1. I define “race” more precisely in Chapter 2. Until then, I use those quotation marks to caution the reader that the concept is fraught with scientific and ethical difficulties, and that I have in mind a rather specific and perhaps not universally accepted meaning when I invoke the term.
2. Since this book deals strictly with U.S. society, I use the terms “blacks” and “African Americans” interchangeably throughout, with no risk of ambiguity.
3. The reader may peruse the Appendix to see some statistical documentation of these claims.
4. I have long thought that applying Goffman’s insights could illuminate the economic study of racial inequality, though I am aware of no serious attempt in this vein.
5. D’Souza 1995 is an especially stark example of this style of argument.
6. It might be objected at this point that even among native-born African Americans there are further distinctions to be drawn—between the children of recent Caribbean or African immigrants and those whose ancestors have been in the United States for many generations; or between those descended mainly from slaves and those whose ancestors were mainly “free persons of color.” (Thomas Sowell makes much of these very distinctions when discussing blacks; Sowell 1981.) But this observation, correct in principle, is of limited relevance to my purposes here. Thus Sowell’s main point in drawing these intra-African-American contrasts is to argue that anti-black discrimination in the larger society takes us only so far when looking for an explanation of the social disadvantage of many blacks. As will become clear, I fully agree with this point. Nor do I presuppose that some monolithic, univalent, adverse perception of blacks by whites is the principal factor accounting for racial inequality.
1. Indeed, recent research in neuroscience suggests that perception of racial difference is deeply rooted in the human brain. Teams of brain scanners and social psychologists working in concert have recently found that a “particular part of the brain becomes more active when people look at members of a different race” (Berreby 2000, p. F3).
2. The phrase “racial classification” is also a term of art in the legal literature—with the Supreme Court pronouncing often on the conditions under which the use of racial classifications by governmental agents can be found consistent with the equal protection requirement of the Fourteenth Amendment. I will have essentially nothing to say here about that important but somewhat technical point of constitutional law.
3. I belabor this point because, in the partisan climate created by political struggle over race-based social policies, some commentators have taken to arguing that we ought to simply move beyond the discredited practice of “counting by race.” In my opinion, an inquiry into the ethics of racial classification that never gets beyond the categorizing acts of individuals, so as to inquire into the virtue of the purposes on behalf of which those acts are being undertaken, is a rhetorical exercise that ought not to be taken seriously.
4. My definition is quite similar to that advanced by Stephen Cornell and Douglas Hartmann. They write: “A race is a group of human beings socially defined on the basis of physical characteristics. Determining which characteristics constitute the race—the selection of markers and therefore the construction of the racial category itself—is a choice human beings make. Neither markers nor categories are predetermined by any biological factors. These processes of selection and construction are seldom the work of a moment. Racial categories are historical products and are often contested” (Cornell and Hartmann 1998, p. 24). With this definition in hand, I henceforth drop the fashionable practice of putting the word “race” in quotes, relying on the reader’s understanding of what I intend with the term.
5. Eloquent and powerful arguments exposing ethical and philosophical problems with the “race” concept can be found in Gilroy 2000 and Appiah 1992. For a discussion of the scientific limitations of the “race” idea see Cavalli-Sforza 2000.
6. I will on occasion use “raced” as an adjective. Though this may strike the reader as odd, it is natural in the context of my argument. I simply want to be able to express with economy the notion that a person’s fate is affected by racial categorization, or that the operation of a significant social institution is influenced by the practice of racial classification.
7. Again, I belabor what may be obvious only because the literature on the ethics of racial identification (self-classification) can be muddled on this point. For instance, Randall Kennedy argues that race-based feelings of loyalty, pride, and kinship are ipso facto morally problematic (Kennedy 1997). That argument is mistaken in my view, or, at best, it is woefully incomplete. It is not a moral error for a person subject to acts of racial classification, however ungrounded in objective taxonomy those acts may be, to take on as an operational and a psychological response some awareness of race that may serve to guide that agent’s interrelations with others. An inquiry into the ethics of racial identification, racial loyalty, racial kinship, and racial attachment that never gets beneath the act of a single individual so identifying and so feeling, that never inquires into the structure within which that identifying, loyalty-feeling, kinship-affecting agent operates, ought not, in my humble opinion, to be taken seriously.
8. The economist A. Michael Spence was among the first to make this point. See Spence 1974.
9. Raising employers’ review standards could also elicit greater effort from workers. But since an impossibly difficult standard makes effort superfluous, a tougher standard must discourage effort after some point. I am implicitly assuming in this thought experiment that by withholding the benefit of the doubt from black workers employers move beyond that point.
10. Bear in mind that this example is a thought experiment, and is not intended to be an explanation of today’s crime statistics!
11. More generally, both types of young men could be discouraged by the drivers’ reluctance to stop. What is required for this example to work is that the proportion of robbers among those hailing cabs rises as the frequency of stops declines. This condition will hold if the law abiding are more sensitive to delay than are those bent on robbery, which is quite plausible.
12. Of course some black buyers may be driven from the market altogether by this dealer behavior. But, as in the taxi example, when opting out is selective—with the toughest black bargainers being the first to withdraw, as seems plausible—the result will be to reinforce the dealers’ stereotypic view of blacks as easy marks.
13. As with the workplace example, a tougher admissions standard could either encourage or discourage effort. Notice, however, that effort is superfluous if the standard is trivially easy or impossibly difficult to meet. Lowering the standard therefore discourages effort when one starts from a point where the standard is not too difficult, which is assumed to be the case here.
14. Another scenario could envision race-preferential admissions policies as having been objectively necessary at the start, but as creating self-confirming incentives that perpetuate reliance on these policies long after they have ceased to be objectively required. In any case, I want again to emphasize that these thought experiments are being offered here to illustrate the logic (and relevance) of my theory of stereotypes, and not to explain (or to criticize!) today’s college admissions policies.
15. Regardless of which account he credits, in the real world of American society, beyond the confines of my thought experiments, a monopolistic observer has another important option: He can simply avail himself of an alternative source of labor exhibiting none of the problems associated with blacks—by moving his operations to a different region, or hiring immigrant workers rather than African Americans. The interesting point here is that negative stereotypes about one group may persist precisely because of the existence of positive stereotypes about another group and vice versa. The two sets of beliefs can be mutually supportive via interacting, self-confirming feedback processes of the sort illustrated here. Although I cannot pursue this complication in this book, the questions it raises are of theoretical interest and, I think, also of practical importance. I am very much indebted to the sociologist John David Skrentny of the University of California at San Diego for comments that stimulated me to think in this direction.
16. The social psychology literature abounds with evidence supporting this claim. See the still useful though somewhat dated survey by Nisbett and Ross (1980).
17. The Yale University economist Hanming Fang has nicely formalized this insight. See Fang 1998.
18. The writer Brent Staples reported that he employed precisely this technique to avoid being reacted to as though he were a criminal while walking the streets of Chicago’s Hyde Park neighborhood in the 1970s (Staples 1994).
1. “Natal alienation” here refers to the circumstance in which slaves do not enjoy unfettered relations with their forebears, offspring, or mates. Rather, these are relations over which masters exert ultimate control.
2. See Skrentny 1996 for an illuminating exploration of how the development of race-preferential policies has been shaped by social meanings attached to race.
3. This point is further illustrated by a recent debate among N.F.L. officials over a proposed ban on players wearing bandannas and stocking caps under their helmets during games. According to the New York Times (George 2001), Dennis Green, the (black) head coach of the Minnesota Vikings, argued in favor of the ban during a closed meeting of owners and coaches in the spring of 2001: “Green, who has not allowed Vikings players to wear bandannas or stocking caps in his 10 years in Minnesota, gave a speech about image, how he would not let his own child wear them and asked the group would they allow their children to wear them. He asked what the value in them was. If the image they projected is O.K., he asked, why do only a few of the league’s hundreds of black players wear them? Would they allow their daughters to date a man with a bandanna?”
Green’s position carried the day, with the owners voting 30–1 in favor of the ban (an exception was allowed for the skullcaps worn by some players to protect their scalps from abrasion). By way of explanation for the owners’ decision, the Times observed: “Bandannas and stocking caps are linked to the hip-hop rap culture and also to gangs. They are, too, simply a fashion statement for some. But when you have been where the N.F.L. has been recently—murder trials, rape trials and boorish player behavior—the league showed its concern by passing stringent sportsmanship rules at the meetings. Clean it up. From head to toe.” It is obvious that N.F.L. officials, black and white alike, are aware of the dire consequences for their financial interests that would ensue if the game’s image were besmirched by the taint of racial stigma.
4. These differentially evolving speech patterns have been studied extensively by the sociolinguist William Labov of the University of Pennsylvania.
5. Of course, the adoption issue is complicated by the fact that many blacks oppose interracial adoptions (and interracial marriages, for that matter). Nor should an individual’s decision (to adopt in Beijing rather than Baltimore, say) be taken as an accurate measure of that person’s racial sympathies. It is likely that some prefer adopting Asian over African-American babies because, given their understanding of the social meaning of race in our culture, the prospect of raising an Asian child through a turbulent adolescence is less daunting than is the same prospect with a black child. However, it is also likely that some adoptive parents feel “safer” with a child not born in the ghetto because of the stigma-driven concern that, despite effective medical screening, an inner-city child may be defective in some way. In any event, I want again to acknowledge that the problems of biased cognition go in both directions—blacks (social workers adamantly opposed to interracial adoption) can misperceive whites (prospective adopting parents). I would, however, venture the generalization that the consequences of biased social cognition for the achievement of a just public order are greatest when those with a distorted or incomplete vision of social reality, whatever may be their race, also control powerful levers of institutional discretion.
6. Documentation of the “hyper-segregation” of blacks in older American cities can be found in Massey and Denton 1993.
7. This story has been brilliantly told in the case of one American city by the University of Pennsylvania historian Thomas Sugrue. See Sugrue 1996.
8. Note that, whereas reward bias is purely an external matter (how disadvantaged persons are rewarded by outsiders for their expressed productivity), development bias may have both internal (to the group) and external sources. I will return to this point in Chapter 5.
9. Tilly 1998 captures precisely what I have in mind when calling attention to discrimination in contact. His parallel concept, “categorical inequality,” is described as follows: “Durable inequality among categories arises because people who control access to value-producing resources solve pressing organizational problems by means of categorical distinctions. Inadvertently or otherwise, those people set up systems of social closure, exclusion, and control. Multiple parties—not all of them powerful, some of them even victims of exploitation—then acquire stakes in those solutions … Through all of these variations, we discover and rediscover paired, recognized, organized, unequal categories such as black/white, male/female, married/unmarried, and citizen/noncitizen” (pp. 7–8).
1. See Mills 1997, a provocative but in my view ultimately unpersuasive book.
2. My argument here is influenced by the work of James Fishkin. In Fishkin 1983 he defines a “tri-lemma” for liberalism, insofar as it is committed at one and the same time to the ideals of equality of opportunity, reward according to merit, and autonomy of the family. He observes that autonomous but differentially endowed families will pass along differential developmental advantages to their children, some of whom, because rewards are distributed in reference to merit alone, will have superior life chances they have not earned, thus defeating the goal of achieving equal opportunity. A difficult choice, he concludes, must be made among these ideals.
3. This critique of liberalism is similar in spirit to the so-called communitarian arguments found in the work of Michael J. Sandel (1982) and Charles Taylor (1992), among others.
4. Thomas Sowell is perhaps the leading exponent of this view. See Sowell 1983.
5. Nozick 1974 provides a prototype of the procedural approach. I hasten to note that Nozick is himself aware of these difficulties. His “entitlement theory” of justice (ch. 7) consists of three parts: (1) an account of the just acquisition of holdings; (2) a theory of the just transfer of holdings; and (3) a principle of rectification which specifies how violations of requirements derived from (1) and (2) are to be handled. Thus Nozick’s full system, if it could be implemented (he offers only a sketch of a theory), would not be incomplete in the sense being criticized here.
6. The argument of this section draws on Loury 1998.
7. See Tonry 1995 for an extended critique of U.S. drug policy along precisely these lines, and for compelling evidence in support of the claim that U.S. drug policy has led to young blacks being imprisoned disproportionately.
8. Obviously, there are also benefits to blacks from enforcement of antidrug laws. This illustration is by no means intended to suggest that those benefits are slight. Taking them into account, and calculating the net impact of the policy on blacks as a group, would be entirely consistent with the spirit of the argument here.
9. Connerly 2000 provides an extended exposition of his views.
10. Some readers may object to my use here of the phrase “even some blacks.” A critic’s race is surely irrelevant, and my mention of it amounts to an ad hominem attack, one might hold. I think this objection is unfounded, and it may be useful to say why: If a critic’s race is used (even implicitly) to give authority to his or her argument, then he or she can hardly demand a race-blind evaluation of that argument. It is a fact about public life in the United States that the meaning of utterances—the sincerity or profundity of them—can depend on a speaker’s race. A black adherent of race-blind liberalism, for instance, by publicly dissenting from a position most blacks endorse, does more than state an opinion. He or she will be understood as having taken a principled stand, contra filial attachment. This posture, more so than its content, may be what gives the criticism its public currency. Ironically, black advocates of race-blindness are busy denying the relevance of race, even as their race helps to make their denials relevant. This observation does not refute their arguments, nor does it prove them guilty of “racial disloyalty.” But surely it is fair to take note of the irony.
11. Moskos and Butler 1996 documents this rationale for racial affirmative action in the U.S. Army.
12. Hopwood v. Texas, 1996, 78 F.3d 932.