The Blackwell Companion to Sociology (46 page)

white female earnings ratio for full-time year-round workers was 90 percent, and for all women it was a slightly lower 87 percent (US Bureau of the Census, 1992, p. 57, table 11). If anything, both types of time series indicated evidence of a decline in the ratio in the 1990s. To extrapolate on the basis of current trends would lead one to conclude that divergence, not convergence, is the trend for the future.

We reiterate that an additional basis for skepticism about the convergence

hypothesis is the fact that data on black family incomes have never displayed the same pattern of gap closure as have the data on individual earnings or on income for persons with earnings or income. Between the mid-1960s and 1980 the

black±white family income ratio stayed in a rather narrow range between 61

and 64 percent. During the 1982 recession the ratio dipped below 60 percent. By

Racial Economic Inequality in the USA

187

1990 the mean black±white family income ratio was 62 percent and the median

black±white family income ratio was only 58 percent (US Bureau of the Census,

1992, p. 56, table 10). In 1980 one of the authors of this chapter made the

following observation, suggesting the superiority of relative family income as a gauge of relative racial well-being:

Although neither income nor earnings inequality between black and white women

measured in per capita terms have been great in the postwar era, inequality in living conditions is probably quite substantial. This is a consequence of one of the short-comings of per capita income as an index of relative well being. It ignores

how people come together socially and hence economically. White women combine

their incomes with white men far more frequently than black women combine their

incomes with white men. In fact, the large proportion of single parent, female-

headed black families indicates that black women combine their incomes with

those of black men less often than white women combine theirs with white men.

(Darity, 1980, p. 176)

The incidence of females heading black families underwent a sharp upturn in

the 1970s, accelerating throughout the 1980s, with no evidence of a peak in

sight. We computed comparative estimates of the percentage of black families

headed by women from the Census Bureau's Current Population Survey for

1976 and 1985. Nationally, the incidence of female-headed families rose from

36 to 45 percent of all black families. Particularly dramatic increases took place in the West and North Central regions. In the West the incidence rose from 25 to 47 percent and in the North Central region the incidence rose from 39 to 52

percent. For white families, the direction of movement was the same but the

magnitude was not the same. White families headed by females rose from 11

percent in 1976 to 13 percent in 1985.

Data from the Current Population Survey (see US Bureau of the Census, 1992,

Table 10, p. 56) indicate that the racial income disparity is much narrower

among married couple families. In 1991, the black±white ratio of mean income

for married couple families was 80 percent; the ratio of medians was 83 percent.

However, less than half of all black families, 47 percent (3.57 million out of 7.47

million), were married couple families. Since a near majority of black families

are now female-headed, evidence about approaching parity in relative incomes

between black and white two parent families is misleading. The growing incid-

ence of female-headed black families signals the effects of a process we refer to as the marginalization of black males. This process bears an intimate relationship

to the fragile status of the entire black population in the United States, whether male or female. Economic manifestations of the marginalization process are

plentiful.

Consider the absolute decline in labor force participation rates (LFPRS) for

males of both races that began in the early 1960s and the relative decline that

occurred after the mid-1960s. Between 1954 and 1966 the ratio of black±white

male LFPRs held fairly constant at 98 percent. But over the course of the next

decade LFPRs fell much faster for black males than for white males. As a result, 188

W. A. Darity, Jr and S. L. Myers, Jr

despite the fact that after 1976 LFPRs for black males seem to have bottomed

out, while white male rates continued to fall, a wide gap persists because of the sharp difference in rates of decline between 1966 and 1976, the very period that Freeman and Smith and Welch have touted as the period of dramatic progress for

blacks.

By 1985 the employment±population ratio for adult black males was only 55

percent, while it was 70 percent for adult white males. Almost one out of every

six black men between the ages of 25 and 54 reported zero earnings for 1984. By

1990 one out of every three black men 15 years old and over reported no

earnings (US Bureau of the Census, 1992, p. 57, table 11). There also appears

to be a bifurcation among black males based on credentials, between those with

a college education or better and those with a high school education or less. The former group has far more earnings experience in the labor market than the

latter. Pecuniary prospects are dim for young black males with a high school

education or less.

In a very interesting and revealing study, Bluestone et al. (1991) have provided a typology of Metropolitan Statistical Areas (MSAs) based on growth rates of

total non-agricultural employment and total manufacturing employment in the

1980s. They classified (1,1) cities as those with the slowest growth rates in

overall and manufacturing employment, (2,2) cities as those with mid-range

growth rates in both categories, and (3,3) cities as the fastest-growing in both categories. They then compared unemployment rates by race for males 20 years

old with less than a college education in each type of city.

The results that they report are provisional because the sample cell sizes

become relatively small for black males. But in the (1,1) cities they find that

among 20-year-olds the black male unemployment rate was 42 percent while

the white male unemployment rate was 4.1 percent. In (2,2) cities the black

male unemployment rate was 26.5 percent while the white male unemployment

rate was 3.5 percent for the same age group. Finally, in (3,3) cities the black male unemployment rate was 17.9 percent while the white male unemployment rate

was only 1.5 percent, twelve times lower than the black rate! (Bluestone et al., 1991).

Conditions became so patently bleak that even Richard Freeman performed a

complete flip-flop on his research agenda. In the 1970s Freeman sought to

explain what he labeled thè`dramatic economic progress''of black Americans.

Now he was asking, in collaboration with John Bound, ``What went wrong?

Why have relative earnings and employment fallen for young black males?''

(Bound and Freeman, 1992).

Exposure to college education does not alter comparative conditions substan-

tially for black males. Black males 25 years of age or older with one to three

years of college education and with positive earnings received, on average,

$22,979 in 1990. White males 25 years of age or older with earnings, who had

dropped out of high school, had mean earnings of $19,270, about $3,700 less.

White males who completed high school earned more than $3,000 more than

black males with some college education. Nor does completing college close the

gap. A black male with four years of college and positive earnings in 1990 would Racial Economic Inequality in the USA

189

have received mean earnings of $32,259, more than $10,000 below the mean of

$43,919 earned by white male college graduates with positive earnings (US

Bureau of the Census, 1992, table 12, pp. 47, 52).

Convergence in educational characteristics need not translate into conver-

gence in earnings. Bound and Freeman (1992, p. 16) acknowledge that the

increasingly depressed labor market experience for young black males is not

due to a corresponding deterioration in their relative quality as potential

employees: ``We find little support for the hypothesis that deteriorated labor

market skills of young blacks due, say, to poor schooling, worsened family

background resources, or increased drug use explains their declining economic

position.''

Two key points need to be made. First, the Bound±Freeman (1992) discussion

implies that the current trajectory will only worsen as more and more black

males come into the labor market from younger cohorts that are now experien-

cing even greater resource deprivation. Second, and more fundamental, inde-

pendent of the relative quality black males present to the labor market, processes persist that exclude them from comparable participation with non-black males

with similar productivity-linked characteristics. To put it bluntly, discrimination in a comprehensive sense lies at the core of matters, not just at the point of

employment but throughout an entire range of stages that affect labor market

outcomes. American society is quite distant from attaining a pure equal oppor-

tunity environment, never mind correcting for historically inherited inequalities.

Current discrimination, while creating a disadvantage for the group against

which it is directed, necessarily creates a corresponding advantage for another

group. In 1991 the Urban Institute conducted an audit study of entry-level job

access in the cities of Chicago and Washington, DC. The researchers paired

young black and white males (19±24 years of age) with fake, identical creden-

tials and had them apply for the same jobs. The subjects applied for a total of

1,052 jobs. Treatment of the applicants appeared to be neutral 73 percent of

the time, but black males faced discriminatory treatment 20 percent of the time, while white males faced discriminatory treatment only 7 percent of the

time. Therefore, black males were three times as likely to face job bias. There

were also higher rates of discrimination against black males in white-collar jobs or jobs involving direct client contact (Lawlor and Pitts, 1991, p. 1A).

Gerald Jaynes has written, ``The heralded recovery of the mid-1980s appar-

ently improved the position of white low skilled men but not black. Why? We do

not know'' (Jaynes, 1990, p. 24). But we can make a highly educated guess. The

changing structure of the US economy was narrowing the availability of well

paid blue-collar jobs, jobs that pay well without requiring advanced educational credentials. White labor's discriminatory behavior would naturally intensify

under such conditions, particularly if the heralded recovery of the mid-1980s

did not mean a sharp reversal of the deindustrialization of the US economy. It

was simply a classic case of protecting one's occupational turf from darker rivals, and it reveals the endogeneity of discriminatory practices. While much economic

research proceeds as if discrimination operates at a constant level of intensity or does not exist at all, a more plausible analysis would indicate that the degree of 190

W. A. Darity, Jr and S. L. Myers, Jr

discrimination fluctuates, in part, with the nature of aggregate conditions of

employment, without disappearing.

The only period where measured levels of in-market discrimination against

blacks declined significantly since Emancipation was during the decade

immediately following passage of the Civil Rights Act of 1964. Thereafter, the

level of labor market discrimination has shown no further tendency to decline

significantly. From about 1975 to the present black men typically have lost

12±15 percent in earnings due to labor market discrimination (Darity, 1998).

The Widening of the Racial Earnings Gap

More recent studies on black±white economic inequality have focused on the

reversal in the pattern of alleged convergence in earnings between black and

white males in the 1980s. The puzzle is: why is there now evidence of a renewed

widening of the gap in black and white male earnings? The thrust of a substan-

tial portion of recent work on black±white economic inequality is to explain

divergence in the 1980s by restoring the pure human capital explanation of

racial economic disparity. Correspondingly, the operation of the labor market

is immunized from playing a role in directly producing racial economic inequal-

ity. The upshot is the claim that the persistent black male lag in earnings is due to deficiencies in productivity characteristics of black males, rather than a failure of the market to offer them rewards parallel to those of white males with equivalent productivity characteristics.

The skills mismatch hypothesis is used in this context in several studies.

According to this line of argument, black males possess significantly fewer skills on average than white males. In the 1980s relative rewards for highly skilled

workers rose vis-aÁ-vis less-skilled workers. Since black males are disproportionately represented among less-skilled workers, their earnings have fallen relative to white males as relative rewards for skills have moved in favor of those with

higher levels of skill acquisition.

Skills are measured in many of these studies by Armed Forces Qualification

Test (AFQT) scores available in the 1980 National Longitudinal Survey of Youth

(NLSY). Inclusion of these scores in regression analyses that seek to examine

racial earnings differences virtually eliminates unexplained residual differences in earnings between blacks and whites. Studies by O'Neill (1990), Maxwell