious racism’ as a fixed property of the individual—or as a diagnostic tool to classify people as ‘having’ racism or sexism (like they might ‘have’ clinical depression)—the IAT is dead.” However dead the concept might be in psychology, it nevertheless appears to be alive in the law.

Its enduring vitality in the courts notwithstanding, Washington’s experiment in Berhe and Henderson is best read as a cautionary example of the limits of implicit bias rhetoric. Nothing in the psychological research that popularized the term “implicit bias” justifies treating it either as a diagnostic tool to detect racist taint in a verdict or as a presumptively causal force in a verdict. Neither Berhe nor Henderson establishes criteria that could be employed reliably to identify tainted verdicts. Nor do the cases establish any means by which implicit bias premises could be useful in such an inquiry. Instead, each case recites premises that prove too much.

Taken at face value, Berhe and Henderson entail that there is no such thing as an unbiased verdict. That may well be true in an abstract sense—everyone brings to court their own experiences and assumptions—but it is not true in any sense useful for analyzing challenges to specific verdicts. Any effort to apply implicit bias premises to vacate specific verdicts will either result in the categorical presumption Washington uses or will be arbitrary. The point may be generalized. If the IAT should not be used to pick jurors in the first place, as IAT research says, then the concepts the IAT popularized should not be used to presume that verdicts are tainted. Distinctions can only be drawn, and sound conclusions reached, based on case-specific facts.

The Henderson court’s dismissive treatment of Thompson illustrates this point. The court suggested that she could not be a victim, even of an inflated demand, because she was liable, and the court effectively put Henderson’s $3.5 million demand back on the table. That would be an unusually large auto insurance policy; an award like that would bankrupt most people. That fact would not impugn an award in that amount if the jury returned it, but it did not.

The court was similarly dismissive of jurors. Unlike Berhe, there was no complaint from any juror about even veiled improper language. Henderson argued bias as an inference from the size of the award—jurors must have been biased because they awarded too little—which is why her initial motion sought either a new trial or additur. Through generalizations about implicit bias, the court’s opinion depicted jurors as little more than racially primed automatons, walking repositories of biases that could be triggered without jurors’ awareness and could be presumed to be the cause of their verdict. The opinion does not work without this premise.

To apply the peremptory challenge standard to post-verdict review is to treat jurors with unacknowledged and unjustified disdain. No sound legal or psychological foundation for such treatment exists. It would be better not to blame jurors even implicitly for what is in fact a rule of conduct directed to lawyers.

Lessons from Washington’s Experiment

At a normative level, some commentators have decried Henderson. One law professor has been quoted as calling it “batshit crazy,” and in conservative circles it appears to be read as a form of political correctness gone off the rails. Because the Washington Supreme Court is wrestling with a difficult and important problem, that kind of criticism is neither fair nor justified. Trials merge and balance many interests, and balancing them is hard. An effort to alter the balance may tip too far in one direction, as Washington has with respect to post-verdict challenges (though not with respect to peremptory challenges), but the remedy is to better the balance, not to ignore the concerns.

In the law, arguments that are accepted are repeated. A logical inference from both Berhe and Henderson is that bias-based arguments will become more common because they have the demonstrated potential to win and because implicit bias concepts have no intrinsic bounds. Implicit bias instructions are common, and one should not be surprised if litigants use them to leverage bias arguments when they can, to make bias objections when they can, and to try to create records to support arguments on appeal. Race is the paradigm case, but there is no reason in principle why litigants cannot attempt to extend such arguments further. Perhaps it will be common for trials to both adjudicate facts and to serve as referenda on stereotypes.

From a law teacher’s perspective, Henderson suggests that students should understand the possibilities of this approach. Perhaps the lesson is that trials are already referenda on social biases, and the only difference Henderson makes is to bring that fact out in the open—though that goal might be achieved without the presumption the court adopted. It is arguable that Henderson goes too far on that point, but perhaps the premise is not wholly wrong.

To this author, the most salient points in the case are the lack of objection at trial, Judge Young’s findings, and the fact that the core complaint was not that the jury rendered a defense verdict (which it didn’t) but with the amount—thus the alternative request for additur. Looking at those facts, traditional legal analysis offers no reason to question Judge Young’s assessment. In her reply brief, Henderson argued that “traditional legal analysis” has “historically ignored the reality of racial bias and contemplates justice from a white-centered perspective. This pivot away from the reality of racial bias is a powerful method of white racial control which results in the protection of white advantage.” Maybe so. But the size of the award is not a proper measure of the merit of traditional analysis. The jury made the award.

Using juror selection rules to conduct post-verdict analysis has the effect of treating verdicts and, by implication, jurors with disdain. Other courts would do well to avoid that path. It is right to be suspicious of attempts to exclude jurors before trial because it is important for everyone in a community to have a chance to be part of the voice of the community expressed through a verdict. For that reason, it would be wrong to use implicit bias premises to exclude jurors, a point on which psychologists agree, and it is wrong to use these premises to treat jurors as bias-ridden ciphers whose decisions are presumptively flawed. That is not what the IAT is about, and certainly not anything it proves.

Increasingly, American society appears to be one in which people talk only to people who are like them and agree with them. Few institutions or places are left where people from disparate backgrounds, races, and classes gather to talk to each other about a common issue. The jury is one such place. A jury is in part an aspiration—that there is such a thing as common decency, humanity, and values, making it possible for people of different views to reason together and for a community to have a voice rather than a shouting match. A jury is in part a demand that any one of us be prepared to submit to the judgment of all of us—that we defend our actions and demands in terms of common values. And a jury is in part an acknowledgment that, like it or not, we are all in this together and we must trust each other to act together. It does not always work. Sometimes evidence will show it has gone wrong. But it is too important to assume away using the rhetoric of implicit bias. A romanticized view, perhaps, but not worthless for being romantic.

Stereotypes and biases are real, ostensibly neutral words can sometimes convey racist meanings, and the law must balance the interests of parties in a fair trial and the interests of jurors in being taken seriously as the voice of the community rather than being dismissed as presumptive victims of forces they cannot perceive. How should the balance be struck? Procedures may vary but certain points are key. Meaning is a contextual, factual issue, and whether a comment that is not explicitly racist was fair argument or a veiled racist appeal is a fact issue for the trial court, to be reviewed for substantial evidence. Trial judges should have independent power to find that an argument was impermissible and to enter appropriate sanctions; no finding of intent should be required. In all but egregious cases in which it would be an abuse of discretion for a trial court not to intervene, absence of an objection should be deemed substantial evidence that a comment was not a veiled racist appeal.

This approach is not perfect, but none will be. It does put most of the weight in the courtroom—where it belongs. The trial judges in both Berhe and Henderson said that they were familiar with implicit bias teachings, and there is no reason to doubt that. They also said that those teachings alone were not enough to decide specific challenges and that verdicts should not be overturned on possibilities or presumptions. They were not wrong.

    The material in all ABA publications is copyrighted and may be reprinted by permission only. Request reprint permission here.

    Dave McGowan

    University of San Diego School of Law

    Dave McGowan is the Lyle L. Jones Professor of Competition and Innovation Law at the University of San Diego School of Law. He teaches legal ethics and wrote Developing Judgment About Practicing Law, an ethics casebook published by West.