The book draws heavily from mainstream science research. However, their analysis is their own. Though Richard Herrnstein had a strong background in psychology, co-author Charles Murray was not an expert in intelligence testing.
The book's title comes from the bell-shaped normal distribution of IQ scores. The normal distribution is the limiting distribution of a random quantity which is the sum of smaller, independent random phenomena. The message in the title is that IQ scores are normally distributed because a person's intelligence is the sum of many small random variations in genetic and environmental factors.
Shortly after publication, many people rallied both in criticism and defense of the book. Some critics denounced the book and its authors as supporting scientific racism. A number of critical texts, including The Bell Curve Debate, were written in response to the book.
It is the only New York Times Bestselling book ever to remain permanently out of print.
The Bell Curve contains 941 pages in the first printing and 879 in the revised paperback. Much of its material is technical and academic. The book's statistical explanations are styled to appeal to a general audience. There are extensive notes, graphs, and tables. The Bell Curve is divided into four sections.
|Category||Parents' socioeconomic class||Cognitive class|
|Very high/Very bright||3||2|
|Very low/Very dull||24||30|
Herrnstein and Murray in many ways follow in the footsteps of UC Berkeley researcher Arthur Jensen, whose controversial article on the subject appeared in 1969 in the Harvard Educational Review. The Bell Curve argues that:
Their evidence comes from an analysis of data compiled in the National Longitudinal Study of Youth (NLSY), a study conducted by the U.S. Department of Labor's Bureau of Labor Statistics tracking thousands of Americans starting in the 1980s. All participants in the NLSY took the Armed Services Vocational Aptitude Battery (ASVAB), a battery of ten tests taken by all who apply for entry into the armed services. Four of those tests comprise the Armed Forces Qualification Test, a measure of cognitive ability comparable to an IQ test. (Some had taken an IQ test in high school, and the median correlation of the AFQT and those tests was .81, which is high.) Participants were later evaluated for social and economic outcomes. In general, IQ/AFQT scores were a better predictor of life outcomes than social class background. Similarly, after statistically controlling for differences in IQ, many outcome differences between racial-ethnic groups disappeared. (See also Significance of group IQ differences.)
|US population distribution||5||20||50||20||5|
|Married by age 30||72||81||81||72||67|
|Out of labor force more than 1 month out of year (men)||22||19||15||14||10|
|Unemployed more than 1 month out of year (men)||12||10||7||7||2|
|Divorced in 5 years||21||22||23||15||9|
|% of children w/ IQ in bottom decile (mothers)||39||17||6||7||-|
|Had an illegitimate baby (mothers)||32||17||8||4||2|
|Lives in poverty||30||16||6||3||2|
|Ever incarcerated (men)||7||7||3||1||0|
|Chronic welfare recipient (mothers)||31||17||8||2||0|
|High school dropout||55||35||6||0.4||0|
|Values are the percentage of each IQ sub-population, among non-Hispanic whites only, fitting each descriptor. Herrnstein & Murray (1994) pp. 171, 158, 163, 174, 230, 180, 132, 194, 247-248, 194, 146 respectively.|
Herrnstein and Murray recommended the elimination of welfare policies that encourage poor women to have babies:
This claim spurred later research in economics and sexology, which considered that welfare programs for females had a doubly negative effect on aggregate IQ within the transfer group, by allowing the female partner to forgo a full consideration of the male's ability to serve as a provider of familial resources, instead placing greater emphasis on desirable physical or social characteristics (presumed to be not as positively correlated with IQ). Neither of these claims, as originally embodied in text and the follow-on research, dealt with race as such, but rather demonstrated concern that large numbers of minorities were positioned as recipients, leading to a continual worsening of the measured divergence in intelligence. However, two years later, the 1996 U.S. welfare reform substantially cut these programs.
In a discussion of the future political outcomes of an intellectually stratified society, they stated that they "fear that a new kind of conservatism is becoming the dominant ideology of the affluent - not in the social tradition of an Edmund Burke or in the economic tradition of an Adam Smith but 'conservatism' along Latin American lines, where to be conservative has often meant doing whatever is necessary to preserve the mansions on the hills from the menace of the slums below" (p. 518). Moreover, they fear that increasing welfare will create a "custodial state" in "a high-tech and more lavish version of the Indian reservation for some substantial minority of the nation's population." They also predict increasing totalitarianism: "It is difficult to imagine the United States preserving its heritage of individualism, equal rights before the law, free people running their own lives, once it is accepted that a significant part of the population must be made permanent wards of the states" (p. 526).
Knowledge of the general intelligence factor (g) is important to evaluate the debates on testing. The g factor is an almost universally supported, and very important, construct in psychometry. In a battery of mental ability tests given to a group of people, all the tests are positively correlated with each other; those who are above average in one will, on average, be above average on the others. Factor analysis can extract a smaller number of factors to account for the variation in the scores; this is possible because the more two tests measure the same thing, the greater their correlations will be. One factor, g can then be extracted (sometimes after another layer of specific factors are removed). The correlation of the test scores with g is its g-loading; a high one is desirable in a test. g is correlated with a wide range of social outcomes; some are such as income, academic achievement, job performance, and career prestige, poverty, dropping out, and out-of-marriage childbirth. g correlates with both speed and consistency of performance on elementary cognitive tasks (simple ones that can be done by everybody without failure). All of this was mentioned in The Bell Curve, and many biological and neurological correlates have been discovered since, in addition to the long known ones such as brain size. These include the frequency of alpha brain waves, latency and amplitude of evoked brain potentials, rate of brain glucose metabolism, and general health as some of the best established ones. Almost all of a test's predictive validity lies in g, as opposed to the more specific factors. The AFQT and IQ tests are very highly g-loaded.
Initially, The Bell Curve received a great deal of positive publicity, including cover stories in Newsweek ("the science behind [it] is overwhelmingly mainstream"), early publication (under protest by other writers and editors) in The New Republic by its editor-in-chief at the time Andrew Sullivan, and The New York Times Book Review (which suggested critics disliked its "appeal to sweet reason" and are "inclined to hang the defendants without a trial"). Early articles and editorials appeared in Time, The New York Times ("makes a strong case"), The New York Times Magazine, Forbes, the Wall Street Journal, and National Review. It received a respectful airing on such shows as Nightline, the MacNeil/Lehrer NewsHour, the McLaughlin Group, Think Tank, PrimeTime Live, and All Things Considered. The book sold over 500,000 copies in hardcover.
While the book's popularity was mostly propelled by its controversial claims regarding race and intelligence, both the accuracy of those claims and the qualifications of the authors soon came under attack in the media. Herrnstein died before the book was released, leaving Charles Murray to the public defense. Although Herrnstein was a psychologist, Murray is a conservative think tank analyst with a Ph.D. in political science and no credentials in psychometrics.
Some scholars have condemned the book. University of Oklahoma Assistant Professor of Anthropology Michael Nunley wrote:
I believe this book is a fraud, that its authors must have known it was a fraud when they were writing it, and that Charles Murray must still know it's a fraud as he goes around defending it. [...] After careful reading, I cannot believe its authors were not acutely aware of [...] how they were distorting the material they did include.
Professor Leon Kamin, a longtime critic of cognitive ability tests, said the book was "a disservice to and abuse of science." Harvard University psychologist Howard Gardner, who teaches in the department of education, called the style of thought "scholarly brinkmanship":
The authors seem to show the evidence and leave the implications for the reader to figure out; discussing scientific work on intelligence, they never quite say that intelligence is all important and tied to one's genes, yet they signal that this is their belief and that readers ought to embrace the same conclusions.
Economist and conservative writer Thomas Sowell criticized the book's conclusions about race and the malleability of IQ, writing:
When European immigrant groups in the United States scored below the national average on mental tests, they scored lowest on the abstract parts of those tests. So did white mountaineer children in the United States tested back in the early 1930s... Strangely, Herrnstein and Murray refer to "folklore" that "Jews and other immigrant groups were thought to be below average in intelligence." It was neither folklore nor anything as subjective as thoughts. It was based on hard data, as hard as any data in The Bell Curve. These groups repeatedly tested below average on the mental tests of the World War I era, both in the army and in civilian life. For Jews, it is clear that later tests showed radically different results—during an era when there was very little intermarriage to change the genetic makeup of American Jews.
The source for the claim that Jews had below average IQ stems from the work of Henry H. Goddard and Carl C. Brigham as these are the only researchers who published data that seemed to indicate this prior to 1930. Goddard wanted to find out if the Stanford-Binet test was as effective at identifying 'high-grade defectives' (the term then used for those with mental ages between eight and twelve) among immigrants as it was among native-born Americans. By 1913, Goddard had translated the Binet test into English and arranged, over a two-and-a-half-month period, for it to be given to a subset of Jewish, Hungarian, Italian, and Russian immigrants "preselected as being neither 'obviously feeble-minded' nor 'obviously normal'" (Goddard, 1917, p. 244,). Among this "unrepresentative" group (178 subjects in all), the tests successfully categorized 83% of the Jews, 80% of the Hungarians, 79% of the Italians, and 87% of the Russians. Goddard (1917) explicitly did not assert that 80% of Russians, Jews, or any immigrant group in general were feeble minded nor that the figures were representative of all immigrants from those nations. Nor did he claim that the feeblemindedness he was measuring was due to heredity. The vast majority of the many immigrants going through Ellis Island were never given mental tests. Nor was a random sample of any national group of immigrants ever tested. The only study by Goddard involving the testing of immigrants begins with the following sentence: 'This is not a study of immigrants in general but of six small highly selected groups...'(1917, p. 243)
In its defense, fifty-two professors, most of them psychologists including researchers in the study of intelligence and related fields, signed an opinion statement titled "Mainstream Science on Intelligence" endorsing the views presented in The Bell Curve. The statement was written by psychologist Linda Gottfredson and published in The Wall Street Journal in 1994 and reprinted in the Intelligence Only seven of the 100 invitees contacted said the statement did not represent the mainstream view of intelligence. Some of the signers were cited as sources for Murray and Herrnstein's book.
Some of the task force's findings supported or were consistent with statements from The Bell Curve. They agreed that:
Regarding Murray and Herrnstein's claims about racial differences and genetics, the APA task force stated:
There is certainly no such support for a genetic interpretation... . It is sometimes suggested that the Black/ White differential in psychometric intelligence is partly due to genetic differences (Jensen, 1972). There is not much direct evidence on this point, but what little there is fails to support the genetic hypothesis.
Regarding statements about other explanations for racial differences, the APA task force stated:
The differential between the mean intelligence test scores of Blacks and Whites (about one standard deviation, although it may be diminishing) does not result from any obvious biases in test construction and administration, nor does it simply reflect differences in socio-economic status. Explanations based on factors of caste and culture may be appropriate, but so far have little direct empirical support.
Regarding statements about any explanations for racial differences, the APA task force stated:
At present, no one knows what causes this differential.
The APA journal that published the statement, American Psychologist, subsequently published eleven critical responses in January 1997.
See also: the discussion of intelligence testing
The initial positive reception of The Bell Curve in media such as newspapers and television talk shows was troubling to critics such as economist Edward S. Herman and evolutionary biologist Joseph L. Graves who felt that it indicated an acceptance of what Herman calls "deterministic racist doctrines.
The second wave of reviews, which did not arrive until much later, was composed of expert opinion in the relevant fields. It provided a belated substitute for the peer-review process to which Murray and Herrnstein did not originally submit their work.
Melvin Konner, professor of anthropology and associate professor of psychiatry and neurology at Emory University, called Bell Curve a "deliberate assault on efforts to improve the school performance of African-Americans":
This book presented strong evidence that genes play a role in intelligence but linked it to the unsupported claim that genes explain the small but consistent black-white difference in IQ. The juxtaposition of good argument with a bad one seemed politically motivated, and persuasive refutations soon appeared. Actually, African-Americans have excelled in virtually every enriched environment they have been placed in, most of which they were previously barred from, and this in only the first decade or two of improved but still not equal opportunity. It is likely that the real curves for the two races will one day be superimposable on each other, but this may require decades of change and different environments for different people. Claims about genetic potential are meaningless except in light of this requirement.
According to Gould, if any of these premises are false, then their entire argument disintegrates (Gould, 1994). Similarly, in "Science" in the service of Racism C. Loring Brace writes that The Bell Curve makes six basic assumptions at the beginning of the book:
Brace proceeds to argue that there are faults in every one of these assumptions. The Nobel Prize winning economist James Heckman writes that two assumptions made in the book are questionable:
Heckman writes that a reanalysis of the evidence used in The Bell Curve contradicts this story. The factors that explain wages receive different weights than the factors that explain test scores. More than "g" is required to explain either. Other factors besides "g" contribute to social performance, and they can be manipulated. Murray responded to a shorter version of Heckman's critique in an August 1995 letter exchange in Commentary magazine.
In the book Intelligence, Genes, and Success: Scientists Respond to The Bell Curve a group of social scientists and statisticians analyzes the genetics-intelligence link, the concept of intelligence, the malleability of intelligence and the effects of education, the relationship between cognitive ability, wages and meritocracy, pathways to racial and ethnic inequalities in health, and the question of public policy. This work argues that much of the public response was polemic and failed to analyze the details of the science and validity of the statistical arguments underlying the book's conclusions.
William J. Matthews writes that part of The Bell Curve's analysis is based on the AFQT "which is not an IQ test but designed to predict performance of certain criterion variables". Nobel Prize in Economics winner James J. Heckman observed that the AFQT was designed only to predict success in military training schools and that most of these tests appear to be achievement tests rather than ability tests, measuring factual knowledge and not pure ability. He continues:
A recent paper in the Psychological Review, " Heritability Estimates Versus Large Environmental Effects: The IQ Paradox Resolved" presents a mechanism by which environmental effects on IQ may be magnified by feedback effects. This approach may provide a resolution of the contradiction between the viewpoint of The Bell Curve and its supporters, and the 'nurture' factors of IQ believed to exist by its critics. Janet Currie and Duncan Thomas presented evidence suggesting AFQT scores are likely better markers for family background than "intelligence" in a 1999 Study.
Herrnstein and Murray report that conditional on maternal "intelligence" (AFQT scores), child test scores are little affected by variations in socio-economic status. Using the same data, we demonstrate their finding is very fragile.Charles R. Tittle, Thomas Rotolo found that the more that written, IQ-like examinations are used as screening devices for occupational access, the stronger the relationship between IQ and income. Thus, rather than higher IQ leading to status attainment because it indicates skills needed in a modern society, IQ may reflect the same test-taking abilities used in artificial screening devices by which status groups protect their domains. Min-Hsiung Huang and Robert M. Hauser write that Herrnstein and Murray provide scant evidence of growth in cognitive sorting. Using data from the General Social Survey, they tested each of these hypotheses using a short verbal ability test which was administered to about 12,500 American adults between 1974 and 1994; the results provided no support for any of the trend hypotheses advanced by Herrnstein and Murray. One chart in The Bell Curve purports to show that people with IQs above 120 have become "rapidly more concentrated" in high-IQ occupations since 1940. But Robert Hauser and his colleague Min-Hsiung Huang retested the data and came up with estimates that fell “well below those of Herrnstein and Murray." They add that the data, properly used, "do not tell us anything except that selected, highly educated occupation groups have grown rapidly since 1940.
According to Christopher Chabris, the most common responses to The Bell Curve involve "phony controversies", adding that "the vast majority of those commenting on The Bell Curve" in books such as The Bell Curve Wars "have little or no scientific authority".
|Values are the average earnings (1993 US Dollars) of each IQ sub-population.|
Murray responded to specific criticisms of the analysis of the practical importance of IQ compared to socio-economic status (Part II of The Bell Curve) in a 1998 book Income Inequality and IQ To circumvent criticisms surrounding their use of a statistical control for socioeconomic status (SES), Murray adopted a sibling design. Rather than statistically controlling for parental SES, Murray compared life outcome differences among full sibling pairs who met a number of criteria in which one member of the pair has an IQ in the "normal" range and the other siblings has an IQ in a higher or lower IQ category. According to Murray, this design controls for all aspects of family background (full siblings share the same family background, growing up together in the same home and the same community).
|Indicator||Bell Curve control for parental SES||Sibling fixed-effect model|
|Annual earnings, year-round workers||5548||5317|
|Years of schooling||0.59||0.45|
|Attainment of BA||1.76||1.87|
|Out of labor force 1+ month||-0.34||-0.3|
|Unemployed 1+ month||-0.52||-0.47|
|Mean years of education||11.4 (10.9)||12.3 (11.9)||13.4 (13.2)||15.2 (15.0)||16.5 (16.5)|
|Percentage obtaining B.A.||1 (1)||4 (3)||19 (16)||57 (50)||80 (77)|
|Mean weeks worked||35.8 (30.7)||39.0 (36.5)||43.0 (41.8)||45.1 (45.2)||45.6 (45.4)|
|Mean earned income||11,000 (7,500)||16,000 (13,000)||23,000 (21,000)||27,000 (27,000)||38,000 (36,000)|
|Percentage with a spouse who has earned income||30 (27)||38 (39)||53 (54)||61 (59)||58 (58)|
|Mean earned family income||17,000 (12,000)||25,000 (23,400)||37,750 (37,000)||47,200 (45,000)||53,700 (53,000)|
|Percentage children born out of wedlock||49 (50)||33 (32)||14 (14)||6 (6)||3 (5)|
|Fertility to date||2.1 (2.3)||1.7 (1.9)||1.4 (1.6)||1.3 (1.4)||1.0 (1.0)|
|Mother's mean age at birth||24.4 (22.8)||24.5 (23.7)||26.0 (25.2)||27.4 (27.1)||29.0 (28.5)|
|Values are "Utopian sample" ("Full sample"). Earning values are the 1993 US Dollars.|