4 475 December 26, 1985 POVERTY IN AMERICA WHA T THE DATA REVEAL INTRODUCTION Poverty continues to be a highly emotional and politicized issue in the United States. Unfortunately for both the poor and the taxpayer, the amount of money voted for anti-poverty programs has been taken as a measure of indi v idual lawmakers' concern for poor Americans. Selective data are used routinely to score political points, yet the broad picture provided by the mass of available statistics is rarely considered when policy decisions are made As Congress prepares to consid e r the FY 1987 federal budget, with the legal requirement that it must balance the budget by 1991, it is essential that program changesare made based on the nature of poverty as revealed by noncontroversial statistics, and not according to anecdotes and un representative emotional appeals.

Such statistics, produced by the Census Bureau and respected academic researchers, show that poverty in America has a very clear pattern four years-indeed the economic expansion since 1982 has caused the first reduction in the poverty rate in six years and the largest reduction since the 1960s. Poverty in the U.S. has much more to do with deep cultural and demographic trends policy can significantly affect the pattern of poverty, it must thus address the issues raised by t h ese trends single-parent-headed households of the nuclear family policy is the cause of this erosion, it is nonetheless the case that poverty programs do concentrate on the consequences of the trend and provide few incentives to keep families together. If Congress is to address poverty seriously, it must deal with these fundamental social It has nothing to do with the economic policies of the past To the degree that federal Poverty is heavily concentrated, for instance, among children in This is a reflecti o n of the erosion While it is unreasonable to claim that federal I aspects of the problem children and single parents with child care responsibilities suggests that the Great Societyls objective of self-sufficiency may be an elusive goal overwhelm even the most numerate lawmaker, it is important for the poor that their true characteristics be kept firmly in mind statistical evidence reveals that the vast majority of the poor are not working long hours at Ilpoverty wages.I1 work force or have only a very ten uous connection with it. This does not imply that the majority of the poor are unwilling to work.

Most are, in fact, children, or elderly, or disabled, or have child care responsibilities, or work part-time because they are poorly educated and have limited job opportunities that programs, as now designed, reach all those in need 40 percent do not receive any public financial assistance at all.

The poor tend to be young: families with heads of households under 24 and children have the highest poverty rates hand, despite the rhetoric, the elderly are not, as a group, in dire straits; they are relatively well off, although certain subgroups of the aged, such as women over 70 living alone, can be very poor closely connected to the fact that paid employment is t he surest way out of poverty find themselves, particularly if they are single parents, raise considerable barriers for them in gaining a foothold in the labor market. The poverty rate for full-time year-round workers is a remarkably low 2.9 percent at the minimum wage would not be enough to raise anyone but a single individual or elderly couple above the poverty line A poverty population composed primarily of In the upcoming budget debate, where statistics will tend to And the Most are eithef not in the Mo r eover, it is a myth As manyzas On the other The fact that women and children are more likely to be poor is But the circumstances in which women with children This is true even though annual earnings The I1feminization1l of poverty also turns out to have n o thing to do with sex discrimination. Rather, it reflects broad societal changes over several decades. Moreover, the underlying reason for increasing poverty among children since the 1960s is clearly related to the marked growth of female-headed households . This presents the greatest single challenge to anti-poverty policy in the 1980s. And the lessons of the 1960s and 1970s make it clear that there are no quick and easy solutions 1. Data in this paper, unless otherwise noted, are from Census Bureau publica t ions especially Monev Income and Povertv Status of Families and Persons in the United States; 1984 (Washington, D.C Current Population Reports, Series P-60, No. 149, 1985 2. See S. Anna Kondratas The Problems of Measuring Poverty," Heritage Foundation Bac k grounder No. 360, November 1, 1984, p. 1 2Rather than engaging in another acrimonious debate on the anti-poverty budget, where the questions center on how much money should be channeled into existing programs, Congress should begin considering structural reforms in the decades-old welfare programs.

The best way for lawmakers to prepare themselves for that task is for them to recognize the nature of poverty revealed by the available data.

CHARACTERISTICS OF THE POVERTY POPULATION Acre Distribution The age groups with the lowest poverty rates predictably are those in their prime earning years between ages 22 and 44 is 11.7 percent (compared with the overall poverty rate of 14.4 percent Above age 55, the poverty rate begins to inch up, but the 12.4 percent r a te for those 65 and over is still below the national average The poverty rate for those The economic situation of the elderly, as a group, is actually more favorable than their official poverty rate indicates. This is because assets are ignored in the off i cial definition of poverty which takes only annual pre-tax cash income into account elderly household has one-third more after-tax income per member than the avesrage baby boom household and three times its financial assets. Thus the popular image of the e lderly as an economically disadvantaged group is false. This does not mean, of course, that segments of the elderly population do not. suffer disproportionately high rates of poverty. About one-fourth of unrelated elderly individuals, for instance, are po or, and about one-half of the elderly poor are aged women who live alone, especially those past 70 The average The higheat poverty rates by age, however, are among the young.

Families where the head of household is between the ages of 15 and 24 have a poverty rate of 29 percent. Children under 18 suffer a poverty rate of 22 percent, and younger children are more likely to be poor.

Black children under 3, for example, have a poverty rate of 53 percent 3. Aldona E. Robbins and Paul Craig Roberts, The Economi c Status of the Aged Imt>lications for Energy Policy, The Institute for Political Economy, Washington, D.C September 27, 1985, pp. 3-4 3The total number of females in poverty is 19.1 million, compared to 14.5 million males more poor women than men, there a re only 6 percent more women than men in the general population groups, including children is also higher than that for men. The most significant difference is in the over-65 group. An elderly woman is nearly twice as likely as a man to be poor (poverty r ates of 15 and 8.7 percent respectively). Only 7 percent of poor men are 65 and over, but 13 percent of poor women are in that age bracket. This reflects traditional work and retirement-benefit patterns of women.

Their greater longevity is also a factor, a s savings become exhausted and health deteriorates While there thus are about 32 percent again The poverty rate for women in all age Work Patterns The data indicate that 12.2 percent of the population over 15 is poor--over 22 million people. Of these, ove r 13 million did not work at all during 19

84. Surveys show that 85 percent of these individuals did not work because they were ill or disabled (2.7 million), keeping house (4.1 million), going to school (2.3 million), or retired (2.1 million work (1.4 mil lion) were unable to find work year, only 2 million worked full-time all year round. In fact, the poverty rate for full-time workers who do not suffer spells of unemployment is a low 2.9 percent (5.2 percent for blacks Families whose household head works f ull-time have a poverty rate of 3.5 percent. Nevertheless, it is possible for a person to work full-time all year and still fall below the official poverty line, particularly if he or she has dependents or not the family is intact, with the potential for a t least two full-time wage earners 6,968; this is below the official poverty level for all but two-earner families usually escape poverty. Indeed, two full-time minimum-wage earners in a family would raise even a five-person household out of poverty Only 1 1 percent of adult Americans in poverty who did not I Of the 9.1 million poor who did work at some time during the I But the most important factor is whether Gross annual earnings at the minimum wage are individuals ($5,278) and elderly couples ($6,282 On the other hand, 1 GeoaraBhical Distribution The South, with 12.8 million in poverty, has the greatest number of poor by far; this compares with 8.3 million in the Midwest, 6.5 million in the Northeast, and 6.1 million in the West. This is partly because t h e South, with 34 percent of the total U.S. population, is now the most populous region. But the rate in the South also is exaggerated by shortcomings of the poverty rate as a measure of actual poverty. In particular, the official poverty definition does n o t take into account geographic cost-of-living differentials the South 6 million and 7.5 million respectively but more than half of America's poor blacks--some 5 million--reside in the South accounting for one of three blacks in that region. Yet the highes t black poverty rate is in the Midwest: nearly two of five Midwestern blacks are poor. And contrary to the myth that Americals poor are an urban underclass, ply 14 percent of the poor live in "poverty areas of central cities The number of poor whites in th e Midwest is nearly as high as in Children and Poverty Whereas the vast majority of poor families have three or fewer children under 18, children affect a family's financial status considerably, as any parent knows. Families with no children have.a very lo w 5.4 percent poverty rate. Families with one child have a poverty rate of 12.7 percent families. The poverty rate for families with four children is 34.5 percent and 52.7 percent for those with five or more As might be expected, the highest poverty rate, 8 7.1 percent is for black female-headed families with five children or more; but there are only some 100,000 such poor families in the U.S The rate rises rapidly for large Education More than half of poor household heads over 24 have less than a high schoo l education. The less educated the head of household, the more likely a family is to be poor only an elementary education or less, nearly one in four families is poor. When the head of household has finished one to three years of high school education, one of five such families is poor. The overall poverty rate drops to one of twenty families if the family head has one year or more of college When the head of household has Black families with similar levels of education to those of families in other'groups a re more likely to be poor, and women are even more so. The poverty rate for black families when the head of household has four years of high school, for example, is 26 percent compared with 10 percent overall for that educational group); for female-headed ' families 27 percent; and for black female-headed families, 44 percent. But although this correlation exists between poverty and educational achievements, skills as such may not be the crucial factor. In the majority of poor families headed by blacks and w omen, the head of household is not only not employed but he or she is 4. William P. O'Hare, "The Myths of Poverty," Focus, Joint Center for Political Studies May 1985, p. 3 5- not even in the labor force. Families headed by white males are far more likely to have employed household heads white male household heads are not in the labor force, compared with 45 percent of poor black males, 60 percent of white females, and 62 percent of black females. For all families with a household head not in the labor for c e, the poverty rate is 23 percent Only 35 percent of poor LONG-TERM AND SHORT-TERM POVERTY I According to a study of the poor conducted at the University of Michigan, there are difference? in the characteristics of the long-term and short-term poor. Over t he lO-year period under study 1968-1978 about 25 percent of the population experienced at least a short spell of poverty did not differ appreciably from those of the general population the persistently poor-less than 3 percent of the population according t o the Michigan researchers--were overwhelmingly in black or female-headed households, and mainly concentrated in black female-headed households were black, and about 61 percent were in female-headed households with considerable overlap The characteristics of these short-term poor But I About 62 percent of the persist-ently poor The Michigan researchers, who believe their results are still applicable, note that more of the short-term poor live in urban areas than in small towns or rural areas, but that the p ersistently poor are far more likely to live in small towns and rural areas. One-third of that Ilsouthern and rural poverty'are much more persistent than is urban poverty.116 One of the major reasons for this, according to their analysis, is that the urba n poor are far more likely to be receiving long-term cash welfare, which is counted as.income in determining poverty status. Earned income, according to the Michigan study, is definitely the surest way out of poverty. This could include, of course, the ind i vidual's own income and that of other family members the long-term poor live in rural areas, they discovered, suggesting I But the vast majority of the persistently poor live in households headed by a person who is disabled, elderly, or has child-care res p onsibilities families headed by able-bodied nonelderly men and fewer than half of the latter work for any substantial periods Only one-sixth of the persistently poor live in 5. Greg J. Duncan et al Years of Povertv. Years of Plentv, Institute for Social R e search, University of Michigan, Ann Arbor, Michigan, 1984, pp. 48-52. "Persistently poor" is defined in the study as having been poor 8 years or longer 6. Ibid, p. 51 6THE FEMINIZATION OF POVERTY The stereotype of a single welfare mother as a woman with m a ny children of uncertain paternity is inaccurate. The mean number of children in poor female-headed households is 2.2, compared to 2.3 in all poor families. Of poor families with no children, as well as those with 4 or more children, the majority are not female-headed.

The majority of poor families with one or two children, and a slim majority of those with three children, however, are headed by women.

A far larger proportion of poor families are headed by women today than was the case 15 or 25 years ago. In 1959, for example only 25 percent of poor whites and 29 percent of poor blacks lived in female-headed households. By 1970, that proportion had risen to 39 percent of whites and 56 percent of b lacks reached 42 percent of whites and 68 percent of blacks.

In 1984 the figures Emlanations of the Pattern This pattern has been popularly attributed to various causes.

One view is that the policy of making welfare more pvailable to women without husband s has discouraged work and marriage. Another school of thougpt blames pay discrimination against women in labor markets. Still another says that there is a dearth of marriageable black men9because of high black male unemployment rates and low earnings.

Of these explanations, the labor market discrimination theory is the weakest. The main reason female-headed families are poor is that the household head is not even in the labor force It is the changes in family structure that underlie this trend, and indee d , these 7. This is essentially the view argued by Charles Murray in Losine: Ground (New York Basic Books, 1984) and George Gilder in Wealth and Povertv (New York: Basic Books 1981 8. Without any empirical evidence, this explanation is advanced by supporte r s of the controversial concept of "comparable worth The "continuing increase in the number of women and children who live at, near, or below the poverty level asserts a bill introduced in the House of Representatives by Rep. Mary Rose Oakar (D-OH is large l y the result of such employment discrimination on the basis of sex 9. Such a possibility is suggested by Duncan OD. cit p. 64, and by June O'Neill An Analysis of Time on Welfare The Urban Institute, June 1984, p. 13 7changes 'lare primarily a response to i mprovements in the employment opportunities of women relative to men, and especially black men.nB10 Poor female-headed households, by and large, tend to be on welfare. Thus, it is not surprising that they are poor, since cash welfare benefits alone are se l dom sufficient to raise family members above the poverty threshold. On the other hand, the official poverty statistics do not necessarily reflect accurately.the relative status of these women, since many, particularly in urban areas, also receive in-kind benefits, which are not calculated in measuring poverty.

Also, welfare recipients underreport their cash welfare income to the Census Bureau by as much as 24 percent, according to Census Bureau calculations The argument about the lack of %arriageable" blac k men has some plausibility opportunities were better for black men 25 years ago than they are today; this did not seem to discourage marriage among blacks then.

But rising employment opportunities for women and rising real welfare benefits over the perio d (real benefits did not begin declining till the mid-1970s) certainly provided women with options they did not have earlier. In this sense, welfare certainly facilitates the nonformation or breakup of nuclear families the rate of female household formati o n, and has thus contributed to the increase of the number of female-headed households in poverty. In a study prepared for the U.S. Department of Health and Human Services David Ellwood and Mary Jo Bane of Harvard University found that the living arrangeme n ts of young mothers Itis sharply influenced by the level of AFDC benefits in a state.Iw" In other words, young single mothers with children who might live with their own parents in the absence of high AFDC benefits tend to set up independent households wh e re benefit levels allow that.. Ellwood and Bane estimate that "a 100 increase in benefit levels nationally would increase the number of independent female heads by as much as 15 percent."12 Harvard researchers found that AFDC benefits had an llimportant i n fluence" on divorce and separation.rates for women under the age of 24, estimating that a $100 increase in benefit levels would result in Yet it would be hard to argue that employment There is no doubt that the availability of welfare has affected I Likew i se 10. Sara McLanahan, "Charles Murray and the Family," in Losinn Ground: A C ritiaue Madison, Wisconsin: Institute for Research on Poverty, Special Report No. 38, August 1985), p 5. As further support of this thesis, McLanahan cites other research done f o r the Institute for Research on Poverty and the Urban Institute 11. David T. Ellwood and Mary Jo Bane, "The Impact of AFDC on Family Structure and Living Arrangements," Harvard University, March 1984, p. 3 12. Ibid p. 34 a- an increase13in the number of d ivorced and separated mothers of about 10 percent.

The Ellwood and Bane study noted, however, "little evidence that AFDC influenced the child-bearing decisions of unmarried women. rr14 This finding has been widely cited as that welfare is not the main caus e of rising illegitimacy in the U.S. Yet the authors themselves have noted some of the difficulties of developing a methodology to test such a hypothesis, and stressed the tentative nature of all their conclusions. It is not necessary, however, to assume t hat welfare causes illegitimacy, since there is no doubt that once illegitimate children are born, welfare levels affect the choice of a life style that enhances the prospects for long-term'welfa&e dependency and poverty, particularly for unmarried black women.

The Root Causes of the Problem The feminization of poverty cannot be blamed entirely on the structure of the welfare system. The increase in female-headed households, for instance, is by no means limited to the poor. This is another reason why it is difficult to argue that the increase in welfare benefits over the past two decades was the only or even main cause of the rise in single motherhood Since 1959, the number of persons in female-headed households has increased 133 percent; the number of per s ons in poor female-headed I declined from 50 percent to 34 percent. Two out of three persons in female-headed households today are not poor. Family dissolution in the 1960s and 1970s increased across income-level lines. But it is family dissolution among the poor that results in the greater social costs households increased only by 58 percent, as their poverty rate I THE POVERTY OF CHILDREN.

Since 1959 the share of childrenl&ving in female-headed households rose from 9 to 20 percent. Because of the inheren tly lower earning capacity of single-parent families compared with 13. Ibid p. 42 14. Ibid p. 6 15. See O'Neill, et al OD. cit 16. House Committee on Ways and Means, Children in Povertv, May 22, 1985, p 57. Much of the subsequent discussion is based on Ch a pters I11 and IV of this volume 9 7 p'two-parent families, female-headed households have significantly higher poverty rates than do those with male heads understandably, the poverty rate for children is disproportionately high--21 percent in 1984, or one c hild in five. Of these, nearly 40 percent of poor white children, and 75 percent of poor black children live in female-headed families Thus There also has been upward pressure on the child poverty rate (as opposed to numbers) because of another phenomenon : the decline in marital fertility and the trend toward smaller two-parent families.

Between 1967 (the first year for which detailed information is available) and 1983, families with children increased by about 4.3 million, anfi almost 4 million of these w ere female-headed households. Even as families increased, the total number of persons in families with children declined by about a million the number of persons in female-headed families increased by 11 million But About 62 percent of the increase in fam i lies with children represents families headed by a divorced or separated woman disquieting is the rise in families headed by women who never have been married. The number of persons in such households has risen 550 percent since 1967 (from 1 percent to ap proximately 5.5 percent).

Families headed by never married women have about a 70 percent poverty rate, and three out of four children of such mothers are poor.

One-fifth of all births in the U.S. in 1980 were to unwed mothers. It is interesting to note th at the increased incidence of child poverty has coincided with this trend. child poverty almost halved in the decade from 1959 to 1969, reaching a low of 14 percent. Although it is still lower than in 1959, it is now more than 50 percent above its Far mor e 1969 .low As for the persistently poor, 90 percent of the children in those families are black. Most lack a father at home, live in the South and are disproportionately rural residents. Overall, 40 percent of children whose mother and father both have no t completed high school are poor. Only 7 percent of the children of high school graduates are below the poverty line.

Thus it is misleading, in a sense, to talk of the phenomenon of child poverty1# as if it were a separate issue and somehow caused by gover nment policy. There is no child poverty without family poverty which is frequently the result of decisions by individual adults who have ignored their parental responsibilities. Almost 90 percent of 17. Statement by David A. Stockman, Director, Office of M anagement and Budget, before the House Ways and Means Subcommittees on Oversight and on Public Assistance and Unemployment, U.S. Congress, September 20, 1984, p. 26 10 children on AFDC, for instance, have able-bodied but absent fathers nearly 50 percent o f those fathers were not married to the mother.

POLICY ISSUES AFFECTING CHILDREN A fundamental question that policy makers will have to address is to what degree public policy aimed at reducing child poverty can, or should, address sweeping societal changes in family structure.

Certainly it is not appropriate for government to try actively to influence fundamentally private decisions concerning living arrangements or reproduction. But then what are the limits of government respons ibility for the unpleasant outcomes of personal decisions? There is a great need (and strong bipartisan support) for significantly strengthening the child support enforcement program to tap the earnings capacity of an absent parent. But this approach has its limits and cannot be expected to solve the problem of child poverty.

Societal assistance to poor children, of course, is justifiable not merely because children are not to blame for parental decisions but also because there is a strong public interest in fostering a well-educated, healthy, and responsible future generation The question is the best way of doing so. There is no denying that "higher family earnings are the primary route out of poverty for children Ill8 In male-headed families, 91 percent o f poverty exits" are the result of increased earnings. The corresponding figure for female-headed families is 60 percent. In both cases, this is usually not because the household head's wage income has increased but because an additional family member has joined the workforce way to help children is to encourage or enable their parents to work whether they are single or married. California, this seems to be the direction of welfare reform Thus the best And from Massachusetts to CONCLUSION Social class in t he U.S. is not a permanent condition is characterized by a remarkably open socioeconomic system significant turnover in the poverty population as individual circumstances improve or worsen persists in poverty The U.S.

There is A very small proportion of th e poor Yet political rhetoric frequently refers to 'Ithe poort1 as an 18. Congressional Research Service Summary of Poor Children: A Study of Trends and Policy 1968-1984 May 22, 1985, p. 10 11 undifferentiated underclass, and partisan politicians sometime s evoke images of Dickensian squalor as if the poor were always victims of economic injustice sanctioned by government policy. The nature of poverty varies in different societies and changes over time in each society. In the U.S for example, poverty in the 1930s was widespread and primarily the result of massive unemployment. Today general unemployment is not the main cause of poverty--fewer than 11 percent of poor adults who did not work in 1984 gave inability to find work as the main cause of their unempl o yment percent children 10 percent elderly, and 8 percent disabled-or faces significant barriers to employment, as is the case for some of the disabled and for many of the 3 million female heads of household with children. Full-time year-round workers, the so-called working poor comprise only 6 percent of the poverty population spells of unemployment during the year (including part-time workers represent another 10 percent.

Policy makers need to develop specific strategies to ameliorate the lives of these d iffering poverty subgroups changes frequently create unexpected consequences-for example skyrocketing divorce and illegitimacy rates have resulted in the feminization of poverty-policy makers should be aware of the implications of such changes for anti-po verty and welfare policy.

They ,also must be aware of how government programs unintentionally may contribute to the impoverishment of individual groups.

The most important actions against poverty are those policies that trigger economic growth. Yet growth alone will not eliminate poverty, though without growth poverty will never be alleviated and will be certain to increase. And much poverty is beyond the government's ability to alleviate, such as that poverty caused by personal decisions leading to illeg itimacy, divorce or dropping out of school.

As such, Americans must be realistic about their government's ability to deal with poverty used. Realistic goals must be set The poverty population is largely unemployable-it includes 39 Workers who suffer Since broad societal Realistic data must be S. Anna Kondratas Schultz Senior Policy Analyst 12 I