Income Inequality and IQ

Charles Murray, Bradley Fellow at the American Enterprise Institute and coauthor with Richard J. Herrnstein of The Bell Curve: Intelligence and Class Structure in American Life, addressed the relationship between intelligence and income inequality at the twelfth seminar in AEI's Understanding Economic Inequality series.

A revolution has occurred in the United States in this century without producing headlines or even attracting public attention. The revolution begins with higher education.

In 1900, the vast majority of very bright young adults did not get a college education, and even by the mid-1920s only about half of those in the top IQ decile were entering college. By the 1960s, the probability that those in the top IQ decile would enter college was almost 100 percent, regardless of whether they were rich or poor, white or black.

At the same time, a stratification took place within higher education, so that the very brightest began to attend a select group of elite institutions, although this had not always been the case. The mean SAT score for entering freshmen at Harvard rose so dramatically, for example, that the average freshman in 1952 would have been in the bottom 10 percent of the class in 1960. While income does not depend directly on IQ (as is illustrated by the case of college professors), a person's initial job placement after college, which in turn has a heavy bearing on eventual earnings, has much to do with where that person went to college.

Furthermore, during the course of the twentieth century the number and proportion of jobs that are highly screened for IQ--such occupations as law, medicine, engineering, accounting, and science--have increased substantially. In addition, it appears that businessmen today tend to be screened systematically for IQ, both by educational requirements and by the cognitive demands of a business environment that depends on complex technologies.

Other changes have made many jobs associated with high intelligence far more lucrative than they were a century ago. In constant dollars, an engineer earned about $30,000 in 1952 compared with $20,000 for a manufacturing worker, which was not much different from the ratio at the beginning of the century. By 1988, the engineer earned almost $75,000 compared with $22,000 for the manufacturing worker.

As a result of these changes, a new kind of class stratification based on brains rather than on initial wealth or social status is emerging in the United States. Put another way, wealth and social status still exist, but they are much more closely linked to intelligence than used to be the case.

Are IQ Tests a Reliable Basis for This Analysis?

Many commentators reject the above analysis on the grounds that IQ tests do not measure intelligence so much as socioeconomic status and educational experience. But data derived from the National Longitudinal Survey of Youth (NLSY) and the Armed Forces Qualification Test confirm the general validity of IQ testing and its relevance to the study of economic inequality.

The NLSY contains a sample of almost 13,000 young people who have been interviewed periodically from 1978 on. By 1992, the subjects were twenty-eight to thirty-five years old. The data on siblings in the sample, from identical social backgrounds but often with different IQs, are particularly revealing.

The data on siblings applied to nearly 3,000 sibling pairs who had the same biological parents, lived with their biological parents at least through the younger sibling's seventh year, and had valid scores on the Armed Forces test. Where the IQ differed, the results were striking. For example, where both children had attended elementary and secondary schools for the same number of years, only 18 percent of the siblings with "normal" IQs (in the 90 to 109 range) got bachelor's degrees, while 83 percent of their brothers or sisters in the very bright category (IQ of 125 or above) did so.

Occupations told the same story. A measure called the Duncan scale, which runs from 1 to 100, ranks occupations according to pay, prestige, educational requirements, and similar factors. For the siblings with normal IQs, the median Duncan score was 41. For the brightest brothers and sisters it was 62--indicating managerial, administrative, and professional positions--and for the dullest it was 11.

As would be expected, the same patterns emerged when the measure was income. Median income was $19,000 for the middle IQ range, for the brightest it was $30,000, and for the dullest it was $7,500. What is more, since these people are still comparatively young, the income gap will grow. Particularly for those in the brighter groups, incomes can be expected to increase--in some cases vastly--as they get older, while those in occupations lower in the Duncan scale cannot expect such increases later in life. The relative importance of IQ in determining income appears to be similar for all ethnic groups.

Inequality in intelligence is part of the human condition and inescapably contributes to a substantial degree of income inequality that greater equality of opportunity cannot circumvent.