Teenage Parents and Welfare Reform: Findings from a Survey of Teenagers Affected by Living Requirements.

Collins, Mary Elizabeth ; Stevens, Joyce West ; Lane, Terry S. 等

An important component of welfare reform aimed at minor teenage parents is a requirement that the teenager live in an appropriate living situation, typically with a parent or guardian. The State of Massachusetts developed Teen Living Programs as an alternate living arrangement for minor teenage parents receiving welfare who are unable to live with family members. This article reports on a survey of 199 teenage parents who lived in the Teen Living Programs. Three research questions were asked: Who are the teenagers served by these programs? What services were provided? To what extent have teenagers attained key outcomes? Outcomes measured included educational attainment, employment, welfare status, homelessness, and subsequent pregnancy. Implications for social policy, further program development, and clinical intervention are discussed.

Key words: evaluation; teenage parents; welfare reform

Recent welfare reforms have resulted in substantial and fundamental changes at the federal and state levels. At the federal level the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 (P.L. 104-193) went into effect October 1, 1996, replacing Aid to Families with Dependent Children (AFDC) with Temporary Assistance for Needy Families (TANF). Because of growing concerns about the prolonged use of welfare by teenage mothers, several provisions of the legislation targeted teenage parents. Specifically, to receive assistance a minor teenage parent is now required to live in the home of a parent, adult relative, or guardian and to pursue a high school diploma or GED.

Although the new legislation allows states greater flexibility in designing programs, through the use of federal waivers several states had begun to experiment with welfare changes before the passage of the federal legislation. In Massachusetts, welfare reform was signed into law in February 1995. Compared with federal policy, Massachusetts state policy included a shorter time limit on benefits (two years rather than five) but less demanding work requirements (including exemptions for most parents of preschool children or those caring for a disabled family member). With regard to teenage parents, Massachusetts implemented living arrangement and school requirements consistent with the federal legislation. Unique among states, Massachusetts developed, funded, and implemented Teen Living Programs (TLPs) for teenage parent welfare recipients who are unable to meet the living requirement because of parental "abuse, neglect, addiction, or other extraordinary circumstances" (Massachusetts Department of Social Servi ces [MDSS], 1996, p. II-2). The TLPs are designed to "enable teen parents to develop, in a safe and supportive setting, the requisite skills and knowledge to be competent parents and to lead independent and productive lives after the completion of the program" (MDSS, 1996, p. II-2).

There are several reasons for this special intervention for teenage parents. Like most reforms, the living requirement may reduce program costs either by removing an incentive for teenagers to enroll on welfare or by sanctioning teenagers who do not comply with the requirement. In addition to cost savings, however, the living requirement might benefit young mothers and their children if they gain needed support in either a parents' or guardians' home or in a TLP. Furthermore, the provision of the TLP alternative recognized that not all young mothers have a safe home environment. As an intervention, the TLPs demonstrate a recognition of the complexity of young lives and the often substantial needs of young mothers. Thus, the program is designed to be comprehensive in terms of programming and length. Administered by the Department of Transitional Assistance (DTA) and MDSS, the program emphasizes the development of parenting skills and the protection of young parent and child from maltreatment in addition to th e goal of self-sufficiency. Because of concern that some teenagers would not abide by the living requirement, MDSS also implemented an Outreach Program to locate, assess, and provide services to teenage parents whose TAFDC [Temporary Assistance to Families with Dependent Children] cases have been closed or denied.

The profession of social work emphasizes the importance of evaluating public policies that affect vulnerable populations. Videka-Sherman and Viggiani (1996) called attention to the need to examine the effects of welfare reform, in part through documenting state level variation. Because of its unique and comprehensive approach to the complex issues of teenage parents on welfare, the Massachusetts policy response requires examination. This article reports the findings from a survey of teenage parents who lived in the TLPs. It answers three research questions: (1) Who are the teenagers served by these programs? (2) What services do teenagers receive? (3) To what extent have teenagers attained key outcomes after leaving the programs?

Background

Teenage parents have received special policy attention for several reasons; they are at particularly high risk of long-term welfare dependence and whether on welfare or off, they are likely to be poor, and thus, they and their children are at risk of the negative consequences of poverty. Research has found that 55 percent of all AFDC mothers were teenagers at the time of their first birth, and 44 percent of AFDC mothers were unmarried teenagers (Wertheimer & Moore, 1998). Teenage mothers also tend to have long spells on welfare and to be among the poorest of those on welfare (U.S. General Accounting Office [GAO], 1994). Furthermore, teenage parents who have not completed high school are at particularly high risk of remaining on public assistance (Duncan, Harris, & Boisjoly, 1997). Because of their longer spells, teenage parents also receive a disproportionate amount of welfare funding over time (Alan Guttmacher Institute, 1994).

Because families on welfare have incomes below the poverty line, long-term welfare dependence raises additional concerns related to the profound risks for the child's health and development resulting from chronic poverty. Substantial developmental research has found child poverty to be a risk factor for several negative outcomes, in the areas of physical health, cognitive ability, school achievement, and emotional and behavioral outcomes (Brooks-Gunn & Duncan, 1997). Poverty also has a strong relationship with child maltreatment (Sedlak & Broadhurst, 1996).

Several programmatic efforts have been developed to increase the self-sufficiency of teenage mothers. Historically, such programs (for example, New Chance, Teenage Parent Welfare Demonstration, Ohio's LEAP program) have aimed to support the young women in continuing school or preparing for employment. For example, the Teenage Parent Welfare Demonstration was a mandatory program for first-time teenage parents on welfare that included case management, preparatory job activity, workshops to enhance personal skills and prepare for education and employment, GED courses, child care, and transportation subsidies (Maynard, 1995). New Chance was a voluntary program that included educational instruction, GED preparation, skills training, job placement, child care, access to health care, health and family planning classes, and parenting and life skills classes (Manpower Demonstration Research Corporation, 1997).

A 1995 GAO report conducted a synthesis of rigorous evaluations of five such programs and found that three--LEAP, New Chance, and Jobstart--increased high school or GED completions. However, long-term increases in earnings and decreases in repeat pregnancies have not been demonstrated (Maynard, 1995). Among the conclusions of the literature regarding teenage child bearing and welfare, Maynard (1995) stated that strong case management is essential to effectiveness, programs with mandatory participation need not be punitive when support services and case management are also provided, a traditional school-to-work transition has better outcomes than second-chance education and job training, and support services such as child care and transportation are essential.

The diversity of adolescent pregnancy experiences and the resilience of many young mothers and children certainly make the attainment of positive outcomes possible. Long-term studies examining the life course of adolescent mothers and their children have found large percentages attaining positive outcomes later in life (for example, Furstenberg, Brooks-Gunn, & Morgan, 1987; Horwitz, Klerman, Kuo, & Jekel, 1991). Although many teenage mothers eventually achieve positive outcomes, the search for social interventions that maximize the likelihood of well-being remains an important social goal. Heretofore, most policy interventions have focused on school attendance and employment. The enforcement of living requirements and the development of alternate housing arrangements is a new policy response. An implicit assumption is that the comprehensive nature of a residential program has greater potential to produce better outcomes.

Massachusetts' Teen Living Programs

At the time of the evaluation, the TLP network had 22 sites throughout the state and a total of 110 beds. Individual program sites are operated by several private social service agencies through contracts with MDSS. The program settings include congregate care facilities, small group homes, and supervised apartments. A comprehensive approach to services is intended; thus, programs must include 24-hour skilled staffing and supervision of residents and their children; access to licensed child care (preferably on-site); access to educational, counseling, and health services; case management; and a curriculum of parenting and independent living skills (MDSS, 1996). Teenagers are allowed to stay in the program until age 20; although they may exit at age 18.

Entry into the program begins when a teenage parent applies for welfare benefits. If the teenager states that she is unable to live at home or with an adult relative or legal guardian, she is referred for an assessment conducted by MDSS. If the teenager does not accept the recommendation (to either return to her parents' or guardians' home or to enter a TLP), she is no longer eligible for TAFDC; the case is closed after 30 days and referred to an outreach program.

The desired outcomes of TLP services include completion of high school or high school equivalency program (GED); increased parenting skills and promotion of a nurturing relationship between parent and child leading to a reduction in the numbers of supported reports of child abuse and neglect; reduction in repeat pregnancy rates; improved health outcomes for teenage parents and their children; attainment of healthy growth and development of the children; and an increased ability to live independently, including the reduction of welfare dependence (MDSS, 1996).

Study Method

Only young women who spent at least 30 days in a TLP were included in the sample. Of the 288 clients who had been in the program, 199 were interviewed. A total of 127 interviews were completed with former TLP residents, and 72 interviews were conducted with current program residents. Several strategies were used to locate the teenagers: A toll-free number was established, and each teenager was sent two mailings inviting her to call to be included in the study. Individual program sites and teenagers completing interviews were asked if they had information about the whereabouts of other teenagers. The last known address of each young woman was visited.

The research team determined that 18 teenagers had moved out of the state or out of the country and that four of them were "underground" because of their domestic violence abuse histories. MDSS, DTA, and the sites were unable to provide any locating information for 17 women. The total response rate was 65 percent overall, and 74 percent when those out-of-state, underground, or without any address information were excluded. Analysis of available data (age, region, priority status) found no differences between those interviewed and those not interviewed.

In-person interviews were conducted during a two-month field period in spring of 1998. The interviews lasted approximately one hour and were conducted by Boston University School of Social Work masters' degree students in a setting chosen by the teenage parent, usually in her home. The survey instrument consisted primarily of closed-ended questions in the following areas: background information (for example, demographics, education and employment status, living situation), financial information, housing situation, education and training, employment, health and pregnancy, substance use, service needs, support systems, and client assessment of services.

Results

Client Characteristics

All 127 clients were female. The average age at entry into the program was close to 18 (M = 17.83), with a range from 14.90 to 19.60. Nearly two-fifths (37 percent) of clients were Latina, 30 percent were black, 24 percent were white, and 9 percent were "other," primarily biracial. The primary language of clients was English (78 percent), followed by Spanish (19 percent), with a few whose primary language was Haitian Creole (3 percent) or "other" (1 percent). Most teenagers (67 percent) were already parenting at the time of intake into the TLP. One-fifth of the clients (19 percent) were pregnant, and 15 percent were both pregnant and parenting at intake.

Before entering the TLP most of the 127 young women (57 percent) had lived with their families in a self-described "permanent" living situation (that is, "a place you plan to stay indefinitely"), whereas 43 percent lived in a "temporary" situation (that is, "a place you don't plan on staying very long"). Of those in a temporary situation, 27 percent lived with relatives or friends, and 16 percent lived in a setting such as a foster home, transitional housing program, or shelter.

The 127 teenagers were asked why they were unable to live with a parent. One-third (35 percent) stated that they could not get along with their parents, 16 percent said their parents' home was too small, 15 percent had been kicked out because of their pregnancy, 12 percent did not have parents available (deceased, out-of-state, or incarcerated), 10 percent reported abuse, 7 percent wanted their own apartment, 5 percent said there were drugs in the home, and 19 percent stated "other" reasons.

Regarding previous work experience, 77 percent of the 127 young women reported having worked at some time. Typical of teenage employment, these jobs were in the areas of retail, fast food, and other low-skilled occupations. One-fourth (24 percent) of the teenagers were working at the time of entry into the TLP; of these, 30 percent were working full-time. Regarding previous schooling, 42 percent reported that they had been attending school at entry into the TLP. The average grade completed was slightly below the 10th grade. An additional 26 percent were in a GED program.

Finally, most of the 127 teenage parents reported some sources of social support, although 13 percent reported that there was no one they felt they "could depend on" or who "was there" for them when they "needed someone." The most common sources of support were mothers (39 percent), father of baby (36 percent), friends (28 percent), relatives (27 percent), father (20 percent), siblings (20 percent), and father of baby's family (17 percent). In addition, 28 percent of the 127 young women reported being involved in their church, and 23 percent reported being involved in community activities.

Services Provided

Data on services provided and outcomes attained are based only on those who have left the program (N = 127). The average length of time teenage mothers had been in the program was almost eight months (M = 7.85, SD = 5.66) with a range of one to 30 months. Some services are designed for all teenagers regardless of their individual circumstances. Other components are not appropriate for all teenagers (for example, day care is not yet needed by pregnant teenagers, GED services are not needed by those who have finished high school). (See Table 1 for clients' perceptions of services received.) In addition to these individual program components, all teenagers received basic care (shelter, food, and transportation assistance), case management, and 24-hour access to staff. Additional information about the program and issues in services delivery can be found elsewhere (Collins, Lane, & Stevens, 1999).

Description of Outcomes

The average length of time teenage mothers had been out of the program was approximately one year (M = 361 days).

Education and Training. Two-thirds of the 127 young women (65 percent) reported engaging in some type of education or training since having left the TLP. Most commonly this was GED training (28 percent), followed by job training (19 percent), college course (16 percent), high school (8 percent), and "other" (4 percent). Slightly more than a third (35 percent) reported completing their education or training program (Table 2). More specifically, 30 percent had obtained their GED since leaving the TLP. An additional 8 percent had obtained their GED and had attended some college. The highest level of education achieved (not including those reporting GED attainment) was almost the 11th grade (M = 10.68). Most of the 72 young women (87 percent) also reported plans for additional education and training. Typically this additional training focused on job training (48 percent) or college (45 percent).

Employment and Income. Slightly more than two-fifths (44 percent) of the 127 teenagers reported having been employed since the TLP. Describing their most recent or current job, most were in retail, fast food, or grocery work. One-fifth (20 percent) reported having health benefits at their most recent job. Although the number of hours worked per week was fairly substantial (M = 32.52), generally the tenure of the job was relatively short (M = 76 days). The average wage, too, was fairly low (M = $6.67 per hour). One-fourth of the young women (25 percent) reported current employment on the day of the interview. Nearly three-fourths of the 127 of teenage parents (71 percent) reported receiving TAFDC at the time of the follow-up interview. They also frequently received food stamps (61 percent) and Supplemental Food Program for Women, Infants, and Children (58 percent). Other sources or income were reported as well. One-fourth (24 percent) were receiving income from employment, 21 percent were receiving money from the father of the baby, and smaller percentages were receiving financial assistance from other family members and friends. (Although 25 percent of the young women were employed, one had not yet received a paycheck at the time of the interview and did not report income from her job.) The average monthly income was $657, and the amount of savings was quite low (M = $55.83).

Housing. More than half of the 127 teenage mothers reported living in a temporary housing situation (58 percent) as opposed to a permanent situation (42 percent). Of those reporting a temporary arrangement, most (60 percent) were living with family, friends, or the father of their child. Of those living in permanent housing, most (53 percent) lived in public housing, 11 percent lived in Section 8 housing, more than one-fourth (26 percent) lived in unsubsidized rental housing, 6 percent lived in an owned home, and 4 percent lived in an "other" setting. The average number of places lived since the TLP was 1.86, and the average monthly housing cost was $310.

Twenty teenage mothers (16 percent) reported having been homeless at some point since leaving the TLP. The median number of months they were homeless was 3.59. A description of the nature of their homelessness includes living in shelters as well as "floating around" between friends and relatives. The circumstances surrounding the homeless spells included an inability to afford their own place, an inability to live with others or in programs, the tenuous nature of the teenagers' relationships, and a combination of these factors.

Health and Pregnancy. Most of the 127 teenage mothers (91 percent) reported having Medicaid insurance. Four percent had private health insurance, and 5 percent reported no health insurance. A large majority of the 127 young women (85 percent) had sought medical care for themselves; of these, 31 percent reported seeking care from a physician for a medical problem, and 82 percent reported seeking care from a physician for a check up. Slightly more (87 percent) reported having sought medical care for their child; of these, 40 percent sought care from a physician for their child's medical problem, 86 percent reported having sought care from a physician for their child's check up.

One-third of the 127 teenage mothers (28 percent) reported that they had been pregnant since leaving the TLP. Of these, 31 percent have since delivered the child, and 40 percent were still pregnant. It is unknown to what extent the remainder aborted or miscarried.

Teenage mothers were asked about various difficulties they may have experienced since the TLP. Of the 127, 32 percent reported having sought counseling; 14 percent reported that they had suffered physical, emotional, or sexual abuse. Nearly one-fifth (18 percent) stated that a report for suspected child abuse had been filed on behalf of their child, and 2 percent reported that such a report had been filed on their own behalf.

Discussion

Evaluation activities are useful for providing information to aid in further policy and program development. Although lack of a comparison group limits the conclusions about program effectiveness, "improvements can be made in programs when discrepancies are noted between what is observed and what was planned, projected, or needed" (Posavac & Carey, 1997, p. 2). Thus, the descriptive data presented about the teenagers who lived in the program can be compared to program expectations, and the outcomes attained can be compared with relevant data on similar populations to guide program improvements.

It is not surprising that all clients are female, that educational attainment is low, and that work experience is limited; these characteristics are typical of welfare caseloads. Of greater interest were the findings related to race, age, and reason for program entry. The programs had a smaller percentage of white clients and larger percentages of Latina and black clients than in the general caseload. An analysis of data from the entire DTA caseload of teenage parent recipients in the year following welfare reform found that 52 percent of recipients were white, 28 percent were Hispanic, 18 percent were black, and 2 percent were from "other" racial backgrounds. Several possible reasons might explain the overrepresentation of Hispanic and black teenagers. First, there may be reasons why Hispanic and black teenagers are less able to live in the home of a parent or guardian than white teenagers (for example, a higher frequency of abuse or neglect in the home, fewer family resources in terms of safe and stable ho using, or the unavailability of parents and guardians). Second, it is possible that white teenagers are less likely to disclose abuse or neglect in their homes, and consequently remain living in unsafe situations. A third reason may be the location of programs. Few programs are located in the rural areas where the predominant racial group is white, and more programs are located in the urban areas where larger percentages of poor people are black or Hispanic. Future research should investigate these as possible reasons for the disproportion.

Another unexpected finding about the population was its relatively mature age. It is not surprising that there were few young teenagers in the program (ages 13 to 16) because young teenagers are not a large proportion of the overall caseload. Evidence from site visits and client interviews (Collins et al., 1999) suggested that because of the numerous program rules, teenage clients are eager to leave programs when they turn 18 (at which time they are no longer required to abide by the living requirement). The finding that the mean age of the group is close to 18 suggests that some teenagers are interested in entering and staying in the program beyond this age.

Finally, although the program was designed for those teenage parent welfare recipients unable to live at home because of "abuse, neglect, addiction, or other extraordinary circumstances" the findings identify the type and frequency of the various reasons for entering the program. In addition to some expected situations (for example, abuse, drugs in home, and parent unavailability), other substantial percentages reported reasons such as parents' home was too small, conditions were unsafe or unsanitary, and teenager wanted her own apartment. The first two reasons, "home too small" and "unsafe conditions," suggest problems with the physical housing situation rather than the relationship with family members. Consequently, improvements in the general housing stock available for low-income families may reduce the need for the teenage parent to live outside of a family. Although the stated reason "wanted own apartment" is on the face of it not an acceptable reason for entry into the program, it may be indicative of underlying reasons related to abuse, neglect, or other more appropriate reasons for not living at home. However, anecdotal information collected during program site visits suggested that some teenagers erroneously view the program as a means of getting their own apartment.

The data describing outcomes provides important information about strengths and limitations of the program. Outcomes related to health, education, and TAFDC receipt appear to be most positive; outcomes related to employment and income, housing, pregnancy, and violence, are more disappointing. Stronger conclusions about program effects could be made if a control group had been available or if it were possible to isolate the effect of living requirements from the broader context of large-scale and rapid policy change at both state and federal levels. However, the implementation of reforms in a natural setting precludes such controlled experimentation. In the absence of an experimental design, we compared the outcomes with existing literature to make assessments about areas of program strength and weakness.

As stated, outcomes related to health care were quite positive. Few clients were uninsured, most reported seeking preventive care, and the vast majority of medical problems reported were not serious. While they reside in the TLPs, teenagers and their children have easy access to health care providers, and staff provide education and modeling regarding health and nutrition. Research has shown that the frequency of negative maternal health outcomes is reduced for teenagers who receive comprehensive prenatal care (Klerman, cited in Coley & ChaseLansdale, 1998). Consequently, it is realistic to believe that the direct provision of health care services and the controlled setting of the programs (that is, nutritious meals and the prohibition of alcohol and substances) contribute to the positive health of teenagers and their children.

Large percentages of teenage mothers reported engaging in educational and training activity. Although historically teenage parenting has been found to be negatively related to educational attainment, recent studies have found increases in the educational attainment of teenage parents, in part because of increasing availability of general equivalency degree (GED) programs, schooling requirements for welfare receipt, and more progressive school policies on accepting pregnant students (Coley & ChaseLansdale, 1998). GEDs were the favored educational option of the parents in this program, often because they were behind their peers in grade level and were reluctant to engage in formal education. Unfortunately, a GED does not have the same future economic return as a high school diploma, in part because students entering college with a GED are more likely to drop out than those entering with a traditional diploma (Cameron & Heckman, 1993).

There is a clear decline in the percentage of teenagers receiving TAFDC. At program entry 100 percent of the young women were receiving TAFDC, at the time of the follow-up interview, 71 percent reported receiving TAFDC. Other research using administrative data examined the cohort of teenagers enrolling on welfare in the year after welfare reform (N = 2,006) and found that 56 percent of those living in, or waiting to enter, a TLP (n = 91) were receiving TAFDC exactly one year later (Collins, in press). This indicates that a greater percentage of the number surveyed were still on welfare compared with this first-year cohort. The analysis of the first-year cohort included those waiting to enter and those staying less than one month, thus, possibly explaining the lower percentage of those receiving TAFDC as a result of sanctioning of those who did not enter or stay in the program. Although data from the client survey are not available to explain the reason for leaving assistance, we compared the monthly income of those receiving TAFDC with those not receiving TAFDC to ascertain whether those having left welfare were better off financially. This analysis found that those having left TAFDC had nonsignificantly higher monthly income. Thu s, those having left TAFDC are neither better nor worse off financially than those remaining on assistance. Not surprisingly, whether on welfare or off, the teenagers are likely to be poor.

Less promising outcomes were those related to employment and income, homelessness, subsequent pregnancy, and violence. Although nearly half of the young women reported having worked since leaving the TLP, only a quarter were working as of the date of the interview. This appears to be a lower percentage than the general population, in which just over one-half of all mothers with infants under one year of age are working and almost two-thirds of mothers of preschoolers (ages three to five) are employed (U.S. Bureau of Labor Statistics, cited in Brooks-Gunn, Smith, Berlin, & Lee, 1998). Moreover, the type of jobs reported by members of the sample were generally low-paying, without benefits, and of fairly short duration.

As discussed earlier, adolescent parents appear to desire education and training and to engage in such activity after program completion. However, the link to adequately paying employment and consequent escape from poverty remains tenuous. In part, progress in education, employment, and economic stability requires prolonged effort. The length of follow-up in this study averages about one year and therefore is unlikely to capture experiences with advanced education and its effect on employment. Moreover, in this sample, a fair percentage of teenagers reported experiences with homelessness, violence, and subsequent pregnancy, which can interrupt the attainment of education or disrupt the transition from education to employment.

The lack of adequate income is likely to influence directly the instability of housing for teenage parents and their children. Although a large number reported temporary housing arrangements, many of these situations reflected stepping stones to a better housing situation (that is, living with a roommate until the teenager could afford her own place) rather than a tenuous arrangement. More troublesome is the fairly large percentage (16 percent) who reported having been homeless at some point since leaving the TLP. This is a larger percentage than two recent investigations of teenage homelessness. A study of New England high school students found that 5 percent to 10 percent reported homelessness or severe housing distress in a year's period (Vissing & Diament, 1997), and a nationally representative survey of youths found that almost 8 percent had at least one homeless episode in the past 12 months (Ringwalt, Greene, Robertson, & McPheeters, 1998).

Repeat pregnancy was not uncommon; more than one-fourth of the young women reported a subsequent pregnancy since leaving the TLP. Maynard (1995) reported that in a follow-up to the Teenage Parent Welfare Demonstration over half of the young mothers were pregnant within two years after enrolling in the study sample. Thus, the percentages reported in this study appear to be no better or worse than those reported elsewhere.

Finally, teenage parents appear to be at high risk of victimization and their children at high risk of maltreatment. Domestic violence has been identified as a serious problem among high school girls (Bergman, 1992; O'Keefe, Brockopp, & Chew, 1986), and a high prevalence of sexual victimization (Boyer & Fine, 1992) and revictimization (Collins, 1998) has been found in studies of adolescent mothers. Violence can impede the transition from welfare to work by physically preventing women from engaging in training activities that are necessary for self-sufficiency and because the effects of violence may make young women unable to take advantage of opportunities.

In a fairly large percentage of cases, reports of child maltreatment had been made on behalf of the teenager's child. Although this is a serious outcome, we note that during program site visits staff recounted that maltreatment reports were sometimes made as the teenage mother left the program if she were moving into a situation considered unsafe. This suggests that the substantial percentage of reports was in part a response to the program surveillance of the teenage parents. Moreover, because most teenage parents enter the program because they do not have an appropriate living situation, it is likely that more children are endangered before entering the program than at program end. In comparison, Collins (1999) found that in a randomly selected sample of 447 DTA teenage cases, 10 percent had a report of child maltreatment within a three-to-five-year period after enrollment on welfare. This bolsters the suggestion that TLP cases are at higher risk of a maltreatment report because of the program surveillance . It also should be noted that in the analysis of the 46 cases, all but one involved neglect. Although data is not available about the type of maltreatment reported among the teenagers surveyed, poverty and unsafe housing that the teenagers continued to experience, were likely related to reports for neglect.

Conclusion and Implications

The data reported here lead to several suggestions for policy, program development, and clinical intervention. First, the findings identify limits of the current policy response to teenage mothers receiving welfare. Pervasive and systemic issues related to the broader economy, societal attitudes toward teenage mothers, and values that emphasize power rather than vulnerability, will continue to threaten the wellbeing of teenage mothers and their children. The least successful outcomes (employment and income, housing stability, violence) were those that reflect these disadvantages. Welfare reform living requirements in general and TLPs in particular do not address labor market conditions, the supply of housing, or societal attitudes supportive of violence. Thus, efforts to enhance the likelihood of optimal outcomes must simultaneously aim at educational and training systems, safe and affordable housing, and value orientations that are supportive of adolescent girls and young families. Specifically, in terms of educational and training systems, more emphasis should be placed on training for nontraditional jobs that pay higher wages, flexibility in school environments that accommodate teenage mothers, increased emphasis on traditional high school education, postsecondary educational support in terms of financial aid, and the allowance of postsecondary education and training to count toward the work requirement (currently not the policy in Massachusetts). The most critical policy response in terms of housing is the construction of safe and affordable housing in most communities. Additional programs that combine housing with other social supports and emphasize a transition to independent living are needed and are discussed more later. Finally, although the issues related to reducing violence toward young women are complex, policy that strengthens employment potential and safe housing can help reduce young women's dependence on abusive partners. Moreover, although antiviolence training of young men is needed, so are pro grams of education, employment, and emotional support that may reduce a propensity for violence.

Specific implications for further programming can be made in three areas: program targeting, transitional housing, and follow-up services. Program targeting issues have been discussed. Clearer communication about the program purpose and structure before entry will help to ensure program space is allotted to those most likely to benefit. Attention to geographic location also is necessary to ensure adequate coverage throughout the state. Transitional housing services provided to the teenagers at program end would provide a further step to independence but would continue to provide reasonably priced and safe housing for these still-vulnerable families. At the same time, ongoing staff contact and support can help to provide continued individual and parental growth. Related to this, the results of our study suggest an increased need for follow-up services. Although programs do maintain contact with teenagers after they leave the program, more sustained contact, support, and services appear warranted to provide the needed guidance and support to help the teenage mothers through the extended transition from adolescent girl to young adult with family responsibilities.

Finally, the study offers implications for clinical intervention, specifically related to social support. Scholars agree that extended social support networks, including family and friends, are critical protective factors in the development of resilience in youths. Moreover, resilient youths tend to have a warm and caring relationship with at least one caring adult (Stevens, 1997). In the TLP study, 13 percent of the teenage parents reported that they were not connected to a social support network. A significant factor in clinical work with this client population would constitute the development of strong stable relational connections involving friends, relatives, and community institutions. Supportive healthy relational connections are the basis for a social support network once services with the TLP are terminated. There is some evidence from empirical research to suggest that poor families enhance the social mobility of their offspring when they connect them to indigenous mentors and mobility enhancing co mmunity organizations and institutions (Blechman, 1992). This program has not considered development of a social support network per se as a required skill acquisition for sustained independent community living. More explicit attention to this skill is needed. We hope that for these young parents natural support systems would become a social anchor in their lives for continued resilient development.

In conclusion, living requirement legislation and the development of TLPs are recent responses to address the needs of young families receiving welfare. The Teen Living Programs are comprehensive in terms of programming (physical safety, education, attention to health care and emotional needs, parenting instruction and support, and so forth) and duration (teenagers can stay in the program until the age of 20). The early evidence reported here suggests that teenagers in the program have attained some positive outcomes. A social work response at policy, program, and individual levels is needed in the ways suggested in this article, so that the likelihood of positive outcomes for these vulnerable families is enhanced.

Mary Elizabeth Collins, PhD, is assistant professor, Joyce West Stevens, PhD, ACSW, LICSW, Diplomate, is assistant professor, and Terry Saunders Lane, MSS, is associate dean for professional education and sponsored projects, School of Social Work, Boston University, 264 Bay State Road, Boston, MA 02215; e-mail: mcollins@bu. edu. An earlier version of this article was presented to the third annual conference of the Society for Social Work and Research, January 1999, Austin, TX. This study was completed under contract to the Massachusetts Department of Transitional Assistance. The authors thank research assistants Lisa Street and Cristi Lemon, the research interviewers, and the parents who participated. Correspondence regarding this manuscript should be addressed to Mary Elizabeth Collins, PhD, 264 Bay State Road, Boston, MA 02215; e-mail: mcollins@bu.edu.

References

Alan Guttmacher Institute. (1994). Sex and America's teenagers. New York: Author.

Bergman, L. (1992). Dating violence among high school students. Social Work, 37, 21-27.

Blechman, E. (1992). Mentors for high-risk minority youth: From effective communication to bicultural competence. Journal of Clinical Child Psychology, 21, 160-169.

Boyer, D., & Fine, D. (1992). Sexual abuse as a factor in adolescent pregnancy and child maltreatment. Family Planning Perspectives, 24,4-11.

Brooks-Gunn, J., & Duncan, G. J. (1997). The effects of poverty on children. Future of Children, 7(2), 55-71.

Brooks-Gunn, J., Smith, J., Berlin, L., & Lee, K. (1998). Implementations of welfare changes for parents of young children. Unpublished manuscript, Columbia University, Teachers College, Center for Children and Families.

Cameron, S. V., & Heckman, J. J. (1993). The nonequivalence of high school equivalents. Journal of Labor Economics, 11(1), 1-47.

Coley, R. L., & Chase-Lansdale, p. (1998). Adolescent pregnancy and parenthood: Recent evidence and future directions. American Psychologist, 53, 152- 166.

Collins, M. E. (1998). Factors influencing sexual victimization and revictimization in a sample of adolescent mothers. Journal of Interpersonal Violence, 13(1), 3-24.

Collins, M. E. (in press). Impact of welfare reform on teen parent recipients: An analysis of two cohorts. American Journal of Orthopsychiatry.

Collins, M. E., Lane, T. S., & Stevens, J. W. (1999). Teen Living Programs: A description of program components and issues in service delivery. Unpublished manuscript, School of Social Work, Boston University.

Duncan, G. J., Harris, K. M., & Boisjoly, J. (1997). Time limits and welfare reform: New estimates of the number and characteristics of affected families. Chicago: Northwestern University/University of Chicago Joint Center for Poverty Research.

Furstenberg, F. F., Jr., Brooks-Gunn, J., & Morgan, p. (1987). Adolescent mothers in later life. New York: Cambridge University Press.

Horowitz, S. M., Klerman, L. V., Kuo, H. S., & Jekel, J. F. (1991). School-age mothers: Predictors of long-term education and economic outcomes. Pediatrics, 87, 862-867.

Manpower Demonstration Research Corporation. (1997). New Chance: Final report on a comprehensive pro gram for young mothers in poverty and their children. New York: Author.

Massachusetts Department of Social Services. (1996). Teen Living Program specifications: Request for proposals. Boston: Author.

Maynard, R. (1995). Teenage childbearing and welfare reform: Lessons from a decade of demonstration and evaluation research. Children and Youth Services Review, 17, 309-332.

O'Keefe, N., Brockopp, K., & Chew, E. (1986). Teen dating violence. Social Work, 31, 465-468.

Personal Responsibility and Work Opportunity Reconciliation Act of 1996, P.L. 104-193, 110 Stat. 2105.

Posavac, E. J., & Carey, R. G. (1997). Program evaluation: Methods and case studies. Upper Saddle River, NJ: Prentice-Hall.

Ringwalt, C. L., Greene, J. M., Robertson, M., & Mcpheeters, M. (1998). The prevalence of homelessness among adolescents in the United States. American Journal of Public Health, 88, 1325-1329.

Sedlak, A. J., & Broadhurst, D. D. (1996). The third national incidence study of child abuse and neglect. Washington, DC: U.S. Department of Health and Human Services.

Stevens, J. W. (1997). Opportunity, outlook and coping in poor urban African American late-age female adolescent contraceptors. Smith College Studies in Social Work, 67, 456-474.

U.S. General Accounting Office. (1994). Families on welfare: Focus on teenage mothers could enhance welfare reform efforts (Publication No. GAO/ HEHS-94-112). Washington, DC: Author.

U.S. General Accounting Office. (1995). Welfare to work: Approaches that help teenage mothers complete high school (Publication No. GAO/HEHS/PEMD-95-202). Washington, DC: Author.

Videka-Sherman, L., & Viggiani, P. (1996). The impact of federal policy changes on children: Research needs for the future. Social Work, 41, 594-600.

Vissing, Y. M., & Diament, J. (1997). Housing distress among high school students. Social Work, 42, 34-41.

Wertheimer, R., & Moore, K. (1998). Childbearing by teens: Links to welfare reform. In New Federalism: Issues and Options for States (Publication No. A-24). Washington, DC: Urban Institute.