Ralph Richard Banks & Richard Thompson Ford, (how) Does Unconscious Bias Matter? Law, Politics, and Racial Inequality

• Publication year: 2009

EMORY LAW JOURNAL

Volume 58 2009 Number 5

ARTICLES

(HOW) DOES UNCONSCIOUS BIAS MATTER?: LAW, POLITICS, AND RACIAL INEQUALITY

Ralph Richard Banks^*

Richard Thompson Ford^**

ABSTRACT

During the past several years, psychological research on unconscious racial bias has grabbed headlines, as well as the attention of legal scholars. The most well-known test of unconscious bias is the Implicit Association Test (IAT), a sophisticated and methodologically rigorous computer-administered measure that has been taken by millions of people and featured in major media. Its proponents contend that the IAT reveals widespread unconscious bias against African Americans, even among individuals who believe themselves to be free of racial bias.

In fact, however, the findings of the IAT are ambiguous. The test could just as plausibly be thought to measure racial bias that is simply covert, known to oneself yet intentionally concealed from researchers. On this interpretation, the IAT reveals not that individuals are more biased than they realize, but that they are more biased than they want others to know. The characterization of the IAT as a measure of unconscious bias has practically eclipsed this plausible alternative interpretation. Why?

One possibility is that unconscious bias, even if not incontrovertibly demonstrated by the IAT, warrants attention because it poses a unique challenge for antidiscrimination doctrine. But this explanation for the ascendance of the unconscious bias discourse is wrong. Antidiscrimination law grapples as well, or as poorly, with unconscious bias as with covert bias. Neither statutory nor constitutional doctrine turns on the distinction between the two.

The better explanation for the ascendance of the unconscious bias discourse is that assertions of widespread unconscious bias are more politically palatable than parallel claims about covert bias. The invocation of unconscious bias levels neither accusation nor blame, so much as it identifies a quasi-medical ailment that distorts thinking and behavior. People may be willing to acknowledge the possibility of unconscious bias within themselves, even as they would vigorously deny harboring conscious bias. The unconscious bias claim thus facilitates a consensus that the race problem persists.

Despite its ostensible political benefits, the unconscious bias discourse is as likely to subvert as to further the cause of racial justice. Racial injustice inheres in the entrenched substantive racial inequalities that pervade our society. These disparities are not primarily a consequence of contemporary racial bias. Thus, the goal of racial justice efforts should be the alleviation of substantive inequalities, not the eradication of unconscious bias. Yet, the rhetoric of unconscious bias is so compelling that people are likely to accept it as the goal of racial reform and, consequently, to push the theory in directions that siphon energy away from problems of substantive inequality and that may be undesirable in their own right. The unconscious bias discourse reinforces a misguided preoccupation with mental state, and perpetuates an obsession with antidiscrimination law, rather than policy reform, as a means of realizing racial justice goals. If the goal is to eliminate substantive inequalities, then the task of racial justice advocates should be to explain forthrightly why those inequalities are objectionable and how to address them.

INTRODUCTION ............................................................................................ 1056

I. EMPIRICAL AMBIGUITY .................................................................... 1060

A. The Measurement of Implicit Bias ............................................ 1060

B. The Necessity of Measuring Conscious Bias ............................ 1063

C. The Social Desirability Problem .............................................. 1065

D. Recharacterizing the IAT .......................................................... 1068

II. ANTIDISCRIMINATION DOCTRINE ..................................................... 1072

A. Statutory Law: Title VII ............................................................ 1073

1. Single-Motive Cases: Formalism, Realism, and

Employment at Will ............................................................ 1073

2. Mixed-Motive Cases: Causation and Duty ......................... 1081

3. The Honest-Belief Rule ....................................................... 1086

B. Constitutional Law: Equal Protection ...................................... 1089

C. Disparate Impact, Affirmative Action, and the Burden of

Proof ......................................................................................... 1100

III. POLITICAL APPEAL ........................................................................... 1103

A. The Stigma of Racism ............................................................... 1103

B. The Denial of Discrimination ................................................... 1104

C. The Denial of Bias .................................................................... 1106

D. The Historical Narrative .......................................................... 1108

IV. AGAINST BIAS .................................................................................. 1110

A. The IAT and the Risk of Failure ............................................... 1110

1. Empirical Uncertainty ........................................................ 1110

2. The Scholarly Role ............................................................. 1112

B. The Costs of Success ................................................................. 1113

1. Goal Distortion: Confusion of Means and Ends ................ 1113

2. Undesirable Outgrowths .................................................... 1116

CONCLUSION ................................................................................................ 1121

INTRODUCTION

Jimmy Carter made headlines when he confessed-not to his pastor, psychotherapist, or spouse, but to a national audience-that he had "looked on a lot of women with lust."¹It may sound odd to claim that the 1970s were a more innocent era but let's face it: these days it's enough if a politician doesn't sin in the Oval Office, with an underage congressional page, or using a .gov e-mail address. Virtually no one, let alone a politician, would admit to sinning in one's own heart-the one place on earth no special prosecutor can serve a summons and no intrepid reporter can plant a tape recorder. Yet lots of people want to look just there-in one's own heart of hearts and innermost mind-for sin.

Repressed memories, subconscious phobias, the maledictions of the reactive mind-popular psychology is replete with hypotheses that posit that reality is hidden, that appearances deceive, that the truth is submerged, obscure, invisible to the naked eye. The idea of the unconscious is highly provocative, and appealing, perhaps nowhere more so than as dramatized in Hitchcock's series of psychodramas (most famously Psycho, but also Vertigo,

Spellbound, and Marnie).²The idea of the unconscious has proven appealing as well for legal scholars writing about antidiscrimination law during the past two decades. Early legal analyses drew from the theories of Freud, for whom dreams and free association served as passageways to the depths of the unconscious.³More recently, legal scholars have relied on a growing body of sophisticated and methodologically rigorous experimental studies conducted by social and cognitive psychologists.⁴

The most well-known test of unconscious bias, and the one typically referenced by legal scholars, is the Implicit Association Test (IAT).⁵A computer-administered test available over the Internet, the IAT is a compelling interactive experience that has been taken by millions of people, and featured in print and broadcast media.⁶The IAT measures the strength of the association between social categories (e.g., blacks or whites) and positive and negative attributes (e.g., "joy" and "love" versus "agony" and "evil"). Akin to a computer game for grownups, the IAT requires momentary immersion into the interactive medium. In a series of trials, the participant categorizes images or words that appear on the computer screen by pressing particular computer keyboard keys as quickly as possible. At the end of the exercise, the computer calculates a score that reflects the nature and magnitude of one's implicit bias.⁷

Most participants are found to have an implicit bias against African Americans. The overt racism of the Jim Crow era, the psychological research suggests, has given way to racial bias that is predominantly unconscious.

In fact, the findings of the IAT are ambiguous. The characterization of the IAT as a measure of implicit bias depends on being able to distinguish implicit bias from conscious bias. Yet it is extraordinarily difficult to disentangle the two because, since the disavowal of racism during the civil rights era, research participants have become increasingly unwilling to openly express views that may be condemned as racist. Thus, the IAT could defensibly be viewed as a subtle measure of conscious psychological processes, of attitudes and beliefs that are known to oneself yet intentionally concealed from researchers. This empirical ambiguity has been practically eclipsed by the unconscious bias account. Why?

One possibility is that unconscious bias, even if not incontrovertibly demonstrated by the IAT, warrants the attention of legal scholars because it poses a unique challenge for antidiscrimination doctrine. This explanation for the ascendance of the unconscious bias discourse is intuitively appealing and widely embraced. But it is wrong. Antidiscrimination law grapples as well, or as poorly, with unconscious bias as with covert bias. Neither statutory nor constitutional antidiscrimination law turns on the distinction between the two. While the research cannot distinguish between...