Why Race Matters: A Preview

My main concerns in Race, as its title indicates, are the implications of genetic race differences, not their existence per se, so it was not strictly necessary for me to document these differences. It would have sufficed to ask, purely hypothetically, what follows if they exist. However, two incongruent yet related considerations forced a substantial discussion of the empirical issue. The first is that many people find the consequences of race differences so obvious that they only need to be convinced that blacks and whites do differ genetically to have read enough. At the same time there are many others who regard the possibility of race differences as too fantastic to be worth thinking about. For them, the topic has all the urgency of UFOs. To win their attention, it must be shown that genetic race differences are likely enough to take seriously.

The scientific material in Race is hardly original. Except for a statistical tweak or two, all the evidence for race differences cited in Race can be found in the standard literature of the subject. Since so few people seem to be listening, however, the facts bear repetition.

Race begins with a short exposition of the relevant biological and statistical concepts, particularly the distinction between phenotypes and genotypes. A phenotype is any trait of an organism: IQ, birth weight, and weight in adulthood are examples. An organisms genotype is the genetic basis of its phenotypes, the DNA that produce phenotypes in the environment( s) to which the organism is exposed.

Note the interaction of genotype with environment. Not only may different genotypes produce different phenotypes in the same environment, the same genotype may produce different phenotypes in different environments. Had you been born with the same genes but raised on rice instead of hamburger, your weight would probably differ. Liberal environmentalists make much of this interaction some going so far as to deny that talk of genetic input makes any sense but it also means that phenotypes, and phenotypic differences between individuals or groups, can be identified apart from their possible genetic basis. One need not know whether an observed race difference is genetic or environmental in origin to be sure it is real.

In particular, the evidence that blacks and whites² differ in phenotypic intelligence and motivation is overwhelming. For the eight decades during which IQ tests and related measures of mental ability have been given, white populations have consistently outscored black populations by slightly more than one standard deviation. Despite some talk of the IQ gap narrowing, the most recent studies continue to indicate that the full 1 SD difference emerges by age four.³

IQ tests are often said to measure only acculturation to white society, or to distort black intelligence because whites design them.⁴ If standardized tests picked up only knowledge of white culture, the questions most whites answer correctly the easy ones should be those querying aspects of their culture available to most whites (Who was Thomas Edison?), while harder questions should be those querying white culture's more obscure aspects (What is a niblick?) Yet questions easy for whites are consistently found to be relatively easy for blacks also, and the questions hardest for whites are those hardest for blacks, implying similar acculturation between blacks and whites, so the claim of cultural bias is surely incorrect.

Further evidence that IQ measures an intrinsic mental property manifest in a variety of ways are its associations with non- social variables. For instance, IQ correlates positively with brain size⁵ and efficiency of cerebral glucose metabolization,⁶ although it is unlikely that white thoughts enlarge the brain, or that white children are encouraged to slow the burning of sugar in their frontal lobes. These correlations have so far been established only for whites, but the techniques of modern neurology⁷ could be deployed tomorrow to search for race differences in brain function.

The rank- order and correlational evidence against test bias is indirect, but there is direct evidence. Logically speaking, a test for a trait is biased against blacks when a black must possess more of that trait to earn the same score that a white earns. (Tennis as a test of athletic ability is biased against people unfamiliar with the game, since only an outstanding athlete can play passable tennis the first time.) If IQ tests are biased, therefore, a black whose IQ measures (say) 110 should outperform 110- IQ whites on tasks with a large intellectual component, such as earning good grades in school. Yet standardized tests do not under-predict black performance on criterion tasks, and actually over-predict it⁸ that is, blacks with a given IQ or SAT score typically earn lower grades than whites with the same score. This anomaly suggests a weaker black achievement drive: whites on average try harder than blacks to reach long- term goals, and so whites reach those goals more often when pure cognitive ability is controlled for.

Before moving to the motivation issue, I should note a general point about intelligence stressed in Race: since there is no evidence that blacks are as able as whites, the egalitarian case for racial parity consists entirely of ad hoc conjectures⁹ and definitional objections directed against inegalitarianism. One favorite target is the word race,allegedly too vague for scientific employment. Race's response is simply to operationalize Negroid [i. e. black] as having mostly subSaharan African ancestors, and Caucasoid [i. e. white] as having European ancestors. Erstwhile racial differences in IQ can then be reformulated without loss of empirical content as IQ differences between individuals with different geographical ancestries, and hypotheses about genetic differences between the races can be reformulated similarly. Nothing is lost but a word.¹⁰

Another popular conceptual objection is that no unitary ability answers to the word intelligence. A statistical technique called factor analysis is used to extract a single factor,g, detected by all mental tests, but Stephen J. Gould and others dismiss g as a mathematical artifact. The unitariness question is rather technical, but the upshot is that it is irrelevant to all racial issues. All sides agree that individual variation in IQ can be analyzed as variation in g or in a cluster of more specific factors such as verbal ability and numerical reasoning. Yet so far as explaining black achievement goes, it makes no difference whether whites are (a) more intelligent than blacks or (b) more able verbally and better at numerical reasoning than blacks, for in either case the race differences in literacy, school achievement,participation in science and other socially significant outcomes result from race differences in mental activity rather than racism. Likewise, it does not matter whether genes produce a race difference in intelligence or race differences in verbal ability and numerical reasoning. Either way, once again, genes, rather than racism, explain the difference in cognitive performance and its social consequences. From the moral point of view, finally, whites are innocent whether genes cause a shortfall in black intelligence or shortfalls in a cluster of specific abilities collectively labeled intelligence.

The burden of proof, usually borne by inegalitarians, should be placed on egalitarians. Quite apart from any psychometric tests, blacks seem less intelligent than whites. Black children do far less well in school than white or Asian children; the black adults seen in ordinary life and on television news commit more crimes, parent more illegitimate children, have lower- paying jobs, and boast less- regular work histories. Indeed, the very absence of evidence for racial parity in intelligence, given that evidence would exist (and be publicized by egalitarians) were the races equally able, is evidence against it. The question is not why anyone thinks whites are more intelligent than blacks, but why anyone would think otherwise.

Race differences in motivation are anchored by Walter Mischel's classic study of Dominican children who, when asked to choose between a small candy bar at that moment or a larger bar a week hence, overwhelmingly chose the smaller one. So marked was the black preference for immediate gratification when compared to a matched white sample that Mischel called tests in statistical significance superfluous.¹¹

Orientation to the present is a persistent theme of studies of black behavior. Blacks watch almost 50% more television than whites and, holding income fixed, spend three times as much on movies. Blacks spend about as much on electronic games as whites, but far less on computers. Black scores on the Minnesota Multiphasic Personality Inventory are indicative of impulseridden fantasies.¹² Thomas Kochman reports that in turn- taking situations, such as classroom discussion, blacks are much more likely than whites to burst in out of order.¹³ This pattern of thought and behavior has been described as impulsiveness, lack of self- restraint, lower ego- dominance, and more rapid decay of reinforcement, but, as these terms contain implicit value judgments or causal hypotheses, Race favors the economist's neutral notion of time preference. An individuals time preference is measured by the money he would have to get tomorrow to forego $1 today, and represents the rate at which he discounts the future. In these terms, the mean phenotypic time- preference of blacks is higher than that of whites.

Also, contrary to pop sociology, black self- esteem is generally higher than white,¹⁴ and what is obvious in everyday life the black self- presentational style is more confrontational. Blacks also appear to be less cooperative than whites and less inclined to follow Kant's categorical imperative, popularly known as the golden rule. For instance, blacks are significantly more likely than whites to agree with the statement . It is not hard for me to ask help from my friends even though I cannot return the favor.. Kochman detects a black tendency to put self- expression ahead of the sensibilities of others, as when talking back to the screen in a movie theater: "With the shift in focus from doing unto others to doing for oneself, blacks can also act as their feelings direct without subsequent guilt." The best evidence for what Race calls the lower "Kantianism" of blacks is the black crime rate. Nobody, black or white, wants to be victimized by theft, assault or murder, yet blacks commit these acts far more frequently.

Defiant egotism has the same feel as impulsiveness the traits converge in refusal to wait ones turn but there are deeper links. Cooperation, reciprocal altruism, and kantianism originate in the benefits accruing to everyone when everyone pulls together. However, many cooperative situations present Prisoners Dilemmas: pulling along with everyone else is advantageous, but not as advantageous as pretending to pull while everyone else does the work. This being so, the only selfinterested reason not to cheat is the danger of being caught, barred from future cooperative ventures, and ending up worse off than if one had cooperated. But there is a catch: an individual will choose cooperation only if he can grasp the future losses jeopardized by cheating, and the time-discounted (dis)value of those expected losses exceeds his immediate gain from cheating. In other words, his intelligence must be sufficiently acute and his time- preference rate sufficiently low. Thus, the lower mean intelligence and higher mean time preferences of blacks explains their less cooperative and rule- bound ethos.

Race connects intelligence and time preference to free will via a difficulty in the standard exponential representation of discounting. (15) If you are indifferent between one unit of a good (say, G) today and n units of G tomorrow, then your daily discount factor is 1/ n. (We can think of 1/ n as being the inverse of the gross interest rate: someone with a personal interest rate of 10% has a gross interest rate of 110% principal plus interest. Equivalently, one with a personal interest rate of 10% has a discount factor of 1/ 1.1.) Thus, if 1/ n is the mean discount factor for whites and 1/ n' is the mean discount factor for blacks, then Race is arguing that n' > n.¹⁶

Still, weakness of will remains a puzzle on this analysis. How is it possible to choose a dispreferred alternative, as we all do whenever we eat a dessert we know we should avoid?

Weakness of will becomes more comprehensible in a model where time discounting is hyperbolic.¹⁷ In a hyperbolic timediscounting model, the perceived value of G in the future depends on how much one values a G in the hand, and on a parameter I, known as the "coefficient of impulsivity." To relate this back to the standard model, think of I as being similar to n: the higher the value of I, the lower the weight one gives to the future at any fixed level of current holdings of G. The argument in Race is, therefore, that the mean coefficient I for whites is smaller than the mean coefficient I' for blacks.¹⁸

So if I is the mean white coefficient, and I' is the mean black coefficient, Race argues that I' > I. On this analysis, an agent may prefer G in the hand to a smaller G' in the hand, but, if G is farther in the future than G', the perceived value of G' can temporarily overtake that of G.

Consider the following illustration. Entering a restaurant at 7: 00, I prefer skipping my dessert at 8: 00 to having a dessert at 8: 00. I know at 9: 00, I will be more pleased if I have skipped the dessert than if I have succumbed. However, once my main course has come and gone and the waiter returns to offer me a dessert, my situation changes. If it is now 7: 45, I no longer consider what would have made me happier at 7: 00. That time period has come and gone, and now I only care about what pleases me now and in the future. When the waiter asks if I should like him to bring me the cheesecake, my answer depends on what pleases me now and what will please me later; the more heavily I discount the future, the more I will focus on what pleases me now. I refer to the time period when the short- term pleasure of the dessert tempts me as my "period of weakness." I have weakened despite having remembered what I had told myself at 7: 00, and knowing that I will feel indigestion at 9: 00. It is during this interval that I believe myself not in control, not fully autonomous.

Now, all else equal, larger coefficients of impulsivity will bring about longer periods of weakness. The future becomes more heavily discounted, so there is more room for a discrepancy between the future decision which I initially plan to make (avoiding the dessert) and the choice I actually make in the future (when the waiter arrives with the dessert menu). The argument in Race, that the coefficients of impulsivity are greater for blacks than for whites, would therefore imply that blacks will suffer longer and more frequent such episodes.

The fact that behavior during periods of weakness is less than free is explained by John Locke's insight that free action must not only express preferences all action, even of children, animals and madmen, does that but that the preferences expressed must meet the agents approval. Race expounds a conception of free will according to which an agent is free when he is able to step back, examine his desires, and act on those desires he prefers to act on.¹⁹ This is why choosing a prospectively dispreferred but transiently attractive alternative is not free; I did what I wanted, but not what I wanted to want. But a more significant corollary is that freedom increases with reflectiveness and self- restraint. With insight into his own motives, the intelligent man can anticipate preferences he does not wish to act on, and thereby take steps to forestall their becoming dominant. Cognizant that he will be tempted by dessert, he may drink a glass of water, or shame himself into declining dessert by assuring his companions beforehand that he will. Self- restraint, for its part, gives a man time to contemplate the consequences of yielding to questionable impulses.

Race concludes that blacks are, on average, less free and responsible than whites. Further, it argues, awareness of this disparity in autonomy explains why liberal fretfulness about the root cause of black crime never matched by fretting about root cause of stock fraud, or lynching, or the Holocaust. Liberals, without quite admitting it even to themselves, think of black crime as akin to an amoral natural force. This awareness also explains liberal lenience, since penalizing black criminals seems to punish them for what they cannot help. In this, we will shortly see, liberals are wrong.

Although their self- esteem is higher, blacks consider themselves more powerless than whites.²⁰ At the same time, perception of powerlessness is associated with loss of self- esteem among whites. This discrepancy offers some empirical support for the greater heteronomy of blacks, for it can be interpreted as showing that blacks regard the self as less a "locus of control" than whites do.²¹ This is turn makes sense if, for blacks, the self is a less significant mediator of behavior.

Having discussed phenotype race differences, I now turn my attention to the evidence for genetic causes of those differences. I group the evidence into seven categories.

1) The high interindividual heritability of IQ and personality

Holding group differences to one side, twin studies indicate that genetic rather than environmental variation explains most of the difference in intelligence between individuals, and much of their personality differences. A pair of randomly selected individuals will differ in IQ by about 17 points, but identical twins reared apart typically differ by about six. Identicals raised apart are also highly similar on dozens of personality traits, including impulsivity. There are ad hoc environmentalist explanations for some of these concordances,²² but only genetic factors seem able to explain them all.

At this juncture environmentalists can be relied on to insist that a genetic account of individual differences does not entail a genetic account of group differences. After all, genetically identical seeds will yield plants of different heights in different soil. This is true enough, but it is also clear that some mean group differences are too large to be explained plausibly by any environmental divergence. Bonsai might all be undernourished Redwoods, but it is not very likely. Numbers crunched in Race show how unlikely it is that differences in black and white environments have produced the racial IQ gap. Speaking qualitatively, this conclusion is very natural. American whites and blacks would seem to inhabit quite similar environments they speak the same language, watch the same TV shows and movies, attend schools with identical curricula, vote in the same elections. Black slums are of course less salubrious than white suburbs, but since a neighborhood is created by its residents, the disarray in black slums counts as an effect rather than a cause of race differences. (Race discusses such gene/environment correlation at length.) Indeed, the fixity of the IQ gap during a century that has seen black and white environments converge leaves genes the only remaining source of variance.

2) Early appearance of race differences.

The appearance of the IQ gap before age four rules out most environmental causes: low teacher expectation cannot affect the development of a black infant still in his cradle. One environmentalist hypothesis consistent with the early onset of race differences is poor nutrition for black children, but this must be rejected on other grounds: (1) black babies are physically more precocious than white; (2) Department of Agriculture surveys indicate that black children do get the minimum requirement of vitamins and minerals; and (3) blacks dominate sports from early youth on. If, however, there is truth in the claims that black children receive poor nutrition, then the failure of black mothers to feed their children properly in a society inundated with messages about good health may itself indicate lower parental intelligence, which might then be passed to offspring.

3) Physiological race differences

IQ correlates with brain size, and, controlling for body size, black brains are about 4% less massive than white.²³ To argue that brain size means one thing for whites and another for blacks would countenance race differences in the functional organization of the brain, a position as biological as genetic differences, and presumably as repugnant to environmentalists. In any event, the IQ/brain size correlation appears to hold between as well as within races, since black and white head sizes hence, inferentially, brain sizes are identical when IQ is controlled for. Although I have heard environmentalists conjecture that racism shrinks black brains, this finding is difficult to explain non- genetically, since eventual adult brain size is fixed by early childhood.

With regard to temperament, young black males exhibit relatively high levels of serum testosterone, a mediator of aggression and libido. A race difference has also been found in a gene controlling alcoholism.

4) Transracial adoption

Sandra Scarr and her colleagues tracked over one hundred black and mixed- race infants adopted in the late 1960s by upper middle- class white families. The mean IQ of the black adoptees at age 7 was 97, but had fallen to 89 by age 17. More significantly, at both junctures the IQs of the black cohort remained a standard deviation behind those of the birth children and the white adoptees of the adoptive families. Scarr takes these figures to show the persistence of racism in the surrounding culture, but she does not specify what racist influences the adoptive families failed to block. A genetic analysis is obviously simpler.

Incidentally, the decline in black IQ with age in the adoption study is consistent with the known increase in the heritability of intelligence across the life span. The currently accepted theory is that individuals increasingly shape their own environments as they mature, so that their environments come increasingly to reinforce innate tendencies. The parents of a genetically dull child may read to and otherwise stimulate him to a reasonable level of school achievement, but as an adult on his own he will cease reading and choose nonintellectual companions, leading to less- impressive intellectual performance.

5) Failure of Intervention

Like transracial adoption, early intervention programs such as Head Start control for many of the environmental variables to which lower black intelligence is commonly attributed. Facilities designed by highly motivated psychologists to stimulate children as young as six months are not likely to omit any factor crucial for mental development. Yet these programs have produced no lasting gains in the IQs or academic achievements of black children (although Head Start has produced small but measurable gains in the academic achievement of white children).

6) Achievement of other Minorities

Despite functioning in more deprived environments, many groups, including American Indians, outscore blacks on IQ tests, particularly on the least culture-bound items. Jews, persecuted more viciously than blacks for many centuries, have consistently produced a disproportionate number of scholars wherever they have lived. Chinese and Japanese in the US, two highly successful groups, are as recognizably non- white as blacks, undercutting the hypothesis that black failure is related to identifiability. Explaining the differential success of various groups in terms of their cultural values is circular, as it begs the question of why different groups inculcate different values.

7) Africa

There was no white racism in Africa during the millennia that blacks have dominated that continent, yet the mean IQ of African blacks is considerably lower than that of American blacks. Again contradicting conventional wisdom, Race crunches some more numbers to show that the 100,000+ years during which Africans and Eurasians evolved apart was enough to permit the African/ Caucasoid divergence.

No sub- Saharan society has ever developed mathematics, a written language, formal educational institutions, or the wheel, while white and Asian societies have done so many times. Once again, attribution of this failure to culture is circular, begging as it does the question of why whites and Asians, but not blacks, built cultures in which these indicia of civilization appeared. Race takes the strongly individualistic line that culture and society are not independent causal factors shaping their members, but effects of their members. individual properties effects, that is, of phenotypes expressing underlying genotypes in ambient physical environments. Cultural attainments thus mirror innate factors.

Thus, looking at these seven categories of genetic data, to say that blacks are genetically less intelligent and more impulsive than whites is not racist, since racism is by definition bad, and facts, however unwelcome, are morally neutral. Mention of genes and race in the same breath triggers hysteria not because it is a reminder of Hitler (nobody can seriously fear a Nazi resurgence),but, Race proposes, because genes determine personal identity. The reader is asked first to imagine his zygote, with his genes, born into an environment unlike his natal one and becoming an adult with phenotypes unlike his; I suspect he will agree that the adult described would still be him. The reader is next asked to imagine a zygote with different genes somehow developing into an adult with precisely his present phenotypes; I suspect he will agree that that being would nonetheless not be him. Judgments of identity thus track genotypes, making genetic but not environmental factors essential to the self. For this reason, genetic deficiencies are seen as essential, and talk of them consequently more wounding, than deficiencies imposed by the environment.

Before it turns to any concrete implications of the race differences, Race devotes two chapters to the topic of norms. My concern that this material might pull the discussion too far toward pure philosophy was overridden by several worries that prompted its inclusion.

For one thing, I wished to make clear that no empirical facts about race imply that whites are better than blacks, a judgment so often imputed to hereditarians that only a full airing of the issue of value can put the imputation to rest. To this end Race presents a resolutely "naturalistic," non- realist view of values. There being no empirical phenomena requiring objective value for their explanation, we have no more reason to believe in it than to believe in phlogiston, or the little man that wasn't there. A God's-eye view favors neither high nor low time preferences; investing $150 may be more prudent than spending it on running shoes, but it is not inherently better. Greater intelligence is likewise no better or worse than or equal in value to less intelligence. The mean intelligence levels of whites and blacks were adaptations to selectional pressures at work in Africa and Eurasia, just as the lion's strength and the gazelle's speed are evolved responses to selectional pressures in their niches. And just as the lion's talons are neither better nor worse than the gazelles speed each creature simply is what it is whites are not better or worse than blacks.

Race is similarly neutral toward morality itself. An individual's "moral" values are construed as those of his preferences that he wants everyone to adopt (and wants everyone to want everyone to adopt); and a group's morality is the set of moral values shared by most of its members. Given the advantages of mutual trust and the temptations to exploit trust by lying, it is likely that every group reinforces truthfulness to some extent. However, groups may differ in the moral values they adopt and the intensity with which they reinforce them. Moral codes, too, are adaptations over evolutionary time to differing environmental conditions, none better (or healthier) than any other. In particular, it is suggested, individuals had a better chance of surviving by their solitary efforts in the milder sub- Saharan climate than in Eurasia, leading to less intense selection for cooperativeness and reciprocity. That is why blacks appear to be less cooperative or committed to the golden rule than whites. Strictly speaking, the black/ white difference is not so much a divergence in morality - in the rules blacks and whites universally prescribe - as in concern for morality itself. The notion of a moral value has universality built into it by definition, so relative indifference to the golden rule amounts to greater amorality. But once again, amorality for the naturalist is no better or worse than morality. Wolves are amoral, but there are no grounds for judging a world with moral agents but without wolves better or worse than one with wolves but lacking moral agents.

It is proposed that this evolved divergence in morality helps explain the curious racial guilt felt by whites who have done nothing to blacks, and the deference of many whites to the most outlandish black demands. Whites, having evolved to interact with other whites, interpret black signals to mean what they would mean coming from whites. Whites are programmed to interpret and respond to black indignation as if it signaled the more serious injuries that would elicit that level of indignation from whites, so constantly overestimate the seriousness of black injuries.

Race insists that explaining honesty and reciprocity by their inclusive genetic fitness²⁴ does not reduce morality to disguised selfishness. Inclusive fitness theory, despite much bad press, does not deny that people are morally motivated to play fair and sometimes make sacrifices. It claims, rather, that maximization of genetic fitness is why these motives exist. A tendency to cooperate in Prisoner's Dilemmas and even sacrifice oneself for others is an evolutionarily stable strategy, but the motive of cooperation selected is nonreductively unselfish.

Skeptical naturalism also raises the practically pressing worry of how to draw normative consequences from race differences. Can a skeptic criticize affirmative action or endorse freedom of association if neither is intrinsically good or bad? Race's answer exploits two old philosophical saws: (a) Much apparently normative disagreement is actually disguised factual disagreement, and (b) all genuinely normative suasion assumes some prior agreement on normative principles. Option (a) lets naturalists participate in the many moral disputes resoluble by empirical investigation: whether Jones owes Smith damages for his broken leg, for instance, depends on whether Jones did in fact break it, a factual issue the naturalist may address. In the case of essentially normative disputes, option (b) counsels the naturalist to bring his interlocutor around by appealing to his interlocutor's values. Show a man that your view is implicit in his commitments, and he must accept your view on pain of inconsistency. Such an appeal is no weaker for being ad hominem, since everyone thinks his own values are correct.

In light of this, the pragmatically most fruitful perspective from which to discuss race is that of my reader who, I assume, accepts Caucasoid values. After seeking to epitomize Caucasoid values (they revolve, I say, around the golden rule), Race applies them to a cluster of issues at the heart of America's race problem.

Those few on the Right (Herrnstein and Murray²⁵) or Left (Christopher Jencks²⁶) willing to discuss psychological race differences say their origin does not matter. One aim of the pivotal chapter, Biology and Justice, is to show how issues of racial justice hinge on whether these differences are biological in origin.

The quota debate is patently about compensatory justice, since quotas are meant to redress injuries said to have been inflicted on blacks by whites. Lyndon Johnson introduced affirmative action for government contractors via the analogy of a man released from shackles required to run a foot race, and who (Johnson reasoned) deserves a head start to make up for his unjust handicap.

Many quota advocates say they reject the redress rationale. Johnson's metaphor has lost its vigor as it has been realized that an edge for the lame runner cheats his competitors, who are in no way responsible for his plight. In literal terms, affirmative action, particularly when state- imposed, is recognized as prima facie unfair to whites who never discriminated.²⁷ To dodge this objection, defenders of affirmative action now say it is necessary to create role models or prevent renewed discrimination. Ronald Dworkin offers an elaborate rationale based on a distinction between personal and external preferences. Race spends a number of pages arguing that virtually all of these rationales tacitly rely on compensation, or share the compensatory premise that the race gap in attainment is due to harm done to blacks by whites. The compensation rationale and its many avatars are no firmer than the causal premise.

Ironically, most critics of affirmative action also leave this premise unchallenged, perhaps in hope of avoiding the epithet "racist." Affirmative action is bad, they say, because it stigmatizes blacks, or costs too much, or replaces equality of opportunity with equality of result. There is something to many of these criticisms, but Race takes pains to show why, by missing the logical heart of the matter, they are also rhetorically ineffective. The causal question is all.

A propos cost, it is curious that redress for racial wrongdoing is always said to require placing blacks where they should have been, even though their inferior qualifications incurs great inefficiency. After all, redress is normally constrained by feasibility: when literal restoration of the complainant to his uninjured condition would be too disruptive, an equivalent is thought sufficient. A dancer who loses his foot in an accident asks the earnings the accident cost him, not the right to be a onelegged King Kastchei. Race speculates as to why champions of redress accept no cheaper substitutes, such as lump-sum reparation payments, for compensatory hiring.

Because affirmative action rests on a theory of the cause of the racial attainment gap, Race argues, only the race differences in intelligence and motivation, which explain the gap more plausibly than racism, offers a convincing case against it. Negatively, racial discrimination cannot explain black failure because there is not enough of it. Slavery ended over 130 years ago; segregation in the public schools ended over 40. Private discrimination has been a federal tort since 1964, and government- sponsored preference for blacks has pervaded American society since 1970. Whites frequently favor black candidates in elections. Positively, the IQ gap predicts the contours of black failure with remarkable accuracy. For instance, according to the IQ data, only about 13 blacks in any one- year cohort are as intelligent as the average mathematician, and in fact, blacks earn only about a half- dozen doctorates in the mathematical sciences annually, a impressive fit given the small numbers involved. Black over- representation in classes for the mentally retarded falls right out of the IQ numbers. As one might expect in a quota- ridden society, blacks are over- represented in prestigious areas when IQ is fixed.²⁸ There is no quantitative data on the contribution of impulsivity to the attainment gap, but Race cites many commentators on the low level of black entrepreneurship and willingness to work.

That the attainment gap is explained by race differences in phenotypic intelligence and motivation allows heads- up compensation theorists to reply (as some have) that this phenotypic variance is itself caused by racism. Blacks don't try because trying is pointless when the deck is stacked against them. And were racism thus indirectly responsible for the attainment gap, the compensation argument would reenter at one remove, with blacks deserving redress for their wrongfully caused dysfunctional traits. That is why letting the issue rest at the phenotypic level is inconclusive. The case for reparation can be judged only by looking at whether racism offers a better account than biology of the lower IQs and higher time preferences of blacks. The topic of genes is unavoidable, Race insists. No matter what the black shortcoming, some will insist that racism is its cause, or the cause of its cause, or the cause of that cause. Sooner or later, the genetic question must be faced.

In fact - this theme permeates Race - play of the gene card, far from being a gratuitous swipe at blacks, has been forced on defenders of justice by the constant diabolization of whites. It is impossible to be silent when silence amounts to an admission of guilt. When Smith limps into court, berates Jones for breaking his leg, and demands damages, Smith must be prepared to hear Jones deny the charge. Smith has opened the door to alternative hypotheses about the cause of his deformity, for instance that it runs in his family, and he must be prepared to face them. Smith cannot accuse Jones and then call him tactless for pleading innocent - exactly what liberals do when they blame whites for black woes, then call whites who deny the charge insensitive and try to silence them with speech codes.

Actually, Race notes two commonly neglected reasons why genetic arguments may not be strictly necessary to rebut the compensation argument. First, the suffering of the North and the non- slave- holding South during a Civil War fought to give black slaves their freedom may have made further white sacrifices superfluous (i.e., non- slaveholding whites have already done more than their fair share). Second, subsequent black behavior may have canceled any remaining debt. The high black crime rate and the preference of black offenders for white victims have subjected whites to many more murders and robberies than would have been inflicted by a white subpopulation of equal size. In addition, blacks have marred or destroyed enormous stretches of real property created by whites, as is obvious from a tour of any large black neighborhood. On balance, blacks may owe whites compensation.

Race pursues the issue of justice beyond compensation, to the distributive fairness of racial inequalities. Many writers simply assume that "we" (unspecified) should keep trying environmental interventions in hope of finding one that raises black IQs to white levels.

The obvious objection is that such regimens may not exist. Champions of "let's keep trying" treat what is possibly possible as if it were actual, whereas the racial gap may be, like differences in eye color, ontogenetically fixed. But there are deeper problems. Whites alone can afford to fund efforts to raise black IQs, and it is not obvious, once compensation is set aside, why whites have any duty to channel resources to that end. Virtually no one believes that equality is good per se.²⁹ The environments of black children are less stimulating than those of white children, but, Race insists, the distributive baseline for where American blacks "should" be, namely the sub- Saharan African milieu they would have occupied had they never encountered whites, is far less stimulating than a plantation or a segregated school in the Jim Crow south, to say nothing of modern schools. Perhaps everyone has an equal right to resources when they are literally the unprocessed bounty of nature, but Caucasoid morality decrees that wealth created by human hands belongs to the hands that made it.

Whereas it is always clear who should finance correction of a compensatory injustice,namely the tort- feasor or his legatees, appeals to distributive justice obscure the cost question. In fact, may proponents of distributive justice, like John Rawls, explicitly divorce distributive rights to wealth from contribution to it, and are duly chastised in Race for this. Critics of inequality, like Jonathan Kozol, who complain of more money being spent on predominantly white than predominantly black schools are not merely uninformed, although they are that - more is now spent per capita on black children than on white children in the nation's public schools because blacks need more special education and counseling. Worse, these critics are morally obtuse: there would be nothing wrong with white children receiving more resources, given that these resources are controlled almost exclusively by white parents. Per capita, blacks pay about half the income tax whites do, and 20% of the property taxes, the main source of school funding. Statistics cited in Race show Southern whites during the Jim Crow era paying $15.60 for black schools for every $1 paid by blacks. It takes considerably imaginative effort to appreciate it, but the segregated schools of the American South expressed white generosity,and integration has given blacks access to schools, with their labs and computers, even further beyond what blacks could create by their own efforts.

The chapter on justice ends with glimpses at utilitarianism and Rawls's Difference Principle, that permissible inequalities must benefit the worst off. While neither is consistent with ordinary conceptions of justice, it is noteworthy that utilitarian bang- for- the- buck calculations and (rather surprisingly) the Difference Principle may both steer resources away from blacks. Over time, the average well- being and the prospects of the worst off have been most markedly lifted by the inventiveness of the intellectual elite. Directing contested resources to society's most gifted predominantly white and Asian students may thus help everyone, blacks included, more than directing those same resources to blacks, a conjecture confirmed by the improved white performance unmatched by improved black performance returned on the investment in Head Start.

Race begins its discussion of crime with data showing that blacks commit violent felonies at a rate ten times that of whites, and decidedly prefer white victims. Despite the enormous publicity given to any white- on- black felony, a black is on average about twenty- five times more likely to kill a white than vice- versa. The question is what responses are appropriate.

Race divides the question into private and public. Its view of appropriate private measures, not shocking to libertarians but scandalous in many quarters, is that individuals have a right to consider race when assessing risk, and to be especially careful when dealing with blacks. It is rational for a man on a deserted street to be more leery of a black heading his way than he would be of a white, and to cross the street or turn around. It is rational for a cabbie to refuse to pick up young black males. Unfortunately, such permissible avoidance is now often illegal. Cabbies who refuse black fares lose their licenses.

Libertarians will wonder why a right to avoid blacks needs any defense at all, since it falls under voluntary association and self- defense. Here emerges a theme prominent in the last third of Race: the rationality of exercising rights race- consciously. Americans respect freedom of association, but they also want assurance that it is not being abused. One reason for passage of the Civil Rights Act, the Fair Housing Act, and other infringements of liberty has been the conviction that a wish not to hire or otherwise associate with blacks must be absurd, vicious, or insane. This conviction has forced friends of liberty to say, in effect, "There is no good reason to avoid blacks, but people have a right to do so anyway," which, while perfectly consistent (a right is a right, after all), leaves non- libertarians dissatisfied. The statistics on race and crime show that avoidance of blacks is perfectly rational.

A right to flee danger is not the only truism defended at inordinate length in this context; several abstruse issues about statistical inference and risk are analyzed to prove that, since blacks commit more crime than whites, an arbitrarily chosen black is more apt to be a criminal than is an arbitrarily chosen white. Where race is concerned, people will deny what they elsewhere find obvious, necessitating argumentative overkill.

Libertarians may join liberals in rejecting race- conscious state enforcement of rights against aggression, but I argue that this too is permissible.³⁰ For instance, a policeman may frisk a young black male under circumstances in which he may not frisk a young white male, on the statistical grounds that the black is more likely to be up to no good.

Race's preliminary argument for this conclusion is the precedent offered by suspect profiles for interdicting smugglers and other statistical algorithms used by the state to deter crime. The central argument is appeal to the Hobbes/ Locke theory of the state. While Locke and Hobbes of course disagree about the existence of natural rights, they are at one that the purpose of government is to guarantee security; the state exists to prevent aggression. Moreover, this function is normally understood to allow some preemptive action against merely potential aggressors. I may disarm a man gesturing menacingly with a gun, and the police, my agents, may detain him. Since race is a valid predictor of threat, the state may consider race in pursuing its preemptive function. I emphasize that race- based screening meets the Supreme Court's "strict scrutiny" standard, which permits only those racial classifications necessary to achieve a compelling government interest. Controlling aggression is the state's raison d'être, a compelling interest if any is, and crime statistics show that attention to race may be necessary for its discharge.

An aside in this context about Nozick's Anarchy, State and Utopia may interest libertarians. My defense of race- based screening is patently "rights- maximizing," permitting aggression against possible innocents to minimize aggression against innocents overall. (I have argued earlier that ordinary Caucasoid morality maximizes rights in extreme cases.) As is well known, Nozick claims to reject such utilitarianism of rights in favor of a nearly absolute prohibition against aggression. I note, however, that his own derivation of the state is also rights- maximizing, for it lets the dominant voluntary protective association impose itself on independents because of the anxiety independents create for association subscribers. That a deontologist such as Nozick should find himself relying on minimizing intuitions does not vindicate race- conscious screening, but it does show how deeply such intuitions inform the ordinary concept of the state.

As the root cause of black crime seems to be so urgent a problem to so many, evidence is reprised that black crime is caused by the relatively low genetic intelligence, self- control and kantianism of blacks.³¹ An important misconception addressed here is that the sharp increase in black crime over the last sixty years, presumably unaccompanied by any major genetic change, proves that black crime is environmental in origin. Relaxed sanctions against wrongdoing have certainly facilitated an increase in black crime, so the cause of the increase in black crime in particular may be deemed environmental. However, this easing of sanctions has produced no corresponding increase in white crime. What is more, the net overall change in the criminal justice system has been toward an increasing similarity (and lenity) in the treatment of blacks and whites. Divergent responses to a similar environment imply genetic differentiation, so the post- 1960 rise in black crime together with the stability of white crime makes the race difference in crime rates likely to be genetic.

Tracing black crime to genetic factors leads back to free will and responsibility. Race defends the view that, while the lower mean level of responsibility of blacks makes them by definition less punishable than whites, it permits other measures to control black crime. After all, homicidal maniacs, although not punishable, are not let roam free. Race suggests some possible deterrent measures, including swifter imposition of harsher sanctions on blacks than whites for the same infraction, in light of higher black time preferences, and (in light of the apparent more rapid maturation of blacks) treatment of blacks as adult offenders at an earlier age than whites. It is also noted that if white mistreatment has so affected the psyches of blacks that they are not accountable for their actions, as some apologists assert, blacks must also forego civil privileges associated with full rationality, such as voting and jury service.

The chapter on crime seeks above all to discredit the idea that black crime is an excusable payback for white misdeeds, an idea which encourages black criminals and weakens the will to resist them. Excuses for black crime would be equally poisonous even if blacks have been as mistreated as liberals believe, Race concludes, because no society can function with one group exempt from rules against theft and violence.

By now, many readers of Race will have wondered at my willingness, as in the case of selective police attention, to let people be judged on the basis of their race. Does not Caucasoid morality demand that each person be judged as an individual? Arguably the issue of individualism should have been faced at the outset, but it is best discussed in light of the facts and distinctions accumulated in the meanwhile. I conclude that the principle of individualism as conventionally formulated is preposterous, and when made plausible it allows race consciousness.

I stress at once that the principle of individualism is unavailable to supporters of quotas, who tend to trumpet it in other contexts. Every rationale for affirmative action is shown to classify by race and draw race- based statistical inferences about individuals.³² This is not merely hypocrisy on the part of quota enthusiasts; it reflects their recognition that all judgments, all descriptions, categorize. Calling Jones optimistic, surely an individual trait, groups him with all the other optimists. Affirmative action errs not in grouping all whites together, but in grouping incorrectly. It would be perfectly proper to reason . Most whites profit from oppression; Jones is white; so Jones probably profits from oppression. were the generalization true; affirmative action is wrong because the generalization is false.

People cannot be judged by the traits that make them unique. Attempting to do so would produce such absurdities as friendships and hiring decisions based on fingerprints, or on precise places and times of birth. The shibboleth of individualism means, if anything, judging people by their important traits, which may not be unique - Jones's optimism properly influences how he is treated, but Jones is not the lone optimist - and anyway leaves important undefined. At the end of the day, Race argues, important traits are either those valued for their own sake, or (what is more relevant in the racial context) those that are predictively valid. Optimism is part of "who someone is" because it tells others something about his sense of humor, behavior under pressure, and overall personality. By this standard, a person's race also is important, for it supports numerous reasonable expectations. Taking someone's race into account is thus consistent with "treating him as an individual." It classifies him with all other members of his race, true, but noticing someone's optimism classifies him with all the other optimists. Of course, one should when possible look at more than race, just as one should try to learn as much about any individual as is relevant to the evaluation at hand. But sometimes learning more isn't possible, as when you are deciding whether to proceed ahead on the running track, or what to do when your neighborhood school becomes predominantly black. In any case, it caricatures awareness of a person's race to equate it with disregard of all his other traits.³³

An almost- reflexive objection to race- consciousness is that race is involuntary, and that it is unfair to judge people by traits they cannot help. Yet most everyday treatment of people is based on involuntary, immutable traits. Mates are chosen because of their appearance and personality, neither of which are chosen. Personnel decisions are task- driven, and there is no reason to expect the ability to perform a valued task to be voluntary. Managers pick the rookie infielder with the fastest reflex is, even though reflexes are not subject to the will, because infielders must snag line drives. Conversely, many voluntary traits are irrelevant to how individuals should be treated. For example, the Admissions Committee for a law school will be unmoved by an applicant's beer can collection.

Race speculates that the artificial link between voluntariness and importance was forged mainly by the civilrights movement. Once it was decided to ban racial discrimination, some characteristic of race was needed to justify the ban; since race is involuntary, discrimination based on involuntary traits was proscribed (a proscription that subsequently rationalized banning free association based on sex, handicap and many another trait). The link has been sustained by the perception of all goods as rewards; since only voluntary behavior merits reward or punishment, goods like employment and housing (it then seems) must not be allotted by involuntary attributes like race. In truth, very few goods are bestowed as rewards. A rookie's place on the team is not a prize for quick reflexes, however much it feels that way to him, but an amoral management decision dictated by the goal of a winning team, dictated in its turn by the need to attract paying fans.

If race- consciousness is permissible and rational, what of action based on it, i.e. discrimination? Focusing first on the private sphere, Race distinguishes positive discrimination harming someone because of his race from negative discrimination, the refusal to benefit someone because of his race. Positive discrimination is indeed wrong, but because of its aggressive rather than discriminatory character. Negative discrimination is not wrong at all, although it can be made to seem so when represented as harm, for instance by describing a black turned away from a job as a "victim of discrimination." However, an applicant denied a job whether because of credentials, his race, or pure whim, is no worse off than he was prior to the refusal, hence not injured.

Libertarians will once again wonder at the fuss over the obvious, here the right to refuse an association for any reason, or no reason at all. Yes, Race acknowledges this right (deducing it from the categorical imperative); the deeper question it asks is why so many Americans have surrendered it, and taken it from those not willing to surrender it, in the name of civil rights. A couple of answers have already been broached. The assumption that blacks are just like whites, or would be but for discrimination, makes refusal to deal with blacks appear wholly irrational. By this point, I hope, readers of Race will be ready to agree that race differences in intelligence, behavior, and Kantianism show reluctance to deal with blacks to be quite rational. A second answer, it is argued, is that the perception of refusal to benefit blacks as coercing them is traceable to blacks' own limitations. As blacks could not enjoy Caucasian technology TVs, multi- story dwellings, automobiles unless whites sold or gave it to them, white refusal to do so appears to keep blacks from these goods. Race tries to make clear that refusal to give or sell someone something he cannot make himself leaves his productive capacities undiminished, so does not harm him.

What of public discrimination? Having already argued that the state may act race- consciously to protect rights against attack, Race suggests that race- conscious welfare policies may be in order. While all welfare states may self- destruct sooner or later, as an increasing number of people batten off the public treasury, most whites, at least so far, prefer working and parenting children within marriage over the seductions of welfare. Consequently, a limited safety net might be stable in an all- white population. But the very different response of blacks to welfare incentives, including an illegitimacy rate nearing 70%, is by now familiar, making a safety net for blacks clearly unstable. If we must have welfare, Race suggests, blacks should be denied it or held to more stringent criteria. A general caution for libertarians is that institutions which don't work in black or multiracial populations (public education is an example) need not, by that token, be intrinsically flawed. The problem may not be the institution, but the population.

Race contends that the "equal protection" clause of the Fourteenth Amendment is consistent with state raceconsciousness. The Fourteenth Amendment excludes only racial classifications irrelevant to any vital government function. (Inborn, involuntary and immutable traits might well be relevant, as when the state denies driver's licenses to the congenitally blind.) The Constitutional powers- that- be have decided that compensating blacks is a sufficiently central government function to support laws burdening whites. Race presses the logical symmetry of racial classifications burdening blacks.

Instead of ending Race with a conventional authorial compendium of conclusions, I asked myself what I would want an American President to say in a major speech about race. That speech concludes Race.

The chief message of my imaginary President is honesty. The facts about race must be faced unflinchingly, and whites must stop being blamed for everything. More concretely, my imaginary President lays out three basic approaches to race differences.

Minimizing envisions large- scale social engineering daycare, childhood enrichment, racial preferences, job training to reduce the race gap as much as is possible. This approach is not categorically ruled out, but the President makes clear that it will generate legitimate white resentment and probably not work even on its own terms. Eugenic measures such as tax- funded incentives to encourage the more intelligent to reproduce are also rejected as taking too long, and not being the state's business.

The second approach, control, would seek to reduce disruptive black deviation from white norms by instilling self- discipline, respect for law, and the work ethic. It would ease the regulatory burden on black businesses, restrict or eliminate welfare, and impose swifter and more effective punishment for crime. Its downside is infringement of civil liberties and possible creation of civil disorder.

The laissez- faire approach is the policy, or non- policy, of leaving everything to the market. It would end all forms of income transfer on moral as well as practical grounds, repeal all laws against private discrimination, and abolish the minimum wage, allowing the many blacks with limited skills to find gainful employment. A small government would operate virtually race- neutrally on the economically efficient basis of merit. The market has the great advantages of not aiming at any pre- set outcome which biology may have put out of reach, and of imposing discipline without coercion. My imaginary President does not explicitly endorse laissez faire (he is, after all, President of all the people, including liberal statists), but his preference is unmistakable.

One might think that acknowledging biological race differences entails enhanced private and public race consciousness, but a suggested alternative is "realistic race blindness": racial criteria should not be used in making public or private decisions, but the disparate racial impact of almost any decision is to be anticipated and disregarded as a natural consequence of race differences. A good idea apart from its racial impact is a good idea- period. If, for instance, it is pedagogically wise to track schoolchildren by ability, and the only reason not to do so is white and Asian dominance of the fast groups and black over- representation in the slow groups, then children should be tracked. However, it might be appropriate to weigh the racial impact of policies whose effect on any specific individual is unpredictable, but to which the races can be expected on average to react in significantly different ways. The President warns that welfare and penology may demand "realistic race conscious."

Mindful of the radical (by today's lights) character of some of his proposals, he cites Claudius's precept that "diseases desperate grown By desperate measures are relieved, or not at all."

Although Race makes some effort to explain my personal motives in taking up these highly charged issues, it may still be said that my views are defeatist, and that we should all continue to proceed on the assumption of racial equality. In the long battle between enlightenment and obscurantism, however, unpleasant truth has always proved more useful than well- intentioned lies.

*Michael Levin is Professor of Philosophy at the City College of New York and the Graduate Center of the City University of New York.

¹ Michael Levin, Why Race Matters: Race Differences and What They Mean (Westport, Conn.: Praeger, 1997). Documentation for all claims unreferenced in the present paper may be found there.

² And Asians, but that is not my topic.

³ C. Peoples, J. Fagan and D. Drotar, "The Influence of Race on 3- year- old Children's Performance on the Stanford- Binet: Fourth Edition," Intelligence 21 (1995); J. Brooks- Gunn, P. Klebanov and G. Duncan, "Ethnic Differences in Children's Intelligence Test Scores: Role of Economic Deprivation, Home Environment and Maternal Characteristics," Child Development 67 (1996).

⁴ By such an argument, one might claim that yardsticks of white manufacture unable to measure the height of blacks, or that blacks should poorly at games invented by whites, like basketball.

⁵ L. Willerman, R. Schultz, and J. Rutledge, "In vivo Brain Size and Intelligence," Intelligence 15 (1991).

⁶ R. Haier, B. Siegel, K. Nuechterlein, K. Hazlett, J. Wu, J. Paek, H. Browning, and M. Buchsbaum, "Cortical Glucose Metabolic Rate Correlates of Abstract Reasoning and Attention Studied with Positron Emission Tomography," Intelligence 12 (1988).

⁷ Chiefly positron emission tomography and magnetic resonance imaging.

⁸ W. Garner and A. Wigdor, Ability Testing (Washington, D. C.: National Academy Press, 1982), 1, p. 72.

⁹ E. g. that the criteria for validating IQ, and the criteria for validating those criteria, are all biased.

¹⁰ Race cites egalitarians who contest the use of some supposedly unclear word ( "race," "intelligence" ), and then contest its abandonment.

¹¹ W. Mischel, "Preference for Delayed Reinforcement: An Experimental Study of a Cultural Observation," Journal of Abnormal and Social Psychology 56 (1958); "Preference for Delayed Reinforcement and Social Responsibility," Journal of Abnormal and Social Psychology 62 (1961); "Father- Absence and Delay of Gratification: Cross- Cultural Comparisons," Journal of Abnormal and Social Psychology 63 (1961).

¹² See W. Dahlstrom, D. Lachar and L. Dahlstrom, MMPI Patterns of American Minorities (Minneapolis: University of Minnesota Press, 1986).

¹³ T. Kochman, Black and White: Styles in Conflict (Chicago: University of Chicago Press, 1983).

¹⁴ A. Tashakkori and V. Thompson, "Race Differences in Self- Perception and Locus of Control during Adolescence and Early Adulthood," Genetic, Social and General Psychological Monographs 117 (1991).

¹⁵ In Race, the following material is divided between an appendix and the chapter on crime.

¹⁶ A standard discounting model might make sense when discussing finance, where it makes little difference whether one is computing the present value of $100 or $1,000,000. That is because the interest rates we see in financial models are typically taken to be market equilibrium rates, not reflections of an individual agent's preferences. When we talk about individual behavior, we might expect that the way one discounts the future depends largely on what one has in the present. People may be more inclined to discount future utility, rather than simply discounting future consumption.

¹⁷ The standard discussion is G. Ainslie, Picoeconomics (Cambridge: Cambridge University Press, 1992).

¹⁸ The most elegant expression for perceived value is G/[ I×t + Z], where Z is a constant to keep perceived value from becoming infinite when t= 0.

¹⁹ The account is compatibilist, i.e. consistent with universal determinism.

²⁰ A. Tashakkori, "Race, Gender and pre- Adolescent Self- Strucure: A Test of Construct- Specificity Hypothesis," Personality and Individual Differences 14 (1993).

²¹ Tashakkori and Thompson, "Race Differences."

²² I. e., separated twins get placed in similar families; twins reunited by psychologists influence each other.

²³ Data is cited in Race; J. Rushton, Race, Evolution and Behavior (New Brunswick, N. J.: Transaction, 1994) contains a thorough review and analysis.

²⁴ Measured by the number of genetic relatives carrying partial copies of its genes which an organism leaves behind, as opposed to individual fitness, which is an organism's own probability of survival.

²⁵ R. Herrnstein and C. Murray, The Bell Curve (New York: The Free Press, 1994).

²⁶ C. Jencks,Rethinking Social Policy (Cambridge, Mass.: Harvard University Press, 1992).

²⁷ Assuming that discrimination is an injury, which I do assume in this chapter.

²⁸ here the work of Herrnstein and Murray, and Linda Gottfredson is called on.

²⁹ The discourses of some prominent soi- disant egalitarians like John Baker and Michael Walzer are deconstructed to this effect.

³⁰ A libertarian can recast the issue as the permissibility of race- conscious action taken by his protective association.

³¹ A description offered by a group of young black males of their social relations bears a striking similarity to Hobbes's account of the state of nature.

³² Another self- evident truth that requires banging away at.

³³ Race reviews a number of caricatures of race consciousness.