Status Report on Research on the Outcomes of Welfare Reform, 2001. Competitive Grant Programs


Grants to States and Localities to Enhance Studies of Welfare-Related Outcomes

The purpose of these grants is 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. Grants to states are being used, for example, to add additional survey waves to measure longer-term outcomes, collect data to support greater sub-group analyses, and/or gather more detailed information on non-respondents. To be eligible, states had to have an existing survey that had been administered at least once, so that the grants can facilitate real improvements, without paying for basic startup costs. Survey findings should fill an important knowledge gap that could not be filled with states' existing data, and will cover a variety of welfare reform outcomes, such as measures of family hardship and well-being, barriers to employment, poverty status, and utilization of support programs. When measuring welfare reform outcomes, the surveys and data analyses will focus on subsets of the low-income population including long-term welfare recipients, child-only cases, former recipients, potential recipients, welfare leavers with little or no reported income, and other special populations affected by state TANF policies. The funded proposals include:

Alameda County, CA (2000)

Alameda County builds on its existing survey of current TANF recipients and TANF leavers who were interviewed at baseline and at 15 months. Under this project, researchers are conducting a 27- month follow-up survey and maintain the same detailed focus on health barriers to employment, including issues related to mental health and substance abuse. Researchers from the Public Health Institute will conduct in-person interviews and anticipate drawing a sample of 512 cases with a response rate of 72 percent. As with their earlier rounds of this survey, the data will be linked to the state's administrative data systems to gain information on demographics, earnings and program participation. The questionnaire has been finalized and is being translated into Spanish and Vietnamese. Data collection is expected to run from early spring to August 2001.

Estimated Completion Date: February 2002

Iowa (2000)

Iowa builds on an existing study of its Family Independence Program (FIP), conducted by Mathematica Policy Research (MPR) and partially funded by ASPE. Their new study, which began in September 2000, consists of two components, one focusing on vulnerable families and one focusing on longer-term outcomes. The component on vulnerable families focuses on two groups that are not clearly depicted in existing data: survey non-respondents and families who report very low incomes. The study will use intensive search techniques and other methods to conduct interviews with approximately 47 non-respondent cases from Wave 1 of their survey, targeting a response rate of roughly 60 percent. Information from these interviews will be used to assess the representativeness of survey data on welfare outcomes and the implications for interpreting findings. In their study of families with very low incomes, MPR will conduct in-depth interviews of 16 families reporting no more than $500 in total income per month, including those with no TANF and no employment, and those with low levels of TANF and/or employment. These interviews will focus on possible income sources that were missed or incorrectly measured, coping strategies and family well-being. The second component of their analysis will add an additional wave to their existing survey of welfare leavers to observe longer-term outcomes. This wave will gather information on outcomes 20 to 23 months after case closure for approximately 380 cases (assuming a response rate of 85 percent). This component of the project will also incorporate administrative data to track outcomes for approximately 950 cases. MPR has secured significant funding from foundations in addition to the ASPE grant for both components of the project.

Thus far, MPR researchers have finalized their research designs for studying a) long-term recipients, b) nonrespondents, and c) vulnerable families. The survey questionnaire for these three studies is nearly complete, and postcards have been sent to sample members as a first step in their contact and location efforts. Next steps include fielding the questionnaire and assembling administrative data on the sample members.

Estimated Completion Date: February 2002

Missouri (2000)

Missouri is building on its ASPE-funded study of former TANF recipients who left the rolls in 1996 and 1997, and is adding a cohort of recipients who have remained on TANF for at least 36 months. The study seeks to characterize and contrast the self-sufficiency outcomes and barriers for current and former TANF recipients, and to identify which factors are most predictive of successfully transitioning off welfare, as well as those characteristics most predictive of exhausting the time limit. The existing study (funded in FYs 1998 and 1999) follows a sample of 1,200 former recipients, and the new cohort of stayers will consist of 400 cases. Survey data will be linked with administrative data on TANF, food stamps, child care, Medicaid, and some community-based assistance. Thus far, the contractor for the study, the Midwest Research Institute, has selected the sample of long-term recipients and developed and pre-tested the questionnaire on barriers to work. Findings from both the leavers and stayers studies will be combined in the final report.

Estimated Completion Date: December 2001

San Mateo County, CA (2000)

This study will use both administrative data and survey data to study child-only cases, including cases that have left TANF and those that remain on the rolls. The study seeks to better understand the characteristics and outcomes of these families, many of whom are headed by immigrant parents. The study will take place in San Mateo and Santa Clara counties, and is being conducted by the SPHERE Institute. Their survey targets a response rate of 70 percent among leavers for approximately 430 cases, and 80 percent among stayers for approximately 750 cases. Their study also will draw on administrative data from county case files, wage records and Medicaid eligibility data. Thus far, researchers have developed their research design, coordinated with various county offices, developed the survey instrument, and determined what administrative data will be needed. Next steps include programming the survey questions into the computerized interviewing system, identifying the population of child-only cases, selecting a sample, and initiating contact and location efforts.

Estimated Completion Date: August 2001

Wisconsin (2000)

This project will add a third wave of interviews to the Institute for Research on Poverty's existing study of a cohort of TANF applicants in Milwaukee, Wisconsin. The new results will reflect outcomes for this cohort approximately two years after the baseline data were collected. The study is based on a sample of applicants, and the survey will contain results for those who have left TANF, those still receiving TANF, and some who never received TANF. Adding a third wave to the applicant survey will support analysis of a significant number of cases who have reached the time limit, and a significant number of cases who have cycled off and on the rolls. The researchers will examine a large number of outcomes related to employment, well-being and program participation. They anticipate that of the 1200 respondents from Wave 1, approximately 900 will complete interviews for Wave 3.

Estimated Completion Date: December 2001

South Carolina Welfare Outcomes Grant (1998 and 2000)

This project continues ASPE's support of a multi-year effort by South Carolina's Office of Budget and Control Board's Office of Research and Statistics to link administrative data and additional data from surveys of former welfare recipients. The funds provided through an ACF cooperative agreement will allow South Carolina to continue its contract for the expansion of the follow-up studies. The first report on welfare leavers has been received(1), as well as a draft report which focuses on those diverted from cash assistance. A longer-term follow-up (at 24 and 36 months) of welfare leavers and divertees will be the final report of this project.

Estimated Completion Date: May 2002

Researcher Initiated Grants on Welfare Outcomes

In FY 1999 ASPE awarded approximately $807,000 in grants in FY 1999 to support seven researcher-initiated proposals to study important questions related to the outcomes of welfare reform. Through these grants, we are supporting efforts to analyze a variety of information about low-income individuals (both adults and children) and their families, including their economic and non-economic well-being and their participation in government programs. Issues that are being examined under these grants include caseload dynamics, the impact of spatial distribution of economic opportunities, health insurance and health care utilization, the use of food stamps, living arrangements, maternal and child health, domestic violence, and quality-of-life issues.

In FY 2000 we continued this grant program, in cooperation with the Administration for Children and Families, focusing on use of state and federal administrative data, and on current and former TANF recipients and other special populations affected by state TANF policies. Priority research interests centered on issues likely to be of concern during TANF reauthorization discussions, including the composition of the caseload, patterns of government program use, sub-populations, non-working welfare leavers, sanctions, employment stability, marriage and family structure, TANF flexibility, barrier identification and service utilization, and entry effects and welfare dynamics. Approximately $1.3 million was awarded to 10 applicants. In general, ASPE funding is supporting research and secondary data analysis efforts that will be completed within 12 months covering a variety of information about adults, children, and families, including economic and non-economic well-being and participation in government programs. ACF awarded an additional $1.2 million in FY 2000 to support continuation of two of the projects beyond this first year and seven other longer-term projects involving primary data collection. Brief descriptions of the ASPE-funded projects follow. When available, final reports from the grantees will be posted on the ASPE website at <> or you may fax us your request c/o Ethel Norris, fax: (202) 690-6562.

RAND: Entry, Exit and the Changing Composition of the Caseload (2000)

This project is exploring the role of the economy in explaining the welfare caseload declines. It will address the following questions: 1) What is the relative importance of changes in the rates of entry, exit, and re-entry in explaining the observed caseload declines? 2) What is the role of expenditures on welfare programs in explaining these declines? 3) To what extent is the caseload becoming harder to serve as the total caseload declines? Researchers will also explore how the answers to these questions vary by race-ethnicity (white, black, Hispanic, Asian), and welfare program (two-parent, one-parent, child-only.) The project will use California administrative data from 1987 through mid-2001.

Estimated Completion Date: September 2001

Baruch College, City University of New York: Effects of Welfare Reform on Investments in Human Capital and Family Formation (2000)

This study is investigating whether the behavior of teens and young adults ages 16 to 21 has changed as the result of welfare reform. Researchers will use data from the 1979 and 1997 National Longitudinal Surveys of Youth (NLSY) to compare cohorts (both between and within) that entered these ages prior to and following welfare reform, describing differences in outcomes and behaviors such as high school completion, teenage and non-marital child bearing, employment and welfare receipt. They will then investigate the role of welfare reform in bringing about the observed changes.

Estimated Completion Date: February 2002

University of Oregon: TANF and Household Savings (2000)

This project studies the impact of new savings incentives offered to participants in the TANF program. Specifically, researchers are addressing the following questions: 1) Has saving increased among those low-income households who reside in states that have increased the liquid-asset and vehicle equity limits for program eligibility? 2) Has saving increased among those low-income households who reside in states that have introduced Individual Development Accounts? 3) What is the impact of time-limited benefits on household savings? 4) Are there differences by race, marital status, and poverty status in the response to the new saving incentives? The study uses data from the 1989, 1994, and 1999 wealth supplements of the Panel Study of Income Dynamics. A first draft of a paper has been produced.

Estimated Completion Date: August 2001

University of Michigan: Mental Health, Substance Abuse, and Domestic Violence Service Utilization by Welfare Recipients (2000)

This project is analyzing the impact that spatial proximity to social service providers and individual-level characteristics have on service utilization rates among welfare recipients in the three-county Detroit metropolitan area. The project will address the following questions: 1) How are social service providers spatially distributed in the Detroit metropolitan area? 2) Where do welfare recipients live relative to the location of mental health, substance abuse, and domestic violence services? 3) Are service utilization rates correlated with spatial proximity to providers?

Researchers will use data from the Mother's Well-Being Study (MWS), a survey of welfare recipients in the Detroit metro area, and link data from the MWS to data on the geographic location of mental health and substance providers.

Estimated Completion Date: Summer 2001

Manpower Demonstration Research Corporation (MDRC): An Analysis of Caseload Composition and the Non-Working Welfare Leavers (2000)

Researchers are examining three groups of low-income populations (those who leave welfare for work, those who remain on the welfare rolls, and non-working welfare leavers) to address the following questions: 1) In what ways are the families who remain on welfare different than the ones who have left? 2) What are the characteristics and circumstances of people who leave welfare and are not working? The project uses data from seven evaluation studies of welfare programs (six used random assignment) conducted by MDRC.

Estimated Completion Date: September 2001

Case Western Reserve University: The Effect of Job Accessibility and Neighborhood Characteristics on the Employment Stability of Welfare Leavers in an Urban Labor Market (2000)

The study is examining womencs employment stability, earnings, and wage trajectories over a 13-month period following their exit from TANF in the Cleveland metropolitan area. The following questions will be addressed: 1) What is the geographic distribution of jobs held by women leaving welfare, and do labor market success and job stability differ by whether jobs are located in high job growth or slow job growth sections of the metro area? 2) How residentially mobile are former welfare recipients once they have gained employment, and is that mobility related to the location of their jobs? 3) How do the residential locations of former welfare recipients and their proximity to entry-level job openings and the implied access to public transportation affect their labor market success? 4) How do the social and economic conditions in their residential neighborhood affect labor market outcomes for women leaving welfare?

Researchers will use three data sets: an ongoing longitudinal study of women leaving welfare, a regional labor market data set, and a database containing measures of neighborhood distress.

Estimated Completion Date: September 2001

Urban Institute: The Link Between Marriage and Low-Income Family Well-Being (formerly How Important is Marriage to Low-Income Family Well-Being?) (2000)

This project, which began in September 2000, is examining the interactions between marital status, household status, and economic well-being to better understand whether increases in marriage among the low-income population would increase economic security and reduce poverty. The primary research question is: Does marriage between two biological parents, as well as other family forms, bestow economic benefits and other advantages to families with children over other family types, including single parent families and other families headed by individuals with low educational attainment and/or low earnings capacity? The study will examine conventional family types, such as two-natural- parent married family, step-family, cohabiting families and single parent families, as well as an expanded set of family types that include visiting relationships and extended families. Both economic and non-economic outcomes will be considered. Researchers have assembled the data (the 1997 and 1999 rounds of the National Survey of America's Families), constructed the necessary variables, and tabulated basic descriptive statistics regarding the incidence of various family types of family unions and measures of well-being. They are currently working on more complicated analyses and gathering and reading articles for the literature review.

Estimated Completion Date: September 2001

Columbia University: Fragile Families and Welfare Reform (joint with ACF) (2000)

This study will describe the conditions and capabilities of vulnerable mothers and fathers in the first few years following enactment of PRWORA and begin an evaluation of the impact of TANF and child support policies. Specifically, researchers will document the composition of the actual and eligible welfare caseload, how unwed mothers are packaging various forms of support and government programs, and how well families are doing as a result of individual efforts and social policies. Researchers will also conduct subgroup analyses on teenage parents and immigrants. Researchers will use data from the Fragile Families study, a random sample of new unmarried mothers and fathers in 20 large cities across the United States. Currently, researchers are gathering data and working on the first report under this grant, a baseline report that will provide demographic and descriptive statistics, as well as compare mothers who are receiving public assistance to those who are not. The first report, including baseline data and some descriptive statistics, is expected to be released in Fall 2001.

Estimated Completion Date: Summer 2003

Washington University: Employment, Earnings and Recidivism: How do Entrants to TANF Differ from Entrants to AFDC? (2000)

The project will examine factors related to welfare exits, employment stability, earnings mobility, and recidivism among welfare recipients in North Carolina, comparing the experiences of black, white and Hispanic AFDC and TANF participants. Specifically, the study will report on the demographics, welfare participation, employment retention, and post-exit earnings of five cohorts of welfare recipients in North Carolina. It will compare outcomes for those who entered welfare before TANF (1995), in the early implementation of TANF (1996 and 1997) and in the later stages of TANF implementation (1998 and 1999). It will also report on the longer-term labor market outcomes of the earlier cohorts, as well as the types of jobs AFDC/TANF recipients in North Carolina obtain, the range of wages for these jobs, and the potential for on-the-job skill development. The project will use state and county administrative data from North Carolina.

Estimated Completion Date: December 2001

UCLA/RAND: A Proposal to Examine the Reporting of Welfare Benefits in the SIPP Using Matched Administrative Records in California (joint with ACF) (2000)

This study examines the accuracy of self-reports of program participation in survey data. In particular, researchers are comparing self-reported program participation among Californians interviewed in the Survey of Income and Program Participation (SIPP) with California administrative files of program participation for the same individuals. Researchers will document the degree of misreporting in a variety of programs, including AFDC/TANF, Medicaid, and Food Stamps, and investigate the implications of misreporting for conclusions about the dynamics of welfare participation.

Estimated Completion Date: September 2001

University of Illinois: Young Mothers' Transitions On and Off TANF: How do Child Care Assistance, Job Training, and Social Supports Influence These Decisions? (2000)

This project will identify the likelihood that Ayoung mothers will go on, stay on, leave, and stay off TANF@ given use and/or availability of child care, job training, and other social programs in their community. The research will analyze three subgroups of young mothers (ages 18-24) who lived in the Chicago metropolitan area between January 1, 1997, and June 30, 2000. There are three major components of the study: 1) geographic analyses of local area job-related resources, 2) event history models of TANF participation, and 3) process-oriented models of TANF participation. These components will utilize ZIP-code level data on the availability of child care, job training, and other social services; state administrative data to examine when mothers received TANF; and detailed questionnaire data.

Estimated Completion Date: September 2001

UCLA/University of Wisconsin: The Effects of the Work Pays Demonstration, EITC Expansions and the Business Cycle on the Labor Market Behavior of the California Caseload (2000)

This project is will examine the effect of: 1) welfare changes, 2) the 1990 and 1993 expansions of the EITC, and 3) changes in the business cycle on three specific issues concerning the California welfare population. These issues include: 1) how do these factors contribute to the economic well-being of families; 2) how do they affect labor market and transfer program participation; and 3) how do they affect employment changes and earnings trajectories? The project will use California administrative data drawn from the welfare, unemployment insurance, and tax systems.

Estimated Completion Date: September 2001

University of Wisconsin: Toward Understanding the Longitudinal Health Insurance and Food Stamp Status of Short- and Long-Term Welfare ALeavers@ (1999)

This study is relying on Wisconsin administrative data to examine the apparent eligibility for and participation in Medicaid and, to the extent possible, private health insurance, among two samples of AFDC and TANF clients: those who were receiving AFDC in September 1995 and those who were receiving Wisconsin Works (W-2) in September 1997. A similar analysis will be done with regard to receipt of food stamps. The research will focus on longitudinal use through December 1999 of Medicaid and food stamps by three groups of AFDC and W-2 participants (AFDC and W-2 Aleavers,@ Arecidivists,@ and Astayers@) across a variety of subgroup definitions, including: location; race; pregnancy; age of children; number of children; SSI recipiency; level of earned income; apparent eligibility for Medicaid and food stamps; and apparent access to private health insurance. Because of the wide range of Medicaid-funded health insurance programs, each with a different set of eligibility criteria (e.g., SCHIP, Medicaid, spenddown), it should be possible to assess how the existence of these programs affects the likelihood of Medicaid uptake among different population subgroups. In addition, for those persons for whom a wage record exists, researchers will obtain information about whether the employer offers health insurance to employees, and construct a variable to estimate the length of continuous employment with the same employer which will permit them to estimate whether a family has access to private health insurance.

Early findings concerning the determinants of food stamp take-up include: families who live in Milwaukee, have mothers with less education, are non-white, have more children, or have a longer welfare history, are more likely to take up food stamps than other families. Selected findings concerning the determinants of health insurance coverage include: children who are older, have younger siblings, moved more frequently, or have mothers who work full time are more likely to be uninsured at some point in the year. The data also indicate that children are more likely to be privately insured if the child's mother has higher educational attainment, is currently married, or works full time. Children from families living outside of Milwaukee, and with family income greater than the poverty threshold are also more likely to be privately insured. Use of public insurance is more likely among children who are young, living in Milwaukee, from families with income below the poverty threshold, hand have unmarried mothers. Future work will attempt to determine which of these characteristics are the most salient.

Estimated Completion Date: September 2001

Continuation of 1998 Grants to States and Localities to Study Welfare Outcomes (1998 and 1999)

Thirteen states and large counties(2) were awarded approximately $2.9 million in grants in FY 1998 to study the outcomes of welfare reform on individuals and families who leave welfare. Some of the grants also included studies of families who applied for cash welfare but never enrolled and families who appear to be eligible but not enrolled. Three grantees — Arizona, Missouri, and a consortium of San Mateo, Santa Cruz, and Santa Clara counties — received additional funding in FY 1999 to extend the studies and administer a second or third wave of interviews, allowing analysis of longer-term outcomes for former recipients. For individual project descriptions and links to available reports, see <>.

As of March 2001, all 15 studies had released preliminary reports based on administrative data findings, and 12 of the 15 also had released reports with more detailed findings from follow-up surveys. Highlights from these reports are presented in Chapter II, with a focus on outcomes in employment and earnings, recidivism and program participation, and household income and family well-being.(3) Despite some methodological differences in study design, as noted below, these reports show a surprising amount of consistency among findings across sites and different TANF programs, particularly in the areas of employment and recidivism.

ASPE's research strategy combined local flexibility in study design with some national direction and coordination. Most of the projects used administrative data to track an early cohort of individuals who left welfare around 1996 or 1997. Projects also used a combination of administrative and survey data to track the economic status and general well-being of at least one cohort who left welfare one to two years later, after the transition from AFDC to the TANF program. Projects varied in the number and types of administrative data sets examined and the design of the surveys of former recipients. Final survey sample sizes varied from 277 to over 3,500 cases, response rates ranged from 23 to 81 percent, and approximate time of interview varied from 6 to 30 months after exit. All researchers were encouraged to collect data across multiple dimensions, including employment, program participation, economic status, family structure, child well-being, material hardship, barriers to employment, etc. Grantees designed their own survey instruments, however, which differed in wording and emphasis. While this diversity poses challenges for summarizing results nationally, it has allowed states to meet the demands of elected officials for timely information on families leaving their state's welfare program.

Estimated Completion Dates: vary by project

Grants to States and Localities to Study Welfare Reform Outcomes, with an Emphasis on Diversion (1999)

One of the Congress's major objectives in providing welfare outcomes money to ASPE over the last several years is to 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.(4)  To this end, ASPE issued a request for applications from states and large counties in April 1999 with an emphasis on the study of applicants and potential applicants to the TANF program. ASPE awarded seven grants under this announcement, six of which specifically support state efforts to gather a variety of information about individuals and their families who are formally or informally diverted from TANF. In addition, several of the leavers studies funded in FY 1998 have significant applicant components to their projects.(5)

ASPE is particularly interested in learning about 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. Below are summaries of the grants provided to states and large counties in fiscal years 1998 and 1999 with a particular emphasis on TANF diversion. "Diversion" this context is not limited to participation in formal diversion programs, but also includes "informal" divertees. These are usually defined as individuals who began the application process but were either deemed ineligible for non-monetary reasons, withdrew voluntarily after completing the process, or failed to complete the process for some other reason. These project updates are current as of March 2001.

Arizona (1999)

Arizona is building on their FY 1998 study of leavers by looking at informal divertees and entrants to TANF. The study uses a wide range of administrative data (including data on child care subsidies) to track second quarter 1999 divertees and recipients for 12 months, and includes surveys of 400 individuals in both populations at three and nine months after the application period (second quarter of 1999). Some of the subgroups on which the state will be focusing include urban vs. rural applicants and applicants who are initially denied but eventually reapply for TANF.

Arizona has collected administrative data from a number of different sources, including a data warehouse established as part of the FY 1998 ASPE leavers grant. The first wave of the survey resulted in a response rate of 71 percent. In the second wave of surveys, completed in September 2000, researchers were able to find 85-90 percent of individuals interviewed during the first wave. The state expects to submit a draft of the final report, as well as the public use data files for the project to ASPE in Summer 2001.

Estimated Completion Date: Summer 2001

Contra Costa County and Alameda County, California (1999)

Contra Costa and Alameda Counties are located in the East San Francisco Bay area of California and contain the cities of Oakland and Richmond. This project is studying TANF leavers from both counties, as well as formal and informal divertees in Contra Costa County. Researchers at the SPHERE Institute have been able to take advantage of these counties' Case Data System (CDS), which includes every TANF application that is initiated in the two counties. The CDS allows SPHERE to uncover the reasons individuals were diverted or left TANF, as well as make comparisons across the two counties. They used the CDS both to link all applicants with other administrative databases and to draw their survey sample of 850 leavers and 150 divertees from the third quarter of 1999. The research plan calls for comparisons between divertees and leavers and between the two counties.

The first wave of surveys was administered at six months after exit/diversion, with a response rate for informally diverted families in Contra Costa County of 64 percent. The second round of survey data collection was completed in November 2000, with a response rate of 54 percent. A preliminary draft of findings from both rounds of survey data and merged administrative data should be available in Summer 2001.

Estimated Completion Date: Summer 2001

Illinois (1999)

Illinois is focusing this study on applicants; the state is particularly interested in learning about families who fail to complete the application process. The population to be studied includes one month of approved, denied, and withdrawn applications. Although the state has no formal diversion policy, the study will assess Illinois' new intake process, which emphasizes employment, assessment, and prompt referral to needed services. Administrative data analysis is ongoing for the entire population of approximately 6,000 families, and a survey is being administered to the sample of 1,200 divertees approximately two to four months after application. The study also includes surveys of program administrators at six local welfare offices to help evaluate the new intake process.

After receiving the ASPE grant, Illinois issued a request for proposals for contractor assistance in conducting the study and selected MAXIMUS as the contractor. Survey administration is currently ongoing, and initial findings from the project are expected by Fall 2001.

Estimated Completion Date: Fall 2001

New York (1999)

New York, which also received a FY 1998 leavers grant from ASPE, has included divertees, all other denials, and entrants in their sample for this study. Their analysis will focus on comparing TANF applicants who were diverted with those who received cash assistance. Twenty-one local districts are participating in the study, including New York City and other sites ranging from large urban to rural areas. In most districts, the project uses administrative data to track a March 2000 sample of divertees, denials, and entrants for 12 months after the application.

The sample was drawn through intercept interviews with TANF applicants in each of the local districts. This methodology allowed New York to include individuals who entered the TANF office with the intent to apply but who did not submit written applications. The state' s contractor, ORC Macro, is currently administering the survey to the sample of 864 families, evenly split between diverted applicants and entrants. Their goal is a response rate of 70 to 75 percent. The state expects to report results in Summer 2001.

Estimated Completion Date: Summer 2001

Texas (1999)

This project represents the combined efforts of the Texas Department of Human Services, the Texas Workforce Commission, and the University of Texas-Austin. It focuses not only on informal divertees, but also on potential TANF applicants who are formally diverted by the state, either through a one-time lump sum payment or by redirection into work. The administrative data analysis incorporates a wide variety of sources, and tracks both applicants who are redirected into work or denied for non-financial reasons and participants in the lump-sum diversion program. The state has also added leavers to both the survey and administrative samples.

Intercept surveys were conducted with 30 applicants who were denied TANF for non-financial reasons, with a second wave survey currently being administered at four months after application. The interviews with leavers, redirects, and formal divertees took place in 1999 and early 2000. Follow-up on individuals in these samples will be done using administrative data. A preliminary report was issued in early April 2001, with a final report anticipated in late Summer 2001.

Estimated Completion Date: Summer 2001

Washington (1999)

Following up on the leavers grant that they received in FY 1998, Washington is studying formal and informal divertees and entrants. The state hopes to compare the experiences of individuals who participated in the state' s Diversion Cash Assistance program, those who entered TANF, and those who were diverted and received assistance from neither program. They will be providing an analysis of administrative data for the full populations of each of these groups from the fourth quarters of 1997, 1998, and 1999, including data from up to 12 months prior to and 12 months after the selection quarter. The state has nearly completed administrative data collection.

The state completed its survey of individuals who applied for TANF or Diversion Cash Assistance between July and October of 1999. The survey effort, which was completed in May 2000 and took place between four and eight months after the time of application, resulted in a response rate of 84 percent. The state researchers hope to provide ASPE with a draft report containing both survey and administrative data by Summer 2001.

Estimated Completion Date: Summer 2001

Wisconsin (1998)

This study of individuals applying for Wisconsin Works (W-2) assistance in Milwaukee has been undertaken by the Institute for Research on Poverty (IRP) at the University of Wisconsin at Madison. This portion of the study focuses on three subgroups of applicants: those who request assistance and subsequently participate in the W-2 program, those who request assistance but are determined to be ineligible for program participation, and those who request assistance, appear to be eligible, but do not participate in W-2. A six-month cohort of applicants is being tracked through a combination of linked administrative data (e.g., public assistance, quarterly earnings, child support, foster care, and mental health data) and three waves of surveys, the third being funded by ASPE in FY 2000.

The first two waves of the survey, conducted at the time of application and 12 months afterward, are complete, and IRP is currently analyzing the survey results and using state administrative data to supplement the survey data. A report detailing findings from the first wave of the survey has been completed and will be posted at <>.

Estimated Completion Date: Spring 2002

Leavers Studies that also Examine Diverted Populations (Florida, San Mateo, and South Carolina) (1998)

Three FY 1998 grants that have a primary focus of studying outcomes for families leaving welfare also include research on families that were formally or informally diverted from entering TANF. These three grants are Florida, a consortium of California counties, and South Carolina.

The Florida study, undertaken by researchers at Florida State University, examines three groups of individuals from the second quarter of 1997: TANF leavers, individuals who began the application process but who either withdrew voluntarily after completing the process or failed to complete the process ("diverts" ), and individuals who receive food stamps or Medicaid, have minor children, and have income and assets below the cash assistance limit but who do not receive cash assistance ("opt-nots"). Findings from Florida' s final report, released in November 2000, indicated that the "diverts" and "opt-nots" look very similar to leavers in terms of employment rates. However, leavers appear to have slightly higher earnings and slightly lower use of government services than the other two groups.

The study in San Mateo, Santa Clara and Santa Cruz counties in California was funded primarily as a leavers study. However, because the administrative system in all three counties includes all applications, and not just those for individuals who receive TANF, researchers at the SPHERE Institute were able to also study informal divertees. Analysis of administrative data is being supplemented by surveys administered at six, 12 and 18 months after "case closure" (when either the applicant withdraws from the application process or the TANF recipient leaves the program). A draft report summarizing the first wave of survey data along with linked administrative data was released in December 2000. At six months after case closure, leavers were better off than informal divertees in terms of household earnings, receipt of health insurance, and returns to cash aid. However, divertees had fewer barriers to employment and incidences of other hardships, such as food insecurity and child risk behaviors, than leavers. A final report from SPHERE incorporating the second and third waves of survey data is expected in Summer 2001.

The project in South Carolina also is focused primarily on leavers, but state researchers have also used food stamp records to identify families that appeared to be eligible for cash assistance but were not enrolled. Surveys were conducted with families who went on food stamps between October 1998 and March 1999 and who did not apply for TANF at any time in the following one year period. Each of these families had dependent children and was eligible for TANF based on gross income, but did not enroll in the TANF program. The state and their contractor, MAXIMUS, achieved a response rate for the surveys of 71 percent. A report detailing South Carolina' s findings has been completed and will be posted at <> .

Estimated Completion Dates: Complete (Florida and South Carolina); Summer 2001 (San Mateo)