|
Office of the
|
|
|
|
|
|
Interim Status Report on
Research on the Outcomes of Welfare Reform:
[ Report’s Main Page |
Introduction & Overview |
1999 Projects |
1998 Projects |
This Appendix |
Appendix B ]
The grants awarded to states and counties to study the outcomes of welfare
reform on individuals and families who leave the TANF program, who apply
for cash welfare but are never enrolled because of nonfinancial eligibility
requirements or diversion programs, and/or who appear to be eligible but
are not enrolled have several attributes in common. All were funded
under a competitive announcement of availability of grant funds, and the
major purpose of all the grants is to enable grantees to track and monitor
how individuals and their families do in the first year after they leave
welfare and to provide a foundation for longer follow up.
In their project descriptions, all 14 grantees proposed to use a combination
of linked administrative data and surveys to study welfare reform’s
outcomes on families leaving the TANF program. Each study is expected
to include at least two cohorts. All grantees plan to track leavers
in linked administrative data for both cohorts. All grantees proposed
to survey sample members, generally drawn from the second cohort. Most
surveys are expected to take 20-30 minutes. Proposed sample sizes are
generally between 600 and 1,200 completed interviews. With one exception,
grantees proposed to conduct a mixed mode survey, with first contact made
by telephone when a working telephone number was available. All funded
proposals agreed either to provide a public use file, to provide limited
access to project datasets through some form of research review process,
or both.
A large number of potential subgroup analyses were proposed by grantees.
Grantees plan to use administrative data, survey data, or a combination of
both types of data to address a number of important research questions.
For the purposes of this summary, these research questions are grouped into
eight general research topic areas: employment and
earnings, case closures and recidivism,
other income supports, health insurance,
child care, child wellbeing,
barriers to selfsufficiency,
insecurity/deprivation, and other research
topics. Each of these topics is discussed below, followed by a
description of each of the 14 projects.
Almost all grantees plan to use a combination of administrative data and
survey data to examine such issues as employment status, quarterly earnings,
wage levels, hours worked, and types of jobs or occupations. In addition,
10 of the 14 grantees include questions on fringe benefits on their surveys,
and several include questions about jobrelated training. Figure
A-1 shows, by employmentrelated topic area, the number of grantees
using survey and administrative data to study each topic.
All grantees are using administrative data to study the reason for case closure,
as shown in Figure A-2. In addition, 9 of the 14 grantees are including
survey questions about the recipients’ explanation for why the case
closed. Likewise, all grantees are using administrative data to study
returns to TANF, with most grantees proposing survey questions about the
reasons for such returns.
Almost all grantees plan to study receipt of food stamps and child support,
through administrative data, survey data, or a combination of both.
All states and counties also ask questions about resources from family
members. Other types of income supports specifically addressed in proposed
surveys or proposed research plans include Supplemental Security Income (SSI),
general assistance (GA), housing assistance, energy assistance, and the Earned
Income Tax Credit (EITC). See Figure A-3 for the number of grantees
examining the various income supports using either administrative or survey
data.
Figure A-4 shows that all grantees plan to use administrative data to examine
Medicaid coverage, and most grantees also include questions about Medicaid
and employerprovided health insurance on their surveys. Only
four grantees, however, specifically discussed examining use of the
Children’s Health Insurance Program (CHIP) — two through
administrative data, and two through surveys. Over half the grantees
plan to ask former recipients if adults and children in the family have some
type of health insurance coverage.
Figure A-5 shows that most (12) of the grantees proposed including survey
questions about the child care used by families after they leave welfare.
Typical questions focus on the type of provider used, the parents’ awareness
and use of subsidies and child care costs. Half (7) of the grantees
also ask parents about their perceptions of child care quality. In
addition, six grantees are linking TANF records to the state or county subsidized
child care data system, as another source of information about the type and
cost of care, as well as the use of subsidies.
Almost all grantees proposed to study some indicators of child wellbeing,
as shown in Figure A-6. The most common source of administrative data
is the child welfare system (i.e., data on child abuse and neglect, foster
care, or both). In addition, most grantees are proposing to include
some type of survey questions about children’s living arrangements and/or
interactions with the child welfare system. Nine grantees also plan
to examine some measure of child health status (typically through a survey
question), seven grantees plan to ask questions about children’s school
attendance, and five grantees plan to include questions addressing child
behavior.
All grantees proposed to examine barriers to selfsufficiency.
Primarily through survey questions, grantees will be gathering data about
such subjects as disability, maternal depression, substance abuse, illiteracy,
domestic violence, and lack of education/skills. In a few instances,
administrative data are available. Illinois, for example, will use
administrative data on publicly funded substance abuse services, and Washington
and Wisconsin have administrative data related to domestic violence.
As Figure A-7 shows, most include specific questions about barriers related
to lack of transportation and lack of child care.
Figure A-8 shows that, for the most part, grantees are planning to use survey
questions to query whether former recipients encounter severe problems in
meeting basic needs. These include issues associated with hunger, access
to health care, use of emergency services, and housing problems (homelessness,
lack of money to pay rent or living with relatives). Two grantees,
Georgia and Missouri, have access to administrative data on food pantries/food
banks to examine deprivation issues.
Half of the grantees are surveying former recipients about their attitudes
toward TANF or welfare reform. Additionally, Figure A-9 shows that
five are surveying former recipients about their attitudes toward work.
Many grantees also are including survey questions to probe for recipients’
awareness of transitional child care (TCC) benefits and transitional Medicaid
benefits. Finally, 12 of the 14 grantees are surveying individuals
that leave TANF about changes in household composition after program exit,
such as birth of a new child. Over half explicitly ask about changes
in marital status, and about half use administrative and/or survey data to
examine changes in household residence.
A description of each state/county project follows.
This project, the TANF Cash Exit Study, involves cohorts
from the last calendar quarter of 1996 and the first calendar quarter of
1998, with administrative data from multiple systems for both cohorts and
a sample survey of 800 cases from cohort two. Arizona has waivers which
allow for 24 months of transitional child care and Medicaid. The project
is looking at whether people are taking advantage of these benefits and the
reasons why or why not. Arizona has also recently implemented progressive
sanctioning, ending in fullfamily sanctions. They are interested
in learning whether these sanctions are succeeding in motivating participation
or employment. The project is being run by the Research and Reporting
Office of Maricopa County.
Estimated Completion Date: The draft interim report has been
submitted and is expected to be available very soon. It will include
preliminary results from their administrative data analysis covering 12 months
of data for October - December 1996 and six months of follow-up data for
the January - March 1998 cohort. Their survey instrument has been finalized
and piloted; administration of the final instrument was scheduled to begin
in March. The end date for the project is August 1999.
This project, conducted by the Manpower Demonstration Research Corporation
(MDRC), will provide an addition to MDRC’s Urban Change project with
a focus on leavers. Cohorts will be drawn from the third calendar quarters
of 1996 and 1998. The project will include analysis of up to ten years
of full population administrative data developed for the Urban Change project,
as well as a small mixedmode sample survey drawn from cohort two and
surveyed between July and December 1999. This project and the Los Angeles
project, also by MDRC, will effectively be a twosite study that will
allow comparisons while controlling for study design.
Estimated Completion Date: A draft interim report based on
the administrative data for the first cohort was submitted in late April
1999, and should be available soon. The final report is expected in
June 2000.
This project is being managed by the Urban Institute, with subcontracts for
a survey and focus groups. They will use administrative data on TANF,
Food Stamps, Medicaid and foster care from their current integrated system,
eliminating the need for difficult linking. This administrative data
is available over time going back to 1992. In addition, there will
be a series of focus groups with members of the first cohort of welfare leavers
(those who left in the first quarter of 1997) and a mixed modesurvey
of the second cohort (cases that exited in the fourth quarter of 1998).
Approximately 675 families will be surveyed nine months after exit, between
July and September 1999.
Estimated Completion Date: Uncertain. The project was
delayed in start-up by about four months because the District’s contract
with the Urban Institute was not signed until February 1999.
Florida is looking at all three populations listed in the program
announcement: those who leave the WAGES (Work and Gain Economic
SelfSufficiency) system, those who are diverted (informally, by withdrawing
voluntarily or failing to complete the application process), and those who
appear to be eligible but do not enroll. The project involves both
administrative data linking and a very large telephone survey effort (15,000
households both from statewide and across four geographic regions).
ASPE funds are supporting 10,000 of the 15,000 interviews, including all
of the interviews for cases of diversion and the eligible but not
enrolled. Findings from the statefunded surveys were released
by Florida’s subcontractor, Florida State University, in March 1999.
Estimated Completion Date: February 2000.
This project assessing the impact of welfare reform on women leaving TANF
in Georgia will add a new component to a larger ongoing study. The
current effort involves two cohorts: a sample of 2,000 leavers from
January to October 1997 tracked in administrative data; and a telephone survey
of 200 leavers per month from July 1998 to June 2001, for a total of
7,800. ASPE funding will allow the state to add a twowave survey
of sanctioned cases, before exit and again six months later, which will provide
for nonexperimental analyses of the impact of leaving TANF. The
State also received separate funding for in-person interviews of 500 persons
whom they are unable to contact by telephone, addressing the very real problem
of nonresponse bias in telephone surveys of lowincome populations.
Estimated Completion Date: The contractor, Georgia State
University, experienced initial delays in receiving necessary data from the
state, creating difficulties in finding the “leavers” and causing
implementation of their survey instrument to be delayed about three
months. Their final survey instrument has been tested and
implemented. An interim report of administrative data analysis was
mailed in late April. The end data for their project is January 2000.
Through a contract with the University of Illinois, the Illinois Department
of Human Services is sponsoring a study of how Illinois families fare after
they leave cash assistance, with an emphasis on understanding the factors
associated with success in maintaining selfsufficiency. Through
a combination of state funds ($350,000) and federal funds granted under this
Appropriation ($250,000), the project is studying cases that closed between
July 1997 and December 1998. The Illinois project integrates administrative
data from multiple systems with interview data from selected families.
The administrative data, tracked over a 6- to 23-month follow-up period,
includes data on earnings, TANF, food stamps, Medicaid, Project Chance, child
care, foster care, child protective services, and other programs. The
survey data are being collected from samples of families that exited in December
1997, June 1998, and December 1998. The federal grant has been used
to extend the study cohort by six months (through December 1998), extend
the period of administrative data tracking by six months (through June 1999),
and fund a survey of 800 cases closing in December 1998. Those survey
participants who cannot be reached by traditional telephone procedures will
be targeted by more intensive field work, in order to raise survey response
rates beyond those of the earlier statefunded cohorts.
Estimated Completion Date: A final report is expected in the
fall of 1999.
This project by the Manpower Demonstration Research Corporation (MDRC) provides
an addition to their Urban Change project with a focus on leavers.
Cohorts are being drawn from the third calendar quarters of 1996 and 1998.
The project will include analysis of up to ten years of full population
administrative data developed for the Urban Change project, as well as a
small mixedmode sample survey drawn from cohort two and surveyed between
July and December 1999. This project and the Cuyahoga County, OH project,
also by MDRC, will effectively be a twosite study that will allow
comparisons while controlling for study design. Los Angeles also received
supplementary funding of $225,00 from the Department of Housing and Urban
Development to draw a larger sample of welfare leavers who receive housing
assistance, both through public housing and Section 8.
Estimated Completion Date: A draft interim report based on
the administrative data for the first cohort is expected to be completed
in May or June 1999. The final report is expected in June 2000.
This project looks at two cohorts of families who left welfare. The
first cohort is a full population sample (20,000) of families who left welfare
between January and June, 1997. This cohort is currently being tracked
using the Massachusetts Department of Revenue’s existing Longitudinal
Database as well as a completed sample survey (350 cases). The results
of the survey are scheduled to be released shortly. For the second
cohort, they plan to draw a full population sample (15,000) of administrative
data for families that exited welfare between December 1998 and February
1999. This will be supplemented with a detailed, mixedmode survey
of 600 cases conducted by the Center for Survey Research at the University
of MassachusettsBoston. This second cohort is particularly
interesting because it will include the first group of TANF recipients to
“hit” Massachusetts’ time limits.
Estimated Completion Date: February 2000.
This project is based on two valuable databases: a comprehensive
statewide collection of administrative data from health and human services
programs, employment and training programs and wage records maintained by
the University of Missouri, and a unique database of nonprofit emergency
assistance in metropolitan Kansas City created by the Mid America Assistance
Council, a consortium of nonprofit agencies using shared forms and
intake procedures. These existing databases will be linked to each
other and to a twocohort survey of TANF leavers by the Midwest Research
Institute.
Estimated Completion Date: The survey instrument has been
finalized and piloted; interviews of fourth quarter 1996 TANF leavers began
in March 1999. They expect to complete their preliminary administrative
data analysis by the end of April and to begin linking survey data with
administrative data by July. The projected end date of the project
is October 2000.
This is a twoyear project for which the state requested less than
onethird of total funding in 1998. The study includes both TANF
exiters and individuals under sanction, whether or not they leave
assistance. The goals of the project include: determining the
frequency of outcomes such as employment, job retention, use of transitional
assistance and returns to assistance; identifying barriers to
selfsufficiency; examining the effectiveness of sanction policies in
changing behavior; and developing a longitudinal tracking capacity for welfare
outcomes in New York City. The State plans to use full population
administrative data from TANF, SSI, Food Stamps, Medicaid, foster care, child
support, and wage records to analyze cohorts from the first quarter of 1997
and the first or second quarter of 1999. Assuming second year funding
is awarded, the 1999 cohort will also involve a 900 case survey in the first
quarter of 2000.
Estimated Completion Date: The final report in this project
is expected in October 2000.
This project involves a consortium of three contiguous counties: San
Mateo, Santa Cruz, and Santa Clara, together with the SPHERE Institute.
They are using full population administrative data for the fourth quarter
1996 cohort, and a combination of full population administrative data and
two 30-minute mixed mode surveys for the fourth quarter 1998 cohort.
The project includes linking of several administrative databases and an elaborate
sampling plan in an attempt to look at a large variety of policy relevant
subgroups across the three counties, including “diverted” families
who initiate TANF applications but do not enroll in the program. The
survey instrument draws questions from a variety of wellknown national
surveys, and researchers are seeking a response rate of 85-90% of those sampled.
Estimated Completion Date: A draft interim report was submitted
in April 1999. The proposal was for a twoyear project, although
secondyear funding is contingent on performance and availability of
funds.
This project will involve cohorts leaving the Family Independence Program
(FIP) program from the first half of 1996 and the first half of 1997 using
linked administrative data and 1,000 in-person interviews. They will
examine the outcomes of four groups: leavers due to earnings, sanctions
or timelimits and eligibles who never enroll. Each of the initial
interviews will have a follow-up one year and two years later. The
study focuses particular attention on issues of child wellbeing and
use of the child welfare system. The project will also include 40 in-depth
family stories about 10 families from each group who will participate in
a more detailed and extensive survey.
Estimated Completion Date: Uncertain. USDA funds were
recently added to the project to study individuals and families leaving the
Food Stamp Program. An evaluation contractor for both populations is
about to be selected. The survey questionnaire will be modified to
also include questions related to food stamps. The sample size will
be increased deal include the food stamp leavers.
Washington is studying all three populations listed in the announcement:
those who leave welfare, those who are diverted (in this case through
lumpsum payments), and those who appear to be eligible but do not
enroll. Three cohorts are being analyzed by linked administrative
data: a preTANF cohort from the fourth quarter of 1996; an
earlyimplementation cohort from the fourth quarter of 1997; and a
fullimplementation cohort from the fourth quarter of 1998. TANF
data will be linked to data from Medicaid, food stamps, child support, child
welfare, unemployment insurance, and the State Basic Health plan. Each
cohort will include a sample of continuing cases for comparison. A
mixed mode survey of 1,000 cases is planned for the third and final
cohort. The survey sample is split evenly between individuals leaving
TANF and TANF entrants, with an emphasis being placed on leavers in rural
areas of the state.
Estimated Completion Date: A draft interim report was submitted
in late April 1999. The final report, containing both the survey and
administrative data analyses, is scheduled to be submitted by November 30,
1999.
Wisconsin is using its “leavers” funding to build on and expand
two projects. First, Wisconsin has used state funds to build a longitudinal
database to study families who either left AFDC prior to implementation of
Wisconsin Works (W-2) or who did not convert during the transition.
This data is being used in conjunction with a series of surveys of families
who left W-2 in 1998. The state funded a survey of clients who exited
in the first quarter of 1998, the results of which were released in January
1999. Nearly half of their funding under this announcement will be
used to extend this survey to clients who exited during the last three quarters
of 1998.
Estimated Completion Date: The final reports on each leavers
cohort will be released approximately one year from the end of the quarter
being studied, making the final report due on or around December 31, 1999.
Wisconsin is using more than half of its 1998 HHS funding (and plans to use
additional funding from the Administration for Children and Families in FY
1999) to expand an applicant diversion study undertaken by the Institute
for Research on Poverty (IRP) at the University of
WisconsinMadison. This portion of the study focuses on three
subgroups of applicants: (1) those who request assistance and subsequently
participate in the W-2 program; (2) those who request assistance but are
determined to be ineligible for program participation; and, (3) those who
request assistance, appear to be eligible, but do not participate in W-2.
A sixmonth 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 two waves of surveys.
Where to?
[ Report’s Main Page |
Introduction & Overview |
1999 Projects |
1998 Projects |
This Appendix |
Appendix B ]
[ HSP Home Page |
ASPE Home Page |
HHS Home Page ]