Focus on Explicit Disparities Instead of Implicit Biases
The theory of implicit bias will not help us understand or reverse patterns of segregation. It will only mire us in all-too familiar, intractable controversies over mental state. Rather than trying to sell a skeptical public on a novel theory of prejudice, we should focus on the institutionalized conditions that produce and sustain inequality and segregation.
Professor Kang argues that the concept of implicit bias can reveal ongoing discrimination of which most Americans are unaware and this revelation will inspire a sense of moral urgency to address the problem of segregation. But this assumes that most of the bias responsible for segregation is unconscious and unwitting, as opposed to deliberate but concealed. Suppose that although most Americans abhor the type of discrimination they associate with unrepentant bigots, many believe that they are justified in considering lamentable but undeniable racial dynamics. A classic example of such a soft form of discrimination is Reverend Jesse Jackson’s notorious admission that even he worries when approached by a group of young black men on a deserted street at night. Many people who would not consider themselves bigots and would not discriminate in other areas of their lives probably have similar reactions. A showing of implicit association can’t distinguish unconscious bias from this kind of concealed bias. In fact, the implicit association research began as a way of smoking out such concealed bias: the original concern that inspired implicit association research was that as prejudice became unacceptable, test subjects would try to conceal their biases from researchers. We won’t be able to stir up moral urgency about racial generalizations that many people are aware of and believe are justified.
Consider one of Professor Kang’s examples in this light: studies show that many people perceive minority neighborhoods to be more dangerous and disorderly than objectively similar white neighborhoods. Is the problem here that people are unaware that they are making race dependent evaluations, or is it that they are unwilling to admit that they are making race dependent evaluations? The idea of a poor, high crime minority neighborhoods is so pervasive in our society that I suspect many people consciously believe that it’s reasonable to presume that a minority neighborhood is more likely to be dangerous and disorderly than a predominantly white one. When forced to make a high stakes decision (like purchasing a home—the largest investment most people will ever make) under conditions of uncertainty, most people will consider any available information, even if they know the information is imperfect. Race is a highly imperfect proxy for crime or disorder, but sadly there is some correlation. A rational person could believe that considering race is justified, (although, given anti discrimination laws and social mores, she may be unwilling to admit it.)
Even people who do not credit such racial generalizations may well think it prudent to take account of the likelihood that other people will act on them, driving down real estate values and making investments in minority neighborhoods less reliable. It’s easy to imagine the vicious cycle that will result: unstable property values make the neighborhood a bad risk, causing banks to withhold prime mortgages; property owners faced with unattractive loan terms are more likely to default; higher foreclosure rates lead to higher numbers of abandoned properties, which are easy marks for vandals, squatters, and illegal enterprises; and, of course, these activities make the neighborhood more dangerous and disorderly than it would otherwise have been, vindicating those who associate minority neighborhoods with crime and disorder. This kind of downward spiral can get going with only a relatively small number of people directly acting on racial stereotypes, supplemented by a larger number of people responding to the risk that other people will act on racial stereotypes or acting on the objective differences created by downward spiral once it gets going.
In the case of individual buyers, this type of bias is not actionable and even in the cases of real estate agents, landlords and lenders subject to antidiscrimination laws, it is very hard to prove. Moreover, bias is far from the only cause of discriminatory practices: non-biased people will contribute to the problem when they react to racially unequal market dynamics such as unstable property values in minority neighborhoods. Focusing on implicit bias will only reinforce the widespread and misguided idea that inequality demands redress only if it is caused by a specific prejudiced mental state, conscious or unconscious.
Eliminating racial injustice will require us to attack the institutional practices and structural incentives that now make racial discrimination rational. Instead of focusing on a subjective and ultimately unknowable mental state, we should focus on objective inequalities and the objective circumstances surrounding contested decisions. We don’t need an account of the cognitive mechanism that produced a race or sex dependent decision—it is sufficient to demonstrate objectively that the decision is in fact race or sex dependent. Here the social science developments that are most promising are not in psychology, but in econometrics and empirical analysis, which can offer increasingly refined techniques to isolate race, sex and other variables and determine which decisions depend on them. An objective approach to questions of inequality would blur, if not eliminate, the distinction between disparate treatment and disparate impact, and look for unexplained and unjustified disparities in workforces, renter populations, mortgagees, and other groups covered by anti discrimination laws. Most importantly, a focus on objective unjustified inequality would lead us to look beyond individual cases to broad, institutional solutions that would improve the fairness of housing, financial and labor markets generally. These solutions will not involve convincing individuals that they are unwittingly prejudiced—instead they will involve institutional changes and incentives for socially constructive behavior.