While marriage rates have been declining among all Americans, one of the most visible trends is the apparent difference in first marriage patterns between African Americans and Whites. These variations are reflected in the lower proportion of African American women who marry. Marriage rates have declined substantially across cohorts for African American women but modestly across cohorts for White women (Bennett, Bloom, & Craig, 1989; Rodgers & Thornton, 1985; Schoen & Kluegel, 1988). According to an analysis conducted by Norton and Moorman (1987), from 1975 to 1985 there was a widening gap between the marriage rates of African American and White women, with the likelihood of African American women never marrying increasing. These decreasing marriage rates have not been accompanied by decisions by younger women to delay parenthood (Cherlin, 1988).
It is the weakening link between marriage and childbearing that has captured the attention of social scientists. Family formation patterns are changing for Whites as well as African Americans, yet the most pronounced changes are occurring among African Americans. Jencks (1989), studying the marriage and fertility trends of African American and White women from 1960 to 1986, found that the proportion of children born to unmarried mothers rose from 2% to 16% among Whites compared with an increase from 23% to 61% among African Americans. A recent analysis of United States census data compared trends in marriage and fertility between 1982 and 1992. According to this source, the rate of nonmarital births increased from 49% to 67% among African Americans and rose from 10% to 17% among White women (U.S. Department of the Census, 1993).
Since World War II, the number of mother-only families has increased for both Blacks and Whites. However, when social scientists have carefully evaluated the reasons underlying the changing family structure, racial differences have again emerged. For example, Garfinkel and McLanahan (1986) examined changes in mother-only families from 1950 to 1980 and found that for Whites there was an increase in the prevalence of formerly married mothers. For African Americans, however, there was a decline in formerly married mothers, offset by an increase in households headed by never-married mothers. When Cherlin (1992) evaluated the fertility patterns of African American women from 1960 to 1980, he found that a "far higher proportion of Black children are born to young, unmarried mothers than is the case for White children" (p. 96).
While these changes in marriage and fertility patterns were occurring within the African American family, there was a concurrent shift in emphasis in discussions about poverty, focusing on the characteristics of poor families. During the 1980s, attention has been increasingly directed to the intergenerational correlation of persistent poverty and the marital status and race of the mother. Bane and Ellwood's (1983) longitudinal study of welfare spell duration found that women who had never been married received welfare benefits for a significantly longer duration than divorced or separated mothers. Further examining the relationship between marital status and persistent poverty, David Ellwood (1986, 1989) found that never-married African American mothers were less likely to earn their way off of welfare.
When Greg J. Duncan (1984) evaluated Panel Study of Income Dynamics (PSID) data collected between 1969 and 1978 on two isolated groups of low income families, he found that while most of the temporarily poor had an adult male in the household, the much smaller group of persistently poor families were headed by women. Households that were both African American and headed by women comprised 31% of the persistently poor.
The relationship between race, poverty, and family structure is perplexing to social scientists in that African American women have historically had high rates of labor force attachment when compared with White women. …