Should Race Matter? Unusual Answers to the Usual Questions.

• Author: Alexander, Larry
• Position: Book review

SHOULD RACE MATTER?: UNUSUAL ANSWERS TO THE USUAL QUESTIONS. By David Boonin. (1) Cambridge, Cambridge University Press. 2011. Pp. vii + 441. $99.00 (cloth), $34.99 (paper).

One frequently hears that America has a race problem. We agree, but the race problem we identify is not what is usually meant by those who invoke it. It is not discrimination, intentional or otherwise, but rather obsession with race that is America's more consequential "race problem" today. America has vanquished slavery, segregation, and long-standing racial discrimination only to succumb to an almost equally destructive race obsession. Despite the biological arbitrariness of dividing a single, interbreeding biological species into "races," despite the sorry history legally and socially of the use of race, and despite the Civil Rights Movement's original ambition to substitute the content of character for the color of skin as the basis of decision making, America today is in many ways as race conscious as it was in the era of Jim Crow.

For that reason we welcome David Boonin's Should Race Matter? Boonin takes up five topics that constitute a good portion of the current obsession--reparations, affirmative action, hate speech, hate crimes, and profiling--and he subjects each to philosophical scrutiny. Boonin is sober and fair-minded in tone, and purports to be careful and comprehensive in method. Unlike many discussions of race, Boonin's tries to shed light, not heat. He deserves to be read by everyone who takes a serious interest in public policy as it bears on race.

Boonin's book has its limitations, as we will suggest. Moreover, Boonin discusses only race, not sex, ethnicity, nationality, religion, disability, or sexual preference, although most of the policies he considers have been urged or actually extended beyond race to some or all of these other categories. Nonetheless, Boonin's analyses of these policies as they bear on race would have direct implications for these other categories. Given that Boonin takes 350 pages to examine five racial policies, we think limiting his focus to race was quite justifiable.

Although we believe Boonin's is a worthwhile treatment of contemporary racial policies, we take issue with him on several points. We think that his arguments in support of affirmative action and hate crimes are incomplete and thus unpersuasive, and we consider his case for reparations a failure on its own terms. Nonetheless, we admire the effort at fair-mindedness and the care with which he makes the case for these policies.

1. HATE SPEECH

   We begin with the topic on which Boonin and we are in full agreement: hate speech. Boonin is opposed to bans on racial "hate speech" because he believes such bans can only be justified by repudiating current free speech doctrines that we would and should be loath to reject. In his exemplary two-chapter discussion of the issue, he canvasses the major rationales that are offered to support bans on racial hate speech and finds them all wanting. In the first chapter on this subject (Chapter Six), Boonin analyzes arguments that try to assimilate racial hate speech to categories of expression that are already deemed unprotected speech by today's constitutional free speech doctrines. Not all racial hate speech is a true threat; and true threats are already prohibited (pp. 210-13). Not every instance of racial hate speech constitutes "fighting words," which are insults rendered face to face and likely to provoke a violent response (pp. 216-17). (4) Racial hate speech cannot be assimilated to the libel of some corporate entity (pp. 217-25), nor is every instance of it a case of actionable harassment (pp. 226-29). In sum, a broad ban on racial hate speech could not be justified under current free speech doctrines.

   In Chapter Seven, Boonin then considers and rejects justifications for banning racial hate speech as such, rather than as instances of other categories of legally unprotected expression. "Words that wound" is a justification that sweeps in far too much expression that a free society would want to protect (pp. 230-36). Nor can hate speech be properly construed as a "subordinating speech act" (pp. 236-41): either the speaker lacks the authority required to subordinate, or the notion of authority has to be expanded to the point where speech that almost everyone would want to protect would be deemed subordinating. Finally, hate speech cannot be banned on the ground that it "silences" without again sweeping in lots of quite legitimate speech (pp. 241-45).

   At the conclusion of Chapter Seven, Boonin asks whether the fact, if it is so, that hate speech wounds, subordinates, and silences makes a case for banning it even though each of those harms, individually, would not do so (pp. 245-48). He compares a white student's calling a black student a "nigger" in the presence of other students with a Catholic student's publicly calling another Catholic student a "dangerous heretic" for supporting abortion and gay rights. He concludes that if the first student's speech wounds, subordinates, and silences, the second student's can do so as well. If the second student's speech should be protected--and Boonin believes that it surely should be--then so, too, should the first student's. The "combination" of harms justification for hate speech bans fails.

   Depending on the circumstances, the first student's speech might be banned as "fighting words" under well-established constitutional law. But in principle, we are in agreement with Boonin on the issue of hate speech. Arguments for hate speech bans consist of some normative premise (speech with characteristic X should not be protected as free speech) and a factual premise (racial hate speech has characteristic X). In Chapter Six Boonin takes on arguments in which the factual premise fails to hold. In Chapter Seven he takes on arguments in which the normative premise fails. Although racial hate speech is ugly and regrettable, so too is much other speech. The power to cleanse public dialogue of ugly and hurtful speech is a dangerous power that, for reasons Boonin adumbrates, would likely be used either selectively in a divisive, partisan way, or much too broadly. (5)

2. RACIAL PROFILING

   In the final two chapters of the book, Boonin examines the vexed topic of racial profiling. In the book's penultimate chapter he asks whether racial profiling is rational, and he concludes that it can be. In the final chapter he asks whether racial profiling, even if rational, is nonetheless immoral, and he concludes that it is not. With some qualifications regarding the moral question, we believe Boonin is essentially correct in both chapters.

   In thinking about the rationality of racial profiling, it is useful to be clear about what profiling is. (6) When we profile or stereotype--these are essentially synonymous--we use a given trait that is relatively easy to identify as a proxy for the trait in which we are ultimately interested. The relation between the proxy trait and the target trait is a probabilistic one. The existence of the proxy trait makes it more likely that the person who possesses it has the target trait than a person picked at random. The rationality of using a given proxy trait depends upon its correlation with the target trait and the relative costs and benefits of using a different proxy or a more refined proxy that has a higher correlation with the target trait. But there is no question but that using proxy traits to predict target traits is rational. Indeed, we must do it, for we must act on the basis of imperfect information about others all the time. It is not only insurance companies predicting our life expectancy, our health, or our traffic accidents, or political pollsters predicting our likely votes, or sociologists predicting behavior more generally who must rely on proxies. All of us do, all the time. It is inconceivable that we could dispense with proxies, so the rationality of their use is beyond question.

   That still leaves open the question of whether the use of race as a proxy is ever rational. Race is, after all, an imprecise characteristic, and its use will require some arbitrary judgments. Nonetheless, despite the arbitrary boundaries of the proxy, in some contexts its use can be quite rational. If one is testing for sickle cell anemia, it makes sense to focus on persons who appear to be descended from the pre-colonial peoples of sub-Saharan Africa, just as if one is testing for Tay-Sachs, it makes sense to focus on Ashkenazi Jews. And even those who purport to object to racial profiling probably take greater precautions for their personal safety when in some neighborhoods rather than others, with the racial or ethnic composition of the neighborhood serving as their proxy for relative dangerousness.

   Of course, if there are more predictive proxies available at a low enough cost to obtain--or if, again at a cost-benefit justified cost to obtain, there are ways to refine the proxy by combining it with other traits--then race will cease to be the most rational proxy for its purpose. A neighborhood's ethnic makeup might be less predictive of danger than its wealth or poverty. Or its economics combined with its racial makeup may be more predictive than either proxy by itself.

   Boonin's focus is on a paradigmatic use of racial profiling: the decision by the police to stop, among drivers who are violating traffic laws, a disproportionate number of black offenders on the ground that they are more likely than the average offender to be committing drug or weapons offenses. If it is true that they are more likely than average to be committing these offenses, and given the finite resources of the police, then Boonin concludes that this type of racial profiling is rational.

   Boonin first considers some arguments that contest the claim that black drivers are more likely to be committing drug or weapons offenses, and he finds them to be unsound (pp...