|
|
|
|
Political Science Quarterly, Fall 1997 Teenage Childbearing and Personal Responsibility: An Alternative View Arline T. Geronimus
The nature and scope of the welfare reform debate resulting in the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 (PRWORA) reflect a growing convergence toward the position that poverty is the result of poor people's values and behaviors. Among examples of this convergence is a perspective on teenage childbearing that informed some provisions of PRWORA"Teenage childbearing" operates as a uniquely effective symbol of the failure to act responsibly. At first glance, everyone can agree that teenage childbearing is a "bad thing," unambiguously harmful to all involved -- the young parents, the innocent children, and the larger society. Because of its perceived social costs and the fact that children are involved, the invocation of incentives and punishments to alter this behavior is now considered to be within the legitimate purview of public policy. This particular form of social engineering garners enthusiastic support from representatives of different political persuasions. To the extent that it represents concern for the well-being of vulnerable children it appeals to those with the desire to help the needy. And because to many it represents extramarital sexuality it ignites the interests of those who seek to reaffirm social order. But if this apparent consensus siphons energy and resources away from searching debate about the nature of poverty, it presents a definite social danger.
In this paper, I reopen discussion of widely shared assumptions about teenage childbearing. I begin with a brief discussion of its place in recent welfare reform proposals and legislation. I then argue that the scientific evidence of causal relationships between teenage childbearing and welfare dependance, or between teenage childbearing and the health and development of the children of teen mothers, is more modest and equivocal than conventional wisdom allows. I argue further that the current evaluation of teen mothers is crafted through the prism of middle class nuclear family experience. This prism may ill-reflect the traditions or, more importantly, the environmental contingencies and life expectancy faced by members of poor families who hope to provide for children’s well-being.
Based on these arguments, I question the merit of placing the reduction of teenage childbearing as an important goal of welfare reform. Teenage mothers and their children suffer from many disadvantages that public policy might address, as do all poor families. But current efforts to reduce teenage childbearing in the poor population may be ineffective, at best, or even harmful. Meanwhile energetic focus on reducing teen childbearing may distract from the development of more effective antipoverty policies.
Teenage Childbearing and Welfare Reform “One objective of welfare reform, shared by both political parties, is to reduce teenage childbearing.” Representative Eva M. Clayton (D-North Carolina), February 27, 1996.
In the recent national welfare reform debate, both Democrats and Republicans explicitly endorsed the view that successful welfare reform requires reducing teenage childbearing. Furthermore, they both came to generally interpret teenage childbearing as “a bedrock issue of character and personal responsibility.” In June 1994, the Clinton administration offered the 103rd Congress a proposal for welfare reform that highlighted teenage pregnancy prevention. The proposal included provisions for a national campaign -- sometimes referred to as a national mobilization or crusade -- against teenage pregnancy. The 1994 proposal was unsuccessful, but the 104th Congress took up welfare reform with greater zeal and severity, including continued focus on teen childbearing.
As welfare reform was debated, some proposals highlighted the direct use of welfare regulations to discourage teenage childbearing, either by placing extra requirements on teen mothers who receive welfare benefits or by eliminating eligibility for welfare altogether among specified segments of the population of teenage mothers. For example, the 1994 Clinton administration proposal included provisions to require every school-age parent or pregnant teenager who received or applied for welfare to finish high school or enroll in a JOBS program; to live with a responsible adult, preferably a parent; and to cooperate in the effort to establish paternity. Any welfare recipient under age 23 would be reqired to search for a job during her first twelve weeks on the welfare rolls or undergo job training. Tougher proposals were suggested by moderate Democrats and by Republicans. A group of Democrats modified clinton's plan to include a provision to deny all AFDC and food stamp benefits to unwed mothers under age 21 and their children. The Welfare and Teenage Pregnancy Reduction Act introduced by Representative Jan Meyers (R-KS) proposed that no AFDC benefits be permitted unless both the mother and the father were at least 18 years old and unless paternity was established.
The provisions of the Republican Personal Responsibility Act proposed that states be compelled to deny cash assistance when paternity was not established, and to deny family assistance when a parent did not cooperate in establishing paternity. A compromise advanced by a bipartisan coalition of governors in February 1996 proposed empowering states to deny benefits to children born to unwed women under the age of 18. Massachusetts requested a waiver to enable it to deny cash assistance to mothers under age 20 who did not live with a parent or other caretaker relative. At the most extreme, conservative legislators floated the idea of enforced fosterage of children with financially needy teenage mothers. The welfare bill signed into law by President Clinton in August 1996 avoided codifying the harshest proposals into law, but does single out teen parents who hope to be welfare beneficiaries for extra requirements. Except in special cases, states are prohibited from providing assistance to teen parents or their children if the parents fail to meet these requirements. Individual states may choose to add to these prohibitions. The law also calls for the Secretary of Health and Human Services and the attorney general to implement programs aimed at teenage pregnancy prevention and for the former to implement a strategy by 1997 to assure that at least 25 percent of U.S. communities have teen pregnancy prevention programs in place.
In his 1996 State of the Union Address, President Clinton gave voice to the logic underpinning the link between teen childbearing and welfare reform. He argued that the “welfare system has undermined the values of family and work,” and that “to strengthen the family we must do everything we can to keep the teen pregnancy rates going down.” He announced that a group of prominent Americans had been convened to launch such a bipartisan national campaign. Its mission is “to reduce teenage pregnancy by supporting values and stimulating actions that are consistent with a pregnancy-free adolescence” and to do so by one-third by the year 2005.
Past arguments connecting teenage childbearing to welfare did not focus prominently on questions of values. Instead, they alleged a sizeable link between teenage childbearing and welfare costs as the rationale. In Congress, these arguments were more often made by family planning advocates than by advocates of welfare reform. Family planners who sought to place teenage pregnancy on the national agenda advocated the view that teen pregnancy was often caused by (or was a symptom of) poverty (more than vice versa) and was an educational or medical problem to be solved by increased access to contraception, abortion, and sex education. Now, reducing teenage childbearing is also seen as a measure of moral renaissance. Members of a moderate Democrat think tank argue that the campaign against teen childbearing must stand firm, that such behavior is "morally wrong," and "cast single motherhood as a selfish act." Remedies include measures that can be seen as punitive -- such as denial of welfare benefits and reinvigoration of statutory rape laws -- and national discussion of religion, culture, and public values. The legislation earmarks $400 million for abstinence-only education programs for teens, but eliminates the mandate making family planning available to welfare beneficiaries.
In terms of scientific evidence, either perspective is at best incomplete and oversimplified. Family planning advocates blur the distinction between teenagers at risk of pregnancy who avoid childbearing when given adequate information and technology and those who would bear children even provided the same. Meanwhile, welfare reformers focus narrowly on teenage childbearing and leave unexamined other important causes of persistent poverty. There is no doubt that teenage childbearing is associated with many social and public health problems. And, high rates of teen childbearing in socioeconomically disadvantaged communities likely result from the severe limits placed on the options available to the young for pursuing important goals. But none of these observations leads necessarily to the conclusion that reducing teenage childbearing itself, in the absence of other social changes, is either an attainable goal or one that would result in other social improvements. In targeting teen childbearing as a major activity for welfare reform, the basic question of cause and effect is consistently glossed over by policy makers and advocates.
The question policy makers and social scientists need to answer is the following: Would social problems be alleviated if the same women who become teen mothers postponed childbearing to older ages? This question poses a serious challenge to scientific investigation, since the same women cannot have their first births twice. Nor can we do a controlled experiment and randomly assign women to their childbearing age. Instead, analysts must try to approximate an experimental approach by comparing teen mothers to older mothers who otherwise resemble them in every possible respect. But we know that those who will become teen mothers differ from the larger population in countless and consequential ways. For example, teen mothers are far more likely to have grown up in extremely impoverished circumstances and to have experienced prior school failure. These disadvantages lead both to early motherhood and to the subsequent problems we now attribute to teen childbearing. To assess the effects of teen motherhood per se, we need to consider whether an already disadvantaged group is further disadvantaged by early motherhood. Without such reasoning, any conclusions about teenage motherhood are apt to be misleadingly dire.
Teenage Childbearing and Welfare Dependence
When proposing the Welfare and Teenage Pregnancy Reduction Act, Representative Jan Meyers (R-KS) stated, "My bill ensures that the responsibility of having a child belongs to the mother and father, rather than to the mother and U.S. taxpayers." Similarly, Senator Arlen Specter (R-PA) when debating family planning and discussing welfare reform observed, "welfare payments are enormous when we talk about teenage pregnancy, which may be the greatest domestic social problem America faces today." Statements such as these describe the simple statistical association between teen childbearing and poverty or welfare receipt and refer to public spending on women who had their first births in their teen years. However, this spending is incorrectly implied to represent the amount the government would save if teen mothers delayed their first births. That causal inference does not jibe with evidence from those studies on the relationship of teen childbearing to poverty or welfare dependence that most adequately control for the selection into teen childbearing of already disadvantaged teenagers. Simple associations may be easier to accept because they reinforce peoples’ intuition that early childbearing is detrimental; yet, the extent of measureable background differences between the teens most likely to become mothers and more typical teens nationwide suggest the importance of proceding cautiously before relying on intuitions developed from experience with typical teens.
Many studies employ standard multivariate statistical techniques to control for specified, measurable background characteristics. These studies improve over simple statistical associations, but are likely to overestimate the negative consequences of teen childbearing since there are many unmeasured factors that jointly determine a girl’s likelihood of becoming a teen mother and of being poor in adulthood. A subset of recent studies compare mothers of different ages who are similar on a fuller range of potentially important antecedents of teen childbearing than can be measured directly. For example, investigators have compared outcomes among national samples of sisters who had their first births at different ages.
Unlike more typical studies, comparing sisters, in effect, controls for such factors as general access to resources throughout childhood (not simply income at a specific point in time), cultural environment, neighborhood, or school system. Sisters comparisons do not control for personal antecedents of early childbearing that can affect longterm economic success (such as low academic ability), but vary among teens in the same neighborhood, school or even family. Sisters comparisons, then, can be seen as producing estimates of the effects of teen childbearing on subsequent economic success that remain upwardly biased, but illuminate the extent to which unobserved factors associated with family background account for the poor economic fortunes of teen mothers.
Another approach taken to account for unmeasured differences between teen and older mothers is to compare outcomes between teenage mothers and teenagers who became pregnant but miscarried. Comparing teen mothers to those who would have become teen mothers, but for the accident of miscarriage, approximates random assignment to teen motherhood and, thus, yields estimates that approximate true effects. This approach provides a more complete assessment of the full magnitude of upward bias in estimates produced by cross-sectional studies than do sisters comparisons, but does not offer evidence on what factors other than age-at-first- birth may be responsible for the economic difficulties associated with being a teen mother.
For key economic outcomes bearing directly on the question of longterm economic self- sufficiency findings from studies using these methods challenge well-entrenched beliefs. As a group, they suggest substantially smaller effects of teen childbearing on economic outcomes than more typical studies, often reducing effects estimated by more traditional methods by at least one-half. Some findings suggest no independent effect of teen childbearing on important economic outcomes. For example, Sanders, Korenman and I reported no relationship between teen childbearing and high school graduation or subsequent family income in our analysis of sister pairs in the National Longitudinal Survey of Young Women. Using more recent data from the Panel Study of Income Dynamics, Mary Corcoran and James Kunz found teen mothers to be no more likely to be welfare recipients after age 25 than their sisters who became mothers at older ages. So too, Joseph Hotz et al. comparing teens who gave birth to those who miscarried in recent data from the National Longitudinal Survey of Youth concluded that a teen mother is no more likely to participate in welfare programs, to have her labor market earnings reduced, or to experience significant losses in spousal earnings than the same woman would have experienced if she had delayed childbearing.
Not all investigators using state-of-the-art methods conclude that teen childbearing has no independent effect on economic outcomes, although that possibility cannot be ruled out. There is general agreement that the alarmist view is unsubstantiated. Yet, those debating welfare reform accepted it uncritically. Even if one prefers a conservative interpretation of recent findings that teen mothers do pay some penalty in their future income -- although lower than previously thought -- this alone would not justify targeting teen childbearing prevention ahead of addressing other contributors to poverty as an important focus of welfare reform. And it is possible to imagine a scenario (as I will suggest later) where forgoing some income is a reasonable trade-off, if early childbearing were positive in other important respects.
Teenage childbearing is believed to result in other negative outcomes for the children of teen mothers, in addition to long-term poverty or welfare dependence. These outcomes include child abuse or neglect by teen mothers, the propensity of children born to teen mothers to engage in antisocial acts such as crime in adolescence or young adulthood, infant poor health (indexed in terms, say, of low birthweight or infant mortality), and the lower developmental or academic achievement of children born to teen mothers. As the following quotes indicate, concern about such outcomes is evident in the reasoning of elected officials:
We need to make teens better understand that their actions have very serious consequences for which they are ultimately responsible....Becoming a parent is a bad deal for children. Representative Lee H. Hamilton, D-Indiana, January 25, 1995.
Tragically, high pregnancy rates among adolescents take a toll not only on young parents but also on their children. Children born to teens are at greater risk of early death and are more likely to perform poorly in school and suffer abuse and neglect, President Bill Clinton, announcing National Teen Pregnancy Prevention Month, May 1996.
Does the research literature support these beliefs?
Abuse or neglect of children by teen mothers is overrepresented in the population of maltreating parents in some samples, but not in others. Associations between teenage motherhood and maltreatment often persist after adjustment for simple demographic characteristics, such as race. However, several studies that employ more detailed multivariate analyses -- for example, including control for poverty -- fail to show teen mothers to be at increased risk of abusing or neglecting their children. No researcher of this topic has applied a research design to account for unmeasured factors. But, even based on her cross-sectional research findings, Carol Rippey Massat concluded that the prevailing wisdom that adolescent parents are more likely to maltreat their children than are older parents appears to be a myth entrenched in the popular culture. Resources might better be used to cope with harms that research has shown to occur.
There is very little evidence on the effect of teen motherhood on adolescent or young adult antisocial behavior. Kristin Moore, Donna Morrison and Angela Greene use standard cross-sectional techniques to analyze a national sample of 18-22 year olds and find no association between one’s probability of having a teen mother and having committed delinquent acts (such as assault, theft, arson, or prostitution) or used illegal drugs in the past year. They estimate an increased risk of running away from home in the previous year for 18-22 year olds whose mothers were 18 or 19 at their birth, but not for sample members whose mothers were 17 years old or younger. On the specific outcome of incarceration, Jeff Grogger tries to control for unmeasured background factors. His findings are difficult to interpret. Although he is speaking to the question of whether incarceration rates would be reduced if teen mothers delayed their first births, as Grogger acknowledges, he actually estimates the effect of maternal age, not the effect of maternal age-at- first-birth. Grogger concludes that, "even large changes in young teen mothers’ age at first birth would have a relatively modest effect on their sons’ incarceration risk." In the absence of strong causal evidence linking teen childbearing to child abuse, neglect, delinquency, or crime, one is left to assume that the firm belief that teen childbearing contributes to these problems is largely stereotypical.
On the relationship of teenage childbearing to infant health or outcomes in early childhood there is stronger evidence that refutes the idea that the association between teenage childbearing and poor infant or child outcomes is due to maternal age per se. The most robust evidence comes from studies on the health of African American infants. In contrast to President Clinton’s comments, among African Americans, rates of low birthweight and infant mortality are lowest for babies whose mothers are in their mid to late teens, and such risk among the first born tends to increase with advancing maternal age from the teens through the twenties and early thirties. There is some evidence that the increase in infant risk with advancing maternal age is steepest in poor communities. For example, in Harlem, infant mortality rates for teens are half those for older mothers (see Table 1). So, too, among first births in the poorest African American communities in Michigan, infants with 15-year-old mothers are one-half as likely to be low birthweight than those whose mothers are 25, and one-third as likely as those whose mothers are 35. One reason for this is early deterioration in maternal health. The health of poor African American women deteriorates in measurable ways as early as the mid-20s, perhaps the consequence of long-term severe socioeconomic disadvantage.
Empirical findings related to child development and school achievement also fail to provide consistent or strong endorsement for the political viewpoint that teen childbearing harms children. For example, in a national sample Moore and Nancy Snyder find no association between teen motherhood and young children's performance on the Peabody Picture Vocabulary Test among African Americans or Hispanics. Among whites, crude associations become insignificant, once the researchers control for other maternal characteristics that antecede teen childbearing. In a multivariate, cross-sectional study of a wider range of child development indicators, Moore, Morrison, and Greene arrive at conflicting estimates of the magnitude of any effects of teen childbearing depending on the outcome, the maternal age groups compared, or whether they focus on black or nonblack children. On balance, their evidence of negative effects is modest in comparison to the tenor of public concern. They even find some evidence of positive effects. Among black children in their sample of 4-14 year olds, those whose mothers were 18 or 19 at their birth performed better in reading and math than those whose mothers had been in their early 20s. This finding is of particular interest given that black mothers in their late teens account for more than half of black and almost one-quarter of all unwed teen mothers.
Moore et al.'s generally modest and conflicting estimates were produced after control for measurable aspects of maternal family background, leaving open the question of their stability in the presence of more comprehensive controls. Evidence on this question is provided by Geronimus, Korenman, and Hillemeier who also study the performance of preschool and elementary school age children in a national sample on several standard tests of cognitive development and achievement by age of mother. To control for preexisting differences between teen and older mothers, they compare the test scores of the children of sisters who experienced their first births at different ages. Once maternal background characteristics are controlled by this method, the differences between children with teen versus older mothers are generally insignificant. Those tests where statistically significant differences are estimated usually favor the children of teen mothers: such assessments include indices of verbal ability and achievement in reading and math. And this general pattern of findings is suggested even when the teen birth group is restricted to young teen mothers (those <18 years old).
Social Costs
I have argued that political statements of deleterious effects of teen childbearing for teen parents, their infants, or children are out of line with the more modest and conflicting social scientific evidence. However, if there were demonstrably large social costs, policy aimed at reducing teen childbearing might still be justifiable . The belief that teen childbearing is very expensive to society clearly underlies political sentiment and action. For example, Senator Specter noted in 1995 that "Society spends an estimated $34 billion on behalf of families in which the first birth occurred when the mother was a teenager. When we look at the problem of low-birthweight babies...they carry scars for a lifetime. Frequently those lifetimes are not very long, but are very expensive to society, costing in the range of $200,000 a child and thousands more each year. It costs society multiple billions of dollars." And a finding stated in PWORA is, "Between 1985 and 1990 the public cost of births to teenage mothers under the aid to families with dependent children program, the food stamp program, and the medicaid program has been estimated at $120,000,000,000, Although large numbers are quoted by politicians and advocacy organizations have announced evidence of large social costs, in truth, such figures are speculative. They are calculated using estimates of the effects of teen childbearing that are almost certainly exaggerated. The most recent example was a summary report released at a White House Press Conference by the Robin Hood Foundation. It claimed enormous social costs resulting from teenage childbearing (ranging from 6.9 to 29 billion dollars per year). Although the report’s conclusions were widely publicized without qualification, the social cost estimates were arrived at by ignoring cautions offerred by many of the investigators whose research formed the basis for the report. In fact, the researchers studying the key question of government outlays for public assistance estimated that if all current teen mothers delayed childbearing “the total expenditures on public assistance would increase slightly.” Moreover, since they estimated that lifetime earnings of teen mothers would decrease if they delayed childbearing, and, hence, so would their contribution to the tax base, they calculated “the net (of taxes) annual outlays by government for cash-assistance and in-kind transfers to these women would actually increase by 35 percent, or $4.0 billion. This increase in net expenditures associated with delaying childbearing would amount to over $1,200 per teen mother.”
The possibility that delaying childbearing by teen mothers might, on balance, increase spending on public assistance was not considered in the debate over welfare reform. Instead, the presumption that preventing teen childbearing would result in substantial cost savings was taken as axiomatic. Yet, a full and careful accounting of social costs remains to be performed. A prudent accounting would include sensitivity analyses based on the range of estimates produced by state-of-the-art evidence for economic and child health or developmental outcomes. Such an accounting would likely result in social cost figures that are far lower than those quoted above to rationalize welfare reform. For example, the reviewed evidence suggests that if current teen mothers postponed childbearing, the social costs paid for income transfers or food stamps might be decreased only marginally -- if they were decreased at all. Meanwhile the formidable expenses associated with low birth-weight babies -- including for medicaid, for funding of public hospitals, and for uncompensated care -- might well increase.
Family Support
The dominant cultural ideal in the United States is for a baby to be born into an intact nuclear family. A critical reason teenage childbearing has been problematized is its perceived link to single parenting. Senator Specter observed that, "It may well be the most important problem as we grapple with teenage pregnancy where we have a family coming into existence without any family structure at all. Sociological and anthropological evidence suggest that categorizing families as either nuclear or nonexistent is overly simplistic, as I discuss below. However, the tendency to do so in the case of teen mothers has been re-enforced by overdrawn popularized images. Because a teen mother is more likely to be unmarried now than in the past, the popularized image of teen parenthood is of an absent father and an immature, problem-prone girl raising her (unwanted) child alone. Given this image, common sense suggests that teen childbearing must jeopardize the health and well-being of children and be costly to society.
These nagging doubts might be assuaged by reconstructing this image to be more consistent with the literature. Teen mothers in the United States tend to be older (two-thirds of teen mothers are 18-19 years old; only 2 percent are under age 15), are more likely to be married (about 40 percent are married and teens constitute less than one-third of all unmarried mothers), or, if not married, to benefit from the support and guidance of others than the popularized image suggests. That is, welfare policy makers -- along with many middle class Americans -- fail to distinguish between the function of providing stability, care, and economic support to children that is most often provided by married couples in the United States from the form of marriage itself.
Within the United States, population-variation in family structure is well-documented, both for the current period and historically. Some scholarly theories to explain this variation take non-nuclear family structures as deficient, prima facie, while others attempt to appraise their strengths, weaknesses, and origins without prejudgement. This latter academic tradition was ignored in the welfare reform debate. Yet, consistent with this tradition, some research suggests it may be more accurate to think of a poor teen mother (at least in African American communities that have been studied the most extensively) as an emerging adult participant in active multigenerational social networks than as a rebellious adolescent set apart from her elders, raising her baby alone.
Ethnographic observation and analyses of survey data provide evidence in African American communities that mothers, fathers (or father surrogates), grandmothers, aunts, uncles, and others from both maternal and paternal sides (regardless of parental marital status) actively contribute to the support and nurturance of the young. This approach to shared childrearing, in effect, pools risk among poor families, providing social insurance against common risks. These include risks of severe income shortfall due to the unpredictability of wages, employment (or welfare benefits); risks of hunger, homelessness, or early adult disability; and risks of physical separation between children and any specific individual adult (for example, inflexible work requirements, incarceration, or premature death). If we concern ourselves with family functioning rather than form, it is possible that in some circumstances greater stability, care, and economic support for children may be realized through family systems and forms of union that are not nuclear and that do not place high value on legal marriage, per se.
This argument may appear to run counter to recent studies suggesting that children raised by single parents fare worse than those raised by two parents. Such studies provide a scientific basis to support policies to enhance the attractiveness of marriage. Yet, there are two general problems with this policy response in the context of teen mothers. First, the empirical findings are based on cross-sectional evidence and do not answer the key question of whether children of single parents would do better if their parents stayed together (if divorced) or married (if unmarried). Second, in the select populations where teen childbearing is common, early parental death or disability may be a major threat to a child's chance of being raised by two parents. Encouraging shotgun weddings is unresponsive to this problem. To the extent that welfare reformers are concerned that welfare rules themselves have served as a disincentive towards marriage, it would be reasonable to expand programs such as AFDC-UP and in-kind benefits for which two parent families qualify or to change restrictions such as the 100 hour rule that may encourage fathers with low-wage jobs to leave their families. Such an approach would remove welfare-driven disincentives toward marriage without penalizing the poor or members of minority groups if they maintain multigenerational extended families. Of course, theoretically, kin network participation and legal marriage are not mutually exclusive. Many teen mothers are married. But even among those who are not legally married, ethnographers have observed in specific poor, African American communities that a father and his kin often comprise an important segment of a mother’s kin network. The father’s kin may remain as a resource to a mother and children even if the father is out of the city, out of the state, or out of the picture. Empirical sociologists also find evidence to question the tendency to infer father absence from the lack of legal marriage. They argue that norms related to marriage and nonmarital fertility are and have been more fluid among African Americans than among U.S. whites. In addition, based on his longitudinal analysis of a national sample of children Frank Mott concludes that “ the concept of a discrete discontinuity between father’s presence and absence, particularly for black children, is singularly inappropriate whether used as an analytical tool or as a policy instrument.”, In addition to having kin support, a typical teen mother may be more mature and experienced in childcare than popular images suggest. As noted, the majority of teen mothers are 18 or 19 years old. Older teenagers in poor communities may be more mature than the typical American teenager. They are more likely to be economic contributors to their families and to have had to grow up fast and assume a range of responsibilities, including responsibilities for babies and young children in their social networks. More generally, the relationship between late teen age and social maturity is variable. Being a teenager and being an adolescent are not always synonymous. For example, although we now assume that teenagers are never developmentally ready to be responsible parents, many of our grandmothers bore their first children in their teen years.
The tendency to categorize poor teens in terms of their adolescent status instead of their socioeconomic group may also be misleading if it conceptually sets the individual teenager (or teenager and her boyfriend) as having impulses opposed to the viewpoints of more mature elders, taking risks and making pregnancy resolution decisions in a vacuum apart from mature guidance. Yet, at least in specific socioeconomically disadvantaged local settings, there is evidence to suggest that teens generally decide how to resolve unplanned pregnancies in collaboration with elders. In some instances, teens are encouraged to become mothers by elders who speak from life experience. In these instances, ethnographers interpret the elders as drawing on their socially-suited knowledge that poor families must rely on each other for caretaking as well as for economic support. As poor families configure themselves, elders may make judgments about which teens are promising economic providers and which ones are better at caretaking. They may recruit teens for whichever aspects of kinwork are currently required. In communities where babies are highly valued, longevity is in question, and the experience of grandparenthood is longed for, elders may also encourage early childbearing. More speculatively, if as the research suggests in extremely disadvantaged communities, early childbearing maximizes the health and well-being of infants, a caring elder might encourage it.
Poor Teens and Life Prospects
Middle-class teens have opportunities stretching well beyond age 18 or 19 to become better educated, better skilled, and thus eventually more employable. If they bore children as teenagers, their childcare obligations might well conflict dramatically with their ability to utilize these opportunities. And, given such opportunities, middle-class youth have every reason to believe that they will be better providers for their children if they delay parenthood. But this is not true for the young women most likely to become teen mothers. Poor youth and minority youth in general, as well as those who have suffered early school failure, in particular, face more restricted educational or labor market opportunities. Poor families do appear to mobilize to support those teenagers believed to possess the skills necessary to overcome chronic barriers to achievement and upward mobility, including discouraging them from engaging in the dating activities of their peers. But other poor teens may have less to gain from postponing childbearing. Meanwhile, to the extent they have access to kin support for childcare the trade- offs between school, work and childcare may be diminished. In targeting policy to address unequal opportunity among American youth, it is likely that social forces other than early childbearing constitute the major impediments to labor market success among poor teen mothers.
Differences in life prospects between poor and better-off teenagers are evident at their most basic and literal level. The most profound inappropriate assumption that leads us to believe that postponing childbearing well beyond the teen years is always in children’s best interests is the assumption that teens from all social classes face a predictable future with death far off in its “logical position...at the close of a long life.” Yet, in the United States, one of the most disturbing expressions of social inequality is the variation across socioeconomic groups in rates of premature adult mortality. For example, Table 2 shows the probability that a 15-year-old male or female will survive to age 45, 55, or 65 for U.S. whites and for African American residents of two poor communities: Harlem and Chicago’s south side. 95 percent of white women, but less than 80 percent of Harlem or Chicago women, survive to their 55th birthday. More than one-third of Harlem or Chicago women die by age 65. For men, the statistics are more grim. Less than three-quarters of Harlem or Chicago men survive to age 45, compared to 94 percent of U.S. whites. Little more than half of Harlem or Chicago men can expect to survive to age 55, and almost two-thirds who reach their 15th birthday will not live to see their 65th. (This represents less than half the probability that the typical white 15-year-old-male nationwide will survive to age 65.) In Harlem and Chicago, 15 year old men have less chance of surviving to age 45 than the typical white 15 year old nationwide has of surviving to age 65. Deaths from chronic diseases, rather than the more publicized instances of homicide, are the primary reason.
Lower, but still strikingly high rates of early adult mortality for women and men have been described for other urban areas such as Watts in Los Angeles or Central City Detroit. In predominantly African American poor urban populations -- some of the same populations where early childbearing is common and engenders the most concern in the general population -- the probability of premature mortality appears sufficiently high that, unlike the average American teenager, teens in these areas cannot confidently expect to survive through or even to middle adulthood.
Without such confidence, to postpone central goals such as childbearing is to risk forgoing them. And if one believes that responsible parenthood includes maximizing the chance that a parent will survive to see and help her child grow up, then insecurity about one’s own longevity would be a serious consideration when contemplating whether to defer parenthood.
When these aspects of the life experience of the poor are ignored and teen mothers are viewed merely as youth, the state may appear justified in moving in, in loco parentis, to regulate the behavior of poor teens, whether through the “tough love” approaches advocated by conservatives or the paternalistically supportive approaches suggested by liberals. As Representative Linda Smith (R-WA) argued on March 23, 1995: “The most compassionate thing we can do for these little kids and their kids is to not give them cash grants, to not go on and reward the wrong decisions, to not reward sometimes their mothers who encourage them in some tenement house to go get pregnant so they can get the welfare that they have learned to live on.” But if compassion is what we aim for, then this approach of constructing all sexually active teenage women as unsupervised adolescents and using that image to justify government policy as surrogate parent is deeply problematic. It pathologizes the role of elders in the decisions of poor teens. It definitionally attributes the same psychosocial developmental trajectories, opportunities, and human resources for helping to rear children (now or in the future) to all teenagers. But, as we have seen, there are marked variations across social groupings in all of these factors. Generalizing from middle class to poor teens may lead to mistaken characterization of the motivations of unmarried, sexually active teens who are poor and of the opportunities and resources dependably available to them for realizing more socially approved goals than early childbearing. It is not only incorrect in its premise that every teen, by definition, is unprepared to be a responsible mother, but also in its implication that every teen, by definition, will be in a better position to be one at a later date.
An Alternative View: The Rationality of Childbearing
The perspective on teen childbearing that helped to inform the PRWORA is not only inconsistent with important empirical evidence, but it is also only one among several theoretical perspectives on the relationship of teenage childbearing to socioeconomic disadvantage. Here I outline one alternative perspective that is consistent with existing empirical evidence, but leads to a different conclusion regarding the rationality or “responsibility” of early fertitlity timing, at least among an important subset of teen mothers. I focus on poor, urban African Americans for two reasons. First, there is more research evidence about this subset of the disadvantaged. Second, insofar as public concern about teen childbearing is now racialized, teenage childbearing is construed as one in a constellation of pathological behaviors particularly engaged in by an urban, African American “underclass.” I argue instead that the generally earlier fertility exhibited in these communities relative to the national average has a quite different meaning. It does not represent the abandonment by teens or their elders of the mainstream “family value” that responsible parents strive to bring children into the world when they are most prepared to provide for their children’s well-being, broadly defined. Instead, it expresses attempts to embrace this view of the appropriate time for childbearing within extremely adverse circumstances that constrain and qualitatively alter the routes available for achieving this goal.
What if a poor, African American teen wants the following: to have two or three healthy children who will continue to thrive and will be provided for materially and emotionally until they reach adulthood. Moreover, she (or he) wants to achieve this result, while minimizing the chance of needing government assistance in order to do so. If this is what she values, what age would she or the elders who advise her think would be the most appropriate to begin childbearing?
What if, as the literature suggests, in pursuing these goals, there are complicating factors? The poor teen might die at any time, might well not survive through middle age, and is likely to suffer from a chronic disease or disability starting at even younger ages. If she delays childbearing, this early health deterioration will jeopardize her chances of bearing normal-weight infants who survive infancy. If she does bear a low birthweight baby, that child is at increased risk of long-term developmental deficits. If she has a “special needs” child, that will be a drain on family resources and may depress the child’s school performance and future economic prospects. The possibility of early death or disability also characterizes the people a poor mother depends on to support and care for her children -- be they her baby’s father or members of a broader social network. And a child with a disabled family member is more likely to exhibit behavior problems. The research findings suggest that a poor mother faces particular uncertainty if she follows the nuclear family model and depends primarily on a husband for economic, practical or emotional support. His chance of early death is distressingly high. Meanwhile, his earnings potential is likely to be low and unreliable.
In fact, whether or not she marries before her first birth, a poor mother must expect to contribute substantially to the financial support of her children. But her chances for labor market attachment are also unreliable. If she finds employment, the wages and benefits she can command may not offset the costs of being a working mother. She cannot expect paid maternity leave; nor is accessible or affordable daycare available that would free her from reliance on kin for childcare once she does return to work. Moreover, she faces the social expectations that she help care for her kin as their health falters. Postponing childbearing increases the chance that her young children compete with ailing elders for her energies and decreases the chance that their father will survive through much of their childhood. Her greatest chance of long-term labor force attachment will be if her children’s pre-school years coincide with her years of peak access to social and practical support provided by relatively healthy kin. Her best chance of achieving her stated goals is by becoming a mother at a young age. Note that, the kin-network approach to childrearing is motivated, among other reasons, by the community-level social expectation that adults work whenever they can, however they can, and that others pitch in to make this possible. Put another way, if the poor place value on economic self-sufficiency through gainful employment then they often must accept jobs that separate them from their children, sometimes for long periods. Low-skill jobs provide income and other practical and psychological rewards, but they may also exacerbate disease risk and the chance of early disability or mortality. Working-poor parents may experience legitimate worries of becoming compromised in their capacity to provide for their children or of leaving them orphaned. If along with their work ethic, parents also value providing consistent emotional and material support for their children, then they must develop, invest in, and rely on informal social capital. By doing so they do their best to insure their children caring adults willing and able to supplement or even substitute for parental support.
In sum, one need not assume the abandonment of personal responsibility to explain persistently high rates of early childbearing in poor communities. Instead, the scenario described here suggests the possibility that poor teens in collaboration with their elders may be making rational and responsible decisions. Instead, the scenario described here suggest the possiblity that poor teens in collaboration with their elders may be making reational and responsible decisions. Instead of being a sign that their values place them apart from themainstream, the behavior of teen mothers can be read as representing mainstream values whose phenotypic manifestations take different forms as suits the environmental contingencies that must be addressed. Of course, it is not being assumed that poor teens or their elders know the precise statistical odds they face of restricted opportunities, early death and disability, or infant mortality. But it is equally unreasonable to assume that their life experiences do not impress some version of these facts upon them.
The perspective that teenage childbearing represents trade-offs made in order to maximize children’s well-being in hard circumstances may also explain why early childbearing has persisted in extremely disadvantaged communities, despite increased access to contraceptive technology and abortion and in the face of very public disapproval of the behavior. It would suggest that it may continue to persist in the face of welfare reform.
Implications for Social Policy
The harshest measures against teen childbearing proposed or adopted by welfare reformers are not supported by a careful review of the social scientific research on the costs and consequences of teen childbearing. The literature is sufficiently equivocal as to render radical policies or the conclusions on which they are based exceedingly premature. Nor does the literature justify the current equation of antipoverty policy with welfare reform. Indeed, the scenario I paint suggests nonmarital teenage childbearing is sometimes a strategy for coping with economic uncertainty emanating from many sources, including, but not limited to, welfare generosity. It can also be a strategy for insuring against the adverse health affects of poverty. To the extent this scenario applies, welfare policies meant to be disincentives to early childbearing would at best play a small role in reducing teen marital or birth rates. If teen birth rates were reduced by welfare policy alone, we might witness some increase in other social problems such as low birthweight rates and their sequelae, at least among African Americans. However, whether or not these policies will reduce teen birth rates is open to question. Despite two decades of active attempts to reduce teenage childbearing in poor communities, there has been little evidence of a successful targeted program. Meanwhile teen childbearing rates have fallen and risen along with secular trends in childbearing obtaining to greater or lesser extent among mothers of all ages. Trends in nonmarital childbearing rates among teens also appear to mimic more general trends in nonmarital childbearing rather than be particularly responsive to expansions or retrenchments in welfare generosity. Nonmarital birth rates did begin rising in the 1960s when AFDC was relatively generous but they continued to rise through the 1970s and 1980s when eligibility requirements became more stringent and the real value of benefits declined.
Like everyone else, poor teens and their elders respond to incentives when making fertility-related decisions. But the most important incentives to them may not be narrow carrot- and-stick incentives of the kind envisioned and manipulated by welfare reformers. Instead, they may derive from larger, more enduring aspects of their life experience. For example, faced with this trade-off, a woman might accept reduced welfare benefits, if by early childbearing she increases the chance that her child will be born healthy and have able-bodied caretakers who survive through his childhood.
I argue that the motivations underlying high rates of teenage childbearing in poor communities may be versions of exactly those values that many now believe the poor, generally, and African American residents of central cities, in particular, lack. These motivations can be understood in light of the problems that together form the larger context in which teen childbearing takes place. These problems merit policy attention. They include problems in educational systems, labor market opportunities, child care, housing, and health that impede the productivity and shorten the lives of young through middle-aged adults in some poor communities, making it difficult for them to escape the anxieties that accompany the profound uncertainties they face. These anxieties, in turn, may move responsible, future-oriented, and caring adults to arrange and invest in elaborate systems of social insurance, which in turn exert pressure towards earlier childbearing and away from marriage.
Through PRWORA, President Clinton and the 104th Congress have set policies in motion that are likely to exacerbate anxiety among the poor, perhaps providing indirect incentives toward early childbearing. Meanwhile, the Congress did little to improve job or educational prospects for the poor and appeared to be set on dismantling the public health service. Increasingly stringent Medicaid and Food Stamps eligibility critieria included in provisions of PRWORA, threaten the health and longevity of the poor and may effectively deny the option of nursing home care to some segments of the poor population. In turn, young adults in these segments may be increasingly called upon to care for ailing elderly. If early childbearing reduces the stress of juggling this obligation with work and child care responsibilities, and if ailing elders long to be grandparents, social expectation in these communities may move in the direction of teenage childbearing.
How did this happen? Certainly, the so-called “Republican Revolution” contributed importantly to the configuration of the recently enacted welfare reform bill. Yet, some liberal advocacy organizations have been among the most industrious in their efforts to maintain a public misperception that was important for its success. The misperception is that irrefutable scientific evidence shows teenage childbearing to be highly costly to poor teens, their children, and society. Liberal advocates may hope that the broad-based social consensus that teen childbearing is destructive can be used to funnel resources to poor women and children that they could not otherwise mobilize. Or they may hope to take advantage of this consensus -- combined with the common failure to distinguish between teenage childbearing and teenage pregnancy -- to protect the reproductive freedom of middle-class teens to avoid conception or childbirth. A full discussion of the merits of such tactics is beyond the scope of this paper. However, there is little evidence that any such strategy has succeeded. The premature consensus that teenage childbearing is a major social ill is now exploited to legislate abstinence education programs, while mandated support for family planning programs and abortion has eroded. And this consensus has been effectively used to help undercut support for the social safety net and other antipoverty programs.
For whatever reasons, liberals have played an active role in perpetuating myths about teen childbearing. These myths, alongside others, have been used to portray the poor as morally marred and have helped justify proposals by radical Republicans to suspend rights and freedoms of poor teens and to direct resources away from their children. They have put Democrats in the position of making proposals and voting for or signing legislation that are now seen as compromises, but which in a different era would have been unthinkable by liberals or moderates. The result is antipoverty policy that is estimated to push 1.1 million children -- and 2.6 million people overall -- into poverty. And PRWORA is an act that is likely to increase material hardship of the most severe forms including homelessness and infant and premature adult mortality. At a time when local governments will have limited resources to address expanding urgent needs, they now must commit some of those resources to developing political, administrative, and operational responses to devolution as well as to activities whose success in reducing teen birth rates are highly speculative, such as conducting programs on the problem of statutory rape.
When the crusade against teen childbearing fails, will that failure be interpreted as evidence that the poor are morally even more far removed from the mainstream than they are already thought to be? Will that failure provide further ammunition for those who argue that public spending on social programs is wasteful? Or will we take the opportunity to learn that the poor continue to suffer from many compelling problems, problems that could be reduced by public intervention? Rather than being a cause of public problems, teenage childbearing may be emblematic of the price paid by the poor themselves -- not by the larger society -- as they work actively to fulfill the values of self-sufficiency, hard work, and responsibility to children and elders in a hostile environment that wears away their health and limits their life chances.
Arline T. Geronimus, Sc.D. is Associate Professor, Department of Health Behavior and Health Education, School of Public Health, and Faculty Research Associate, Population Studies Center, University of Michigan. This work was undertaken while the author was a Visiting Scholar at the Russell Sage Foundation. The author gratefully acknowledges the support of the Russell Sage Foundation, the William T. Grant Foundation, and the Centers for Disease Control and Prevention, grant #U83-CCU51249. I am indebted to Jim Caraley for encouraging me to write this paper. I thank Robert K. Merton, Sylvia Tesh, Alice O'Connor, Sherman James, James Rule, J. Phillip Thompson, Carol Stack, Mary Corcoran, John Bound and two anonymous reviewers for helpful comments on previous drafts; Adam Becker for helpful discussions; Kimberly Giamportone and Marianne Hillemeier for research assistance and help with the production of the manuscript; and Jamie Taylor for editorial assistance. Teenage Childbearing and Personal Responsibility: An Alternative View