Among the most dangerous arguments for racial profiling are the most rational. They are persuasive because they are by definition based on logic and statistics. The premise is that a stereotype is true, or more probably true than false, or at least more true of the group subjected to it than of other populations.
These utilitarian calculations can be about Chinese immigrant scientists who are alleged to be spies or African American youngsters who are alleged to be thugs. The same principles are at stake. The more successful we are eliminating old-school prejudice, the more it mutates into education-resistant forms. Instead of angry demagogues shouting obscenities to threaten violence, we now are confronted by calm academics displaying PowerPoint slides with percentages which are color coded literally and figuratively.
People who appear to be anything but extremists prefer these justifications for their attitudes. They even deny that they are bigots; they are merely reviewing evidence. Some use fancy terms, such as “Bayesian inference,” which is an important technique for revising predictions based on additional information, but which also can be abused like any other intellectual tool.
They assert openly that individuals of Chinese heritage have a propensity toward loyalty to the Chinese government, out of ethnic affinity, even if they are United States citizens. They add that the Chinese government is recruiting agents through such appeals. Thus, they insist, targeting Chinese Americans is appropriate. If one is guilty, others must be too.
The problem is that we should believe what is rational — and that leads us to suppose what is rational also is right. Yet there are many claims and public policies that are superficially reasonable yet morally wrong. That extends beyond discrimination. Philosophers have a term for it, “the naturalistic fallacy,” which consists of confusing descriptive claims about the world with normative claims about our ideals. People realize this: we would be appalled by an advocate for euthanizing children who are disabled, because cost-benefit analysis proves society would be better off economically. We can concede for the sake of discussion that the descriptive claim (disabled children cost more than they benefit, materially) is accurate, without concluding that the normative claim (they should be killed) follows.
Much that looks rational also is not. Thanks to “confirmation bias,” our tendency to exaggerate patterns to select the data conforming to our pre-existing assumptions, we discount warnings about risks.
Experiments, not to mention experience, demonstrates that the same behavior exhibited by whites and blacks is processed by observers very differently. The white youngster is acting up because they need treatment for attention deficit disorder, while the black kid is revealing tendencies toward future law breaking that should be punished preemptively, even if objectively their temper tantrums are identical. The suggestion that African Americans are biologically or culturally oriented toward criminality, because they happen to be overrepresented in the prison system, verges on comedy and arguably crosses the line except for its tragedy. It requires willful disregard of the causes of disparities, which include the self-fulfilling prophecy of racism.
Even if we were disinterested in civil rights and hostile to diversity, sophisticated empiricism would include collateral consequences. A “thick description,” to use the phrase of social scientists, of how the world works would incorporate the penalty to productivity caused by false positives. A person who wished only to maximize the gains generated by law enforcement investigations would account for the damage done by accusations which turn out to be erroneous if not malicious. That is not limited to the person who is affected, whose life may be ruined, and the community she belongs to, which suffers the cumulative losses, because others have invested in training that person who is cast aside without respect.
The crux of racial profiling is not improved with rationalism. At their heart, invidious stereotypes depend on using the identity of people (the classification of race or ethnicity) to deduce the actions of persons (espionage or theft). The Supreme Court, by and large, has come around. Legislation challenged for constitutionality must be rational. The judges err on the side of the government. Laws, however, that operate by dividing people according to a “suspect classification” such as race or ethnicity are said to trigger “strict scrutiny.” The judges demand an explanation.
Thanks to a movement for Black equality, which not only eliminated the formal, legal versions of bias but also created a consensus about informal, everyday norms, we continue our progress as a democracy. If we are fooled by bigots because they have lowered their voices and armed themselves with numbers, we will regress.
Frank H. Wu is the William L. Prosser Distinguished Professor at the University of California Hastings College of the Law and adjunct professor at Johns Hopkins University.