Daniel Altman

The Supreme Court's decisions on affirmative action brought a raft of social issues to the fore last week. But what of the economic consequences?

By its nature affirmative action takes innate attributes that may have nothing to do with ability and makes them part of an educational or professional selection process. An economist's knee-jerk reaction to any policy like this is usually that it will keep a market from matching buyers and sellers in the most efficient way. In other words you had better have a darn good reason to do it.

But given what economists know about educational and professional attainment, this point of view is naive and incomplete. There are definite connections between generations; a parent or grandparent's success, material or otherwise, can directly influence a child's prospects in life. In addition, bad blood within communities can harm the smooth workings of a society, not to mention an economy. The theory seems especially inadequate when faced with the reality. Careful studies of the economic effects of affirmative action have found mostly an upside.

Research conducted in the 1980s by Jonathan Leonard, chairman of the Haas Economic Analysis and Policy Group at the University of California at Berkeley, found that affirmative action had increased the employment of minorities and women in the 1970s.

In the 1990s, Harry Holzer a professor of public policy at Georgetown University, and David Neumark, a professor of economics at Michigan State, looked at how workers in those demographic groups fared. Examining data from four major metropolitan areas, they found that companies with affirmative action policies employed blacks and Hispanics with slightly less education or credentials than those employed by companies without the policies. But the professors also discovered that the performance of minority workers hired under affirmative action was no worse than that of those who were not.

At least indirectly, these studies give credence to an idea long held by supporters of affirmative action: that most standard measures of aptitude are biased against minorities. In a market for labor based on these measures minorities would be underemployed relative to their productive potential.

One might argue that the labor market in the United States is much more subtle, taking account of many different markers of future success. Yet can it delve beyond all superficial traits to measure a person's true potential, without bias? If it could, there would be much less need for affirmative action.

Research by Holzer and Neumark showed that affirmative action might help fine-tune the labor market. It appears to increase the use of recruitment and screening procedures for job applicants. It also leads to more hiring of "stigmatized" applicants including those who have criminal records or have been on welfare, who may still be productive workers. In a 1999 working paper, the professors concluded that "the empirical case against affirmative action on the grounds of efficiency is weak at best."

They also noted that affirmative action could still have hurt some groups. Whites, for example, might have missed opportunities to work or learn as a result of affirmative action as the plaintiffs in the Supreme court cases argued. But if this occurred because colleges and companies had begun to value the potential minorities and women more accurately, then economic efficiency was probably served.