The Race-IQ Non-Controversy

J. Philippe Rushton and Arthur R. Jensen, “Thirty Years of Research on Race Differences in Cognitive Ability,” Psychology, Public Policy, and Law, Vol. 11, No. 2, June 2005.

In 1969, Arthur Jensen resurrected the scientific study of racial differences in IQ, which had fallen into disrepute after the Second World War. His 120-page article in Harvard Educational Review created a controversy that shook the country. It was the beginning of Prof. Jensen’s career as perhaps the most feared and hated — but deeply respected — scientist of our time. Since then, there has been tremendous progress in the study of race and intelligence, and Prof. Jensen has been joined by a score of other scholars willing to endure persecution for studying a subject their colleagues have declared beyond the pale.

The latest issue of Psychology, Public Policy, and Law is devoted to a review of the last 30 years’ findings in this field. The main article, written by Arthur Jensen and Philippe Rushton, is a 60-page survey concluding unambiguously that environment alone cannot account for persistent black-white differences in intelligence. Genes, they argue, must contribute at least 50 percent of the one-standard-deviation gap in IQ test scores that has remained largely unchanged since the early 20th century.

This main article is followed by three critiques, one supporting article, and a final summation by Profs. Rushton and Jensen. These are not papers that present new findings, but in combination they are the most concise, up-to-date summary of the state of the debate now available. As is so often the case in the scientific treatment of politically sensitive subjects, the weight of the evidence is not what most Americans would expect. With only a few exceptions, the arguments of those who support the conventional view are almost laughably weak.

The Data

Profs. Rushton and Jensen review a great deal of research in a concise way that makes few concessions to non-scientists. They begin by pointing out that barring something extraordinary, the same combination of genetic and environmental factors that explain individual differences in IQ should explain racial differences too. There would have to be some race-specific and very powerful environmental force at work in order for the well-understood genetic contribution to individual differences in IQ to play no role at all in racial differences.

Needless to say, that force is alleged to be “racism,” past, present, blatant, and subtle. Genes therefore contribute to individual IQ differences among blacks, but “racism” is so powerful it shoves the entire bell curve of black IQ distribution 15 points to the left. For those who promote the “culture-only” theory — that genes play no role in race differences in intelligence — blacks and whites are, essentially, groups of identical twins separated at birth, but blacks are subject to unique and debilitating environmental forces that hold them down.

There is little argument that genes contribute a great deal to individual IQ differences within races. Family environment can push a child’s IQ up or down, but adolescents increasingly seek out environments that match their own inclinations, with the result that in adults, the genetic contribution to IQ is overwhelming.

When full siblings are reared together, their IQs are most similar when they are children, while they are still within a family environment over which they have no control. This similarity drops to a correlation of .49 in adulthood, almost exactly the figure one would expect for people who have half their genes in common. For full siblings reared apart, the correlation in IQ is only .24 in childhood — the family environments are different — but the similarity increases as they choose their own environments, to the point that the correlation likewise becomes.49 as adults.

When unrelated children are adopted into the same home, their shared home environment gives them a correlation in IQ of .25, but this correlation drops to .01 in adulthood — the figure of essentially zero one would expect for two people picked at random from a population. The lesson is clear: By the time we are adults, whatever boost or handicap we may have had as children from our families has washed away, and our intelligence reflects our genes far more than it does our environments.

Of course, even children reared in the same family do not have exactly the same environment. Accidents, disease, chance friendships, as well as genetic predispositions that push a child one way rather than another can make the same household seem a very different place even for twins. By statistically separating out the effects of shared environment (common family characteristics like social class, geographic location, type of housing) from unshared environment (experiences unique to each child), social scientists get the results shown on the graph on this page. The shared, household environment has a strong effect on children while they are still at home, but not after they leave. By the time they are adults, genes and only unique childhood experiences contribute to differences in IQ. These results hold for people of all races.

This pattern is difficult to explain if environment alone explains racial differences in IQ. If environment, at least as it relates to race, is so powerful, one would expect its effect on individuals to increase over time, not decrease as it actually does. Furthermore, as Profs. Rushton and Jensen point out, the one-standard-deviation difference in black-white IQs appears as early as age three, on the Sanford Binet IV test for children, even for black and white children whose mothers have the same level of education. If blacks suffer from some pernicious environmental effect, it must enter their homes soon after they are born, and do its work before most black children could have had any meaningful contact with “racism.”

At the same time, culture-only theorists usually point to white advantages in the shared family environment (parents’ income, good neighborhoods) as important reasons for high white IQs. Within races, however, these advantages essentially count for nothing in the long run; why should they produce enduring differences between the races?

The culture-only theory fails to predict virtually all of the research results of the last several decades. Ability, aptitude, and IQ tests differ greatly in whether their content is “cultural” and can be learned, or consists of unusual challenges virtually no one has practiced or could anticipate. Profs. Rushton and Jensen note that the forward and backward digit tests are a good example of this difference. In the first, a tester recites longer and longer strings of numbers, and the subject repeats them back as best he can. It is essentially a memory test. In the reverse digit test, the subject listens to strings of numbers but must repeat them back in reverse order — a much harder task. There is a small white advantage on the forward digit test but a large white advantage on the reverse digit test. Repeating numbers in order is part of our “culture” but reversing them mentally is not. If whites have a “cultural” advantage over blacks, why is the ability gap so much greater in the reverse test, for which nothing in the environment of whites specially prepares them?

Moreover, scientists have shown that ability on the more demanding test is more heritable than on the easier test, which is to say that environment has little effect on how well someone does. Performance on the reverse digit test is largely a matter of native ability and is not improved by fancy environments. This is consistent with findings in intelligence testing across the board: The most heritable abilities, as well as those that most directly test for intelligence (or g  — the general factor for intelligence) are the ones for which the black-white gap is greatest. This is exactly the reverse of what we would expect if white performance were due to better environments rather than genes.

Something else for which the culture-only theory has no explanation is the fact that the children of high-income blacks have lower IQ scores than the children of low-income whites. Surely, if IQ improves with good surroundings, rich blacks should be able to give their children better, more IQ-boosting conditions than poor whites give theirs.

In families of all races, rich people tend to have smarter children than poor people, but the environments they provide account for only a small part of this. By and large, the wealthy are wealthy because they are smart and know how to make money, and their children are smart because they got their genes from smart parents, not because they live in big houses and take harp lessons.

In general, when blacks and whites are matched for social and economic status, the IQ gap decreases, but only by about a third. The black-white gap for children of the wealthiest parents is considerably greater than the gap for the children of poorer parents.

Regression towards the mean — a strictly genetic phenomenon — best explains what is happening. Nature has a tendency to temper extremes. Very tall people tend to have tall children but not as tall as themselves. Very short people are likely to have children not quite so short as themselves, and this tendency to retreat from extremes is called regression towards the mean. It applies to intelligence too, and means that the children of very smart or very stupid parents tend to drift back towards the average for the population. Study after study has shown that black and white children seem to be regressing towards different means: averages of 100 for whites and 85 for blacks.

This tendency is apparent among siblings, and is actually a better test of regression than parent-child comparisons, because parents and children may have had very different environments. A very smart or very stupid child is something of a genetic freak in any race, but very smart children are less freakish among whites and very stupid children are less freakish among blacks. Prof. Jensen demonstrated this with a sample of black and white children, all with high IQs of 120. The siblings of the smart white children had IQs that averaged about 110 while the siblings of the smart black children averaged about 100. An IQ of 120 is abnormal in either race, but is a more dramatic departure from the norm for blacks than it is for whites — as the IQs of the siblings demonstrated.

The effect was the same with very low IQs. For whites and blacks matched for IQs of 70, black siblings had average IQs around 78 while white siblings had average IQs around 85. From both extremes, whites are regressing to a higher average IQ than blacks, a finding the culture-only theory cannot explain.

Intelligence is correlated with a number of clearly biological factors, the most obvious of which is brain size. On average, East Asians have 17 cm³ more brain than whites, who have 97 cm³more brain than blacks. Brain size has about a .4 correlation with intelligence, so these differences alone would explain five points of the 15-point average IQ gap between blacks and whites. Interestingly, when blacks and whites are matched for intelligence, they have the same sized brains, but simply matching for brain size does not produce a match in IQ. Brain size seems to be only one part of the biological puzzle.

Another purely physical indicator of intelligence is the amount of glucose someone’s brain metabolizes when he is trying to solve a problem. Glucose, or sugar, is the main fuel for the brain, and intelligent people require less of it than dim people; their brains use fuel more efficiently. Tests of impulses along individual nerves also show that smart people have faster-acting nerves than less smart people. Intelligence is also associated with levels of acidity in the brain as well as an electrical characteristic known as average evoked potential. Some of these biological factors have been studied only in whites — and correlate with individual differences in intelligence — but the studies that considered race have found the group differences one would expect.

Arthur Jensen has been a pioneer in what are called reaction-time tests. These require a subject to perform very simple tasks that everyone can usually get right 100 percent of the time: push a certain button when a particular light goes on, for example. Intelligence is correlated with speed and consistency of reaction time, and there are racial differences in results. It would be hard to argue that the “culture” of whites somehow better prepares them than blacks for doing this.

Tests have generally shown that American mulattos, who have a significant percentage of white genes, have higher IQs than blacks with little or no white ancestry. The same is true in South Africa, where mixed-race “coloreds” consistently score at an intermediate level between whites and pure-bred blacks. The best evidence suggests that for African blacks with no white ancestry, the average IQ is about 70. This extremely low figure is consistent with the low level of development that characterizes black-run countries.

IQs in this range are found even in the United States, in those parts of the rural South where blacks have very little white admixture. Profs. Rushton and Jensen cite the remarkable case of a school district in Georgia, in which the average black IQ was 70. The whites in the district had an average of 101, so there was not something odd about the district itself that pulled down IQ scores.

The same kinds of evidence that show a genetic component in the black-white IQ gap show a genetic component in the smaller white-North Asian gap. Asian children adopted by whites are often more intelligent than the white average. One study designed to evaluate the effects of early malnutrition tested a large number of Asian adopted children who had suffered various degrees of deprivation. As expected, IQ scores were higher among the Asians who had suffered the least, but even the worst-nourished Asians were above the white average, and the best-nourished Asians were far above average. In this connection, Profs. Rushton and Jensen ask why alleged white “racism” holds down black IQs whereas Asians (and Jews), who have had histories of mistreatment, have higher IQs than gentile whites.

Prof. Rushton has done important work in tabulating the many and varied ways in which the major races differ from each other. The table on the following page summarizes differences that all point to a particular pattern: At one extreme, Asians give birth to larger-brained, slower-maturing children who receive a great deal of care from their parents. Blacks are more careless about producing their smaller-brained, rapidly maturing children, and whites are intermediate between the two. Differences in average intelligence are just one part of what can be seen as different approaches the races have followed for successful reproduction.

A more casual approach to child-rearing may explain why blacks seldom do the one thing generally known to raise a child’s intelligence: breast feed him. Black mothers are three times more likely than white mothers to give their children formula. This would be a classic example of an environmental disadvantage for black children, but one that reflects the genetic predispositions of their parents rather than white malevolence.

Profs. Rushton and Jensen conclude that it is a great disservice to society to promote the view that whites are responsible for the failures of blacks if, in fact, the causes are genetic. Moreover, it is futile to expect equal results from populations that are not themselves equal: “Ultimately, the public must accept the pragmatic reality that some groups will be overrepresented and other groups underrepresented in various socially valued outcomes.” Race relations cannot be improved if the fundamental assumptions about them are false.

Robert Sternberg, a prominent Yale psychologist who has spent much of his career combating race-related research, wrote the first critical response to Profs. Rushton and Jensen. It is an embarrassment. First, he blames the authors for even choosing to study “so-called races,” and seems to be saying that such research can only reflect prejudice and self-delusion.

He notes that the authors say low IQ is associated with unhappy results like going on welfare and divorce. Not so, says Prof. Sternberg. In Mexico there is no welfare, so IQ cannot be associated with that, and in some countries divorce is forbidden, so IQ has nothing to do with that either. In like manner, if racial discrimination were eliminated in America, everything we know about race and IQ would turn out to be wrong. In yet another off-the-mark argument, he says high IQ isn’t such a great thing after all, because only smart people can build chemical weapons or get away with terrorist acts. These are some of Prof. Sternberg’s central arguments, not uncharacteristic lapses from otherwise rigorous thinking.

Richard Nisbett of the University of Michigan does considerably better. He points to evidence that the black-white IQ gap is decreasing (with data no more recent that 1998) and predicts that at this rate black children will read as well as whites in 25 years and equal them in science in 75 years. He also points to a 1994 study that claimed significant improvement in black IQ that lasted through age 12, but concedes that it involved no less than eight hours a day of intensive pre-school instruction. He argues that enrichment programs can boost intelligence (and presumably make up for bad black environments) even up to the college level.

Prof. Nisbett also cites several studies of mulattos that did not show that white ancestry increased IQ. Also, according to a 1974 study, children of white mothers and black fathers had IQs nine points higher than the children of black mothers and white fathers. Presumably the mulattos lived with their mothers, and there was a benefit to living with a white woman. He also cites a 1984 study that suggests black orphans end up smarter if they are adopted by whites than if they are adopted by blacks. Among other things, white adoptive mothers reportedly encouraged intellectual development and were more forgiving of mistakes. Prof. Nisbett writes that this sort of evidence is so powerful that all of Profs. Rushton’s and Jensen’s other arguments can be ignored.

In the article that follows, Linda Gottfredson of the University of Delaware points out some of the flaws in the rather old adoption and miscegenation studies on which Prof. Nisbett relies, but agrees with him that this line of investigation is a potentially valuable source of convincing data on the genetic contribution to the black-white gap. The trouble, she points out, is that no one is likely ever to do the large-scale, methodologically sound studies needed to lay the question to rest, because the subject frightens people. She also brings Prof. Nisbett up to date by citing more recent findings on the black-white intelligence gap that suggest it did appear to narrow for some time, but has started widening again.

Prof. Gottfredson also notes that culture-only theorists have repeatedly taken up one position, only to abandon it for a different but weaker one when their original claims were shown to be wrong. Several decades ago, it was fashionable to claim tests were biased against blacks. Now that testing advocates have thoroughly discredited that view, opponents now almost universally assert the much vaguer view that blacks suffer from ill-defined but powerful IQ-depressing conditions of some kind.

Ultimately, she writes, intellectuals seem to agree with Nathan Glazer, who wrote in 1994 about the controversy over The Bell Curve: “I ask myself whether the untruth is not better for American society than the truth.”

The final commentary on the main article is by Lisa Suzuki and Joshua Aronson of New York University. They make the very lightweight argument that genes and culture are so interwoven it is impossible to distinguish their effects. They even argue that brain size, glucose requirements, and nerve conduction speed are partially or even completely determined by environment. One wonders just what it is about the environment of less intelligent people that slows down their nerves and shrinks their brains.

Profs. Suzuki and Aronson are also big fans of “stereotype threat” theory. This is the argument that blacks don’t do well on tests because they know they have a reputation for being dim, and the pressure to disprove this “stereotype” so unnerves them they muff the tests. This theory has a very spotty replication record, and is fatally undermined by the fact that black children do badly on tests long before they are old enough to know anything about “stereotypes.” Likewise, blacks do no better on tests for which there are no consequences and no pressure than they do on tests they know could make a big difference in their lives. Also, if “stereotype threat” keeps test scores down, what is it that depresses to an equal degree the school and job performance the tests are supposed to predict? Yet more stereotype threat? Finally, as Profs. Rushton and Jensen ask, how is stereotype threat supposed to work in black countries where everyone with power or prestige is black? This is the kind of quackery people flock to when they are desperate.

This issue of Psychology, Public Policy, and Law (PPPL) is, in short, a crushing defeat for the conventional — perhaps we should say obligatory — view of why blacks do not do as well as whites. And that, no doubt, is why it has been met with almost complete media silence. The official publication of the American Psychological Association has now put before the public a peer-reviewed compilation of arguments that blow to bits one of the country’s most cherished illusions. The articles that are supposed to defend orthodoxy are so weak, and ignore so much of what Profs. Rushton and Jensen wrote, that only the most frothing believers in culture-only theory could think the result was anything but a humiliation.

This issue of PPPL should have touched off a frenzy of press coverage. Imagine what would happen if a sober, top-tier, peer-reviewed journal published overwhelmingly compelling evidence that exercise causes cancer, say, or that nuclear waste makes good lawn fertilizer. When a prestigious journal presents scientific findings that overturn the conventional view on important questions, the press is supposed to pay attention. And, indeed, Profs. Rushton and Jensen issued press releases and distributed hundreds of copies of the journal in the expectation that the press would pay attention. But this is, after all, the 21st century, and the subject is race differences. What would be top news were it about any other subject goes unremarked.

Some day, genetic researchers will isolate the genes that contribute to human intelligence. If they are brave enough to look, they will find that these genes are not distributed equally among all population groups. Maybe a top-rank journal will even publish these findings. Will anyone pay attention?