The Bell Curve: Intelligence and Class Structure in American Life
Reviewed by Sandra Scarr
Sandra Scarr is Commonwealth Professor of Psychology at the University of Virginia.
"All men are created equal" is a political statement, not a scientific one. Individual variability is the biological norm of populations, and humans are no exception. The Constitution guarantees Americans political and legal equality, not biological identity. Confusion in the media and among scholars about the fundamental difference between equal citizenship rights and unequal human characteristics is behind much of the furor this book has aroused.
The Bell Curve is both an optimistic American story of a rising intellectual meritocracy and a pessimistic tale of a cognitive underclass, doomed by low intelligence not only to limited success at school but also to high rates of poverty, unwed births, and crime. The authors show that for the past 50 years selection into cognitively elite universities and high-status, well-paid occupations has increasingly reflected high IQ rather than privileged family background. Similarly, low socioeconomic status and social problems have increasingly become associated with low IQ rather than low social status of one's parents.
Most Americans are willing to accept the scientific finding that individuals' intellectual abilities play a critical role in determining their socioeconomic success. After all, we like to believe that the country was built on individual talents and initiative. Furthermore, there is broad recognition that people differ in their abilities to cope with educational, social, and occupational demands. The finding that IQ differences are moderately to highly heritable is disturbing to some social scientists, but most Americans accept the idea that individual differences in talents, motivation, and personality are genetically variable to some degree.
In summing up the research on white North American and European populations, Herrnstein and Murray find that IQ differences are 60 to 80 percent due to genetic variability and 15 to 35 percent to differences in rearing environments (with 5 percent attributed to measurement error). What is surprising to those not familiar with the field is that the most important environmental factor is individual or within-family experience, not differences among families. It should be noted that high heritability can be seen as an index of social justice. In a society in which everyone had an equal opportunity to learn, the heritability of IQ would approach 100 percent, because there would be no significant environmental differences among individuals. If there are no environmental differences in opportunities to learn, then all individual differences in IQ must arise from genetic differences.
IQ is obviously not the only factor that influences an individual's socioeconomic prospects, but Herrnstein and Murray point out that other personal characteristics are also highly heritable. For example, motivation and personality are about 50 percent heritable, with die rest of the reliable variation due to within-family environmental differences. Some of the popular discussions of the book have questioned its review of the science. I find the book's discussion of IQ heritability among whites to be a fair summary of the past 40 years of research in behavior and genetics.
Public acceptance of the book's findings breaks down when the authors report significant differences in the distribution of IQ scores among ethnic or racial groups, especially when those group differences in intelligence have important real-life implications. Specifically, the finding that on average African-Americans score significantly lower than whites on IQ tests and the assertion that low IQ can explain the lack of economic success and prevalence of social pathologies among African-Americans enraged many Americans. The long history of racial discrimination undermines U.S. claims of equal opportunity and makes the book's treatment of African-American's lower IQ scores and achievements seem overly simple. Further, the ugly political mood of the country makes it unlikely that any scientific studies of race differences will be calmly received.
We should remember, however, that educational achievements can be improved through better schooling and greater parental investments. Other countries do a better job than we do at assuring nearly universal literacy and numeracy. We can and should improve education, but we should understand that we have not yet found any reliable method for raising IQ.
Race becomes an issue in discussion of IQ because African-Americans score, on average, 15 IQ points lower than whites. Whereas 20 percent of whites score below 85 on IQ tests, about 50 percent of African-Americans do. Considering the unfavorable social outcomes predicted for low-IQ whites, the high percentage of African-Americans in this category is particularly disturbing. For unexplained reasons, the authors did not conduct comparable analyses of the relationships of IQ to meritocratic success or to social pathology for the large number of African-Americans in the NLSY sample. Are African-Americans as efficiently selected into the cognitive elite as are whites? Lower correlations between IQ and social class among blacks suggest that they are not. Because of affirmative action programs in higher education and in employment, the authors show that at every level of college or university and middle to high occupational levels, African-Americans' IQ scores fall below those of comparable whites by at least a standard deviation. But less efficient selection for social mobility by IQ within the African-American population would mean that, unlike whites, some high-IQ African-Americans remain in low status occupations. Does low IQ have the same correlations with social pathology among African-Americans? Herrnstein and Murray provide no specific evidence.
On causes of between-race differences in IQ, the authors favor a mixture of genetic and environmental differences as the most likely explanation. Although they admit that no research on race differences in IQ is conclusive, they interpret the Minnesota study of transracial adoption as support for their position. In this study, Richard Weinberg and I found that socially-identified black children (some had two African-American birth parents, and others had one) who were adopted in infancy and early preschool years by middle class white families scored as well as white adoptees and well above average for African-Americans at age seven. At age 18, their IQ scores were lower than those of white and Asian adoptees in the same families. Adoptees with two African-American birth parents earned IQ scores that were not notably higher than IQ scores of black children reared in black families. It must be emphasized that the study's results do not tell one anything about the cause of the relative decline in IQ. The low scores of the black adoptees at age 18 could be the result of social discrimination or racial genetic differences, or both.
Tellingly, the authors fail to address evidence from our study of IQ and African ancestry within a large sample of black adolescents that contradicts the hypothesis that IQ and race are closely correlated. All of the children in our study were socially classified as black but differed individually in their proportion of African ancestry. We found no relationship between blood group and serum protein markers of African ancestry and cognitive test scores. If more African ancestry is not related to lower scores among socially-classified blacks, then African ancestry can hardly be an explanation for IQ differences between black and white groups. The authors also fail to address evidence for lower heritability of IQ scores within a large sample of black identical and fraternal twins, which suggests that environments are less equally distributed among blacks than among whites.
Latinos and Asians play only minor roles in press accounts but major roles in the book. Both groups score higher on tests than blacks and have better social outcomes, in keeping with the authors' thesis about an IQ-social pathology link. In the NLSY data set, Latinos score lower on ability tests and have higher rates of social pathology than whites, but higher scores and lower rates than blacks. Asians were not sampled in significant numbers in the NLSY, but other research shows them to have an equal or higher average IQs than whites, and very low rates of social pathology.
In my view, the public rejection of the book's presentation of the science of IQ is misguided. There are weaknesses in the discussion of the racial aspects of IQ differences, but the more fundamental case for the heritability of IQ and
The Bell Curve is divided into four parts: the rise of the meritocracy, scientific research on the relationship between IQ and social behavior, IQ differences among ethnic groups and their social consequences, and the authors' libertarian social policy prescription. Everyone who has read newspaper accounts or seen television reports will tell you that The Bell Curve is about race and IQ, but it is not primarily so.
The Bell Curve is mostly about the strengthening relationship between high intelligence and many forms of success in the United States and the hidden role of low IQ in many forms of social pathology. Intelligence matters in earning a place in society, whether intelligence is measured by IQ tests, the Scholastic Aptitude Test (SAT), the Armed Forces Qualification Test, or school achievement tests. The authors focus on the theory and measurement of general intelligence, or g, which has proved exceedingly useful in predicting outcomes for intellectually varied populations.
For example, the military uses g in deciding which enlistees to train for the most demanding occupations. Among more narrowly defined intellectual groups - Yale freshmen or corporate executives - g is less useful as a predictor because virtually all members of the group score high in general intelligence, and other factors determine who will be most successful.
The less well known finding about g is its relationship to social behavior. Using the large data set known as the National Longitudinal Survey of Youth (NLSY), the authors analyze the progress of some 9,000 white youths over 15 years into adulthood. They find that those with IQ scores below 85 (one standard deviation below the white mean of 100) are more likely to drop out of school, be unemployed, commit crimes, be dependent on welfare, bear children out of wedlock, have low-birth-weight babies, and be poor parents. This cognitive underclass, the authors predict, will be loathed and feared by a society sick of crime and social problems. We can all recognize that this is not a pretty picture.
Furthermore, the authors foresee the "dumbing" of America, because the cognitive underclass is outbreeding Americans of average to high IQ and the country is admitting too many immigrants with below-average IQ. Data on birth rates to mothers of lower and higher IQs confirm their premise. This information, combined with research showing that IQ is heritable and that low-IQ mothers provide measurably poorer rearing environments for their children, provides both genetic and environmental reasons for concern about this trend.
Murray and Herrnstein also worry that the country is becoming more stratified along IQ lines. Because selection for the best schools and best jobs is increasingly likely to be made on the basis of intellectual ability (at least for whites) rather than class or family background, Herrnstein and Murray believe that intellectually demanding professions now absorb nearly all high-IQ white Americans, leaving fewer bright workers in manual trades today than in earlier decades.
The cognitive elite are becoming progressively more isolated from the rest of society, whom they govern, hire and fire, but never live among. This isolation begins early with public schools tracking students into classes according to IQ and ends with the cognitive elite sending their own children to private schools.
A critical element of the book is the discussion of efforts to raise IQ by improving an individual's environment. The authors are properly skeptical about the effectiveness of educational interventions to raise IQ. For children in very impoverished or abusive circumstances, educational enrichment can raise cognitive functioning, but there is little evidence that current programs have any long-lasting effects on IQ scores of children in adequate environments. Critics of heritability studies emphasize that some untried interventions could conceivably raise everyone's IQ (malleability is unrelated to heritability, they say), but that theoretical possibility does not change the fact that at this time any intervention to provide a known favorable environment will have small effects on IQ differences. the link between IQ and social outcomes is sound. However, the controversy over the book's policy recommendations is justified. The authors' solution to the social and economic consequences of low intelligence is to cut off welfare and affirmative action and to admonish everyone to find a "valued place" in his or her "clan" or local community. Their eugenic concerns are reflected in the call to eliminate public policies that provide incentives for poor, unwed mothers to reproduce.
Charles Murray did not need the scientific literature on individual and group differences in intelligence and achievement to propose the abolition of welfare support for single mothers or the abolition of affirmative action for African-Americans. He had already proposed those policies to a skeptical Congress many years ago. Likewise, since the 1980s Murray has argued that affirmative action requirements that force companies and universities to accept unqualified minority candidates are economically harmful because they lead to higher training costs and lower productivity. Further, he maintains that selection of less-qualified minorities cheapens the achievements of those African-Americans whose qualifications are equal or superior to those of whites.
Even the authors do not suggest that their public policy proposals arise necessarily from the scientific data. In fact, one could argue that knowing that low IQ is largely heritable calls for more, not less support for those so disadvantaged through no fault of their own. Low IQs reduce individuals' chances to be fully self-sufficient in a technologically advanced society; therefore, compensation for low IQ is a plausible U.S. policy. Compensation for disabilities derives from our concerns with social justice for those whose disadvantages are not their fault. Social welfare democracies such as the European Union and the Nordic Countries have the same IQ distribution as the United States, yet their public policies prevent abject poverty, homelessness, and medical neglect among those with low IQs. Charles Murray's "valued place" is hard to build on hunger, poverty, and despair.
The justification for affirmative action programs that select and promote less qualified minorities over higher-IQ whites depends entirely on the value society attaches to ethnic diversity in educational and employment settings. If providing African-Americans with adequate representation throughout all sectors and levels of American society has a high value, then selection models can be devised to make that happen. Scientific research on race differences in IQ scores can inform us about how to carry out that policy, but science cannot define the values that make ethnic diversity important. Policies proposed in the book are based on the value judgment that economic efficiency is more important than social harmony.
The Bell Curve is a challenging book that compiles a vast research literature on IQ and social outcomes and spells out the troubling facts on ethnic group differences. Unfortunately, the authors' radical proposals to abolish welfare and affirmative action will lead many people to reject everything in the book. This would be a pity because scientists and policymakers will be tempted to dismiss the very data that could inform more benign and effective public programs to address the needs of low-IQ Americans.
|Netscape 5.0||27/03/106 06:25:30||User: 1||Last visit: Mon Mar 27 00:21:02 2006|