CONTACT: Stanford University News Service (650) 723-2558
'IQ does not explain black-white income differences,' economist says
STANFORD--Recent claims that IQ at birth determines one's fortunes in life are based on flawed interpretations of statistical data, according to labor economist Martin Carnoy, a Stanford professor of education.
In Faded Dreams, to be published this month by Cambridge University Press, Carnoy argues that contrary to Richard Herrnstein and Charles Murray's assertions in The Bell Curve, large changes in African Americans' achievement and income can result, and have resulted, from government action.
Using detailed census data from the past 50 years and many of the same sources cited in The Bell Curve, Carnoy finds that positive changes in the ratio of black to white economic performance mark the period.
"Herrnstein and Murray repeatedly slide between measures of achievement, which can be altered by a different learning environment, and IQ, which they identify as genetically determined and immutable," Carnoy said. "To support their claims of the relationship between IQ and income, they rely on studies of the relationship between individuals' wages and a mixture of their aptitude and acquired knowledge. All three components - wage structure, aptitude and acquired knowledge - can be and are affected by government policies, from programs that improve low-income mothers' prenatal care to early education to the level of minimum wages."
In Faded Dreams, Carnoy suggests that reduced poverty in the 1960s and early 1970s - combined with back-to-basics school reform in the 1970s - produced a major increase in the school achievement of low-income children, especially blacks. No single program caused the change, he said. Rather it derived from the attention Americans paid to the plight of both white and minority poor in those years.
Gains in African American children's school test scores, including SAT scores, relative to whites over the past 20 years were expected to result in rising college graduation rates and higher wages for African American high school and college graduates, he said. Neither occurred, but no one, including Herrnstein and Murray, blames this on IQ differences.
"The Bell Curve recognizes the existence of this huge increase in scores," Carnoy explained, "but it conveniently ignores its implications for their theory on the importance of IQ, which tightly connects IQ and wages. If IQ were so consequential for wages, young black high school graduates should have cut the wage difference with whites by about one-third in the 1980s. In fact, young blacks' wages compared to white graduates' were the same in 1979 and 1989."
Carnoy shows that the achievement-income link is also weak for other groups.
"Asian Americans score higher than whites on tests. If we took The Bell Curve's argument about income determination seriously, Asian American male college and high school graduates should earn significantly more than whites of the same age. Yet they earn less," he said. "Men in general continue to earn more than women even when their labor market experience is the same. Married men also earn higher wages than single men, yet no one would claim that marriage makes you smarter.
"When you add all this up, you quickly understand that there are a host of factors affecting wages that make the IQ argument disingenuous.
"What happened to African Americans in the 1980s shows us even more clearly than before that there is no natural causality between higher achievement and income - that other factors, including labor market discrimination and financial barriers to getting more schooling, have a large effect on a group's life chances," Carnoy said.
As an educator, Carnoy said he has "no doubt that more education increases the possibilities for an individual to do well in work, mainly by giving him or her a sense of confidence and greater trainability. I also agree that higher achievement has an impact on how much education people pursue and an impact - although a much smaller impact - on how much they earn once they complete a degree."
Faded Dreams provides a different explanation for what has happened to blacks by analyzing black-white income and education changes over five decades. One of the problems with the analysis of The Bell Curve, Carnoy said, is that it relies too much on data from one narrow time frame, the 1980s, missing what happened earlier. Thanks to affirmative action in the 1960s and 1970s, large numbers of low-income, minority Americans got college degrees and earned much higher incomes than their parents. "They form what is now the black and Latino middle class," he said.
Government policies beginning with the 1964 Civil Rights Act also worked to raise African American children's performance in school. But the decline in the 1980s of government grant programs for college and the decrease in anti-discrimination efforts in the labor market are major reasons that higher achievement by younger blacks never translated into more college or better jobs for them. Other explanations, such as one some economists offer about the changing industrial structure, do not fit the data nearly as well, he said.
Evidence for a policy-driven explanation
The evidence for what has happened to low-income minorities can be seen most clearly, Carnoy said, by looking historically at the relation between wages and other variables, including education, for African Americans.
African Americans steadily increased their schooling relative to whites from 1940 to 1990, and when government policies broke down racial barriers to better jobs, they made large wage gains, he said. National school achievement testing, which began in the late 1960s, shows that those blacks born since the civil rights movement also have substantially improved their school performance relative to whites. But neither the enhanced test scores nor high school completion rates of the group now in their 20s has translated into better jobs relative to whites. In other words, the test scores are not reflected in higher wages as it was for blacks with similarly high test
scores who are now in their 40s and 50s.
More stunning, however, is the fact that those who went to college in the 1980s, and who also had much higher SAT scores than blacks who went in the 1970s, have lost more ground to whites than those who only finished high school. Such findings, he said, undermine the argument of former Education Secretary William Bennett and other conservatives who say that African Americans are responsible for their own economic problems by not taking enough advantage of schooling.
"Black and Hispanic children made big gains relative to whites on test scores from 1975 to 1989. There is still a substantial gap, but blacks cut the gap in half at all school ages on national tests of reading and by more than one third on math. If you really believe that achievement is primarily what determines wages, you should see that progress in the wage statistics," Carnoy said.
"When you don't see it, you have to start to ask, does the achievement score really play a role in the hiring process as many education and business groups have been claiming it does?"
The Bell Curve relies on the Armed Services Vocational Aptitude Battery test for all kinds of questionable claims about worker productivity and wages, Carnoy said, "but the elements of the ASVAB that have had the biggest effect on wages in a tested group were mechanical skills and computational speed, not exactly what most people think achievement or IQ tests measure."
Much of the difference in average family incomes of racial groups can be explained by differing educational levels, Carnoy said. But he and doctoral student Hugh Daley found that 13 percent of the difference between young white and black workers' wages could not be explained by work experience, education, or the industries and geographic regions where whites and blacks work. A gap about half of the difference between whites and blacks exists between whites and both native-born Latinos and Asian Americans, he said. In this work, the researchers estimate the hypothetical incomes for each group in various census years by using the wage payoffs of schooling, experience and other variables as they occur for whites.
Carnoy said he believes that these income differentials are at least partly attributable to continued racial discrimination by employers.
Go to college still good advice
Neither arguments about IQ nor about discrimination mean that parents and teachers should stop giving children the advice to stay in school as long as possible and to study hard, Carnoy cautions. The correlation between test scores and wages that is suspect applies only to large statistical groups and not single individuals.
"You never advise your children against finishing college," he said, "because college broadens their skill base. You need a broad skill base because the individual has no control over the labor market situation. It is changing so fast that jobs you thought you would have are gone before you know it."
Since the Civil War, African Americans have placed a high emphasis on using education to catch up, he said, and today's data simply suggest that "Booker T. Washington had it right: More education works best in improving blacks' economic condition when they have access to better jobs."
African Americans and Latinos were more able to reduce their income gap with whites in the 1940s and 1960s than in other decades, he said, not only because they were getting more education, but because of positive factors in the economy and government. "In the '40s, those factors included a rapidly growing industrial sector increasingly accessible to blacks and a more equal income distribution among workers. In the 1960s, economic growth and government action to limit wage discrimination came into play."
Further, Carnoy said he believes the improved conditions for black adults in the 1960s led to educational gains by their children later. Test score gains in the 1970s and 1980s can probably be attributed, he said, to the increased education of low-income parents and reduced poverty, as well as to improved school quality, especially in the South as a result of rapid economic growth and desegregation.
Conversely, the deteriorating conditions in African American communities in the 1980s could mean falling achievement in the 1990s, Carnoy said. "The first indication of this effect appeared in the lower 1990 reading scores for black 9-year-olds born in 1981, when poverty rates had risen sharply."
Fluctuating college costs and premiums
Educational requirements for jobs are rising, Carnoy said, not as some argue just because the information age requires workers to have more skills. "The average education is also rising in jobs that have not changed that much," such as clergy, nurses and janitors, he said. This change can be attributed to the fact that more Americans are earning high school diplomas and college degrees.
"In 1976, college enrollment for blacks and native-born Latinos peaked at 33 percent of 18- to 24-year-old high school graduates - the same proportion as for whites," he said. "As more college graduates became available, employers eventually figured out they could now get people with college for the same jobs they had been hiring high school people for."
College enrollments then fell briefly for all groups, he said, "in response to this smaller premium for college degrees."
The United States in the 1970s was "one of the few countries in the world to see a decline in what is called the college premium - the income that college graduates earn over and above that of high school graduates," he said. "At one point, the college premium totally collapsed, but it came back with a vengeance in the '80s.
"Ironically, just as the income payoff to college education rose sharply for everyone, including minorities, low-income groups saw their incomes fall, and it also became more difficult for them to get grants and loans to attend college. This made it less obvious for young blacks and Latinos that going to college could work for them financially. The military became their best shot for more training."
When the premium began increasing, he said, lower-income blacks and Latinos did not return to college in the same proportion as whites for three reasons:
"Economists know that an important cost of college is forgone income, and forgone income is particularly important to poor families," Carnoy said. "If Dad loses his job or is not present and Mom has a lousy job at low pay, everybody in the family needs to work. I think that's what happened to minority enrollments, and probably to low-income whites who were eligible for college, too."
Government policies vs. world economic changes
Two forces, Carnoy said, have combined to undermine the purported link between educational attainment and income. These are a more integrated, competitive global economy and changes in government policy that tilted strongly against labor, such as holding down the federal minimum wage and supporting employers' anti-union action.
Global change has led to American job growth concentrated in low- end and high-end service jobs and to a falling away of middle-range, mostly manufacturing jobs. This structural change has been credited with limiting the American dream of upward social mobility and causing the emergence of a permanent class of both working and unemployed poor.
Theories by conservatives such as Thomas Sowell, a senior fellow at Stanford's Hoover Institution, as well as those of liberals such as University of Chicago sociologist William J. Wilson, Labor Secretary Robert Reich and Sen. Patrick Moynihan (D-N.Y.) all tend to point to a "skill mismatch" - the combination of lower levels of college graduation by African Americans and the shift in demand to high-skilled jobs - as the main reason for their failure to move up economically in the 1980s.
Carnoy is a liberal who once ran for Congress as a Democrat and was Robert Kennedy's Washington, D.C., campaign manager in 1968, but he finds fault with liberal theorists for overstating the effect of skill mismatch and understating the politics of wage and employment policies, which he said have had a major impact on minority workers at all education levels. He also finds fault with conservatives who claim government can do nothing about inequities or that, worse, when it did something in response to the civil rights movement, it only made matters worse for blacks.
"A number of new studies now support the position that the Civil Rights Act and subsequent enforcement of affirmative action had a major impact on blacks' relative wages," Carnoy said. His own analysis suggests that 65 percent of the gain in relative income of full-time-employed black males from 1959 to 1973 and 50 percent of the gain for full-time-employed Latino males can be attributed to a direct reduction in pay discrimination. For part- timers and minority women, there also were gains.
What to do
Minority educational gains followed by a period of a growing U.S. economy are necessary for minorities to make economic gains relative to whites, many economists say. Carnoy faults Reagan-Bush-era conservatives for suggesting that national political leadership and government play no role.
Full employment and greater educational attainment were not sufficient for blacks to raise their income relative to whites in the 1920s, 1950s and 1980s, he said. In contrast, the 1940s and 1960s saw laws passed that had not only direct effects on minority access to jobs and education but the indirect effects of sending a message to the general public about the morality of private discriminatory actions.
Since the 1970s, he said, Americans have lost confidence in government's ability to help solve socioeconomic problems, and "little will be done until political leaders work to restore faith in government as a tool.
"The data suggest that government had to intervene in their favor before blacks could take advantage of educational opportunities. That is still true: Impressive differences continue to exist between learning conditions accorded urban black and suburban white children.”
This is an archived release.
This release is not available in any other form.
Images mentioned in this release are not available online.
© Stanford University. All Rights Reserved. Stanford, CA 94305. (650) 723-2300.