National Academies Press: OpenBook

Urban Change and Poverty (1988)

Chapter: Family Structure, Poverty, and the Underclass

« Previous: Part II: Background PapersIncome, Opportunities, and the Quality of Life of Urban Residents
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 102
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 103
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 104
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 105
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 106
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 107
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 108
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 109
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 110
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 111
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 112
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 113
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 114
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 115
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 116
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 117
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 118
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 119
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 120
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 121
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 122
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 123
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 124
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 125
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 126
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 127
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 128
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 129
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 130
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 131
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 132
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 133
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 134
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 135
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 136
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 137
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 138
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 139
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 140
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 141
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 142
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 143
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 144
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 145
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 146
Suggested Citation:"Family Structure, Poverty, and the Underclass." National Research Council. 1988. Urban Change and Poverty. Washington, DC: The National Academies Press. doi: 10.17226/1096.
×
Page 147

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Family Structure, Poverty, and the Underclass SARA MCLANAHAN, IRWIN GARFINKEL, and DOROTHY WATSON INTRODUCTION AND CONCLUSIONS Family structure and household composition have changed dra- maticaDy during the past two decades. Young adults are more likely to live apart from their parents today than they were 20 years ago, the aged are less likely to live with relatives, and children are more likely to live in households headed by single women (Bureau of the Census, 1984~. Of all these changes, the growth of mother-only fam- ilies is perhaps the most striking and has certainly stimulated the most concern. This concern arises in part because of the economic insecurity of these families nearly half are poor, and most of these poor fannies are dependent on welfare and in part because mother- only families may be linked to the growth of an underclass. Over half of all children born today will spend some time in a mother-only family, which means that this family form is playing a major role in shaping the next generation of Americans (Bumpass, 1984~. This paper begins by describing the increase in the number of families headed by single women over the past several decades. We examine both overall trends and trends in central cities to determine whether the change in family structure is more or less prominent in urban areas than in suburban and nonmetropolitan areas. Following the description of trends, we discuss the causes of the growth in such families, including the effects of increases in income transfer programs 102

FAMILY STRUCTURE, POVERTY, AND THE UNDERCLASS 103 and of changes in men's and women's employment opportunities. We conclude that for whites the major cause of growth has been the increase in women's employment opportunities; for blacks it has been the decline in mate employment. Increases in welfare benefits accounted for about 10 to 15 percent of the growth in mother-only families between 1960 and 1975. Some analysts have equated the growth of families headed by women with the Feminization of poverty and the emergence of an "underclass. In the second section of the paper, we examine these claims and attempt to clarify the relationships among single motherhood, poverty, and economic dependence. We conclude that the feminization of poverty is not a particularly useful concept for understanding the economic status of mother-only families inasmuch as it implies an increase in poverty during a time when rates actually declined. Moreover, a large proportion of mother-only families are not poor, even though they may have experienced large drops in income as a consequence of marital disruption. This is not to say that poverty is not a serious problem. Mother-only families are more likely to be poor than any other major demographic group, and they stay poor longer than other groups. An empirical analysis of the sources of income for mother-only fannies indicates that a major cause of poverty is the low earnings of single mothers. Despite their status as family head, single mothers earn an average income that is between 30 and 40 percent of the earnings of married fathers. The absence of child support from noncustodial parents and low welfare benefits in most states also contribute to income insecurity and poverty. Have mother-only families contributed to the growth of an un- derciass? This question is addressed in the third section of the paper. Although the term underclass has been used in a variety of ways, we define it as a group that is both persistently detached from the legitimate work force and socially isolated. Persistence must include intergenerational persistence as well as a long-term condition for the individual. We find that only a small minority of white mother-only families fits any of the criteria for an underclass. The picture for blacks is somewhat cloudier. A substantial proportion of black single mothers exhibit persistent nonattachment to the labor force as measured by long-term welfare dependence, and a substantial proportion of their daughters become single mothers and dependent on welfare. On the other hand, only a small proportion of the black children living in

104 Sara McLan~an, Irunn Garfinicl, and Dorothy Watson mother-only families are socially isolated in severely poor urban black neighborhoods. Furthermore, although this proportion increased during the 1970s, it is ironic that the deteriorating condition of poor black neighborhoods resulted from economic and social gains made by the black population in general. On balance, we conclude that the growth of mother-only families is not associated with the growth of an underclass among whites, but it may be among some blacks. The extent to which the growth of a black underclass threatens to undermine the more general progress of blacks in America is worthy of further research. Government has always played some role in reducing the poverty and economic insecurity of mother-only families. As a consequence, policy-makers have continually faced and sought to resolve the dilem- ma over whether to give priority to reducing the poverty of single mothers or to reducing their prevalence and dependence. In the last section of the paper, we examine three recent developments in American federal income transfer policy and their effect on the economic well-being, self-reliance, and prevalence of mother-only families. The three policies are: (1) the reduction in the real value of transfer benefits, (2) the increase in the public enforcement of private child-support obligations, and (3) the increase in the public enforcement of work requirements for welfare recipients. We find that the falling level of benefits for single mothers brought about by a combination of inflation and budget cuts has had a substantial impact on the extent of their welfare dependence and only a trivial impact on prevalence. The work requirement leg- isIation has the potential to increase the earnings of single mothers if training is provided and jobs are available or guaranteed. It has not yet been implemented on a large scale, however, and questions re- main as to whether its potential will ever be realized. Child-support legislation, as currently enforced, will probably not have a big im- pact on poverty, welfare dependence, or prevalence. It does have the potential to substantially reduce poverty axed dependence if higher awards are secured and enforced in many more cases. PREVALENCE, GROWTH, AND CAUSES OF MOTHER-ON[Y FAMILIES In 1983 there were over 5.7 million families headed by single mothers in the United States, representing about 19 percent of all fannies (Bureau of the Census, 19843. Among whites, these families

FAMILY STRUCTURE, POVERTY, AND THE UNDERCLASS 105 accounted for 14.3 percent of all families; among blacks, for about 48 percent. Altogether, mother-only families were fairly evenly dis- tributed across residential areas: about 42 percent lived in central cities, 32 percent lived in surrounding suburbs, and the remaining 27 percent lived in nonmetropolitan areas. The concentration among blacks was much greater, however, with about 64 percent living in central cities compared with 18 percent each in suburban and non- metropolitan areas. For blacks, then, the experiences (and problems) of mother-only families are closely related to the experiences of urban life. These figures on prevalence are based on cross-sectional data and offer only a snapshot of the proportion of families headed by single mothers at one point in time. This view understates the proportion of women and children who will ever live in a mother-only family because it misses all families in which the mother has remarried (or the children have grown and gone) and all families in which a marital disruption (or premarital birth) has not yet occurred. Demographers estimate that about 42 percent of the white children and about 84 percent of the black children born in the late 1970s will live for some time with a single mother before they reach the age of 18. The median duration in a mother-only family is 6 years for children of formerly married mothers and even longer for children born to never-married mothers (Bumpass, 1984; Hofferth, 1985~. Lends in the proportion of families headed by single women are depicted in Figure 1 for 194~1983. Trends for blacks and whites are quite similar, although single motherhood has always been more common among blacks. For whites, the proportion of mother-only families grew 37 percent during the 1960s and 40 percent during the 1970s; for blacks, the proportions were 37 percent and 35 percent, respectively. Figure 1 also presents the trends since 1960 for black and white families living in central cities. The upper lines for each group show that the trend in central cities tends to parallel that of the entire group, although the absolute level is higher in the former. During the 1970s, however, the trend lines for both races appear to rise faster in central cities than in the general population. Numerous explanations have been put forward to account for the growth of families headed by single mothers during the past few decades, and there is a vast literature of empirical studies that at- tempt to test many of these arguments. In the following sections, we briefly review and evaluate three major explanations for the in- creases in single motherhood: (1) increases in welfare benefits, (2)

106 Sara McLanahan, Irwin Garfinicl, arid Dorothy Watson 60 55 50 45 40 25 20 10 5 o - White - Black Central City, White Central City, Black _— ~ f . ~ ~~ , I_; : - :/./ .~,r~ . 1 1 1 1 J - 1940 1950 1960 1970 1980 1983 Year FIGURE 1 The growth of mother-only families, 194~1983. SOURCE: Gar- finkel and McLanahan (1986~. increases in women's employment opportunities and marital conflict, and (3) declines in men's employment opportunities, especially those of young black men. Creases In Welfare Benefits Both common sense and economic theory suggest that raising public benefits to single mothers and their children will increase the number of mother-only families. Higher benefits increase the finan- cial ability of single mothers to establish their own households and thereby to become household heads. They enable a single mother to choose to keep her baby or have the baby adopted rather than have an abortion. Higher benefits also increase the ability of poor married mothers to choose divorce rather than remain in an unde- sirable relationship. In short, increases in benefits should increase the prevalence of single motherhood, all else being equal. Neither economic theory nor common sense, however, tells us how big any of these effects will be. The relationship between welfare and single motherhood has

107 been examined in numerous studies, including time-series analy- ses, longitudinal studies, and cross-sectional comparisons. Some researchers have compared welfare benefits across states with the proportion or "stock" of families headed by women; others have compared benefit leveb with the "flows in and out" of single mother- hood—for example, marital disruption and remarriage rates, illegiti- macy rates, and the propensity to establish independent households. Not surprisingly, studies that examine the correlation between welfare benefits and the stock of mother-only families are more likely to find effects than are studies that examine the effects of benefit levels on particular flows in and out of single motherhood. Studies of stocks conducted by Honig (1973), by Ross and Sawhill (1975), and most recently by Danziger and his colleagues (1982) all find an association of benefit levels with the number of female-headed households. Studies of flows, on the other hand, suggest that the as- sociation is due primarily to ejects on living arrangements and rates of remarriage (Cherlin, 1976; Hoffman and Holmes, 1976; Hutchens, 1969; Moore and Waite, 1976~. In response to the Honig study, Cut~rright and Madras (1976) demonstrated that benefit levels were associated with the proportion of single mothers who head their own households but not with the percentage of women who were divorced or separated. A more recent study by Ellwood and Bane (1984) confirmed these findings. After examining the effect of benefit levels on living arrangements, marital breakup, and premarital births, they con- cluded that the major consequence of welfare is that it allows single mothers to establish independent households. Furthermore, Ellwood and Bane also found that benefit levels are related to the proportion of divorced and separated mothers In the population but not to di- vorce rates. These results are consistent with an earlier finding by Hutchens (1969) and suggest that welfare affects the "flow outs of female headship (remarriage) but not the "flow ins (divorce). The empirical studies can be used to estimate the effects of increases in welfare benefits on the prevalence of single mothers. Be- cause some of the studies find no effect, a lower-bound estimate would be that the increase ~ benefits had no effect on prevalence between 1960 and 1975. If we use the highest estimate in the literature- Honig's estimate for blacks in 196~we estimate that the 196~1975 increase in welfare led to a 42 percent increase in single motherhood. In our judgment, however, the studies by Ellwood ~d Bane (1984) and by Danziger and his colleagues (1982) provide the most FAMILY STRUCTURE, POVERTY, AND THE UNDERCLASS . ,

108 Sara McLanahan, Irwin Garfinicl, and Dorothy Watson. reliable sources from which to estimate the effect of increased govern- ment benefits on the formation of mother-only families: the Ellwood and Bane study because it is comprehensive and distinguishes be- tween effects on prevalence and eEects on living arrangements, and the Danziger study because it models the effects of alternative oppor- tunities. Using these studies, we estimate that the growth in benefits increased the prevalence of single motherhood by between 9 percent and 14 percent from 1960 to 1975. In view of the fact that the prevalence increased approximately 100 percent during this period, increases in welfare benefits account for no more than one-seventh of the overall growth. In short, although increased benefits may have led to a measurable increase in prevalence, they account for only a small portion of the total growth in mother-only families.) That the increase in government benefits played only a small role in the overall growth in families headed by single women does not mean that the effects of benefits on single motherhood should be ignored. It seems reasonable to assume that welfare benefits played little or no role in the marital decisions of women in the top half of the income distribution. If so, welfare must have played a bigger role in the decisions of those in the lower half. Thus, if the growth in benefits accounted for 15 percent of the total growth in single motherhood, it could possibly account for 30 percent of the growth within the bottom part of the income distribution. Moreover, as documented later in this paper, women who have grown up in mother-only families are more likely to become single parents themselves, illustrating how the effects can mushroom over time. Finally, the effects of increased welfare benefits on living arrangements are a cause for concern because there is some evidence that children in families with ~ The much publicized results from the Seattle-Denver income maintenance experiment (SIME/DIME) have been interpreted to show that the effect of welfare benefits on divorce is much greater than the foregoing summary indicates (Groeneveld et al., 1983~. The SIME/DIME results, however, say nothing about the effects of raising or lowering the welfare benefits available to single mothers. The experiment was implemented in a world that already had a welfare system, and families in both the experimental and control groups retained whatever eligibility they would have had in the absence of the experiment. Many single mothers in the control group and some in the experimental group received welfare. Consequently, whatever effect the experiment had on behavior, it cannot be attributed to the availability of additional income to women who became single heads of households. If divorce rates were higher in the experimental groups, this was due to something about the treatment other than an "independence" effect.

FAMILY STRUCTURE, POVERTY, AND THE UNDERCLASS 109 other adults do better than children in families in which the mother is the only adult. Changes in Women's Employment Many people believe that the growth of mother-only families is due to greater participation in the labor force by women and, in particular, by married women with children. Some point to an ~inde- pendence effects that arises from increases in women's employment opportunities; others emphasize the wrote conflict" that accompanies the renegotiation of the traditional roles of husband and wife. Clearly, employment provides women with an alternative means of gaining financial security and thus competes with marriage and economic dependence on husbands. It also competes with traditional ideas about husband/wife roles by reducing the amount of time available for women to spend on housework and child care. The body of published empirical research in this area is nearly as large as the literature on welfare. It is also based on a variety of approaches, including analyses of time series, aggregate-level data, and survey data. For example, Preston and Richards (1975) examined the 100 largest standard metropolitan statistical areas (SMSAs) in the United States in 1960 and found that job opportunities, women's earnings, and unemployment were all good predictors of the marital status of women in the population. These researchers concluded that changes in job opportunities for men and women between 1960 and 1970 could account for about half of the decline in marriage during this period, or about half of the increase in single women. In her replication of the Preston and Richards study, however, White (1981) did not find a similar relationship for blacks. Another way to look at the question is to follow married women over time to see if working mothers are more likely to divorce and less likely to remarry than nonworking mothers. Several studies based on data from the Michigan Pane} Study of Income Dynamics (PSID) and the National Longitudinal Survey of Labor Market Experience (NES) have found that married women who work or who have higher earnings potential are more likely to divorce than more dependent women. Ross and Sawhill (1975) found that, controlling for hus- band^'s income and other factors, an increase of $1,000 in a wife's earnings was associated with a 7 percent increase in separation rates. Similarly, Cherlin (1976) found that a higher ratio of wife's earn- ings capacity to husband's earnings was a strong predictor of marital

110 Sara McLanahan, Irwin Garfinicl, and Dorothy Wanton disruption. Taken together, these studies indicate that the increase in economic opportunities for women can account for a substantial part of the increase in single motherhood among whites. For black women, the change in employment is much smaller, and the overall effect appears to be much weaker. Changes in Men's Employment The most widely discussed hypothesis concerning male employ- ment comes from Moynihan (1965), who argued in the early 1960s that unemployment among black men was causing a breakdown of the black family. Moynihan's graphs for male unemployment rates and single motherhood rates showed a close relationship throughout the 1950s and early 1960s. During the late 1960s, however, the trends diverged. Extending Moynihan's time series into the 1970s, we find that both unemployment and single motherhood continue to rise, but overall the relationship is not as close as during the 1950s. In a time-series analysis (using lagged variables) for the the post-WorId War IT period, South (1985) found a statistically significant relation- ship between unemployment rates and divorce rates. He also found a positive and statistically significant effect of women's employment on divorce. Additional evidence for an effect of male unemployment on single motherhood comes from microleve! analyses of longitudinal and cross-sectional surveys. Using data from the PSID, several re- searchers found that the probability of marital disruption is greater for families in which the husband has been unemployed (Hoffman and Holmes, 1976; Ross and Sawhill, 1975~. Cherlin (1976) and Moore and Waite (1976), in separate studies based on the NES, found that the husband's working less than full-time as well as earning low wages increased the probability of marital disruption. A problem with these studies is that a third factor such as alcoholism may be leading to both unemployment (or low wages) and divorce. Presumably, however, there is less chance of such an omitted variable being correlated with aggregate variations in unemployment rates across cities. Again, numerous aggregate-level studies have found a relationship between high unemployment rates and low wages on the one hand and high single motherhood and divorce rates on the other hand (Honig, 1973; Minarik and Gol~farb, 1976; Ross and Sawhill, 1975). The most recent version of the male employment argument has

FAMILY STRUCTURE, POVERTY, AND THE UNDERCLASS 111 been proposed by William Julius Wilson and his colleagues (Aponte et al., 1985; Darity and Myers, 1983; Wilson, 1985b; Wilson and Aponte, 1985; Wilson and Neckerman, 1986~. Like Moynihan, these researchers focus on black fannies and attribute the recent growth of mother-only families to increases in joblessness among black men. Their indicator, the "mate marriageable pool index," is the ratio of employed men per 100 women of similar age in the population. It is somewhat broader than indicators used by previous researchers because it takes into account not only unemployment but also par- ticipation in the labor force and sex differences in mortality and incarceration rates (Wilson and Neckerman, 1986~. Wilson points out that declines in the pool of marriageable black men between 1960 and 1980 were greatest in the North Central and Northeast regions of the country. These regions also showed the greatest growth in mother-only families. Wilson and his colleagues note that declines in employment among blacks were due initially to a shift in unskilled jobs from the South to the North and later to a loss of jobs in central cities in the North where blacks are highly concentrated.2 For example, during the 1970s, the number of un- skilled jobs declined by more than 30 percent in some cities (e.g., New York, Philadelphia, and Baltimore). Although the loss of Tow-skilled jobs in these areas was offset somewhat by an increase in higher- paying jobs, this shift worked to the disadvantage of black males, who are less likely to have a high-school degree (Kasarda, 19853. Given that the increase in single motherhood hap been especially pronounced among black women who have low levels of education- women whom we would have expected to marry men in low-skilled jobs the researchers conclude that the loss of jobs in the central cities is a major factor in the growth of mother-only families. THE FEMINIZATION OF POVERTY One of the most serious problems facing mother-only families is 2 The West, which accounts for only 9 percent of the total black population, did not fit the pattern. The marriageable pool of men in the West remained fairly constant while the number of female-headed families increased substantially. Wilson and Neckerman (1986) attribute this anomaly to the fact that black female heads of families in the West are more likely than are blacks in other parts of the country to be middle class and to behave more like whites. Thus, they should be expected to be more like whites in that they will respond to increases in opportunities for women rather than to declines in opportunities for men.

112 Sara McLanahan, Irwin Garfinicl, and Dorothy Watson 60 50 o o _ 30 o UP 0— 20 10 o Mother-Only Familles -ad Disabled _ —._ Aged Two-Parent Familles I I I I I I I I I 1 1 1 1 ~ 67 69 71 73 75 Year 77 79 81 83 FIGURE 2 Trends in poverty rates for mother-only families, two-parent fami- lies, persons over 65, and disabled persons, 1967-1983. SOURCE: Ross (1984~. poverty. Although not all of these fannies are poor, they face a much higher risk of poverty than other demographic groups. Roughly one out of two single mothers is poor, according to the official government definition of poverty. Figure 2 shows trends in the prevalence of poverty for mother-only families, two-parent families, aged persons, and disabled persons for 1967-1983. The figures include income from cash transfer programs such as Aid to Families with Dependent Children (AFDC), Social Security, and Disability Insurance (Ross, 1984~. Women and children in mother-only fannies are the poorest of all these groups, and the gap has been widening. In relation to the elderly and the disabled, their economic position has declined steadily cluring the past two decades. This does not mean that their absolute income has deteriorated, however. In fact, the poverty rate of those living in mother-only families actually declined until the late 1970s, only to rise again after 1978. If the econorruc status of mother-only families has not declined, why do we observe what some have called the "feminization of poverty"? The concept was introduced in 1978 by Diana Pearce (1978) and refers to the period between 1967 and 1978 when the pro- portion of the poor living in mother-only households was increasing. In 1967, 21.4 percent of the non-aged poor were living in households headed by single mothers, compared with 41.4 percent in two-parent households. By 1978, the pattern was reversed: 35 percent of the

FAMILY STRUCTURE, POVERTY, AND THE UNDERCLASS 113 poor were living in mother-only families, compared with 29.S percent in two-parent families. A trend such as the one described by Pearce can occur for three reasons: (1) the standard of living of mother-only families has declined, (2) the standard of living of other poor groups has improved, or (3) the proportion of persons living in mother-only families has increased. The feminization of poverty that occurred between 1967 and 1978 was due entirely to the second and third factors (Garfinke} and McLanahan, 1985~. The number of mother-only families grew dramatically during this period, as outlined above, and that growth increased the proportion of poor mother-only families relative to other poor subgroups. In addition, while the economic status of single mothers and their children more or less stood still during this period, the situation of other groups improved substantially. Poverty among the elderly dropped by about half between 1967 and 1974 in response to major increases in Social Security benefits. Poverty among the disabled declined temporarily between 1967 and 1975. Poverty among two- parent families also declined during the early part of the 1970s. About half of the feminization of poverty was due to increases in the number of single mothers, and half was due to the improvement in the living standards of other groups (Garfinke} and McLanahan, 1985~. Apart from poverty, mother-only families also face economic in- stability and loss of income as a result of marital disruption. Duncan and Hoffman (1985) found that the income of mothers and their children a year after divorce is only 67 percent of their predivorce income, but the income of divorced fathers stays about the same. Job change, change of residence, and unemployment are all more com- mon in mother-only families than they are in married-couple families (McLanahan, 1983~. A comparison of the sources of income available to different family types suggests three reasons why mother-only families are more likely to be poor: (1) lower earnings of the head of the family, (2) no child support from the second parent, and (3) meager public transfers. Low Earnings of Single Mothers The major source of income for all family types apart from those headed by widows is the earnings of the household head. Approxi-

114 Sara McLanahan, Irunr~ Garfinkel, and Dorothy Watson TABLE 1 Average Income Receipts (in dollars) of Two-Parent and Mother-Only Families, by Race, 1982 Whites Blacks Mother- Mother- Two-Parent Only Two-Parent Only Income Source Families Families Families Families Total cash income 30,814 12,628 23,913 9,128 Earned income of family head 21,932 7,666 13,508 5,363 Earnings of other family members 6,377 928 8,096 837 Alimony and child support 227 1,246 253 322 Social Security, pensions, and other unearned income 2,171 1,782 1,720 907 Public assistance and food stamps 174 1,399 1,838 2,573 SOURCE: Garfinkel and McLanahan (1986). mately 60 to 70 percent of a family's total income is accounted for by the household head's earnings, which suggests that the earning ability of single mothers is a critical determinant of their economic status. Table 1 shows the average income from different sources received by two-parent and mother-only families. Female breadwinners bring in only about a third as much as married fathers, partly because they work fewer hours and partly because they have lower hourly earnings. Much of the difference in poverty rates among different family types is due to the fact that single mothers work fewer hours than married fathers. The significance of not working is profound. David Ellwood (1985) has shown that only about 6 percent of single mothers who worked full-time, year round, during the previous decade were poor in any given year as compared with more than 70 percent of nonworking women. These findings should not be interpreted to mean, however, that if all single mothers worked full-time, only 6 percent of them would be poor. To some extent, the apparent advantage of working mothers reflects the selection process that channels women with higher earnings capacity into the labor force and women with Tower earnings capacity into homemaker status. On this point, SawhiT! (1976) found that most of the women on welfare in the early 1970s had a very Tow earnings capacity and that even if they worked full-time, more than half would still earn less than their welfare grant.

FAMILY STRUCTURE, POVERTY' AND THE UNDERCLASS 115 Although the Sawhill study has not been replicated with more recent data, there is good reason to believe that a large proportion of women on welfare would be unable to earn their way out of poverty or significantly improve their economic position even if they worked full-time, year round. For example, a woman working 2,000 hours a year at the minimum wage of $3.35 an hour would earn only $6,700 a year, which is less than the $7,050 poverty level for a family of two. To earn more than $8,850, the poverty level for a family of three, a woman working 2,000 hours a year would have to earn more than $4.40 an hour. The lower wages of women, then, may be just as important as their labor force participation rates in explaining the high incidence of poverty in mother-only families. The wage gap between women and men has not narrowed, and occupational segregation is still widespread despite increased partic- ipation by women in the labor force in recent decades. The median earnings of women who work full-time, year round, as a percentage of men's fell from 63.6 in 1955 to 60.2 in 1981 (Blau, 1984~. Using an index of segregation that represents the minimum proportion of one group that would have to be shifted in order for its occupational distribution to be identical with that of the other, Reskin and Hart- mann (1986) cite the index as .41 for segregation by sex for whites and .39 for sex segregation of nonwhites in 1981. A study by Bielby and Baron (1984) of nearly 400 California establishments suggests that the extent of segregation by institution may be even greater than the occupational segregation figures would suggest. There are two major competing hypotheses that seek to explain occupational segregation and earnings inequality: (1) human capital (i.e., education, training, job experience) and (2) discrimination by employers. The human capital hypothesis is that (a) for various reasons (specialization within households, differences in lifetime labor force participation, time preferences, labor force attachment, etc.), women and men differ in their decisions to invest in human capital; (b) this leads to sex differences in human capital stock and, hence, productivity; and (c) rational employers paying workers according to their marginal productivities will pay women less than men. In short, women earn less than men because they are less productive. The discrimination hypothesis is that women earn less than men because employers discrirn~nate against them.3 3 Becker's (1957) discrimination hypothesis was postulated to deal with race discrimination, but it can be adapted to deal with discrimination against

116 Sara McLa?~ahar~, Irmsr~ Garfrzlel, and Dorothy Watson Studies that use human capital variables such as education, train- ing, and work experience typically explain between 10 percent and 44 percent of the wage gap between women and men (Treiman and Hartmann, 1981~. Work history variables (e.g., years of full-time labor force experience and years of on-thejob trainings are the most important factors in accounting for the portion of the wage gap that is explained (Corcoran and Duncan, 1979~. Differences in human capital are clearly important in accounting for earnings differences between women and men.4 It is also significant, however, that they account for less than half of the wage gap. What seems to be missing in recent research on the earnings gap between women and men is a comprehensive model that takes women. His hypothesis is based on the idea that employers, coworkers, or cus- tomers have an externally acquired "taster for discrimination, so that employers will employ workers against whom there are discriminatory tastes only at a dis- count. In the discrimination model, occupational segregation is accounted for by variations in tastes for discrimination. In the case of occupational segregation by gender, as Blau (1984) notes, an extreme distribution of tastes would have to be assumed to account for the observed extent of segregation. Blau suggests that rather than employers, coworkers, or customers having a preference for male employees in all jobs, it may be the case that certain jobs are viewed as "socially appropriates for women. Employers would be willing to employ women in these jobs but would only employ women in Inappropriate jobs at a discount. England (1982) points out that the greater the proportion of women in an occupation, the lower the wage rate contrary to what the discrimination model would predict. If certain jobs were seen as inappropriate for women, one would expect them to be employed in such jobs only at a discount; thus, women in male-dominated jobs would be paid less than women in the (presumably more appropriate) female-dominated jobs. Bergmann's (1974) ~overcrowding" hypothesis gives a possible explanation for this discrepancy. She argues that occupational segregation will not automatically lead to pay differentials but will do so if supply exceeds demand in the female sector. The low wages will encourage the use of more labor-intensive technologies, which would result in female workers being less productive because they have less capital to work with. 4 Human capital theories have been less successful in explaining the ex- istence of widespread occupational segregation. For example, some research suggests that different occupations have different atrophy rates that is, they differ in the extent to which market skills are lost while workers are out of the labor force—and that women who expect their labor force participation to be intermittent will tend to choose occupations with low atrophy rates (Mincer and Polachek, 1974; Polachek, 1977~. The empirical evidence, however, shows no difference in atrophy rates between the earnings of women who are in male- dominated jobs and those in female-dominated jobs (England, 1982~.

FAMILY STRUCTURE, POVERTY, AND THE UNDERCLASS 117 account of the insights from both the human capital and discrim- ination approaches. Choices made by women and the recruitment preferences of firms are important, but the interaction between the two may well be the most important factor: the choices women make given their expectations of the behavior of firms, and the employ- ment practices of firms given their expectations of how women will behave. Inadequate Child Support The second reason for the greater poverty of mother-only families is that in most cases only one parent contributes to the faintly income. In two-parent households, according to Table 1, the earnings of white wives account for about one-third of family income. In mother-only farn~lies, child support payments from noncustodial fathers account for only one-tenth of family income for whites and less than one- twentieth for blacks. Income from child support is low partly because a large proportion of single mothers receive no child-support. The average amount received by single mothers who do receive child support payments is $3,129 for whites and $1,145 for blacks. These payments are much lower than the contribution of fathers in two- parent families and are also lower than the contribution of the other adults in two-parent families. Thus, even though a woman's earnings capacity is lower than a man's, and even though the contribution of the second parent is needed more in mother-only families than in two-parent families, absent fathers contribute a smaller proportion to child support in the former than women do in the latter. When a family splits, it loses the economies of scale that result from living together in one household. Two homes must be bought or rented and furnished, heated, and maintained. Even if all non- custodial fathers paid a reasonable amount of child support, such payments would not compensate fully for the economic contribution of the father who is present. Yet most noncustodial fathers do not pay reasonable amounts of child support. National data on child support awards indicate that only about 60 percent of the children poten- tially eligible for child support even have an award from the courts. Nearly 30 percent of those with awards receive nothing (Bureau of the Census, 1981a). To get an idea of what the poverty status of mother-only families (excluding widows) would be if noncustodial fathers paid a bigger

118 Sara McLanahan, Irwin Garfinbel, and Dorothy Watson share, Oellerich and Garfinke} (1985) simulated the effect of col- lecting 100 percent of the child-support obligation that the state of Wisconsin has set as a standard: 17 percent of the noncustodial parent's gross income for one child, 25 percent of income for two children, 29 percent for three children, 31 percent for four children, and 34 percent for five or more children. Their estimate indicated that the poverty gap the difference between the incomes of poor families headed by women and the amount of money they would need to move above the poverty level would be reduced by more than a quarter. Low Welfare Payments A final cause of poverty in mother-only families is the relatively meager public transfers these families receive. This effect can be seen in the contrast between the poverty status of widows and other single mothers. Fifty-one percent of all mother-only families are poor, com- pared with 34 percent of families headed by widows. This difference is largely due to the differences in benefits between Survivors Insur- ance, for which only widows are eligible, and Aid to Families with Dependent Children, for which all single mothers are eligible. White widows are far better oh than any other group of single mothers, not because they earn more but because they receive a large proportion of their income through Social Security. Welfare plays a much smaller role in the family income of the nonwidowed single mother than Survivors Insurance plays for the widowed mother. First, the proportion of all widows who receive Survivors Insurance is much higher than the proportion of other single mothers receiving welfare. Among widows, nearly 90 percent of whites and 70 percent of blacks receive Survivors Insurance. Only 22 percent and 33 percent, respectively, of white and black divorced women report receipt of welfare, and the proportion of separated and never-marriecI women who do so ranges from 38 percent to 58 percent. Second, the average level of benefits in Survivors Insurance is much higher than the average level of welfare benefits: for whites, it is more than double; for blacks, the difference is smaller but is still a sizable 20 percent more. Third, benefits for the child from Survivors Insurance are not reduced when the custodial parent works. By drastically reducing benefits as earnings increase, welfare programs replace rather than supplement earnings. Even when the AFDC program contained work incentives, as it did between 1967

FAMILY STRUCTURE, POVERTY, AND THE UNDERCLASS 119 and 1982, the gains from working were slight. The choice faced by poor single mothers is not an attractive one: become dependent on welfare, or work full-time and achieve, at best, a marginally better economic position. In addition, work is discouraged because it increases the risk of losing valuable in-kind benefits such as health care and public housing. In summary, although single motherhood is not synonymous with poverty, the risk of long-term poverty is substantially greater for mother-only families. Their growth between 1960 and 1978 ac- counted for only one-half of the observed feminization of poverty, the remainder being accounted for by the lack of improvement in the living standards of families headed by single women at a time when other groups experienced increasing incomes. Finally, the principal reasons for poverty in families headed by women are the low earnings capacity of the single mother, lack of adequate (or any) support from the noncustodial father, and relatively low public transfers. THE UNDERCLASS Poverty and income insecurity are serious problems in their own right. Some analysts, however, go even farther in arguing that the mother-only family is responsible for the growth of an "underclass" in American cities (Auletta, 1982~. To demonstrate such a relationship, one would have to show that mother-only families are more likely to be in the underclass than two-parent families, or that children who grow up in the former are more likely to be in the underclass than children who grow up in the latter. Defining the Underciase The term underclass has been used in a variety of contexts and defined in a variety of ways.5 Based on his review of the literature and discussions with poverty researchers, Auletta (1982) identifies 5 Many of the usages of the term underclass bear little relation to the concept of class in either the Weberian or the Marxist senses. Sociologists in the Weberian tradition see classes as groups of people with common economic "life-chances" based on their relative control over goods and skills in the marketplace (Giddens, 1973; Parkin, 1971; Weber, 1922~. The MaIxist concept of class is defined in terms of common structural positions within the social organization of production based on ownership or nonownership of factors of production (Bukharin, 1921; Lenin, 1914; Wright, 1985~. (See Wright l1979l for an overview of the different concepts of social class.) The Weberian concept of class tends to be gradational, emphasizing the relative positions of the classes

120 Sara McLanahar~, Irunn Garfinkel, and Dorothy Wat~orz the underclass as those who suffer from "behavioral as well as income deficiencies" and who "operate outside the mainstream of commonly accepted values." For him, this includes four distinct groups: hostile street criminals, hustlers, welfare mothers, and the chronically ill.6 Wilson and his colleagues speak of the underclass in the context of urban ghettos (Aponte et al., 1985; Wilson, 1985b; Wilson and Aponte, 1985; Wilson and Neckerman, 1986~. They point to changes that have taken place in ghetto communities, including deindustrial- ization and the exodus of middie-cIass blacks, and contend that the groups left behind are collectively different from those who lived in the communities in the 1940s and 1950s. Wilson argues that poverty rates in ghetto neighborhoods were high throughout the first half of the century but that rates of joblessness, teenage pregnancies, out-of-wedIock births, female-headed families, welfare dependence, and serious crime did not reach alarmingly high levels until the mid- 1970s. Although long-term welfare fannies and street criminals are distinct groups, the fact that they live and interact in the same com- munity and are increasingly isolated from the mainstream patterns of norms and behaviors is a cause for concern (Wilson, 1985b). In his analysis of urban problems, Wilson relies heavily on the work of Kasarda (1985), who shows that poor inner-city minorities have been especially vulnerable to the economic transformation oc- curring in central cities that is changing them from centers of produc- tion and distribution of physical goods to centers of administration, information exchange, trade, finance, and government services. Since (e.g., lower class, middle class). The Marxist concept of class tends to emphasize the relationships of classes to each other and the possibility of conflict between them. The definition of the underclass outlined in this paper is closer to the Marxist than to the Weberian concept of class because it emphasizes the nonownership of salable labor power by members of the underclass. A fuller analysis of the relationship of the underclass to the working class and to the capitalist class goes beyond the scope of this paper but would, at a minimum, involve a discussion of their role as a "reserve army of labor. 6Auletta's data come from case studies of participants in supported work programs run by the Manpower Demonstration Research Corporation in New York City and Appalachia. The eligibility requirements for the New York program included being a recently released ax-offender, a recent ax-addict, a long-term welfare-dependent mother, or an unemployed high-school dropout— categories markedly similar to Auletta's four main divisions of the underclass. Auletta's arguments and methods are very close to those of the earlier culture- of-poverty approach (Lewis, 1966) in his emphasis on the behavioral deficiencies of the poor and in his reliance on case studies.

FAMILY STRUCTURE, POVERTY, AND THE UNDERCLASS 121 1948, vast numbers of unskilled jobs in wholesale, retail, and manu- facturing in the nation's central cities have been lost, and this process has accelerated since the late 1960s. The new jobs in the growing service sector usually require levels of education and training that poor inner-city residents do not possess. The cumulative impact, according to Wilson, has been increasing ghettoization character- ized by poverty, institutional problems (services and schools of poor quality), and an increase in social dislocations (joblessness, crime, mother-only families, teenage pregnancies, and welfare dependence). Weak Attachment to the Labor Force What these seemingly diverse groups have in common is a weak attachment to the legitimate labor force.7 The underclass can be seen as made up of family units with no readily salable labor power or other factors of production. Labor power may be difficult to sell for a variety of reasons: disability, lack of human capital or salable skills, discrimination, or low demand. To the extent that members of the underclass do participate in the labor force, they tend to be in "secondary jobs that offer little stability of employment.8 Attachment to the labor force can be direct or indirect, and both kinds of attachment carry a "right" to income. Direct attachment involves the sale of one's own labor power with the corresponding right to a wage or salary income. Indirect attachment includes links to the labor force through former employment- in which the income is in the form of pensions or Social Security benefits and through 7Perhap~ a more precise definition would see the underclass as comprising families and individuals with a relatively weak attachment to the production process, because one would not wish to include owners of capital in the under- class. Given that few people have enough capital or wealth to make the sale of labor power unnecessary, however, the definition in terms of attachment to the labor force is adequate for most purposes. actual labor market theory suggests that a dichotomization of the Ameri- can labor market has occurred over time, forging two separate labor markets a 'iprimaryn market and a "second oryx market in which workers and employers operate by fundamentally different behavioral rules (Gordon, 1977~. The pri- mary labor market is characterized by high wages, good working conditions, employment stability and job security, equity and due process in the administra- tion of work rules, and chances for advancement. The secondary labor market has low wages, poor working conditions, unstable employment, harsh and often arbitrary discipline, and little opportunity for advancement (Piore, 1977~.

122 Sara McLar~ahar`, Irunn Garfinicl, arid Dorothy Warrior' family status in which a spouse or dependent shares in the income of another family member who is attached to the labor force.9 Given their lack of access to the usual means of gaining a liveli- hood, members of the underclass must find another source of income such as public transfers or the underground economy (e.g., crime). It is their common need to find an alternative to wage or salary income, and the social stigma that accompanies their position, that allows such diverse groups as poor female-headed families, criminals, the unemployed, and the mentally ill to be grouped together. Persistence and Social Isolation In addition to weak attachment to the labor force, the notion of an underclass implies a persistence or permanence in status. Thus, individuals who are temporarily unemployed, ill, or dependent on welfare are not viewed as part of the underclass, whereas long-term welfare recipients and those with unstable work histories are prime candidates. Persistence may also occur across generations, and much of the concern over female-headed families arises from the belief that children who grow up in such families are more likely to become single parents and welfare recipients when they grow up. The final characteristic of the underclass is the isolation of its members from a community activity. Isolation and persistence of nonemployment, be they in urban ghettos or rural areas of the South, are of concern because they are believed to encourage the develop- ment of a hostile or alienated subculture. We know, for example, that the longer an individual remains outside the labor force, the less likely he or she is to become employed. Many believe that this relationship is due to a change in values and a loss of motivation that goes along with being marginal to the labor force. In particular, there is concern over whether the children of welfare mothers are developing the motivation and the practical knowledge required to find and hold a steady job. Not all analysts agree about whether the underclass has a unique culture. In fact, since the mid-1960s, liberal scholars have tended to avoid describing any behavior on the part of the poor or minori- ties that might be construed as unflattering or stigmatizing. Wilson 9 Indirect links to the labor force typically bring a lower return than do direct links and tend to be seen as less legitimate. The indirect links that have traditionally received the blessing of social approval are those of the child through the parent and the wife through the husband.

FAMILY STRUCTURE, POVERTY, AND THE UNDERCLASS 123 (1985a) notes that the debate over the underclass was dominated by the conservatives until the recent resurgence of interest in the "un- derciass~ and welfare dependency, partly in response to the failure of equal opportunity legislation to eradicate poverty and partly because of the increasing concentration of the poor in central cities. In addition to the revival of research and scholarship oriented toward an understanding of the urban underclass, liberals recently have begun to discuss the consequences (rather than just the causes) of the distinct cultures that emerge in ghetto areas. Emphasizing the social structural constraints and opportunities that give rise to subcultures that are distinct from the mainstream, Wilson (1985a) nevertheless argues that the liberals of the 1960s did not give suffi- cient attention to the role that culture plays in influencing behavior. Culture, once it has emerged in response to structural conditions, can itself become a constraining or a liberating factor. Unlike the earlier culture-of-poverty theorists, Wilson places more stress on the lack of contact between ghetto residents and people with good jobs (leading to a lack of information about job openings and few role models for the young) than on the role played by distinct values in reproducing the ghetto underclass. Being poor and being in the underclass are not the same, how- ever, although the two statuses overlap a good deal. A family may be poor even though the parents (or parent) work full-time. Or, as noted earlier, a family may be poor because the parent is temporar- ily disabled or unemployed. In neither case would such a family be classified as being in the underclass. Similarly, not all members of the underclass are poor. In particular, some Hustlers" and street criminals may have incomes well above the poverty line. What dis- tinguishes the underclass from the poor is the persistence of their nonemployment and their isolation from the values and behavior of the broader community. Are Female-Headed Families Part of the Underclass? We now turn to the question of whether families headed by single mothers are part of the underclass by virtue of their own position or that of their offspring. To what extent do single mothers have a persistent, weak attachment to the labor force? To what- extent is this status passed on to their children? And to what extent are these women and their families socially isolated?

124 Sara McLanahan, Irwin GarfirUccl, and Dorothy Watson Persistent, Weak Attachment to the Labor Force Single mothers can be linked to the labor force in three ways: directly, indirectly, or tenuously. First, mothers who have relatively stable employment are directly linked to the labor force. Second, widows who receive Survivors Insurance and divorced or separated women who receive child support are indirectly linked to the labor force through the past or present participation of their former hus- bands, whether or not they themselves work. (This link to past employment, albeit indirect, may account for the generally higher benefits received by widows as compared with welfare benefits, and for the widespread perception that such income is theirs by right.) Third, women who neither work consistently nor have indirect links to the labor force through a spouse have the most tenuous right to income and are most likely to be viewed as part of the underclass. Because data on work history are less readily available than data on long-term welfare dependency, the latter can be used as an index of weak attachment to the labor force and persistent dependency. Cross-sectional data on welfare participation indicate that within the last decade, half of single mothers received some welfare in any year (Garfinke! and McLanahan, 1986~. Most of these mothers have a very high degree of dependence while they are on welfare. Eighty- five percent of those receiving benefits do not work, and most have no sources of income other than food stamps, Medicaid, and, in some cases, public housing. Furthermore, EDwood (1985) found that 65 percent of AFDC recipients at any point in time are in the midst of a welfare spell of 8 or more years. In short, about 30 percent of mother-only families can be classified as long-term dependents, which eliminates at least 70 percent of single mothers and their children from the underclass. Intergenerational Dependence To what extent are the children who grow up in families headed by single women likely to be part of the underclass? The literature on the consequences of faintly structure for children is quite large, and we do not attempt to cover all of it here. Rather, we have limited our discussion to outcomes that are directly related to labor force attachment: education and occupational attainment, family formation behavior, employment status, and welfare status. Low levels of education and low-status occupations are both associated with a higher risk of unemployment, and they help explain, as we

FAMILY STRUCTURE, POVERTY, AND THE UNDERCLASS 125 have seen, the weak ties to the labor force of female-headed families. To the extent that the offspring from mother-only families are more likely to leave school sooner, to get low-status jobs, or to have out- of-wedIock births, their risk of being in the underclass is increased.~° Studies of educational attainment based on cross-sectional data have consistently found that offspring from single-parent families complete fewer years of schooling. Otis Dudiey Duncan and his colleagues (1972) found that growing up in a married-couple family added between 0.6 and 1.2 years of schooling to the educational attainment of white males and about 0.4 to 0.8 years to that of blacks. Similar results were reported by Featherman and Hauser (1978) in their analysis of the more recent data from the Occupational Change in a Generation (OCG) study. Recent analyses based on the PSID and the NES data are consistent with earlier findings. McLanahan (1985) found that living in a female-headed faintly increased the probability of dropping out of high school by 42 percent for whites and 70 percent for blacks, and Shaw (1979) and Krein and Belier (1986) reported similar results using the NES data. Children from single-parent families are also disadvantaged with respect to occupational status. Duncan et al. (1972) found that off- spring from one-parent families had lower occupational status scores than those from married-couple families. The average score for white men from two-parent families was 45.12 compared with 40.28 for those from single-parent families. Among blacks, the scores were 21.8 and 17.93, respectively. Featherman and Hauser (1978) re- ported similar results and noted that differences in status were due both to differences in educational attainment and to differences in the returns to education. Perhaps the strongest evidence for intergenerational effects comes from research on family formation behavior and intergenerational use of welfare. Several researchers have shown that daughters who grow up in single-parent families are more likely to marry early and to have children early (including both marital and premarital births), both of which are positively related to becoming a single mother. These daughters are also more likely to divorce than children from two-parent fannies (Bumpass and Sweet, 1972; Hogan and Kita- gawa, 1985; McLanahan and Bumpass, 1986~. McLanahan (1987) moor major reviews of the literature on intergenerational consequences, see Herzog and Sudia (1973); Hetherington et al. (1983); Ross and Sawhill (1975); and Shinn (1978).

126 Sara McLanahan, Irwin Garfinkel, and Dorothy Cavort also found that daughters of single mothers were more likely than daughters from two-parent families to become single heads of families themselves and to receive welfare. Part of the reason for the disadvantages suffered by children from mother-only families is low family income. Differences in in- come account for much of the difference in educational attainment among children from one- and two-parent families (Hetherington et al., 1983; Krein and Belier, 1986; McI.anahan, 1985; Shaw, 1979~. Yet income explains very little of the reproduction of single moth- erhood and welfare dependence (McLanahan, 1987~. The residual influence probably comes from the absence of a second parent, which is likely to increase the perceived legitimacy of single parenthood and to reduce the amount of supervision of dating during adolescence, an important factor in accounting for teenage pregnancies (Hogan and Kitagawa, 1985; McLanahan and Bumpass, 1986~. This hypothesis is consistent with the finding that children who grow up in father-only families also appear to fare less well than children from two-parent families tHetherington et al., 1983; McLanahan, 1987; McLanahan and Bumpass, 1986~. A final explanation that must be considered is the possibility that differences between one- and two-parent families are due to selectiv- ity, that is, to preexisting differences between families that break up and those that do not. Because research on single parenthood is necessarily nonexperimental, we cannot rule out the possibility that children in these families would have had more problems even if their parents had not divorced. Selectivity effects could stem from predivorce differences in family socioeconomic status or the nature of intrafam~ly relationships (parent conflict or abuse, parent-child con- flict or abuse), or both. Support for the selection hypothesis comes from studies showing that children in unhappy, two-parent families are similar to children in single-parent families with respect to psy- chological adjustment and behavior (Emery, 1982; Peterson and Zill, 1986~. Conflicting evidence comes from studies showing that widow- hood also has negative consequences for offspring (Levy-Shiff, 1982~. Widowhood is a more random event with respect to preexisting fam- iTy interaction patterns even though it is more likely among poorer segments of the population. Ghettoization If families headed by mothers are to be classified as part of an emerging underclass, they would not only have to have a persistent

FAMILY STRUCTURE, POVERTY, AND THE UNDERCLASS 127 weak attachment to the labor force but would also have to be socially isolated. Although social isolation may be present in a rural area, recent discussions of the underclass have focused on urban areas, and the data we use to crudely measure isolation pertain only to urban areas. Our proxy for social isolation is living in an urban neighborhood that is predominantly poor. The census reports from which we derive these numbers give three classifications of poverty neighborhoods: those in which at least 20 percent, 30 percent, and 40 percent of their populations are poor. Although it seems reasonable to assume that the higher the percentage of poor people in a neighborhood, the greater the degree of cultural isolation, it is di~cult to know exactly where to draw the line. Hence, when possible, we use all three classifications. The top two rows in Table 2 present data on the proportion of families with children that are headed by single mothers and by others living in poverty areas of the 100 largest cities in the United States in 1979. The data are presented both as a proportion of families living in the central city and an a proportion of all families in the United States The bottom pane} presents separate figures for black and white families headed by a single mother. Several aspects of the table are worth noting. First, a comparison of the first and second rows clearly indicates that families headed by single mothers are much more likely than other families to live in poverty areas. The greater the proportion of the neighborhood that is poor, the bigger the difference in the proportion of mother-only and other families that live there. For neighborhoods that are at least 20 percent poor, the proportion of mother-only families is a bit more than twice as much as the proportion of other families. For neighborhoods that are at least 40 percent poor, the proportion of mother-only families is nearly five times as high. In view of the fact that families headed by single mothers are more likely to be poor, this is not surprising. Second, the proportion of all mother-only families that live in poverty areas in central cities is very sensitive to the definition of a poverty area. If a poor population of 20 percent is the cutoff, we find that 56 percent of all mother-only families in central cities reside in poverty areas. If 40 percent is the cutoff, the figure drops to only 19 percent. Third, if the denominator is the proportion of all U.S: families headed by single mothers rather than the proportion of such families

128 o oo - oo a ._ ._ v v oo s" C: o o a, ._ ~n s" ¢ :^ s" a, o P~ ._ bO ._ ._ oo ._ ._ C: o o ._ o o P~ e~ m ¢ E~ D] V O ~ O S" O U3 a, V P~ ~ O - 0D " V O ~ O 03 ~ ~V a~ a~ V ~ ~ O V O O V V P~ ~ O V, ~ ~ 00 C<: ~ ~ V V ~ ~o E~ ·- £ — =~ o~ C. ~ ~ `-E ~ .~ Ce `4 VV~ o~ . — ~ ~ U] =~ o~ ' a, _ ~ ·,. C~ ~ Om ~ ~ ·~ ~ ~ :^ V V ~ o ~ :^ ._ z ¢ ~ ~ z ~ ¢ ~ z z cs e~ o; . . ~ ¢ z z ~ u: CD u, CO u: e~ c~ a~ . - ~= . - ~ ~ ~ 3 c: a) oo :^ ~ .O _ ._ _ o ~ ~ Q, ~ s.4 ~ , a~ i~ o ~4 2 o o ~ o s" ~ ·— ~V V ,S: ~ , ~ 3 ~ 3 U] ~ ~ ~ ~ ._ ~, o ._ ~ _ ~= E-_ ~ V V ~ V C. m C~ _ ~Q o .— o aS o ,, ~ 11 O - ~ ~. o o CO o ._ ~ 3 ~ oo ~ CVC ~, o _ a, ~ . — :~ V ~ ·— oo ~ o o ,= o ~ _ v v L. ~ ~ ._ ~ o o C~ ~ o ._ o 3 ~ ~n V ~ s~ ~, a, 0 ~ o U3 a' ~ _ ~ .~ .. 3 v ~ ~q E~ o ~ o z ~ U) - Q U: oo C~ ~4 U: oo ~ _, ~n oo V o s" m

FAMILY STRUaTURE, POVERTY, AND THE UNDERCLASS 129 living in central cities, the proportions are much smaller ranging from 19 percent to 7 percent. The proportions are smaller because a large percentage of mother-only families do not live in central cities. Thus, even if one accepts the premise that social isolation is serious in neighorhoods that are 20 percent poor, only 19 percent of all families headed by single mothers would fit the classification of belonging to an emerging urban underclass. The last two rows of Table 2 indicate that the proportions differ dramatically by race. The breakdown by race was available only for poverty areas in which 20 percent or more of the families were poor. For whites, only 26 percent of mother-only families in central cities live in poverty neighborhoods; for blacks, the figure is 75 percent. Black families headed by single mothers thus tend to be disproportionately concentrated in poverty areas. On the other hand, black mother-only families who live in these neighborhoods constitute only about 41 percent of all black femaTe- headed families. Furthermore, what the table does not show is that the children in these families represent only 17 percent of all black children. Finally, if the most stringent measure of social isolation is employed that is, 40 percent of the population is poor a crude estimate suggests that only about 15 percent of black children in mother-only families and about 7 percent of all black children are socially isolated in urban ghettos. To what extent did black mother-only families become more so- cially isolated during the 1970s? The data in Table 3 suggest that black mother-only families who reside in poor neighborhoods have become more socially isolated; in addition, the proportion of such families living in all poor neighborhoods has declined somewhat, but the proportion living in the poorest neighborhoods has increased. The two indicators of social isolation are the proportion of families dependent on public assistance in the neighborhood and the propor- tion of men employed. In both the 20 percent and 40 percent poverty areas, the proportion of families dependent on public assistance grew by 40 percent between 1970 and 1980. The percentage of men em- ployed declined by 13 percent and 22 percent, respectively, in the 20 percent and 40 percent neighborhoods. By 1980, less than half of all iiThis estimate is derived by multiplying the 41 percent and 17 percent figures for the 20 percent or more neighborhoods by 7/19 the ratio of the proportion of female-headed families living in 40 percent or more poverty neigh- borhoods to the proportion living in 20 percent or more poverty neighborhoods.

130 Sara McLanahan, Irunn Garfr~kel, and Dorothy Watson TABLE 3 Trends in Social Conditions and Proportion of Black Persons and Families Living in Poverty Areas in Large Central Cities, 1970-1980a . . . Census Tracts with 20 Percent Poor Census Tracts with 40 Percent Poor Percentage Percentage Percentage Percentage 1970 1980 Change 1970 1980 Change Poverty Area Characteristics Employment rate, men aged 16+ 63.3 56.0 -13 56.5 46.0 -22 AFDC recipiency, all families 19.8 28.0 40 30.2 42.0 40 Families and Persons in U.S. Living in Poverty Areas Black families 27.2 25.2 -8 Poor black families 28.6 30.4 6 Black mother-only families 30.5 29.7 -3 Poor black mother-only families 30.7 31.4 2 Black persons 27.2 26.5 -3 6.3 8.3 32 Poor black persons 28.3 30.5 8 9.4 13.1 40 NOTE: Census tracts with 20 percent poor include tracts with 30 and 40 percent poor, and tracts with 30 percent poor also include all those tracts with 40 percent poor. AFDC = Aid to Families with Dependent Children. aThe data are from the cities which were among the 50 largest cities in the United States in 1970. SOURCE: Bureau of the Census (1973, 1985a, and 1985b). men in 40 percent neighborhoods were employed, and nearly half of all families were dependent on welfare. The proportion of both mother-only and all families in the 20 percent poverty neighborhoods (the only areas for which data on black families with children are available) declined somewhat between 1970 and 1980, whereas the proportion of poor black families and poor black mothers-only families increased. Finally, to see if the picture changes when the 40 percent rather than the 20 percent cutoff is used, we present data for all black per- sons for both cutoffs. Whereas the proportion of all black persons in neighborhoods that are 20 percent or more poor decreased by 3 per- cent between 1970 and 1980, the proportion in neighborhoods that are 40 percent or more poor increased by 32 percent. This increase suggests that the proportion of mother-only families living in 40 per- cent areas also increased. Such findings appear to be consistent with

FAMILY STRUCTURE, POVERTY, AND THE UNDERCLASS 131 WiTson's hypothesis that a number of relatively poor neighborhoods have deteriorated as the nonpoor move out; these areas have now become very poor neighborhoods (Wilson and Aponte, 1985~. Thus, both the number of very poor neighborhoods and the proportion of blacks living in such neighborhoods have increased, even though the proportion of blacks in all poor neighborhoods has declined slightly. RECENT PUBLIC POLICIES REGARDING FAMILIES HEADED BY SINGLE WOMEN Children who grow up in families headed by single women are clearly disadvantaged. As adults, they have lower socioeconomic status; they are more likely to become single parents themselves, either through out-of-wedIock births or divorce; and they are more likely to be dependent on government. Many of these problems arise from the economic insecurity and poverty of the families in which they grew up. Government can reduce these problems by increasing the incomes of mother-only families. But doing so may increase both the numbers of such families and their dependence on government. Apart from a brief flirtation with social Darwinism toward the end of the nineteenth century, there has always been a general ac- ceptance in this country of government responsibility for aiding the poor. Accompanying this historical constant has been a concern to avoid encouraging dependence on welfare and growth in the number of the poor. Policy makers today must decide which should receive greater priority: reducing the economic insecurity and poverty of families headed by women or not encouraging the growth of such families and their dependence on government. This dilemma has led to the development of distinctions among groups of the poor, on the basis of which differential aid has been provided. The most important distinction has been between those who are and those who are not expected to work. This distinction has implications for policy toward mother-only families in that ex- pectations about whether single mothers should work have recently changed. Widows—especially war widows have always received better treatment than other single mothers. In the eighteenth century, most single mothers were widows. Divorce, separation, and out-of- wediock births were strongly discouraged by law and custom and were uncommon by today's standards. Widows were treated more favorably for two reasons: first, widowhood was an involuntary state

132 Sara McLar~ahar', Irunr' Garfinkel, arid Dorothy Wats ore and aiding widows was unlikely to increase their numbers; second, widows were commonly seen as more "deserving" because of their husband's service to the community, as was the case with war widows for example. The three most important recent trends in policies that affect families headed by women are (1) the large reduction in public bene- fits, (2) the increasingly strong legislation either to induce or require single mothers without preschool-age children to work, and (3) the strengthening of public enforcement of private child-support obli- gations. Each of these policies is designed to strengthen the links between mother-only families and the labor force either directly, by encouraging or requiring the single mother to work, or indirectly, by increasing support from the noncustodial parent. The nature of each of these trends and their effects on the poverty, welfare depen- dence, and prevalence of families headed by single women is discussed in turn. Reduction in Benefits Throughout the 1970s, the real value of AFDC benefits declined because state legislatures failed to increase benefit levels to keep pace with inflation. Between 1975 and 1980, inflation cut the benefits received by mother-only families by about 13 percent. Apart from these passive benefit cuts, the Reagan administration actively sought to cut benefits. Reagan proposed a series of specific budget cuts that, when taken together, would have added up to large reductions in benefits to families headed by single women. Congress eventually adopted more modest cuts that amounted to about 12 percent of total federal benefits to mother-only families (Garfinke] and McLanahan, 1986~. The reductions in benefits to families headed by single women between 1975 and 1985 were substantial, wiping out more than one- third of the increases that had occurred during the previous two decades. Single mothers who both worked and received welfare ab- sorbed two-thirds of the cuts. The administration abandoned the strategy of extending welfare eligibility to those not previously cov- ered to reward behaviors such as working and marrying, in the hope that such incentives would reduce dependency in the long run. In retrospect, perhaps the reductions in the real value of benefits are not surprising. Over the long run in the United States, the real wage level has been the principal determinant of the average level of

FAMILY STRUCTURE, POVERTY, AND THE UNDER GLASS 133 living and of benefits to the poor (Garfinkel and McLanahan, 1986~. Real wages in the United States fell during most of the 1970s and early 1980s. Although short-run changes in general living standards do not always lead to corresponding changes in benefit levels, that is what happened in 1975-1985. The decline in the real value of public benefits reflected but also exceeded the decline in real wages. By the end of the Reagan administration's first term, Congress was no longer enacting legislation that even modestly reduced ben- efits; indeed, some of the earlier cuts to mother-only families were restored. Moreover, early in Reagan's second term, further cuts in the major programs that aided single mothers were specifically ex- cluded from the stringent Gramm-RuUman-Hollings budget-cutting measures designed to reduce the large federal deficit. In view of both the recent growth in real wages and the recent resistance of Congress to enact further budget cuts, families headed by women will proba- bly not be subjected to additional decreases in benefits in the near future. The evidence suggests that the budget cuts increased the poverty of mother-only families by nontrivial amounts in return for small-to- trivial reductions in dependence and prevalence. Almost by defini- tion, the elimination by the Reagan administration of eligibility for AFDC for many working mothers decreased their dependence on welfare. Some analysts have argued that these cuts may increase welfare dependence in the long run by encouraging those whose ben- efits were cut to leave work and return to AFDC and by discouraging nonworking mothers from working their way off the welfare rolls. Studies of the behavior of working mothers after the budget cuts, however, indicate that only a minority of recipients who were taken off the rolls returned to AFDC (Cole et al., 1983; Joe et al., 1984; Usher and Griffith, 19833. The decrease in benefits during 1975-1985 was accompanied by a noticeable decrease—more than one-sixth- in the dependence of mother-only families on welfare. This decline re- versed the increase in welfare dependence that had accompanied the expansion of government benefits during the previous two decades. Conversely, just as the large benefit increases between 1955 and 1975 caused, at most, a modest increase in the proportion of mother- only fannies during that period, the smaller reductions in benefits during 1975-1985 had little if any eject on the prevalence of these families. If one can assume that a cut in average benefit levels has the same effect as an equal-sized cut that is targeted primarily at those who work, estimates derived from the two most careful and

134 Sara McLanahar~, Irunr~ Garfir~kel, arid Dorothy Watson comprehensive studies suggest that the 12 percent cut in welfare benefits would have reduced prevalence by between 0.9 percent and 1.6 percent. Further extreme cuts in benefits would have at most a modest effect on the proportion of families headed by single women. Work Requirements Although able-bodied men have always been expected to work, expectations regarding poor women who head families have changed considerably. Until the early twentieth century, single mothers were expected to work. For the next half-century, the stated objective of government policy was to provide enough aid to enable them to imitate what was then the current middle-cIass ideal of the mother who refrains from market work and stays at home to Took after her children. Not until the late 1960s, however, was sufficient aid provided to make the objective achievable, and by that time, both the middIe-cIass ideal and beliefs about whether poor women who head families should work had changed. There was increasing emphasis on the idea that single mothers should work. At first, the federal government tried to induce AFDC mothers to work by creating work incentives within AFDC. When this failed to have much impact on either work or caseloads, Congress began legislating requirements for mothers with no children under age 6 to work. The Carter administration proposed a combination of a guaranteed jobs program and assistance that would have, in effect, required mothers without preschool-age children to work. The Rea- gan administration rejected the approach of creating work incentives within the AFDC program in favor of the pure work requirement. The administration sought to reinforce working by limiting benefits to women with no income from work and by requiring those who received benefits to work for them. Congress agreed to much but not all of this strategy. t2The work incentive provision under AFDC that ignored the first $30 of earnings plus $1 out of every $3 in excess of this amount each month was limited to 4 months. Families with incomes of more than 150 percent of the ~tate's need standard (the level of income for determining initial eligibility) were made ineligible for benefits. Individually determined work-related expense deductions with no upper limits were replaced by a hat rate of $50 per month for part-time work and $75 per month for full-time work. These restrictions were liberalized somewhat in 1984, and some of the benefit cuts were restored. The $30 set-aside was extended from 4 to 12 months, the $75 per month deduction for expenses

FAMILY STRUCTURE, POVERTY, AND THE UNDER CLASS 135 Since 1972, AFDC mothers without children under age 6 have been required to register for work and training, but only a minority of those who registered ever received services to help them find jobs. The Reagan administration proposed that states require AFDC mothers to work for their relief checks in community work experience programs (CWEPs). Congress passed legislation that permitted but did not require states to substitute work relief for cash relief. Under workfare, as it is called, participants work off their relief checks. They are paid at the minimum wage, and the hours they must work are limited to the check amount divided by the minimum wage. One anomaly of this procedure for calculating benefits, however, is that mothers with a greater number of children receive higher welfare benefits and thus will be required to work more hours. Congress also passed several other provisions designed to encour- age work. States may now require AFDC recipients to participate in a program of job searches for up to 8 weeks upon application and for an additional 8 weeks for each year benefits are received. States are also permitted to operate work supplementation, or grant diver- sion, programs in which federal funds that would ordinarily finance AFDC cash benefits may be used to subsidize a job for an AFDC recipient. With this program, unlike workfare, a participant's total income exceeds what would have been received on welfare alone. As of September 1985, 37 states had implemented one or more of these options, with workfare programs being the most common. Most pro- grams are being run on a demonstration basis, however, arid in the few states that have statewide programs, only a small proportion of the caseload is being served. To enforce work requirements, the government must create or locate jobs. Some analysts have argued that it is not possible to find or create enough jobs to enforce work requirements when the unem- ployment rate is over 7 percent (General Accounting Office, 1985~. Yet a number of states have already demonstrated their ability to create and find jobs. Furthermore, if it was possible to create 3.5 million WPA jobs during the Great Depression, it must be techni- cally possible to find or create a similar number now with a lower unemployment rate and an expanded economy. Finally, scholars who have explored this issue have estimates] that the supply of jobs is was extended to part-time as well as full-time work, and the eligibility level for those already on welfare was raised to 185 percent of the state's need standard.

136 Sara McLa?~ahan, Irun?~ Ga,fi?~kel, arid Dorothy Watson more than sufficient to warrant enforcement of work requirements (Fechter, 1975~. Apart from the issue of technical feasibility is the question of whether the benefits of enforcing work offset the costs. Studies of work and training programs for women who head families generally report sufficiently large gains in earnings to make the programs prof- itable within 3 or 4 years, although they do cost more initially in comparison to the payment of cash benefits only (Bass) and Ashen- feTter, 1986; Hollister et al., 1984~. The most carefully evaluated job creation program the National Supported Work Demonstration- indicated that the net social benefit per participant was $8,000. Initial costs are higher than welfare costs, however, because the cost of finding or creating jobs must be paid as well as the cash benefit. The effects of enforcing work requirements on the incomes of single mothers depend on the nature of the work programs created, the magnitude of the increased earnings of participants, and the rules of the AFDC program. Single mothers in work and training programs gain an average of $600 to $1,200 in earnings per year (Bass) and Ashenfelter, 1986) an amount equaling one-quarter of the average poverty gap for mother-only families and representing a substantial proportion of the total earnings of poor single mothers. In the National Supported Work Demonstration, the average annual increase in earnings of $900 represented an increase of nearly 50 percent over the earnings of the control group (Hollister et al., 1984; Manpower Demonstration Research Corporation, 1980~. Although the potential gains in earnings can be relatively large, the increase in the incomes of single mothers will be smaller than the increase in earnings because they will lose some AFDC and other transfer benefits. Whether AFDC families realize gains or losses from the enforcement of work requirements will depend on the nature of the key programs that aid poor single mothers and on the attractiveness and availability of jobs in the regular labor market. Even if income increases, however, economic well-being may decrease because, in the absence of work requirements, many mothers may have chosen the combination of lower income from welfare and more time for childrearing, housework, and leisure. The effects of enforced work will almost certainly reduce depen- dence on welfare. Even if the increased earnings of AFDC mothers lead to no decrease in AFDC receipts, the share of total income from welfare must decrease. Work programs, however, will probably de- crease the total amount of welfare received because there will always

FAMILY STRUCTURE, POVERTY, AND THE UNDERCLASS 137 be some recipients who are at the margin between choosing welfare and another alternative. Some research also suggests that if the labor market is strong, a nontrivial proportion of single mothers on AFDC need little more than good professional help in locating jobs. For ex- ample, preliminary evidence from San Diego, where unemployment rates are well below average, indicates that job search assistance has been the most profitable service provided there (Gueron, 1986, 1987~. The effect of enforced work requirements on the prevalence of mother-only households depends on the extent to which the economic well-being of single mothers is improved and on the extent to which the relative earnings opportunities of men and women are altered. If enforcing work requirements increases the earnings of single mothers relative to those of men, single mothers will probably become less dependent both on men and on welfare. There are three reasons for caution in interpreting the above ev- idence in favor of compulsory work programs. First, whereas partic- ipation to date in most of the work and training programs evaluated has been voluntary, much of the current public discussion involves making work compulsory. Programs that involve significant elements of compulsion may be less profitable both to the beneficiaries and to society as a whole. Early experience with the workfare programs, however, suggests that to date, at least, enforcing work requirements also seems to be profitable (Gueron, 1986, 1987~. Second, and even more important, few single mothers of families in the work and training evaluations have had preschool-age children. The child-care costs for such children could easily be so high as to offset the earnings gains of the program. Long-run earnings gains could more than make up for child-care costs, but the opposite is equally possible. This issue warrants more experimentation and study. Finally, it may be unrealistic to expect single mothers to work full-time, year round. As Ellwood (1985) points out, the only way that most single mothers can be self-supporting is by working full- time, full-year. Such complete participation in the labor force is the exception rather than the rule among all mothers, contrary to popular belief. Single mothers already work more hours than wives in married-couple households: 35 percent of single mothers with children under 6 work at least 1,500 hours per year, compared with 23 percent for comparable wives. Similarly, 50 percent of single mothers with older children are fully employed, compared with 37 percent of wives. Working 1,500 hours or more remains the exception,

138 Sara McLanahar`, Irunn Garfinicl, arid Dorothy Watson not the rule, for all mothers. Ellwood argues that, because the norm is for mothers to spenc] considerable time with their children, it may be unrealistic to expect behavior from single mothers that deviates markedly from this norm (Ellwood, 1985~. This is especially true because work requirements impose a dual role on single mothers: because only one parent is present, that parent must undertake the roles of both caretaker and breadwinner. Requiring single mothers to work for their welfare checks places a heavy burden on them. Child Support Congressional interest in enforcing child support grew as the proportion of AFDC children with absent fathers grew. The biggest burst of federal legislation on child support followed hard on the heels of the 1965-1975 growth in the welfare rolls. In addition, a consensus had developed that the existing child-support system condoned parental irresponsibility. A special study conducted by the Census Bureau in 1979 found that only 59 percent of women with children potentially eligible for support were awarded payments. Of those awarded payments, only 49 percent received the full amount due them and 28 percent received nothing. In addition, award levels and enforcement efforts were arbitrary and inequitable (Bureau of the Census, 1981a). The milestone 1976 act created federal and state offices of child support enforcement the public bureaucratic machinery to enforce the private obligation to support one's children. During the 7 years that followed, several new acts strengthened this machinery. Then, in 1984, Congress unanimously enacted by far the strongest federal child-support legislation, requiring all states to enact laws that with- hold from wages all future chiTd-support payments once the obligor is delinquent in payments for one month. The legislation also re- quires states to appoint commissions to design statewide guidelines for child-support standards. The 1984 act requires state child-support offices to provide as- sistance to nonwelfare as well as welfare cases. Although states may charge for these services and thereby target subsidies toward the poor, the service itself is provided universally to rich and poor custodial parents. The contrast between the restrictions for AFDC eligibility and the universalization of eligibility for child-support en- forcement services could not be more stark.

FAMILY STRUCTURE, POVERTY, AND THE UNDERCLASS 139 Because the 1984 child-support legislation is so recent, an as- sessment of its effects relies heavily on theoretical expectations and rough empirical estimates. The incomes of mother-only fannies can increase as a result of the withholding of wages of the delinquent supporting parent; they can also increase depending on the state guidelines for determining child-support obligations, the incentives of states to increase non-AFDC collections, and the incentives for interstate collection of child support. The size of the increase will depend on how the 1984 legislation is implemented on both the fed- eral and state levels. There will be few positive effects if the states enact weak standards and neither the number nor the amount of child-support awards increases much; if the states fait to effectively enforce the new law for withholding wages; and if federal, state, and local resources to enforce child support are cut. On the other hand, further strengthening of child-support enforcement could greatly- in- crease the incomes of mother-only families. To estimate the potential effect of child-support enforcement, we explored what would happen if all children potentially eligible for support obtained a child-support award based on some agreed- upon standard, and what the outcome would be if all such children received the full amount due them. According to a simple percentage- of-income standard used in Wisconsin, the child-support obligation is equal to 17 percent of the gross income of the noncustodial parent for one child, 25 percent for two, 29 percent for three, 31 percent for four, and 34 percent for five or more children. (In our calculation, we tax only the first $50,000 of income for child support.) Using this standard, we estimate that the incomes of families headed by women would increase by more than $10 billion (Garfinke! and McLanahan, 1986~. The poverty gap would be reduced by nearly $2 billion. These estimates should be considered an upper bound, however, because even the most efficient collection system would fall short of 100 percent collection. Increased enforcement of child support will raise the incomes of some single mothers who receive AFDC high enough to enable them to leave welfare. The precise effect of the child-support legislation on welfare dependence will vary according to the extent that collections will increase as a result of wage withholding and the new state standards, on the one hand, and the effect of the increased collections on caseloads, on the other. Some crude estimates are that (a) if existing awards are used as a standard, caseloads could be reduced by less than 5 percent; (b) if the Wisconsin standard described above

140 Sara McLanahan, Irwin Garfirzlel, and Dorothy Watson' is used, caseloads would decrease by 25 percent. Again, this estimate is an upper bound because it assumes a 100 percent collection rate. Still, even 100 percent collection of child-support obligations de- rived from any reasonable standard would leave the overwhelming majority of AFDC recipients no better off than they were in the absence of the program. This situation prevails because most non- custodial parents of AFDC children do not earn enough to pay as much child support as their children are already receiving in AFDC benefits. Programs to increase the employment and earnings of poor noncustodial fathers would help. But even the best program imagin- able would still leave a large proportion of the AFDC caseload poor and dependent on government benefits. Most of the increases in collections of child support for families on welfare will accrue to the government in the form of AFDC savings. Low-income families on AFDC can share in some of the i] · ~ ~~ .. ~ . .~ . . . . _ . increased collections ot child support in two ways. one approach is to ignore some of the child support payment in calculating AFDC grants. Congress has required all states to ignore the first $50 per month. That requirement modestly increases the incomes of mother- only families on AFDC in which there is a living, noncustodial father who makes child-support payments. It also increases by a small amount the number of mother-only families who will continue to receive AFDC. An alternative approach is to use the increased child-support col- lections to help fund a nonwelfare benefit that encourages work. This approach is being pursued on a demonstration basis in Wisconsin. Under the Wisconsin child-support assurance system, child-support obligations are determined by a simple legislated formula that was described above. The obligation is withheld from wages and other sources of income in all cases, just as income and payroll taxes are. The child is entitled to receive either the money paid by the noncusto- dial parent or an assured child-support benefit, whichever is greater. Thus, the savings in AFDC that result from increased child-support collections are funneled back into the system, in the form of assured benefits and wage subsidies, to increase the economic well-being of fannies with children eligible for child support.~3 i3The state of Wisconsin is also considering a work expenses subsidy of $1.50 an hour to the custodial parent. Child-support legislation could address two dimensions of the disadvantage suffered by families headed By single women: the low earnings of mothers relative to fathers and the lack of support from the absent parent. Child-support legislation in general attempts to tackle the latter

FAMILY STRUCTURE, POVERTY, AND THE UNDEROLASS 141 We estimate that such a program could reduce the poverty gap among American families potentially eligible for child support by more than a third and AFDC caseloads by more than a half, and even reduce total public expenditures. The effects on poverty and welfare do not depend on how much collection improves, but the costs do. If 100 percent of the Wisconsin standard were collected, the program would save $2.4 billion. If only 70 percent were collected, the net cost would be about $60 million. One criticism of the child-support assurance program is that it will benefit only those mothers who work. For those who are unable to work or who cannot find jobs or who simply prefer to take care of their children full-time, the program provides nothing. By contrast, the $50 per month set-aside that Congress enacted in 1984 provides more for this group than the child-support assurance program. Thus, the success of this latter approach will hinge largely on the extent to which both poor custodial mothers and poor noncustodial fathers work. Enhanced enforcement of child support is, on balance, likely to reduce the prevalence of families headed by single women. It is also likely to reduce out-of-wedlock births by giving men an incentive to take responsibility for birth control. In order for enforcement of child support to have an appreciable effect on out-of-wedIock births, however, there would have to be a sizable increase in the number of cases in which paternity is established. Enforcement of child support may also reduce divorce by making it financially more difficult for the noncusto(lial parent. The impact of stronger enforcement on the behavior of the prospective custodial parent is likely to be smaller because welfare already exists as an alternative means of support. issue by enforcing parental responsibility; in contrast, the Wisconsin assured benefit program represents an attempt to tackle the issue of the mother's earnings as well, both by providing an assured benefit and by providing a wage subsidy.

142 Sara McLanahan, Irunn Garfinkel, and Dorothy Watson REFERENCES Aponte, R., K. M. Neckerman, and W. J. Wilson 1985 Race, family structure and social policy. Ch. 2 of Working Paper 7: Race and Policy. Project on the Social Role. Washington, D.C.: National Conference on Social Welfare. Auletta, K. 1982 The Underclass. New York: Random House. Bassi, L. J., and O. Ashenfelter 1986 The effect of direct job creation and training programs on low-skilled workers. Pp. 133-151 in S. H. Danziger and D. H. Weinberg, eds., Fighting Poverty: What Works and What Doesn't. Cambridge, Mass.: Harvard University Press. Becker, G. 1957 The Economics of Discrimination. Chicago: University of Chicago Press. Bergmann, B. R. 1974 Occupational segregation, wages and profits when employers discrim- inate by race or sex. Eastern Economic Journal l(Apr.-July):103-110. Bielby, W. T., and J. N. Baron 1984 A woman's place is with other women: Sex segregation within organi- zations. Pp. 27-55 in B. F. Reskin, ea., Sea Segregation in the Workplace: lkends, Explanations, Remedies. Washington, D.C.: National Academy Press. Blau, F. D. 1984 Occupational segregation and labor market discrimination. Pp. 117- 143 in B. F. Reskin, ea., Sea Segregation in the Workplace: fiends, Explanations, Remedies. Washington, D.C.: National Academy Press. Bukharin, N. 1921 Historical Materialism. (Authorized English Translation.) Ann Arbor, Mich.: Ann Arbor Paperbacks, 1969. Bumpass, L. L. 1984 Children and marital disruption: A replication and update. Demogra- phy 21:71-82. Bumpass, L. L., and J. A. Sweet 1972 Differentials in marital instability: 1970. American Sociological Review 37(Dec.) :754-766. Bureau of the Census 1973 Low Income Areas in Large Cities. 1970 Census of Population. Subject Report PC(2~-9B. Washington, D.C.: U.S. Department of Commerce. 1981a Child Support arid Alimony, f978. Current Population Reports Special Studies, Series P-23, No. 112. Washington, D.C.: U.S. Department of Commerce. 1981b Household and Family (characteristics: March 1980. Current Population Reports, Series P-20, No. 366. Washington, D.C.: U.S. Department of Commerce. 1984 Household and Family Characteristics: March 1985. Current Population Reports, Series P-20, No. 388. Washington, D.C.: U.S. Department of Commerce. 1985a Money Income and Poverty Statw of Families and Persons in the Uruted States: 1984. Current Population Reports, Series P-60, No. 149. Washington, D.C.: U.S. Department of Commerce.

FAMILY STRUCTURE, POVERTY, AND THE UNDERCLASS 143 1985b Poverty Areas ire Large Cities (And Addendum). 1980 Census of Popula- tion. Subject Report PC-80-2-8D. Washington, D.C.: Department of Commerce. Cherlin, A. 1976 Social and Economic Determinants of Marital Separation. Unpub- lished Ph.D. dissertation, Department of Sociology, University of Los Angeles. Cole, S., S. Danziger, and I. Piliavin 1983 Poverty and Welfare Recipiency After OBRA: Some Preliminary Ev- idence from Wisconsin. Institute for Research on Poverty, University of Wisconsin. October. Corcoran, M., and G. J. Duncan 1979 Work history, labor force attachment and earnings differences between the races and the sexes. Journal of Human Resources 14(Winter):3-20. Cutwright, P., and P. Madras 1976 AFDC and the marital and family status of ever married women age 15-44: United States, 1950-1970. Sociology arid Social Research 60:314-327. Danziger, S. H., G. Jakubson, S. Schwartz, and E. Smolensky 1982 Work and welfare as determinants of female poverty and household headship. Quarterly Journal of Economic 97:519-534. Darity, W., and S. M. Myers 1983 Changes in black family structure: Implications for welfare depen- dency. American Economic Review Proceedings 73(May):59-64. Duncan, O. D., D. Featherman, and B. Duncan 1972 Socioeconomic Background arid Achievemcnl;. New York: Seminar Press. Duncan, G. J., and S. Hoffman 1985 Welfare Dynamics and Welfare Policy: Past Evidence and Future Research Directions. Paper presented at the Association for Public Policy Analysis and Management meetings, October, Washington, D.C. Ellwood, D. T. 1985 Targeting the Would-be Long-term Recipients of AFDC: Who Should Be Served? Unpublished Preliminary Report. John F. Kennedy School of Government, Harvard University. Ellwood, D. T., and M. J. Bane 1984 The Impact of AFDC on Family Structure and Living Arrangements. Report prepared for the U.S. Department of Health and Human Services (Grant No. 92A-82. John F. Kennedy School of Government, Harvard University. Emery, R. E. 1982 Interparental convict and the children of discord and divorce. Psycho- logical Bulletin 92:310-330. England, P. 1982 The failure of human capital theory to explain sex segregation. Journal of Human Resources 17(Summer):358-370. Featherman, D., and R. Hauser 1978 Opportunity arid Charge. New York: Academic Press. Fechter, A. E. 1975 Public Employment Programs. Washington, D.C.: American Enterprise Institute for Public Policy Research.

144 Sara McLanahan, Irwin Garfinkel, and Dorothy Watson Garfinkel, I., and S. S. McLanahan 1985 The Feminization of Poverty: Nature, Causes and a Partial Cure. Dis- cussion Paper No. 776. Institute for Research on Poverty, University of Wisconsin. 1986 Single Mothers and Their Children: A New American Dilemma. Washing- ton, D.C.: Urban Institute Press. General Accounting Office 1985 Evidence Is In~ufficierd to Support the Administration'` Proposed Changes to AFRO Work Programs. Report to the Chairman, Subcommittee on Intergovernmental Relations and Human Resources, House Committee on Government Operations. Washington, D.C.: General Accounting Once. August 27. Giddens, A. 1973 The Class Structure of Aduanced Societal. New York: Harper and Row. Gordon, D. M., ed. 1977 Problems in Political Economy: An Urban Perspective. Lexington, Mass.: D.C. Heath and Company. Groeneveld, L. P., M. T. Hanna, and N. B. Tuma 1983 Final Report of the Seattle-Denver Income Maintenance Experiments, vol. 1, part 5. Menlo Park, Calif.: Stanford Research Institute International. May. Gueron, J. M. 1986 Work Initiatives for Welfare Recipients. March. New York: Manpower Demonstration Research Corporation. 1987 Reforming Welfare with Work. February. Manpower Demonstration Research Corporation. Herzog, E., and C. Sudia 1973 Children in fatherless families. Pp. 141-231 in B. Caldwell and H. Riccinti, eds., Review of Chad Development Research, vol. 3. Chicago: University of Chicago Press. Hetherington, E. M., K. A. Camara, and D. L. Featherman, 1983 Achievement and intellectual functioning of children in one-parent households. Pp. 205-284 in J. Spence, ea., Achievement and achievement Motives. San Francisco, Calif.: W. H. Freeman. Hofferth, S. 1985 Updating children's life course. Journal of Marriage and the Fancily 47:93-1 16. Hoffman, S., and J. Holmes 1976 Husbands, wives, and divorce. Pp. 23-75 in G. J. Duncan and J. Morgan, eds., Five Thousand American Fam~lic~, vol. 4. Ann Arbor, Mich.: Institute for Social Research. Hogan, D. P., and E. M. Kitagawa 1985 The impact of social status, family structure, and neighborhood on the fertility of black adolescents. Amerscan Journal of Sociology 90(Jan.):825-856. Hollister, R. G., Jr., P. Kemper, and R. A. Maynard, eds. 1984 The National Supported Work Demonstration. Madison: The University of Wisconsin Press. Honig, M. 1973 The Impact of AFDC income, recipient rates, and family dissolution. Journal of Human Resources 9:303-322.

FAMILY STRUCTURE, POVERTY, AND THE UNDERCLASS 145 Hutchens, R. M. 1969 Welfare, remarriage, and marital search. American Economic Renew 69:369-379. Joe, T., R. Sari, and M. Ginsberg 1984 Working Female Headed Families in Poverty: Three Studies of Income Families Affected by the AFDC Policy Changes of 1981. Washington, D.C.: Center for the Study of Social Policy. Kasarda, J. D. 1985 Urban change and minority opportunities. Pp. 33-37 in Paul E. Peterson, ea., The New Urban Reality. Washington, D.C.: Brookings Institution. Krein, S. F., and A. H. Beller 1986 Family Structure and Educational Attainment of Children: Differences by Duration, Age, and Gender. Paper presented at the annual meeting of the Population Association of America, San Francisco. Lenin, V. I. 1914 A great beginning. Pp. 482-502 in The Essentials of Lenin. London: Lawrence and Wishart, 1947. Levy-Shiff, R. 1982 Coping with the loss of father: Family reaction to death or divorce. Journal of Family Issues 3 :41-60. Lewis, O. 1966 La Vida: A Puerto Ricar' Family in the Culture of Poverty San Juan and New York. New York: Random House. Manpower Demonstration Research Corporation 1980 Summary and Fin dings of the National Supported Work Demonstration. Cambridge, Mass.: Ballinger. McLanahan, S. S. 1983 Family structure and stress: Longitudinal comparison of male- and female-headed families. Journal of Marriage and Family 45:347-357. 1985 Family structure and the reproduction of poverty. American Journal of Sociology 90:873-901. 1988 Family structure and dependency: Early transitions to female house- hold headship. Demography, forthcoming. McLanahan, S. S., and L. L. Bumpass 1986 Intergenerational Consequences of Family Disruption. Paper pre- sented at the annual meeting of the Population Association of Amer- ica, April 3-6, San Francisco, Calif. Minarik, J., and R. Goldfarb 1976 AFDC income, recipient rates, and family dissolution: A comment. Journal of Human Resources 9(Spring) :243-257. Mincer, J., and S. W. Polachek 1974 Family investment in human capital: Earnings of women. Journal of Political Economy 82(March/April, Part 23:S76-S108. Moore, K. A., and L. J. Waite 1976 Marital dissolution, early motherhood and early marriage. Social Forces 60(September) :20-40. Moynih an, D. P. 1965 The Negro Family: TheCa~eforNat~onalAction. Office ofPoliGy Planning and Research. Washington, D.C.: U.S. Department of Labor.

146 Sara McLanahar~, Irwin Garfinkel, arid Dorothy Watson Oellerich, D., and I. Garfinkel 1985 Distributional impacts of existing and alternative child support sys- tems. Policy Studies Journal 12:119-130. Parkin, F. 1971 Class Inequality and Political Order. New York: Praeger. Pearce, D. 1978 The feminization of poverty: Women, work, and welfare. Urban and Social Change Review 11:28-36. Peterson, J. L., and N. Zill 1986 Marital disruption, parent-child relationships, and behavior problems in children. Journal of Marriage and the Family 48:295-307. Piore, M. 1977 The dual labor market: Theory and implications. Pp. 93-97 in D. M. Gordon, ea., Problem~inPoliticalEconomy: An Urban Perspective. Lexington, Mass.: D.C. Heath and Company. Polachek, S. W. 1977 Differences in expected post-school investment as a determinant of market wage differentials. Pp. 127-148 in P. A. Wallace and A. M. Lamond, eds., Women, Minorities and Employment Discrimination. Lexington, Mass.: Lexington Books. Preston, S. H., and A. T. Richards 1975 The influence of women's work opportunities on marriage rates. De- mography 12:209-222. Reskin, B. F., and H. I. Hartmann, eds. 1984 Women's Work, Men's Work. Sex Segregation on the Job. Committee on Women's Employment and Related Social Issues, Commission on Behavioral and Social Sciences and Education. Washington, D.C.: National Academy Press. Ross, C. 1984 Trends in poverty, 1965-1983. December. Paper prepared for the conference, "Poverty and Policy: Retrospect and Prospects." Insti- tute for Research on Poverty, University of Wisconsin, and the U.S. Department of Health and Human Services. Ross, H., and I. Sawhill 1975 Time of lEan~itior`: The Growth of Families Headed by Women. Washing- ton, D.C.: Urban Institute Press. Sawhill, I. 1976 Discrimination and poverty among women who head families. Signs 2:201-211. Shaw, L. 1979 Does Living in a Single-Parent Faintly Affect High School Completion for Young Women? Columbus, Ohio: State University Center for Human Research. Shinn, M. 1978 Father absence and children's cognitive development, Psychological Bulletin 85:295-324. South, S. 1985 Economic conditions and the divorce rate: A time-series analy- sis of post-war United States. Journal of Marriage and the Fanuly , _ ,~ . ~ _ . 4-' tt e uruary) :51-41 .

FAMILY STRUCTURE' POVERTY, AND THE UNDERCLASS 147 Treiman, D. J., and H. I. Hartmann, eds. 1981 Women, Work, and Wages: Equal Pay for Jobs of Equal Value. Committee on Occupational Classification and Analysis, Commission on Behav- ioral and Social Sciences and Education. Washington, D.C.: National Academy Press. Usher, C. L., and J. D. Griffith 1983 The 1981 AFDC Amendments and Caseload Dynamics. Research Triangle Park, N.C.: Research Triangle Institute. Weber, M. 1922 Economy and Society, edited by Gunther Roth. New York: Bedminster Press, 1968. White, L. 1981 A note on racial differences in the effect of female economic opportu- nity on marriage rates. Demography 18(August) :349-354. Wilson, W. J. 1985a Cycles of deprivation and the underclass debate. Social Science Review 59(December) :541-559. 1985b The urban underclass in advanced industrial society. Pp. 129-160 in 17`e New Urban Reality, Paul E. Peterson, ed. Washington, D.C.: Brookings Institution. Wilson, W. J., and R. Aponte 1985 Urban poverty. Annual Review of Sociology 1 1 :231-2 58. Wilson, W. J., and K. M. Neckerman 1986 Poverty and family structure: The widening gap between evidence and public policy issues. Pp. 232-259 in S. H. Danziger and D. H. Wein- berg, eds., Fighting Poverty: What Works and What Doesn't. Cambridge, Mass.: Harvard University Press. Wright, E. O. 1979 Class Structure and Income Determination. New York: Academic Press. 1985 Classes. London: Verso.

Next: Jobs, Migration, and Emerging Urban Mismatches »
Urban Change and Poverty Get This Book
×
Buy Paperback | $100.00
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

This up-to-date review of the critical issues confronting cities and individuals examines the policy implications of the difficult problems that will affect the future of urban America. Among the topics covered are the income, opportunities, and quality of life of urban residents; family structure, poverty, and the underclass; the redistribution of people and jobs in urban areas; urban economic growth patterns; fiscal conditions in large cities; and essays on governance and the deteriorating state of cities' aging infrastructures.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!