About Us
Child Abuse Training
International Activities
Health Policy Collaboration
Rossi Award for Program Evaluation
UMD Capstone Courses
Mailing List
Contact Us

Three P
Paternalism, Teenage Pregnancy Prevention, and Teenage Parent Services

Rebecca A. Maynard

Teenage pregnancy and parenting are among the nation's greatest tragedies because of the burdens they impose on future generations. With the sharp increases in out-of-wedlock childbearing and a persistently high abortion rate, an especially strong link has emerged between teenage parenthood, long-term welfare dependence, and poor outcomes for children. The growth in the rate of out-of-wedlock childbearing, in turn, has paralleled expansions of the federal welfare safety net. At this juncture, there is substantial research to challenge the wisdom of past programs designed to prevent teenage pregnancy or to mitigate its consequences. Those programs were usually voluntary, nondirective, and focused on providing economic support. Their limitations suggest that more paternalistic policies linking public support for teenage parents and their children to specific parental and child behaviors would better serve both these families and society as a whole.

These conclusions are derived from two strands of empirical research. One is studies of child and adolescent development, which identify teenage pregnancy and parenthood as examples of the types of irrational behaviors common among adolescents, particularly those with inadequate external sources of behavioral controls. Teenagers do not intend to get pregnant and begin families at young ages. Moreover, doing so is at odds with their stated values. Yet more than 1 million become pregnant each year and just over half of those getting pregnant give birth. These statistics have worsened in the past few decades despite increased efforts to educate youths about sex and family planning and to increase both the availability and reliability of contraceptives.

The second strand of research has investigated the common features of the few effective pregnancy prevention or parenting policies and programs for teenagers. Those that have shown promise clearly state values, maintain specific behavioral requirements, and impose sanctions for failure to fulfill requirements.

This chapter first discusses the conflict between the values and actions of teenagers for teenage mothers and the implications of the half million births to teenagers each year for their children and for taxpayers. I then discuss public policies aimed at reducing teenage pregnancy and parenthood and the results. The chapter then examines in greater detail the practical lessons from the past two decades of policy and program evaluation research and assesses the current trend toward highly paternalistic social welfare policies. Not only are more paternalistic policies crucial to maintaining public support for social programs such as welfare, but they also offer the most promise for preventing teenage pregnancy and mitigating adverse consequences when it does occur.

Conflicts between Teenagers' Values and Actions

Each year more than a million teenage women in the United States become pregnant, 88 percent of them unintentionally.1 More than 500,000 of them give birth, more than three-fourths of them for the first time. Teenage pregnancy and parenting are, of course, not new phenomena in the United States, but the face of this problem has changed dramatically in the past thirty years. Most notably, births occurring out of wedlock have risen from less than 30 percent of all teenage births in 1950 to more than 70 percent today. Indeed, only 18 percent of all teenage pregnancies end in the formation of traditional two-parent families.2

The increase in the incidence of these births has been accompanied by a steady decline in economic status and a corresponding increase in welfare dependence among teenage parents. More than three-fourths of teenagers giving birth out of wedlock end up on welfare for an average of eight to ten years.3 Most will be living in poverty ten years after their first child is born.

Teenagers are still children. As a group, those who unintentionally get pregnant and begin parenthood at a young age signal their inability to make decisions that are in their own best interests, the best interests of their children, and the best interests of society. Despite the rapid growth of school- and community-based health and education programs focused on adolescent and reproductive health, more teenagers are having sex at younger ages. And despite greater use of contraception and the availability of more effective means of contraception, the incidence of unprotected sex and the failure rates of contraception remain unacceptably high. As a result, the rate of teenage pregnancy remains unacceptably high.

Teenagers' poor judgment has been compounded by social policies that enable those who bear children in their teenage years to continue to make poor decisions. The federal social safety net provided by AFDC, food stamps, and medicaid has traditionally guaranteed teenage mothers basic economic security. The laissez-faire attitudes in this country toward fathers' responsibilities have enabled fathers of children born to teenage mothers, most of whom are not themselves teenagers, to escape responsibility.4

Pregnancies Are Unplanned

The vast majority of the million-plus teenage pregnancies occurring each year are not planned but result from unprotected sex or contraceptive failure. In the words of young mothers who participated in a federal welfare demonstration program: "I didn't plan it, and then again, I kind of knew that it was going to happen because I wasn't like really taking the pills like I was supposed to. I couldn't remember every day to take the pill. And, I still don't." "I really don't want to take time off . . . right now. . . . But, I'm allergic to birth control pills." "My boyfriend thinks [the pills have] something in there killing him."5

At the same time that teenagers are becoming sexually active at increasingly younger ages, they are intentionally getting married at increasingly older ages (figure 3-1). These changing goals no doubt contributed to their greater awareness and use of contraceptives. For example, in 1982 only 48 percent of teenagers used contraceptives during their first intercourse, but by 1988 the rate had risen to 65 percent.6 This increase in contraceptive use had not, however, been sufficient to keep increased sexual activity from raising teenage pregnancy and birthrates.

Figure 3-1.Sexual Activity and Marriage Rates among Teenagers, 1960_93


fig3-1.JPG (6921 bytes)

Source: Alan Gutmacher Institute, Sex and America's Teenagers (New York, 1994).

The most troubling aspect of the unplanned pregnancies is that more than 40 percent end in abortions, which pose both physical and mental health risks to the teenagers and financial burdens for them, their families, and society. In fact, many teenage mothers have had one or more unplanned pregnancies terminated through abortion before having their first baby.7

Dreams and Realities for Teenage Mothers and Their Children

As a group teenage mothers are no less excited about their newborns than are older mothers. And their goals for themselves and their children mirror those of the general population. They want to be good mothers and provide a good life for their children: "I have to get out of here. I can't stand it. No matter where you turn, all you see is this guy and that guy try'n to

sell drugs. I don't want my son to grow up with them. Sometimes little boys let other people influence them, and I don't want him to be growing up thinking that that's something tha' he has to do to make money." They want good jobs for themselves: "I think mothers should work. For one thing, your child gets to learn how to be with other children. And they learn to do little things that maybe you don't have time to teach them at home. So I think she should work, even just to help the child out a little." And they want supportive relationships with male partners, although many reject marriage as a short-term goal because they believe that married men can be too controlling or are not stable providers for themselves and their children. "When you're single it's better. They [the men] treat you so much better when you're not married, you know. . . . When you're single, it's honey this and honey that. When you're married, do this, do that."

Most also are adamant that they want no more children in the foreseeable future. As one said, "I just want to get into school and to work. I really don't want to take time off for no more children right now. I'm not ready for it now. When I have my own place, a full-time job, but not right now." Another added, "It's different when you don't know, when you don't have a kid. . . . I know how hard it is with one. How in the world would you make it with two?" Yet teenage mothers fail miserably in achieving these goals. They, their children, and the fathers of their children all suffer from early parenting decisions—the children most of all.

In general, children of teenage mothers exhibit worse health, greater use of medical care, greater poverty, and poorer school performance than their counterparts born to older mothers. Young parenthood itself is responsible for most of these disadvantages. For example, in 1987, 28 percent of former teenage mothers age 21 to 33 were poor compared with 7 percent of the women who had their first baby at age 20 or older. Only 43 percent of this difference is attributable to the fact that the teenage mothers came from more disadvantaged backgrounds.8

The troubles for children of teenage mothers begin before birth and continue into adulthood. Even after controlling for differences in backgrounds, the offspring of young teenage mothers have a 50 percent higher probability of having a low birthweight and subsequently having poorer health as compared with those born to women who begin parenting in their early twenties.9 The consequences of slightly older childbearing are similar in scope, if somewhat less severe. Despite having more health problems, children of teenage mothers, particularly young teenage mothers, get less medical care than if their mother delayed childbearing a few years.10 They also do much worse in school. They are 50 percent more likely to repeat a grade, perform much worse on standardized tests, and have a much lower rate of school completion.11

In part these poor development and educational results may be related to the children's much higher rates of abuse and neglect. There are 110 reported incidents of abuse and neglect per 1,000 families headed by a woman who had her first child before the age of 18, compared with less than half that number among children born to mothers in their young twenties.12 The results are only marginally better for children whose mothers delay childbearing until 18 or 19 years of age.

Not surprisingly, adolescent childbearing also contributes to the high rate of crime among sons and to the repetition of the cycle of teenage parenting. All else equal, simply being born to a mother younger than 17 rather than 21 increases by 13 percent (from 9.1 to 10.3 percent) the likelihood a male child will end up in prison and increases by 20 percent (from 13 to 16 percent) the probability that a female child will herself become a teenage mother.13

Unquestionably, those teenagers who get pregnant tend to have poor life prospects regardless of whether they become teenage parents.14 Still, the added complications of being a young parent, especially a parent before age 18, further disadvantages these young teenagers in ways that may undermine their efficacy as parents. For example, less than 35 percent of teenagers who begin their families before age 18 ever complete high school, compared with 85 percent of those who delay childbearing. Nearly half of this difference is caused directly by the young teenage parenting. Young teenage mothers also spend 60 percent more time as a single parent, thus shifting a disproportionate share of childrearing responsibility to the mothers and depriving the children of the benefits of two parents in the household.15

As a group the fathers of children born to teenage mothers bear little of the cost of adolescent childbearing compared to the mothers.16 Most are not themselves teenagers when they father children and less than 30 percent of them marry the teenage mothers of their children. The more than 70 percent who are absent fathers pay less than $800 annually for child support. Indeed, only a minority formally acknowledge their fatherhood, and fewer than half of those make regular child support payments.17

Spillover Costs for Taxpayers and Society

Unlike many life choices that primarily affect the individual, teenage pregnancy and childbearing affect the whole of society. Society pays for about 400,000 abortions for teenagers each year. It bears an even larger cost of teenage childbearing. One recent study estimated conservatively that childbearing by women under the age of 18 costs nearly $9 billion dollars annually, $7 billion of which is paid directly by the taxpayers.18 Including the costs of childbearing by 18 and 19 year-olds raises these figures by at least 75 percent. Thus the economic rationale for stronger public policies against teenage pregnancy is clear. And this ignores the intangible gains from minimizing the affront to social values represented by unwed pregnancy and abortion.

Teenage Sex, Pregnancy, and Parenting Behavior

In the past three decades the United States has witnessed trends in teenage behavior that would seem to invite more paternalistic social policies. Before 1970 premarital sex was openly discouraged, and those "found out" by getting pregnant faced social sanctions from their peers, adults in their families, and their communities. At the same time, contraceptives were not readily accessible to teenagers, and those methods available (primarily condoms and foams or jellies) tended to be less effective than many of today's options. Abortions were widely illegal. And public assistance was essentially nonexistent for teenage-parent families. Only 27 percent of teenagers were sexually active before age 18, and only 61 percent were sexually active by age 20.19 Although teenage pregnancy and birthrates were high by historical standards, the vast majority of the pregnancies occurred among 18 and 19 year olds who gave birth rather than having an abortion and who married the fathers of their children.20

In the 1970s there were marked improvements in contraceptive technologies and in the accessibility of contraceptives to young people. Abortions were legalized in 1973, although they were still not easily available in many parts of the country and among lower socioeconomic groups. The rate of teenage sexual activity increased dramatically: most teenagers were becoming sexually active before their eighteenth birthday.21 At the same time, teenagers increased their rate of contraceptive use, and they more than doubled their rate of abortion, from 19 per 1,000 teenage girls to 43 per 1,000 by the end of the decade. As a result, the teenage birthrate approached its all-time low of about 50 per 1,000 women age 15 to 19, and the age at first marriage continued to edge up, reaching 22 by the end of the decade.

But the rate of births to unwed mothers steadily rose: by 1980 they accounted for 48 percent of all teenage births. The majority of these mothers turned to AFDC for support. This dramatic rise resulted in a steady increase in the number of out-of-wedlock teenage births despite the decline in the overall teenage birthrate.

The Policy Context

During the 1960s AFDC, which had primarily served divorced and widowed women, began primarily serving never-married mothers and their children. The 1970s marked the start of a new wave in social services; public assistance was more widely available and social service providers promoted empowerment among poor women, particularly poor single mothers. In the 1980s social services adopted a deficit model for service planning and delivery in which teenagers were seen as helpless clients in need of additional resources. Income and social support programs expanded their eligibility criteria and generosity, many of them assuming the qualities of entitlement programs. Between 1984 and 1991 funding for family planning services increased by 41 percent, outpacing the growth in the number of clients, which increased by only 31 percent.22

The 1980s also witnessed an explosion of programs serving the reproductive health needs of teenagers, often through health clinics located in or near schools, as well as an array of programs to support teenage parents as parents and enable them to continue their education.23 Many large comprehensive high schools instituted special programs for teenage parents, often with on-site day care for the babies. Throughout the country there was also a rapid increase in the number of community programs to meet the social, health, economic, and educational needs of teenage parents and their children. Initially the programs were supported primarily through private funds. But by the end of the 1980s many were incorporated into state and local welfare programs. Indeed, the Family Support Act of 1988 mandated that states serve teenage parents on AFDC who have not completed high school as a priority group in their Job Opportunities and Basic Skills Training (JOBS) programs.24

By the middle of the 1980s teenage pregnancy rates and birthrates had stabilized, but the rate of out-of-wedlock births among teenagers continued to rise and by 1990 had reached 70 percent.25

Currently, the United States has entered a period of public policy that psychologists might characterize as codependency. Policies enable teenagers to continue to bear children and suffer few social or economic sanctions. Despite myriad welfare reform initiatives throughout the country, most states still allow teenage parents to qualify for AFDC with only modest reciprocal demands or none. In the Family Support Act, Congress gave states authority to mandate school attendance or serious employment efforts by teenage parents. Yet eight years later most states still had not acted aggressively on this authority.26 Only a small minority of unwed teenage mothers enforce the establishment paternity and child support payments on the fathers of their babies. In addition, current policies do little to address the high rates of rape and incest that have occurred among pregnant and parenting teenagers, especially those who give birth before age 16.27 Finally, policies do nothing to hold the schools accountable for their lack of success with pregnant and parenting teenagers, even those states such as Florida, Ohio, and Wisconsin that have "learnfare" programs.

Incentives and Welfare

Nevertheless, conservatives' views that the welfare system itself promotes teenage pregnancy and parenthood have at best weak support, at least within the range of plans states implemented before passage of the Personal Responsibility and Work Opportunity Reconciliation Act of 1996.28 There is little connection between trends in real welfare benefits and the trends in teenage pregnancy rates and birthrates. Similarly, there is little correlation between states' teenage pregnancy and birthrates and the generosity of their welfare benefits. Instead, during periods when benefits have eroded, the rates have tended to increase and vice versa.29

Nor are changes in welfare incentives a clear answer to preventing teenage pregnancies. More than twenty states have implemented family caps, which end the practice of increasing welfare benefits if families have additional children. In taking this action states are sending a clear message that it is not acceptable for people to have children they cannot support. States are not, however, explicitly imposing penalties or social sanctions on those who bear additional children while on welfare. The states thus are not being truly paternalistic. Study findings are consistent with the fact that family caps impose only modest financial penalties and social sanctions for having more children while on welfare. The early results from the New Jersey Fed Wed program, begun in 1994, show little or no effect on the fertility of welfare recipients (adults and teenagers together) compared with that of control groups.30

Pregnancy Prevention Efforts

In the past decade many program and policy initiatives have sought to encourage teenagers to delay sexual activity and prevent unplanned pregnancies. Most of the more than $600 million in annual expenditures for fertility control flow through a few national programs: medicaid, the Title X Family Planning program, the Maternal and Child Health Services Block Grant, and the Social Services Block Grant. Hundreds of other programs throughout the country also seek to reduce the incidence of teenage pregnancy. These include programs with a primary goal of promoting abstinence, "safe sex," or both. They also include programs with a much broader focus on the social and educational needs of particular groups of at-risk and disadvantaged youth, including those who have dropped out of school or who are likely to do so and those who live in low-income families and in poor neighborhoods. Finally, the programs include school-based initiatives to provide sex education and adolescent health services.

Although the efforts have been many and scholarly studies are replete with descriptions, the programs have produced few answers as to how to reduce the incidence of teenage childbearing or improve outcomes for teenage mothers and their children. The base for judging the efficacy of various intervention strategies is a relatively small number of demonstration and community service programs. Still, the results suggest that the most successful programs have been those that were most directive or authoritative and sent the clearest messages about right and wrong.

National Programs Providing Family Planning Services

The millions of dollars of federal money spent on family planning and reproductive health has lowered rates of infant and neonatal death, decreased fertility rates among low-income women at high risk of tightly spaced births, and reduced the incidence of low birthweight babies.31 Although there is no evidence that these funds have reduced teenage pregnancy rates, they may reduce the teenage birthrates by increasing the rates of abortion.32 The research also shows that sexually active teenagers are somewhat less likely to use contraceptives at their first intercourse if they live in a community with relatively high welfare benefits, perhaps reflecting the greater likelihood that areas with such benefits also may offer teenagers greater access to abortions.33

Demonstration and Service Programs

The limited research on demonstration and service programs covers strategies ranging from promoting abstinence to providing full-service adolescent health education (with abstinence training) in conjunction with social and health services, sometimes including dispensing of contraceptives and counseling about options (table 3A-1). The most promising prevention programs are clear about which values they promote. They also offer ways to resist peer pressure to engage in sex and teach youths how to use contraceptives effectively after they become sexually active. Simply promoting abstinence does not seem especially effective, at least judging from current research. Neither Project Taking Charge, which combined abstinence education and vocational education, nor the Success Express school- and community-based abstinence initiative successfully promoted abstinence among middle school students.34

Programs providing only sex education, without the focus on abstinence (Teenage Talk, Group Cognitive Behavior Curriculum, and Facts and Feelings, for example) have in some cases increased knowledge of contraceptive options and contraceptive use. But they have not succeeded in delaying the onset of sexual activity.35 But programs that combine values training with sex education have shown some success. For example, the Teenage Services Program based on the Postponing Sexual Involvement model, which combines abstinence training with sex education and contraceptive education, has shown evidence of delaying the onset of t of intercourse or reduce the frequency of intercourse . . . [or whether] school-based or school-linked reproductive health services, either by themselves or in addition to education programs, significantly decrease pregnancy or birthrates."37

Only one of the programs studied, the Self-Center, a full-service health education and counseling program offered through a community-based, school-linked health clinic, showed signs of delaying sexual involvement and improving contraceptive use.38 It also appears that compared with the other program models, this one placed greater emphasis on values and used a more directive approach in helping teenagers make wiser choices.

The Quantum Opportunities program, a mentoring and enrichment program for at-risk high school students that emphasizes values and opportunities associated with staying and doing well in school, appears to have had dramatic effects on the teenage pregnancy rate in one of its four sites, a result that will be reexamined in a larger-scale replication study.39 This successful site also appeared to be the most structured and paternalistic, providing greater levels of contact between program staff and students, as well as the most structured peer and mentor support components. In contrast, programs such as the Summer Training and Education program, which combined family planning education with summer employment in a voluntary and highly supportive setting, promoted improved knowledge about contraception but have not affected pregnancy and birth outcomes among teenagers.40

There is some promise that programs taking a more holistic approach to dealing with the social, economic, and educational needs of teenagers from at-risk families will reduce their pregnancy rates and birthrates. One of the best known, the Children's Aid Society Teenage Pregnancy Primary Prevention program in New York, offers strong reproductive health education and counseling in the context of comprehensively addressing the needs of teenagers from disadvantaged backgrounds. The program not only attempts to deal with the economic and social environments of the youths, but it also devotes considerable effort to building strong social values and decisionmaking skills that emphasize long-range goals. Although a descriptive assessment of program outcomes is encouraging, there has not yet been a rigorous evaluation of its effectiveness in preventing teenage pregnancy and parenting.41

Programs for Teenage Parents

There have been rigorous tests of only a few programs that target teenagers who already are mothers by simultaneously attempting to reduce the incidence of repeat pregnancies, increase the likelihood of high school completion and moving into higher education, and promote employment. After more than a decade of research, observers have discovered no magic formula. None of the eight methodologically sound evaluations of such programs covered in this review points to a model program for future policies.42 But together these studies enrich our understanding of the problems and suggest the potential of directive programs.

These eight programs represent four strategies for addressing the needs of teenage parents, albeit with overlapping features (table 3A-2). Two are youth employment and training programs that served, but did not specifically target, significant numbers of teenage parents.

—JOBSTART. This thirteen-site demonstration of education, vocational training, and support services for disadvantaged young school dropouts operated between 1985 and 1988. During this time it served about 2,000 youths between the ages of 17 and 21. About one-fourth were teenage parents. The program was evaluated using an experimental design.43

—Job Corps. A federally funded program offering intensive education, training, and social services to at-risk 16- to 22-year-old school dropouts from disadvantaged backgrounds. Most Job Corps services are provided in residential centers. Job Corps centers were the focus of the most recent completed evaluation of the program conducted in the late 1970s and early 1980s. Only a small percentage of corps members (but relatively large absolute numbers) were teenage parents. The subsample of young parents for the evaluation consisted of 1,008 young women. The evaluators used a matched comparison sample study design.44

Two programs were comprehensive education and training programs targeted on teenage parents exclusively.

—New Chance. This national demonstration of small-scale, intensive, and comprehensive service programs offered education, training, and social support services for teenage parents on welfare who had dropped out of school. Programs operated in sixteen locations in ten states between 1989 and 1992 offering services to about 1,400 young mothers for up to eighteen months. The evaluation used an experimental design.45

—Project Redirection. This was a four-site demonstration of comprehensive services for parents age 17 or younger. The programs, which served about 300 young mothers between 1980 and 1981, were operated by community-based organizations offering education, training, mentoring, job placement, child care, family planning, and parenting training. The evaluation used a comparison-site design.46

Two of the programs were welfare-based education and employment programs directing mandatory education and job preparation services to teenage parent welfare recipients.

—Ohio Learning, Earning and Parenting (LEAP). This ongoing program is designed to keep teenage parent welfare recipients in school through a system of financial incentives and penalties. Some counties offer special support services to facilitate staying in school, but most provide only minimal case management services. LEAP has been evaluated using an experimental design.47

—Teenage Parent Welfare Demonstration. This was a large-scale field test of a change in welfare programs for first-time teenage parents. The programs required teenagers to participate in self-sufficiency-oriented activities to receive the maximum welfare grant. They also provided them with case management as well as education, training, employment, and social support services. More than 3,000 young mothers received services between 1987 and early 1991. The programs were evaluated using an experimental design.48

Finally, two programs focused on health and were targeted at first-time parents, many of whom were teenagers.

—Teenage Parent Health Care Program. This was an intensive health-focused intervention for mothers younger than 17 and their infants. It provided intensive case management by trained social workers for up to eighteen months after delivery. The program served about 120 mothers and infants in the late 1980s. The demonstration was evaluated using an experimental design.49

—Elmira Nurse Home Visiting Program. This was a demonstration of nurse home visitation services for socially disadvantaged women bearing their first child. The program served 400 women, 47 percent of whom were teenagers. It was evaluated using an experimental design.50

These programs ranged in cost from a few dollars per person per service month (Ohio LEAP) to more than $1,500 per service month (New Chance) (table 3A-2). They not only included programs serving very small portions of the teenage parent population (Job Start, Job Corps, New Chance, Project Redirection, and the two health-focused interventions), but also those serving the universe of teenage mothers going onto welfare (Ohio LEAP and the Teenage Parent Welfare Demonstration).

The Need for Mandatoriness

For years, the prevailing view among policymakers and social service deliverers alike was to prefer voluntary programs. The idea was that public resources should be reserved for the "most motivated," who were also presumed to benefit most. But today the trend is to mandate participation. This shift in emphasis arose in part from the frustration of taxpayers who were increasingly concerned about rising welfare expenditures and in part from the ineffectiveness of unstructured programs. These forces led to the first serious testing of paternalistic policies that establish general expectations: for example, that all welfare recipients should participate in work-oriented effort or that teenage parents who have not completed high school should attend school full-time despite their parenting responsibilities.

The evidence from these trials suggests that levying such expectations is constructive for teenage mothers and their children. First, experience from the two major field tests of mandatory programs for teenage parents, the Teenage Parent Welfare Demonstration and Ohio's LEAP, demonstrates clearly that most teenage parents will not participate voluntarily in any type of education or training program. For example, of the 89 percent of recipients who participated as required in the Teenage Parent Welfare Demonstration, just under half did so after they received their first notice from the welfare department informing them of their obligation to report to the program. Another 42 percent participated after the notification by the welfare department that their grants were about to be cut by 50 percent on average if they did not report. Finally, about 7 percent participated only after their welfare grants had been reduced.51

Similar experience with front-end participation occurred in Ohio's LEAP program, which offered a $62 a month bonus if the recipient complied with school attendance and performance standards, and a $62 a month penalty if she did not (mediocre performance meant no adjustment). Less than 10 percent of the teenagers in Ohio's LEAP program were unaffected by the policy, and more than half of the teenagers received at least one sanction. Indeed, the program's effects were large and consistently significant only in the site that made the greatest use of sanctions (68 percent of all teenagers versus only 56 percent in the other sites).52

In contrast, New Chance, which offered many more services but in which participation was voluntary, had enormous difficulty filling small programs even in areas that had large numbers of teenagers giving birth monthly. Participation patterns among the volunteers in these programs may look similar to those of the Teenage Parent Welfare Demonstration and the other more mandatory programs, with an average participation level of 4 hours a week (300 hours over eighteen months).53 But the mandatory programs achieved this level for a much greater share of those eligible.54

Program Effects

These programs differ widely in the teenage parents they target, program size, intensity and mix of services, setting, and cost. Three larger demonstrations—New Chance, Ohio's LEAP, and the Teenage Parent Welfare Demonstration—allow the examination of variations in effectiveness in the same program across sites and the features that might explain them.

Employment-related outcomes. Employment-related outcomes for teenage-parent populations served by these programs tended to be poor, whether or not the teenagers participated in the programs. Between one-fifth and one-third of participants engaged in any type of job training after enrolling in the programs. Between 20 and 60 percent of the young mothers were employed at some time following their program experience (the higher rates were in sites with longer follow-ups). Average earnings were less than $300 a month and were as low as $76 a month in one demonstration (New Chance), largely because of the low employment rates and part-time nature of the jobs. Not surprisingly, given the alternative of welfare support, wages among those employed tended to average $5 to $6 an hour.

The seven programs for which employment-related outcomes were measured show at best modest results, and the effects varied considerably across programs (table 3-1). In three of the four programs—New Chance, Project Redirection, and the Teenage Parent Welfare Demonstration—that

Table 3-1.Estimated Employment-Related Outcomes of Seven Welfare Programs
Percent of control/comparison group mean unless otherwise specified


Monthly earnings (1995 dollars) Job Employment training
Estimated program impact
Job Start n.a. 0.4 7.5
Job Corps n.a. -40.5 a -44.4 a
New Chance 48.0 a -4.9 -33.4
Ohio Learnfare -9.3 20.0 b 7.6
Project Redirection 12.4 15.1 51.5 b
Teenage Parent
Welfare Demonstration
18.6 a 11.8 a 20.0 a
Elmira Nurse Home
Visiting Program
n.a. 46.8 n.a.

Participant group mean

Job Start n.a. 49.1 260
Job Corps n.a. 40.0 c 448
New Chance 0.3 42.6 85
Ohio Learnfare 17.5 33.2 143
Project Redirection 21.8 60.1 436
Teenage Parent
Welfare Demonstration
26.8 48.2 108
Elmira Nurse
Home Visiting
n.a. 21.0 d n.a.

Sources: See appendix table 3A-1. Data for the Job R

Table 3-2.Estimated Education-Related Outcomes of Seven Welfare Programs
Percent of control/comparison group mean unless otherwise specified


High school enrollment Received diploma/GED   Measured Received basic GED skills
Estimated program impact
Job Start n.a. 57.3a 67.1 a n.a.
Job Corps -80.3a, b 666.7 a n.a. n.a.
New Chance 84.9 a 43.7 a 74.9 a < 0.1
Ohio Learnfare 15.5 a , c 6.6 32.1 n.a.
Project Redirection 16.7 0.1 n.a. n.a.
Teenage Parent
Welfare Demonstration
42.0d 4.0e 19.2 < 0.1
Teenage Parent
Health Care
< -0.1 n.a. n.a. n.a.
Elmira Nurse
Home Visiting
-45.2d n.a. n.a. n.a.

Participant group mean

Job Start n.a. 42.0 39.1 n.a.
Job Corps 6.6 f 6.3 f n.a. n.a.
New Chance 33.0 f 43.1 36.8 7.8
Ohio Learnfare 67.4 34.0 11.1 n.a.
Project Redirection 18.6 48.3 n.a. n.a.
Teenage Parent
Welfare Demonstration
41.6 48.4 5.7 8.1 g
Teenage Parent
Health Care
55.6 n.a. n.a. n.a.
Elmira Nurse
Home Visiting
62.0 d n.a. n.a. n.a.

Sources: See table 3-1.

n.a. Not available.
a. Statistically significant at the 5 percent level.
b. Months 0_6 after enrollment.
c. Continued through tenth month postpartum.

d. Measured in grade equivalent scores.
e. Statistically significant at the 10 percent level.
f. Percent of time.
g. Measured 18 months after enrollment.

pants relative to controls). Only participants in Ohio LEAP had lower rates than those in a control group, presumably because of their younger ages and the emphasis on their attending school.

The gains in job training in Project Redirection and the Teenage Parent Welfare Demonstration were accompanied by net increases in employment and earnings. Ohio LEAP also led to modest (20 percent) increases in employment and, although not statistically significant, the point estimate of the employment increase for the Elmira Nurse Home Visiting Program is large (47 percent over forty-six months). None of the remaining three programs led to increases in employment. Indeed, participation in Job Corps led to large (40 to 44 percent), statistically significant reductions in employment and earnings.55

Education-related outcomes. In general the programs were successful in promoting increased school attendance and even degree attainment. All except the two health-focused ones and Job Corps increased high school enrollment. New Chance increased enrollment by 85 percent, from 18 to 33 percent of possible school weeks over an eighteen-month follow-up period (table 3-2).56

Four of the programs, including Job Corps, significantly increased the percentage who earned a general educational development certificate or a high school diploma, but mainly a GED. Interestingly, the increase in school enrollment among participants in Ohio's Learnfare did not translate into higher rates of school completion or GED attainment three years after enrollment. And the increased enrollment and degree attainment in New Chance and the Teenage Parent Welfare Demonstration did not translate into higher measured basic skills. This finding is consistent with the results of California's welfare JOBS program, Greater Avenues for Independence, in which there was no correspondence between effects on GED attainment and increased measured basic skills. The GAIN results also showed no correspondence between earnings gains and either GED attainment or measured basic skills.57

Fertility outcomes. Rates of repeat pregnancies soon after a first birth are very high among teenage parents, especially those on welfare. Recognizing that additional pregnancies and births would interfere with education and employment goals, all programs except the two focused on youth employment devoted considerable resources to helping the young mothers control their fertility by offering family planning services and counseling as part of the basic intervention.

Still, repeat pregnancy rates were higher than 50 percent in all but one of the programs that measured them. In the Teenage Parent Health Care Demonstration the rate was 28 percent over eighteen months. Pregnancy rates increased significantly among participants in the Job Start and the New Chance demonstrations relative to rates among those in control groups (by 13 and 8 percent, respectively). Only the two health-focused programs were successful in reducing the repeat pregnancy rates (table 3-3). The Teenage Parent Health Care Demonstration reduced the repeat pregnancy rate by an estimated 57 percent, and the Elmira Home Visiting Demonstration reduced the rate by 43 percent. One theory about the greater effectiveness of these programs is that nurses are trained to follow strict service delivery protocols and to be much more direct than welfare caseworkers in their dealings with clients. The nurse home visitors in these programs may simply have been more willing to tell clients to use birth control and to follow up to ensure they were not only using contraceptives but using them correctly.

For New Chance participants the abortion rate also increased sufficiently to offset the higher pregnancy rate, suggesting that the program may have empowered the young mothers to engage in unprotected sex without the threat of an unplanned birth (table 3-3). In contrast, in Project Redirection and the Teenage Parent Welfare Demonstration, the abortion rate declined sufficiently to show increases in the birthrates among program participants, even though the repeat pregnancy rates had not increased significantly. For the Teenage Parent Welfare Demonstration at least, this result may stem from the program's overt policy against abortion counseling. Participants in Ohio LEAP had higher birthrates than control group counterparts, but the difference was not statistically significant.

Table 3-3.Estimated FertilityRelated Outcomes of Seven Welfare Programs
Percent of control/comparison group mean


Pregnancies Abortions Births
Estimated program impact
Job Start 12.7a n.a. 17.1 a
New Chance 7.5 b 34.2 a 8.4
Ohio Learnfare n.a. n.a. 4.3
Project Redirection n.a. 41.5 b 20.0 a
Teenage Parent
Welfare Demonstration
0.1 16.9 6.6 b
Teenage Parent
Health Care
57.1 a n.a. n.a.
Elmira Nurse
Home Visiting
43.1 a n.a. n.a.

Participant group mean

Job Start 76.1 n.a. 67.8
New Chance 57.0 14.9 28.4
Ohio Learnfare n.a. n.a. 26.7
Project Redirection 3.3 0.3 2.4
Teenage Parent
Welfare Demonstration
1.0 0.16 0.64
Teenage Parent
Health Care
28.0 n.a. n.a.
Elmira Nurse
Home Visiting
0.7 n.a. n.a.

Sources: See table 3-1.

n.a. Not available.
a. Statistically significant at the 5 percent level.
b. Statistically significant at the 10 percent level.

Variations in Programs and Effects

One of the most troubling findings from this review of the evidence on program effectiveness is that the results vary significantly among sites participating in the same demonstration, supposedly implementing the same model, and generally having access to similar levels of demonstration or program resources. The pattern of site differences in three of the demonstrations suggests that a key to effectiveness in delaying repeat pregnancies is delivering a consistent message about expectations for the teenagers. The most striking example was in New Chance, where average expenditures for services ranged from $17,000 per participant to $4,400, even though, theoretically, sites had access to similar levels of resources. The impact estimates are even more variable, with little to no correlation between program costs and efficacy. At the other extreme the Teenage Parent Welfare Demonstration had a more consistent and prescriptive model of services, lower cost and cost variance, and generally qualitatively similar results across sites.

New Chance

Schooling results for New Chance participants generally were positive. However, estimates range from no effect in Detroit to more than a 24 percentage point increase in GED or high school diploma attainment in Inglewood, Minneapolis, Pittsburgh, and San Jose. The results of the demonstration are much more mixed, however, if judged by effects on the fertility and earnings of the young mothers. Of the sixteen sites, Denver and Salem increased earnings or reduced the incidence of repeat pregnancies or both. The programs in Allentown, Detroit, and Portland experienced large increases in the rates of repeat pregnancies or significant reductions in earnings or both. San Jose had large reductions in the rate of repeat pregnancies but witnessed large earnings losses.

Explanations for these variations range from the way programs were implemented to how they were targeted. For example, the San Jose program, which increased school completion rates and reduced pregnancy rates but decreased earnings, emphasized education for non_high school graduates in connection with the California welfare JOBS program (GAIN). Participants were enrolled in both New Chance and GAIN.58 As a result, the young mothers had access to welfare services and also may have had stronger incentives to participate in New Chance and the GAIN education services, thus delaying employment.

The Denver program that increased school completion and earnings was tightly linked to the local job training program operating under the Job Training Partnership Act, which may account for its stronger education and employment focus. The JTPA program in Denver also seemed to have fairly strict attendance policies.

Both the Denver and the San Jose programs had more intensive family planning components than most others, which possibly accounts for their greater effectiveness in preventing pregnancy. New Chance general program guidelines called for one group session each month on family planning, but the Denver program scheduled four each month and the San Jose program two.59

Weak implementation and a lack of emphasis on values training are likely explanations for the three sites that performed especially poorly. One of the sites moved a long distance from its client population midway through the demonstration period. In addition, these programs tended to celebrate rather than to sanction pregnancies and births. Many seem to have emphasized teenage empowerment without the complement of strong social values training.

Ohio's LEAP

In most counties the learnfare policy consisted simply of monetary incentives and penalties. In Cleveland and Cleveland East, however, financial incentives were complemented with case management and social support services, much like those offered in the Teenage Parent Welfare Demonstration. All of the positive results of LEAP were concentrated in Cleveland and Cleveland East. None of the sites that simply modified the benefit policy to account for teenage parents' school attendance and performance resulted in measurable changes in behavior. This suggests that the paternalistic actions of case managers, who reinforced the signals sent by the change in benefit policy, were essential to improving the behavior of the teenagers.

The Teenage Parent Welfare Demonstration

Results of the Teenage Parent Welfare Demonstration were somewhat more consistant across sites than were those of New Chance or LEAP. All of the sites showed modest estimated increases in employment and earnings and reductions in welfare benefits.60 None had significant effects on pregnancy rates, and the effects on educational attainment were positive for two sites and negative for the third.

Likely explanations for the greater consistency relative to the New Chance programs are the clear participation requirements and program accountability. The federal project staff held the programs accountable for keeping young mothers in the demonstration actively engaged in some activity pursuant to employment (school, work, job training, or an activity preparatory to one of these). If a young mother chronically failed to meet this obligation, program staff were required to issue a request for a reduction of more than $150 a month in her welfare grant. The staff members were strongly motivated to work with the clients to avoid the imposition of sanctions; the clients were motivated to maintain their participation for similar reasons.61

Program Effects and Levels of Coercion

The Teenage Parent Welfare Demonstration offers a unique look at how applying different levels of enforcement of the program's participation requirements might affect its effectiveness. Activity rates differed somewhat among the demonstration participants who enrolled under varying degrees of coercion. Those coming in voluntarily were active 37 percent of the following two years. Those coming in after being threatened with sanctions were active 32 percent of the time. Those coming in only after receiving a financial sanction were active 29 percent of the time.

Yet from a policy perspective, what is important is whether the program was differentially effective for those who came in more or less voluntarily. The effects of the program were roughly similar for all groups of participants, regardless of the level of coercion needed to get them in (table 3-4). Where there were differences in effects, those who were least likely to come into the program on their own tended to do relatively better than those requiring less coercion.62 Teenagers who came into the program only after being sanctioned had employment rates two years later that were more than twice those of their control group counterparts. Those who came in under less coercion experienced much more modest gains. The program achieved beneficial effects for many more clients than it would have had it been less demanding.

Table 3-4. Estimated Effects of the Teenage Parent Welfare Demonstration, by Level of Encouragement to Enroll in Program
Percent unless otherwise specified
Outcome Level of encouragement
Follow-up with no sanction Routine call-in Sanction
Estimated effects of the new welfare regime
Time active 7.6a 6.4b 6.1c
Employed month 24 6.9a 6.2b 38.1 c
Monthly earningsd 24.6a 25.8e 31.0f
AFDC benefits month 24d -16.0a -16.0e -20.9c
Repeat pregnancy, all sites 5.1 2.5 . . .
Repeat pregnancy, Camden -14.6b 3.0 . . .
Participant group means
Time active 37.1 32.2 28.7
Employed month 24 40.8 35.8 62.5
Monthly earningsd 176 158 131
AFDC benefits month 24d 302 306 312
Repeat pregnancy, all sites 67.1 74.6 . . .
Repeat pregnancy, Camden 77.4 77.5 . . .
Number in sample 2,637 2,258 380

Source: Follow-up surveys administered an average of 28 months after sample intake.

a. Statistically significant at the 10 percent level, two-tailed test.
b. Statistically significant at the 5 percent level, two-tailed test.
c. The difference for the routine callin and sanction groups combined is statistically significant at the 5 percent level.
d. 1995 dollars.
e. The difference is nearly statistically significant at the 10 percent level.
f. The difference for the routine callin and sanction groups combined is statistically significant at the 10 percent level.

A part of the explanation for the comparable success rates across groups is the effect of participation mandates on operational practice. The sanction policy altered significantly the accountability of the case managers to the clients and, consequently, their attention to the clients' needs.63 Case managers were accountable for helping all the young mothers meet their participation requirements. Failing to help meant that they might have to request a grant reduction for the mother. Although at first the caseworkers considered the policy punitive and attempted to ignore it, they quickly came to appreciate its power in gaining access to the mothers and creating clear expectations for them.

The clients had not generally known such consistency in message and expectations in any part of their lives, and they responded well. They reported that they considered the participation requirements fair, particularly since case managers and support services were available to assist them in overcoming real and perceived barriers. But for a large portion of them to comply with the requirements, they first needed to be convinced that the stated consequences of noncooperation were real.

The first time they sent me a letter, I looked at it and threw it away. The second time, I looked at it and threw it away again. And then they cut my check, and I said, "uh, oh, I'd better go." I was like, "Oh, my goodness, these people really mean business. And I'd better go down there and see what this is all about."

At first I didn't go. They used to send me letters and call me. I still wouldn't go. And then they sent this man [a case manager] out to my house. And I was like, I'll go and see what it is all about. Then the first time I went I didn't like it because they would ask me little personal questions. Then after I did that I never came back, and they came out to my house again and called. "Could you please come to the program." And I finally went, and then after that I went and I liked it then. I really liked it then.

Program participation requirements seem, however, to have been generally ineffective in helping to prevent repeat pregnancies and births. The program reduced rates of pregnancies and births in one site, had no effect in a second, and increased rates in the third. In considering the outcomes according to the extent of voluntarism among the participants, the researchers found that in the more successful site the program had large deterrent effects among those who responded on first call-in, but not among those requiring more coercive efforts to enter the program. Yet in the site where pregnancy rates increased among the program group relative to a no-service control group, the adverse effects were concentrated among those who came into the program voluntarily. Program staff suggested the following possible explanation for the perverse results.

Our key to family planning was case manager counseling to reinforce the information presented in the up-front workshops. Case managers generally focused their attention on the "difficult, less compliant" cases. Their emphasis with the more voluntary participants was on building independence and on encouraging the mothers whom they judged to be least likely to have a repeat pregnancy to take charge of their lives and strive for independence. We more or less expected them to be compliant and follow the family planning advice provided in the workshops. We paid much more attention to reinforcing the messages of the workshops with the less cooperative clients and those who had exhibited less compliant behavior in the past.

After learning of these results the program staff realized that family planning needed to be emphasized with all participants. This is consistent with the fact that the more successful site offered a richer, six-week family planning workshop for all clients and had smaller overall caseloads, thus permitting staff to provide relatively intensive case management for all clients, regardless of the extent of voluntarism in participation.

Discussion and Conclusion

Now more than ever, the United States needs to consider much more paternalistic policies to address the problems of teenage pregnancy and parenting. At a political level the policies are essential to maintain public support for services. Taxpayers are simply no longer willing to provide unconditional assistance to the poor and disadvantaged. More important, however, paternalistic policies seem to offer the best hope to deter teenage pregnancy and thereby protect teenagers and their children from increasingly restrictive welfare policies.

The rich array of social support services available to teenage parents in the past is fast disappearing. The future will have more limited economic support, which may not include cash assistance particularly for young teenage mothers. Schools are becoming less willing to provide special services for teenage parents. Health care services are increasingly being offered to poor families through more paternalistic managed care systems. And under the current welfare reform policies, there will be greater competition for the already scarce community services that in the past have helped teenage parents cope.64

The silver lining in the current wave of welfare reforms is that this may be the setting in which, out of necessity, we discover more successful policies for reducing teenage pregnancy. Among the reasons for optimism are, first, there is now scientific evidence that the consequences of teenage pregnancy and childbirth are at odds with the long-run goals of the teenagers themselves and with the short- and long-run welfare of the nation. The teenagers' decisions are based, at best, on calculations of the likely benefits and rewards of sex or childbearing that are too short term. Second, experience with both prevention and social service programs suggests that the more paternalistic programs tend to outperform programs aimed at empowerment. Or, put another way, the clearer a program is about the behavioral response it expects and will reward, the more likely it will succeed. Third, clarity of expectations, boundaries, and consistency are elements of successful policies and also qualities that build social competence and economic independence in adolescents and young adults.65

In designing policies we should keep in mind that marginal changes in economic incentives will have little or no effect on either teenage pregnancy rates or birthrates; that simple programs teaching how babies are made, preaching abstinence, or providing universal access to contraceptives will not appreciably lower the teenage pregnancy rate; that teenage parents most in need of assistance are unlikely to walk into a volunteer program in search of help; and that full-service programs alone will not significantly improve outcomes for them.

The evidence suggests that future policies and programs should emphasize multipronged strategies that are much more paternalistic, strategies that emphasize prevention, establish behavioral expectations not unlike those expected of adult parents, and propound clear consequences for not fulfilling the behavioral expectations.

Table 3A-1.Selected Programs Aimed at Teenage Pregnancy Prevention, by Primary Focus of Intervention


Setting focus and project Study design Duration of intervention Sample size Program effects
Abstinence programs

Project Taking Charge

2-site, school-based (7th graders and their program (randomly selected likelihood of being parents) classrooms) Experimental design 6 week program 91 No change in values or the  sexually active after six months

Success Express

10 site school program (middle schoolers) Quasiexperimental community based design (pretest-    posttest) 6-week program 848 Effects on sexual activity ambiguous

Sex education

Facts and Feelings 2 urban school districts (7th and 8th graders) and 2 rural  
counties in Utah
Experimental design 6-unit video curriculum 548 No effects on sexual intentions or behaviors after 12 months
Teen Talk 7-site community prevention program in Calif. and Texas (13_19-yearolds) Experimental design 8_12 hour curriculum 1,288 No significant effect on teenage pregnancies.   Increased knowledge of contraception options
Group Cognitive Behavior Curriculum Community health center in Minn. (14_18-yearolds) Experimental design 8 two-hour sessions 107 Lower rates of unprotected sex among males. Results decayed over 1-year follow-up for males

Sex education, abstinence, and safe sex

Postponing Sexual Involvement School-based program in Georgia (8th graders) Quasiexperimental design (comparison by 12th grade. schools) 10 classes 536 Decreased pregnancy rates.  Delay in onset of sexual activity
Reducing the Risk 13 high schools (9th and 10th graders) in Calif. Quasiexperimental design (comparison classrooms) 3 weeks, 15  classes 758 No effect on pregnancy rates

Education and family planning services

SelfCenter Program 2 schools and community health clinic (middle and high school 
Quasi-experimental design (comparison in Md. schools) Up to 3 years 3,944 Decrease in pregnancy rates among older teens. Increase in contraceptive use. Delay in becoming sexually active
SchoolBased Health Clinics Schools in 6 states (high school and before and
Comparison schools after comparison of school-level outcomes Ongoing 7,033 No change in pregnancyrates. No change in rates of sexual activity or contraceptive use
St. Paul School Based Health Clinics 6 schools in St. Paul (high school students) Before and after comparison of school clinics level outcomes Ongoing . . . No evidence of effect on pregnancy and birth rates
Employment and education programs with prevention services
Summer training and education 5 communities (14_15-year-olds) Experimental design Summer 4,800 No effects on pregnancy rates. Increased rates of   reported use of contraceptives
Quantum Opportunities Program 4 communities (9th graders) Experimental design 4 years 216 Decrease in pregnancy rates in one of the four cities

Source: Table draws on program summaries in Brown and Eisenberg (1995) and Moore, Sugland, Blumenthal et al. (1995), as well as primary information sources reported in table 3A-3.

Table 3A-2.Selected Programs Serving Teenage Parents
Program Setting Cost (1995 dollars) Study design Sample size
Youth employment and training programs
Job Start 13 community-based organizations, vocational schools, and Job Corps programs $1,073 a month (5.9 months) Random assignment 508a
Job Corps 78 Job Corps centers $1,256 a month Matched comparison sample 1,008a
Comprehensive education and training programs for teenage parents
New Chance 16 community-based organizations, PICS, and schools $1,706 a month (6 months) Random assignment 2,088
Project Redirection 4 community-based organizations $475 a month Comparison sites 277
Welfare-based education and training programs
Ohio LEAP County welfare program $47.00 per eligibility month Random assignment 4,225
Teen Parent Welfare Demonstration Welfare offices in 3 cities $244 a month (9 months a year) Random assignment 5,297
Health-focused programs
Teenage Parent Health Care Program Urban community health center n.a. (probably high) Random assignment 243
Elmira Nurse Home Visiting Program Rural community health center n.a. (probably high) Random assignment 51a

Source: See table 3A-3. Data for the Job Start evaluation pertain to four years after sample enrollment; for Job Corps to four years after enrollment; for New Chance to 18 months after enrollment; for Project Redirection to five years after enrollment; for Ohio Learnfare to three years after enrollment; for the Teen Parent Welfare Demonstration to two years after enrollment; for the Teen Parent Health Care Demonstration to 18 months after enrollment; and for the Elmira Nurse Home Visiting Demonstration to 46 months after enrollment.

n.a. Not available.
a. Teenage mother subsample.

Table 3A-3.Sources of Evaluation Data, Impact Estimates, and Costs

Program Sources

Prevention initiatives

Project Taking Charge S. R. Jorgensen, V. Potts, and B. Camp, "Project Taking Charge: SixMonth Followup of a Pregnancy Prevention Program for Early Adolescents," Family Relations, vol. 42 (October 1993), pp. 401_06

Success Express F. S. Christopher and M. Roosa, "An Evaluation of an Adolescent Pregnancy Prevention Program: Is `Just Say No' Enough?" Family Relations, vol. 39 (January 1990), pp. 68_72

Facts and Feelings B. C. Miller and others, eds., Preventing Adolescent Pregnancy: Model Programs and Evaluations (Newbury Park, Calif.: Sage Publications, 1992)

B. C. Miller and others, "Impact Evaluation of Facts and Feelings: A Home-Based Video Sex Education," Family Relations, vol. 42 (October 1993), pp. 392_400

Teen Talk M. Eisen and G. Zellman, "A Health Beliefs Field Experiment," in B. Miller and others, eds., Preventing Adolescent Pregnancy (Newbury Park, Calif.: Sage Publications, 1992)

M. G. Eisen, M. G. Zellman, and A. McAlister, "A Health Belief Model—Social Learning Theory Approach to Adolescents' Fertility Control: Findings from a Controlled Field Trial," Health Education Quarterly, vol. 19, no. 2 (1992), pp. 249_62

Group Cognitive L. D. Gilchrist and S. Schinke, "Coping with Behavior Curriculum Contraception: Cognitive and Behavioral Methods with Adolescents," Cognitive Therapy and Research, vol. 7 (October 1983), pp. 379_88

Postponing Sexual M. Howard, " Delaying the Start of Intercourse Involvement among Adolescents," Adolescent Medicine: State of the Art Reviews, vol. 3, no. 2 (1992), pp. 181_93

M. Howard and J. B. McCabe, "Helping Teenagers Postpone Sexual Involvement," Family Planning Perspectives, vol. 22 (January 1990), pp. 21_26

Program Sources

Reducing the Risk R. P. Barth and others, "Enhancing Social and Cognitive Skills," in Miller and others, eds.,  Preventing Adolescent Pregnancy

D. Kirby and others, "Reducing the Risk: Impact of a New Curriculum on Sexual Risk Taking," Family Planning Perspectives, vol. 23 (1991), pp. 253_63

SelfCenter L. Zabin, "Addressing Adolescent Sexual Behavior and Childbearing: Self-Esteem or Social Change?" Women's Health Issues, vol. 4, no. 2 (1994), pp. 92_97

L. Zabin, "SchoolLinked Reproductive Health Services: The Johns Hopkins Program," in Miller and others, eds., Preventing Adolescent Pregnancy 

L. Zabin and others, "The Baltimore Pregnancy Prevention Program for Urban Teenagers: How Did It Work?" Family Planning Perspectives, vol. 20, no. 4 (1988), pp. 182_87

Girls Incorporated L. T. Postrado and H. Nicholson, "Effectiveness in Delaying the Initiation of Sexual Intercourse of Girls Aged 12_14: Two Components of the Girls Incorporated Preventing Adolescent Pregnancy Program," Youth and Society, vol. 23 (March 1992), pp. 356_79

St. Paul school-based D. Kirby and others, "The Effects of School Based health clinics Health Clinics in St. Paul on Schoolwide Birth-rates," Family Planning Perspectives, vol. 25, no. 1 (1993), pp. 12_16

School-based health D. Kirby, C. Waszak, and J. Ziegler, "Six School-clinics Based Clinics: Their Reproductive Health Services and Impact on Sexual Behavior," Family Planning Perspectives, vol. 23 (1991), pp. 6_16

Summer Training and G. Walker and F. VilellaVelez, "Anatomy of a Education Program Demonstration: The Summer Training and Education Program (STEP) from Pilot through Replication and Postprogram Impacts," Philadelphia: Public/Private Ventures, 1992

Program Sources

Quantum Opportunities A. Hahn and others, "The Quantum Opportunities Program Demonstration," Brandeis University, 1994

Programs serving teenage parents

Job Start G. Cave and others, "JOB START: Final Report on a Program for School Dropouts," New York: Manpower Demonstration Research Corp., 1993

Job Corps C. Mallar and others, "Evaluation of the Economic Impact of the Job Corps Program: Third Follow-up Report," Princeton, N.J.: Mathematica Policy Research, 1982

Ohio LEAP D. Long and others, "LEAP: Three-Year Impacts of Ohio's Welfare Initiative to Improve School Attendance among Teenage Parents: Ohio's Learning, Earning, and Parenting Program, " New York: Manpower Demonstration Research Corp., 1996

D. Bloom and others, "LEAP: Interim Findings on a Welfare Initiative to Improve School Attendance among Teenage Parents: Ohio's Learning, Earning, and Parenting Program," New York: Manpower Demonstration Research Corp., 1993

New Chance J. Quint and others, "New Chance: Interim Findings on a Comprehensive Program for Disadvantaged
Young Mothers and Their Children," New York: Manpower Demonstration Research Corp., 1994

Project Redirection D. Polit and C. White, "The Lives of Young, Disadvantaged Mothers: The Five-Year Follow-up of the
Project Redirection Sample," Saratoga Springs, N.Y.: Humanalysis, 1988

Teenage Parent Welfare R. Maynard, ed., "Building Self-Sufficiency among   Demonstration Welfare-Dependent Teenage Parents," Princeton, N.J.: Mathematica Policy Research, 1993

R. Maynard, W. Nicholson, and A. Rangarajan, "Breaking the Cycle of Poverty: The Effectiveness of Mandatory Services for Welfare-Dependent Teenage Parents," Princeton, N.J.: Mathematica Policy Research, 1993

Program Sources

Teenage Parent Welfare A. Hershey and M. Silverberg, "Program Cost of the Demonstration Teenage Parent Demonstration," Princeton, N.J.: (continued) Mathematica Policy Research, 1993

R. Maynard and A. Rangarajan, "Contraceptive Use and Repeat Pregnancies among WelfareDependent Teenage Mothers," Family Planning Perspectives, vol. 26 (September-October 1994), pp. 198_205

Teenage Parent Health A. O'Sullivan and B. Jacobsen, "A Randomized Trial   Care Program of a Health Care Program for First-Time Adolescent Mothers and Their Infants," Nursing Research, vol. 41, no. 4 (1992), pp. 210_15

Elmira Nurse Home D. Olds and others, "Improving the Life-Course Visiting Program Development of Socially Disadvantaged Mothers: A Randomized Trial of Nurse Home Visitation," American Journal of Public Health, vol. 78 (November 1988), pp. 1436_45


1. Kristen Moore, Facts at a Glance (Washington: Child Trends, 1996); and U.S. Bureau of the Census, "Fertility of American Women: June 1994," Current Population Reports, Series P-20, no. 482 (Department of Commerce, 1995), table 4.

2. Susan McElroy and Kristen Moore, "Trends Over Time in Teenage Childbreating," in Rebecca A. Maynard, ed., Kids Having Kids: Economic Costs and Social Consequences of Teen Pregnancy (Washington: Urban Institue Press, 1997), pp. 40_45.

3. David T. Ellwood, Poor Support: Poverty in the American Family (Basic Books, 1988); and Jon Jacobson and Rebecca A. Maynard, "Unwed Mothers and Long-term Welfare Dependency," in Addressing Illegitimacy: Welfare Reform Options for Congress (Washington: American Enterprise Institute, 1993).

4. Only 40 percent of the fathers of children born to teenage mothers are themselves teenagers. Twenty six percent are six or more years older than the mothers. Alan Guttmacher Institute, Sex and America's Teenagers (New York, 1994), figures 41 and 42.

5. Unless otherwise noted, all quotations in this chapter are from the teenage parents who participated in the Teenage Parent Welfare Demonstration sponsored by the Department of Health and Human Services in the late 1980s and evaluated by Mathematica Policy Research, Inc. Denise Polit, Barriers to Self-Sufficiency and Avenues to Success among Teenage Mothers (Princeton, N.J.: Mathematica Policy Research, 1992); and Rebecca A. Maynard, prepared statement, Teen Parents and Welfare Reform, Hearing before the Senate Committee on Finance, 104 Cong. 1 sess. (Government Printing Office, 1995).

6. This reflects primarily an increase in condom use. Alan Guttmacher Institute, Sex and America's Teenagers, figure 27.

7. Philip Gleason and others, "Service Needs and Use of Welfare-Dependent Teenage Parents," report PR93-17, Mathematica Policy Research, Princeton, N.J., February, 1993; and Janet Quint and others, New Chance: Interim Findings on a Comprehensive Program for Disadvantaged Young Mothers and Their Children (New York: Manpower Demonstration Research Corp., 1994).

8. Alan Guttmacher Institute, Sex and America's Teenagers, figure 52.

9. Kristin A. Moore, Donna R. Morrison, and Angela D. Greene, "Effects on the Children Born to Adolescent Mothers," in Maynard, ed., Kids Having Kids, pp. 145_80.

10. Barbara Wolfe and Maria Perozek, "Teen Children's Health and Health Care Use," in Maynard, ed., Kids Having Kids, pp. 181_203.

11. Moore, Morrison, and Greene, "Effects on the Children."

12. Robert M. Goerge and Bong Joo Lee, "Abuse and Neglect of the Children," in Maynard, ed., Kids Having Kids, p. 211.

13. Jeffrey Grogger, "Incarceration-Related Costs of Early Childbearing," in Maynard, ed., Kids Having Kids, pp. 240_44; and Robert H. Haveman, Barbara Wolfe, and Elaine Peterson, "Children of Early Childbearers as Young Adults," in Kids Having Kids, chap. 9 (figures in text derive from weighed estimates in table 9.6, panel 1).

14. Arline Geronimus, "Mothers of Invention," Nation, August 12, 1996, pp. 6_7.; and V. Joseph Hotz, Susan W. McElroy, and Seth G. Sanders, "The Impact of Teenage Childbearing on the Mothers and the Consequences of Those Impacts for Government," in Maynard, ed., Kids Having Kids, pp. 55_94.

15. These estimates of direct effects are based on a methodology that controls for possible selection effects better than has been possible in most previous studies. The authors used teenagers who miscarried and so had a forced (somewhat random) delay in the timing of the first birth as the control group. Hotz, McElroy, and Sanders, "Impacts of Teenage Childbearing," pp. 58_61.

16. Michael J. Brien and Robert J. Willis, "Costs and Consequences for the Fathers," in Maynard, ed., Kids Having Kids, pp. 95_143.

17. Rebecca A. Maynard, "The Costs of Adolescent Childbearing," in Maynard, ed., Kids Having Kids, pp. 328_29, 336; Congressional Budget Office, Sources of Support for Adolescent Mothers (1990); and Rebecca A. Maynard, Anu Rangarajan, and Reuben Snipper, "To Sanction or Not: Are We Shortchanging Welfare Recipients through Laissez-Faire Attitudes toward Participation in Employment-Related Activities?" paper prepared for the Association for Public Policy Analysis and Management meeting, Washington, October 1993.

18. These are lower-bound estimates that take account of only the costs that are directly attributable to early childbearing. They measure the costs relative to delaying childbearing until age 20 or 21, still lower than the average age at first birth, and they do not include all costs. For example, these estimates include the higher health care costs for the children, but not for the parents; the foster care costs associated with higher placement rates, but not the protective services devoted to family preservation or reunification; and the construction and operation costs of prisons associated with the higher incarceration rates of the older male children of teenage parents, but not the other costs associated with elevated crime rates. So too, they omit the higher educational costs resulting from the greater incidence of health and developmental problems among children born to teenage parents, and such social costs as those associated with the high rates of poverty, single parenthood, and school failure. See Maynard, "Cost of Adolescent Childbearing."

19. Alan Guttmacher Institute, Sex and America's Teenagers, figure 12.

20. Susan W. McElroy and Kristin A. Moore, "Trends over Time in Teenage Pregnancy and Childbearing: The Critical Changes," in Maynard, ed., Kids Having Kids, figures 2.9 and 2.11.

21. Alan Guttmacher Institute, Sex and America's Teenagers, figure 10.

22. Leighton Ku, Freya L. Sonenstein, and Joseph H. Pleck, "Neighborhood, Family, and Work: Influences on the Premarital Behaviors of Adolescent Males," Social Forces, vol. 72 (December 1993), pp. 479_503; and Sarah S. Brown and Leon Eisenberg, eds., The Best Intentions: Unintended Pregnancy and the Well-Being of Children and Families (Washington: National Academy Press, 1995), p. 141.

23. Joy G. Dryfoos, Adolescents at Risk: Prevalence and Prevention (Oxford University Press, 1990).

24. Sheila Smith, ed., Two Generation Programs for Families in Poverty: A New Intervention Strategy, Advances in Applied Developmental Psychology, vol. 9 (Norwood, N.J.: Ablex Publishing, 1995).

25. McElroy and Moore, "Trends over Time," figure 2.3.

26. Jan L. Hagen and Irene Lurie, "Implementing JOBS: Progress and Promise," Nelson A. Rockefeller Institute of Government, Albany, 1994.

27. Debra Boyer and David Fine, "Sexual Abuse as a Factor in Adolescent Pregnancy and Maltreatment," Family Planning Perspectives, vol. 24 (January_February 1992), pp. 4_19; and Peggy Roper and Gregory Weeks, "Child Abuse, Teenageage Pregnancy, and Welfare Dependency: Is There a Link?" Washington State Institute for Public Policy, Evergreen State College, October 1993.

28. Robert Haveman and Barbara Wolfe, Succeeding Generations: On the Effects of Investments in Children (Russell Sage Foundation, 1994); Greg J. Duncan and Saul D. Hoffman, "Welfare Benefits, Economic Opportunities, and Out-of-Wedlock Births among Black Teenage Girls," Demography, vol. 27 (November 1990), pp. 519_35; Haveman, Wolfe, and Peterson "Children of Early Childbearers as Young Adults"; Shelly Lundberg and Robert D. Plotnik, "Adolescent and Premarital Childbearing: Do Economic Incentives Matter?" Journal of Labor Economics, vol. 13 (April 1995), pp. 177_200; and Shelly Lundberg and Robert D. Plotnick, "Effects of State Welfare, Abortion and Family Planning Policies on Premarital Childbearing among White Adolescents," Family Planning Perspectives, vol. 22 (November-December 1990), pp. 246_51.

29. Maynard, prepared statement, Teen Parents and Welfare Reform, Hearing; and Kristin Anderson Moore, Donna R. Morrison, and Dana A. Glei, "Welfare and Adolescent Sex: The Effects of Family History, Benefit Levels, and Community Context," Journal of Family and Economic Lives, vol. 16 (Fall 1995).

30. This finding is consistent with research showing that there is no relationship between the proportions of female teenagers who are sexually active and state welfare benefit levels relative to average communitywide incomes. Michael J. Camasso and others, "New Jersey's Family Cap Experiment," paper prepared for the Conference on Addressing Illegitimacy, American Enterprise Institute, Washington,

Pregnancy Prevention Program," Family Relations, vol. 40 (October 1991), pp. 373_80; Jorgensen, "Project Taking Charge: Six-Month Follow-up of a Pregnancy Prevention Program for Early Adolescents," Family Relations, vol. 42 (October 1993), pp. 401_06; Mark W. Roosa and F. Scott Christopher, "A Response to Thiel and McBride: Scientific Criticism or Obscurantism?" Family Relations, vol. 41 (October 1992), pp. 468_69; Roosa and Christopher, "Evaluation of an Abstinence-Only Adolescent Pregnancy Prevention Program: A Replication," Family Relations, vol. 39 (October 1990), pp. 363_67; and Karen S. Thiel and Dennis McBride, "Comments on an Evaluation of an Abstinence-Only Adolescent Pregnancy Program," Family Relations, vol. 41 (October 1992), pp. 465_67.

35. Brent C. Miller and others, "Pregnancy Prevention Programs, Impact Evaluation of Facts and Feelings: A Home-Based Video Sex Education Curriculum," Family Relations, vol. 42 (October 1993), pp. 392_400.; Marvin Eisen, Gail L. Zellman, and Alfred L. McAlister, "A Health Belief Model—Social Planning Perspectives, vol. 20 (July-August 1988), pp. 188_92; Zabin and others, "The Baltimore Pregnancy Prevention Program for Urban Teenagers: How Did It Work?" Family Planning Perspectives, vol. 20 (July-August 1988), pp. 182_87; and Zabin and others, "Dependency in Urban Black Families Following the Birth of an Adolescent's Child," Journal of Marriage and the Family, vol. 54 (August 1992), pp. 496_507.

39. Andrew Hahn with Tom Leavitt and Paul Aaron, Evaluation of the Quantum Opportunities Program (QOP): Did the Program Work? A Report on the Post Secondary Outcomes and Cost-effectiveness of the QOP Program (1989_1993) (Brandeis University, 1994).

40. Gary Walker and Frances Vilella-Velez, ng Teenage Parents," Manpower Demonstration Research Corp., New York, April 1996.

48. Rebecca A. Maynard, ed., Building Self-Sufficiency among Welfare-Dependent Teenage Parents: Lessons from the Teenage Parent Demonstration (Princeton, N.J.: Mathematica Policy Research, 1993).

49. Ann O'Sullivan and Barbara Jacobsen, "A Randomization Trial of a Health Care Program for First-time Adolescent Mothers and their Infants," Nursing Research, vol. 41 (July-August 1992), pp. 210_15.

50. David L. Olds and others, "Improving the Life-Course Development of Socially Disadvantaged Mothers: A Randomized Trial of Nurse Home Visitation," American Journal of Public Health, vol. 78 (November 1988), pp. 1436_45.

51. Eleven percent of the teenagers required to participate in the program never did so, primarily because they had other means of support and so left welfare rather than participate.

52. Long and others, "LEAP: Three-Year Impacts," pp. 26_30.

53. Quint, New Chance: Interim Findings, pp. 58_61.

54. Most of the New Chance programs had difficulty filling their modest-size programs (generally 100 teenage parents a year). In its Chicago site, for example, New Chance had difficulty enrolling 100 teenage parents annually in a catchment area where, each month, more than 150 teenagers have their first child and go onto welfare.

55. Some have speculated that these decreases in employment and earnings may stem from the residential nature of the Job Corps, which takes young mothers out of their communities during the service period. However, the program did succeed in increasing the earnings of men in the residential programs. So it seems likely that this is not the explanation.

56. Participation in and completion of education programs varied widely among teenage mothers in the eight programs. Very few Job Corps participants were in regular high schools because the program did not offer this option. But a large proportion of those in Ohio Learnfare were in school, presumably because there were financial penalties of $62 or more a month for nonenrollment.

57. One possible explanation for the low correspondence of GED attainment and improvement of basic skills is that the basic skills tests are too unreliable at the low ends of the performance distribution to pick up gains that may have occurred. Elena Cohen and others, "Welfare Reform and Literacy: Are We Making the Connection?" background briefing report for the Family Impact Seminar, Washington; and National Center on Adult Literacy, University of Pennsylvania, June 1994.

58. Quint, Fink, and Rowser, "New Chance: Implementing a Comprehensive Program."

59. Ibid., table 6.2.

60. The estimated effects on earnings are statistically significant only for the Chicago site. Those for welfare and school enrollment are significant for all sites. However, the effects on educational attainment are significant only in the two sites where the estimates were positive. Maynard, Rangarajan, and Snipper, "To Sanction or Not."

61. Ibid.

62. Ibid.

63. Alan M. Hershey and Rebecca A. Maynard, "Designing and Implementing Services for Welfare Dependent Teenage Parents: Lessons from the DHHS/OFA-Sponsored Teenage Parent Demonstration," Education, Training and Service Programs for Disadvantaged Teens, Hearing before the Subcommittee on Human Resources of the House Committee on Ways and Means, March 6, 1992; and Maynard, ed., Building Self-Sufficiency.

64. Dryfoos, Adolescents at Risk; and Congressional Budget Office, Sources of Support for Adolescent Mothers.

65. Marilyn Benoit, "Instrinsic Psychological Issues in Teenage Pregnancy," paper prepared for the Seminar on Programs for Unwed Teen Mothers, American Enterprise Institute, Washington, 1994; Judith S. Musick, "The Psychological and Developmental Dimensions of Adolescent Pregnancy and Parenting: An Interventionist's Perspective," Rockefeller Foundation, New York, December 1987; and S. Shirley Feldman and Glen R. Elliott, At the Threshold: The Developing Adolescent (Harvard University Press, 1990).

Back to top