Bergmann, Barbara. In Defense of Affirmative Action . New York: Basic Books, 1996: 36-41

There are, of course, those who claim that women's position in the labor market is and has always been exactly what women wish it to be. Men would not dream of oppressing their sisters, wives, and mothers in the labor market. In this view, it is not sex discrimination that keeps women out of "men's jobs" but the free desire of most women to give priority to home duties.' As we shall see, there is good evidence that white and African American women do suffer considerable discrimination on the job because of their sex.

In judging the conflicting claims about the state of the labor market, it is useful to start by looking at how much change has actually occurred. Chart 2.1 shows the weekly wages of those who worked full-time in the years 1967-95, corrected to eliminate the effect of inflation. The inflation-corrected wages of white women have been on a downturn since the mid-1970s. However, white men have not lost their superior position in the labor market: a substantial gap remains between their wages and those of white women and black men and women. Given the slowness of change in the labor market, as shown in chart 2.1, that gap will not close anytime soon.

Modest reductions have been made in that gap since 1967.
Black men's wages were 69 percent of white men's in 1967. By
1976 their wages had risen to 79 percent of white men's. Since then, they have been losing rather than gaining ground on white men. The loss of manufacturing jobs, some of them unionized and thus relatively well-paying, has hit both white and black men, but the latter have been particularly hard hit. ' White women gained no ground on white men until the early 1980s; they have been gaining in the years since. In 1995 their wages were 73 percent of white men's, compared with 6l percent in 1967. Black women have made gains throughout the period, but recently their gains have not matched those of white women In 1995, black women's wages were 63 percent of white men's.

The fall in black men's wages relative to white men's over the last twenty years suggests that whatever help they have received from affirmative action has been modest at best, and has not been enough to counterbalance the effects of their buffeting from market forces. The globalization of the labor market has reduced the demand by U.S. employers for the labor of the less skilled both black and white-- and black men have suffered disproportionately. While affirmative action has allowed some college-educated black men to enter the middle class, the deterioration of the labor market for non-college-educated black men has been disastrous. It has made their lives increasingly precarious; their decreased chances for decently paid work have contributed to the fall in the black marriage rate, the increase in single parenthood, and the recruitment of black men into crime and the drug trade in the inner-city.

An end to the long-term slide in the wages of less-skilled workers would obviously be highly desirable for its own sake. That slide is having many unfortunate effects, of which the worsening climate for affirmative action is only one. The drastic social and political consequences of a continuation of that slide can easily be foreseen. There is no way to attack it that is guaranteed to improve matters and that. does not pose formidable political difficulties. That said, it has to be realized that there is never an easy time to make efforts at reducing the disadvantages of groups that have been subordinated. Allowing those efforts to stall or go into reverse would exacerbate our social and political problems.

Not all of the gap in wages between white males and other workers is due to discrimination in the workplace; a wage gap may also reflect the differences between workers in education, experience, and other factors, such as residence in low-wage regions. Statistical techniques allow us to measure how much of the gap is due to nondiscriminatory factors and to remove their effect from the gap, leaving a "residual gap," which is a better approximation of the extent of wage loss due to current workplace discrimination. Table 2.1 presents two estimates of the residual gap between white males and other groups, based on two different sources of information on conditions in the labor market. Neither source tells the whole story, but taken together, they give us an approximate idea of the size of the penalty that current discrimination exacts from each group.

In table 2.1, one set of estimates of the residual gap derives from information on people's wages collected by the National Longitudinal Survey of Youth (NLSY). The NLSY tells us about the workers' sex and race, as well as their education, age, experience, area of residence, and scores on a test of cognitive skills. The second, considerably larger set of estimates of the residual gap derives from Census Bureau data. A reasonable estimate of the effect of discrimination on the earnings of each group probably lies somewhere between these two sets of figures, both of which pertain to 1991. The NLSY-derived estimates are too low because they cover only workers in the age range 26-33. Much of the effect of discrimination shows up after age 33. In their thirties, some white men are being promoted into well-paid and high-status jobs, whereas women and African Americans who, in their twenties, were placed in dead-end jobs miss out on such promotions for the most part. If the figures derived from the NLSY are too low, the figures derived from the Census data are too high. That data set, unlike the NLSY figures, contains no information on individuals' cognitive skills. Part of the difference in pay between blacks and whites is legitimately due to their differences in this regard: a higher proportion of blacks attended poor schools as children and their families and neighborhoods have few resources to help them develop their skills.

The true penalties of discrimination suffered by average full time black and female workers lie somewhere between the high and the low figures shown in table 2.1: about $3,000 a year for black men and $5,000 for black and white women. These estimates, though only approximate, suffice for our purpose, which is to decide whether discrimination effectively ended some time in the past. The evidence on earnings from the Census Bureau and the NLSY suggests that discrimination is still very much alive.

Carl Gutierrez-Jones,

Department of English

University of California

Santa Barbara, CA 93106

E-mail: carlgj@humanitas.ucsb.edu