Why Race Matters: Race Differences and What They Mean by Michael Levin, Westport, Praeger, 1997.
Edward M. Miller Department of Economics and Finance University of New Orleans 504-286-6913 (work) 504-286-6397 (fax) 504-283-3536 (home) E Mail: emmef@uno.edu
The Journal of Social, Political, and Economic Studies, Vol. 223, (Fall 1998) No. 2, 360 - 366.
Levin's new book, Why Race Matters, is probably one of the most politically incorrect books of recent years. It argues both that there are racial differences in ability and personality, and that these differences matter for public policy. Professor Levin's views have been controversial for years, and he has successfully defended in court his right to express them.1 The book is organized into three major sections. The first documents that differences exist and that they are genetic. The second and third discuss why these differences matter for economic and social policy, hence the title of Why Race Matters.
The empirical evidence for racial differences is strong, although seldom discussed in the academic literature, being often dismissed as due to stereotypes. Levin deals with the stereotype issue, asking why over the centuries certain beliefs about race have prevailed. The simplest answer is observation; the traits have been repeatedly observed. A summary of the data on intelligence describes the differences between races.
As a professional philosopher, Levin frequently focuses on issues of definition and logic. As an example of Levin's style he says (p. 51), "Calling athletic ability. . . 'intelligence' no more changes the intelligence of athletes, . . . than call dogs 'horses' will make them whinny. The world is what it is no matter how it is described. That is why, when dispute focuses on a word, the word is best dropped and the facts restated without it. Should someone insist that whether Albert Einstein or Babe Ruth were equally intelligent is culture-relative, it is best to say that Einstein was better . . . at abstract reasoning, and Ruth better . . . at hitting baseballs, whatever these traits are called."
Levin points out how many who "make a point in argument of not understanding 'intelligence' invariably understand it in all other contexts." One of his examples is how Gould (who pretended not to understand the concept) who in his book The Mismeasure of Man2 asserts that psychometricians had "reified" intelligence. But Levin points out that whether an abstraction like intelligence is useful depends on whether it correlates with other variables of interest, and that intelligence does meet this criterion. Speaking as a philosopher, Levin points out how positivism or operationalism (the idea that concepts have not meaning beyond the operations used to measure them) has become unpopular in philosophy of science circles. As he puts it. "At the present time, an objection like 'IQ tests predict' but so what? would be dismissed as doctrinaire in any context other than psychometrics. Nobody belittles Maxwell's equations by saying 'they predict radio waves, but so what?" Radio waves are taken to show the existence of the fields these equations describe. Likewise, the correlates of IQ are important, not because correlation exhausts scientific knowledge, but for the opposite reason, that they indicate the underlying reality.
Having argued that intelligence is a meaningful concept, Levin then goes on to deal with the evidence for racial differences and related questions. He takes up test bias, and the claim that the racial differences are only due to the tests being biased.
Taking Gould on again, Levin points out that even Gould3 agreed that tests are not biased against blacks in the statistical sense (as Levin points out, this admission is in effect a retraction of Chapter 5 of his earlier 1981 book), but argued that people are really interested in whether "blacks average 85 and whites 100 because society treat blacks unfairly." This is replacing the usual meaning of bias with a new question. As Levin points out repeatedly, an effect can be real but unjustly caused, and whether or not an effect is real, and whether it is justly caused are logically separate questions.
He presents evidence that tests predict academic performance, and job performance, as well for blacks as for whites. Interestingly, black performance is usually over predicted, not under predicted. Given the same test score, the blacks typically perform worse on the job or in employment.
There are a number of reasons for this. As was pointed out in this journal4, Bayes' Theorem implies that the best estimate of the true ability of a person will be a suitably weighted average of their ability as tested, and the average ability of the group. This theorem has the politically unfortunate implication that when seeking the best employees, those from a low scoring group should have points deducted. Levin cites the working paper version of this, but due probably to the time delays between the first drafts of his book and its final publication, failed to update the reference.
As to evidence that the tests are unbiased it is found that blacks lag whites by about two years in their performance on tests and Asians by three years, but that the nature of the errors made (and the relative difficulty of the questions) at the same level of mental ability are similar.5 If differential exposure to "white culture" is the problem, the nature of the errors made and the relative difficulty of questions would vary greatly between the races, but they do not. Also, while it is plausible that blacks are exposed to a somewhat different culture than those (typically middle-class whites) that make up the tests, the Asians are even more culturally different and yet outperform whites.
The racial difference in IQ is well known to specialists (even if not to the media) and well discussed by them. There has been less attention paid to racial differences in personality. Levin describes two of special interest.
One concerns self esteem. Contrary to popular belief, studies show black self esteem to typically be higher than that of whites. The wide spread belief to the contrary among the intelligentsia may be because these people believe they would have low self esteem if they had the status and school abilities of blacks.
Another very important racial difference is in time preference. In a classic experiment children were asked to fill in a very simple questionnaire, and then offered a choice of rewards. They could have a small piece of candy now, or a bigger piece next week. Black children in Trinidad were found to prefer the immediate piece more than Indian (i.e. ancestors from India) children in Trinidad.6 Banfield7 has shown how many life decisions made differently by inhabitants of inner cities can be traced to differences in time preference. For instance, deciding not to study, or to steal a purse now, involves trading off immediate gratification for future gratification. Levin states that blacks watch 73 hours of TV per week versus whites' 50 hours, and spend twice as much per capita on movies in spite of lower incomes, which he implies is related to this trait. The fact that even at the same income level, blacks have typically accumulated less wealth than whites or Asians appears to be another reflection of this trait.
After an introduction to basic genetics, Levin presents the evidence (from twin and adoption studies typically) that within races, intelligence and personality traits show considerable variability. He starts out by quoting the Snyderman and Rothman8 study which showed three times as many experts thought the difference between blacks and whites in intelligence was both genetic and environmental as thought it was only environmental. Even among the editors and journalists surveyed, 27% were in the both genetic and environmental camp.
Levin then goes on to present some of the evidence that has persuaded so many experts that at least part of the differences between the race are genetic.9 He starts out by pointing out that racial differences in intelligence appear from about age three. Many of the cultural explanations (schooling for instance) have not had an effect before then. He recognizes the theoretical possibility that genetics are very important for within race differences (as is well documented), but that the causes for racial differences are wholly environmental.
With a wholly environmental theory it is implausible that the heritability of a specific tests would be correlated with the racial differences in that test. Yet it has been found that the mental tests that are most heritable are those that show the greatest black white differences. Environmental factors would normally be expected to be greater for vocabulary tests than for picture arranging, but the racial differences on vocabulary tests are less, even though one would expect vocabulary to be among the most culture sensitive of tests.
Professor Levin is a professional philosopher (City University of New York). His views on philosophy are interesting, and to me at least, original. His approach is sociobiological and assumes that human nature evolved. Recognizing that moral feeling have probably evolved, he also recognizes the possibility that they may be different in different races. He feels that the white race has evolved to have a lower rate of time preference (i.e. gives higher weights to future events), and a greater degree of altruism and willingness to obey rules. He discusses the possibility that the hunting required for survival in prehistoric Europe required group cooperation, and that this led to greater altruism among its inhabitants. Gathering required less cooperation. At one point (p.176) he argues that "Since it is easier for female gatherers in a warm climate to support their offspring, there will be less intense selection for females who prefer fidelity in their mates, hence, by what Darwin called sexual selection, less intense selection for males disposed to conform to this female demand." This is part of Miller's (1994) differential paternal invesment theory, which Levin had seen and discussed with the author, although it is not mentioned. Since Levin believes there is no logical basis for saying some beliefs or traits are better than others, he notes "Hunters may regard gatherer sexual morality as loose, while gatherers regard hunter sexual morality as inhibited."
One of the interesting ideas in the book is that the degree of altruism and the rates of time preference in a population should vary together. He goes through the evolutionary logic of cooperation, discussing the famous prisioner's dilemma game. He recognizes that in such situations non-coperating (cheating) is optimal, if one will never see a person again. However, the environments humans evolved in were ones where people lived in small bands. They regularly saw the same people repeatedly. Cheating someone hurts in the long run, because that person will not cooperate with you next time you needed his help.
The last part of the book deals with questions of why race matters. Here again, Levin's role as philosopher comes in. He argues that whether the causes of black poverty and suffering is genetic is indeed important. If it is due to something that white people did, possibly the blacks have some claim for compensation (although the question would still have to be addressed of whether the whites now living were the ones who owed the compensation). However, Professor Levin argues that the black problems are due to genetic causes. Since no one is responsible for their own genes or for the genes of other people, this absolves the white community (and white individuals) of responsibility for black problems.
He uses the example of a man with an inherited limp in one leg. Out of politeness we may pretend not to notice his handicap. However, if he sues us claiming we caused his limp, it is a perfectly proper defense to present evidence that he was born with the problemLevin ends his book with a discussion of crime, which deals both with the causes of racial differences in crime (after documenting that they do indeed exist), and with the provocative idea that perhaps it would be justified for citizens and for the police to take the known racial differences into account. Why Race Matters is the most politically incorrect academic book of recent years. However, in spite of this the reader will find many provactive facts and ideas in it.
Footnotes
1. See Pearson, R. (1997) Race, Intelligence, and Bias in Academe, Washington: Scott, Townsend, for an account of the pressures that Levin had to face.
2. Gould, S. (1981) The Mismeasure of Man, New York: Norton.
3. Gould, S. "Curveball". The New Yorker, Nov. 28, 139-149.
4. See Miller, E. (1994) The Relevance of Group Membership for Personnel Selection: A Demonstration Using Bayes Theorem, Journal of Social, Political, and Economic Studies, Vol. 19 (Fall 1994) No. 3, 323-359.
5. Jensen, A. R. (1980) Bias in Mental Testing, London: Methuen.
6. Mischael, W. (1961) "Father-absence and delay of gratification: cross-cultural comparisons." Journal of Abnormal and Social Psychology 63, 1, 116-124.
7. Banfield, E. (1974) The Unheavenly City Revisted. Boston: Little Brown.
8. Snyderman, M. and Rothman, S. (1988) The IQ Controversy, the Media and Public Policy. New Brunswick, Transaction Books.
9.For the Best and most up-to-date work on this, see Jensen, A. R. (1998) The g Factor. Westport: Praeger.
