TANF 5th Annual Report to Congress

XIII. TANF Research and Evaluation

Highlights of Major Research and Evaluation Findings

Consequences of Welfare Reform: A Research Synthesis
Welfare Time Limits: State Policies, Implementation, and Effects on Families
National Evaluation of Welfare-to-Work Strategies: How Effective Are Different Welfare-to-Work Approaches? Five-Year Adult and Child Impacts for Eleven Programs
Grants to States and Localities to Study Welfare Reform Outcomes, with an Emphasis on TANF Applicants and Diversion
Overview of Additional Research and Evaluation Efforts
Evaluating Welfare Reform
Promoting Healthy Marriage and Responsible Fatherhood
Employment Retention and Advancement
Effective Strategies for Serving the Hard-to-Employ
Rural Welfare-to-Work Strategies
Addressing the Needs of Other Special Populations
Evaluation of the Welfare-to-Work Grants Program

This chapter highlights a number of the Department of Health and Human Services’ (HHS) key research and evaluation initiatives pertaining to welfare reform and summarizes findings from recent research reports. The Department’s research agenda in this area has two main goals:

To contribute to the success of welfare reform by providing timely, reliable data to inform policy and program design and management, especially at the State and local levels where much of the decision-making has taken place.
To inform the nation of the effects of policies and programs on low-income children, families, and communities, and the nation as a whole.

The research undertaken to achieve these goals is done primarily by the Administration for Children and Families’ (ACF) Office of Planning, Research and Evaluation (OPRE) and the Office of the Assistant Secretary for Planning and Evaluation (ASPE), also in the Department. OPRE and ASPE carefully coordinate their research agendas with each other and with other government agencies and private foundations. Many projects involve collaboration and partnerships.

The National Academy of Sciences convened a Panel on Data and Methods for Measuring the Effects of Changes in Social Welfare Programs to provide HHS with unbiased scientific recommendations for studying the outcomes of recent changes in the welfare system. The panel’s conclusions and recommendations on research questions and populations of interest, evaluation methods and issues, and data needs and issues were presented in its 2001 report, Evaluating Welfare Reform in an Era of Transition. The Academy applauded the Department’s broad-based welfare reform research agenda, while recommending improvements and expansions in data collection and the development of research questions and methodology, particularly with respect to evaluating the effects of welfare reform, i.e., how welfare reform has changed the outcomes for families and individuals relative to what would have happened in the absence of reform.

This chapter summarizes recent research and evaluation findings and provides an overview of additional research and evaluation initiatives related to the Temporary Assistance for Needy Families (TANF) program undertaken by HHS. The following section highlights results from four major research and evaluation projects designed to improve understanding of programs and policies serving low-income families, and the section after that provides an overview of additional HHS projects in this area. Major research and evaluation initiatives include a study by RAND synthesizing the current state of knowledge about the net effects of TANF, a comprehensive summary of what has been learned to date about time limits, the National Evaluation of Welfare-to-Work Strategies (NEWWS, formerly known as the Job Opportunities and Basic Skills Training program (JOBS) Evaluation), and grants to States and localities to study welfare reform outcomes.

Highlights of Major Research and Evaluation Findings
Consequences of Welfare Reform: A Research Synthesis

To inform public debate on issues pertaining to reauthorization of the TANF legislation, and to help States improve their TANF programs, ACF contracted with RAND to synthesize the current state of knowledge regarding the effects of the TANF legislation and the programs of individual States. The report considered a range of outcomes, including TANF caseloads, employment and earnings, use of other government programs, fertility and marriage, household income and poverty, food security and housing, and child well-being. The primary focus was on the net effects of TANF, after taking into account other factors such as the economy and other policy changes. As does the literature on which it was based, the synthesis considered both the effects of specific policies, such as benefit structures, time limits, work requirements, and sanction policies, as well as the effect of the TANF reforms as a whole. Key findings are as follows:

The effect of welfare reform on the use of welfare has been well studied. Over a dozen econometric studies have attempted to estimate the effects of welfare reform as a whole, and all but a few conclude that reform had substantial effects on reducing the caseload. Most of the reforms that were introduced in the 1990s had positive effects on employment and earnings. Thus, it is likely that welfare reform is responsible for a portion of the increase in work and earnings among single mothers during the last decade, as nearly all of the experimental and econometric evidence suggests. Some welfare reform components raise incomes and reduce poverty, although this does not occur with all policy components, and there is reason to believe that some favorable effects will not persist over time. Generous financial work incentives – high earned income disregards inside the welfare system or earnings supplements outside the welfare system – generate the strongest income gains and anti-poverty effects.
There is little information available about the effects of welfare reform on the use of other government programs or on family structure. Research generally indicates that welfare reform has caused part of the recent decline in participation in the Food Stamp Program, but it does not explain the mechanisms underlying this. Only a few studies analyze the effects of welfare waivers on the use of school nutrition programs, and fewer still have considered housing subsidies. Existing evidence does not determine whether welfare reform caused part of the initial decline in Medicaid enrollment that followed the implementation of TANF. Similarly, evidence from both experimental and econometric studies is insufficient to draw firm conclusions regarding the effects of welfare reform on marriage or fertility.
There is evidence of both favorable and unfavorable effects of various components of welfare reform on child well-being. Such effects were observed for indicators that capture socio-emotional behavior, academic performance, and health. The most favorable effects are associated with financial work incentives, most likely due to the increase in family income that results from combining work and welfare. Even for these programs, there is evidence of unfavorable impacts for some subgroups of participants, particularly for adolescent children and for younger children of parents who do not experience large income gains. Work requirements do not appear to have strong impacts on children, although here, too, there is evidence of unfavorable impacts for adolescents, especially in school performance.

Welfare Time Limits: State Policies, Implementation, and Effects on Families

ACF contracted with the Manpower Demonstration Research Corporation (MDRC) and the Lewin Group to survey State welfare administrators, conduct site visits in five States, and synthesize existing research on time limits. The report provides a comprehensive summary of what has been learned to date about time limits, including State policies and their implementation, their effect on employment and welfare receipt, and the circumstances of families whose welfare cases were closed once time limits were reached. It also provides the first national summary of the early experience of individuals reaching the 5-year Federal time limit. Key findings are as follows:

States have developed widely varying approaches to time limits. This is because States have broad flexibility in designing time-limit policies and because the Federal 60-month time limit does not apply to State-funded benefits. As a result, States can continue to provide TANF or other cash benefits to families, or to the children in these families, using State-only funds. Currently, 40 States have time limits that can result in the termination of families’ welfare benefits; 17 of those States have limits of less than 60 months. However, nearly half of the national welfare caseload is in a State that either has no time limit (Michigan and Vermont) or a time limit that reduces or modifies benefits when the limit is reached (Arizona, California, Indiana, Maine, Maryland, New York, Rhode Island, Washington, and Washington, D.C.).
Only a small fraction of those who could have reached a time limit have done so. Nationally, as of December 2001about 231,000 families had reached a time limit, of which approximately 54,000 had reached the Federal time limit; at least 93,000 families had their welfare case closed due to a time limit, and another 38,000 have had their benefits reduced. Most of the case closures have been in States that have time limits of less than 60 months. In December 2001, families began reaching the Federal time limit in fewer than half the States. Even in States where families began to reach the 60-month time limit, relatively few families had actually reached it, as most recipients do not remain on welfare for 60 consecutive months. Eighty one percent (44,000) of the families nationwide who have reached the Federal time limit reside in New York.
When combined with work requirements and generous earnings disregards, time limits increased employment in the period prior to reaching the time limit. While this combination of requirements and services did not reduce welfare receipt in this period, some evidence suggests that time limits, by themselves, would have reduced welfare receipt. Approximately the same number of recipients worked immediately after reaching time limits as worked immediately prior.
All States allow exemptions (which stop the time-limit clock), extensions, or both. Child-only cases, which now account for about one-third of all welfare cases nationwide, are not subject to the Federal time limit, and no State has chosen to impose a time limit upon them. Disability, caring for a disabled family member and domestic violence are other common bases for an exemption. In many States, recipients who comply with work requirements but are unable to obtain employment can receive extensions, although States define and assess compliance in different ways. As a result, some States routinely grant extensions to recipients reaching time limits, while others close most of these cases.
The circumstances of families who left welfare due to time limits are diverse and depend on States’ policies. In some States, most recipients whose cases have been closed due to time limits were already working while on welfare. In other States, time limit leavers are more heterogeneous. Most studies find that time-limit leavers do not consistently experience more or fewer hardships than those who left welfare for other reasons. Large proportions of time-limit leavers continue to receive food stamps and other assistance.

National Evaluation of Welfare-to-Work Strategies: How Effective Are Different Welfare-to-Work Approaches? Five-Year Adult and Child Impacts for Eleven Programs

ACF and ASPE, with support from the Department of Education, funded The National Evaluation of Welfare-to-Work Strategies (NEWWS), a comprehensive study, with a five-year follow-up period, of the effectiveness of 11 mandatory welfare-to-work programs, with MDRC and the Child Trends as evaluation contractors. The study compared the effects of two strategies to help welfare recipients move toward self-sufficiency: programs that emphasize short-term job search assistance and encourage people to find employment quickly (known as Labor-Force Attachment or LFA), and programs that emphasize basic education and skill-building activities (known as Human Capital Development or HCD). The effects of the two strategies were estimated through the random assignment of individuals to one or more program groups (with services) or to a control group (who were eligible for welfare as usual).

The evaluation studied programs in seven locations (Atlanta, Georgia; Columbus, Ohio; Detroit and Grand Rapids, Michigan; Oklahoma City, Oklahoma; Portland, Oregon, and Riverside, California). Data were collected on over 40,000 single parents (mostly mothers) and their children. The in-depth study of young children in three sites (Atlanta, Grand Rapids and Riverside) was conducted by Child Trends. Key findings are as follows:

HCD, which is more expensive to administer, did not produce added economic benefits for clients. Moreover, LFA moved welfare recipients into jobs more quickly than the training approach. The LFA approach did not affect parents’ financial well-being or that of their children any differently than the HCD approach. Overall, there were few indications of harm or benefit to the well-being of children, but effects varied by the age of the child.
The most successful program, implemented in Portland, Oregon, used an approach that was strongly employment-focused while emphasizing both HCD and LFA. This finding, along with other past research, suggests that a "mixed" approach may be most effective.
Over a five-year follow-up period, approximately three-quarters of single-parent welfare recipients who did not participate in a NEWWS program (the control group) nevertheless found jobs, and more than half left the welfare rolls. Few of the NEWWS programs improved on this already-high rate of finding a job, but nearly all programs helped single parents work during more quarters of the follow-up period and earn more than they would have in the absence of any program. Family income, however, was not affected by the presence or absence of a program, as increased earnings were offset by decreased receipt of welfare benefits. Moreover, the programs did not show any significant effects on such family measures as marriage and household composition.

Grants to States and Localities to Study Welfare Reform Outcomes, with an Emphasis on TANF Applicants and Diversion

In fiscal years 1998 and 1999, ASPE funded ten separate studies that looked at applicants and/or potential applicants to the TANF program and included individuals who were diverted from applying for TANF assistance by receiving other assistance or referrals. ASPE provided technical assistance to grantees as they determined the cohorts to be studied and the administrative data measures to be used, and developed surveys specific to their State or locality. The ten grantees were Arizona, Florida, Illinois, New York, South Carolina, Texas, Washington, Milwaukee County in Wisconsin, and two consortia of counties in the Bay Area of California (Contra Costa and Alameda Counties, and San Mateo, Santa Clara, and Santa Cruz Counties). Eight of these grantees (all except New York and Washington) have released final reports detailing their findings.

Cross-study comparisons have been difficult because there was not a common definition of "applicant" or "diverted" populations across States or grantees. Depending on the grantee, "applicants" or "divertees" were defined as individuals who either: (1) participate in the State or county’s formal diversion program; (2) begin the application process but fail to complete it; (3) complete the application process and are deemed eligible for but are not enrolled in TANF; (4) are ineligible for TANF for non-financial reasons; or (5) receive food stamps and/or Medicaid and appear to be eligible for TANF but do not receive it. Despite the differences in the definitions of the study populations among the ASPE grantees, a number of themes have emerged across these studies:

There appears to be a significant dip in earnings immediately prior to application to TANF. While this dip and the decision to apply for TANF might be related, the drop in the employment rate during this period is much less pronounced than the dip in earnings.
There is moderate use of food stamps and Medicaid among this population, decreasing gradually in the year following the application to or diversion from TANF.
A significant number of diverted applicants reapply for TANF and eventually receive cash assistance in the year following the initial diversion.

Overview of Additional Research and Evaluation Efforts
Evaluating Welfare Reform

Over the past decade, HHS has made significant investments in research and evaluation focussed on the implementation and impacts of State welfare reform initiatives. These have included projects focussed on reforms carried out to test welfare-to-work strategies under the Job Opportunities and Basic Skills Training (JOBS) program, a variety of policies to promote work and personal responsibility implemented by individual States under waiver demonstrations, assessments of reforms enacted under TANF, and the projects described above.

Prior to the passage of TANF, 43 States and the District of Columbia obtained waivers of certain program requirements in title IV-A of the Social Security Act (the Act), as authorized under section 1115 of the Act. ACF required an evaluation component as a part of each approved waiver. Continuation of evaluations begun under waivers was permitted but not required under the TANF statute. Twenty States were funded to complete ongoing evaluations, either as originally planned or modified. Only one of these projects (Minnesota) continued to receive funding in FY 2002, while other states carried out the process of completing evaluations with prior year funds. During 2001, Arizona, Florida, and Iowa released interim or final reports under this series of grants. Through September 2002, Connecticut, Iowa, Maryland, Nebraska, Vermont and Virginia have released final reports, Montana released an interim report, and Indiana released a report on the impacts of welfare reform on children. These reports outline lessons learned regarding the implementation of welfare reform programs in terms of work requirements, time limits, and enhanced earnings disregards. Additional State reports are expected to be released in late 2002 or early 2003 (California, Indiana, Montana, Nebraska, New Hampshire, Ohio, and Texas).

Findings from reports on the waiver evaluations are particularly relevant to TANF, since these demonstrations first implemented many of the policies now incorporated under State TANF plans. Of particular note are findings from the Florida and Connecticut studies, since both States instituted short time limits for cash assistance and earnings disregards to promote work. Both States also augmented their evaluations to focus more directly and consistently on appropriate measures of child health and well-being. Iowa and Indiana also augmented their evaluations to capture information on child well-being. Both of the latter states released reports in 2002 of their findings on child outcomes (Iowa as part of the final report on their waiver evaluation). Vermont, which also released the final report of their waiver evaluation this year, had no time limits on the receipt of cash assistance, but did provide a demonstration of the potential impacts of financial work incentives.

Findings from Florida are as follows:

Over the four-year follow-up period, Florida’s Family Transition Program (FTP) increased employment and earnings, reduced welfare receipt, and modestly raised participants’ income. The pattern of FTP’s impacts on employment, welfare receipt, and income shifted significantly over the four-year follow-up period. FTP’s impact on employment emerged early in the follow-up period, peaked in years two and three, and then disappeared in year four. In the last few months of follow-up, the FTP group was less likely to receive welfare than the control group, but no more likely to work, and the two groups had about the same total income. Employment and earnings gains were concentrated among less disadvantaged sample members, and FTP had little or no impact on employment or earnings for more disadvantaged groups.
FTP had no impact on a range of measures of family structure and economic well-being, although, on a few indicators, living conditions appeared to be slightly better at the four-year point for the FTP group compared to the control group. Levels of material hardship were high for both groups. Most of the employed people in both the FTP group and the control group worked full time or close to full time in jobs that paid low wages and offered few fringe benefits.
More than three-fourths of the FTP group received benefits for less than the 24 or 36 months allowed under their time limit, while two-thirds of those who received 24 or 36 months of benefits - one sixth of the participants - had their welfare grants canceled due to the time limit. The FTP participants who reached the time limit were a diverse group, and were not necessarily the most disadvantaged participants. The post-welfare experiences of families whose grants were canceled varied considerably. Most struggled financially, but no more than families who left welfare for other reasons.
Overall, FTP had few effects across a range of measures of parenting and child well-being for 5- to 12-year-olds. There were a few negative impacts on school-related outcomes for adolescents, however. Surprisingly, FTP generated some negative effects for children in the least disadvantaged families - the subgroup with the largest earnings impacts. FTP children spent more time in child care than their control group peers, and were more likely to have contact with their non-custodial fathers.

In Connecticut’s Jobs First demonstration, findings are as follows:

The majority of the treatment group reached the time limit during the study period, but half of these received benefit extensions. During the study, about one-third of the treatment group cases, most of whom were working, were closed due to the time limit.
The program significantly increased employment and earnings, with the largest impacts among individuals facing the greatest barriers to employment. Income was increased significantly prior to individuals reaching the time limit, but these gains disappeared after large numbers of cases were terminated upon reaching the time limit.
Over the follow-up period Jobs First did not affect cash assistance receipt, since generous income disregards allowed working recipients with income below the poverty level to remain on the rolls until their time limit expired. However, the Jobs First group derived more of its income from earnings.
The program had no consistent effect on a range of indicators of material well-being for either group. During the entire study period, absolute levels of hardship (defined as at least one material hardship in the past year, such as unmet medical needs) remained high for the majority of both the treatment and control group cases.
Jobs First generated some positive effects on behavior among elementary school-age children, but no consistent impacts on school performance or engagement. Effects on adolescents were mixed, with the treatment group doing worse in school than their counterparts, but less likely to have been convicted of a crime.
During the study, savings from decreased welfare payments did not offset government costs.

The Iowa Family Investment Program (FIP) combined program changes designed to ease a family’s transition from welfare to work, such as more generous earnings disregards, with strict requirements that recipients participate in the development and execution of a social contract, the Family Investment Agreement (FIA), which details the steps parents will take to achieve economic self-sufficiency, the financial assistance and services that the State will provide to facilitate that process, and the time frame for doing so. Families that opt not to develop a FIA or fail to follow through with the agreement are placed on a 6-month Limited Benefit Plan (LBP) followed by the complete loss of cash assistance for the subsequent 6-month period. Findings from the evaluation are as follows:

Welfare reform resulted in higher rates of participation in the employment and training program and higher rates of combining employment with the receipt of cash assistance. For cases on assistance when Iowa’s welfare reform began, it increased the earnings of the heads of households in the long run, which was accompanied by reductions in FIP and Food Stamp benefits. Welfare reform also instilled in FIP participants greater responsibility for the well-being of themselves and their families.
The findings also indicate significant negative impacts on cases that applied for assistance after FIP was implemented. Welfare reform raised the rates of participation in FIP, Medicaid, and public/subsidized housing, while it reduced household income earnings and total income, the latter by $200 per month. Welfare reform had a negative impact on marriage, which may have led to this reduction of household income.
In addition, FIP also had negative impacts on a number of other measures of family and child well-being, such as financial strain, doubling up of households, domestic abuse, school engagement of children and the proportion of children in foster care.

Indiana's welfare reform program included a strong work-first focus, combined with a 24-month time limit on adults' receipt of assistance and a family cap policy. In addition, adults were required to comply with a number of personal responsibility provisions. Findings on family and child well-being effects were as follows:

Overall, Indiana's welfare reform program did not have large impacts on elementary school age children (those between 5 and 12 years old at the time of the survey). There were no statistically significant impacts on education or health outcomes. With respect to behavioral outcomes, there was only one statistically significant impact - an increase in arrest rates among 10 to 12-year-olds; it is not clear, however, whether this is a real effect of welfare reform or only a chance artifact of multiple significance tests.
Although the survey focused primarily on 5 to 12-year-olds, it did include some questions pertaining to adolescents (i.e., those 13 to 17 years old at the time of the survey). For all but one of the small number of outcomes available for adolescents, the study found no statistically significant impacts of welfare reform. The exception was a small adverse effect on adolescents' school performance. This finding is consistent with other recent studies showing some modest adverse effects of welfare reform on adolescents.

To achieve its aim of increasing employment and reducing reliance on welfare, Vermont's Welfare Restructuring Project (WRP) provided financial incentives to encourage and reward employment, as well as requiring single-parent welfare recipients to work in a wage-paying job after 30 months of benefit receipt (after 15 months for two-parent families). The State helped recipients find work and provided subsidized, minimum-wage community service positions to those who could not find regular jobs. Vermont's policies differed from the welfare reform policies in other States and at the Federal level in two major ways. First, WRP did not have a time limit on welfare receipt and did not require single-parent recipients to participate in any work-related activities during their first 30 months on the rolls. Second, when a recipient failed to comply with the work requirement, the State did not lower the welfare grant. Instead, it took control of the grant, using the money to pay the recipient's bills, and required the recipient to attend frequent meetings at the welfare office. Findings from the evaluation were as follows:

WRP increased employment and reduced reliance on cash assistance for single-parent families. The WRP group was more likely to work than the control group. The WRP group earned an average of $500 (9 percent) more per year than the control group and received $300 (12 percent) less per year in cash assistance. WRP had few effects for two-parent families, who make-up a small percentage of Vermont's welfare caseload.
The program's work requirement was needed in order to generate impacts. WRP's financial incentives alone did not lead to increases in employment or income.
WRP had little effect on family income, material hardship, or child well-being. The WRP group's higher earnings were largely offset by their lower welfare payments; as a result, average income for the WRP group was about the same as average income for the control group. However, members of the WRP group derived a greater share of their income from earnings and a smaller share from public assistance. Because WRP did not raise family income, it is not surprising that it also had few effects on hardship. WRP also had few effects on child outcomes.
WRP saved money for taxpayers. The WRP group received few services that were not also available to the control group. Thus, the program's net cost was low and was more than offset by the public assistance savings it generated.

Wisconsin also released a report in 2001 based on a waiver demonstration that allowed the State to "pass-through" to TANF families the entire Federal and State share of child support collected rather than some of the funds collected in order to offset the costs of providing welfare benefits. The evaluation provided substantial evidence of the expected direct effects of the child support policy. Findings are as follows:

Mothers eligible for a full pass-through received about $150 to $200 more per year than did those in the control group, and a significantly higher percentage of nonresident fathers paid child support on behalf of experimental group families than control group families.
The experimental group experienced significantly higher rates of paternity establishment.
Evidence was less consistent for secondary effects. For example, although child support reduced the need for cash payments, and experimental group fathers tended to have less informal earnings, there were few differences in mothers’ employment or earnings, child well-being, and relationships between parents.

ACF and ASPE are also attempting to understand better welfare reform in urban areas by supporting The Project on Devolution and Urban Change, a multi-disciplinary, longitudinal study of the implementation and impacts of welfare reform in four large urban areas - Cleveland (Cuyahoga County), Ohio; Philadelphia, Pennsylvania; Los Angeles, California; and Miami, Florida. All of the sites applied work participation requirements to a larger proportion of the welfare caseload than they had prior to TANF, and succeeded in engaging a higher proportion of recipients in welfare-to-work program activities. The most common work activity was unsubsidized employment. Recent findings are as follows:

The combination of employment and welfare status was the most consistent predictor of respondents’ health-related outcomes, even after controlling for site differences and demographic characteristics. Women who were employed at the time of the interview were substantially less likely to have physical or mental health barriers, or to experience material hardships than those who were not working. In addition, children of working women were also more likely than children of women not working to be healthy. Yet nearly all the women who worked faced one or more challenges, such as health problems (theirs or a child’s), lack of education, or depression. This suggests that these characteristics are not absolute barriers to employment, but they reduce the probability of stable employment.
MDRC surveyed prior welfare recipients in neighborhoods with high concentrations of poor and welfare-dependent households. Surveys found this to be a disadvantaged population, as respondents and their children consistently did poorly on a range of health outcomes compared to national comparison samples. Nonetheless, about half of the survey respondents were working at the time of the interview, and another quarter had worked at some point in the past two years. However, most had low-wage jobs with earnings that left their families below the poverty level.
Welfare status was closely associated with access to health care. Welfare leavers, especially those who were not working, had significantly greater problems with health care access than current recipients. The most important reason for loss of non-welfare benefits, including food stamps and Medicaid, was failure to appear for TANF eligibility redetermination.
The most recent report synthesized data on welfare reform in Cuyahoga County from all of the components of the Urban Change study: longitudinal administrative data, survey data, an implementation study, neighborhood indicators, and an ethnographic study. Similar synthesis reports on the other three sites in Urban Change (Miami-Dade County, FL; Los Angeles, CA; and Philadelphia, PA) will follow over the next year.

Promoting Healthy Marriage and Responsible Fatherhood

ACF and ASPE have developed a multi-pronged approach to increasing knowledge in this important subject area. In FY 2001, ACF awarded a contract to Mathematica Policy Research, Inc., to develop a conceptual framework for interventions that promote positive family functioning among unwed parents and, for those who wish to marry, offer services to help them enter and sustain healthy marriages. The contract also provides for technical assistance to State and local agencies interested in developing interventions and for process evaluations or case studies. In late 2002, ACF intends to use the above findings to launch a large-scale evaluation of demonstrations of alternative strategies to increase child well-being by helping unwed couples with newborns who wish to marry, enter into and sustain healthy marriages; increase positive family functioning for all families; and help new parents address their life-course objectives. These demonstrations will examine the opportunities for strengthening family connections and support earlier in a couple’s relationship than in previous demonstrations conducted under the Parents’ Fair Share (PFS) program. Two final reports on the impact of PFS on paternal involvement and low-income fathers’ employment were released in early FY 2001.

In addition, ASPE funded the Lewin Group to produce an inventory of State-level policies that support or promote marriage through various means, including media campaigns, special commissions, marriage and relationship preparation and education, youth education and development, divorce laws and procedures, court initiatives, State tax policies, welfare and other related policies, and specialized programs. The report, State Policies to Promote Marriage: A Preliminary Report, was released in May 2002. Findings are as follows:

All States have undertaken work in at least one of these areas and some States have activities underway in multiple areas.
More work has been undertaken in some areas than others. For example, 33 States provide TANF benefits to two-parent families using the same eligibility rules as for single-parent families. Four States include marriage or relationship education and skill building as part of their home visiting programs for high risk families.
Approaches to promoting and strengthening marriage include creating incentives for participation in pre-marital education, providing marriage education for adults, and using TANF funds for marriage and relationship preparation for low-income and other families.
Seven States are undertaking improvements in their State-level divorce and marriage data in order to better measure outcomes relating to strengthening marriage.
Other activities include school-based marriage education, marriage emphasis in fatherhood programs and in programs for incarcerated parents, respite care programs for parents of children with disabilities, and relationship strengthening components in home-visiting and other mentoring and counseling programs for high-risk families.

ASPE also funded a project with the Urban Institute, which began in September 2000, entitled The Link Between Marriage and Low-Income Family Well-Being. This study will examine the interactions between family formation status and economic well-being, and the extent to which marriage is a protective factor against economic hardship. The study is examining various types of family formation, including single, married and cohabiting parents, and observe measures of poverty as well as material hardship. Utilizing various data sources, the project will produce four reports. The first report, Marriage as a Protective Factor Against Economic Hardship was completed in August 2001 and found that, overall, marriage reduced the likelihood that parents would experience poverty or material hardship. The second and third reports, Married and Unmarried Parenthood and the Economic Well-Being of Families and The Relationship Between Marriage and Other Family Structures and the Material Hardship Experienced by Households with Children, were completed in August 2002.

ACF and ASPE also are jointly funding evaluation of the Partners for Fragile Families demonstrations in nine States in order to document the effects of interventions on poor, young, unwed parents. ASPE and ACF also have funded additional analysis of data from the Fragile Families and Child Well-Being Survey, a longitudinal survey of new unwed parents beginning at the birth of their child. To build the knowledge base further, ACF awarded a contract to Abt Associates, Inc., in FY 2001 to conduct an in-depth review of the literature on the patterns, determinants, and correlates of marriage and cohabitation to family formation decisions among low-income populations, including quality and stability of relationships. Work under the contract will also focus on knowledge gaps and areas in which further research is most needed.

In FY 2002, ACF expects to award a contract that will provide a comprehensive examination of a broad spectrum of marriage-related services, from premarital counseling and assessment and marriage enrichment services to divorce prevention services. The project will explore the potential for evaluating such services by identifying relevant funding streams, examining alternative organizational missions, and assessing receptivity to evaluation among key stakeholders and opinion leaders. To address a broader range of issues, ACF will fund a contract to provide technical assistance, performance measurement, and community impact assessment for a demonstration initiative on marriage, fatherhood and family formation. The demonstration will be designed to develop comprehensive, community-level intervention strategies that promote and maintain healthy marriages, fatherhood, and family formation, and that reduce the number of non-marital births.

Employment Retention and Advancement

Over the last eight years ACF has committed research funds to address varied issues related to increasing employment among welfare recipients. ACF’s latest major initiative in this area, The Employment Retention and Advancement Evaluation (ERA), builds on earlier experience in order to test experimentally a new generation of approaches to promoting employment retention and advancement.

The goal of this multi-year demonstration and evaluation project is to gain knowledge about how best to help low-income families sustain attachment to, and advancement in, the labor market. ACF and the contractor, MDRC, have worked with participating development sites to fully implement their programs. By the end of 2002, we expect a total of 15 intervention strategies in nine States (California, Illinois, Minnesota, New York, Ohio, Oregon, South Carolina, Tennessee and Texas) to have begun operations. A first report on the ERA project was released in October 2002.

All sites are randomly assigning participants to control and experimental groups, and several important variations on the retention and advancement themes are being tried:

Cleveland is experimenting with a cooperative effort between employers and the TANF agency that offers career progression.
New York will attempt to enhance employment retention and advancement among substance abusers.
Texas is offering significant cash incentives to increase participation in its post-employment services program.
Minnesota is providing enhanced services for participants with acute mental health problems.
Portland, Oregon is targeting participants who experience repeated employment loss and cycle through the welfare system multiple times.
South Carolina is aggressively reaching out to former welfare recipients who have been off the rolls for at least nine months, attempting to locate and work with ex-TANF recipients who may need employment-related assistance in order to avoid recidivism.
Los Angeles, California is testing a variation of "work first" in the form of enhanced job club activities.

Effective Strategies for Serving the Hard-to-Employ

State and local TANF officials and other service providers have expressed the need for more information and guidance as they develop employment-focused strategies to work more effectively with TANF recipients who face substantial barriers to employment. These include adults with substance abuse and/or mental health problems, physical or developmental disabilities, learning disabilities or very low basic skills, who have experienced domestic violence, or have a general history of low and intermittent employment. In many instances, agencies will need new methods and strategies to meet the needs of individuals facing one or more of these barriers in order for them to enter and succeed in the labor market.

ASPE, ACF, and the Department of Labor are funding a major evaluation project that builds on lessons from earlier work and is intended to increase knowledge about the most effective strategies for helping hard-to-employ parents find and sustain employment and improve family and child well-being. The project is a multi-year, multi-site effort that will begin by identifying agencies and organizations already working or interested in working with such parents and helping those agencies design and implement programs that address barriers to employment by using state-of-the art methods and approaches. The evaluation will utilize an experimental design to assess program effectiveness and will document the implementation and operational lessons from the perspective of program operators, administrators, and participants.

Rural Welfare-to-Work Strategies

ACF is investing resources to learn how best to help TANF and other low-income rural families enter into and sustain employment. This study will help identify effective rural welfare-to-work strategies, operational challenges, and solutions that can be used by State and local TANF agencies and others.

In FY 2000, ACF awarded a contract to Mathematica Policy Research, Inc., for an evaluation of rural welfare-to-work strategies employing random assignment, experimental design. Illinois, Nebraska, and Tennessee have been selected to participate in the evaluation. The evaluation will highlight promising models and determine the effectiveness and cost-benefits of various welfare-to-work strategies in rural areas.

The Illinois Future Steps Program, implemented in five counties in southern Illinois beginning in July 2001, has an intensive employment and case management program tailored to people with low incomes.
The Building Nebraska Families Program is an education-based developmental program that began in March 2002 and will work with participants in 37 rural counties throughout the State. It provides one-on-one instruction and assistance in clients’ homes focused on helping Nebraska’s TANF/Employment First clients who have not found or sustained employment through regular program activities get enhanced services.
In the Tennessee First Wheels Revolving Car Loan Program, beginning in October 2002, current and former TANF Families First participants will have the opportunity to apply for an interest-free loan to purchase a vehicle needed for transportation to work, so that caretakers can become self-sufficient and improve the quality of life for their families.

Addressing the Needs of Other Special Populations

In FY 1999, ACF and ASPE contracted with the Urban Institute to examine critical issues in the screening and assessment of TANF and/or Welfare-to-Work recipients who experience barriers to employment, with a focus on substance abuse, mental health or illness, low basic skills, physical or developmental disabilities (including learning disabilities), and domestic violence. The March 2001 report, Screening and Assessment in TANF/Welfare-to-Work: Ten Important Questions TANF Agencies and Their Partners Should Consider, discusses the important issues and challenges faced by TANF agencies and administrators as they develop screening and assessment strategies for identifying TANF recipients with significant barriers to employment. Three regional meetings were held in 2000 and 2001 to further discuss these issues, and site visits to six States and localities were undertaken to describe local approaches to screening and assessment and service delivery. Discussion of the site visits is contained in Screening and Assessment in TANF/Welfare-to-Work: Local Answers to Difficult Questions, published in December 2001.

In order to learn more about the mental health services developed by State and local welfare offices to address the mental health needs of welfare recipients, the options for delivering these services, and the key implementation challenges and lessons learned in this area, ASPE contracted with Mathematica Policy Research, Inc., to study the approaches taken in Florida, Oregon, Tennessee, and Utah. The final report, Providing Mental Health Services to TANF Recipients: Program Design Choices and Implementation Challenges in Four States, identified key program design elements such as the types of services provided, the population targeted for services, the range of personal and family challenges addressed, strategies for identifying clients in need of assistance, integration of mental health and employment services, administrative and service delivery structure, and funding. The sites provided mental health services in a variety of ways, and there was no evidence that a single model for providing these services is better than any other. Across program designs, the sites shared many of the same challenges:

Encouraging employment case managers to refer clients to mental health services;
Integrating mental health services into work activities and encouraging client participation;
Creating a professional support network for mental health staff and maintaining client confidentiality; and
Monitoring and tracking client participation in mental health services.

Evaluation of the Welfare-to-Work Grants Program

While the welfare reform law of 1996 heralded a new emphasis on work first and time limits, policy makers also acknowledged that those most likely to remain on welfare would be those with the greatest barriers to employment. Congress created a temporary program, the Welfare-to-Work (WtW) Grants Program administered by the Department of Labor, to provide additional resources targeted at getting the hardest-to-serve members of the TANF population, including the non-custodial parents of children on TANF, into work. The Balanced Budget Act of 1997 authorizing the WtW grants program also required HHS to evaluate the program’s effectiveness and report findings to Congress. ASPE, the lead for HHS on this project, has worked closely with the Office of Management and Budget, the Departments of Labor and Housing and Urban Development, and ACF in designing and implementing the five-year study which began in August 1998.

The evaluation consists of a description of the implementation status and structure of the grantees’ programs; an enhanced process and implementation study that describes implementation and operational realities and examines administrative and survey data for a sample of WtW program participants, in order to measure changes in their employment and well-being outcomes over time; and a separate process and implementation study of Tribal WtW programs. The evaluation also focuses on sub-populations of particular interest, such as non-custodial parents. Of the two evaluation reports to Congress required under the law, one has been completed and the other, a report on participant outcomes, will be published in 2003. Additional reports will provide information on program costs and effective program designs. Findings to date are as follows:

Program Structure and Service Delivery in Eleven Welfare-to-Work Grant Programs: This interim implementation report shows that WtW grant programs have been innovative despite initial delays in start-up and enrollment; that most WtW services are delivered through highly decentralized systems; that most programs serve all WtW-eligible individuals, but some focus on particular subgroups, such as non-custodial parents or those with particularly difficult barriers to work; that activities are primarily work-focused; and that most WtW administrative agencies are workforce development agencies but often have formal ties with TANF.
Serving Non-custodial Parents: A Descriptive Study of Welfare-to-Work Programs. This study examines 11 WtW grants programs with a focus on serving non-custodial parents, and finds that such programs tend to be more successful when recruitment activities are well-designed, multiple agencies collaborate in service provision, a combination of positive inducements and pressures for participation are present, and a variety of employment support services are made available.
Learning from Tribal Experience: The Evaluation of the Tribal Welfare-to-Work Grants Program. This report describes ways in which Tribes are implementing welfare reform and WtW welfare-to-work programs, including integration of services, partnerships with State TANF agencies, provision of support services such as transportation and child care, and responding to low education levels and substance abuse on reservations.
The Evaluation of the Tribal Welfare-to-Work Grants Program: Initial Implementation Findings. This report from the Tribal component of the evaluation documents the socioeconomic circumstances of Tribes, their approaches to implementing WtW and welfare reform, the kinds of work support services offered, and the ways in which WtW funding can help support child care, transportation and economic development for Tribes.
Understanding the Costs of DOL Welfare-to-Work Grants Programs. This report examines the costs of 18 WtW programs in nine in-depth evaluation sites to identify the cost structure of these programs and the factors that influenced their costs. The costs per participant in these programs ranged from $1,887 to $6,641, reflecting differences in the package of services that participants received.
Implementation of the Welfare-to-Work Program. The report describes the implementation of the WtW program in 11 sites based on information collected through two rounds of site visits in 1999 and 2001, and management information system data maintained by the programs. The programs fell into three general models for delivery of services: Enhanced Direct Employment, Developmental/Transitional Employment, and Intensive Post-Employment Skills Development.

Fifth Annual Report to Congress (February 2003)

XIII. TANF Research and Evaluation