Status of Research on the Outcomes of Welfare Reform, 2000. Activities that Support Conferees' Recommendations


Collection and Use of State-Specific Surveys and State and Federal Administrative Data

Activities that support and enhance the collection and use of federal and state administrative data and state-specific surveys are critical to monitoring and understanding the outcomes of welfare reform. ASPE has devoted substantial resources to supporting state-level data collection efforts and administrative data linking, making certain that national survey instruments are responsive to policy changes and needs, supporting and maintaining a wide range of Census Bureau data collection efforts, and supporting secondary analyses of state and national-level data to add to our understanding of the effects of welfare reform.

For example, the states and counties awarded grants to study welfare reform outcomes in FYs 1998 and 1999 are using a combination of linked administrative data and surveys to study welfare reform's outcomes on families leaving the TANF program. A separate FY 2000 project will support the storage of the research data sets that combine the state-specific administrative and survey data they have collected in a controlled environment where confidentiality can be protected. In FY 1999 we also awarded grants to states and large counties to support efforts to gather a variety of information about individuals and their families who are formally or informally diverted from TANF. Through a combination of administrative data and surveys, these projects are assessing the degree to which TANF applicants receive, or are aware of their potential eligibility for, Medicaid, food stamps, and other programs and services that are important in helping low-income families make a successful transition to work.

We have provided support in both FYs 1999 and 2000 for a project designed to match Social Security earnings records with samples of adult welfare recipients and non-recipients from Census surveys to help assess employment and earnings patterns and outcomes on the basis of baseline characteristics. We have also provided funds to offset one-time infrastructure costs and programming time associated with using the National Directory of New Hires (NDNH) database to link with samples drawn for research projects to improve the quality of information on employment outcomes. In cooperation with the Census Bureau, we convened a group of researchers, policy makers and local practitioners to explore potential uses of the new American Community Survey (ACS) data and to explore the development of econometric models that combine ACS data with local area administrative data and local business economic data to provide local area data.

In cooperation with the Administration for Children and Families, we funded a total of 17 researcher-initiated grants in FYs 1999 and 2000 to study various aspects of welfare outcomes. These grants focus primarily on secondary analyses of state and federal administrative data about low-income individuals (both adults and children) and their families. A separate project will use California administrative data from the welfare, unemployment insurance and tax systems to examine the effect of welfare changes, EITC expansions, and business cycles on the economic well-being of families, participation in the labor market and/or transfer programs, and employment changes and earnings trajectories.

We also funded a project in FY 2000 that will use information from the Current Population Survey (CPS), Panel Study of Income Dynamics (PSID), the Survey of Income and Program Participation (SIPP) and administrative data to compile a book of tables showing trends in income, poverty and other economic measures, such as access to health insurance and food and housing security. Separately, we funded a project to analyze data from the Urban Institute's National Survey of America's Families, supplemented with other national data sets like the CPS and SIPP, to address the effects of state TANF policies, including whether they have resulted in unmet needs for services. Data from the National Survey of Family Growth will be used to create simulation models that may clarify which factors are associated with changes in teen pregnancy and births, and how possible future changes in these factors might affect teen pregnancy and birth rates. Projects such as these support the overall goal of ensuring that good data are available to create an integrated picture of the low income population, especially families with children, in the wake of welfare reform.

Improve the Capabilities and Comparability of Data Collection Efforts

As stipulated in the FY 1998 appropriations language accompanying the targeted policy research funds to study welfare outcomes, the Department submitted its research plan to the National Academy of Sciences (NAS) for guidance on research design and recommendations for future research on the effects of welfare policy changes. We have provided funding in FYs 1998, 1999 and 2000 for the 30-month NAS Welfare Outcomes Panel Study which was established to evaluate the design of current, proposed and future studies of the effects of the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) of 1996. In its Interim Report, released in September 1999, the panel offered the Department an initial set of short-run recommendations (2) for welfare evaluation strategies, including placing a high priority on improving the capabilities of data collection efforts and working toward cross-state and within-state comparability of data and research on welfare program effects. Our strategy of providing grants to states and local TANF agencies is explicitly designed to increase state and local capacity for data collection efforts. In addition, we have been working to facilitate greater comparability in state and local level studies through grantee meetings, list-serve discussions, guidance to states and a technical assistance contract.

In FY 2000 we awarded competitive grants to five states and counties specifically to enhance state-specific surveys of populations affected by welfare reform, by expanding or improving data collection activities, including efforts to improve cross-state comparability. These grants were designed to facilitate real improvements of existing or close-to-final surveys, without paying for basic start-up costs, and will be used by the grantees to add additional waves of data collection, raise response rates, analyze non-response bias, increase sample sizes, and expand the content of instruments. We also committed FY 2000 funds to an effort to improve cross-study comparisons of the prevalence of multiple barriers to employment, by critically assessing current studies that measure barriers to employment and convening experts to address issues in identifying an optimal caseload survey design, with the goal of laying out the scope and content of a "model" caseload survey of welfare recipients.

We also funded two technical assistance projects in both FYs 1999 and 2000. One project funds a contractor to provide technical assistance to the welfare outcomes grantees. The focus of this year's technical assistance under that contract is to improve the quality, comparability, documentation and accessibility of the research data sets each grantee will be preparing and submitting from the state-specific administrative and survey data they have collected on former, current, and potential TANF recipients. The other project is devoted to providing researcher technical assistance to states that received grants to promote child indicator work. Technical assistance has been provided on conceptual and methodological issues in identifying and measuring child health and well-being indicators within and across states and ways of creating or using survey and administrative data and combining several data approaches.

Develop and Report Reliable State-by-State Measures of Family Hardship and Well-Being and of the Utilization of Other Support Programs

Welfare devolution and increased flexibility in the design and delivery of program benefits have created substantial new challenges for data collection and analysis to monitor welfare outcomes. To meet these challenges, new and better data are needed at the state and local level. Over the last three years, ASPE has committed funds for several projects geared toward developing state-level data on hardship and/or program utilization.

In FY 2000, funds are being provided to support the longitudinal New Immigrant Survey which will collect data on program utilization and hardship and well-being over time among newly arriving low-income immigrant families in different states. Support is also being provided for the design of a framework for research samples to be drawn from the Federal Parent Locator Service (FPLS) database, with the goal of obtaining representative samples for all states that can identify socio-demographic characteristics and program participation data. Our support for the administration of a welfare participation question in the State and Local Area Integrated Telephone Survey (SLAITS) will yield a data element that, when combined with other data available from the survey, will permit the development of state-level estimates of the incidence of special health care needs of children of current and former welfare recipients, plus the health insurance status (including Medicaid and the State Children's Health Insurance Program, SCHIP) of current and former recipients.

The "leavers" grants awarded to states and counties in FYs 1998 and 1999 to study the outcomes of welfare reform are using a combination of linked administrative data and surveys to study welfare reform's outcomes on families leaving the TANF program. Their research questions span eight broad research topic areas, including employment and earnings, case closures and recidivism, other income supports, health insurance, child care, child well-being, barriers to self-sufficiency, insecurity/deprivation, and other topics. FY 2000 funds have been committed to synthesize the findings from these projects and conduct secondary data analyses of welfare outcomes measures drawing on the state-specific public use data sets. The "diversion" grants awarded in FY 1999 are supporting state efforts to gather information about individuals and their families who are diverted from TANF. These projects are assessing the degree to which TANF applicants receive, or are aware of their potential eligibility for, Medicaid, food stamps, and other programs and services that are important in helping low-income families make a successful transition to work. Finally, the "child indicators" grants awarded to states in FYs 1998 and 1999 promote state efforts to develop and monitor indicators of the health and well-being of children as welfare reform and other policy changes occur.

Measure Outcomes for a Broad Population of Low-Income Families, Welfare Recipients, Former Recipients, Potential Recipients, and Other Special Populations Affected by State TANF Policies, including Diversion Practices

Consistent with one of the recommendations from the National Academy of Sciences Interim Report to include a broad population of recipients, former recipients and potential recipients in our welfare reform research, studies we funded in FYs 1998 and 1999 cover a broad array of topics, including the examination of diversion practices, an assessment of those leaving the TANF rolls, and participation in other programs such as Medicaid. Projects have also been funded to study the effects of welfare reform on special populations (e.g., child-only cases, disability, substance abuse). Examples from our FY 2000 research agenda that continue and expand on this trend follow.

One study will use data from the Current Population Survey and data on industries and occupations to investigate the prevalence of alternative or contingent work arrangements (such as work through a temporary help agency, working for a contract company, or working on-call) among the low-skilled and low-income populations (including current, former and potential welfare recipients) to determine the rates and trends in nonstandard work and overlap with welfare receipt. Another will look at coping mechanisms and examine a variety of factors that may help or hinder a low-wage working family's efforts to be self-sufficient, including formal and informal support services, social support networks, time management, money management, and other life skills. A separate project will study strategies that states and communities are pursuing to provide high quality child care and other support services for welfare and working poor families.

We funded a study that will examine the way in which special populations, particularly individuals of different backgrounds and limited English-language abilities, are treated in local offices, and the extent to which program services, agency culture and caseworker discretion may lead to differences in approval rates, work assignments, access to support services, or sanctions. Another study will examine ways in which state TANF programs identify, refer and treat welfare recipients with mental health problems and assist them in obtaining treatment and finding and keeping employment. We are also providing continuing support for an evaluation of the effectiveness of a substance abuse research demonstration project that includes coordinating screening, referral and treatment with employment and training or vocational services.

Another project will investigate what is known about the effects of welfare reform on families affected by having incarcerated mothers and fathers. We are looking separately at various approaches states are taking to help children in families in which a grandparent or another relative has taken over parental responsibilities. These and other projects will help us achieve our goal of covering all important low-income population groups in our grant and contract research programs, from welfare leavers to stayers to potential applicants who are diverted or do not apply.

Support Efforts at Iowa State University to Develop State-Level Data on Low-Income Families that can be Integrated with National Data Collection Efforts

As directed by the conferees, ASPE provided funding to Iowa State University to continue support for efforts to develop state-level data on low-income families that can be integrated with national data collection efforts. Iowa State began working with ASPE in FY 1999 to develop an approach for state-level surveys that is relevant for local welfare program design, implementation, and evaluation (particularly in less densely populated areas) and that can be integrated into the Census Bureau's Survey of Program Dynamics (SPD). Current work is exploring the feasibility of extending and expanding the SPD to capture state-level reliable samples for use in exploring the outcomes of federal and state policies, as well as economic conditions of low-income families. Continued funding will support further feasibility work on the extended survey, including conducting a 20-minute telephone survey of Iowa households to pre-test a questionnaire that includes a module from the SPD along with a module on transportation issues to help address the needs for data in a rural setting. We believe this is a promising approach, given the need for state-specific questions and data, but the desire for a national framework.

These projects and the others described in later chapters of this report reflect the broad-based approach we have taken in advancing the Department's welfare outcomes research agenda. Chapter II summarizes the results and findings from projects funded in FYs 1998 and 1999. Chapter III provides brief descriptions of the FY 2000 welfare-outcomes funded projects, including more information on the studies highlighted above. Finally, Chapter IV describes the current status and target completion dates of continuing projects funded by the targeted policy research appropriations in FYs 1998 and 1999.