Honesty Is the Best Policy: Addressing Implicit Bias in the Judiciary

Indiana Journal of Law and Social Indiana Journal of Law and Social

Equality Equality

Volume 9 Issue 1 Article 4

1-26-2021

Honesty Is the Best Policy: Addressing Implicit Bias in the Honesty Is the Best Policy: Addressing Implicit Bias in the

Judiciary Judiciary

Catie Wheatley

[email protected]

Follow this and additional works at: https://www.repository.law.indiana.edu/ijlse

Part of the Law Commons

Publication Citation Publication Citation

Catie Wheatley, Honesty Is the Best Policy: Addressing Implicit Bias in the Judiciary, 9 Ind. J.L. & Soc.

Equality 94 (2021).

This Comment is brought to you for free and open access

by the Maurer Law Journals at Digital Repository @

Maurer Law. It has been accepted for inclusion in Indiana

Journal of Law and Social Equality by an authorized

editor of Digital Repository @ Maurer Law. For more

information, please contact [email protected].

COMMENT

Honesty Is the Best Policy: Addressing Implicit Bias in the Judiciary

Catie Wheatley

INTRODUCTION

People are likely to have an image of the perfect judge. Perhaps a wise old

man in a long flowing robe sitting on high and handing down a measured and wise

judgment. King Solomon, for instance, is often referred to as an idealized impartial

judge.

Depictions show King Solomon on a throne with the requisite beard and

robes, and portray him as wise, authoritative, and unambiguously patriarchal. In

real life, however, judges are not godlike kings but mere mortals. And this is a good

thing. By allowing judges to be human we can escape a narrative that continues to

cast the aged, white, male patriarch as an infallible archetype. King Solomon’s

fallibility is evident: a judge who is willing to split a baby in half with a sword and

use maternal instincts against a woman in order to solve a fairly straightforward

problem.

But, if mythical perfection is out of the question, what does this mean for

judicial impartiality or the “capacity to decide cases with an open mind and without

bias for or against those who appear before [them]”?

It means that impartiality is,

as it always has been, an aspiration. It is a worthy aspiration. One on which it is

incumbent upon the legal profession to strive towards. But, as Charles Geyh et al.

describe it, a realistic assessment of impartiality is not a question of whether a

judge is impartial, but whether a judge is “impartial enough.”

In keeping with Geyh’s realistic view of impartiality, I argue for a realistic

view of implicit judicial bias, i.e. that it is inevitable but addressable. Part I will

situate judicial bias within current social science research on implicit bias and

establish an illustrative case study from Judge Richard Posner’s decision in Wassell

v. Adams.

Part II addresses why judicial bias matters based on the normative and

practical concerns of fairness and legitimacy. Part III then discusses current

prohibitions against bias in the American Bar Association’s (ABA) Model Code of

Judicial Conduct (“Model Code”) and the Code of Conduct for United States Judges

(“Code of Conduct”) and demonstrates that while they address explicit bias, they are

not intended to address implicit bias and cannot do so adequately. And, Part IV

proposes practical safeguards based on social science research and tailored to

judges’ unique role.

See, e.g., Madsen v. Women’s Health Ctr., Inc., 512 U.S. 753, 784 (1994).

1 Kings 3:16–28.

CHARLES GARDNER GEYH, JAMES J. ALFINI, STEVEN LUBET & JEFFREY M. SHAMAN, JUDICIAL CONDUCT AND

ETHICS 1-3 (5th ed. 2013).

Charles Gardner Geyh, The Dimensions of Judicial Impartiality, 65 FLA. L. REV. 493, 493 (2013).

865 F.2d 749 (7th Cir. 1989).

2021] Honesty Is the Best Policy 95

I. IMPLICIT BIAS IN THE JUDICIARY

A. Current Social Science Research and Its Application to the Judiciary

For over eighty years, social science research on bias has contributed to a

growing body of literature.

Part of this research has given rise to the concept of

implicit bias and how it affects behavior.

Because judges are responsible for their

own behavior and their job is to make decisions on other’s behavior, implicit bias

research should inform how we view the judicial role.

To begin, Jerry Kang differentiates explicit from implicit bias as that which is

“introspectively accessible” versus that which is “introspectively inaccessible.”

As a

result, while people may be unwilling to report explicit bias, people are unable to

report implicit bias.

Using the Implicit Association Test (IAT), Mahzarin Banaji

and Anthony Greenwald measured unconscious mental associations by measuring

response times to words and pictures.

They found that “schema-consistent

pairings,” for instance white faces with positive words, took less time for people to

recognize than “schema-inconsistent pairings,” or black faces with positive words.

Similarly, they found that schema-consistent pairings of women with the word

“family” took less time and fewer errors for people to recognize than schema-

inconsistent pairings of women with “career.”

Further, according to Banaji and Greenwald’s IAT research, implicit bias

levels significantly exceed reported, or explicit, bias levels.

Other researchers,

using measures like facial electromyogram and galvanic skin response techniques

also found that implicit bias measures sharply diverged from reported bias.

Scholars have described implicit and explicit bias as “dual attitudes” that exist side

by side; while the explicit attitude requires conscious retrieval, the implicit attitude

is automatic.

In analyzing implicit bias, Kang unequivocally states “no one is immune,” in

part because brains are good at sorting things.

With this in mind, determining

MAHZARIN R. BANAJI & ANTHONY G. GREENWALD, BLINDSPOT, at xiv (2013).

Id.

IU Maurer, Fuchs Lecture: Jerry Kang “Intending Equality,” YOUTUBE (Apr. 17, 2015),

https://www.youtube.com/watch?v=4CqrGmuwWrs [hereinafter Kang Fuchs Lecture].

Id.

BANAJI & GREENWALD, supra note 6, at 39–41.

Jerry Kang, Mark Bennett, Devon Carbado, Pam Casey, Nilanjana Dasgupta, David Faigman, Rachel

Godsil, Anthony G. Greenwald, Justin Levinson & Jennifer Mnookin, Implicit Bias in the Courtroom, 59

UCLA L. REV. 1124, 1130 (2012).

BANAJI & GREENWALD, supra note 6, at 115.

See id. at 47, 67.

John F. Dovidio, Kerry Kawakami & Kelly R. Beach, Implicit and Explicit Attitudes: Examination of the

Relationship Between Measures of Intergroup Bias, in BLACKWELL HANDBOOK OF SOCIAL PSYCHOLOGY:

INTERGROUP PROCESSES 175, 183 (Rupert Brown & Samuel L. Gaertner eds., 2001).

Linda Hamilton Krieger & Susan T. Fiske, Behavioral Realism in Employment Discrimination Law:

Implicit Bias and Disparate Treatment, 94 C

AL. L. REV. 997, 1041 (2006) (citation omitted).

Kang Fuchs Lecture, supra note 8.

96 Indiana Journal of Law and Social Equality [9:94

whether judicial bias really exists is a simple syllogism: all people have implicit

biases, judges are people, therefore judges have implicit biases. The views of leading

jurists correspond with this conclusion. Justice Cardozo said, “deep below

consciousness are other forces, the likes and the dislikes, the predilections and the

prejudices, the complex of instincts and emotions and habits and convictions, which

make the man, whether he be litigant or judge.”

In keeping with this view, social

scientists have confirmed that five prevalent “cognitive illusions” affect judicial

decision-making

and have also found that judges can associate negative attitudes

with black people that influence judicial decisions.

In explaining their findings on

cognitive illusions, Chris Guthrie et al. concluded that “[j]udges, it seems, are

human . . . [and subject to] cognitive illusions that can produce systematic errors in

judgment.”

As a result, implicit bias can result in the unequal, albeit unintentional,

treatment of negatively stereotyped group members. To see how this could play out

for a judge deciding a case, consider Linda Krieger et al.’s statement that

“stereotypes can bias decision making implicitly by skewing the manner in which

inherently ambiguous information about the stereotyped target is perceived,

characterized, attributed, encoded in and retrieved from memory, and used in social

judgment.”

Because of this, a decision-maker like a judge could apply a negative

stereotype to a person while still “believ[ing] that his judgment and resulting

decision were based entirely on legitimate nondiscriminatory reasons.”

Banaji and

Greenwald have established a correlation between implicit bias and prejudiced

behavior that substantiates Krieger et al.’s conclusion.

However, simply because jurists draw implicit associations and can

rationalize them after the fact does not mean that they act on a particular bias

consistently. In fact, Krieger et al., drawing on psychological research spanning over

fifty years, found that whether an individual acts according to bias is situational.

Krieger et al. explains that people operate under a certain “tension” as a result of

dual attitudes.

In other words, some perceptions, thoughts, and motivations pull

toward one course of action while others simultaneously pull toward another.

This

means that a person will react differently to her stereotypes depending upon

BENJAMIN N. CARDOZO, THE NATURE OF THE JUDICIAL PROCESS 167 (1921).

Chris Guthrie, Jeffrey J. Rachlinski & Andrew J. Wistrich, Inside the Judicial Mind, 86 CORNELL L. REV.

777, 778 (2001).

Jeffrey J. Rachlinski, Sheri Lynn Johnson, Andrew J. Wistrich & Chris Guthrie, Does Unconscious Racial

Bias Affect Trial Judges, 84 NOTRE DAME L. REV. 1195, 1210 (2009).

Guthrie et al., supra note 18, at 778.

Krieger & Fiske, supra note 15, at 1036.

Id.

BANAJI & GREENWALD, supra note 6, at 48–50.

Krieger & Fiske, supra note 15, at 1062.

Id. at 1040 (citation omitted).

Id. (citation omitted).

2021] Honesty Is the Best Policy 97

whether the implicit or explicit attitude is dominating her judgment while she is

forming an impression about the stereotype’s target.

Jerome Frank, a prominent legal realist, is said to have quipped that judges

base their decisions on “what [they] had for breakfast.”

This is not my claim.

Rather, I take the view, in Geyh’s words, that “[i]n a post-realist age . . . the best

empirical work to date shows that the decisions judges make cannot be divorced

from the judges who make them because judicial decision-making is subject to a

complex array of legal and extralegal influences.”

Of these extralegal influences, I

focus on implicit bias because it is harder to quantify and identify than explicit

bias,

its effect on the common law can be felt for generations, and, despite the

current and unequivocal social science research on implicit bias, we have yet to

address fully its effect on judges.

B. Wassell v. Adams: An Illustrative Case Study in Implicit Judicial Bias

To illustrate implicit bias’s effect on judicial decision making, we can look at

Judge Richard Posner’s opinion in Wassell v. Adams

—a 1989 case that is in many

first-year torts curriculums. The plaintiff was Susan Wassell, a twenty-one-year-old

woman who was visiting Chicago to see her fiancé graduate from basic training at a

nearby naval base.

While visiting, Susan was attacked and raped in her hotel

room.

Late one night, hearing a knock and thinking it might have been her fiancé,

Susan opened the door to a man outside who asked for a glass of water.

She got

the water and brought it back; the man, saying the water was not cold enough, then

went into the bathroom.

When he came out, he was naked from the waist down,

and Susan ran from the room.

The man chased Susan out, caught her, and

dragged her back into the room, where he raped her several times.

After several

hours Susan managed to escape.

The hotel was four blocks away from a high crime area, and the owners

occasionally warned female guests not to walk in the neighborhood at night by

themselves.

However, the owners did not warn Susan.

Susan sued the hotel for

Id. at 1041.

See, e.g., Gregory S. Parks, Judicial Recusal: Cognitive Biases and Racial Stereotyping, 18 LEGIS. & PUB.

POL’Y 681, 682 (2015).

Charles Gardner Geyh, Why Judicial Disqualification Matters - Again, 30 REV. LITIG. 671, 715 (2011).

Kang Fuchs Lecture, supra note 8.

865 F.2d 849 (7th Cir. 1989).

Id. at 850.

Id. at 851.

Id.

98 Indiana Journal of Law and Social Equality [9:94

negligence as a third party to her rape.

Under a comparative fault statute,

the

jury came back with the verdict that she was ninety-seven percent at fault for her

own rape, and the hotel owners only three percent at fault, thus reducing her

recovery.

Susan requested that the judge grant judgment for her notwithstanding

the verdict either because she was legally nonnegligent or the hotel owners had

wantonly disregarded her safety.

Alternatively, Susan requested a new trial

because the fault apportionment was contrary to the manifest weight of the

evidence.

The court denied both motions, and Susan appealed.

Upholding the verdict, Judge Posner, while not expressing any explicit

discourtesy,

rested on some concerning description and reasoning. For instance,

early in the analysis, Judge Posner identified the assailant as a “respectably

dressed black man.”

The man’s race (or his attire) does not appear to be relevant

to the law or facts implicated in the case. But, mentioning traits like race when not

otherwise relevant can be evidence of either explicit bias or the unconscious

expression of implicit bias.

On the other hand, if race was relevant to the court’s analysis, then its

mention is even more concerning, because in the context of a comparative fault

analysis, it suggests that the court’s view was that Susan should have known the

stranger was a threat because he was a black man. This does not appear to be

explicit bias because Judge Posner does not make any such connection in his

analysis. However, the unnecessary disclosure of the man’s race juxtaposed with the

assignment of Susan’s fault in her own rape points to implicit bias. It is especially

suggestive of implicit bias given the extant backdrop of the pervasive social-sexual

myth of black men raping white women,

and even more so when coupled with the

fact that rapes are substantially more likely to be intraracial than interracial.

Id. at 852.

Id.

There were times, in fact, when Judge Posner showed a sensitivity to Susan’s plight. For instance, he called

the defendant’s claim that the hotel could not post a warning because it would be too expensive, “absurd[].”

Id. at 853. In addition, when he wrote that the plaintiff pretended, post-rape, to like the rapist, he clearly

stated that she did it “[t]o save herself” and refrained from suggesting that this cast doubt on her rape

assertions. Id. at 851.

Id.

See MODEL CODE OF JUDICIAL CONDUCT r. 2.3 cmt. 2 (AM. BAR ASS’N 2010).

See Kimberle Crenshaw, Demarginalizing the Intersection of Race and Sex: A Black Feminist Critique of

Antidiscrimination Doctrine, Feminist Theory and Antiracist Politics, 1989 U

NIV. CHI. LEGAL FORUM, no. 1,

at 159 n.54. Although Judge Posner does not disclose Susan’s race, she was white. Amy H. Kastely, Out of

the Whiteness: On Raced Codes and White Race Consciousness in Some Tort, Criminal, and Contract Law,

63 U. CIN. L. REV. 269, 283 & n.52 (1994).

From 1973–82, reported rapes with a black offender and white victim represented twenty-three percent of

rapes. P

ATSY A. KLAUS & MARSHALL DEBERRY, BUREAU OF JUSTICE STATISTICS, U.S. DEP’T OF JUSTICE, PROD.

NO. NCJ 96777, THE CRIME OF RAPE 3 (1985). The reporting guidelines changed in the 1980s and later

research found that intraracial rape accounted for about eighty-eight percent of rapes. LAWRENCE A.

2021] Honesty Is the Best Policy 99

Finally, Judge Posner upheld the jury verdict in spite of, and without even a nod to,

the fact that using comparative fault in rape litigation is painfully resonant with

the age old victim blaming epitaph that “she was asking for it.”

While these

implications each arguably fall short of overt bias, their cumulative resonance with

racism and misogyny nonetheless suggests a type of bias at work “deep below

consciousness.”

Of course, the appellate judge’s role is often to review for abuse of

discretion.

As a result, the easy explanation is that Judge Posner was simply

observing his proper role by deferring to the trial court and jury verdict. However,

Judge Posner is not shy about either correcting an application of law or pointing out

that he is limited to his role, even though he may not like it.

In fact, there is one

point in Wassell that Judge Posner is uncomfortable with, but his discomfort is

focused exclusively on whether the jury got the percentage of fault correct because

the jury asked a question about the hotel’s duty to warn that went unanswered.

But why is there no concern for the gender and racially problematic aspects

of the opinion or its applicable law? Judge Posner could have broken new ground in

third-party rape suits by applying a different standard.

For instance, one

alternative is the no-duty approach, which is to “recognize important differences in

the situation of the two actors that justify treating them differently.”

Under this

approach, courts would: (1) “see a difference in failing to prevent injury to oneself as

opposed to others,” (2) “recognize the different kind of costs that each actor pays for

not being more cautious or prudent,” and (3) “decide whether reducing or barring

recovery for rape victims will likely serve the social policy of curtailing high rates of

sexual assaults.”

Of course, it would have required innovation to institute a no-duty exception

in comparative fault doctrine for third-party rape suits. However, Judge Posner has

written innovative opinions on many other matters,

and innovation is consistent

GREENFELD, BUREAU OF JUSTICE STATISTICS, U.S. DEP’T OF JUSTICE, PROD. NO. NCJ-163392, SEX OFFENSES

AND

OFFENDERS: AN ANALYSIS OF DATA ON RAPE AND SEXUAL ASSAULT 11 (1997).

See Martha Chamallas, Symposium: Who Is the Reasonable Person? Gaining Some Perspective in Tort Law:

A New Take on Third-Party Criminal Attack Cases, 14 L

EWIS & CLARK L. REV. 1351, 1381 (2010); Leah M.

Slyder, Note, Rape in the Civil and Administrative Contexts: Proposed Solutions to Problems in Tort Cases

Brought by Rape Survivors, 68 CASE W. RES. 543, 567–569 (2017).

CARDOZO, supra note 17.

Wassell v. Adams, 865 F.2d 849, 856 (7th Cir. 1989).

See State Oil Co. v. Khan, 93 F.3d 1358 (7th Cir. 1996) (following antitrust precedent even though he

referred to it as “moth-eaten,” “wobbly,” and “unsound”); Ind. Harbor Belt R.R. Co. v. Am. Cyanamid Co.,

916 F.2d 1174 (7th Cir. 1990) (lowering the standard of legal liability a railroad company faced for a

chemical spill); George M. Cohen, Posnerian Jurisprudence and Economic Analysis of Law, 133 U. PA. L.

REV. 1117, 1117 (1985) (discussing Judge Posner’s “innovative methodological approach”).

See Wassell, 865 F.2d at 856.

A third-party rape case is a subset of tort cases where a plaintiff sues a third party to an attack for

permitting dangerous conditions or failing to take other reasonable precautions to prevent such an attack.

Chamallas, supra note 52, at 1354.

Chamallas, supra note 52, at 1383 (citing Ellen M. Bublick, Citizen No-Duty Rules: Rape Victims and

Comparative Fault, 99 COLUM. L. REV. 1413 (1999)).

Id. (citation omitted).

Cohen, supra note 55.

100 Indiana Journal of Law and Social Equality [9:94

with how he views the judge’s role.

In his words, a flexible, innovative approach is

required because the “multi-layered character of American law . . . thrust[s] on the

courts a responsibility for creative lawmaking that cannot be discharged either by

applying existing rules to the letter or by reasoning by analogy.”

And yet, in

Wassell this approach is notably absent. Judge Posner did not acknowledge the

policy implications of allowing courts to find victims responsible for their own rape,

express personal distaste for the outcome or implication, mention misogyny or

racism’s possible effect on the jury, or discuss whether finding Susan ninety-seven

percent responsible for her own rape could possibly vindicate her individual right to

redress.

Based on these analytical gaps, and Judge Posner’s view of rape published in

Sex and Reason in 1992,

I posit that Judge Posner did not see the role that

misogyny and racism played in allowing Wassell’s jury to find the white plaintiff

ninety-seven percent responsible for her own rape by a black man—or the common

law rule that permits victim fault allocation in third-party rape suits in the first

place. After Sex and Reason’s publication, feminist scholars struggled with Judge

Posner’s rape discussion, particularly his conclusion that rape is “primarily a

substitute for consensual sexual intercourse rather than a manifestation of male

hostility toward women or a method of establishing or maintaining male

domination.”

One critic paraphrased that in this view, rape and other acts of

subordination are not about subordination at all, but are merely efficient practices

based on “successful adaptations to biological market conditions.”

Excavating an old opinion through literary analysis is not proof that Judge

Posner wrote Wassell under the influence of unacknowledged bias.

Rather, I use it

here for illustrative value. Judge Posner was a fiercely academic, rational,

and

RICHARD A. POSNER, OVERCOMING LAW 12–13 (1995).

Id. (emphasis added).

While Wassell presents a rape involving a black man and a white woman, under a theory of

intersectionality, black women may have a different experience than either black men or white women,

particularly when it comes to racism, feminism, misogyny, and rape. See Crenshaw, supra note 50, at 139,

140, 157–160.

RICHARD A. POSNER, SEX AND REASON (1992).

Id. at 384; see, e.g., Robin West, Review Essay: Sex, Reason, and a Taste for the Absurd, 81 GEO. L. J. 2413,

2440–41 (1993). Naturally, research available at the time of Sex and Reason’s publication, and since then,

suggests that the motivation to rape is much more complicated. See A. Nicholas Groth, Ann Wolbert

Burgess & Lynda Lytle Holmstrom, Rape: Power, Anger, and Sexuality, 134 A

M. J. PSYCHIATRY 1239, 1239

(1977) (finding that interviewers of 133 convicted rapists concluded that their crimes “were not rapes in

which sex was the dominant issue; sexuality was always in the service of other, nonsexual needs”);

Crenshaw, supra note 50, at 158 (critiquing the exclusive focus on rape as an expression of male power over

women while identifying the rape of black women as part of a racially motivated terror campaign); Heather

Murphy, What Experts Know About Men Who Rape, N.Y.

TIMES (Oct. 20, 2017),

https://www.nytimes.com/2017/10/30/health/men-rape-sexual-assault.html (identifying that while exclusive

focus on convicted rapists may skew data, researchers continue to find that hostility to women and power

are relevant factors in explaining rape).

West, supra note 65, at 2445–46.

See Bruce A. Green, Legal Discourse and Racial Justice: The Urge to Cry “Bias!,” 28 GEO. J. LEGAL ETHICS

177, 202 (2015) (stating that no social scientist would base a conclusion on one data point).

See Michael B. Brennan, Book Review, 79 MARQ. L. REV. 329, 329 (1995).

2021] Honesty Is the Best Policy 101

dedicated jurist. He was not likely to say and do things that were overtly racist or

misogynistic, and there is no reason to suspect a history of disciplinary proceedings

for improper conduct toward women and racial minorities.

But if elements of the

Wassell opinion can be explained by implicit bias and a person as academic and

rational as Judge Posner can inadvertently manifest such bias, then without doubt,

any jurist is susceptible.

In state trial courts, examples of possible implicit bias play out in the records

of disciplinary proceedings. In In re Jordan, a disciplinary commission sanctioned a

judge for twice referring to an attorney appearing before him as “little girl.”

that case, the commission focused on the second time the judge said “little girl”

because it found that he used the phrase intentionally to demean the attorney.

primary interest here, however, was the commission’s conclusion that the first use

of “little girl” may have been “an inadvertent expression of unconscious prejudice.”

This is evidence that disciplinary commissions at least tacitly recognize that

implicit bias is qualitatively different than explicit bias and are hesitant to

discipline it in the same way.

While in In re Jordan, the bias’s target was an attorney, there are also many

instances where trial court judges seemingly exhibit implicit bias toward litigants.

For instance, a New York trial judge, in a case dealing with international students,

told a prosecutor “you better deport these people” and “you better get them out.”

The disciplinary commission waffled about the words’ import.

Because it wasn’t “a

continuous pattern of . . . conduct” they were unwilling to conclude that bias was at

work and hedged their language about what “appear[ed] to indicate distrust and

dislike of all those from outside his community.”

However, distrust or dislike is

precisely the kind of passive antipathy that implicit association researchers have

identified and we are concerned with here.

II. WHY IMPLICIT JUDICIAL BIAS MATTERS

Given the strength of contemporary social science research on implicit bias, it

is clear that unconscious bias exists, and the judiciary is not immune from it. The

question, however, is does it matter. I argue that bias in the judiciary does indeed

matter because judges play a substantial role in a common-law system, because of

normative concern for legitimacy and fairness, and because of the related practical

concern for compliance and procedural justice.

While complaints issued against federal judges are not made immediately public, substantiated complaints

can lead to public censure and other sanctions that may become public knowledge. See 28 U.S.C §§ 351–55

(2018).

GEYH ET AL., supra note 3, at 3-37.

Id.

Id. at 3-26.

Id.

Id. (emphasis added).

See infra text accompanying notes 85–87.

102 Indiana Journal of Law and Social Equality [9:94

In a common-law system, judges have the unique role of applying legal

precedent to new facts. As a result, the common law is dynamic, and judges have

significant power in shaping it.

While this is true for judges presiding over courts

with both original and appellate jurisdiction, it is perhaps especially true for the

appellate courts. According to Judge Roger Traynor, the “intellectual process of

appellate judgment, [is] of the greatest public importance.”

At the same time, the legitimacy of government institutions and actors is a

primary concern because of a representative democracy’s unique theoretical

underpinnings

and because of the “wondrous” power that legitimacy has to

“convince[] losers in law and politics to accept their losses.”

As a result, it is

unsurprising that in our governmental system, the judiciary’s role is consistently

subject to legitimacy concerns,

many of which are tied to impartiality, both actual

and its appearance.

Because unchecked bias both directly reduces a judge’s ability

to decide impartially

and gives the appearance of partiality,

it stands as a

legitimacy measure.

One of implicit bias’s dangers is that it can consistently tip justice’s scales to the favor

of one group and the disfavor of another. This is the case even though scholars generally

think that people express explicit bias less now than at previous points.

For instance,

“rather than antipathy, many now show ambivalence and avoid members of stereotyped

See, e.g., VICTOR E. SCHWARTZ, KATHRYN KELLY & DAVID F. PARTLETT, PROSSER WADE AND SCHWARTZ’S TORTS:

CASES AND MATERIALS 1 (13th ed. 2015) (establishing that torts have largely been the creation of judicial

opinions).

Roger J. Traynor, Some Open Questions on the Work of State Appellate Courts, 24 U. CHI. L. REV. 211, 212

(1957).

In a representative democracy, the government should reflect the will of the people. See ROBERT NOZICK,

ANARCHY, STATE, AND UTOPIA 290 (1974). For this governmental system to have any meaning, the people

must understand the institutions of government to be acting legitimately, that is to say, discharging the

people’s will. See id. at 134 (stating that at the very least, “a legitimate government is one that most of its

subjects view as legitimately ruling.”).

JAMES L. GIBSON, ELECTING JUDGES: THE SURPRISING EFFECTS OF CAMPAIGNING ON JUDICIAL LEGITIMACY 130

(2012).

See Charles Gardner Geyh, Preserving Public Confidence in the Courts in an Age of Individual Rights and

Public Skepticism, in BENCH PRESS: THE COLLISION OF COURTS, POLITICS, AND THE MEDIA 21, 22 (Keith J.

Bybee ed., 2007).

See id. at 24, 30–31.

See supra Part I.

See John Sides, White People Believe the Justice System is Color Blind. Black People Really Don’t., WASH.

POST (July 22, 2013, 12:20 PM), https://www.washingtonpost.com/news/wonk/wp/2013/07/22/white-people-

believe-the-justice-system-is-color-blind-black-people-really-dont/?utm_term=.7f330f928deb (reporting that

sixty percent of African Americans disagreed with the statement that “courts give all a fair trial” and that

perceptions of procedural injustice informed these attitudes).

See, e.g., Victor D. Quintanilla, Beyond Common Sense: A Social Psychological Study of Iqbal's Effect on

Claims of Race Discrimination, 17 MICH. J. RACE & L. 1 (2011), at 17–18. According to Quintanilla, “[w]hile

blatant racism has waned, prejudice still persists . . . the field of social psychology has shown a continuing

divide between words and deeds toward people of color.” However, we should not construe this to mean that

all bias is subtle. See Joan C. Williams, Double Jeopardy? An Empirical Study with Implications for the

Debates over Implicit Bias and Intersectionality, 37 HARV. J.L. & GENDER 185, 188 (reporting that while

some gender bias was subtle “much was not subtle at all”). In addition, as Williams points out, once people

recognize a pattern of bias, even implicit bias is then easy to spot. See id. at 227.

2021] Honesty Is the Best Policy 103

groups . . . resulting in decreased helping behavior and cooperation, passive harm, and

neglect.”

At the same time, “many exhibit liking and trust toward in-group members,

resulting in facilitation and cooperation with other majority group members.”

In a complex evaluative process, these mental errors can result in courts chronically

undervaluing out-group members, such as women and racial minorities, and chronically

overvaluing in-group members, such as white men. In Wassell, how can we explain the

possibility that Judge Posner just did not assign value to irrelevant racial descriptions, the

negative policy implications or plain injustice of holding a victim ninety-seven percent

responsible for her own rape, or to racism and misogyny’s effect of legitimizing and

encouraging human suffering? It could be that Judge Posner was simply unable to mentally

recognize that his opinion implicated issues of racial and gender equality;

he instinctively

valued them less than some other principles or policy outcomes,

or both.

There is good reason to think that implicit bias is especially likely to cause disparate

valuing when there is ambiguity in the law or a particularly close question. Despite implicit

racial prejudices, people are less likely to discriminate when there is a clear choice that

would associate them with biased attitudes and behaviors versus one that would not.

However, in the law, there are many instances when there are no clear choices. Karl

Llewellyn illustrates the point when he says that courts can decide a question, correctly, in

twenty-six ways and goes onto say “the major defect in [our] system is a mistaken idea . . .

that the cases . . . plus the correct rules . . . provide one single correct answer to a disputed

issue of law.”

Instead, Llewellyn says, “the available correct answers are two, three, or ten”

and the salient question is “[w]hich of the available correct answers will the court select—

and why?”

Because judges in a common law system have considerable discretion in

applying the law, this opens the door to bias. Llewellyn describes the “felt need” and

“felt sense” that judges use in their decision-making

and instincts are undoubtedly

as essential to the judicial profession as they are in any other endeavor. However, it

is very likely that in ambiguous cases, the type of split-second subconscious

categorization described by implicit bias research informs “felt” needs and sense.

Quintanilla, supra note 85, at 18.

Id.

See supra discussion accompanying notes 50–56, 61–63.

Judge Posner’s self-professed take on feminism is suggestive here:

[L]aw or government should [not] prescribe a particular role for women or discourage them

from exercising free choice regarding occupation, marriage, and style of life. Nor, however,

do[es] [he] believe that women should be put ahead of men or encouraged to lead separate

lives from men, or that ‘what’s good for women’ should be the lodestar for social governance

instead of ‘what’s good for the United States’ or ‘what’s good for humankind.’

Richard A. Posner, Conservative Feminism, 1989 U.

CHI. LEGAL F. 191, 191.

Quintanilla, supra note 85, at 20 (citing John F. Dovidio & Samuel L. Gaertner, Aversive Racism, 36

ADVANCES EXPERIMENTAL SOC. PSYCHOL. 1, 4–7 (2004)).

Karl N. Llewellyn, Remarks on the Theory of Appellate Decision and the Rules or Canons About How

Statutes Are to Be Construed, 3 VAND. L. REV. 395, 396 (1950).

Id. (emphasis omitted).

Id. at 398.

See supra Part I.

104 Indiana Journal of Law and Social Equality [9:94

In addition, information-rich environments or even the semblance of information

give people adequate cover to act on bias.

Because judges often have substantial

information supplied by the record, and, as both Llewellyn and Judge Posner

suggest, law’s application to facts is often ambiguous,

judicial decision-making is

fertile ground for implicit bias.

Judge Posner recognized the relationship between ambiguity and bias in

McReynolds v. Merrill Lynch.

In describing corporate practices in a discrimination

case, he said that the white brokers in question “are more comfortable teaming with

other white brokers” but that “they would doubtless ask a superstar broker to join

their team regardless of his or her race.”

Where there is uncertainty, however,

Judge Posner said, “people tend to base decisions on emotions and

preconceptions.”

To allow the law to consistently favor an in-group while consistently

disfavoring an out-group is fundamentally unfair. Fairness is central to judicial

impartiality and, not surprisingly, Merriam-Webster defines fair as “marked by

impartiality.”

100

In the first place, we should care about unchecked implicit bias’s

unfair effects on judicial decision-making for normative reasons. Our sense of

fairness helps us intuitively gauge how people treat us against how we want them

to treat us. In Llewellyn’s words, fairness is part of our “felt sense,” and there is

significant collective recognition that we should treat others as we would like to be

treated.

101

However, fairness, like legitimacy, is not just a normative concept to aspire to

for its own sake. For the rule of law to provide order, there must be compliance.

102

Procedural justice, resting on perceptions of both fairness and legitimacy, helps

explain why people comply with the law despite adverse outcomes.

103

To do so,

procedural justice captures how individuals subjectively experience the legal

process.

104

Researchers have found that people care both about outcome and,

Krieger & Fiske, supra note 15, at 1050–51.

See supra text accompanying notes 60, 91–92.

McReynolds v. Merrill Lynch, 672 F.3d 482 (7th Cir. 2012).

Id. at 489.

Id. One question is how Judge Posner could recognize this preference in McReynolds but not in Wassell.

One obvious difference is that the opinions were written twenty-eight years apart. Another possible

explanation is the “bias blindspot,” which is the name given to individuals’ tendency to think they are more

objective than others. See Kang et al., supra note 11, at 1173.

100

Fair, MERRIAM WEBSTER’S COLLEGIATE DICTIONARY (11th ed. 2003).

101

Lisa Capretto, Scholar Points to One Thing Nearly All Major Religions Have in Common, HUFFPOST (Aug.

17, 2016), https://www.huffingtonpost.com/entry/principle-in-nearly-all-major-

religions_us_57b370a1e4b0edfa80d9e0e6 (identifying the Golden Rule as a principle common to most of the

world’s major religions).

102

RICHARD H. MCADAMS, THE EXPRESSIVE POWERS OF LAW: THEORIES AND LIMITS 3 (2015).

103

Tom Tyler, What is Procedural Justice?: Criteria Used by Citizens to Assess the Fairness of Legal

Procedures, 22 L. & SOC’Y REV. 103, 103 (1988).

104

Rebecca Hollander-Blumoff, The Psychology of Procedural Justice in the Federal Courts, 63 HASTINGS L.J.

127, 134 (2011).

2021] Honesty Is the Best Policy 105

independently of outcome, the fairness of the process.

105

Based on this, procedural

justice scholars conclude that how an individual perceives the legal process’ fairness

is “critical to assessments of legitimacy and deference to legal authority.”

106

Our perceptions of fairness rest on unbiased, impartial decision makers. Tom

Tyler has identified seven factors that determine whether citizens feel judges acted

fairly.

107

Two factors are relevant to actions taken by the parties: the chance to have

representation and the opportunity to appeal. Five of the seven factors directly

pertain to the judge. They are the judge’s (1) effort at fairness, (2) honesty, (3)

ethical conduct, (4) quality of decisions, and (5) bias.

108

Tyler not only enumerates

bias as a stand-alone concern, but it is also implicit in the other four judicial

conduct factors. Thus, bias is highly relevant to legitimacy and compliance.

Moreover, litigation’s very purposes include other practical concerns that

raise the threat of lawlessness. Its purpose is largely to avoid vigilantism, act as a

deterrent, encourage desirable social conduct, make the injured person whole, and

provide justice for an individual’s right to redress.

109

Judicial decision-making

thwarts these purposes if subconscious undervaluing is taking place during a

judge’s evaluation process. We can easily imagine if a court undervalues certain

people’s claims, it will be unable to make them whole or fully validate their

individual rights to redress. At the same time, if the court overvalues other people’s

claims, it is less likely to deter them completely or acknowledge how socially

undesirable their conduct is. The disparity between one group’s lack of redress and

recovery against another group’s diminished deterrence and limited social sanction

can only diminish the legitimacy and power of our legal system . Simply put,

unchecked implicit bias is not only unfair—it threatens the rule of law.

III. BIAS PROHIBITIONS IN THE CODE OF CONDUCT AND MODEL CODE

The ethical rules in both the Code of Conduct and the Model Code currently

contain prohibitions against bias. The Code of Conduct’s Canon 3(A)(3) specifies

that “[t]he duty to be respectful includes the responsibility to avoid comment or

behavior that could reasonably be interpreted as harassment, prejudice or bias.”

110

Similarly, the Model Code’s rule 2.8 prohibits undignified and discourteous

behavior

111

and rule 2.3 insists that “[a] judge shall perform the duties of judicial

office, including administrative duties, without bias or prejudice.”

112

Examples of

discourteous conduct that have given rise to discipline are verbally abusing and

intimidating parties in a courtroom, subjecting people to unwarranted searches,

105

Id. at 132.

106

See, e.g., id.

107

See Tyler, supra note 103.

108

Id.

109

See SCHWARTZ ET AL., supra note 77.

110

CODE OF CONDUCT FOR UNITED STATES JUDGES Canon 3(A)(3) (JUDICIAL CONF. 2014).

111

MODEL CODE OF JUDICIAL CONDUCT r.2.3, 2.8 (AM. BAR ASS’N 2010).

112

Id.

106 Indiana Journal of Law and Social Equality [9:94

criticizing jurors after a verdict, and using profanity in the courtroom.

113

Sanctions

for these acts include public reprimands, public reproval, suspension, removal,

etc.

114

Examples of racial bias that have given rise to disciplinary proceedings

include using racial epithets, making racial “jokes,” referring to certain crimes in

racial terms, stereotyping racial groups, and using racially charged language.

115

Examples of gender bias include using gender loaded language and insults to

disparage, making inappropriate comments during domestic abuse cases, referring

to women as children, pontificating that a woman’s place is in the home, calling

women terms of endearment in a professional context, and commenting on women’s

appearance.

116

Sanctions in bias proceedings are similar to those resulting from

discourtesy and include reprimands, public reproval, suspension, and removal.

117

In these examples, there are two common themes: 1) the conduct in question

is overt, and 2) the biases most readily addressed tend to be explicit, or at least

perceived to be explicit.

118

While the current rules are effective when used to correct

overt conduct that expresses explicit bias, they are a poor fit for addressing implicit

bias. For instance, the prohibitions against bias and discourtesy do not easily reach

a Wassell-type situation. Recall that in Wassell, Judge Posner was not discourteous

and did not engage in any conduct that was clearly prejudiced toward women or

racial minorities.

119

Where a judge’s implicit associations may surreptitiously allow

his brain to value women and black men less, fail to recognize racism at play in

irrelevant references to skin color, or refuse to acknowledge the potential for

misogyny and racism to inform a legal rule or jury verdict there is a gossamer-like

quality that makes comparisons with using the “n-word” in open court

problematic.

120

Moreover, much of the Model Code and the Code of Conduct’s concern for

impartiality is focused on conflicts, not on bias per se.

121

For instance, in Caperton

v. A.T. Massey Coal Co., the Court used the word bias often,

122

but its context was

violation of a disqualification rule similar to Model Code rule 2.11.

123

And the issue

in the case was whether a West Virginia Supreme Court judge had a conflict in a

113

GEYH ET AL., supra note 3, at 3-3 n.5.

114

Id.

115

Id. at 3-23 to 3-28.

116

Id. at 3-36 to 3-41.

117

See id. at 3-23 to 3-28.

118

See Alex Kozinski, The Real Issues of Judicial Ethics, 32 HOFSTRA L. REV. 1095, 1095 (2004) (“The canons

focus on the tensions and potential conflicts that are most easily detected by an outside observer.”).

119

For instance, Judge Posner was seemingly sympathetic to Susan’s efforts to do whatever she could to get

away from her attacker. See supra note 47. Additionally, while “respectably dressed black man” is certainly

suspect in the context of the opinion, the language itself is not inherently prejudiced. See supra text

accompanying notes 48–52.

120

Compare supra Section I.B. with supra text accompanying notes 113–16.

121

See Green, supra note 67, at 183–84.

122

See Caperton v. A.T. Massey Coal Co., 556 U.S. 868 passim (2009).

123

“Bias” is used similarly in the disqualification rule for federal judges. See CODE OF CONDUCT FOR UNITED

STATES JUDGES Canon 3(C) (JUDICIAL CONF. 2014).

2021] Honesty Is the Best Policy 107

lawsuit before his court because one of the litigants made very large contributions

to the judge’s campaign.

124

As the Supreme Court stated in Caperton,

disqualification is normally the

redress for a conflict that gives the appearance or has the actual result of impeding

a judge’s impartiality.

125

However, disqualification is not a viable option for cases

involving implicit bias. Because all people subconsciously categorize others, we

would have no judges left to adjudicate if we required judges to disqualify every

time their brains engaged in an implicit association.

126

Finally, while social science research splits biases into the neat categories of

explicit and implicit, it also recognizes that people may not always report their own

biases or may attempt to hide them.

127

As a result, we should be careful about

assuming that all instances of judicial bias will be easily identifiable as implicit or

explicit, fit in a neat analytical framework, or be addressed mechanically through

the measures proposed in Part IV. As In re Jordan suggests, it is possible for a

judge to simultaneously express an implicit bias alongside an explicit bias,

128

and it

may be hard to distinguish the two. We should also assume that judges are

sometimes able to hide their explicit biases or might be able to use implicit bias as

cover for what is in fact explicitly biased behavior. My goal is not to give biased

judges a get out of jail free card, but rather to curb true implicit associations when

these associations are married to a judge’s good faith commitment to deliver

impartial justice. I do not intend my analysis to reach judges who deliberately hide

their equally deliberate bias.

For the bulk of jurists who are acting in good faith and are subject to implicit

associations, the need for tailored guidelines is especially pressing because scholars

have noticed that while overt racism has waned, “the manifestation of prejudice has

grown more subtle.”

129

As a result, while the bias prohibitions in the Model Code

and the Code of Conduct generally do what they were intended to do, we need

additional mechanisms to continue working toward our aspiration of an impartial

judiciary.

IV. BEYOND THE STATUS QUO: NOT LETTING PERFECT GET IN THE WAY OF GOOD

Before turning to the recommendations, it is worth noting a subtle difference

between the purpose of the Model Rules and Code of Conduct provisions just

discussed and the measures I propose. The guides to professional ethics exist to

govern misconduct.

130

Because implicit bias is a split-second neurological response

124

See Caperton, 556 U.S. at 872.

125

See id. at 888–89.

126

See Green, supra note 67, at 184.

127

See supra text accompanying notes 24–27; see also infra text accompanying notes 159–60.

128

See supra text accompanying notes 70–72.

129

Quintanilla, supra note 85, at 17.

130

See GEYH ET AL., supra note 3, at ix.

108 Indiana Journal of Law and Social Equality [9:94

that occurs unconsciously,

131

we should be slow to term it misconduct by rote and

automatically discipline it as such for two reasons. First, the culpability for

manifesting implicit bias as compared to explicit bias is less. How can a person be

morally responsible in any meaningful sense when she does not know what she is

doing?

132

Second, we intend ethical rules and their corresponding disciplinary

processes to act as deterrents.

133

Because individuals are not initially in the driver’s

seat when it comes to their own implicit bias, ethical rules and discipline will be less

effective as deterrents.

However, just because implicit judicial bias is unintentional does not mean it

cannot cause grievous injustice. Sentencing’s racially disparate effect, especially

death sentences, suggest just how immoral the result of unchecked implicit bias can

be.

134

While we may not have initial control as individuals over the quick

categorization that our brain goes through when faced with different types of

people,

135

collectively we have control over what we do from there—and willful

ignorance is not a sufficient excuse. With that in mind, the Model Code and Code of

Conduct should incorporate the methods below with their other aspirations and

expectations for judicial conduct. With these expectations in place, disciplinary

committees could sanction a judge for not engaging proactively in curbing her

implicit bias but not for having an unconscious bias to begin with.

The question is whether, since implicit bias occurs in an unconscious process,

there is anything we can do about it. The research is clear that our implicit

associations do not always lead to biased decision-making.

136

Surveying social

science research on behavior across many decades, Krieger et al. concluded that

“whether attitudes predict behavior depends on the attitude, the context, and the

person.”

137

In examining the correlation between bias and behavior, there are two key

factors that influence when they will be closely correlated: motive and opportunity

to deliberate.

138

When motive and opportunity are present, attitudes and behavior

correlate less.

139

Krieger et al. build out these two factors into the saliency of social

131

See supra text accompanying notes 8–12.

132

In analyzing the purposes underlying our criminal law, arguably our legal system’s most punitive area,

Stuntz and Hoffman identify three overlapping purposes, including retribution. WILLIAM J. STUNTZ &

JOSEPH L. HOFFMAN, DEFINING CRIMES 11 (2nd ed. 2014). For the purpose of retribution, “punishment . . . is

a moral good because crime is a moral wrong.” Id. We can analogize between crime and judicial misconduct

here to the extent they are both socially undesirable activities that cause harm.

133

See Randall T. Shepard, Foreword to CHARLES GARDNER GEYH, JAMES J. ALFINI, STEVEN LUBET & JEFFREY M.

SHAMAN, JUDICIAL CONDUCT AND ETHICS, at v-vi (2013) (stating that while the rare judge might be a “crook”

many others are trying to discharge the duties of their office faithfully and use the ethical rules as a

roadmap to handle avoidable pitfalls with “greater ethical sensitivity”).

134

See Kang et al., supra note 11, at 1148.

135

See supra Section I.A.

136

See Krieger & Fiske, supra note 15, at 1049.

137

Id.

138

Russell H. Fazio & Michael A. Olson, Implicit Measures in Social Cognition Research: Their Meaning and

Use, 54 A

NN. REV. PSYCHOL. 297, 304 (2003).

139

Id.

2021] Honesty Is the Best Policy 109

norms, the decision-maker’s perceived sense of control, the decision-maker’s desire

to avoid the bias, and relevant information.

140

The checks proposed below

incorporate these concepts and, because the research shows that these factors

predict the greatest variance when each is present,

141

my solutions will be most

effective when used together. While the proposed solutions are not enough to root

out all implicit bias, as this section’s title suggests, my goal is to provide a

framework for taking a meaningful step forward. I focus on using systems that are

already largely in place in order to minimize the force of objections to moving

forward.

The first question is how to provide judges with evidence of their own implicit

bias so that they will be motivated to address it. Kang et al. suggest that one step is

having judges take the IAT just like other individuals and professional groups do.

142

In order to promote participation and avoid unnecessary shaming, the assessments

should be confidential with respect to individual judges, but the aggregated data

over time should be collected by court administrators in order to target and refine

efforts. Kang et. al. also suggest judicial training on cognitive illusions, the

fallibility of heuristics, and how to become sensitized to biases similar to the type

used in other professional settings.

143

Most state court systems can build this into

their current continuing legal education requirements

144

so that they require

implicit bias training hours each year. In fact, Kang has already supplied a primer

on implicit bias for courts in partnership with the National Center for State

Courts.

145

While there are optional training programs for the federal judiciary, there is

no mandated training,

146

so it would take an extra step to implement this

suggestion for the federal bench. The Federal Judicial Center is the body charged

with providing training to federal courts and it currently offers workshops on

implicit bias.

147

For both state and federal courts then, access to adequate training

materials is unlikely to be an issue.

148

Because annual judicial conferences already

140

Krieger & Fiske, supra note 15, at 1050.

141

See Fazio & Olsen, supra note 138; see also Alexandra Kalev, Erin Kelly & Frank Dobbin, Best Practices or

Best Guesses? Assessing the Efficacy of Corporate Affirmative Action and Diversity Policies, 71 AM. SOC. REV.

589, 589 (2006) (finding that diversity measures were most effective when training, assigned responsibility,

and steps to reduce minority isolation were employed together).

142

Kang et al., supra note 11, at 1175–77.

143

Id.

144

Forty states have continuing education requirements. Legal Education/Bar Admission Resource Guide,

NAT’L CTR. FOR ST. CTS., https://ncsc.contentdm.oclc.org/digital/collection/legserv/id/86 (last updated May 16,

2019).

145

JERRY KANG, NAT’L CTR. FOR STATE COURTS, IMPLICIT BIAS: A PRIMER FOR COURTS (2009),

https://cdm16501.contentdm.oclc.org/digital/collection/accessfair/id/248.

146

S.I. Strong, Symposium: Judicial Education and Regulatory Capture, 2015 J. DISP. RESOL. 1, 14 (noting that

judges “are largely in control of their own educational agendas”).

147

Programs and Resources for Judges, FED. JUD. CTR., https://www.fjc.gov/education/programs-and-resources-

judges#SF (last visited Sept. 1, 2020).

148

In addition to the other resources mentioned, the ABA has a resource page dedicated to implicit bias

training for attorneys and judges. Diversity and Inclusion 360 Commission, A

M. BAR ASS’N,

https://www.americanbar.org/diversity-portal/diversity-inclusion-360-commission.html (last visited Sept. 1,

110 Indiana Journal of Law and Social Equality [9:94

exist and judges typically attend them,

149

the easiest way to administer this

requirement may be to push the training out in partnership with these conferences.

However, the key is to require the education because if an external body does

not require it, judges are less likely to opt into this training. A mandate helps

establish norms and contributes to motivation.

150

In addition, people who are

uncertain in their knowledge are likelier to make more effective educational

choices.

151

While a confident judge is undoubtedly helpful in many contexts, because

people are unable to perceive their own implicit bias,

152

and tend to think that they

are less biased than others,

153

a lack of uncertainty is especially problematic in this

context. By requiring the training, judges are not left to self-diagnose what social

scientists know is a typical blind spot.

The question is then which external body should mandate implicit bias

training for federal judges. The answer is not the legislature. While there are some

areas in which the legislature must exercise judicial oversight,

154

opening up this

sensitive and nascent oversight role to the politics between the federal judiciary and

Congress and the politics between Congress and its constituents,

could affect the

judiciary’s independence and perceptions of its legitimacy,

155

would be excruciating

to watch,

156

and would almost certainly result in the false and unfair cry of “Bias!”

of such concern to Bruce Green.

157

With this in mind, the best suggestion is for a

dedicated Judicial Conference committee to work with the chief judges to determine

the most effective way to execute and enforce a training requirement.

Accountability and transparency are essential to successful efforts to

counteract implicit bias because of the role of social norms in negating the

2020). While the efficacy of these trainings is largely untested, there are at least some interventions

workshops that, while only tested in small numbers, are showing positive effects. See Jessica Nordell, Is

This How Discrimination Ends?, A

TLANTIC (May 7, 2017),

https://www.theatlantic.com/science/archive/2017/05/unconscious-bias-training/525405/ (reporting on an

implicit bias workshop at the University of Wisconsin).

149

See, e.g., JUDICIAL COUNCIL OF THE NINTH CIRCUIT, NINTH CIRCUIT UNITED STATES COURTS 2009 ANNUAL

REPORT 22 (2009), https://www.ca9.uscourts.gov/judicial_council/publications/AnnualReport2009.pdf.

150

See infra notes 158–63 and accompanying text.

151

Strong, supra note 146.

152

See supra text accompanying notes 8–9; Strong, supra note 146, at 15.

153

See Kang et al., supra note 11, at 1173.

154

For instance, Congress is the only body that can establish the lower federal courts. See U.S. CONST. art. III,

§ 1.

155

See CHARLES GARDNER GEYH, WHEN COURTS AND CONGRESS COLLIDE 280–82 (2008).

156

For instance, consider the political grandstanding on both sides of the aisle accompanying the recent

Kavanaugh nomination hearings. See Cavett Feazel, Kavanaugh’s Confirmation Hearings a Display of

Political Grandstanding, CIVILIAN (Oct. 9, 2018), https://sites.law.lsu.edu/civilian/2018/10/kavanaughs-

confirmation-hearings-a-display-of-political-grandstanding-september/.

157

Because implicit bias affects almost everyone but is hard to determine on a case-by-case basis, it would be

hard to prove an implicit bias accusation, but also equally hard to disprove. See Green, supra note 67, at

180. As a result, it is unwise to use such accusations as a political club, which is almost certain to occur, at

least periodically, if the legislature has oversight over this area. See GEYH, supra note 155, at 51–52

(describing the periodic upheaval that occurs between the public, Congress, and the Supreme Court in

relation to disfavored opinions); Green, supra note 67, at 181 (discussing the way people might respond to

disfavored opinions with unprovable bias allegations).

2021] Honesty Is the Best Policy 111

expression of bias. To see how this plays out, consider the case of the aversive racist.

Aversive racism describes racial bias among people who claim to hold egalitarian

beliefs but who nonetheless act on “indirect, subtle, and less obvious racial

biases.”

158

Notably, aversive racists do not discriminate if strong norms against

discrimination are present and the discrimination would be apparent to themselves

or others.

159

Without transparency, there is no mechanism of visibility by which

bias can become apparent to self or others.

160

At the same time, accountability helps

reinforce and establish norms because it signals their relevancy and strength.

161

Adopting language in the Model Code and Code of Conduct will promote

salient norms, transparency, and accountability in relation to our expectation that

judges take active steps to reduce the impact of their implicit biases on their judicial

decision-making. Expressing norms in ethical guidelines increases focus on specific

behaviors, which serves the interest of transparency because its very idea depends

on the assumption that another actor is watching.

162

Moreover, the Model Code and

Code of Conduct’s language should be updated so that judges, while not subject to

discipline for experiencing an implicit association, should be subject to discipline for

not participating in good faith efforts to curb their implicit associations. If the

guidelines have no teeth, there is nothing that will trigger formal accountability

and the efforts will be less effective.

163

In addition to norms and motivation, having the opportunity for “mindful

reflection on individuating features,”

164

is essential to overcoming an unconscious

reaction.

165

One way to create the opportunity to deliberate is for judges to make

the internal commitment before and during each case, bench memo, brief, hearing,

opinion, and order to not make decisions influenced by bias. In the first place, this

creates the opportunity to consider “individuating features.” In the second, research

has demonstrated that people will not make decisions based on bias when the choice

is clearly biased.

166

Judges can further operationalize this commitment by looking

for situational explanations to replace individual stereotypes

167

and deliberately

considering “countertypical associations” before and during judicial decision-

making.

168

Countertypical association is exposure to members of stereotyped groups

that we would respond to positively or exposure to a member of a negatively

158

Dovidio et al., supra note 14, at 179.

159

Id.

160

See Krieger & Fiske, supra note 15, at 1050–51.

161

See id.; MCADAMS, supra note 102, at 7, 16, 62 (“By publicly endorsing a particular behavior, law tends to

make that behavior salient, thereby producing self-fulfilling expectations that it will occur.”).

162

See Krieger & Fiske, supra note 15, at 1050–51.

163

See Kalev et al., supra note 141, at 590 (finding that corporate diversity efforts were most effective when

responsibility was assigned).

164

Quintanilla, supra note 85, at 24.

165

See Kang et al., supra note 11, at 1177–78.

166

Krieger & Fiske, supra note 15, at 1050.

167

Nordell, supra note 148.

168

See also infra text accompanying notes 196–97 (discussing the use of indirect countertypical associations to

reduce bias in lieu of diversity).

112 Indiana Journal of Law and Social Equality [9:94

stereotyped group that does not share the stereotypical trait.

169

Judges could

incorporate countertypical associations by a screensaver, desk photo, or work of art

in their office.

170

Granted, there is no way to enforce compliance or ensure that a judge is

meaningfully deliberating, but as Alex Kozinski succinctly pointed out, “[j]udicial

ethics, where it counts, is hidden from view.”

171

We already count on judges to do

their best to discharge their duties faithfully and there is good reason to believe

that if court systems adopt norms protecting against implicit judicial bias, judges

will act responsibly to self-regulate in this area, just as they do in many others.

172

One potentially intransigent obstacle in creating the opportunity to

deliberate is the workload for judges. Over the past 100 years, cases in district

courts have shown an eleven-fold increase.

173

For federal appellate courts, the

increase was sixty-four fold.

174

In contrast, the population only increased by a factor

of four, district judges by a factor of ten, and circuit judges by a factor of six.

175

the state level, growth in cases over time has outpaced the increase in judges as

well with state judges increasing at only half the rate of cases.

176

I recognize that

this limits the opportunity to deliberate, but pending costly and systemic change,

these proposals are crafted with the judicial system’s status quo ante in mind.

A final method to curb implicit judicial bias is increasing the diversity on the

bench and in chambers. In the first place, it may be that judges who are members of

out-groups experience fewer implicit associations favoring in-groups than in-group

members.

177

Studying a small pool of judges, Jeffrey Rachlinski et al. found that

eighty-seven percent of white judges had a pro-white bias in contrast to forty-three

percent of black judges.

178

In addition, black judges had a significantly weaker pro-

white bias than their white counterparts.

179

At the same time, research shows that

countertypical associations, or exposure to people who run counter to our implicit

schemas, reduce the expression of implicit bias.

180

Lastly, some commentators

169

Kang et al., supra note 11, at 1173.

170

See id.

171

Kozinski, supra note 118, at 1106.

172

See Geyh, supra note 29, at 696–97 (discussing the judge’s responsibility to decide her own disqualification

motion); Shepard, supra note 133.

173

Albert Yoon, Law Clerks and the Institutional Design of the Federal Judiciary, 98 MARQ. L. REV. 131, 133

(2014).

174

Id.

175

Id.

176

CHARLES GARDNER GEYH, COURTING PERIL: THE POLITICAL TRANSFORMATION OF THE AMERICAN JUDICIARY 26

(2016).

177

Rachlinski et al., supra note 19.

178

Id.

179

Id. (finding that black judges made the schema-inconsistent match twenty-six milliseconds faster than the

schema-consistent match while white judges made the schema-inconsistent match 216 milliseconds faster

than the schema-consistent match).

180

Kang et al., supra note 11, at 1170 (citing Nilanjana Dasgupta & Shaki Asgari, Seeing Is Believing:

Exposure to Counterstereotypic Women Leaders and Its Effect on the Malleability of Automatic Gender

Stereotyping, 40 J.

EXPERIMENTAL SOC. PSYCHOL. 642, 649–54 (2004)).

2021] Honesty Is the Best Policy 113

postulate that people in the in-group and people in the out-groups can “see two

different realities.”

181

For instance, “[w]hite people . . . might only hear a racist

remark, [but] people of color might register subtler actions, like someone scooting

away slightly on a bus—behaviors the majority may not even be aware they’re

doing.”

182

Currently, white men hold judgeships in the greatest number. In federal

courts, white men are 45.6% of circuit court judges and 49.3% of district court

judges.

183

In comparison, women hold thirty-seven percent of circuit judgeships and

thirty-four percent of district judgeships.

184

By race, black people are thirteen

percent of circuit judges and fourteen percent of district judges.

185

Even more

concerning is the fact that one circuit has only one woman out of nine judges, seven

of thirteen circuits have only one black person sitting on them (this includes the

Eleventh Circuit which has the largest population of black people in the country),

and one circuit has no black judges at all.

186

For district courts, the disparity is

worse. Eleven of ninety-one courts have no women on them and forty-four have no

black people serving on them.

187

Even starker, black women hold only 4.4% of

circuit judgeships and 6.1% of district judgeships.

188

State court judges and clerks in both federal and state courts have a higher

ratio of white men, too. For instance, black judges are between seven to nine

percent of state court judges across all levels

189

and female judges occupy thirty-one

to thirty-six percent of state judicial seats across all courts.

190

At the same time, the

181

Nordell, supra note 148.

182

Id.

183

BARRY J. MCMILLION, CONG. RESEARCH SERV., R43426, U.S. CIRCUIT AND DISTRICT COURT JUDGES: PROFILE OF

SELECT CHARACTERISTICS 8, 20 (2017), https://fas.org/sgp/crs/misc/R43426.pdf.

184

Id. at 4, 15.

185

Id. at 5, 17.

186

Id. Other racial minorities are also underrepresented. For instance, there are four circuit courts with no

Hispanic judges and nine circuit courts with no Asian American judges. Id. at 5.

187

Id. at 15, 17. As with circuit courts, Asian American and Hispanic judges are also underrepresented. Asian

Americans hold 2.8% of district judgeships despite being 5.8% of the total U.S. population and Hispanic

judges hold about 10% of district judgeships and represent 18% of the U.S. population. Id. at 20; The Rise of

Asian Americans, P

EW RES. CTR. (June 19, 2012), http://www.pewsocialtrends.org/2012/06/19/the-rise-of-

asian-americans/; Antonio Flores, How the U.S. Hispanic Population Is Changing, PEW RES. CTR. (Sept. 18,

2017), http://www.pewresearch.org/fact-tank/2017/09/18/how-the-u-s-hispanic-population-is-changing/. By

comparison, black people make up about 13% of the population and women about 50.8%. Kristen Bialik, 5

Facts About Black Americans., PEW RES. CTR. (Feb. 22, 2018), http://www.pewresearch.org/fact-

tank/2018/02/22/5-facts-about-blacks-in-the-u-s/; QuickFacts, U.S.

CENSUS BUREAU,

https://www.census.gov/quickfacts/fact/table/US/SEX255217#viewtop (last visited Aug. 29, 2020).

188

MCMILLION, supra note 184, at 8, 20.

189

National Database on Judicial Diversity in State Courts, AM. BAR ASS’N,

http://apps.americanbar.org/abanet/jd/display/national.cfm (last visited Sept. 9, 2019)

[https://web.archive.org/web/20190909143807/http://apps.americanbar.org/abanet/jd/display/national.cfm].

190

2017 US State Court Women Judges, NAWJ, https://www.nawj.org/statistics/2017-us-state-court-women-

judges (last visited Aug. 30, 2020).

114 Indiana Journal of Law and Social Equality [9:94

estimated percentage of minority members in judicial clerkships is 6.5% and women

in judicial clerkships is about eighteen percent.

191

Perhaps because these numbers are so stark, Kang et al. recognize the value

in diversifying the bench but also recognize that it would be a costly endeavor.

192

While increasing judicial diversity at the federal level may require fewer economic

resources because an appointment process is already in place,

193

it would still take

significant political resources for alignment around diverse candidates. It is

possible, however, as the Obama Administration’s record shows.

194

In contrast,

because many states select judges through election,

195

ensuring diverse results

would be a very complicated and resource-laden endeavor that would likely require

revamping entire selection processes through constituent and legislative support.

With this in mind, one possible stopgap measure to true diversity is to stimulate

countertypical associations by using pictures, videos, and other indirect contact.

196

Indirect countertypical associations are not a substitute for actual diversity, and

cannot be expected to promote accountability, transparency, and other norm-

reinforcing interactions among peers to the same degree. However, they are likely,

as Kang et al. suggest, worth experimenting with.

197

Finally, it is possible that judges will be unable to acknowledge their implicit

bias because they will not want to lose face or delegitimize their rulings.

198

However, drawing a broad conclusion like this is suspect given that judges

occasionally admit general error in writing in legal opinions

199

or, notably,

acknowledge bias even after leaving office.

200

It might be helpful to remember our

original premise, which is that judges are human. And humans are capable of

change.

201

191

ELIZABETH CHAMBLISS, IILP REVIEW 2017: THE STATE OF DIVERSITY AND INCLUSION IN THE LEGAL PROFESSION

16, 23, http://www.theiilp.com/resources/Pictures/IILP_2017_Demographic_Survey.pdf.

192

Kang et al., supra note 11, at 1170 (“[A]ny serious diversification of the bench, the bar, and staff would take

enormous resources, both economic and political.”).

193

See FAQs: Federal Judges, U.S. CTS., http://www.uscourts.gov/faqs-federal-judges.

194

Under President Obama, sixty-nine percent of circuit court appointees and sixty-two percent of district

court appointees were nontraditional appointees, i.e. not white males. MCMILLION, supra note 184, at 9, 22.

195

Forty-two states provide for the election of at least some judges. Charles Gardner Geyh, Why Judicial

Elections Stink, 64 OHIO ST. L. J. 43, 52 (2003).

196

Kang et al., supra note 11, at 1171.

197

Id.

198

See MCADAMS, supra note 102, 220–21.

199

See City of Erie v. Pap’s A.M., 529 U.S. 277, 316 (1999) (Souter, J., dissenting) (acknowledging a contrary

position to one earlier stated and explaining it as “[i]gnorance, sir, ignorance” (quoting McGrath v.

Kristensen, 340 U. S. 162, 178 (1950) (Jackson, J., concurring)).

200

See James C. McKinley, Jr., Ex-Brooklyn Judge Seeks Reversal of His Verdict in 1999 Killing, N.Y. TIMES

(Dec. 12, 2013), https://www.nytimes.com/2013/12/13/nyregion/ex-brooklyn-judge-seeks-reversal-of-his-

verdict-in-1999-murder-case.html.

201

For instance, consider that in 2016, a C-SPAN caller made headlines after identifying as a white man from

North Carolina and asking how he could become less biased, so he could “be a better American.” Paul

Walter, Caller Admits Racism and Is Gently Advised, C-SPAN (Aug. 22, 2016), https://www.c-

span.org/video/?c4618001/user-clip-caller-admits-racism-gently-advised.

2021] Honesty Is the Best Policy 115

CONCLUSION

Because the days of casting judges as perfectly impartial are gone, and

because the social science research on implicit bias is clear, we should embrace a

view of the judiciary that allows them to be human but also promotes fairness,

legitimacy, and procedural justice. This means acknowledging that judges

experience implicit bias, while neither concluding that impartiality is dead nor

avoiding responsibility for implicit bias’s effects. Rather, just as the ermine cloak

used to be the litmus paper on which we gauged our judges’ purity, now we should

use their actions in response to, and in the face of, implicit associations to prove

their sincerity in curbing unfairness resulting from unchecked implicit bias.

The methods I propose are intended to help us, and our judges, do just that.

Court systems can employ these steps through existing administrative, ethical, and

disciplinary frameworks by requiring judges to take the IAT, mandating implicit

bias training, increasing judicial diversity while employing other opportunities for

countertypical associations, and operationalizing mindful deliberation. In addition,

norm-reinforcing steps like disciplinary sanctions for not participating in these

measures and updates to the Model Code and Code Conduct should be employed.

These efforts are unlikely to be perfect, but I share Kang et al.’s view that “[e]ffort is

not always sufficient, but it ought to count for something.”

202

Kang et al., supra note 11, at 1169.