Racist. This book is racist. You don’t need to contemplate or evaluate its content because it is racist. Avoid exposure to the ideas because they are racist. Let’s not disturb our safe slumber with racist ruminations. Carry on.
What’s that? The term “racist” has been so overused and misused that it is now meaningless to you? And what’s that!? You prefer painful truths to comforting misapprehensions? Oh dear. Then proceed with caution, troublemaker, if you must. But you’ve been warned.
Given the inflammatory content of the book, I’ve done my best to objectively summarize what it actually says, chapter by chapter, without any commentary from me.
Introduction - Intelligence is as universal and ancient as any understanding about the human condition. Intelligence varies from person to person and impacts everyone's life in myriad ways. Intelligence tests have been refined and analyzed since the mid-19th century by a long list of thinkers: Charles Darwin, Sir Francis Galton, Alfred Binet, Charles Spearman, Karl Pearson, Cyril Burt, Lewis Terman, Jean Piaget, David Wechsler, B.F. Skinner, Raymond Cattel, Arthur Jensen, William Shockley, Richard Herrnstein, Charles Murray, and many others.
IQ tests were widely used during World Wars I and II for soldier assignments. The U.S. Supreme Court permitted eugenic sterilization based on behavior and IQ ("Three generations of imbeciles are enough." Buck v. Bell, 274 U.S. 200 (1927), J. Holmes.). United States immigration policy at the beginning of the 20th century favored Nordic stock over eastern and southern European immigrants in an attempt to avoid lowering the national average IQ.
As egalitarian ideology ascended in the 1960's and 1970's, an assault on the validity of IQ and IQ testing began. IQ's academic credibility held fast in many corners, but others challenged it on moral grounds. Public perception faltered. Arthur Jensen was excoriated for a 1969 article in the Harvard Educational Review in which he concluded that remedial education programs failed because they targeted individuals with low IQ—namely blacks who tended to have lower IQ's than whites—and educational success was highly dependent on IQ. The Supreme Court prohibited the use of standardized ability tests in employment for all practical purposes in 1971 (Griggs v. Duke Power Co., 401 U.S. 424 (1971)). The National Education Association called upon schools to abolish IQ testing, and a number of colleges abolished use of the Scholastic Aptitude Test. Stephen Jay Gould's 1981 book, "The Mismeasure of Man," was the capstone of the assault on the integrity of IQ testing as a discipline.
Out of the public eye, research continued. Three schools of thought regarding IQ emerged. First, the classicists conceptualize intelligence as a structure best represented by g, a statistical aggregate of scores on various mental tasks such as hearing then reciting a string of numbers backwards and mentally rotating images. High g indicates high cognitive ability. Second, the revisionists view intelligence as a process rather than a structure. They focus on how the brain processes sensory inputs, allocates resources to analyze those inputs during decision-making, and routinizes performance of novel tasks into familiar ones. They also look at how intelligence pertains to real world functioning. Revisionists believe they are mining richer veins of inquiry than classicists. Third, the radicals have theories of multiple intelligences. They compare intelligence to beauty or justice rather than height or weight. Radicals argue that without a clear and firm definition, the claims of IQ testers are naive at best and vicious at worst. The most prominent proponent of this school of thought, Howard Gardner, hypothesized seven different intelligences: linguistic, musical, logical-mathematical, spatial, bodily-kinesthetic, interpersonal intelligence, and intrapersonal intelligence.
This book draws most heavily from the classical tradition. The authors view IQ as a useful but limited tool. They reject the radicals' multiple intelligence approach as more akin to talents than intelligence. They accept but are less familiar with revisionist theories.
The authors find six conclusions regarding tests of cognitive ability beyond significant technical dispute:
1. There is such a thing as a general factor of cognitive ability on which human beings differ.
2. All standardized tests of academic aptitude or achievement measure this general factor to some degree, but IQ tests expressly designed for that purpose measure it most accurately.
3. IQ scores match, to a first degree, whatever it is that people mean when they use the word “intelligent” or “smart” in ordinary language.
4. IQ scores are stable, although not perfectly so, over much of a person's life.
5. Properly administered IQ tests are not demonstrably biased against social, economic, ethnic, or racial groups.
6. Cognitive ability is substantially heritable, apparently no less than 40% and no more than 80%.
Part I - The Emergence of a Cognitive Elite
Ch. 1 - Democratization of college admissions in the 20th century resulted in ever more efficient funneling of the cognitive elite into elite universities and ever more stratification of American society according to cognitive ability.
Ch. 2 - A handful of occupations (lawyers, doctors, accountants, engineers, scientists, mathematicians, business executives) require high IQ for success, and people with high IQs have been increasingly concentrated in those positions since World War II.
Ch. 3 - Spearman's general intelligence factor, g, predicts to a high degree an employee's job performance regardless of occupation. A smarter employee is, on average, a more proficient employee. Employers that pick applicants with the highest IQs can realize large economic gains.
Ch. 4 - Cognitive partitioning, driven by an economy that values intellect, will continue through education, occupation, physical segregation (offices vs. construction sites, rich vs. poor neighborhoods), and assortative mating (smart people marry and breed with other smart people). As environmental and political inequalities decrease, the role of inherited genes increases.
Part II - Cognitive Classes and Social Behavior (all analysis in Part II focuses only on white people)
Ch. 5 - Low IQ is a stronger precursor of poverty than low socioeconomic background. If you had to choose between being born smart or rich, you are less likely to end up poor if you choose to be smart.
Ch. 6 - Low IQ is related to lower educational achievement. Almost no one in the top quartile of IQ fails to graduate high school, no matter how poor, and the dropout rate for those in the top half of IQ is very low. The advantage of a high IQ outweighs the advantage of a high socioeconomic status in getting a college degree. Similarly, the disadvantage of a low IQ outweighs the disadvantage of low socioeconomic status in getting a college degree.
Ch. 7 - Low IQ is related to joblessness and physical disability. Among men in their late 20's and early 30's, both unemployment and being out of the labor force are more strongly predicted by low cognitive ability than by socioeconomic status or education. Cognitive ability is a strong predictor of disability—stronger even than the type of job held. Nine out of ten men who described themselves as too disabled to work were in the bottom quarter of the IQ distribution.
Ch. 8 - Low IQ individuals get married at lower rates, get divorced at higher rates, and have far more illegitimate children. Intelligence is a better predictor of all three than socioeconomic status or education.
Ch. 9 - Low IQ strongly predicts welfare dependency. Of those women on welfare within one year of birthing their first child, 75% were in the bottom quartile of IQ. White women of above average IQ or socioeconomic status do not become chronic welfare recipients. But low socioeconomic status is a better predictor than low IQ of chronic welfare dependency.
Ch. 10 - Low IQ relates to low birth weight, worse home environments, problematic temperament in the baby, low scores on a friendliness index, and poorer motor and social development of toddlers. Socioeconomic status also played a role, but the independent variable of IQ persisted. The worst environments for raising children are concentrated in homes in which the mothers are at the low end of the IQ spectrum.
Ch. 11 - Low IQ relates to criminality. The average white criminal IQ was 92. More serious or chronic offenders have lower IQs than more casual offenders. Offenders who have been caught do not score much lower than those who have not. Socioeconomic status is a negligible risk factor when IQ is taken into account.
Ch. 12 - The data linking IQ to civility and citizenship is harder to acquire and assess than other links. In general, low IQ children are less likely to read about, discuss, or participate in the political process. For adults, political engagement is more closely connected to education than socioeconomic status; the authors theorize that is because education has been used as a proxy for IQ along this dimension. Low IQ young people are less likely to stick with school, be employed, and stay loyal to their spouses.
Part III - The National Context
Ch. 13 - Large human populations differ culturally and biologically. They differ in skin color, hair textures, bone density, and muscle mass. IQ is substantially heritable, somewhere between 40% and 80%, meaning that much of the observed variation in IQ is genetic. Ethnic differences in cognitive ability are neither surprising nor in doubt. Those differences in average IQ are consistently confirmed by data from around the world and can be roughly summarized as follows:
Ashkenazi Jews - 115
East Asians (Chinese, Japanese, and probably Koreans) - 110
Whites - 100
Latinos - 90
African Americans - 85
African Blacks - 75
The average African American and white differ in IQ at every level of socioeconomic status, but they differ more at higher levels than lower ones. The results are not due to test bias; in fact, the more highly g loaded the test, the larger the gap. East Asians tend to perform relatively better on tests of nonverbal skills (aka visuospatial) than verbal skills vis-a-vis whites. Men and women have the same mean IQs, but men's are more widely distributed; so, there will be more men at both the higher and lower ends of the spectrum. One standard deviation in intelligence is about 15 IQ points.
Ch. 14 - Latinos and whites with similar IQs have similar social behavior and economic outcomes. For blacks and whites with similar IQs, blacks have higher educational attainments and higher entry into prestigious professions but lower wages by a few hundred dollars per year. For blacks and whites with similar IQs below the poverty line, blacks have lower family income, higher unemployment, lower labor force participation rate, lower marriage, and higher illegitimacy. More research needs to be done in these areas, and the authors call for focus on cognitive ability as it relates to these outcomes.
Ch. 15 - Demographic trends are exerting downward pressure on the distribution of IQs in the U.S. Higher IQ women receive more education and delay marriage and motherhood; lower IQ women have more babies and sooner. These dysgenic pressures apply to all races and ethnic groups, but they are stronger for blacks and Latinos than whites, which could lead to further divergence in the IQ distributions between races. Moreover, immigration policy and practice no longer selects self-starters with high IQ as efficiently as in the past, meaning the average immigrant IQ of about 95 is lower than the national average. Any IQ gains from positive environmental changes—such as the Flynn effect (the observation that standardized IQ test scores tend to drift up slowly over time for unknown reasons), better nutrition, better health, better education, etc.—are sailing against a countervailing demographic head wind.
Ch. 16 - A cardinal rule in statistics is that correlation does not mean causation. Causal relationships are complex and hard to establish. The issue addressed is not whether low IQ causes social problems; but those people experiencing the worst social problems are heavily concentrated in the lower portion of the IQ distribution. Solutions to social problems must take this fact into account to be effective. The social problems in this chapter are poverty, high school dropouts, unemployment, nonworkers, crime, welfare dependency, illegitimacy, low birthweight, neglect and deprivation at home, and developmental problems.
Part IV - Living Together
Ch. 17 - Raising intelligence significantly, consistently, and affordably might circumvent many social problems. But the story of attempts to raise intelligence is one of high hopes, flamboyant claims, and disappointing results. Improving nutrition, better schools, and stimulating environments intuitively seem like they would improve IQ, but the data are mixed and inconclusive. For example, Head Start, an early education program, has not been shown to be effective in improving cognitive functioning, nor has the link between nutrition and cognitive functioning been proved. The only environmental intervention that works consistently to improve IQ is adoption at birth from a bad family environment to a good one, which yields a gain of about 6 IQ points.
Ch. 18 - Students of average intelligence are probably better prepared academically than ever before because, since the 1960's, the focus of public education has been on the average and below average students rather than the gifted students. American educators dumbed down the curricula: Textbooks were made easier, and requirements for courses, homework, and graduation were relaxed. Only .1% of all federal funds spent on elementary and secondary education go to programs for the gifted. Gifted students do not have to develop their full potential to succeed in school. As a policy matter, the federal government should support parental choice of schools through vouchers, tax credits, or choice within public schools. Policy should also reallocate funds away from disadvantaged students and towards the gifted students because the fate of society is so dependent on them.
Ch. 19 - Affirmative action as practiced in college admissions cannot survive public scrutiny. It is not a minor nod to blacks and Latinos in close calls; it is an advantage so large it puts them in a separate admissions competition. In undergraduate and graduate programs across the country, the average black student is at the 10th or 15th percentile in cognitive ability among the white students on the same campus. Blacks from affluent socioeconomic backgrounds are given a substantial edge over disadvantaged whites. Asians suffer admissions penalties at elite universities. The average Asian student at elite universities is at the 60th percentile in cognitive ability among the white students at the same campus. Blacks and Latinos represent a small portion of all students on campus but a large portion of those doing poorly. In society at large, college degrees do not have the same meaning for white and black students. Affirmative action, if practiced at all, should be based on socioeconomic status rather than skin color.
Ch. 20 – Tests of cognitive ability are one of the best and cheapest selection tools for employers to hire the best workers. Job hiring and promotion procedures that were truly fair and unbiased would produce racial disparities. Public policy since the 1960’s has tried to prevent such disparities through affirmative action in the workplace. Since the 1960’s, blacks have been overrepresented in white collar and professional occupations relative to their numbers within the IQ range from which these jobs are usually filled. Aggressive affirmative action produces large racial disparities in job performance in a given workplace. The benefits to productivity and fairness from eliminating antidiscrimination laws and affirmative action are substantial.
Ch. 21 – The authors predict increasing cognitive stratification in American life and government with the following consequences: increasingly isolated cognitive elite, merging of the cognitive elite with the affluent, and a deteriorating quality of life for those with low IQ. Underclass communities arise from a spatial concentration of single mothers with low IQ. With a growing underclass, a custodial state will emerge, accompanied by a new and more virulent form of racism.
Ch. 22 – Instead of government addressing any fallout from an emerging cognitive elite and underclass, a return to individual freedom and responsibility will better deal with it. We should jettison equality as an ideal because people are only equal in terms of legal rights; they are not equal in terms of cognitive ability or any other innate characteristics. People can and should find their appropriate occupation within society. But occupation is not the sole determinant of a virtuous life. A shift in culture to uphold simple, eternal truths and to celebrate individual dignity for laudable actions (like raising a family, helping the elderly with their groceries, attending church or sporting events, being neighborly, avoiding crime), mundane though they may be, will help ensure everyone has a comfortable place in society even if they are not among the cognitive elite.
Afterword – The book cautiously discusses mainstream science. It was viciously attacked as racist upon its release, almost always by people who had not read it. The attacks came and continue because it challenges widely held beliefs about society as the cause of problems, government as the solution, and the manipulability of human beings in reaching the goals of equality. The attacks backfired. By generating notoriety, they caused many people to investigate the claims made. Nearly all of the claims have been substantiated because they were intentionally understated. Attempts to discredit the claims in the book have unintentionally confirmed them.
Howard Gardner’s “Frames of Mind: The Theory of Multiple Intelligences” and Stephen Jay Gould’s “The Mismeasure of Man” are the most commonly cited refutations of “The Bell Curve.” They both reject g. Gardner insists that there are multiple intelligences but has not quantitatively demonstrated them, has not shown that they are independent, and has not shown that they are useful in predicting anything about an individual in the same way as g. Gould criticized g as a statistical concept having no basis in reality under the factor-analytic argument. His refutation appeared as a review of “The Bell Curve” in New Yorker magazine. Murray dismisses Gould’s criticism as having no scholarly standing at the time he wrote it, and having been more widely discredited with additional research. An article appearing in the Wall Street Journal on December 13, 1994, entitled “Mainstream science on intelligence” was signed by 52 scholars who endorsed the findings of “The Bell Curve” but did not mention its critics by name. Murray states that g exists and is verifiable even after all of Gould’s criticisms are duly entered.
Murray believes the claims in his book are beyond significant scientific dispute. He thinks the biggest sleeper effect of his book is the degree to which it undermined both socioeconomic status as a way of interpreting social problems, and also the rationale for many social policies that came into vogue in the 1960’s. Social sciences riddled with taboos and self-censorship must be disinfected with truth. Only the truth can possibly lead to better social policy and genuine progress.
My thoughts on the book are in the comments section below because I ran out of space here.